Ecologia Mediterranea - Université d`Avignon
Transcription
Ecologia Mediterranea - Université d`Avignon
6 mm (96 p) ecologia mediterranea Vol. 41 (2) – 2015 00-ecol-med-vol41(2)-couv-corTC_Mise en page 1 27/01/16 08:33 Page1 Vol. 41 (2) – 2015 Sommaire – Contents Éditorial – Editorial .......................................................................................... 3 MEDECOS special issue Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain V. THOMAS PARKER ............................................................................................... Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; a Key Factor for Successful Restoration of Mediterranean Ecosystems T. TSITSONI, N. GOUNARIS, A. B. KONTOGIANNI, V. XANTHOPOULOU-TSITSONI 33 ....................................................... Creation of an Integrated System Model for Governance in Urban MTEs (Mediterranean-Type Ecosystems) and for Adapting Cities to Climate Change – Preliminary Results T. TSITSONI, M. TSAKALDIMI, M. GOUSIOPOULOU ......................................................... ....................... Vol. 41 (2) – 2015 Revue internationale d’écologie méditerranéenne International Journal of Mediterranean Ecology 5 15 B. D. PEDRA, J. GODOY PUERTAS, L. FUENTES LOPEZ ecologia mediterranea 25 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Caractérisation du fonctionnement des steppes d’Alfa marocaines par la méthode de l’analyse fonctionnelle du paysage ............ 45 M. DERAK, F. T. MAESTRE, J. L. QUERO, V. OCHOA, C. ESCOLAR, S. SOLIVERES, P. GARCÍA-PALACIOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) 61 ...................................... Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) 73 ............................................................................ Résumés de thèses – Ph. D summaries 85 .......................................................................................... Hommage à Jacques Gamisans 92 .......................................................................... 95 P.DELIPETROU, D. GHOSN, G. KAZAKIS, P. NYKTAS, E. REMOUNDOU, I.N. VOGIATZAKIS A. ABABOU, M. CHOUIEB, A. BOUTHIBA, D. SAIDI, K. MEDERBAL R. LIBOIS, R. GHALMI, A. BRAHIMI C. CALVET, T. FRÉJAVILLE Revue indexée dans Pascal-CNRS et Biosis ISSN 0153-8756 ecologia mediterranea RESEARCH PAPERS Editors-in-Chief: Dr Élise Buisson & Dr Brigitte Talon Institut méditerranéen de biodiversité et écologie (IMBE) Mediterranean Institute of Biodiversity and Ecology Naturalia Publications 00-ecol-med-vol41(2)-couv-corTC_Mise en page 1 27/01/16 08:33 Page2 ecologia mediterranea Éditrices en chef : Dr Élise Buisson et Dr Brigitte Talon UMR CNRS IRD IMBE Université d’Avignon, IUT Site Agroparc, BP 1207 84911 Avignon cedex 09 France Instructions aux auteurs ecologia mediterranea publie des articles de recherche originaux sur des sujets se rapportant à l’écologie fondamentale ou appliquée des régions méditerranéennes. La revue exclut les articles purement descriptifs ou de systématique. ecologia mediterranea privilégie les domaines scientifiques suivants : bioclimatologie, biogéographie, biologie de la conservation, biologie marine, biologie des populations, écologie des communautés, écologie forestière, écologie génétique, écologie marine, écologie microbienne, écologie du paysage, écologie de la restauration, écologie végétale et animale, écophysiologie, paléoclimatologie, paléoécologie. La revue accepte également des articles de synthèse, des notes/communications courtes, des comptes rendus d’ouvrages, des résumés de thèses, ainsi que des commentaires sur les articles récemment parus dans ecologia mediterranea. La revue publie aussi des actes de colloques faisant l’objet d’un numéro spécial. Dans ce cas, prendre contact avec les éditrices. Comité éditorial Dr Pierre CHEVALDONNÉ, CNRS, Université Aix-Marseille, Marseille, France Dr Marc CHEYLAN, EPHESS, Montpellier, France Dr Cécile CLARET, Université AixMarseille, Marseille, France Dr Bruno FADY, INRA, Avignon, France Pr Thierry GAUQUELIN, Université AixMarseille, Marseille, France Dr Grant WARDELL-JOHNSON, Université Western, Australie Dr Raphaël GROS, Université AixMarseille, Marseille, France Dr Frédéric GUITER, Université AixMarseille, Marseille, France Pr Serge KREITER, SupAgro, Montpellier, France Dr Audrey MARCO, École nationale supérieure du paysage, Marseille, France Pr Frédéric MÉDAIL, Université AixMarseille, Marseille, France Pr François MESLÉARD, Université Avignon-Tour du Valat, France Dr Tom PARKER, San Francisco State University, États-Unis Dr Philippe PONEL, CNRS, Université AixMarseille, Marseille, France Dr Roger PRODON, EPHE, Montpellier, France Dr Sandra SAURA-MAS, Autonomous University of Barcelona, Espagne Dr Isabelle SCHWOB, Université AixMarseille, Marseille, France Dr Thekla K. TSITSONI, Aristotle University of Thessaloniki, Grèce Dr Éric VIDAL, IRD, France Dr Mercedes VIVAS, Universidad of Concepción, Chili Dr Ioannis VOGIATZAKIS, Open University of Cyprus, Chypre Les manuscrits sont soumis à des lecteurs spécialistes du sujet. La décision finale d’accepter ou de refuser un article relève des éditrices. L’article proposé doit être envoyé en version électronique à [email protected] (version doc(x) ou rtf). Pour la mise en forme du document, voir les instructions qui suivent. Une fois leur article accepté, les auteurs devront tenir compte des remarques des lecteurs, puis ils renverront leur texte corrigé sous deux mois toujours sous format électronique (doc(x) ou rtf). Passé ce délai, la seconde version sera considérée comme une nouvelle proposition. TYPES DE MANUSCRIT À préciser sur la première page lors de la soumission d’un manuscrit. Article de recherche : contribution inédite découlant d’une étude complète. Ce type d’article fait typiquement une vingtaine de pages et environ 6 000 à 8 000 mots. Note/communication courte : observation nouvelle ou rapport d’expérience dans un contexte pertinent avec les sujets visés par la revue. Ce type d’article fait typiquement une dizaine de pages et environ 3 000 à 4 000 mots. Article de synthèse : revue critique et originale de sujets spécifiques d’actualité ou d’un champ de recherche de pointe dans le domaine de l’écologie méditerranéenne. Ce type d’article fait typiquement une vingtaine de pages et environ 6 000 à 8 000 mots. Commentaire : avis sur des sujets déjà publiés dans ecologia mediterranea ou réflexion critique sur des problèmes d’intérêt général en écologie méditerranéenne. Ce type d’article fait typiquement une à cinq pages et environ 1 000 à 3 000 mots. Compte rendu d’ouvrage : revue critique d’ouvrages (livres, monographies, manuels, etc.) dans le domaine de l’écologie méditerranéenne. Les auteurs d’ouvrages souhaitant voir un compte rendu publié dans ecologia mediterranea doivent envoyer un exemplaire de l’ouvrage en question aux éditrices en chef. Résumé de thèse : résumé d’une thèse soutenue récemment dans le domaine de l’écologie méditerranéenne. Auteur, année, titre, spécialité et université (e.g. Thèse de doctorat en écologie soutenue le 3 avril 2012 à l’université de Padoue, Italie, laboratoire xxx), composition du jury, mots clés, résumé de 1 000 mots maximum. TEXTE Les articles (dactylographiés en Times 12, double interligne, format A4) doivent être rédigés en anglais ou en français. Si l’article soumis n’est pas rédigé en anglais, il est demandé (en plus des résumés) une version anglaise abrégée ainsi qu’une traduction en anglais des titres des figures et tableaux. L’article doit être complet : type de manuscrit, titres anglais et français, auteur(s) et adresse(s), résumés en anglais et en français (au minimum), version anglaise abrégée (si le texte n’est pas en anglais), mots clés anglais et français, texte, puis remerciements, bibliographie, liste des titres des figures et tableaux puis les figures et tableaux (un(e)/page). Ainsi, pour la soumission du manuscrit, les illustrations seront intégrées au document et non envoyées ISSN 0153-8756 http://ecologia-mediterranea. univ-avignon.fr ABONNEMENT (contact : [email protected]) France Europe Monde Abonnement à adresser à : SARL Transfaire Immeuble Wanad F-04250 TURRIERS Frais de port Total 60 € 60 € 60 € 6€ 12 € 16 € 66 € 72 € 76 € Code banque 19106 Code guichet 00839 numéro de compte 13995626000 Domiciliation : CA SISTERON IBAN : FR76 1910 6008 3913 9956 2600 062 BIC : AGRIFRPP891 PREMIÈRE PAGE La première page contient : 1) le type de manuscrit visé (article de recherche, communication courte, etc.) ; 2) le titre de l’article ; 3) le nom et prénom des auteurs ; 4) l’adresse de chaque auteur sera indiquée avec le courriel de l’auteur pour la correspondance. Dans le cas où la publication est le fait de plusieurs auteurs, il doit être précisé lors du premier envoi la personne à qui doit être retourné l’article après lecture ; 5) le nombre total de mots de l’introduction à la bibliographie. RÉSUMÉS, MOTS CLÉS ET VERSION ABRÉGÉE Les résumés doivent comporter 300 mots au maximum et la version abrégée (français si le manuscrit est en anglais et anglais si le manuscrit est en français) 1 000 mots (environ une page). Le nombre de mots clés est limité à six, dans la langue des résumés ; ils ne doivent généralement pas figurer dans le titre. BIBLIOGRAPHIE La bibliographie regroupera toutes les références citées et elles seules. Les références seront rangées dans l’ordre alphabétique des auteurs et de façon chronologique. Les abréviations internationales des titres des revues doivent être utilisées (ISI Journal Abbreviations Index). Vérifier attentivement le manuscrit pour s’assurer que toutes les références citées dans le texte apparaissent bien en bibliographie et inversement. Article Andow D.A., Karieva P., Levin S.A. & Okubo A., 1990. Spread of invading organisms. J. Ecol. 4: 177-188. Ouvrage Harper J.L., 1977. Population biology of plants. Academic Press, London, 300 p. Chapitre d’ouvrage May R.M., 1989. Levels of organisation in ecology. In: Cherret J.M. (ed.), Ecological concepts. Blackwell Scientific Public, Oxford: 339-363. Acte de conférence Grootaert P., 1984. Biodiversity in insects, speciation and behaviour in Diptera. In: Hoffmann M. & Van der Veken P. (eds.), Proceedings of the symposium on “Biodiversity: study, exploration, conservation”. Ghent, 18 November 1992: 121-141. Rapport et thèse Jaouadi W., 2011. Écologie et dynamique de régénération de l’Acacia tortilis (Forsk.) Hayne subsp. raddiana (Savi) Brenan var. raddiana dans le parc national de Bouhedma (Tunisie). Thèse de doctorat de l’Institut national agronomique de Tunisie, 180 p. Editors-in-Chief: Dr Élise Buisson & Dr Brigitte Talon UMR CNRS IRD IMBE Université d’Avignon, IUT Site Agroparc, BP 1207 84911 Avignon cedex 09 France ecologia mediterranea publishes original research reports and syntheses in the fields of fundamental and applied ecology of Mediterranean areas, except for descriptive articles or articles about systematic. The editors of ecologia mediterranea invite original contributions in the fields of: bioclimatology, biogeography, conservation biology, marine biology, population biology, community ecology, forest ecology, marine ecology, genetic ecology, landscape ecology, microbial ecology, restoration ecology, plant and animal ecology, ecophysiology, palaeoecology, palaeoclimatology. The journal also publishes reviews, short communications, book reviews, Ph. D. thesis abstracts and comments on papers recently published in the journal. ecologia mediterranea invite conference organizers to get in touch with the editors for special issues as part of conference/symposium proceedings. Manuscripts are peer-reviewed by appropriate referees. The final decision to accept or reject the manuscript is made by the editors. To submit a paper, please send an electronic version of your paper to [email protected] (doc(x) or rtf). Please read the following guidelines to prepare your manuscript. When the article is accepted, the authors should take reviewer’s comments into consideration. They will send back to the journal Editorial Office, within 2 months. After this deadline, the manuscript will be considered as a new submission. Editorial Board Dr Pierre CHEVALDONNÉ, CNRS, Université Aix-Marseille, Marseille, France Dr Marc CHEYLAN, EPHESS, Montpellier, France Dr Cécile CLARET, Université AixMarseille, Marseille, France Dr Bruno FADY, INRA, Avignon, France Pr Thierry GAUQUELIN, Université AixMarseille, Marseille, France Dr Grant WARDELL-JOHNSON, Université Western, Australie Dr Raphaël GROS, Université AixMarseille, Marseille, France Dr Frédéric GUITER, Université AixMarseille, Marseille, France Pr Serge KREITER, SupAgro, Montpellier, France Dr Audrey MARCO, École nationale supérieure du paysage, Marseille, France Pr Frédéric MÉDAIL, Université AixMarseille, Marseille, France Pr François MESLÉARD, Université Avignon-Tour du Valat, France Dr Tom PARKER, San Francisco State University, États-Unis Dr Philippe PONEL, CNRS, Université AixMarseille, Marseille, France Dr Roger PRODON, EPHE, Montpellier, France Dr Sandra SAURA-MAS, Autonomous University of Barcelona, Espagne Dr Isabelle SCHWOB, Université AixMarseille, Marseille, France Dr Thekla K. TSITSONI, Aristotle University of Thessaloniki, Grèce Dr Éric VIDAL, IRD, France Dr Mercedes VIVAS, Universidad of Concepción, Chili Dr Ioannis VOGIATZAKIS, Open University of Cyprus, Chypre ISSN 0153-8756 Les mots « figures » et « tableaux » annoncés dans le texte sont écrits en toutes lettres et en minuscules. Indiquer le nom d’auteur et l’année de publication (mais indiquer tous les auteurs dans la bibliographie). Exemples : “Since Dupont (1962) has shown that...”, or “This is in agreement with previous results (Durand et al. 1990; Dupond & Dupont 1997) ...”. Le numéro de page de la citation n’est mentionné que dans le cas où elle est entre guillemets. Si la publication est écrite par plus de deux auteurs, le nom du premier doit être suivi par et al. http://ecologia-mediterranea. univ-avignon.fr L’usage d’un acronyme ou d’une abréviation technique doit être précédé de sa signification lors de sa première apparition. Les codes de nomenclature doivent être respectés selon les conventions internationales. Les mots latins doivent être mis en italiques (et al., a priori, etc.), et en particulier les noms de plantes ou d’animaux. Lors de la première apparition du nom d’une espèce, il est demandé d’y faire figurer le nom d’auteur (exemple : Olea europaea L.). FIGURES ET TABLEAUX clé RIB 62 Guidelines for authors CITATIONS ET RENVOIS APPELÉS DANS LE TEXTE ABRÉVIATIONS, NOMENCLATURE ET MOTS LATINS 1 an = 2 numéros Abonnement séparément. Le texte des articles de recherche doit comporter quatre parties non numérotées : introduction, méthodes, résultats, discussion. Les sous-titres ne sont pas numérotés. Par exemple : Introduction Méthodes Site d’étude Plan expérimental Analyses des données Résultats Discussion L’emploi de mots soulignés est à proscrire. Les noms d’auteurs cités figureront en minuscules dans le texte comme dans la bibliographie. En français, n’utilisez les majuscules que pour les noms propres, sauf exception justifiée. Les ponctuations doubles ( : ; ? ! ) sont précédées d’un espace, contrairement aux ponctuations simples ( , . ). En revanche, toutes les ponctuations sont suivies d’un espace. La mise en forme définitive du texte sera assurée par la revue. ecologia mediterranea Pour la soumission du manuscrit, les illustrations seront intégrées au document, à la suite des références bibliographiques (voir la partie « Texte »). Une fois l’article accepté, les illustrations doivent être envoyées séparément du texte, prêtes à l’impression (résolution 300 dpi). Tous les documents devant être insérés doivent être annoncés dans le texte, numérotés dans l’ordre d’apparition et légendés. TIRÉS À PART Il n’est plus fourni de tirés à part mais un pdf par article. TYPES OF MANUSCRIPTS Please, mention the type of manuscript you are submitting on the first page of your submission. Research article: research report of broad scope that is an original contribution to science. The typical length of research papers is about 6,000 to 8,000 words. Short communication: brief report of new observations or of important findings that deserve publication before broader studies are completed. The typical length of short communications is about 3,000 to 4,000 words. Reviews: critical appraisal of broad areas of investigation or research in Mediterranean ecology. The typical length of reviews is about 6,000 to 8,000 words. Commentaries: opinion on topics recently published in ecologia mediterranea or essays on topics of general interest in Mediterranean ecology. The typical length of short communications is about 1,000 to 3,000 words. Book review: critical appraisal of a Book interesting to the readers of ecologia mediterranea. Editors who wish to see their book reviewed in ecologia mediterranea should get in touch with the editors of the journal. Ph. D. thesis abstracts: abstract of the Ph. D. thesis recently defended in the fields covered by ecologia mediterranea. Author, Year, Title, University (e.g. Ph. D. thesis defended on April, 4th 2014 at University of California, Irwin, USA, Lab xxx), Composition of board, Keywords, Abstract (1,000 words max.). MANUSCRIPT PREPARATION Manuscripts (typewritten Times 12, with double line spacing) must be written in English or in French. If the language is not English, you should join an English short version and English titles of figures and tables. The manuscript must be complete: e.g. title in English and French, author(s) and address(es), abstract in English and French, an English short version (if English is not the language used in the article), keywords in English and French, text, acknowledgements, references, figures and tables (one / page). For the submission, figures and tables are integrated in the SUBSCRIPTION (contact: [email protected]) 1 year = 2 issues word document and not sent separately. For research papers, the text should normally consist of 4 sections: introduction, methods, results, discussion which are not numbered. Subtitles are not be numbered either: Introduction Methods Study site Experimental design Statistical analyses Results Discussion Use lower-case letter type for names in the text and in the reference section. Do not underline any word. In English, there is one blank after any punctuation, never before. Copy editing of manuscripts is performed by the journal. FIRST PAGE The first page contains: 1) the type of manuscript (research paper, short communication, etc.); 2) the title of the article; 3) the names of the authors; 4) the address of each author + an email at least of the corresponding author; 5) the total number of words from the introduction to the references included. ABSTRACTS, KEYWORDS AND SHORT VERSION Abstracts should be no longer than 300 words. The English short version should not exceed one page long (1,000 words). Do not use more than six keywords. Keywords should not be present in the title. REFERENCES All publications cited in the text should be presented in a list of references following the text of the manuscript. The list of references should be arranged alphabetically on author’s names, and chronologically for each author. You should abbreviate the titles of periodicals mentioned in the list of references (except if you are not sure of it; ISI Journal Abbreviations Index). Check the manuscript to make sure that all references are cited and that all citations in the text are included in the references. Use following system to write the references: Journal article Andow D.A., Karieva P., Levin S.A. & Okubo A., 1990. Spread of invading organisms. J. Ecol. 4: 177-188. Book Harper J.L., 1977. Population biology of plants. Academic Press, London, 300 p. Book chapters May R.M., 1989. Levels of organisation in ecology. In: Cherret J.M. (ed.), Ecological concepts. Blackwell Scientific Public., Oxford: 339-363. Conference proceedings Grootaert P., 1984. Biodiversity in insects, speciation and behaviour in Diptera. In: Hoffmann M. & Van der Veken P. (eds.), Proceedings of the symposium on “Biodiversity: study, exploration, conservation”. Ghent, 18 November 1992: 121-141. IN-TEXT CITATIONS The words “figure” and “table” announced in-text should be written in extenso and with lower-case letter type. In the text refer to the author’s name and year of publication (followed by pages only if it is a quotation). If a publication is written by more than two authors, the name of the first author should be used followed by “et al.” (this indication, however, should never be used in the list of references: first author and co-authors should be mentioned in it). Examples: “Since Dupont (1962) has shown that…”, or “This is in agreement with previous results (Durand et al. 1990; Dupond & Dupont 1997)…”. ABBREVIATES, NOMENCLATURE AND LATIN WORDS Subscription Postage Total 60 € 60 € 60 € 6€ 12 € 16 € 66 € 72 € 76 € France Europe World Explanation of a technical abbreviate is required when the first use. International convention codes for nomenclature should be used. Latin words should be in italic (et al., a priori, etc.), particularly for plants or animals’ denomination (the first time, please precise author’s name: for example, Olea europaea L.). FIGURES AND TABLES Mail this subscription to: SARL Transfaire Immeuble Wanad F-04250 TURRIERS Bank: CA SISTERON IBAN: FR76 1910 6008 3913 9956 2600 062 BIC: AGRIFRPP891 All illustrations should be included in the word document after the section “References”. Once the paper is accepted for publication, illustrations should be sent separately (3,000 dpi). All the illustrations should be cited and should have a legend. REPRINTS A pdf version will be supplied free of charge for each paper. 00-ecol-med-vol41(2)-debut-3-corTC_Mise en page 1 26/01/16 11:03 Page1 ecologia mediterranea Revue internationale d’écologie méditerranéenne International Journal of Mediterranean Ecology Sommaire/Contents Vol. 41 (2) – 2015 Editorial – Éditorial ............................................................................. 3 MEDECOS special issue Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions V. THOMAS PARKER ..................................................................................... 5 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain B.DUGUY PEDRA, J.GODOY PUERTAS, L. FUENTES LOPEZ ............................................ 15 Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; a Key Factor for Successful Restoration of Mediterranean Ecosystems T. TSITSONI, M. TSAKALDIMI, M. GOUSIOPOULOU ................................................... 25 Creation of an Integrated System Model for Governance in Urban MTEs (Mediterranean-Type Ecosystems) and for Adapting Cities to Climate Change – Preliminary Results T. TSITSONI, N. GOUNARIS, A. B. KONTOGIANNI, V. XANTHOPOULOU-TSITSONI .................... 33 RESEARCH PAPERS Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) ........... 45 M. DERAK, F. T. MAESTRE, J. L. QUERO, V. OCHOA, C. ESCOLAR, S. SOLIVERES, P. GARCÍA-PALACIOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 P. DELIPETROU, D. GHOSN, G.KAZAKIS, P. NYKTAS, E.REMOUNDOU, I.N. VOGIATZAKIS Caractérisation du fonctionnement des steppes d’Alfa marocaines par la méthode de l’analyse fonctionnelle du paysage ecologia mediterranea – Vol. 41 (2) – 2015 1 00-ecol-med-vol41(2)-debut-3-corTC_Mise en page 1 28/01/16 09:38 Page2 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) A. ABABOU, M. CHOUIEB, A. BOUTHIBA, D. SAIDI, K. MEDERBAL ................................. 73 Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) R. LIBOIS, R. GHALMI, A. BRAHIMI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Résumés de thèses – Ph. D summaries C. CALVET, T. FRÉJAVILLE ............................................................................... Hommage à Jacques GAMISANS ............................................................ Revue indexée dans Pascal-CNRS et Biosis Journal indexed in PASCAL-CBRS and Biosis http://ecologia-mediterranea.univ-avignon.fr/ Acknowledgments – Remerciements The editorial committee thanks the associate editors and reviewers who have participated in this volume for their advices, corrections and opinions. Le comité éditorial de la revue remercie les éditeurs associés et les relecteurs qui ont participé à ce numéro pour leurs conseils, corrections et avis. Editorial Board – Comité éditorial © ecologia mediterranea Fabrication : Transfaire, 04250 Turriers Imprimé en Europe Dr Pierre CHEVALDONNÉ, CNRS, Université Aix-Marseille, Marseille, France Dr Marc CHEYLAN, EPHESS, Montpellier, France Dr Cécile CLARET, Université Aix-Marseille, Marseille, France Dr Bruno FADY, INRA, Avignon, France Pr Thierry GAUQUELIN, Université Aix-Marseille, Marseille, France Dr Grant WARDELL-JOHNSON, University of Western Australia, Australie Dr Raphaël GROS, Université Aix-Marseille, Marseille, France Dr Frédéric GUITER, Université Aix-Marseille, Marseille, France Pr Serge KREITER, SupAgro, Montpellier, France Dr Audrey MARCO, École nationale supérieure du paysage, Marseille, France Pr Frédéric MÉDAIL, Université Aix-Marseille, Marseille, France Pr François MESLÉARD, Université Avignon-Tour du Valat, France Dr Tom PARKER, San Francisco State University, États-Unis Dr Philippe PONEL, CNRS, Université Aix-Marseille, Marseille, France Dr Roger PRODON, EPHE, Montpellier, France Dr Sandra SAURA-MAS, Autonomous University of Barcelona, Espagne Dr Isabelle SCHWOB, Université Aix-Marseille, Marseille, France Dr Thekla K. TSITSONI, Aristotle University of Thessaloniki, Grèce Dr Éric VIDAL, IRD, France Dr Mercedes VIVAS, Universidad of Concepción, Chili Dr Ioannis VOGIATZAKIS, Open University of Cyprus, Chypre 92 95 00-ecol-med-vol41(2)-debut-3-corTC_Mise en page 1 26/01/16 11:03 Page3 Editorial – Éditorial Élise BUISSON et Brigitte TALON Éditrices en chef Editors-in-Chief This issue, the second of the year marking the 40th anniversary of ecologia mediterranea, half consists of articles from the MEDECOS conference held in Olmué, Chile, from Oct. 6 to 9, 2014. This is an opportunity to remind readers of ecologia mediterranea that ISOMED, the International Society for Mediterranean Ecology promotes research, conservation and public awareness of the biological diversity of the world’s Mediterranean-climate regions since 1971 (http://www.incomme.org/isomed.html). The next ISOMED conference, known as MEDECOS conference, will take place in Seville (Spain) from 31 January to 4 February 2017. The rest of the issue is devoted to research articles, which from Crete to Algeria and Morocco, take us to forests and alfa steppes and tell us about the dietary habits of the otter. This 40th anniversary is also tarnished by the demise, this fall, of two great botanists: Jacques Gamisans and Pierre Quézel, the latter being the founder in 1975 of the journal ecologia mediterranea of which he was the editor-in-chief from 1975 to 1996. Both have greatly contributed to the knowledge on the flora and to the development of plant ecology in the Mediterranean Basin. Both were attached to ecologia mediterranea, in which they published numerous articles since the 1970s. Both will greatly be missed by the scientific community. We would like to conclude with a thought for the victims of Nov. 13 attacks in Paris, which included nationals of many countries that contribute to ecologia mediterranea: Algeria, Australia, Chile, Spain, USA, Italy, Morocco, Portugal, Tunisia. Ce nouveau numéro, le deuxième de l’année marquant le 40e anniversaire d’ecologia mediterranea, est pour moitié constitué d’articles issus de la conférence MEDECOS qui s’est tenue à Olmué, au Chili, du 6 au 9 octobre 2014. Ceci est l’opportunité de rappeler aux lecteurs d’ecologia mediterranea qu’ISOMED, la Société internationale d’écologie méditerranéenne promeut la recherche, la conservation et la sensibilisation du public à la diversité biologique depuis 1971 (http://www.incomme.org/isomed.html). La prochaine conférence d’ISOMED, connue sous le nom de conférence MEDECOS, portant sur des problématiques liées aux régions au climat méditerranéen, aura lieu à Séville (Espagne) du 31 janvier au 4 février 2017. Le reste du numéro est consacré à des articles de recherche, qui, de la Crète au Maroc en passant par l’Algérie, nous conduisent des forêts aux steppes à Alfa et nous renseignent sur les habitudes alimentaires de la loutre. Mais ce quarantième anniversaire de la revue s’est teinté d’une grande tristesse cet automne avec la disparition à un mois d’intervalle de deux grands botanistes, Jacques Gamisans et Pierre Quézel, ce dernier étant le fondateur en 1975 de la revue ecologia mediterranea, dont il fut l’éditeur en chef de 1975 à 1996. Tous deux ont grandement contribué à la connaissance de la flore et au développement de l’écologie végétale du bassin méditerranéen. Tous deux étaient fidèles à ecologia mediterranea, dans laquelle ils ont publié de nombreux articles dès les années 1970. Tous deux vont beaucoup manquer à la communauté scientifique. Nous finirons par une pensée pour les victimes des attentats du 13 novembre à Paris, parmi lesquelles figurent des ressortissants des pays qui contribuent à faire vivre ecologia mediterranea : Algérie, Australie, Chili, Espagne, États-Unis, Italie, Maroc, Portugal, Tunisie. ecologia mediterranea – Vol. 41 (2) – 2015 3 00-ecol-med-vol41(2)-debut-3-corTC_Mise en page 1 26/01/16 11:03 Page4 Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions La divergence de banque de semences entre Arctostaphylos Adans. (Ericaceae) et Ceanothus L. (Rhamnaceae) suggère différentes interactions de prédateur de semences V. Thomas PARKER Department of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, CA 94132, U.S.A. Corresponding author: [email protected] Received: 15 January, 2015; First decision: 11 February, 2015; Revised: 5 March, 2015; Accepted: 19 April, 2015 Abstract Two of the dominant shrub genera in California chaparral, Arctostaphylos and Ceanothus, utilize persistent soil seed banks for post-fire recovery of their populations. Seed production and seed rain is subject to considerable predation by the animal community. These genera differ significantly in their seed size, and contrary to seed bank theory, Arctostaphylos with the larger seeds are able to accumulate substantially larger seed banks. Literature and field data are used to propose a model to account for these differences in seed bank densities, specifically, that the animal community involved in seed predation differs between the two genera. Résumé Deux genres d’arbustes dominants dans le chaparral de Californie, Arctostaphylos et Ceanothus, utilisent les banques persistantes de semences dans le sol pour le rétablissement de leurs populations après un incendie. La production de semences et la pluie de semences est sujette à une prédation considérable par la communauté animale. Ces genres diffèrent de manière significative dans la taille de leur Keywords: scatter-hoarding, mutualism, bird community, wildfire, mammal community. ecologia mediterranea – Vol. 41 (2) – 2015 semence. Contrairement à la théorie de banque de semences, les Arctostaphylos, avec les semences les plus grosses, peuvent accumuler des banques de semences sensiblement plus grandes. Des données provenant de la littérature et du terrain sont utilisées afin de proposer un modèle permettant d’expliquer les différences dans les densités de banque de semences, et notamment le fait que la communauté animale impliquée dans la prédation de semences diffère entre les deux genres. Introduction Arctostaphylos and Ceanothus are the only two genera in California chaparral that have developed both fire-stimulated persistent soil seed banks and a large number of obligate seeders. Research on life history evolution and fire regimes suggests these species are quite similar in their responses (Keeley et al. 2012). However, early studies of chaparral Mots clés : banque de graines, caches de nourriture, communauté de mammifères, communauté d’oiseaux, feu, incendie, mutualisme. 5 V. Thomas Parker seed banks indicated that Arctostaphylos and Ceanothus have substantially different seed bank densities (Keeley 1977; Parker & Kelly 1989). Arctostaphylos species produce relatively large seed banks ranging from around 300 to 10,000 seeds m-2 (Parker & Kelly 1989). Ceanothus species, in contrast, exhibit much smaller seed banks, ranging from close to no seed in some years (Quinn 1994) up to only several hundred seeds m-2 (Keeley 1977; Zammit & Zedler 1988; Parker & Kelly 1989; O’Neil & Parker 2005). No hypotheses have been put forward to explain these differences nor their implications. While chaparral seed banks are a key stage for life history dynamics, they are subject to considerable seed predation (Keeley & Hays 1976; Kelly & Parker 1990; Quinn 1994; O’Neil & Parker 2005). Consequently chaparral seed banks can be used as a probe into the granivore community and to determine the post-dispersal fate of seed (Thompson 2000, Vander Wall et al. 2005). Ceanothus greggii seed banks, for example, show no consistent stand age-correlated pattern (Zammit & Zedler 1988), while no change occurred in persistent soil seed bank densities for a pair of Arctostaphylos species over a 10 years period (Keeley 1987), both studies implying considerable post-dispersal predation. Overall, seed predation studies indicate variable but significant removal of from 30% in short term studies to nearly all fruit or seed of Arctostaphylos and Ceanothus offered (Keeley & Hays 1976; Conard et al. 1985; Kelly & Parker 1990; Quinn 1994; O’Neil & Parker 2005; Deveny & Fox 2006). While studies generally lack definitive evidence, researchers have assumed that rodents are the most important seed predators (Quinn 1994; Deveny & Fox 2006; Huffman 2006). With respect to most studies in shrublands, research has tended to focus on single species or to treat the animal community as a whole, not attempting to differentiate among different types. However, studies distinguishing between invertebrates and vertebrates indicate that there can be quite different impacts (Hulme 1997) or that impacts are additive (Zammit & Westoby 1988; O’Neil & Parker 2005). Additional evidence suggests that herbivores and granivores may differentially impact plant genera, with differential herbivory on resprouting plants or post-fire germination; for example, herbivores consume seedlings and browse young plants of 6 Ceanothus more than Adenostoma Hook. & Arn. (Mills 1983; Quinn 1994). The difference in seed bank densities between Arctostaphylos and Ceanothus suggest the potential for differential animal community dynamics. Animals play a major role in the dynamics of shrub-dominated communities by impacting plant species either positively, for example by dispersal, or negatively by herbivory or seed consumption (Quinn 1994). Animal communities are a well-established component within chaparral (Quinn 1994; Schwilk & Keeley 1998; Parker et al. 2015) and other Mediterranean-shrub communities (e.g., Bond et al. 1980; Bond 1984; Zammit & Westoby 1988; Cowling et al. 1996; Hulme 1997; Forget & Vander Wall 2001; Midgley et al. 2002; Rusch et al. 2014) and can have differential impacts across landscapes (e.g., Pons & Pausas 2007a, b). Animal foraging also may be critical to the burial of seed for these plant community dominants. Seeds must be deep enough to escape the killing heat pulse of a fire in the upper portions of the soil (Odion & Davis 2000). Rodent species found in chaparral are frequently scatter-hoarders (Vander Wall 1990; 2001). If these animals represent a large proportion of the animal community, their activity potentially could provide a mechanism of burial. However, many birds frequently found in chaparral are considered direct seed consumers (e.g. California and Spotted Towhees, California Quail). Seed predators, for example, may be more effective at reducing the density of post-dispersal Ceanothus seeds compared to Arctostaphylos. In field experiments, birds appear to be the dominant seed predator of Ceanothus compared to rodents with Arctostaphylos (Warzecha & Parker 2014). This study examines the potential for differential impact on post-dispersal seed predation and seed banks of two dominant genera of California chaparral, Arctostaphylos and Ceanothus, both of which depend on seed banks for post-fire recruitment. Ironically, most Arctostaphylos seed structures (seeds within nut-like endocarps) (2-12 mm diameter) are much larger than Ceanothus seed (1.5-3 mm diameter) and Arctostaphylos having greater seed bank density contradicts general seed bank theory (Thompson et al. 1993; Bekker et al. 1998; but see Leishman & Westoby 1994; Moles et al. 2000). ecologia mediterranea – Vol. 41 (2) – 2015 Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions Figure 1 – Ceanothus and Arctostaphylos species differ in the type of fruit produced. Illustrated is A.) an example of the ballistic fruit found in Ceanothus species (in this case C. papillosus Torr. & A. Gray) with three seed chambers, and B.) an example of the dry drupes produced by Arctostaphylos species (in this case, A. gabilanensis V.T. Parker and M.C. Vasey). My principal objective is to develop a hypothesis that differences in seed banks reflect differences in the animal granivore community. Differential animal impact could be critical for long-term plant dynamics (Moreno & Oechel 1991; Tyler 1996), and reciprocally, if animals generally associate with different plant genera, then there is also the potential for a longterm impact on animal populations. Animal community structure has an influence on the assembly of plant communities through direct and indirect interactions such as differential herbivory, granivory, and different foraging behavior and preferences (Davidson 1993; Hulme 1998). How animal community composition impacts plant community composition is complex and variable yet important to predict how plant communities will respond to future environments (Götzenberger et al. 2012). Here I ask if the animal community differentiates by different types of foraging behaviors in Ceanothus versus Arctostaphylos dominated chaparral. Methods Study sites: three study sites in California were used, UC Ft. Ord Reserve near Marina in a stabilized dune area, Monterey Co.; Bolinas ridge of the Golden Gate National Recreation Area, Marin Co.; and Bonny Doon Ecological Reserve (BDER), Santa Cruz Co. Seed sizes and seed bank density: Arctostaphylos produces drupes that range in diameter from 3-15 mm. These drupes have a dry ecologia mediterranea – Vol. 41 (2) – 2015 exocarp, dry, pulpy mesocarp (or mesocarp absent), and a hardened endocarp divided into multiple small nuts surrounding individual seeds. Ceanothus produces explosive fruit; the fruit exhibit three segments that each usually contain a single seed (Figure 1). To assess seed size and seed bank sizes, data were collected from published sources. Algorithms were developed for both genera based on measurements of sample fruit and seed, and a constant was developed to multiply into all species fruit diameters to estimate seed sizes. For Arctostaphylos, constants were based on measurements of fruit and seed for five species (A. sensitiva, A. canescens, A. viscida, A. glandulosa, A. glauca) that provided a range of fruit and seed size. Different constants were developed for fruit with separate seed from those with partially or fully fused fruit. For Ceanothus, a similar procedure used measurements of seed from five species representing a range of seed sizes and from both subgenera (C. papillosus, C. thyrsiflorus, C. cuneatus, C. jepsonii, C. prostratus). Seed sizes for both genera were then estimated by multiplying the constants into published fruit diameters from the most recent treatments (Parker et al. 2012; Wilken 2012). Seed bank sizes were taken from available published literature sources (Keeley 1977; Zammit & Zedler 1988; Parker & Kelly 1989; Quinn 1994; O’Neil & Parker 2005; Warzecha & Parker 2014). Values were tested for differences in means using a t-test for unequal variances and unequal sample sizes (Welch two sample t-test) (R Development Core Team 2013). 7 V. Thomas Parker Figure 2 – S ize differences between dispersing propagules of Ceanothus (seed) versus Arctostaphylos (seed enclosed in nut-like endocarps). Differences were highly significant using a t-test accounting for differences in variance and sample size (p = 5.591E-10). Boxes are 25–75% quantile, horizontal line indicates median, open circles indicate outliers. See methods for how data were estimated. The animal community was estimated using preliminary field data and literature. Fruit from Arctostaphylos were collected 4-6 weeks before experiments. Fruit were placed out in petri dishes. During these experiments, live traps were set out on one night (Gage Dayton, University of California, Santa Cruz) to help identify species. In the Ceanothus stand, seed were collected using seed traps. In the study sites, seed was offered to animals for one day to several weeks and video surveillance used motion-sensored Bushnell 119455C Trophy Night Vision Trail Cameras (Bushnell Corporation, Overland Park, Kansas, USA). In earlier experiments, seed or fruit was placed in petri dishes in the middle of large 35 cm diameter pans covered with a layer of fluorescent paint and animal activity was analyzed by examining footprints. These methods are limited in their ability to reliably identify many of the smaller rodents, but these animals generally fall into the same functional group. The field portion was restricted to coastal sites at relatively low elevation (< 450 m) and restricted latitudinal distribution (~38°N – 36°N). Chaparral, however, is a geographically widespread community in California and it has considerable elevation range as well (Parker et al. 2015). Consequently, a literature survey was conducted of common granivores or frugivores in chaparral to generalize the results of this study, even though they could not be assigned to dominants (sources were Lawrence 1966; Soule et al. 1988; Fellers 1994; Price et al. 1995; Schwilk & Keeley 1998; Laakkonen 2003) (see Parker et al. 2015). Results Figure 3 – D ensity differences between seed banks of Ceanothus versus Arctostaphylos. Data are from published sources (Keeley 1977, Zammit & Zedler 1988, Parker and Kelly 1989, Quinn 1994, O’Neil & Parker 2005, Warzecha & Parker 2014). Differences in density were highly significant using a t-test accounting for differences in variance and sample size (p = 0.002). Boxes are 25–75% quantile, horizontal line indicates median, open circles indicate outliers. 8 Estimated seed sizes differed significantly between Ceanothus and Arctostaphylos (Figure 2) (t = -7.1689; df = 71.432, p = 5.591E10). The average diameter for Ceanothus seed was 2.46 mm (+ 0.10 S.E.) and for Arctostaphylos was 5.16 mm (+ 0.36 S.E.). From published data, thirteen seed bank density values were found for Ceanothus and eleven for Arctostaphylos, and for these data, seed bank densities were significantly different between Ceanothus and Arctostaphylos (Figure 3) (t = -4.0894, df = 10, p = 0.002). Ceanothus seed banks were smaller (226.69 + 55.20 seeds m-2) compared to Arctostaphylos seed banks (4382.83 + 1014.81 seeds m-2). ecologia mediterranea – Vol. 41 (2) – 2015 Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions Chaparral contains a diversity of animals and yet there seems to be some differentiation between Arctostaphylos-dominated compared to Ceanothus-dominated chaparral (Table 1, Table 2). Both types of chaparral contain a diverse group of seed consumers, most of which are recognized as scatter-hoarders. Even species thought to be principally herbivorous, for example Neotoma fuscipes Baird (wood rat) will take fruit of Arctostaphylos or seed of Ceanothus when available. One trail camera caught a wood rat burying a small cache, which was not expected. The table includes other species common in chaparral, but in the sites investigated, chipmunks, pocket mice, California mice, and Boyle’s mice were the most common visitors, in addition to wood rats. These species all exhibit scatter-hoarding to some extent. Other mammals were observed, but only occasionally, and these are thought to be principally herbivorous (e.g., the brush rabbit). One major difference between the stands was the common occurrence of granivorous birds in stands of Ceanothus, especially the California and Spotted Towhees. Quail were observed on multiple occasions, but towhees were the most frequent of all representing over 80% of bird samples. During daylight, towhees were present the greatest number of times, followed by quail and sparrows. These types of birds were rarely seen in stands of Arctostaphylos. Other animals were also observed such as the fox sparrow (Passerella iliaca Merrem), the hermit thrush (Catharus guttatus Pallas) and the wrentit (Chamaea fasciata Gambel). These birds eat mixed diets but only the fox sparrow eats seed to any extent. Other vertebrates included the coast garter snake (Thamnophis elegans terrestris Fox), the Pacific coast rattlesnake (Crotalus oreganus Holbrook), brush rabbits (Sylvilagus bachmanii Waterhouse), bobcats (Lynx rufus Mearns), and evidence of coyotes (Canis latrans Say). The brush rabbits are generally strictly herbivorous and so do not directly impact seed banks. The snakes and bobcats are essentially carnivores, and also would only affect seed banks indirectly. Coyotes are not seed predators, but they do consume fruit and frequently eat fruit of Arctostaphylos. The seed passes through unharmed, however, and so they act as long-distance dispersal agents, but do not otherwise directly affect seed banks. Table 1 – Common small mammals found in chaparral by species and common name organized by their general food habits. Frequency refers to how widespread and common species is reported in the literature, while the last column refers to whether evidence was found in this study for their presence in different stands. C = Ceanothus stands; A = Arctostaphylos stands; A “?” means the determination is uncertain. Species Common name Frequency Tamias merriami J.A. Allen T. sonomae Grinnell T. quadrimaculatus Gray Chipmunks Frequent, common at higher elevations Callospermophilus lateralis Say Ground squirrel Higher elevations Otospermophilus beecheyi Richardson Ground squirrel Occasional Chaetodipus californicus Merriam C. fallax Merriam Pocket mouse Frequent to common Dipodomys venustus Merriam D. agilis Gambel D. heermanni Le Conte Kangaroo rat Coastal Ranges Peromycus boylii Baird P. maniculatus Wagner P. californicus Gambel P. truei Schufeldt Boyle’s mouse Deer mouse California mouse True’s mouse Common Reithrodontomys megalotis Baird Harvest mouse Frequent Microtus sp. Shrank Vole Frequent post-fire Neotoma sp. Say & Ord Woodrats Common Sylvilagus sp. Gray Brush rabbit Common Lepus sp. L. Jackrabbit Common Ceanothus or Arctostaphylos Principally frugivores or granivores C/A C C/A C C/A C/A? Principally herbivores but also consume seed C/A Sources are from this study or from published data (see Methods). ecologia mediterranea – Vol. 41 (2) – 2015 9 V. Thomas Parker Table 2 – Birds found in chaparral by species and common name organized by their general food habits. Frequency refers to how widespread and common species is reported, the last column refers to whether evidence was found in this study for their presence in different stands. C = Ceanothus stands; A = Arctostaphylos stands. Species Common name Frequency Ceanothus or Arctostaphylos Callipepla californica Shaw California quail Common C Piplio maculates Swainson Spotted towhee Common C Melozone crissalis Vigors California towhee Common C Melozone fuscus Swainson Canyon towhee Frequent Chamaea fasciata Gambel Wrentit Common Toxostoma redivivum Gambel California thrasher Common Aphelocoma californica Vigors Western scrub jay Common Passerina amoena Say Lazuli bunting Occasional Amphispiza belli Cassin Sage sparrow Occasional Aimophila ruficeps Cassin Rufous-crowned sparrow Occasional Spizella atrogularis Cabinis Black-chinned sparrow Occasional Thryomanes bewickii Audubon Bewick’s wren Common Psaltriparus minimus Townsend Bushtit Frequent Polioptila californica Brewster California gnatcatcher Occasional Geococcyx californianus Lesson Road runner Occasional Principally granivores Principally omnivorous C/A A C Principally carnivorous C/A Sources are from this study or from published data (see Methods). Discussion Seed bank densities of Arctostaphylos species average about 20 times that of Ceanothus species (Figure 2). Because Arctostaphylos has significantly larger seeds on average (Figure 1), such differences require attention because that appears to violate seed bank theory (Thompson et al. 1993; Bekker et al. 1998). This suggests that some other process is involved that inverts the seed bank results. Differential dominance of principal seed predators may be the source of that process. The major difference between animal communities in stands of Arctostaphylos and Ceanothus dominated chaparral is the frequency and abundance of granivorous birds. Both types of chaparral contain a variety of rodents that are also seed consumers, but generally no consistent difference in rodent composition exists between the two chaparral types. The similar mammal composition may still have differential impacts because rodents might directly consume the smaller seed of Ceanothus more frequently than cache them. Caches of Ceanothus do occur, but apparent caches are quite rare in low elevation coastal sites 10 (Parker, pers. obs.), although they do occasionally occur in higher elevation sites, either as scatter-hoarded caches or larder-hoarded caches (Parker, pers. obs.; Vander Wall, pers. comm.). Scatter-hoarding is being recognized as a common process among Mediterraneanclimate shrublands (Forget & Vander Wall 2001; Midgley & Anderson 2005). Historically, most researchers have assumed rodents are the principal seed predators for both genera (e.g., Quinn 1994; Deveny & Fox 2006; Huffman 2006). Generally, studies have not attempted definitive experiments to distinguish between rodents and birds (but see Warzecha and Parker 2014), and have assumed rodents were the most important or made no distinction among animals (e.g., O’Neil and Parker 2005). This distinction is important as granivorous birds are visual seed predators and rigorously work through the litter during and after seed dispersal. Evidence from this study and another indicates that they can be more effective seed predators of Ceanothus than rodents (Warzecha and Parker 2014), and thus may be the missing dimension explaining the differences in seed bank densities between the two genera. ecologia mediterranea – Vol. 41 (2) – 2015 Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions Escaping seed predators for Ceanothus seed and becoming incorporated into soil seed banks is based potentially on the small size allowing them to percolate into the soil, while for Arctostaphylos, scatter-hoarding burial would be the critical process. Because the seeds of both genera are essentially dormant until stimulated by fire, the differential animal composition and activity in chaparral suggest a number of predictions. One is that seed banks should come into a rough equilibrium based on their detectability to the granivore community, and generally, larger seeds should be easier to detect, suggesting lower seed bank densities for larger fruit/seed for both genera. Because of the larger range in fruit/seed size, if detectability is a principal process, this further suggests that there should be a larger range in the density of Arctostaphylos seed banks relative to the density range of Ceanothus seed banks. Some evidence exists to support that contention (Parker & Kelly 1989; O’Neil & Parker 2005), but further research will be necessary to substantiate it. Seed size may be simultaneously under selection by animals to remain undetected, yet also selected to have enough reserves to establish seedlings able to survive summer drought, both within and among species. Both rodents and birds can have selective preferences on seed sizes and shapes in species, complicating the dimensions of these interactions (Pons & Pausas 2007c; Muñoz et al. 2012; Rusch et al. 2013). Finally, overall lower seed bank densities suggest Ceanothus seedlings should be under selection for faster growth rates or tolerance of water stress to assure post-fire establishment more than Arctostaphylos. Some data may support this last suggestion as Frazer & Davis (1988) found greater tolerance to summer drought for Ceanothus seedlings compared to Adenostoma seedlings. Also, in general, seedlings emerging from smaller seeds or suffering higher rates of first-year mortality generally tend to have faster relative growth rates (Gilbert et al. 2006; Kitajima & Myers 2008; Moles & Leishmann 2008) The interactions between the plant community and the resident animal community range from facilitory to inhibitory in terms of the long-term success of the dominant plants. The role of soil seed banks in these two genera is fundamental to the dynamics and persistence of their populations. Given that most species in both Arctostaphylos and Ceanothus are obligate seeders, their ability to replace themselves post-fire from seed banks is critical and ecologia mediterranea – Vol. 41 (2) – 2015 their interactions with animals subject to considerable selection. In the context of changing climates, which in California is predicted to be considerably warmer and drier (Hayhoe et al. 2004; Cayan et al. 2008), how both plant seed banks and animal community behavior and composition respond may determine the long-term dynamics of these communities (Keeley et al. 2012). Already the composition of animal communities in California is changing (Moritz et al. 2008); thus we might expect future local dynamics could include both the enhancement or the extirpation of obligate seeder populations in conjunction with wildfire or extreme weather events. Animal activity is intense within chaparral with respect to interactions with fruit and seed and the difference in seed bank densities may result from the shifts in dominance of the animal communities. Seed burial processes appear to differ between these genera due to the behavioral differences of the principal seed predators, the effective granivorous birds versus scatter-hoarding rodents. These compositional shifts suggest predictions about seed size:seed bank relationships within genera and predict differences in relative growth rates between the two genera. These results and the review of published data support the hypothesis that Arctostaphylos and Ceanothus dominated chaparral stands differ in the composition and abundance of their animal communities. However, because few stands were investigated and the number of survey days differed considerably among sites, these conclusions should be considered preliminary. Acknowledgements Critical field help was provided by E. Herbert, J. van den Berg, B. Warzecha, and B. Peterson. Support for the work in part came from the U.S. National Science Foundation and the Office of Research and Sponsored Projects at San Francisco State University. 11 V. Thomas Parker References Bekker R.M., Bakker J.P., Grandin U., Kalamees R., Milberg P., Poschlod P., Thompson K. & Willems J.H., 1998. Seed size, shape and vertical distribution in the soil: indicators of seed longevity. Funct. Ecol. 12: 834-842. Bond W.J., 1984. Fire survival of Cape Proteaceae – influence of fire season and seed predators. Vegetatio 56: 65-74. Bond W.J., Ferguson M. & Forsyth G., 1980. Small mammals and habitat structure along altitudinal gradients in the southern Cape mountains. S. Afr. J. Zool. 15: 34-43. Cayan D.R., Maurer E.P., Dettinger M.D., Tyree M., Hayhoe K., 2008. Climate change scenarios for the California region. Climatic Change 87 (Suppl 1):S21-S42. Conard S.G., Jaramillo A.E. & Cromack K., 1985. The role of the genus Ceanothus in western forest ecosystems. USDA For. Serv. gen. tech. rep. PNW-182, Oregon State University, Corvallis, 72 p. Cowling R.W., MacDonald I.A.W. & Simmons M.T., 1996. The Cape Peninsula, South Africa: physiographical, biological and historical background to an extraordinary hot-spot of biodiversity. Biodivers. Conserv. 5: 527-550. Davidson D.W., 1993. The effects of herbivory and granivory on terrestrial plant succession. Oikos 68: 23-35. Deveny A.J. & Fox L.R., 2006. Indirect interactions between browsers and seed predators affect the seed bank dynamics of a chaparral shrub. Œcologia 150: 69-77. Fellers G.M., 1994. Species diversity, selectivity, and habitat associations of small mammals from coastal California. Southwest. Nat. 39:128-136. Forget P-M. & Vander Wall S.B., 2001. Scatter-hoarding rodents and marsupials: convergent evolution on diverging continents. Trends Ecol. Evol. 16: 65-67. Frazer J., Davis S., 1988. Differential survival of chaparral seedlings during the first summer drought after wildfire. Œcologia 76: 215-221. Gilbert B., Wright S.J., Muller-Landau H.C., Kitajima K. & Hernandéz A., 2006. Life-history trade-offs in tropical trees and lianas. Ecology 87: 1281-1288. Götzenberger L., de Bello F., Bråthen K.A., Davison J., Dubuis A., Guisan A., Lepš J., Lindborg R., Moora M., Pärtel M., Pellissier L., Pottier J., Vittoz P., Kristjan Zobel K. & Zobel M., 2012. Ecological assembly rules in plant communities – approaches, patterns and prospects. Biol. Rev. 87: 111-127. Hayhoe K., Cayan D., Field C.B., Frumhoff P.C., Maurer E.P., Miller N.L., Moser S.C., Schneider S.H., Cahill K.N., Cleland E.E., Dale L., Drapek R., Hanemann R.M., Kalkstein L.S., Lenihan J., Lunch C.K., Neilson R.P., Sheridan S.C. & Verville J.H., 2004. Emissions pathways, climate change, and impacts on California. Proc. Natl. Acad. Sci. USA 101: 12422-12427. Huffman D.W., 2006. Production, losses, and germination of Ceanothus fendleri seeds in an Arizona ponderosa pine forest. West. N. Am. Naturalist 66: 365-373. Hulme P.E., 1997. Post-dispersal seed predation and the establishment of vertebrate dispersed plants in Mediterranean scrublands. Œcologia 111: 91-98. 12 Hulme P.E., 1998. Post-dispersal seed predation: consequences for plant demography and evolution. Perspect. Plant Ecol. 1: 32-46. Keeley J.E. & Hays R.L., 1976. Differential seed predation on two species of Arctostaphylos (Ericaceae). Œcologia 24: 71-81. Keeley J.E., 1977. Seed production, seed populations in soil, and seedling production after fire for two congeneric pairs of sprouting and nonsprouting chaparral shrubs. Ecology 58: 820-829. Keeley J.E., 1987. Ten years of change in the seed banks of the chaparral shrubs, Arctostaphylos glauca and A. glandulosa. Am. Midl. Nat. 117: 446-448. Keeley J.E., Bond W.J., Bradstock R.A., Pausas J.G. & Rundel P.W., 2012. Fire in mediterranean ecosystems. Cambridge Univ. Press, Cambridge. 522 p. Kelly V.R. & Parker V.T., 1990. Seed bank survival in sprouting and nonsprouting Arctostaphylos species. Am. Midl. Nat. 124: 114-123. Kitajima L. & Myers J.A., 2008. The seedling as part of a plant’s life history strategy. In: Leck M.A., Parker V.T. & Simpson R.L. (eds.), Seedling ecology and evolution. Cambridge Univ. Press, Cambridge: 172-188. Laakkonen J., 2003. Effect of arboreal activity on species composition and abundance estimates of rodents in a chaparral habitat in southern California Am. Midl. Nat. 150:348-351. Lawrence G., 1966. Ecology of vertebrate animals in relation to chaparral fire in the Sierra Nevada foothills. Ecology 47: 278-291. Leishman M.R. & Westoby M., 1998. Seed size and shape are not related to persistence in soil in Australia in the same way as in Britain. Funct. Ecol. 12: 480-485. Midgley J.J. & Anderson, B., 2005. Scatterhoarding in Mediterranean shrublands of the SW Cape, South Africa. In: Lambert J.E., Hulme P.E., & Vander Wall S.B (eds.), Seed fate: Predation, dispersal, and seedling establishment. CABI Publ., Wallingford: 197-204. Mills J.N., 1983. Herbivory and seedling establishment in post-fire southern California chaparral. Œcologia 60: 267-270. Mills J.N., 1986. Herbivores and early postfire succession in Southern California chaparral. Ecology 67: 1637-1649. Moles A.T. & Leishman M.R., 2008. The seedling as part of a plant’s life history strategy. In: Leck M.A., Parker V.T. & Simpson R.L. (eds.), Seedling ecology and evolution. Cambridge Univ. Press, Cambridge: 217-238. Moles A.T., Hodson D.W. & Webb C.J., 2000. Seed size and shape and persistence in the soil in the New Zealand flora. Oikos 89: 541-545. Moreno J.M. & Oechel W.C., 1991. Fire intensity and herbivory effects on postfire resprouting of Adenostoma fasciculatum in southern California chaparral. Œcologia 85: 429-433. Moritz C., Patton J.L., Conroy C.J., Parra J.L., White G.C. & Beissinger S.R., 2008. Impact of a century of climate change on small-mammal communities in Yosemite National Park, USA. Science 322: 261264. Muñoz A., Bonal R. & Espelta J.M. 2012. Responses of a scatter-hoarding rodent to seed morphology: links between seed choices and seed variability. Anim. Behav. 84: 1435-1442. ecologia mediterranea – Vol. 41 (2) – 2015 Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions O’Neil S.E. & Parker V.T., 2005. Factors contributing to the seed bank size of two obligate seeding Ceanothus species in Northern California. Madroño 52: 182-190. Odion D.C. & Davis F.W., 2000. Fire, soil heating, and the formation of vegetation patterns in chaparral. Ecol. Monogr. 70: 149-169. Parker V.T. & Kelly V.R. 1989. Seed banks in California chaparral and other mediterranean-climate shrublands. In: Leck M.A., Parker V.T. & Simpson R.L. (eds.), Ecology of soil seed banks. Academic Press, New York: 231-255. Parker V.T., Pratt J.B. & Keeley J.E., 2015. Chaparral. In: Mooney H.A. & Zavaleta E. (eds.), Ecosystems of California – A source book. Univ. of California Press, Berkeley: 479-507. Parker V.T., Vasey M.C. & Keeley J.E., 2012. Arctostaphylos. In: Baldwin B.G., Goldman D.H., Keil D.J., Patterson R., Rosatti T.J. & Wilken D.H. (eds.), The Jepson manual: Vascular plants of California (2nd ed.). Univ. of California Press, Berkeley: 689-699. Pons J. & Pausas J.G., 2007a. Acorn dispersal estimated by radio-tracking. Œcologia 153: 903-911. Pons J. & Pausas J.G., 2007b. Rodent acorn selection in a mediterranean oak landscape. Ecol. Res. 22: 535-541. Pons J. & Pausas J.G., 2007c. Not only size matters: acorn selection by the European jay (Garrulus glandarius). Acta Oecol. 31: 353-360. Price M.V., Waser N.M., Taylor K.E. & Pluff K.L., 1995. Fire as a management tool for Stephen’s kangaroo rat and other small mammal species. In: Keeley J. E. & Scott T. (eds.), Brushfire in California: ecology and resource management. Intn’l Assoc. of Wildland Fire, Fairfield: 51-61. Quinn R.D., 1994. Animals, fire, and vertebrate herbivory in californian chaparral and other mediterranean-type ecosystems. In: Moreno J.M. & Oechel W.C. (eds.), The role of fire in mediterranean-type ecosystems. Springer, New York: 46-78. R Development Core Team, 2013. R: a language and environment for statistical computing. The R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project. org/. Rusch U.D., Midgley J.J. & Anderson B., 2013. Rodent consumption and caching behaviour selects for specific seed traits. S. Afr. J. Bot. 84: 83-87. Rusch U.D., Midgley J.J. & Anderson B., 2014. Seasonal fluctuations in rodent seed caching and consumption ecologia mediterranea – Vol. 41 (2) – 2015 behaviour in fynbos shrublands: Implications for fire management. S. Afr. J. Bot. 93: 217-221. Schwilk D.W. & Keeley J.E., 1998. Rodent populations after a large wildfire in Calfornia chaparral and coastal sage scrub. Southwest. Nat. 43: 480-483. Schwilk D.W. & Keeley J.E., 1998. Rodent populations after a large wildfire in California chaparral and coastal sage scrub. Southwest. Nat. 43:480-483. Soule M.E., Bolger D.T., Alberts A.C., Wright J., Sorice M. & Hill S., 1988. Reconstructed dynamics of rapid extinctions of chaparral-requiring birds in urban habitat islands. Conserv. Biol. 2: 75-92. Thompson K., 2000. The functional ecology of soil seed banks. In: Fenner M. (ed.), Seeds: The ecology of regeneration in plant communities (2nd ed.). CABA Publishing, Wallingford: 215-235. Thompson K., Band S.R. & Hodgson J.G., 1993. Seed size and shape predict persistence in soil. Funct. Ecol. 7: 236-241. Tyler C.M., 1996. Relative importance of factors contributing to postfire seedling establishment in maritime chaparral. Ecology 77: 2182-2195. Vander Wall S.B., 1990. Food hoarding in animals. Univ. Chicago Press, Chicago, 445 p. Vander Wall S.B., 2001. The evolutionary ecology of nut dispersal. Bot. Rev. 67: 74-117. Vander Wall S.B., Forget P-M., Lambert J.E. & Hulme P.E., 2005. Seed fate pathways: filling the gap between parent and offspring. In: Forget P-M., Lambert J.E., Hulme P.E. & Vander Wall S.B. (eds.), Seed fate: Predation, dispersal and seedling establishment. CABI Publ., Wallingford: 1-8. Warzecha B. & Parker V.T., 2014. Differential postdispersal seed predation drives chaparral seed bank dynamics. Plant Ecol. 215: 1313-1322. Wilken D.H., 2012. Ceanothus. In: Baldwin B.G., Goldman D.H., Keil D.J., Patterson R., Rosatti T.J. & Wilken D.H. (eds.), The Jepson manual: Vascular plants of California (2nd ed.). Univ. of California Press, Berkeley: 1153-1162. Zammit C.A. & Westoby M., 1988. Pre-dispersal seed losses, and the survival of seeds and seedlings of two serotinous Banksia shrubs in burnt and unburnt heath. J. Ecol. 76: 200-214. Zammit C.A. & Zedler P.H., 1988. The influence of dominant shrubs, fire and time since fire on soil seed banks in mixed chaparral. Vegetatio 75: 175-187. Zammit C.A. & Zedler P.H., 1994. Organization of the soil seed bank in mixed chaparral. Vegetatio 111: 1-16. 13 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain Développement d’équations allométriques volume-biomasse pour aider à la caractérisation du combustible dans le nord-est de l’Espagne Beatriz DUGUY PEDRA*, Jesús GODOY PUERTAS, Laura FUENTES LOPEZ Departament de Biologia Vegetal, Facultat de Biologia, Universitat de Barcelona, Av. Diagonal, 643, Barcelona, Spain. * Corresponding author: [email protected] Received: 26 January, 2015; First decision: 11 March, 2015; Revised: 14 June, 2015; Accepted: 14 July, 2015 Abstract As a consequence of the drastic land use changes which occurred in Spain, as in other northern Mediterranean countries, throughout the 20th century, high amounts of fuels accumulated in landscapes and large fires increased in the last decades causing the degradation of ecosystems and landscapes. This trend will likely be enhanced by climatic change, which makes necessary the design at the landscape level of new fire management strategies. Fire models appear as crucial tools for supporting that task. FARSITE is a spatially-explicit fire growth and behavior model initially validated in the USA and successfully calibrated for Mediterranean conditions afterwards. The goodness of its predictions is however strongly dependent on the accuracy of the fuel-related inputs. In that sense, the description of customized fire behavior fuel models for Mediterranean plant communities is essential. When describing custom fuel models a crucial parameter is fuel load for which plant aboveground fine biomass can be considered as a proxy. The development of allometries is a rapid and nondestructive method used for estimating aboveground biomass. The main objective of this study is to establish allometric equations accurately describing Keywords: allometric relationship, fire model, customized fuel model, Mediterranean ecosystems. ecologia mediterranea – Vol. 41 (2) – 2015 the relationship between the apparent volume and aerial total biomass, as well as fine and coarse biomass fractions, for two shrub species, which are very common in dry Mediterranean shrublands of Eastern Spain: Pistacia lentiscus and Ulex parviflorus. Several volume approaches (total and canopy volume) and geometric shapes were explored for estimating the apparent volume based on field-collected dimensional variables. For a given species and a given volume approach, all shapes led to very similar volume-biomass relationships. In most cases, the apparent volume was a very good predictive variable for the aerial biomass, with r2 values generally larger than 0.7. The power-function model applied with the canopy apparent volume predicted better the total and fine biomass fractions. The smallest predictive power was observed for allometric equations predicting the U. parviflorus coarse aerial biomass, which is likely due to the small relative presence of this fraction in this species. The volume-fine biomass allometric relationships appear as suitable tools for contributing to the description of shrubland-type custom fuel models that would result in more accurate fuel model maps and more reliable fire modeling predictions. Mots clés : relation allométrique, programme de simulation du feu, modèle de combustible adapté, écosystèmes méditerranéens. 15 Beatriz Duguy Pedra, Jesús Godoy Puertas, Laura Fuentes Lopez 16 Résumé Introduction Les changements drastiques d’usage des terres qui ont eu lieu tout au long du xxe siècle en Espagne, comme dans d’autres pays de la Méditerranée nord, ont provoqué l’accumulation de grandes quantités de combustible dans les paysages et l’augmentation des grands feux de forêt dans les dernières décennies, ce qui a causé la dégradation des écosystèmes et des paysages. Cette tendance sera probablement accentuée par le changement climatique, ce qui rend nécessaire la conception de nouvelles stratégies de gestion du feu à l’échelle du paysage. Les modèles d’incendies sont des outils essentiels pour contribuer à cette tâche. FARSITE est un programme spatialement explicite de simulation de la propagation et du comportement du feu qui a été initialement validé aux États-Unis, puis calibré avec succès pour les conditions méditerranéennes. Néanmoins, l’adéquation de ses prédictions est fortement dépendante de la qualité des entrées associées aux combustibles. En ce sens, la description de modèles de combustible adaptés (i.e. non standard) pour les communautés végétales méditerranéennes est essentielle. Pour la description de modèles de combustible adaptés, la charge de combustible est un paramètre essentiel et la biomasse en éléments fins en surface peut être considérée comme un proxy. Le développement d’allométries est une méthode rapide et non destructive utilisée pour estimer la biomasse en surface. L’objectif principal de ce travail est d’établir des équations allométriques qui décrivent avec précision la relation entre le volume apparent et la biomasse totale en surface, ainsi que celle des fractions fines et grossières, pour deux espèces d’arbustes qui sont très communes dans les formations arbustives méditerranéennes de l’est de l’Espagne : Pistacia lentiscus et Ulex parviflorus. Plusieurs approches du volume (volume total et de canopée) et formes géométriques ont été explorées pour estimer le volume apparent à partir des variables dimensionnelles mesurées sur le terrain. Pour une espèce et un type de volume donnés, toutes les formes ont mené à des relations volume-biomasse très similaires. Dans la plupart des cas, le volume apparent a été une très bonne variable prédictive de la biomasse en surface, avec des valeurs de r2 généralement supérieures à 0,7. Le modèle potentiel appliqué avec le volume apparent de canopée est celui qui prédit le mieux les biomasses totales et éléments fins. La plus faible capacité de prédiction a été observée avec les équations allométriques de la biomasse grossière en surface d’U. parviflorus, ce qui est probablement dû à la faible présence relative de cette fraction dans cette espèce. Les relations allométriques volume-biomasse en éléments fins semblent donc des outils très appropriés pour contribuer à la description de modèles de combustible adaptés qui conduiraient à des cartes de modèles de combustibles plus précises et à des prédictions plus fiables des programmes de simulation du feu. The profound socioeconomic changes that occurred in Spain, as in other northern Mediterranean countries during the 20th century, led to intense land abandonment (FernándezAles et al. 1992; Vallejo et al. 2012). The decrease in agricultural and grazing activities combined with large-scale afforestation policies often caused the disappearance of the former mosaic-like landscape and the accumulation of fuel in landscapes (Lloret et al. 2002; Duguy 2003), which in turn resulted in a drastic alteration of fire regimes (Fernandes et al. 2014; Moreno et al. 2014). Large severe fires increased in the 1970s becoming a major cause of environmental degradation in most Mediterranean ecosystems and landscapes (Díaz-Delgado et al. 2002; Vallejo et al. 2012). The magnitude of the problem, which has very negative economic and ecological effects, will likely be enhanced by climatic change (Piñol 1998; Pausas & Keeley 2009), making necessary the implementation of landscape-level designed fire management strategies (Duguy et al. 2013). Given the difficulty and obvious limitations in implementing large-scale and long-term experiments on fuel treatments and in assessing their performance in relation to fire control, the design of effective landscape-level fuel management policies requires the use of theoretical and modeling-based approaches that may allow to minimize the arbitrariness in the planning processes while optimizing the effectiveness of fuel treatments (Hiers et al. 2003; Finney 2004; Fernandes 2006). Spatially-explicit fire models, such as FARSITE (Finney 1998), have shown to correctly project fire growth and behavior of hypothetical fires through real landscapes and have become very powerful decision-support tools for land managers (Stratton 2006; Thompson et al. 2011). FARSITE was developed and initially calibrated in the USA (Finney & Ryan 1995; Finney 1998) where it has been widely applied for evaluating the effects of different fuel treatment options on reducing fire hazard (Stephens 1998; Finney 2001; Finney et al. 2007; Schmidt et al. 2008). It has also been successfully parameterized for Mediterranean landscapes (Arca et al. 2007; Duguy et al. 2007). The reliability of its predictions strongly depends, however, on the accuracy of fuel-related inputs (Arca et al. 2007) and that of the fuel model map, in particular. In ecologia mediterranea – Vol. 41 (2) – 2015 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain that sense, the characterization of custom fuel models developed for Mediterranean plant communities is crucial (Arca et al. 2007; Rodríguez y Silva & Molina-Martínez 2012) and more efforts have to be devoted to the rigorous calibration of these models aiming to maximise their performance, i.e. the fit between observed and predicted fire behaviour (Cruz & Fernandes 2008; Ascoli et al. 2015). Such calibrated custom fuel models are expected to improve fire models predictions and thus to promote the use of fire modelingbased approaches for assessing the effectiveness of landscape-level alternative fuel treatment strategies in relation to fire control in the Mediterranean region. The accurate characterization of custom fuel models requires extended datasets of fieldcollected information about several fuel structure variables. The compilation of such datasets is often impossible due to temporal and economical limitations. One of the essential structural parameters required to characterize a plant community as a fuel complex is the aerial fuel load for which aboveground biomass, and more specifically its fine fraction (i.e. branch diameter < 6 mm), may be often considered as a suitable proxy. It is important, thus, to develop rapid and nondestructive methods, such as allometric relationships, for estimating aboveground biomass and its fine and coarse fractions in the most extended plant communities across the Mediterranean region. In the framework of a broader project, in which this study is included, we intend to develop allometric regression equations to predictaboveground biomass fractions (total, fine and coarse) as a function of the volume for various shrub species, which are often among the most abundant in dry Mediterranean shrublands of eastern Spain. Such allometric relationships are very scarce for Mediterranean shrubs so far. The first stage of this work is presented in this study and focuses on Pistacia lentiscus L. (Anacardiaceæ) and Ulex parviflorus Pourr. (Fabaceæ). The former species reaches the largest average specific cover and phytovolume in the shrub layer of the studied plant community, whereas the latter is critical in relation to fuel modeling since it accumulates fine dead phytomass with development stage, which strongly contributes to the increase in fire risk (Baeza et al. 2006). ecologia mediterranea – Vol. 41 (2) – 2015 Methods Study site The study site is located in the municipality of El Perelló, Tarragona province, north-eastern Spain (40º 54’ 19.3’’ N, 00º 40’ 53.0’’ E). The height ranges from 230 to 245 m.a.s.l. (ICGC 2015). It has a dry Mediterranean climate, with a mean annual temperature of 15.9 ºC and a mean annual precipitation of 565 mm, after data registered by the nearest weather station, El Perelló, for the 1999-2013 period (SMC 2015). The soil studies carried out by Mora Betancort (2013) in the study area showed that the dominant soils are Lithic xerorthents and Lithic haploxerepts, after keys to soil taxonomy (Soil Survey Staff, 2010), over limestone, and are generally shallow (< 50 cm). The area is dominated by Pinus halepensis Mill. forests planted in 1970, with an understory of either macchia of Querco-Lentiscetum Br.-Bl. et al. 1935 em. A. et O. Bolòs 1950 (dominated by Quercus coccifera L., Chamaerops humilis L. and Pistacia lentiscus L.) or Rosmarino-Ericion Br.-Bl. 1931 shrublands (dominated by Rosmarinus officinalis L. and Erica multiflora L.). The herbaceous layer is dominated by Brachypodium retusum (Pers.) Beauv. Other abundant species are Aphyllanthes monspeliensis L. and Carex halleriana Asso. The digital fire perimeter maps produced by the regional administration (ICC 2015) showed that no fire occurred in the study site during the 1986-2013 period. This information combined with the revision of the available digital orthophotos showed that the studied plant communities have been present for at least the last 28 years. Experimental design Field sampling Field work was carried out in February, March and April 2013 in the shrub understory of a Pinus halepensis stand in which Pistacia lentiscus and Ulex parviflorus were two of the dominant species. Twenty individuals of each of these two shrub species were randomly selected taking into account the whole dimensional range (focusing on the apparent volume, more specifically) 17 Beatriz Duguy Pedra, Jesús Godoy Puertas, Laura Fuentes Lopez of each species in the studied plant community. Each individual was measured (maximum height, crown base height, minimum and maximum crown diameters), cut (at the basis of the stem), weighted (for obtaining the total aboveground biomass) and reweighted in the field after removing the fine fuel fraction (< 6 mm diameter). The aerial plant biomass could be thus separated into fine and coarse categories. The former is required for the characterization of fire behavior fuel models, which do not consider coarse live vegetation (Scott & Burgan 2005). One U. parviflorus individual had to be discarded later due to measurement incongruities. Estimation of the apparent volume and search of the best regression equations For each species, three geometric shapes (elliptical cylinder, rectangular prism and elliptical cone) and two volume approaches (total and canopy volume) were used for estimating the apparent volume (explanatory or independent variable) of each individual, based on the four field-collected dimensional variables (Table 1). The selection of these geometric shapes was based on the similarity between them and the studied species’ morphologies as well as on previous studies (Usó et al. 1997). For each species and each geometric shape, four regression functions (linear, power, exponential and logarithmic) were tested with the two volume approaches, looking for the ‘best’ regression equation between the volume and any of the three considered aboveground biomass fractions (total, fine and coarse). This “best-fitting” equation is defined as the one with the largest coefficient of determination, or r2 value. Two additional statistics on model performance were also computed: the mean absolute error (MAE) and the mean absolute Table 1 – Geometric shapes, plant apparent volumes and regression functions used for developing the allometric equations. Function Geometric shape Y = a + bX elliptical cylinder Y = aXb Y = c + aexp(bX) Apparent Volume total volume rectangular prism canopy volume Y = aLogb(x) + c 18 elliptical cone percentage error (MA%E) (Cruz & Fernandes 2008). We thus explored 24 allometric equations for each combination of a given species and geometric shape (Tables 2 and 3), that is 72 equations for a given species considering the three tested shapes. The assumptions of normality and homocedasticity were tested for all variables (independent and dependent) in SPSS Statistics v22.0.0. The level of significance for all tests was set to α = 0.05. Both assumptions could only be met through transformation of the variables. In the case of P. lentiscus, the volume variables (total and canopy volumes) were fourth root transformed as well as the total and coarse aerial biomasses. The fine aerial biomass was ln transformed. In the case of U. parviflorus, all volume and biomass variables were square root transformed, except the total aerial biomass, which was fourth root transformed when predicted with the total volume. Results are presented using the transformed variables meeting the assumptions (Tables 2 and 3). Results For a given species, one type of apparent volume (total or canopy), a given biomass fraction and one of the four regression functions, the three tested geometric shapes always led to very similar volume-biomass relationships and coefficients of determination. We will therefore only present and comment the results obtained with the elliptical cylinder (Tables 2 and 3). For both species, for a given dependent variable and regression function, only slight differences were observed when comparing the r2 values obtained with each type of apparent volume (Tables 2 and 3). For a given biomass variable, no significant differences were found indeed when comparing the values predicted with each of the two tested apparent volumes (Table 4). In the case of Pistacia lentiscus, the coefficient of determination was always higher than 0.8 (Table 2). The best-fitting equations obtained for predicting the total, fine and coarse aerial biomasses reached r2 values of 0.971, 0.88 and 0.962, respectively, and were all developed with the power function and the canopy apparent volume (Figure 1). The lowest r2 values, although still high, were observed among the relationships predicting the fine aboveground biomass and, more specifically, ecologia mediterranea – Vol. 41 (2) – 2015 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain Table 2 – Allometric equations obtained with the elliptical cylinder for Pistacia lentiscus (n = 20). TAV and CAV (total and canopy apparent volume, respectively) are in cm3 and 4th root transformed. MAE: mean absolute error; MA%E: mean absolute percentage error. Dependent variable (and transformation applied), in g Allometric equation Total aerial biomass, TAB (4th root) Fine aerial biomass, FAB (ln) Coarse aerial biomass, CAB (4th root) r2 MAE (g) MA%E (%) TAB = 0.1684*TAV + 1.2708 0.958 0.48 6.10 TAB = 0.1737*CAV + 1.8488 0.958 0.53 7.04 TAB = 0.361*TAV0.843 0.967 0.46 5.66 TAB = 0.522*CAV0.772 0.971 0.47 5.56 TAB = 3.0868e0.0213*TAV 0.906 0.85 10.36 TAB = 3.3329e0.0218*CAV 0.897 0.97 11.76 TAB = 6.3568*ln(TAV) – 14.562 0.927 0.78 10.62 TAB = 5.8303*ln(CAV) – 11.832 0.934 0.75 10.30 FAB = 0.0723*TAV + 4.2629 0.842 0.53 7.41 FAB = 0.0742*CAV + 4.5254 0.834 0.57 7.90 FAB = 1.667*TAV0.403 0.879 0.45 6.31 FAB = 1.990*CAV 0.880 0.47 6.55 FAB = 4.6485e0.0101*TAV 0.820 0.59 8.02 FAB = 4.8243e0.0104*CAV 0.809 0.63 8.53 FAB = 2.8195*ln(TAV) – 2.8553 0.869 0.45 6.51 FAB = 2.5803*ln(CAV) – 1.6248 0.872 0.46 6.60 CAB = 0.1642*TAV + 0.1486 0.954 0.58 10.62 CAB = 0.1697*CAV + 0.7036 0.956 0.60 11.45 CAB = 0.135*TAV1.0567 0.960 0.59 9.04 CAB = 0.215*CAV0.9664 0.962 0.59 9.22 CAB = 2.0341e0.0261*TAV 0.860 1.19 18.73 CAB = 2.2373e0.0267*CAV 0.849 1.20 19.31 CAB = 6.239*ln(TAV) – 15.432 0.934 0.72 12.27 CAB = 5.7265*ln(CAV) – 12.767 0.943 0.66 11.11 0.368 Figure 1 – B est-fitting apparent volume-biomass regression equations for all aerial biomass fractions of Pistacia lentiscus (n = 20). (Any transformation applied to any variable is indicated in the corresponding axis.) for those obtained with the exponential function (0.82 and 0.809, using the total and the canopy apparent volume, respectively). When considering a given independent variable and regression function, the smallest r2 was always observed when predicting fine aerial biomass (Table 2). The statistic error confirms the good performance of all models (Table 2). For the three dependent variables, and whatever the independent variable, the smallest MAE values ecologia mediterranea – Vol. 41 (2) – 2015 were generally obtained with the power function, although the logarithmic and linear functions led to very similar results, in the case of fine and coarse aerial biomass, respectively (Table 2). In the case of Ulex parviflorus, r2 ranged between 0.726 and 0.878, when considering the total and the fine aerial biomasses, but it was generally smaller, ranging between 0.363 and 0.779, for the coarse aerial biomass equations (Table 3). The highest r2 values for the 19 Beatriz Duguy Pedra, Jesús Godoy Puertas, Laura Fuentes Lopez Table 3 – Allometric equations obtained with the elliptical cylinder for Ulex parviflorus (n = 19). TAV and CAV (total and canopy apparent volume, respectively) are in cm3 and square root transformed. MAE: mean absolute error; MA%E: mean absolute percentage error. Dependent variable (and transformation applied), in g Allometric equation Total aerial biomass, TAB (4th root with TAV; sq. root with CAV) Fine aerial biomass, FAB (sq. root) Coarse aerial biomass, CAB (sq. root) r2 MAE (g) MA%E (%) TAB = 0.0041*TAV + 2.0555 0.789 0.28 7.55 TAB = 0.0346*CAV + 2.7328 0.854 1.63 12.14 TAB = 0.2273*TAV0.4672 0.867 0.25 6.42 TAB = 0.1033*CAV0.8484 0.878 1.64 11.70 TAB = 2.2652e0.0012*TAV 0.767 0.29 8.06 TAB = 5.4205e0.0027*CAV 0.807 1.83 14.10 TAB = 1.6054*ln(TAV) – 5.804 0.853 0.23 5.81 TAB = 10.609*ln(CAV) – 46.27 0.854 1.64 10.58 FAB = 0.0239*TAV + 2.822 0.726 1.92 15.36 FAB = 0.0284*CAV + 2.9963 0.825 1.48 12.08 FAB = 0.0741*TAV0.8529 0.835 1.83 13.51 FAB = 0.1297*CAV 0.873 1.44 11.12 FAB = 4.9247e0.0021*TAV 0.739 2.15 16.68 FAB = 5.1031e0.0025*CAV 0.803 1.67 13.87 FAB = 9.2897*ln(TAV) – 42.505 0.762 1.67 12.28 FAB = 8.7465*ln(CAV) – 37.413 0.830 1.33 9.58 CAB = 0.0187*TAV – 0.8976 0.754 1.34 16.86 CAB = 0.0207*CAV – 0.1974 0.743 1.35 18.34 CAB = 3E-07*TAV2.7904 0.560 3.79 46.79 CAB = 7E-06*CAV2.3202 0.476 2.62 36.27 CAB = 0.319e0.0062*TAV 0.402 3.08 41.08 CAB = 0.4487e0.0066*CAV 0.363 2.79 36.73 CAB = 7.1924*ln(TAV) – 35.921 0.779 1.21 17.05 CAB = 6.3352*ln(CAV) – 29.452 0.742 1.27 19.00 0.7869 Figure 2 – B est-fitting apparent volume-biomass regression equations for all aerial biomass fractions of Ulex parviflorus (n = 19). (Any transformation applied to any variable is indicated in the corresponding axis.) total, fine and coarse aerial biomass were 0.878, 0.873 and 0.779, respectively (Figure 2). They were obtained with the power function and the canopy apparent volume for the first two variables, but with the logarithmic function and the total apparent volume in the case of the coarse aerial biomass. For a given predictive variable and regression function, the lowest r2 value was generally found with the equation predicting the coarse biomass fraction. For this biomass fraction, the goodness-of-fit was particularly low with the 20 power and the exponential functions (with the latter function, r2 was only 0.402 and 0.363 for the total and the canopy apparent volume, respectively). In fact, considering both species, the smallest r2 value observed for a given aerial biomass fraction was generally associated to the exponential function. Generally, for a given species, apparent volume and regression function, the best-fitting equation was the one established for predicting the total aerial biomass (Tables 2 and 3). ecologia mediterranea – Vol. 41 (2) – 2015 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain Table 4 – Results of the t-test applied for comparing the dependent variables predicted either with total apparent volume or canopy apparent volume. (In each case, the best-fitting volume-biomass regression was used.) Dependent variable (transformation applied) Levene’s test for equality of variances F P t-test for equality of means t df P P. lentiscus_total aerial biomass (4th root) 0.005 0.943 -0.007 38 0.994 P. lentiscus_fine aerial biomass (ln) 0.011 0.917 -0.002 38 0.999 P. lentiscus_coarse aerial biomass (4 root) 0.002 0.962 -0.006 38 0.995 U. parviflorus_total aerial biomass (sq. root) 0.018 0.894 -0.009 36 0.993 U. parviflorus_fine aerial biomass (sq. root) 0.047 0.830 -0.026 36 0.979 U. parviflorus_coarse aerial biomass (sq. root) 0.000 0.992 0.033 36 0.974 th df: degrees of freedom The differences observed in the goodness of prediction among the various biomass fractions were often slight, though, except for U. parviflorus and the coarse biomass fraction. In that case, both the power and exponential functions led to much smaller r2 values (from 0.363 to 0.56) than the ones reached by the predictive equations of total or fine aerial biomass fractions with these same functions. The statistic error showed, as the r2 values did, a lower performance of the models for U. parviflorus than P. lentiscus (Table 3). The smallest goodness-of-fit was again observed among the equations predicting U. parviflorus coarse aerial biomass. The largest MAE and MA%E values were found with the power and the exponential functions (Table 3). Discussion Our results indicate that the total aerial biomass and the two explored biomass fractions (fine, coarse) can be suitably predicted by the apparent volume for both Pistacia lentiscus and Ulex parviflorus. Only few studies that focused on the development of allometries worked so far with Mediterranean shrubs and used the apparent volume as the explanatory variable in their allometric models (Usó et al. 1997). Although obtaining rather poor results for U. parviflorus (r2 value always ≤ 0.58), Pereira et al. (1995) still considered their volume-biomass allometric models as adequate for estimating U. parviflorus fuel loadings at the forest stand level and for mapping understory fuels in Portuguese Pinus pinaster stands dominated by this shrub species. ecologia mediterranea – Vol. 41 (2) – 2015 Working in shrublands under P. halepensis in Alicante province (eastern Spain), Usó et al. (1997) reported that the allometric regression equations describing the relationship between aerial biomass and apparent volume were a sufficiently accurate method for estimating the total aboveground biomass of Rosmarinus officinalis and Cistus albidus, two common Mediterranean shrubs. Baeza et al. (2006) worked in shrublands of U. parviflorus at different development stages (young, mature and senescent) with no tree layer and developed allometric equations for predicting total aerial phytomass using the stem diameter. They found high determination coefficients for all stages, globally ranging from 0.88 to 0.95. The r2 values corresponding to the young (n = 71) and mature (n = 44) stages (0.88 and 0.94 respectively) were very similar or slightly larger than the best r2 value found in our study for estimating the total aboveground biomass of this species (0.878). Although our plant communities were different, most U. parviflorus in our study site could be classified in the young (sometimes mature) stage as defined by Baeza et al. (2006). The use of stem diameter as the predictive variable for aerial biomass appears, thus, as very suitable also. However, it should be pointed out that the use of this variable may be more difficult than the use of the apparent volume in many situations, as previously remarked by Usó et al. (1997). Its accurate measurement can be obviously quite complicated when working with prickly species, such as U. parviflorus, but also dealing with species that have several stems. We found no studies using either the apparent volume, or any dimensional variable, for developing allometric equations predicting 21 Beatriz Duguy Pedra, Jesús Godoy Puertas, Laura Fuentes Lopez the aerial biomass of P. lentiscus. Given the high presence of this species in many types of Mediterranean shrublands, which are usually very extended, and, therefore, its high importance in relation to fire behavior studies, it seems crucial to carry out more studies aiming to accurately estimate its aerial phytomass through easy methods. Our results also show that in the case of P. lentiscus the apparent volume-biomass equations are slightly better at predicting the coarse aerial biomass than the fine aerial biomass. It is the contrary for U. parviflorus for which the fine biomass is clearly better predicted. Although more studies are required in order to know if this result is a general rule that applies for both species in any plant community or not, we can still hypothesize that it might be related to the relative abundance of these two biomass fractions (fine, coarse) in each species. This relative abundance is clearly different between the two studied shrubs. Based on the data measured in the field, the average values of the ratio fine aerial biomass over total aerial biomass were 0.52 (± 0.23) and 0.78 (± 0.1) for P. lentiscus and U. parviflorus, respectively. These results suggest that the aerial biomass fraction reaching the largest proportion in a given species would be better predicted by the apparent volume-biomass equation approach. The poor predictive power for the U. parviflorus coarse aerial biomass that we obtained with several allometric equations would be, therefore, a consequence of the rather small relative presence of this fraction in the individuals of this species that we sampled. Besides, the total apparent volume was a better predictive variable for this U. parviflorus coarse aerial biomass than the canopy apparent volume. It is probably due to the fact that a large part of the standing coarse fuels are not considered when working with the canopy apparent volume since the canopy of U. parviflorus is mostly made up by fine fuels. It is noteworthy that we did not find many differences among the coefficients of determination of the various tested geometric shapes. It is likely that we could have found more differences if we had worked with a larger number of shapes. Usó et al. (1997) observed some differences between the circular or elliptical cylinders and the rotation paraboloid, but never between the circular and elliptical cylinders. In our study, we did not work with the rotation paraboloid since it was reported that this shape 22 leads to an overestimate of the true volume for shrubs, given that the radius is measured at the base (Whittaker & Woodwell 1968). Our results suggest that the three geometric shapes that we tested were a good choice for both P. lentiscus and U. parviflorus since they were generally very appropriate for estimating aerial biomass based on the apparent volume. However, other shapes could maybe improve the prediction of U. parviflorus coarse aerial biomass and this aspect needs to be further explored. As for the type of apparent volume, the fact that no significant differences were found between the predictions obtained with either the total, or the canopy apparent volume suggests that both approaches are suitable for the two studied species. It is advisable, however, to always explore both approaches since one or the other can maybe lead to better results depending on the plant species. This aspect is probably more critical working with species for which one type of fuel fraction (fine or coarse) is clearly dominating in the aboveground biomass, as it happens for U. parviflorus. Given their good performance, the developed allometries, and more specifically the ones predicting the fine aerial biomass, might contribute to fuel characterization in Mediterranean plant communities dominated by the shrub species considered in this study. Some limitations will have to be taken into account though, and solved on a case-by-case basis with the suitable complementary datasets and methodological approaches. The equations do not apply to the community level but to single plants. In our case study, an ongoing vegetation sampling will result in a detailed description of this understory shrub community through a number of composition and structural variables estimated in 10 × 10 m plots divided in 1-m2 quadrats. The grass and the shrub (total and specific) covers as well as the shrub phytovolumes (after height and diameter measurements for all individuals of the dominant species) are being estimated. These data will allow, once combined with the volume-fine biomass equations established for the dominant shrubs, to estimate quite accurately the fine fuel load present in the community. The fact that the proposed allometric equations for fine aerial biomass do not separate live and dead fuels needs to be considered also. In the studied plant community this aspect may be particularly limiting in the case of U. parviflorus, ecologia mediterranea – Vol. 41 (2) – 2015 Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain which is the only species among the dominant shrubs that accumulates large quantities of standing fine dead phytomass. In this community, however, this accumulation is rather low, suggesting that most U. parviflorus individuals are young (the average height is smaller than 50 cm). Baeza et al. (2006) documented that in young U. parviflorus shrublands the living twigs fraction was the largest component in the phytomass representing 91.8% of it. For this species, however, the quotient between the dead and live phytomass fractions shows high variations with growth phases and a proper discrimination between live and dead fuels will be needed when the fuel modeling work will be implemented in later development stages or in other communities. Acknowledgements We gratefully acknowledge funding from the Spanish Ministry of Economy and Competitiveness to the ForBurn-Land project (AGL2012-40098-C03-02). We thank Dr. C. Vega-Garcia (University of Lleida), S. Costafreda-Aumedes (University of Lleida), M. Prendin Navarro, J. Garcia and A. Cardil who collaborated in the field work. We also sincerely acknowledge the useful comments and suggestions of the anonymous referees who helped improve an earlier version of the manuscript. Dr. Duguy is a member of FORESTREAM (Research Group on Forest and Stream Ecological Links: Watershed Management and Restoration) funded by Generalitat de Catalunya (2014 SGR 0949). Conclusions The apparent volume of a shrub, which can be easily estimated after few dimensional variables measured in the field, appears as a satisfactory predictor of total aboveground biomass, but also of the fine and coarse biomass fractions, for both P. lentiscus and U. parviflorus. The coarse aerial biomass of the latter species seems harder to model with this type of allometric equations, though, which is likely due to the low relative abundance of this biomass fraction in this species. More studies including other shrub species and a larger number of geometric shapes need to be carried out in order to better understand to which degree the relative abundances of biomass fractions determine the goodnessof-fit of the apparent volume-biomass regressions and which combinations of geometric shape, type of apparent volume and regression function may be more appropriate for each species. In any case, our results confirm that apparent volume-biomass allometric equations can be quite an easy approach for accurately estimating shrub fuel loads, and fine fuel loads in particular, and thus may be very useful for the characterization of customized fire behavior fuel models in Mediterranean ecosystems. Such fuel models, once calibrated, would improve the reliability of fire spread and behavior model projections and, therefore, help forest managers for designing more effective landscape-level fire management strategies under climate change. ecologia mediterranea – Vol. 41 (2) – 2015 References Arca B., Duce P., Laconi M., Pellizzaro G., Salis M. & Spano D., 2007. Evaluation of FARSITE simulator in Mediterranean maquis. Int. J. Wildland Fire 16(5): 563-572. Ascoli D., Vacchiano G., Motta R. & Bovio G., 2015. Building Rothermel fire behaviour fuel models by genetic algorithm optimisation. Int. J. Wildland Fire 24(3): 317-328. Baeza M.J., Raventós J., Escarré A. & Vallejo V.R., 2006. Fire risk and vegetation structural dynamics in Mediterranean shrubland. Plant Ecol. 187(2): 189-201. Cruz M.G. & Fernandes P.M., 2008. Development of fuel models for fire behaviour prediction in maritime pine (Pinus pinaster Ait.) stands. Int. J. Wildland Fire 17(2): 194-204. Díaz-Delgado R., Lloret F. & Pons F.X., 2002. Satellite evidence of decreasing resilience in Mediterranean plant communities after recurrent wildfires. Ecology 83: 2293-2303. Duguy B., 2003. Interacción de la historia de usos del suelo y el fuego en condiciones mediterráneas. Respuesta de los ecosistemas y estructura del paisaje. Tesis Doctoral. Facultad de Ciencias, Universidad de Alicante. 281 p. Duguy B., Alloza J.A., Röder A., Vallejo R. & Pastor F., 2007. Modelling the effects of landscape fuel treatments on fire growth and behaviour in a Mediterranean landscape (eastern Spain). Int. J. Wildland Fire 16(5): 619-632. Duguy B., Paula S., Pausas J.G., Alloza J.A., Gimeno T. & Vallejo V.R., 2013. Effects of climate extreme events on wildfire regime and their ecological impacts. In: Navarra A. & Tubiana L. (eds.), Regional assessment of climate change in the Mediterranean: Volume 2. Agriculture, forests and ecosystem services and people. Springer, Dordrecht: 101-134. Fernandes P.M., 2006. Silvicultura preventiva e gestão de combustíveis: opções e optimização. In: Pereira J.S., Pereira J.C., Rego F.C., Silva J.M.N. & Pereira 23 Beatriz Duguy Pedra, Jesús Godoy Puertas, Laura Fuentes Lopez da Silva T. (eds.), Incêndios Florestais em Portugal. ISA Press, Lisboa: 327-354. Fernandes P.M., Loureiro C., Guiomar N., Pezzatti G.B., Manso F. & Lopes L., 2014. The dynamics and drivers of fuel and fire in the Portuguese public forest. J. Env. Manag. 146: 373-382. Fernández-Ales R., Martin A., Ortega F. & Ales E., 1992. Recent changes in landscape structure and function in a mediterranean region of SW Spain (1950-1984). Landscape Ecol. 7(1): 3-18. Finney M.A. & Ryan K.C., 1995. Use of the FARSITE fire growth model for fire prediction in U.S. National Parks. In: Proceedings of the International Emergency Management and Engineering Conference, Sofia Antipolis, France, May 1995: 183-189. Finney M.A., 1998. FARSITE: Fire Area Simulator– model development and evaluation. Research Paper RMRS-RP-4, USDA Forest Service, Fort Collins. Finney M.A. 2001. Spatial strategies for landscape fuel treatments. In: Bento J. & Botelho H. (eds.), Workshop on tools and methodologies for fire danger mapping: proceedings of a workshop. Vila Real, Portugal, Universidad de Traj-os-montes el AltoDouro, Departamento Forestal: 157-163. Finney M.A., 2004. Landscape fire simulation and fuel treatment optimization. In: Hayes J.L., Ager A.A. & Barbour J.R. (eds.), Methods for Integrated Modeling of Landscape Change. Interior Northwest Landscape Analysis System. Ch. 9. Gen. Tech. Rep. PNW-GTR-610, USDA Forest Service, Pacific Northwest Research Station, Portland: 117-131. Finney M.A., Seli R.C., McHugh C.W., Ager A.A., Bahro B. & Agee J.K., 2007. Simulation of longterm landscape-level fuel treatment effects on large wildfires. Int. J. Wildland Fire 16: 712-727. Hiers J.K., Laine S.C., Bachant J.J., Furman J.H., Greene Jr. W.W. & Compton V., 2003. Simple Spatial Modeling Tool for Prioritizing Prescribed Burning Activities at the Landscape Scale. Conserv. Biol. 17(6): 1571-1578. ICC, 2015. Institut Cartogràfic de Catalunya – Cartographic Institute of Catalonia, maps of forest fires: http://www.atlesnacional.cat/icc/atles-nacional/riscos-naturals/zones-cremades-per-incendi-forestal/ ICGC, 2015. Institut Cartogràfic i Geològic de Catalunya – Cartographic and Geological Institute of Catalonia. Topographic maps (1:5.000). http://www. icc.cat/vissir3/ Lloret F., Calvo E., Pons X. & Díaz-Delgado R., 2002. Wildfires and landscape patterns in the Eastern Iberian Peninsula. Landscape Ecol. 17: 745-759. Mora Betancort A., 2013. Efecto de una quema prescrita de intensidad moderada sobre el suelo bajo matorral mediterráneo. Tesis de Master. Universitat de Lleida. 49 p. Moreno M.V., Conedera M., Chuvieco E. & Pezzatti G.B., 2014. Fire regime changes and major driving 24 forces in Spain from 1968 to 2010. Environ. Sci. Policy 37: 11-22. Pausas J.G. & Keeley J.E., 2009. A burning story: the role of fire in the history of life. BioScience 59(7): 593-601. Pereira J.M.C., Sequeira N.M.S. & Carreiras J.M.B., 1995. Structural Properties and Dimensional Relations of Some Mediterranean Shrub Fuels. Int. J. Wildland Fire 5(1): 35-42. Piñol J., Terradas J. & Lloret F., 1998. Climatic warming, wildfire hazard and wildfire occurrence in coastal eastern Spain. Climatic Change 38: 345-357. Rodríguez y Silva F. & Molina-Martínez J.R., 2012. Modeling Mediterranean forest fuels by integrating field data and mapping tools. Eur. J. Forest Res. 131: 571-582. Schmidt D.A., Taylor A.H. & Skinner C.N., 2008. The influence of fuels treatment and landscape arrangement on simulated fire behavior, Southern Cascade range, California. For. Ecol. Manag. 255(8–9): 3170-3184. Scott J.H. & Burgan R.E., 2005. Standard fire behavior fuel models: a comprehensive set for use with Rothermels surface fire spread model. Gen. Tech. Rep. RMRS-GTR-153, USDA Forest Service, Rocky Mountain Research Station, Ogden: 80 p. SMC, 2015. Servei Meteorològic de Catalunya - Catalan Weather Service. El Perelló station (1999-2013): http://www.meteo.cat/wpweb/climatologia/anuarisde-dades-meteorologiques/ Soil Survey Staff, 2010. Keys to Soil Taxonomy. 11th edition. USDA, NRCS, Washington DC: 338 p. Stephens S.L., 1998. Evaluation of the effects of silvicultural and fuels treatments on potential fire behaviour in Sierra Nevada mixed-conifer forests. For. Ecol. Manag. 105(1-3): 21-35. Stratton R.D., 2006. Guidance on spatial wildland fire analysis: models, tools, and techniques. Gen. Tech. Rep. RMRS-GTR-183. USDA Forest Service, Rocky Mountain Research Station, Fort Collins: 15 p. Thompson M.P., Calkin D.E., Finney M.A., Ager A.A. & Gilbertson-Day J.W., 2011. Integrated nationalscale assessment of wildfire risk to human and ecological values. Stoch. Env. Res. Risk A. 25: 761-780. Usó J.L., Mateu J., Karjalainen T. & Salvador P., 1997. Allometric regression equations to determine aerial biomasses of Mediterranean shrubs. Plant Ecol. 132(1): 59-69. Vallejo V.R., Allen E.B., Aronson J., Pausas J.G., Cortina J. & Gutiérrez J.R., 2012. Restoration of Mediterranean-type woodlands and shrublands. In: Van Andel J. & Aronson J., (eds.), Restoration Ecology: The New Frontier. 2nd edition. Wiley-Blackwell, Oxford, UK: 130-144. Whittaker R.H. & Woodwell G.M., 1968. Dimension and production relations of trees and shrubs in the Brookhaven Forest, New York. J. Ecol. 56(1): 1-25. ecologia mediterranea – Vol. 41 (2) – 2015 Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; Key Factors for Successful Restoration of Mediterranean Ecosystems Étude de l’architecture des parties aériennes et racinaires et de la croissance des semis de Quercus ithaburensis subsp. macrolepis ; facteurs clés pour une restauration réussie des écosystèmes méditerranéens Thekla TSITSONI*, Marianthi TSAKALDIMI, Maria GOUSIOPOULOU Laboratory of Silviculture, Department of Forestry & Natural Environment, School of Agriculture, Forestry and Natural Environment Aristotle University of Thessaloniki P.O. BOX 262, University Campus, 54124 Thessaloniki, Greece *Corresponding author: [email protected] Received: 19 January, 2015; First decision: 28 March, 2015; Revised: 2 June, 2015; Second decision: 20 July, 2015; Revised: 28 August, 2015; Accepted: 30 September, 2015 Abstract Quercus ithaburensis subsp. macrolepis is a species of high ecological importance. It forms some of the few deciduous oak forests in the eastern Mediterranean zone, is well adapted to Mediterranean conditions and warm dry periods, develops a deep root system, regenerates easily after fire and is used in many restoration projects. However, information on the nursery production method of Quercus macrolepis seedlings and the growth and architectural development of these produced seedlings is limited. In order to select the best cultivation method to produce high quality seedlings, the objectives of this study were to a) investigate shoot and root architecture of seedlings of Quercus ithaburensis subsp. macrolepis and b) compare the growth rates of the seedlings which were grown as bareroot (i.e. propagated in raised beds) vs. containerized in the nursery. The results of the study showed that both aboveground and Keywords: root growth rate, Valonia oak, first order laterals, container, bareroot. ecologia mediterranea – Vol. 41 (2) – 2015 belowground morphological characteristics and growth rates of container Quercus ithaburensis subsp. macrolepis seedlings outperformed those of bareroot ones. Shoot height, root collar diameter, shoot and root growth rates, number of first order lateral roots and the tap root of the container seedlings, showed significantly higher values than those of the bareroot seedlings. The high number of first order roots and the long central root of containerized seedlings would facilitate seedlings’ uptake of soil water and nutrients and confirm that they have a better chance of survival in degraded areas. However, the results obtained are not enough to discourage the production of Quercus macrolepis bareroot seedlings. The influence of root pruning, seedbed density and fertilization on the bareroot seedlings quality, deserve future investigation. Mots clés : taux de croissance des racines, chêne Valonia, racines latérales de premier ordre, contenant, pot, semis en plein sol, semis en parterre ou plate-bandes. 25 Thekla Tsitsoni, Marianthi Tsakaldimi, Maria Gousiopoulou Résumé Quercus ithaburensis, sous-espèce macrolepis est une espèce d’importance écologique majeure. Elle est à la base des quelques forêts de chênes à feuilles caduques de l’est du bassin méditerranéen ; elle est bien adaptée aux conditions chaudes et sèches de la Méditerranée, développe un système de racines profondes, se régénère facilement après un incendie et est utilisée dans certains projets de restauration écologique. Pourtant, peu d’informations existent sur ses besoins écologiques, et encore moins sur tout ce qui concerne sa propagation et méthodes de production en pépinières. Afin de produire des semis de très bonne qualité, pouvant être utilisés pour des projets de reforestation ou de restauration réussis, les objectifs de cette étude étaient de a) enquêter sur l’architecture des parties aériennes et racinaires de Quercus ithaburensis sous-espèce macrolepis et b) comparer les vitesses de croissance des semis ayant crû plein sol vs. en pot durant l’étape de pépinière. Les résultats de l’étude ont montré que les caractéristiques morphologiques aériennes et racinaires ainsi que le taux de croissance des semis ayant grandi en pot ont surpassé de loin les performances de ceux de plein sol. Plus spécifiquement, les conclusions de cette recherche sont les suivantes : –– la hauteur des parties aériennes, le diamètre des collets ainsi que le taux de croissance des parties aériennes des semis en pot ont montré des valeurs supérieures à celles des semis en plein sol, à toutes les dates de prélèvement des données ; –– le taux de croissance des parties aériennes (Shoot Growth Rate – SRG) de tous les semis étudiés a diminué tout au long de la période d’étude, mais celui des semis en pot est resté supérieur à celui des semis en plein sol, à toutes les dates de prélèvement ; –– les semis en plein sol ont montré un taux supérieur de croissance des racines (Root Growth Rate – RGR) au début, mais par la suite le RGR des semis en pot a dépassé ce premier ; –– le grand nombre de racines de premier ordre et la longue racine centrale des semis en pot assurent ses apports en eau et nutriments provenant du sol et confirment qu’ils ont des meilleures chances de survie dans des zones dégradées ; –– afin d’éviter que des racines ne s’enroulent pendant la production des semis dans des pots cylindriques, l’utilisation de pots plus profonds est conseillée ; –– cependant, plus de recherches sur l’influence 1) des dommages faits aux racines lors de la transplantation, 2) de la densité du lit de semences et 3) des apports de nutriments sur les semis en plein sol sont nécessaires. Introduction Q. ithaburensis subsp. macrolepis grows mainly in the central and eastern part of Mediterranean basin and particularly in SE Italy, S. Albania, Greece, Turkey, Israel, Palestine, Jordan, Syria and Lebanon (Figure 1). Large forests with old Q. ithaburensis subsp. macrolepis trees existed during the past in the Mediterranean zone of Greece, in Peloponnesus, Attica, and the Aegean islands (Grispos 1973). During the last decades, activities such as conversion of forests to agricultural land, illegal lumbering and overgrazing have confined Q. ithaburensis subsp. macrolepis to small-forested patches or to isolated Figure 1 – Geographical distribution of Quercus ithaburensis subsp. macrolepis. 26 ecologia mediterranea – Vol. 41 (2) – 2015 Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; Key Factors for Successful Restoration of Mediterranean Ecosystems individuals in the interior of forested islets in lowland and semi-mountainous agricultural fields (Pantera et al. 2008; Pantera & Papanastasis 2012). Compared to the past, the species currently has an ecological rather than economic importance; it forms some of the few deciduous oak forests in the eastern Mediterranean zone (eu-mediteranean, Quercetalia ilicis zone) and certain of these forests are included in the “Natura 2000” network. It is well adapted to Mediterranean conditions and warm dry periods, develops a deep root system, and regenerates easily after fire (Pantera & Papanastasis 2012). In the past few years, there has been a growing interest for the species to be included in reforestation as well as in restoration projects (Zaady and Perevolotsky 1995; Tsakaldimi et al. 2000; Tsitsoni et al. 2011). However, information on the nursery production methods of Quercus macrolepis seedlings and the growth and architectural development of these produced seedlings is limited. Actually, no relevant research on seedling production and quality of the subsp. macrolepis seedlings was conducted. A detailed knowledge of how the production methods affects seedlings quality of the species could contribute to better seedlings’ outplanting performance. The two important factors that limit the establishment and growth of woody seedlings in Mediterranean environments, particularly in abandoned cropland and in deforested areas, are excessive irradiation and reduced water availability (Rey Benayas & Camacho-Cruz 2004; Valdecantos et al. 2006). Moreover, predation is also a limiting factor in case of direct oak seeding in the field, while herbivory has been showed to strongly impact Quercus saplings in Mediterranean forests. Most seedling mortality occurs in the first dry season of their life cycle and it has been attributed, in addition to other factors, to poor stock quality (Gazal et al. 2004). In Mediterranean Greece, the harsh site conditions in relation to the fact that degraded Mediterranean areas occupy extensive areas in Greece (Ganatsas et al. 2012) increase the need for the production of high planting stock quality, able to cope with the environmental conditions and ensure a successful reforestation or restoration. The poor survival of oak plantings has been linked to factors, such as slow growth, rapid growth of competing vegetation, poor planting methods, and poor seedling quality (McGee & Loftis 1986; Pope 1993; Tsakaldimi et al. 2005; Pantera & Papanastasis 2012). ecologia mediterranea – Vol. 41 (2) – 2015 Oak plantings may be established by direct seeding in the field, but the success of such operations has been inconsistent for many species. For that reason, the vast majority of oak plantings are established using seedlings grown in nurseries (Allen et al. 2001; Jacobs 2003; González-Rodríguez et al. 2011). The establishment of transplanted seedlings of woody species, especially oaks, in degraded Mediterranean environments may be aided by nursery treatments that promote the development of a deep and well-structured root system (Green et al. 2005; Tsakaldimi et al. 2009). Several authors argue that root morphology and growth may provide an effective indicator of seedling performance (Davis & Jacobs 2005; Tsakaldimi et al. 2005) despite the fact that these measurements are destructive, laborious and time consuming. The elongation of the taproot and new root growth during the first season after field planting is extremely important for seedling survival under the Mediterranean climatic conditions. Moreover, root regeneration is of critical importance to the establishment of planted seedlings. New root growth enables the seedling to establish a functional connection with the soil and thereby overcome the water stress imposed by transplanting (Burdett 1990; Krasowski 2003; Grossnickle 2005). Among the commonly assessed root system attributes of hardwood seedlings are the number of primary first order lateral roots (FOLRs) and root system fibrosity (Davis & Jacobs 2005). These parameters are broadly indicative of the structural framework (i.e. mainly involved in support and transport functions) and the fine root component (i.e. mainly involved in water and mineral nutrient uptake) of a seedling root system, respectively. Several authors argue that the FOLRs could be one of the best indicators of seedlings competitive capacity in the field (Schultz & Thompson 1997; Teclaw & Isebrands 1993; Kormanik et al. 1994). A large number of FOLRs is linked to rapid early establishment, improved growth rates and survival of oak seedlings (Ruehle & Kormanik 1986; Schultz & Thompson 1997). The two main types of nursery operations for oak seedling production are bare-root and containerized. Although many technological advances have been made over the last twenty years in production of both bareroot and containerized planting stock to improve planting stock quality, oak containerized seedlings dominate the forest nurseries in 27 Thekla Tsitsoni, Marianthi Tsakaldimi, Maria Gousiopoulou Greece, possibly due to shorter production cycles, stock uniformity and better field performance on harsh planting sites (Tsakaldimi et al. 2005). However, few studies have compared shoot and root growth and architecture of oak seedlings produced by these two nursery treatments. Wilson et al. (2007) reported that container Quercus rubra seedlings had a larger number of FOLRs and greater relative growth rate (RGR) than bareroot seedlings. The objectives of this study therefore were a) to compare shoot and root architecture and growth rates in time of Quercus ithaburensis subsp. macrolepis seedlings which were grown as bareroot or containerized in the nursery phase, b) to ascertain through which production technique we have better seedlings’ quality. This knowledge can contribute to a better understanding of what is the indicated species production technique in order to produce high quality seedlings capable of ensuring a successful reforestation or restoration. Materials and Methods Study Site Acorns of Quercus ithaburensis subsp. macrolepis were collected in January 2012 from the Forest Botanic Garden (6.2 hectares) of Aristotle University of Thessaloniki, Northern Greece (Latitude 40.566266, Longitude 22.969896, 32.3 m.a.s.l). According to data from the meteorological station of Aristotle University of Thessaloniki, the climate of the region is Mediterranean with dry hot summers and mild winters. The average annual air temperature is 15.8 °C with minimum average monthly temperature 5.9 °C (January) and maximum 25.9 °C (July). The average annual precipitation is 449.3 mm and relative humidity is 66.7% (Samara & Tsitsoni 2014) Experimental design Acorns were transported to the laboratory for measurements and experimental trials. Prior to experiments, acorns were visually checked and any acorns that appeared abnormal or obviously defective were discarded (Ganatsas & Tsakaldimi 2013). Then acorns were immersed in water for 24 hours to eliminate dead acorns. 28 120 viable acorns were linearly sown in seedbeds (sowing distance 15 × 15 cm) and as many were sown in plastic bags (25 cm depth and 1.5 L volume) filled with peat-perlite (1:1) for the production of bareroot and container seedlings respectively. Soil bulk density in seedbeds is high enough (1.58 g/cm3 at 0-20 cm depth) and it increases with the increase of soil depth (1.98 g/cm3 at 20-40 cm). Soil pH is neutral (pH 6.77-6.88) and its organic matter ranges from 2.82 (0-20 cm) to 1.74 (2040 cm). All acorns were sown in mid March 2012 in an open nursery and were irrigated with an overhead irrigation system. No shading was used and no nutrients were added. Three and a half months after sowing, five seedlings per treatment were randomly extracted for destructive sampling and the same sample size was extracted every fifteen days for a period of two months (five samples of five seedlings each per treatment, totally). After each destructive sampling, samples were transferred to the laboratory for the following measurements: length of the central root, number of first order lateral roots (FOLRs), length of the longest first order lateral, root collar diameter (RCD), stem height, number of leaves. For root measurements, each sampled root system was separated from the soil. More specifically, before the excavation of bareroot seedlings, sufficient water was poured around the seedlings to loose and wet the soil. In the laboratory, roots from container and bareroot seedlings were repeatedly submerged in water. A sieve was used to collect any root fragments detached from the root system (Tsakaldimi et al. 2013). Height and length measurements and root collar diameters were taken using a ruler and a vernier calliper, respectively. The length of the central root (taproot) and the length of the longest first order lateral were measured on a millimeter paper for more precise measurements. Root growth rate (RGR) estimation was based on the growth of the longest first order lateral root. Statistical analysis Distribution was tested for normality by Kolmogorov-Smirnov criterion. Because the homogeneity of variances was not assumed the non parametric Mann-Whitney test was used, comparing two groups (bareroot vs. container seedlings) in each analysis. Variables ecologia mediterranea – Vol. 41 (2) – 2015 Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; Key Factors for Successful Restoration of Mediterranean Ecosystems Table 1 – Above and below ground growth characteristics of containerized vs. bareroot seedlings of Quercus ithaburensis subsp. macrolepis in different sampling dates (sowing date mid-March). For each sampling date, the means followed by asterisk (*) are significantly different (P < 0.01). Given values are mean (std error of the mean). Sampling dates Seedlings type Shoot height (cm) Leaf number Root collar diameter (mm) Length of the longest FOLR (cm) FOLRs number Tap root length (cm) Bareroot 19.90 (1.68) 19.40 (2.65) 4.60 (0.24) 34.02 (3.18) 7.40 (1.63) 20.02 (5.65) Container 22.40 (2.11) 19.60 (2.42) 4.80 (0.48) 55.74 (5.96) 7.20 (1.57) 23.82 (5.32) Bareroot 21.12 (1.53) 19.80 (2.51) 5.04 (0.42) 46.22* (2.82) 10.00 (0.77) 21.16 (3.96) Container 23.78 (0.59) 21.40 (3.52) 5.80 (0.20) 63.94* (1.94) 12.00 (0.83) 24.38 (3.35) Bareroot 21.50 (1.50) 20.60 (2.50) 5.62 (0.29) 47.46* (2.84) 16.40 (2.73) 22.44 (2.46) Container 24.46 (0.99) 22.20 (3.81) 6.36 (0.49) 66.58* (2.62) 18.00 (2.48) 26.74 (3.36) Bareroot 21.87 (1.48) 21.40 (2.50) 5.86 (0.13) 50.16* (2.54) 18.40* (1.39) 23.36* (0.50) Container 24.98 (1.18) 22.60 (3.73) 6.94 (0.68) 79.50* (6.01) 36.40* (5.23) 35.16* (7.97) 20/06/2012 05/07/2012 20/07/2012 04/08/2012 analyzed were: shoot height, shoot growth rate, leaf number, root collar diameter, length of the longest FOLR, number of FOLRs, taproot length and root growth rate. Statistical analysis was performed with SPSS (SPSS Inc., version 21). Results and Discussion During five month growth in the nursery, the containerized seedlings, raised in 1.5 L container, exhibited significantly greater growth at all sampling dates than the bareroot seedlings. At the final extraction date, the five-month old containerized seedlings were 3.11 cm taller and 1.08 mm greater in root collar diameter (Table 1). The aerial development of container seedlings was probably favoured by both the large volume and the favourable properties of the artificial growing medium peat:perlite 1:1 (a very porous, well-drained, with high exchange capacity and friable medium) compared to those of the nursery field soil. Williams and Strupe (2002) found that bareroot water oak (Quercus nigra) and willow oak (Quercus phellos) seedlings grew taller and had larger RCD than the container seedlings, but the latter were raised in plastic cone containers of very limited volume (vol. 164 cm3). However, much more important was the statistically significant increase in root parameters ecologia mediterranea – Vol. 41 (2) – 2015 of the container seedlings; after five months they presented 29.34 cm longer longest FOLR, 18 more FOLRs and 11.8 cm longer central root than bareroot ones (Table 1). Similar to our results, Wilson et al. (2007) found that container seedlings of red oak (Q. rubra L.) had higher number of FOLRs and were significantly more fibrous than bareroot seedlings throughout the first growing season. They also reported that container seedlings acclimated faster in the field, creating a better relationship between soil and main root than bareroot seedlings due to the high difference in the length of the tap root. The rapid growth of the root system improves the uptake and transport of water which in turn favour the survival and growth of plants (Tsakaldimi et al. 2005). Under dry conditions, trees with large root system can survive better than seedlings with small root system (Hobbs 1984). Johnson (1979) noted that an increased number of first order laterals (FOLRs) in container seedlings compared to bareroots contributed to the increase in total length of roots ten weeks after outplanting. This facilitated water absorption when environmental conditions were suitable for the growth and minimized water stress. Many researchers have noted that a large number of FOLRs and a fibrous root system with large root surface area are important features of high-quality seedling planting (Ruehle & 29 Thekla Tsitsoni, Marianthi Tsakaldimi, Maria Gousiopoulou Figure 2 – S hoot growth rate of bareroot (♦) and container seedlings (○). In each sampling date, statistical significant differences were detected between the bareroot and container seedlings (P < 0.05). The error bars are the std error of the mean. Figure 3 – L ongest FOLR growth rate of bareroot (♦) and container seedlings (○). In each sampling date, statistical significant differences were detected between the bareroot and container seedlings (P < 0.05). The error bars are the std error of the mean. Kormanik 1986; Thompson & Schultz 1995; Tsakaldimi et al. 2005). Concerning the length of the central root in container seedlings, after four months growth, it was found longer (26.74 cm) than the container depth because roots were twisting at the bottom of the container. Generally the roots of oaks tend to exhibit twist root around the interior or the base of the containers when these are cylindrical (Landis et al. 1990). Seedlings of most oak species preferentially invest into producing roots, which initially have a taproot or carrot-like configuration. Such rapid growth means that seedlings’ root systems can quickly exceed the depth of the container 30 and become pot-bound. In general, for better quality of oak seedlings narrow, deep bottom opened containers are recommended. The root deformities such as the twisted roots can persist for decades after planting and create problems of weakness, poor growth and lack of stability years later (McCreary & Canellas 2005). Similarly, Wilson et al. (2007), despite the good performance of red oak container seedlings, observed a great number of twisted roots and they suggested the use larger pots than those commonly used. The shoot growth rate (SGR) of all studied seedlings decreased throughout the study period but was significantly higher in container seedlings than that of bareroot, in all sampling dates (Figure 2). The first 3.5 months after sowing, the bareroot seedlings showed significantly higher root growth rate (RGR) than the containerized seedlings (0.81 cm/day and 0.55 cm/day respectively), however, then the RGR was significantly increased in container-seedlings (P < 0.05). After 4.5 months, the RGR of container seedlings was found 0.86 cm/day while the RGR of bareroots remained close to 0.2 cm/ day (Figure 3). The RGR change of bareroots could be affected by the seedlings’ density in seedbeds. Possibly the competition for space and moisture in seedbeds was increased between bareroot seedlings by that time and had a negative influence in seedlings RGR. On the contrary, such competition was absent in container seedlings where seedlings grew in an autonomous growth space. Similarly, container fir seedlings presented higher RGR than bareroot ones (Scagel et al. 1993). Also, a successful seedling establishment is dependent on the capacity of seedlings to initiate new roots quickly (Grossnickle 2005). Seedlings with higher RGR are expected to survive and perform better, than those with lower RGR because they are more prone to damage under adverse site conditions. The results obtained showed that container seedlings were enhanced by the favourable growth conditions; their better performance could be attributed to the moist and friable artificial medium of sufficient volume that they were grown in relation to the nursery field soil. The artificial medium peat:perlite (1:1) has better structure and aeration than the nursery field soil so it was easier for the roots of container seedlings to absorb the necessary nutrients and water for their growth. Pantera et al. (2010) found that acorn germination of Quercus ithaburensis subsp. macrolepis ecologia mediterranea – Vol. 41 (2) – 2015 Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; Key Factors for Successful Restoration of Mediterranean Ecosystems is not affected by soil compaction but under high soil compaction the seedlings’ growth was negatively affected. Recent studies show that field performance of bareroot seedlings can be improved by increasing the amount of organic matter in beds or decreasing nursery bed seedling densities to produce larger seedlings. Salifu et al. (2009) reported that nutrient loading of oak seedlings in the nursery has the potential to increase seedling performance especially on harsh or degraded sites. Moreover, the undercutting of bareroots during the growing season can cause an increased production of new roots, especially FOLRs (Jacobs 2003, Dey et al. 2004). However, it still remains unknown whether the undercutting of taprooted species may favour or not their outplanting survival and growth in dry Mediterranean soils. From all the above, it seems that both aboveground and belowground morphological characteristics and growth rates of container Quercus ithaburensis subsp. macrolepis seedlings outperformed those of bareroot ones. However, it is unclear if the superior growth performance of container seedlings will persist beyond the establishment phase. Andersen et al. (1989) found that container seedlings had higher survival than bareroot red oak seedlings during the first four years after planting in the field. However, seven years after planting, survival did not greatly differ between the two types of seedlings. Further work is needed to investigate both the influence of pot design in root twisting of container seedlings and the influence of root pruning, seedbed density and nutrients in bareroot seedling growth. Enhanced seedling quality could help young trees of Q. ithaburensis subsp. macrolepis, establish themselves more rapidly, thereby giving them a better chance for successful growth in the field. References Allen J.A., Keeland B.D., Stanturf J.A., Clewell A.F. & Kennedy H.E., 2001. A guide to bottomland hardwood restoration. Gen. Tech. Rep. SRS-40, USDA Forest Service, pp.132. Andersen C.P., Bussler B.H., Chaney W.R., Pope P.E. & Byrnes W.R., 1989. Concurrent establishment of ground cover and hardwood trees on reclaimed mined land and unmined reference sites. For. Ecol. Manage. 28: 81-99. Burdett A.N., 1990. Physiological processes in plantation establishment and the development of ecologia mediterranea – Vol. 41 (2) – 2015 specifications for forest planting stock. Can. J. For. Res. 20: 415-427. Davis A.S. & Jacobs D.F., 2005. Quantifying root system quality of nursery seedlings and relationship to outplanting performance. New For. 30: 295-311. Dey D.C., Lovelace W., Kabrick J.M. & GoldIn M.A., 2004. In: Michler, C.H.; Pijut, P.M.; Van Sambeek, J.W.; Coggeshall, M.V.; Seifert, J.; Woeste, K.; Overton, R.; Ponder, F., Jr., (eds.), Proceedings of the 6th Walnut Council research symposium: Black walnut in a new century. Gen. Tech. Rep. NC-243, USDA, Forest Service, North Central Research Station, St. Paul: 188 p. Ganatsas P., Tsitsoni T., Tsakaldimi M. & Zagas T., 2012. Reforestation of degraded Kermes oak shrublands with planted pines: effects on vegetation cover, species diversity and community structure. New For. 43: 1-11. Ganatsas P. & Tsakaldimi M., 2013. A comparative study of desiccation responses of seeds of three drought-resistant Mediterranean oaks. For. Ecol. Manage. 305: 189-194. Gazal R.M. & Kubiske M.E., 2004. Influence of initial root characteristics on phsiological responses of cherrybark oak and Shumard oak seedlings to field drought conditions. For. Ecol. Manage. 189: 295-305. González-Rodríguez V., Navarro-Cerrillo R.M. & Villar R., 2011. Artificial regeneration with Quercus ilex L. and Quercus suber L. by direct seeding and planting in southern Spain. Annals of Forest Science 68(3): 637-646. Green J.J., Baddeley J.A., Cortina J. & Watson C.A., 2005. Root development in the Mediterranean shrub Pistacia lentiscus as affected by nursery treatments, J. Arid Environ. 61: 1-12. Grispos P., 1973. Greek forest history. Publication 25 of the Forest Applications Service, Athens, Ministry of Nation. Econ., Agricultural Sector, Forest Division, 385 p. (In Greek). Grossnickle S.C., 2005. The importance of root growth in overcoming planting stress. New For. 30: 273294. Hobbs S.D., 1984. The influence of species and stocktype selection on stand establishment: An ecophysiological perspective. In: Duryea M.L. & W. Brown W. (eds.), Seedling physiology and rReforestation success. NijhofB/Dr. W. Junk Pub. Boston: 179-224. Jacobs D.F., 2003. Nursery production of hardwood seedlings. FNR 212, Purdue University, Cooperative extension Service, West Lafayette, USA, 8 p. Johnson P.S., 1979. Growth potential and field performance of planted oaks. In: Holt H.A. & Fisher B.C. (eds), Proceedings: Regenerating oaks in upland hardwood forests. The 1979 John S. Wright Forestry Conference. Purdue University, Indianna: 113-119. Kormanik P.P., Sung S.S. & Kormanik T.L., 1994. Toward a single nursery protocol for oak seedlings. In: Proceedings of 22nd southern forest tree improvement conference, Atlanta, GA 14-17 Jun 1993: 89-98. Krasowski M.J., 2003. Root system modifications by nursery culture reflect on post-planting growth and development of coniferous seedlings. For. Chron. 79(5): 882-891. Landis T.D., Tinus R.W., McDonald S.E. & Barnett J.P., 1990. The container tree nursery manual, Vol. 2. Agriculture Handbook 674. USDA Forest Service, Washington DC: 88 p. 31 Thekla Tsitsoni, Marianthi Tsakaldimi, Maria Gousiopoulou McCreary D. & Canellas I., 2005. Restoration of oak woodlands in Mediterranean ecosystems. In: Stanturf J.A. & Madsen P. (eds), Restoration of boreal and temperate forests, CRC Press: 294-295. McGee C.E. & Loftis D.L., 1986. Planted oaks perform poorly in North Carolina and Tennessee. North. J.Appl. For. 3: 114-115. Pantera A., Papadopoulos A., Fotiadis G. & Papanastasis V., 2008. Distribution and phytogeographical analysis of Quercus ithaburensis ssp. macrolepis in Greece. Ecologia Mediterranea 34: 73-82. Pantera A., Genitsaropoulos C., Fotiadis G. & Mertzanis A., 2010. Soil compaction effects on the germination and growth of valonia oak seedlings (Quercus ithaburensis ssp. macrolepis). In: Proceedings of the 7th panhellenic rangeland congress, Xanthi, 14-16 October 2010: 129-134. Pantera A. & Papanastasis V.P., 2012. Competitive effects of herbaceous species on water potential and growth of Quercus ithaburensis ssp. macrolepis seedlings. Global Nest J. 14(4): 525-531. Pope P.E., 1993. A historical perspective of planting and seeding oaks: progress, problems, and status. In: Loftis D. & McGee C.E., (eds.), Oak regeneration: serious problems, practical recommendations (symposium proceedings). Gen. Tech. Rep. SE-84. USDA, Forest Service, Southeastern Forest Experiment Station, Asheville: 224-240. Rey Benayas J.M. & Camacho-Cruz A., 2004. Performance of Quercus ilex saplings planted in abandoned Mediterranean cropland after long-term interruption of their management. For. Ecol. Manage. 194: 223-233. Ruehle J.L. & Kormanik P.P., 1986. Lateral root morphology: a potential indicator of seedling quality in northern red oak. Research Note SE-344, USDA Forest Service, Southeastern Forest Experiment Station, New Orleans: 1-5. Salifu K.F., Jacobs D.F., & Birge Z.K.D., 2009. Nursery nitrogen loading improves field performance of bareroot oak seedlings planted on abandoned mine lands. Restor. Ecol. 17: 339-349. Samara T. & Tsitsoni T., 2014. Selection of forest species for use in urban environment in relation to their potential capture to heavy metals. Global NEST Journal. 16(5): 966-974. Scagel R.K., Binder W.D., Krumlik G.J., 1993. Sitespecific relationships of root growth potential to outplanting performance of coastal Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) under controlled irrigated field conditions. Research Report RR 93002-HQ, B.C. Ministry of Forests, Research Branch, Victoria: 12 p. Schultz R.C. & Thompson J.R., 1997. Effect of density control and undercutting on root morphology of 32 1 + 0 bareroot hardwood seedlings: five-year field performance of rootgraded stock in the central USA. New For. 13: 301-314. Teclaw R.M. & Isebrands J.G., 1993. An artificial regeneration system for establishing northern red oak on dry-mesic sites in the Lake States, USA. Ann. Sci. Forest. 50: 543-552. Thompson J.R. & Schultz R.C., 1995. Root system morphology of Quercus rubra L. planting stock and 3-year field performance in Iowa. New For. 9: 225-236. Tsakaldimi M., Ganatsas P., Gounaris N., Zagas T., Tsitsoni T. & Chatzistathis A., 2000. Water stress effect on the quality of one-year old container seedlings. In: Proceedings of the 9th Hellenic forestry association conference, Kozani 17-20 October 2000: 170-179. Tsakaldimi M., Zagas T., Tsitsoni T. & Ganatsas P., 2005. Root morphology, stem growth and field performance of seedlings of two Mediterranean evergreen oak species raised in different container types. Plant Soil 278: 85-93. Tsakaldimi M., Tsitsoni T., Ganatsas P. & Zagas T., 2009. A comparison of root architecture and shoot morphology between natural regenerated and container seedlings of Quercus ilex L. Plant Soil 324: 103-113. Tsakaldimi M., Ganatsas P. & Jacobs D.F., 2013. Prediction of planted seedling survival of five Mediterranean species based on initial seedling morphology. New For. 44: 327-339. Tsitsoni K.T., Tsakaldimi M. & Prodofikas Ch., 2011. Root growth potential and seedling morphological attributes of four Mediterranean hardwood species. In: Book of abstracts of the 12th international Mediterranean ecosystems conference. MEDECOS XII. “Linking Science with Resource Management”, Los Angeles, 6-9 September: 220. Valdecantos A., Cortina J. & Vallejo R., 2006. Nutrient status and field performance of tree seedlings planted in Mediterranean degraded areas. Ann. Forest Sci. 63: 249-256. Williams H. & Stroupe M., 2002. First-year survival and growth of bareroot and container water oak and willow oak seedlings grown at different levels of mineral nutrition. Faculty Publications. Paper 203. http://scholarworks.sfasu.edu/forestry/203. Wilson E.R., Vitols K.C. & Park A., 2007. Root characteristics and growth potential of container and bare-root seedlings of red oak (Quercus rubra L.) in Ontario, Canada. New For. 34: 163-176. Zaady E. & Perevolotsky A., 1995. Enhancement of growth and establishment of oak seedlings (Quercus ithaburensis Decaisne) by inoculation with Azospirillum brasilense. For. Ecol. Manage.72: 81-83. ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results Création d’un modèle intégré pour la gouvernance des écosystèmes méditerranéens urbains et pour adapter les villes aux changements climatiques : premiers résultats Thekla TSITSONI1*, Nikolaos GOUNARIS2, Aimilia B. KONTOGIANNI1, Valia XANTHOPOULOU-TSITSONI3 1. School of Forestry & Natural Environment, Laboratory of Silviculture, Aristotle University of Thessaloniki, University Campus, P.O. Box: 262, 54124 Thessaloniki, Greece. 2. CEO of Homeotech Co Company, Grigoriou Lampraki 210, 54352 Thessaloniki, Greece. 3. School of Economics, Department of Development and Planning, Aristotle University of Thessaloniki, University Campus, P. O. Box: 178, 541 24 Thessaloniki, Greece * Corresponding author: [email protected] Received: 20 January, 2015; First decision: 30 April, 2015; Revised: 4 December, 2015; Second decision: 14 December, 2015 Abstract Urban trees grow under adverse conditions, governed by the combinatorial effect of multiple natural and anthropogenic factors. Global climate change adds new challenges regarding urban green management that should be taken into account when designing future green urban policies. The development of a smart Information and Communications Technology (ICT) system and the establishment of a continuously up to date information system regarding urban trees is a key issue for future management which aims at: –– climate change adaptation, by providing an instrument for measuring the cities CO2 emission offsets by CO2 sequestration of the tree biomasses; –– efficient utilization of resources that are spent for urban tree management in order to decrease the cities’ environmental footprint; –– enhancement of the cities’ social fabric by promoting citizen participation during the decision making process regarding urban trees. The project was selected for funding after a competitive process by the General Secretary of Research and Development. The Municipality of Thessaloniki was selected as the key study area. The core of the project is the development of a software suite named GreenTree. Through the GreenTree client Android application, 105 different sets of data are collected for each urban tree. The urban tree inventory includes 37,328 tree sites on the pavements, from which we found 1,239 dead trees, 2,787 empty sites and 937 trees which had to be removed because they had been planted in inappropriate location and they disturbed the circulation of cars or pedestrians. The numbers above clearly show the significant need for the establishment of a reliable and smart monitoring system for urban tree management. This system could also help manage the decision making process. Finally, the numbers show that urban environment can be easily improved by applying fast and cheap measures of tree replanting and replacement. Keywords: urban forestry, green spaces, urban tree management, street trees, GreenTree Software. ecologia mediterranea – Vol. 41 (2) – 2015 33 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni Résumé Les arbres urbains se développent dans des conditions difficiles, régies par l’effet combiné de multiples facteurs naturels et anthropiques. Le changement climatique mondial ajoute de nouveaux défis en matière de gestion urbaine “verte” qui devrait être prise en compte lors de l’élaboration des futures politiques urbaines écologiques. Le développement d’un système intelligent en Technologie de l’Information et de la Communications (TIC) et la mise en place d’un système d’information, mis à jour en continu, concernant les arbres urbains est un enjeu clé pour la gestion future qui vise à : –– l’adaptation au changement climatique, en fournissant un instrument pour mesurer les émissions de CO2 que les villes compensent par la séquestration du CO2 dans la biomasse des arbres ; –– l’utilisation efficace des ressources qui sont dépensées pour la gestion des arbres en milieu urbain afin de réduire l’empreinte environnementale des villes ; –– le renforcement des liens sociaux dans les villes en favorisant la participation des citoyens pendant le processus de prise de décisions concernant les arbres urbains. Le projet a été sélectionné pour financement après un processus compétitif organisé par le Secrétaire Général de la Recherche et du Développement. La municipalité de Thessalonique a été choisie comme zone d’étude principale. Le cœur du projet est le développement d’une suite de logiciels nommée GreenTree. Via l’application android GreenTree, 105 ensembles distincts de données ont été collectés pour chaque arbre urbain. L’inventaire des arbres en milieu urbain comprend 37 328 sites de plantation d’arbres sur les trottoirs, parmi lesquels ont été trouvés 1 239 arbres morts, 2 787 absents de leur site (trous vides sur les trottoirs) et 937 arbres devant être retirés. Les chiffres ci-dessus montrent clairement la nécessité de la mise en place d’un système de suivi fiable et intelligent pour la gestion des arbres urbains. Ce système pourrait aussi faciliter la gestion du processus de prise de décisions. Enfin, les chiffres montrent que l’environnement urbain peut être facilement amélioré par l’application de mesures rapides et bon marché pour la replantation et le remplacement des arbres. Introduction The world is undergoing the largest wave of urban growth in history, as it is calculated that above half of the world population lives in urban regions while by 2030 the urban population is expected to be twice as large as the Mots clés : foresterie urbaine, espaces verts, gestion des arbres en milieu urbain, arbres de rue. 34 corresponding rural (Samara & Tsitsoni 2010, Lang 1999, Sandberg 1999, OECD-ECMT 1995, Lambert & Vallet 1994). Increasingly, urban green space is seen as an integral part of cities, providing a range of services to both the people and the wildlife living in urban areas (James et al. 2009, Wolf 2004, Nowak & Dwyer 2000). So, for the last 50 years there has been a growing realization that the solutions to most of environmental problems reside in making cities more efficient in their consumption of energy and materials and disposing of waste products, and in altering patterns of urban development to reduce the amount of impervious “grey” infrastructure and to increase the amount of “green” infrastructure, particularly trees (Carreiro 2006). With this recognition and resulting from the simultaneous provision of different services, there is a real need to identify a research framework in which to develop multidisciplinary and interdisciplinary research on urban green space (James et al. 2009). This realization has been expressed in the concepts of the eco-cities movement, adopted by many environmentalists and urban designers throughout the world (Register 2002). A major problem that globally has to be faced is the climate change and mostly what deals with global warming. The definition of climate change refers to any significant change in the measures of climate lasting for an extended period of time (EPA 2015). Climate change is caused by factors, such as biotic processes, variations in solar or volcanic eruptions etc., but the main threat, mostly because of human activity is referred to as global warming and that is the recent and ongoing global average increase in temperature near the Earth’s surface. According to NASA, the global temperature increased by 0.77 oC since 1880 while nine of the ten warmest summers on record have occurred the last 15 years (NASA 2015). Natural and human factors cause changes in Earth’s energy balance, leading, among other things to increased greenhouse effects. Within this context, urban green spaces can play a central role in both climate-proofing cities and in reducing the impacts of cities on climate (Gill et al. 2007). Urban vegetation through its physiological features could provide potential help to the cities’ adaptation and mitigation strategies for climate change. Global climate change adds new challenges regarding urban green management that should be taken into account when designing future green urban policies. ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results Adaptation deals with preparing for inevitable climate change while mitigation is the act of limiting further climate change, for example by reducing emissions of greenhouse gases. Urban green spaces can help to alleviate the consequences of climate change mostly by cooling. Trees, especially when located close to buildings can reduce temperatures, acting as natural air conditioners through their evapotranspiration and providing shade with their foliage, therefore reducing energy consumption required to maintain comfortable climatic conditions. Even small green spaces can have a cooling effect – parks of 1 or 2 hectares only have been found to be 2 oC cooler than surrounding areas (Shashua-Bar & Hoffman 2000). The extent of the cooling effect is greatest when temperatures beyond the park are highest (Handley & Carter 2006). At the same time, urban trees and green spaces can mitigate the impacts of climate change through the absorption of the CO2 that takes part in greenhouse gases and act as carbon sinks. Additionally trees that are placed close to buildings lead to lower consumption of energy for mechanical cooling and heating, decreasing CO2 emissions. All green spaces help urban areas adapt to the impact of climatic change regardless of whether they are park, private garden or street trees, with location, structure, composition and spatial configuration of them to influence their ecological qualities and functions (Turner et al. 2005). At the same time vegetation type that is characterized by the plant species, the size of the trees and the mixture of the species and the proportion of urban tree canopy as well influence the impact level. The challenge is to find functional and as low budget as it is possible ways for adaptation and mitigation solutions based on urban greening providing the total of physiological, sociological, economic and aesthetic benefits. Identifying and describing the benefits of urban trees to a community is the first step in gaining support for an urban forestry program of tree planting, maintenance and replacement. An urban tree inventory is essential because the information about the quantity and quality of the existing vegetation and its characteristics are important for urban greening management. Nevertheless, there is a lack of urban tree inventory and monitoring protocols and standards. These protocols could detect change over time and across cities, while providing flexibility required by diverse users. ecologia mediterranea – Vol. 41 (2) – 2015 A great proportion of the urban infrastructure consist of urban streets where people walk, shop, meet and generally participate in many social and recreational activities that make urban living enjoyable (Wolf 2004). The use of trees as an element of the landscape is an important design concept that has been used throughout the world, and continues to shape the aesthetics and function of the streets (Gezer & Gül 2009). Street trees are one of the most important components of urban green space and they play an important role in city’s aesthetics as people’s first impression of a city comes from its street landscape (Jacobs 1993). Street trees are complex to study because this entails technical, aesthetical, biological and ecological knowledge (Küçük & Gül 2005). Moreover there is no homogeneous urban environment or site, as it is a conglomeration of soils, microclimates and other site conditions. Street trees should possess strong apical growth, strong branching angles, overall high aesthetic values, predictable growth rates and, in general, have a potential for a long life span (Gezer & Gül 2009), although they also sometimes grow in adverse environments. Over recent decades, a growing proportion of the commonly used species have shown increasing difficulties in coping with the conditions at urban paved sites (Sjöman et al. 2010). Overall, trees in these environments tend to be greatly exposed to heat, low air humidity, periods of critical water stress, high lime content and high soil pH, limited soil volume, pollutants and deicing salts (Pauleit 2003, Sieghardt et al. 2005) Under this perspective a project has been compiled using as a study case the metropolitan area of Thessaloniki in northern Greece. The project was selected for funding by the General Secretary of Research and Development after a competitive process. The main goal of the project is to create a standard system for the exercise of governmental and decision making in the practice of urban forestry in a holistic and integrated way. The concept of holistic lays in a comprehensive manner, examining the needs and problems of urban green infrastructures taking into consideration as many factors as possible, anthropogenic or not, each of which has different weight. This integrated system includes all processes, services and products considered necessary for the existence of the holistic approach to the exercise of urban forestry. 35 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni The policy that is followed for the goal to be achieved contains the following objectives: 1) to create the appropriate guidelines for the management of urban greening, through the proper treatment of the wooden species; 2) adaptation strategies for the inevitable climate change and mitigation of the factors that lead to climate change; 3) to provide an efficient appliance for measuring the city CO2 emission offsets by CO2 sequestration of the tree biomasses; 4) to provide an efficient utilization of resources that are spent for urban tree management in order to decrease the cities’ environmental footprint; and 5) to enhance the cities’ social fabric by promoting citizen participation during the decision making process regarding urban trees. The main tool that is used is a smart Information and Communications Technology (ITC) system that had been developed for that purpose and the establishment of a continuously up to date information system regarding urban greening. The aim of this paper is to show the preliminary results of this research and try to make an estimation of the sufficiency and the functionality of the existing urban greening of Thessaloniki. Methods Study area The area where the specific project is taking place is the whole territory of the municipality of Thessaloniki, the second-largest city of the country and the capital of central Macedonia, in northern Greece. According to the preliminary results of the 2011 census, the municipality of Thessaloniki today has a population of 322,240, while its urban area has a population of 790,824. Thessaloniki lies on the northern fringe of the Thermaic Gulf on its eastern coast and is bound by Mount Chortiatis on its southeast. Its proximity to imposing mountain ranges, hills and fault lines, especially towards its southeast have historically made the city prone to geological changes. Thessaloniki’s climate is directly affected by the sea it is situated on. The city lies in a transitional climatic zone, so its climate displays characteristics of a mosaic of microclimates. As reported by Hellenic National 36 Meteorological Service, the total character of the climate is humid subtropical climate (Cfa) that borders on a semi-arid climate (BSk) – according to Köppen climate classification. The total annual precipitation is 458.4 millimetres due to the Pindus rain shadow drying the westerly winds. However, the city has a summer precipitation between 21 to 31 millimeters – with August the driest month and November the wettest – which borders it close to a hot-summer Mediterranean climate (Csa). The mean annual temperature in Thessaloniki is 15.6 °C. During the coldest winters, temperatures can drop to -10 °C, while the minimum temperature ever recorded is -14 °C. On average, Thessaloniki experiences frost for about 30 days per year. The coldest month of the year is January. Thessaloniki’s summers are hot with rather humid nights. Maximum temperatures usually rise above 30 °C and sometimes over 40 °C. The maximum temperature ever recorded is 42 °C. On average, Thessaloniki experiences hot waves for about 30 days per year. The hottest month of the year is July. The average wind speed for June and July is 20 km/h while in winter the average wind speed is about 26 km/h (Samara & Tsitsoni 2014). Papakostas et al. (2014) represent the variation of temperature and annual average temperature during the period from 1983 to 2012 as it is shown to the Figures 1a and 1b. The same authors conclude that the total increase in the annual average temperature from the first to the third decade was 1.1 °C with an upward trend. Experimental design From August 2013 to October 2014 and during two growing seasons, every single street tree of the municipality of Thessaloniki was recorded in order to build an integrated street tree inventory. In the inventory qualitative and quantitative information was included. Hence every above-ground characteristic of each individual as well as site features were estimated or measured (Table 1). The development of a smart Information and Communications Technology (ITC) system and the establishment of a continuously up to date information system regarding urban trees is a key issue for future management. The core of the project is the development of a software suite named GreenTree. Its use makes it easy to collect data and make a tree inventory that is based on field measurements and estimations ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results for both trees and their site features. Through the GreenTree client Android application, 105 different sets of data were collected as mentioned above. Breast height diameter and tree height data were measured directly in situ using callipers and laser measurement instruments. Tree crown projection extracted using ellipse equation and four radii from the tree trunk R1 and R2 along and R3 and R4 vertically the sidewalk. The crown volume derived using standard geometry shapes equations (Troxel et al. 2013). 30 1983-1992 1993-2002 2003-2012 Ambient Temperature [°C] 25 20 15 10 5 0 Jan Fer Mar Apr Mai Jun Jyl Aug Sep Okt Nov Dec Month Figure 1a – V ariation of temperature for the period from 1983 to 2012 (Papakostas et al. 2014). 18 Ambient Temperature [°C] Average per year Fitter curve 17 16 15 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Year Figure 1b – A nnual average temperature for the period from 1983 to 2012 (Papakostas et al. 2014). Table 1 – Set of tree measurements and evaluation. Tree Species Tree Location Height Crown height Crown diameters Breast height diameter Tree health data Soil data Proposed future measurements Existence of Utilities’ elements Tree importance (i.e. historical, monumental tree etc) Maintenance data (i.e. existence of grade, irrigation etc) ecologia mediterranea – Vol. 41 (2) – 2015 37 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni Results Street tree characteristics The preliminary analysis of the project data showed that the area of total green spaces under the jurisdiction of the municipality services covers about 906,000 m2, less than 5% of the total city’s area, as analytically it is represented to the Table 2. The total number of tree sites of the pavements that were surveyed is 41,672 belonging on 76 different species. To facilitate data processing, the species with population less than 50 individuals did not participate to the calculations, thus they eliminated from this exhaustive inventory. The revised inventory includes 37,328 tree sites of the pavements. They represented 45 wooden species from 24 families – with the majority of them being forest species. 74.34% of the individuals belonged to 12 species, while 25.66% belonged to 33 species (Tables 3 and 4). From these 33 species, 25 species were not much represented (< 1%). At the family level, from the total number of Table 2 – U rban green area per category and per resident for the territory of the municipality of Thessaloniki. Urban greening Categories Area (m2) m2 / resident % of Municipality’s Total area Parks, road islets 527,475 1.64 2.85% Street trees 378,480 1.17 2.05% 905,955 2.81 4.90% University Campus 132,551 0.41 0.72% Military basis 93,321 0.29 0.50% Hospitals 28,378 0.09 0.15% Open spaces 43,193 0.13 0.23% Streams 216,593 0.67 1.17% Stadiums 63,278 0.20 0.34% Cemeteries 8,898 0.03 0.05% Churches 15,000 0.05 0.08% Total 601,212 1.87 3.25% Private green spaces (yards, gardens) 970,940 3.01 5.25% 2,478,107 7.69 13.40% Municipality’s urban green spaces Total Urban green spaces under special management status Total green area of the municipality of Thessaloniki Table 3 – S pecies occurrence proportion. Species 38 Presence % Species Presence % Species Presence % Sophora japonica 15.52 Cercis siliquastrum 1.91 Alnus glutinosa 0.41 Robinia pseudoacacia 12.78 Albizia julibrissin 1.88 Paulownia tomentosa 0.40 Acer negundo 9.62 Fraxinus sp. 1.83 Ficus carica 0.39 Ligustrum japonicum 7.62 Olea europaea 1.68 Pinus nigra 0.35 Koelreuteria paniculata 6.20 Aesculus hippocastanum 1.34 Broussonetia papyrifera 0.33 Platanus prientalis 4.72 Acer pseudoplatanus 0.91 Ginkgo biloba 0.29 Citrus x aurantium 4.31 Cupressus sp. 0.89 Pinus brutia 0.27 Celtis australis 2.81 Morus sp. 0.81 Magnolia grandiflora 0.25 Liquidambar orientalis 2.81 Catalpa bignonioides 0.77 Quercus sp. 0.24 Tilia sp. 2.73 Prunus cerasifera var. pissardii 0.70 Pittosporum tobira 0.23 Populus x canadensis 2.63 Nerium oleander 0.69 Pinus sp. 0.17 Ulmus sp. 2.59 Chamaerops humilis 0.66 Prunus domestica 0.16 Populus alba 2.22 Populus nigra subsp. thevestina 0.54 Eriobotrya japonica 0.15 Acer campestre 1.99 Laurus nobilis 0.50 Thuja plicata 0.14 Hibiscus syriacus 1.97 Ailanthus altissima 0.43 Acacia sp. 0.13 ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results Table 4 – Number of species and individuals on each family. Family Species Individuals % Species Individuals % Fabaceae 6 12,115 32.46 Cupressaceae 2 384 1.03 Aceraceae 3 4,672 12.52 Rosaceae 3 379 1.02 Oleaceae 3 4,154 11.13 Pinaceae 3 294 0.79 Sapindaceae 2 2,814 7.54 Bignoniaceae 1 286 0.77 Ulmaceae 2 2,016 5.40 Apocynaceae 1 259 0.69 Salicaceae 3 2,011 5.39 Arecaceae 1 248 0.66 Platanaceae 1 1,763 4.72 Lauraceae 1 188 0.50 Malvaceae 2 1,757 4.71 Simaroubaceae 1 162 0.43 Rutaceae 1 1,609 4.31 Betulaceae 1 154 0.41 Altingiaceae 1 1,050 2.81 Paulowniaceae 1 150 0.40 Moraceae 3 574 1.54 Gingoaceae 1 108 0.29 24, 3 of them hold 56.1% of individuals (12 species), 91% of individuals belonged to 10 families, and 11 families had a representation lower than 1%. The 45 species consisted of 5 gymnosperms and 40 angiosperms – 64.5% deciduous and 22.3% evergreen. Approximately 50% of the planted species were large and medium sized trees (Table 5). Finally, 44.5% were introduced species – mainly from Asia – while 48.9% were native to Mediterranean region, Balkan Peninsula or generally to Europe (Table 6). Table 5 – Number of species in life form categories. Life form Number of Species % Large Tree 10 22.22 Large – Medium tree 10 22.22 Medium tree 3 6.67 Medium – Small tree 8 17.78 Small tree 5 11.11 Small tree – Shrub 6 13.33 Shrub 1 2.22 Depend on species 2 4.44 Family The cumulative crown volume that came from 378,480 m2 of urban street trees equalled 2,633,959 m3. For all 45 species, population diagrams of height, breast height diameter (dbh), crown projection and crown volume created are shown on Figures 2, 3, 4 and 5. Breast height diameter (Figure 2) and crown projection area (Figure 3) followed an ascending trend, showing that tree growth in diameter was not affected by pruning and that crown projection areas were less affected than the other two silvicultural characteristics measured (i.e. tree height and crown volume). Tree height and crown volume diagrams showed an abrupt reduction in number of existing trees in the 3rd class in comparison with the 3rd of dbh class, 7.203 vs. 15.401. This result can easily be explained by the intense height reduction interventions on street trees and hence on the crown volume. It is commonly known that these interventions are not always appropriate; the actions of pruning are greatly exaggerated and inappropriate but happen because of lack of personnel expertise and low maintenance cost needs. On the other hand the greenhouse gases sequestration is directly related to total leaf surface and thus directly to the volume of the crown. Table 6 – Species origin. Origin Species Origin Species Mediterranean region 13 Europe, Asia, Africa 1 Asia 13 Australia 1 Northern America 5 Balkans 1 Europe, Asia 3 Northern hemisphere 4 Europe, Asia and North America. 1 Hybrids 2 When only the genre is known depend on species 1 ecologia mediterranea – Vol. 41 (2) – 2015 39 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni Figure 2 – Frequency in diameter classes. Figure 3 – Frequency in crown area classes. Figure 4 – Frequency in height classes. Figure 5 – Frequency in crown volume classes. Estimation of CO2 sequestration At this point an effort was made to obtain a coarse image about the potential augmentation of CO2 sequestration that the replacement of approximately 12,000 trees would provide; this concerns the dead individuals and the empty pavement tree sites. This theoretical study examined the effect that these 12,000 trees would play in greenhouse gases sequestration through a correlation of crown volume distribution with the existing distribution of dbh. A wide range of tolerance of ± 20% was used to cover the great variety of conditions prevailing in city trees. By using complex My SQL queries in database of 36,089 living trees it was found the cases where crown volume distribution of each species deviates from the dbh diameter distribution more than ± 20%. As a result 22 of the 40 tree species showed discrepancies greater than ± 20%. 40 A theoretical model was developed where a virtual transfer from volume categories that show surplus, to them of deficit crown volume categories was performed following species dbh distribution. If there were no surplus categories, the transfer occurred proportionally from the other categories in volume deficit, provided that this should not be exceeded ± 20%. As a result a theoretical increased in crown volume equal of 375,388 m3 or 14.56% of the current crown volume of those 40 species derived (Figure 6). Even if this result is not a spectacular increase in the total urban trees crown volume, it gives an idea of what is achievable by applying low cost managerial measures. For this analysis, My SQL queries and PHP programming language are used. ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results Table 7 – General view of street tree health condition. Excellent Condition Good Condition Moderate Condition Bad Condition Worst Condition Dead Empty pavement tree site 5,554 (13.32%) 13,068 (31.35%) 9,537 (22.88%) 6,236 (14.96%) 3,257 (7.82%) 1,239 (2.98%) 2,785 (6.68%) Table 8 – Preliminary management recommendations. Additionally, tree health was to be excellent to moderate for 68% of the trees, 10% of the tree needed to be replaced immediately, while there are 2,787 empty pits (Table 7). Finally, in situ management recommendations were proposed (Table 8). Conservation Maintenance Removal Replacement 9,849 (23.63) 18,974 (45.53) 937 (2.25) 11,912 (28.60) Discussion Every community tree planting program coordinates several processes in order successfully manage public trees. To support a vigorous population of trees, these programs plan and design planting areas, select species for individual sites, coordinate planting activity, perform regular maintenance and pest management, and remove hazardous trees in a timely manner (Burcham 2009). Figure 6 – Frequency in potential crown volume classes. GreenTree specifications The specific ongoing project involves the collection of reliable and accurate information for each tree managed by the municipality of Thessaloniki, the organization of this information in a functional geospatial database, which will be updated and will be editable for completion of new data fields, if necessary. This database will be analyzed to derive statistics and indicators to 1) identify problematic individuals requiring urgent intervention, for example pruning, removal, etc., 2) assess the health situation and the goal achievement level in a management period and to plan maintenance works, supplementing or replacing individuals and calculating the corresponding cost, 3) compare work cost and quality that has to be done by private contractors, and 4) examine alternative scenarios for the distribution of the annual municipal resources by prioritizing the importance of the planned projects according to their contribution to achieving the objectives of the strategic management plan. The project will also take place actions to inform the public and publicize the results of the project. ecologia mediterranea – Vol. 41 (2) – 2015 In order to implement all the above acts, GreenTree has been developed, as a special software to be a basic tool in urban greening management, scheduled to enter into pilot operation at the end of the project. The purpose of this tool is to help municipality services to manage their urban greening specifically in relation to the potential CO2 sequestration and the regulation of temperatures. As referred above, each entry in GreenTree contains information for each individual like: tree record number, species code and name, health status and management needs, dbh, height, crown measurements and site features, etc. The data provided by the inventory will lead to the evaluation of the silvicultural characteristics that describe the tree as carbon sink. A number of different kinds of tables can be printed that will help the visualization of the data and summarize the results. The software is not designed only for professionals but also for beginners or unskilled workers; as skill levels improve the quality of the input and output increases as well. 41 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni GreenTree stages of implementation During our collaboration with the municipality services, we were faced with 1) a disappointing lack of data and 2) disorganization of the operations related to urban forestry, both due to the absence of proper staff numbers and financial resources. The gap of this procedure is expected to be solved by GreenTree system, as it will make data collecting process much easier. So, the first step that had to be done was to put a set of guiding principles of urban greening management, starting by standardizing the monitoring protocols and the construction of the tree inventory. The challenge was to make protocols as simple as possible and at the same time flexible enough for different situations. These technical guidelines for long-term data collection and urban tree inventory development followed Miller (1997): management planning for street tree population involves an inventory of trees and community values; then, this inventory is used to develop management goals, the next step is to develop a management plan to achieve these goals (selection, establishment and maintenance of street trees), and finally a feedback allows monitoring the entire process. Additionally, guidelines for urban green planning and management under the perception of climate change will be provided, including a national network of public and private organizations as well that deal with urban forestry. General considerations One practical concern faced by all community tree planting programs is the need for biological diversity. The heavy, often exclusive, reliance on a small number of species contributed to the proliferation of speciesspecific landscape pests (Burcham 2009). In the case of Thessaloniki, there are more than 70 tree species participate in urban vegetation. Nevertheless, on the one hand there is an uncontrolled distribution and on the other hand about 30 species are represented with population less than 50 individuals and 11 families have rare appearance, less than 1% of the total population of street trees. So, even though species diversity enhances landscape ecological balance and value, it makes it difficult to organize the spatial and temporal planning of management treatments. Additionally, after in situ observation, there were evidence of inadequate tree species 42 selection for particular features of a site, like damages to pavements or other infrastructure: the size of the mature plant had not been taken into consideration in relation with the size of the available growth space. The life form of approximately 50% of the planted trees in the streets of Thessaloniki belongs to the category of large tree, making the individual inappropriate in the most cases. The large proportion of introduced species shows that no ecological properties and their relations with the environmental conditions were taking into consideration when the selection was made. The inadequate plant selection for urban use also cause health problems to the citizens as the pollen of specific species such as poplar, pines, olive and plane trees can have an allergic effect (Papageorgiou 2003). The existence of power and telephone lines in urban areas is indisputable. Planting trees that are expected to grow high in their maturity leads to inevitable trimming and pruning, that in most of the case is inadequate and without any spatial and temporal planning. Topping and lopping is very often with irreversible impact on tree health. Conclusion In Thessaloniki the basic rules took into account as Miller’s (1997) proposed model for selecting species for urban uses were never applied. The main factors to take into account are site conditions, such as cultural and environmental elements, economic factors like planting and maintenance costs and social factors as functional utility, landscape enhancement and public safety. So, the basic properties of the trees must be: 1) climate adaptation, 2) resistance to diseases, 3) large phenotypic plasticity in the plant materials, 4) root quality, 5) growth potential and form at maturity phase, 6) wind and snow resistance, 7) drought resistance and 8) tolerance of air pollution (Sæbø et al., 2003). Species selection, planting location, and cultural practices all have an impact on the ultimate visual quality, health, and cost of street tree maintaining. The use of appropriate species, the proper location of plantings, and the implementation of a program of preventative maintenance of the street trees, will allow a cost effective tree management system. Action programs related to trees in the ecologia mediterranea – Vol. 41 (2) – 2015 Creation of an Integrated System Model for Governance in Urban MTEs and for Adapting Cities to Climate Change: Preliminary Results urban are: (I) policy making, planning and designing, (II) technical focus, such as selection programs and establishment techniques and (III) management aspects (Konijnendijk & Randrup 2002). Apparently the ecological and functional species selection is of utmost importance but the general point is how to apply an efficient management plan with the lower ecological and social cost. Urban greening, when sustainably managed, can have a central role in climate change mitigation and adaptation. Specifically, from the preliminary results of this project can be concluded that: ––the potential crown volume of the existing trees can be increased to 14.5%; ––the replacement of dead individuals will increase the number of trees within 3%; ––the filling of the empty pavement spaces will increase the number of trees at 6.7%. The numbers above clearly show the significant need for the establishment of a reliable and smart monitoring system for urban tree management. This system could also help in managing the decision making process. Finally, the numbers show that urban environment can be easily improved by applying fast and cheap measures of tree replanting and replacement. References Burcham D.C., 2009. Urban Forest Management for Multiple Benefits: an Analysis of Tree Establishment Strategies Used By Community Tree Planting Programs. Thesis submitted to the Faculty of the University of Delaware. Carreiro M.M., 2006. The growth of cities and Urban Forestry. In: Carreiro M.M. (eds. ), Ecology, Planning, and Management – International perspectives. Springer, New York. EPA, 2015. Glossary of Climate Change Terms. Unites States Environmental Protection Agency. Available on: http://www.epa.gov/climatechange/glossary. html#Climate_change (Accessed in 25-12-2014) Gezer A. & Gül A., 2009. Kent Ormancılıgı-KavramsalTeknik ve Kulturel Boyutu. SDU Orman Fakültesi, Kitap Yayın No 86. Isparta. 29 p. (In Turkish). Gill S., Handley J., Ennos R., & Pauleit S., 2007. Adapting cities for climate change: the role of the green infrastructure. Journal of the Built Environment 33 (1):115-133. Handley J. & Carter J., 2006. Adaptation Strategies for Climate Change in the Urban Environment (ASCCUE). Available on: http://www.sed.manchester. ac.uk/research/cure/downloads/asccue_final_ report_national_steering_group.pdf. (Accessed in 25-12-2014) ecologia mediterranea – Vol. 41 (2) – 2015 Jacobs J., 1993. The death and Life of Great American Cities. New York, 449 p. James P., et al. 2009. Urban Green – Towards an integrated understanding of greenspace in the built environment. Urban Forestry and Urban Greening 8(2): 65-76. Konijnendijk C. C. & Randrup T. B., 2002. Editorial. Urban Forestry & Urban Greening 1(1): 1-4 Küçük V. & Gül A., 2005. Isparta kentiçi yol agaçlandirmalari üzerine bir arastirma, Süleyman Demirel Üniversitesi, Fen Bilimleri Enstitüsü Dergisi. 9-3:111-118 (In Turkish). Lambert, J., & Vallet, M., 1994. Study Related to the Preparation of a Communication on a Future EC Noise Policy. INRETS LRN Report No 9420, INRETS- Institut National de Recherche sur les Transports et leur Securite, Bron. Lang, W.W., 1999. Is noise policy a global issue, or is it a local issue? In: Cuschieri J., Glegg S. and Yan Yong (eds.), Proceedings of Internoise 99 - The 1999 International Congress on noise Control Engineering, December 1999, Fort Landerdale, 1939-1943. Miller R. W., 1997. Urban Forestry: Planning and Managing Urban Greenspaces. Prentice Hall, Englewood Cliffs. NASA. 2014. Global Climate Change. Vital signs of the Planet. Available on: http://climate.nasa.gov/. (Accessed in 25-12-2014) Nowak D. J., & Dwyer J. F., 2000. Understanding the benefits and costs of urban forest ecosystems. In: Kuser, J. (ed.), Urban and Community Forestry in the Northeast. Springer, New York: 11-25. OECD-ECMT, 1995. Urban Travel and Sustainable Development. European Conference of Ministers of Transport, Organization for Economic Co-operation and Development, OECD, Paris. Papageorgiou V., 2003. Inventory of plant polulations and their allergic pollen types in region of N. Moudania Chalkidiki and their clinical importance. PhD Thesis, Aristotle University of Thessaloniki. (In Greek) Papakostas K.T., Zagana-Papavasileiou P., & Mavromatis T., 2014. Analysis of 3 Decades Temperature Data for Athens and Thessaloniki, Greece – Impact of Temperature Changes On Energy Consumption For Heating And Cooling Of Buildings. International Conference ADAPT to CLIMATE. Nicosia, 27-28 March 2014. Pauleit, S. 2003. Urban street tree plantings: Identifying the key requirements. Proceedings of the Institute of Civil Engineers-Municipal Engineers 156(1):43-50. Register, R., 2002. Eco-cities: Building Cities in Balance with Nature. Berkeley Hills Books, Berkeley, 320 p. Samara T., & Tsitsoni T., 2010. The effects of vegetation on screening road traffic noise from a city ring road. Noise Control Engineering Journal. 59 (1), pp. 68-74 Samara T., & Tsitsoni T., 2014. Selection of forest species for use in urban environment in relation to their potential capture to heavy metals. Global NEST Journal. 16 (5): 966-974. Sandberg, U., 1999. Abatement of traffic, vehicle and tire/road noise- the global perspective. In: Cuschieri J., Glegg S. and Yan Yong (eds.), Proceedings of Internoise 99 - The 1999 International Congress on noise Control Engineering, December 1999, Fort Landerdale: 37-42. 43 Thekla Tsitsoni, Nikolaos Gounaris, Aimilia B. Kontogianni, Valia Xanthopoulou-Tsitsoni Shashua-Bar L. & Hoffman M.E., 2000. Vegetation as a climatic component in the design of an urban street: an empirical model for predicting the cooling effect of urban green areas with trees. Energy and Buildings 31 (3): 221-235 Sieghardt, M., E. et al. 2005. The abiotic urban environment: Impact of urban growing conditions on urban vegetation. In: Konijnendijk C.C. et al. (eds.), Urban Forests and Trees. Springer, New York: 281-323. Sæbø A., et al. 2005. The selection of plant materials for street trees, park trees and urban woodlands, In: Konijnendijk C.C. et al. (eds.), Urban Forests and Trees. Springer, New York: 257-280. Sjöman H., et al., 2010. Habitat Studies Identifying Potential Trees for Urban Paved Environments: A 44 Case Study from Qinling Mt., China. Arboriculture & Urban Forestry 36(6): 261-271 Troxel B., Piana M., Ashtona M.S., & Murphy-Dunninga C., 2013. Relationships between bole and crown size for young urban trees in the northeastern USA. Urban Forestry & Urban Greening 12 (2): 144-153. Turner K., Lefler L. & Freedman B., 2005. Plant communities of selected urbanised areas of Halifax, Nova Scotia, Canada. Landscape and Urban Planning 71:191-206. Wolf K. L., 2004. Economics and public value of urban forests. Urban Agriculture Magazine. Issue on Urban and Peri-urban Forestry 13: 31-33. ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Pinelopi DELIPETROU1, Dany GHOSN2, George KAZAKIS2, Panagiotis NYKTAS2, Electra REMOUNDOU2, Ioannis N. VOGIATZAKIS3, * 1. Department of Biology, University of Athens, Greece 2. Department of Geoinformation in Environmental Management, Mediterranean Agronomic Institute of Chania, Greece 3. School of Pure & Applied Sciences, Open University of Cyprus, Cyprus * Corresponding author: [email protected], Tel: +357 22411933 Received: 1 March, 2015; First decision: 15 October, 2015; Revised: 26 October, 2015; Accepted: 10 November, 2015 Abstract We aimed to determine the composition, structure and ecological processes of the vegetation of the coastal dunes with Juniperus spp. in Crete, Gavdos and Chrysi in the South Aegean, Greece, in order to apply sound habitat management and restoration. Vegetation composition, structure and zonation were investigated with plots and transects. Data from seven study sites were classified using TWINSPAN. The major patterns of the vegetation data and their relation to environmental variables were explored by DCA and CCA ordination techniques. Ellenberg indicators for moisture, nutrients, and salt were used to characterise the community types identified. The habitat’s floristic composition includes 142 plant species. Five principal community types were identified. Vegetation distribution was related to geomorphology and disturbance gradients. The analysis of transect data identified 20 vegetation units on incipient dune, foredune, interdune and hind dune. Vegetation and geomorphological data were used to construct sand dune profiles for each site, while a set of 36 keystone and 76 indicator species were identified. The sites examined have varied levels of dune development and face different threats. Habitat management should address grazing and trampling at the local level but also land use changes at the catchment level. Keywords: CCA, DCA, Ellenberg values, geomorphologic units, Natura 2000, PCA. ecologia mediterranea – Vol. 41 (2) – 2015 Introduction Coastal dunes are found on shores all over the world, from polar latitudes to the tropics (Martinez et al. 2004). Plant growth forms, sea and wind are the key factors that create, mould or destroy these structures. Among these factors, plant growth plays a vital role in all stages of coastal dune formation by reducing the effect of the wind, trapping sand, and thus encouraging further dune growth (Musila et al. 2001). Mediterranean coastal dunes are characterized by strong seasonality, with warm wet winters and hot dry summers (Castillo et al. 2002). Vegetation is exposed to stress caused by drought, high evaporative demand, and both high irradiance and temperatures (Garcia Novo et al. 2004). Coastal dunes with Juniperus spp. (juniper) are widespread on the coasts of Europe, but not common. Maritime juniper woodlands (Juniperus macrocarpa Sm.) represent the late successional stage of Mediterranean dunes and cliffs. According to Rivas-Martinez et al. (1980), maritime juniper woodlands with Juniperus phoenicea, represent the mature ecosystem on outer dunes and cliffs of the Mediterranean coasts. They have a high ecological value in relation to their ability to retain sand, and 45 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis provide habitat for flora and fauna. Coastal dunes with Juniperus spp. are vulnerable due to their extreme ecological position (Gehu 1993), withstanding the effects of wind, drought, salt, erosion and pH (Brown & McLachlan 1990). At a european level, coastal dunes with Juniperus spp. are threatened by logging, urban and tourism development, fires, invasive species, erosion, grazing, fragmentation, pollution but also by juniper’s restricted natural regeneration (Anon 1992). Global climate changes may exacerbate these threats through changes in minimum winter temperatures. These changes can modify the distribution of plant species, and can combine with frequent storms resulting in damage to the seaward side of the dune slacks. Changes in dune structure and ecosystems are often cyclical, with periods of loss (erosion) balanced by periods of gain (sand deposition) so these tendencies will be obvious only in the long term (Corre 1991). Junipers usually occur in isolated stands of different extent, but large populations still survive in natural or semi-natural conditions. The protection of coastal woodlands is a priority due to the range of functions they perform for biodiversity, recreation and sand stabilization. In the European Union, coastal dunes with Juniperus spp. are a priority habitat (2250*) of the Habitats Directive (Council of European Community 1992). Worldwide studies of the relationship between vegetation and dune formation (Castillo et al. 1991) show that coastal dunes form a complex system of habitats for plants due to the combined effect of steep environmental gradients which are related to the distance from the shoreline, elevation but also to salinity, nutrients, humidity, wind, and inundation (Ranwell 1972). Plants suited to the dune habitat are highly specialized and able to cope with limiting factors such as water shortage, soil fertility, and salt spray. In this prograding shoreline, a temporal and spatial dune succession is found. In the late 20th century, conservation efforts were focused on plant inventories, threatened species and maintaining as many natural habitats as possible. Today, there is a paradigm shift with emphasis placed on understanding the physical and ecological processes within and beyond these habitats as a means for improved conservation. This is a necessity in the South Aegean which contains a high concentration of sand dunes with junipers, 46 currently facing increased pressures (Heslenfeld et al. 2008). Vegetation is by far the most important component of the biota on sand dunes because it is directly involved in establishing the dune formation and structure (Wolfe & Nickling 1993) while there is constant interaction of dune vegetation and geomorphology (Maun 2009). Despite their ecological value, baseline information on juniper dune woodlands composition, structure and ecology are lacking in Greece. This is crucial if this priority habitat is to be effectively managed and restored where necessary. Moreover, describing plant communities is challenging, due to the lack of characteristic species and the resulting difficulties in phytosociological classification (Mayer 1995). This study focuses on the sand dune systems of the islands of Crete, Gavdos and Chrysi in South Aegean, which are part of the European network of protected areas Natura 2000, as a precursor to sound conservation actions. The focus is on coastal juniper woodlands, which are important indicators of the general physical conditions of the coastal dune environment. The objectives of this paper are to: a) identify and describe the plant communities of juniper shrub and vegetation zonation in the sand dunes in the islands of Crete, Gavdos and Chrysi, b) investigate the relationship between plant communities and environment; c) assess the implications of the results for dune habitat conservation and management. Methods Study areas In Greece this habitat is found in 15 sites mainly in the south covering 1,077 ha, i.e. 4.26% of the total surface of Natura 2000 in the country (Picchi 2008). This study was carried out in all Natura 2000 sites designated as coastal dunes with Juniperus spp. habitats on the islands of Crete (sites of Kedrodasos and Falasarna), Gavdos (sites Agios Ioannis, Lavrakas and Sarakiniko) and Chrysi (Figure 1). Data collection Fieldwork took place in May 2009 (Kedrodasos and Gavdos) and in April-May 2010 (Chrysi and Falasarna). Vegetation ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Figure 1 – L ocation of the 2250* priority habitats in Chrysi, Kedrodasos-Elafonisi, Gavdos and Falasarna (upper left) and the sampling scheme employed for vegetation plots (bottom left) and transects (bottom right). composition and structure were studied in 38 plots of 30 m × 30 m established within the juniper scrub forest in all sites: 6 in Kedrodasos-Elafonisi (E1 – E6), 4 in SarakinikoGavdos (Gs1 – s4), 6 in Ag. Ioannis-Gavdos (Gi1 – Gi4), 8 in Lavrakas-Gavdos (Gl1 – Gl8), 6 in East Chrysi (Ca1 – Ca6), 4 in West Chrysi (Cw1 – Cw4), and 2 in Falasarna (F1 – F2). Vegetation plot locations were selected subjectively so as to a) span the studied areas; b) represent various levels of disturbance; and c) include both Juniperus macrocarpa and Juniperus phoenicea shrub. Plot locations were recorded with a GPS. Two subplots were established in each 30 m × 30 m plot: a 10 m × 10 m subplot including juniper shrub or trees (Figure 1) and a 3 m × 3 m subplot in the open (outside the canopy of shrubs or trees). Percentage cover of vegetation layers (trees, tall shrubs, low shrubs, and herbs) and woody vegetation height were visually estimated in all subplots. The abundance of all plants in each subplot was recorded using the modified Braun-Blanquet 9-grade ecologia mediterranea – Vol. 41 (2) – 2015 cover-abundance scale (Braun-Blanquet 1932; Wikum & Shanholtzer 1978). Environmental data included slope and aspect, geology from 1:50000 scale maps, grazing pressure assessed in the field qualitatively (high, medium, low). Plot use was also assessed qualitatively (high, medium, low) based on the presence of broken branches observed. Vegetation zonation was studied in 12 line transects perpendicular to the seashore, 2 of which in Kedrodasos-Elafonisi, 1 in Sarakiniko-Gavdos, 3 in Agios Ioannis-Gavdos, 2 in Lavrakas-Gavdos, and 4 in Chrysi. Their length varied from 55 to 225 m depending on dune morphology while their locations were selected subjectively so that they were representative of vegetation zonation at each site and were geo-referenced using a GPS. Total plant cover, rock cover and vascular species cover-abundance, using the modified BraunBlanquet 9-grade scale (Braun-Blanquet 1932), were recorded in contiguous 5 m × 5 m quadrats along each transect. 47 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Life forms were based on Raunchier (1937) as modified by Ellenberg & Müller-Dombois (1967) while chorology was based on Jahn & Schönfelder (1995). Voucher specimens of plants were collected and stored in the Herbarium of the Mediterranean Agronomic Institute at Chania, Crete. A list of the total species recorded together with their functional attributes (Böhling et al. 2002; Jahn & Schönfelder 1995; Espejel et al. 2004) is presented in Appendix I. Data analysis included only the Ellenberg indices for moisture (F), nutrients (N), and salt (S), because they presented significant variation among the species, while the indices for light (L, values 6 to 9, semilight to light plants), pH (R, values 7 to 9, weakly basic to basic soils), and temperature (T, values 7 to 8, fairly hot to hot sites) did not present significant variation. Data analysis TWINSPAN analysis (Hill 1979a) in PCORD (Ver. 4, McCune & Mefford 1999) and JUICE© (Ver. 7.0.61, Lubomír Tichý 19982010) was used to classify the vegetation and transect plot data. The 10 m × 10 m and 3 m × 3m plots were analyzed separately. The phi-coefficient (Chytrý et al. 2002) with data standardization estimated by JUICE was used as a fidelity measure. Species with significance < 0.05 according to the Fischer’s exact test were given zero fidelity value. Species were assigned to vegetation classes (Appendix II) according to, Rivas-Martínez et al. (2002). Detrended Correspondence Analysis (DCA; Hill 1976) was employed to evaluate the type of response model for selection. Based on the DCA results, a unimodal response model for the data was accepted (Jongman et al. 1987). Therefore Canonical Correspondence Analysis (CCA) in CANOCO© (ter Braak & Smilauer 1998) was used to assess vegetation environment relationships The CCA model and the significance of the fitted environmental variables were evaluated by the Monte Carlo permutation test. Table 1 – Number of species (in bold) in sand dune juniper shrub vegetation in the studied sites and floristic similarity (Sørensen index) between them. Kedrodasos Falasarna Kedrodasos Falasarna Gavdos 92 0.14 0.43 37 0.25 62 Chrysi 0.43 0.31 0.48 48 Gavdos Chrysi 75 Results Floristic analysis The total number of species in sand dune juniper shrub based on the 30 m × 30 m plots in all the sites was 142. Species numbers varied among sites and floristic similarity was relatively low (< 50%) (Table 1). The highest species richness, in terms of absolute numbers, was observed in Kedrodasos and the lowest in Falasarna. Floristic similarity was higher among Chrysi, Gavdos, and Kedrodasos and lower between Falasarna and all the other sites. The dominant families were Poaceae (14%), Compositae (13%) and Fabaceae (11%). The dominant chorological element was the Mediterranean (26%) followed by the East-Mediterranean (12%), South-Mediterranean (10%) and Mediterranean-Atlantic (9%). The Greek endemic element accounted for 5% of the flora. Chrysi had the highest number of endemics. Vegetation composition and structure/Vegetation plot data analysis The species with the highest frequency in all the plots were the phryganic shrubs Coridothymus capitatus and Phagnalon graecum (26 and 21 plots, respectively); a dry grassland small annual, Hedypnois cretica (22 plots); and the ammophilous annuals Pseudorlaya pumila, Triplachne nitens, Vulpia fasciculata and Lotus halophilus (29, 26, 21, and 24 plots, respectively). These, along with J. macrocarpa and J. phoenicea were considered as characteristic of all the juniper shrub communities of the area of Crete. TWINSPAN analysis of the 10 m × 10 m plots resulted in 5 distinct groups (Figure 2, Appendix II). The vegetation units corresponding to these groups are referred to as “community types” (cf. Mueller-Dombois & Ellenberg 1974), since they have not been placed into existing phytosociological categories. Species with a high constancy (frequency > 50%) and a high degree of fidelity (phi-coefficient > 50%) in each group were termed characteristic (or diagnostic) and were used to name each community type. The life form spectra (Table 2) and Ellenberg number spectra (Table 3) were also used for the characterisation of the community types (Table 5) identified, namely: ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) A1: Juniperus phoenicea-Periploca angustifolia community type. A2: Malcolmia flexuosa-Nigella stricta community type. B1: Silene colorata-Ononis natrix community type. B2: Silene succulenta-Cutandia maritima community type. B3: Elytrigia juncea-Medicago marina community type. Ecological gradients/Vegetation plot data analysis Based on the DCA preliminary analysis a unimodal response model for the data was assumed (ter Braak & Smilauer 1998). The full CCA model explained 1.837 of the total inertia of the species data (3.930) and selected automatically five significant variables (p < 0.05), in decreasing order of significance: conglomerate, dune depth × interdune depth, grazing, plot use, and marl. Thus, the environmental variables best explaining the species data were related to geomorphology and anthropogenic disturbance (Figure 3). In Figure 3 the upper quadrants include sites on marl (e.g. Chrysi), and the lower sites on sandstone (e.g. Gavdos), and conglomerate (e.g. Kedrodasos). Dune-interdune depth and slope increase from top to bottom. On the contrary grazing and plot use increase from bottom right to top left. Juniperus macrocarpa is located at the bottom left quadrant but close to the start of the axes along with the species Figure 2 – Hierarchical classification of the vegetation data (community types A1, A2, B1, B2, B3 as in text). Table 2 – Life form spectra of juniper shrub plant community types. The number of plants and percentage cover (in parenthesis) for each life form are shown. Plant Community Life Form A1 A2 B1 B2 B3 Therophyte scapose 23 (8) 39 (32) 22 (18) 18 (15) 14 (13) Therophyte caespitose 5 (4) 12 (10) 11 (5) 9 (8) 5 (4) 2 (1) 2 (0) 2 (1) 2 (1) 1 (1) Therophyte parasitic Therophyte rosette-forming 2 (1) 5 (2) Hemicryptophyte 3 (1) 3 (2) 1 (4) 4 (4) Hemicryptophyte biennial 2 (0) 2 (1) 1 (0) 1 (0) 2 (0) Geophyte bulbous 3 (1) 3 (2) 1 (1) 1 (0) 1 (1) Chamaephyte fruticose 5 (5) 3 (4) 9 (8) 4 (7) 3 (4) Chamaephyte suffruticose 7 (8) 3 (2) 5 (3) 3 (1) 3 (3) Phanerophyte caespitose 4 (73) 4 (44) 5 (58) 3 (63) 3 (68) Geophyte rhizomatous 4 (2) Phanerophyte scapose 1 (5) Table 3 – Ellenberg number spectra for moisture (F), nutrients (N), and salinity (S) of juniper shrub community types. The number of plants and percentage cover (in parenthesis) for each Ellenberg number are shown. Plant Community Type Ellenberg Number A1 F N A2 S F N B1 S F 11 (6) N F 2 (1) 6 (4) 2 (0) 1 (0) 46 (31) 6 (5) 29 (23) 4 (9) 17 (13) 4 (5) 13 (9) 7 (7) 12 (56) 9 (8) 6 (67) 6 (7) 3 (3) 2 (0) 1 4 (3) 25 (15) 2 13 (78) 1 (0) 11 (11) 15 (42) 3 12 (7) 5 (34) 7 (35) 21 (22) 4 (30) 6 (37) 18 (21) 5 (53) 4 10 (4) 7 (3) 1 (0) 1 (2) 5 5 (1) 14 (8) 1 (0) 4 (3) 9 (5) 2 (1) 6 7 1 (0) 8 4 (1) 4 (1) 20 (16) 10 (7) 5 (6) 11 (11) 8 (65) 3 (1) 7 (54) 10 (14) 2 (67) S F N S 1 (0) 8 (71) 10 (12) 2 (68) 11 (76) 11 (10) 5 (5) 11 (8) 2 (1) 2 (1) 11 (9) 1 (1) 2 (1) 12 (10) 3 (1) 9 (6) 3 (1) 7 (6) 1 (4) 5 (3) 8 (7) 3 (2) 1 (0) 16 (14) 1 (0) 11 (9) 1 (0) 11 (14) (0) 1 (1) 8 (7) 1 (1) 1 (0) 15 (13) 1 (0) 12 (10) 1 (0) 9 (7) 1 (0) 8 (10) 1 (1) 1 (0) 1 (0) 4 (1) 1 (1) 10 (6) 1 (0) 10 (12) 3 (7) 4 (3) 1 (3) 3 (2) 1 (1) 1 (0) 4 (1) 9 x 5 (2) N B3 S 0 5 (4) B2 1 (0) 6 (3) 10 (47) 4 (38) 8 (13) ecologia mediterranea – Vol. 41 (2) – 2015 8 (22) 5 (16) 5 (6) 3 (4) 3 (3) 49 50 F E 27 2110, 2230 F2.2 2260 2110, 2230 2250 F3.1 F3.2 B2.4 F3 2260 F2 26 2250 2110 2250 2250 2250 2250 2250 2250 2250 2250 2250 2190 2230 2110 2250 2110 2260 2110 Habitat Type F2.1 B2.3 B* 19,20, 21 F1 B3.3 18 25 B2 16, 17 24 B2.2 23 A1.1 6 B2.1 B3.2 11 15 D4 10 B1 D3 8,9 12 D2 7 A2 B3.1 4 13, 14 D1 5 C2 3 D C1 1,2 C Vegetation Unit Sub group Group ecologia mediterranea – Vol. 41 (2) – 2015 foredune incipient dune, interdune foredune, hind dune incipient dune, interdune foredune, hind dune foredune, hind dune incipient dune foredune, hind dune foredune foredune, hind dune hind dune, rarely foredune foredune hind dune foredune, hind dune hind dune foredune dune slack interdune incipient dune, rarely interdune foredune incipient dune, interdune incipient dune, foredune incipient dune Zone Gavdos (Lavrakas, Sarakiniko) Gavdos (Lavrakas, Sarakiniko) Kedrodasos, Gavdos (Lavrakas, Agios Ioannis, Sarakiniko) Chrysi (East), Kedrodasos, Gavdos (Agios Ioannis) Gavdos (Lavrakas) Gavdos (Agios Ioannis, Lavrakas) Gavdos (Agios Ioannis, Lavrakas) Gavdos (Agios Ioannis, Lavrakas) Chrysi (West) Gavdos (Agios Ioannis) Chrysi (East), Gavdos (Lavrakas) Gavdos (Agios Ioannis) Gavdos (Lavrakas) Kedrodasos Chrysi (West) Chrysi (West) Chrysi (West) Chrysi (East, West) Chrysi (mainly West), Gavdos (Agios Ioannis) Chrysi (East) Chrysi (mainly East) Kedrodasos, Gavdos (Agios Ioannis) Kedrodasos Site Silene succulenta, Plantago squarrosa, Medicago marina, Medicago littoralis, Hedypnois cretica Mostly low cover vegetation with few ammophilous species, lacking any other shrub species and dry grassland species. Divided into 2 subcommunities based on the presence of shrubs. Silene succulenta-Pseudorlaya pumila Juniper shrub (community type B2) with Ononis natrix, Coridothymus capitatus, lacking any other shrub species on foredune or more inland at the margins of erosion corridors. Ammophilous primary dune vegetation at the frontal zone of Sarakiniko and at large shrub openings on deep dune at Agios Ioannis. Dune shrub with Ononis natrix-Coridothymus capitatus on foredune, behind the frontal zone or on hind dune. Mostly low cover vegetation with an array of ammophilous and dry grassland species. Divided into 3 subcommunities based on the presence of shrubs. Very low cover, fragmental vegetation of frontal dune or shrub openings. Dune shrub with Ononis natrix-Coridothymus capitatus on foredune, behind the frontal zone, or on hind dune. Juniper shrub (community type B2) representing an aspect of B2.1 at more stabilised dune. A group of juniper shrub quadrats with floristic composition between communities B1 and B2 (apparently at transition zones). Juniperus macrocarpa, Plantago squarrosa Frontal, very low cover primary dune vegetation. Juniper shrub (community type B3), on foredune. Juniperus macrocarpa, Limoniastrum monopetalum Juniperus macrocarpa, Limonium elaphonesicum, Ononis natrix, Coridothymus capitatus Juniper shrub (community type B2) with few species. Juniperus macrocarpa, Cutandia maritima Limonium elaphonesicum Juniper shrub (community type B2) with a few ammophilous species and stable presence of synanthropic vegetation species, mainly at hind dunes highly disturbed by human use. Juniperus macrocarpa, Mercurialis annua, Cakile maritime Juniper shrub (community type A1) at inland dunes. Juniperus phoenicea, Asparagus stipularis, Periploca angustifolia Juniper shrub (community type B2) , lacking sclerophyllous shrub and dry grassland species, on foredune (frontal) or more inland at the margins of erosion corridors. Juniper shrub (community type B3), on foredune (frontal) or more inland at the margins of erosion corridors or at steep slopes. Juniperus macrocarpa, Pancratium maritimum, Silene sedoides Juniperus macrocarpa, Silene succulenta, Limonium elaphonesicum Dune slack with Juncus heldreichianus, at flat wet dunes, behind the front vegetation zone. Juncus helreichianus Juniper shrub (community type B1). Low cover dune grassland occurring on loose sand at inland shrub openings; Juniperus macrocarpa seedlings are frequent. Triplachne nitens, Erodium lacianiatum Juniperus macrocarpa, Ononis natrix, Erica manipuliflora Ammophilous primary dune vengetation similar to D1 but lacking ammonitrophilous plants (e.g., Zygophyllum album, Cakile maritima) at the front zones and rarely at interior large shrub openings with moving sand or blowouts. Silene succulenta, Euphorbia paralias Juniper shrub (community type A2). Juniper shrub (community type B3), on foredune (frontal) or more inland at the margins of erosion corridors or at steep slopes. Juniperus macrocarpa, Zygophyllum album Juniperus macrocarpa, J. phoenicea, Nigella stricta Ammophilous primary dune vegetation including sand binders (Elytrigia juncea, Medicago marina, Limonium graecum, Limoniastrum monopetalum) at the front zones or at interior large shrub openings with moving sand or blowouts. Dune shrub, occasional presence of ammophilous plants (Centaurea pumilio, Limonium elaphonesicum) Anthyllis hermaniae Silene succulenta, Zygophyllum album Ammophilous primary dune vegetation on flat dune with very low cover. Description Centaurea pumilio, Pancratium maritimum Characteristic species Table 4 – Vegetation units identified by the transect plot data. Habitat type code as in Appendix III. Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) that occur in many plots and are not highly affected by the environmental variables tested such as Pseudorlaya pumila, Lotus halophilous and Pistacia lentiscus. These are candidates for keystone species due to their abundant and stable presence in the habitat. The species closely related to larger dune depths are candidates for indicators of deep dunes and for keystone species are those that characterize communities B2 (Silene succulenta, Cutandia maritima) and B3 (Elytrigia juncea, Medicago marina) and also Salsola kali, Cakile maritima and Lycium schweinfurthii. Juniperus phoenicea is located at the upper quadrant since it occurs at shallower dunes along with Periploca angustifolia and Ononis reclinata. It is not an indicator of dunes on marls since it also occurs on sandstone. The species closely related to marls (e.g. Paronychia macrosepala, Helianthemum stipulatum) are candidates for indicators and for keystone species of shallow and moderate depth dunes on marl. The species closely related to conglomerates and higher rock cover are candidates for indicator and keystone species and are the characteristic species of community type A2, Malcolmia flexuosa, Nigella stricta, and Minuartia mediterranea. Vegetation zonation TWINSPAN analysis of the transect plot data resulted in four well distinguished large groups, whose segregation was strongly influenced by the locality: Group C – Anthyllis hermaniae-Centaurea pumilio (mostly Kedrodasos); Group D – Silene succulentaPancratium maritimum (mostly Chrysi); and two Silene succulenta-Pseudorlaya pumilaCutandia maritima groups; Group E – Juniperus macrocarpa (juniper shrub, all sites); Group F – Ononis natrix-Coridothymus capitatus (mostly Gavdos). Silene succulenta, Table 5 – Community types resulting from TWINSPAN analysis of plot data. Community Type Description A1: Juniperus phoeniceaPeriploca angustifolia Shrub community with Juniperus phoenicea often mixed with J. macrocarpa on shallow/moderately deep dunes with shallow/ moderately deep interdunes, on marls. It is typical of flat, inner dunes and of the transition zones to the shrublands of the inner non dunal vegetation. It was found in ungrazed sites of Chrysi, but also in Gavdos. Floristically characterised by phanerophytes (notably P. angustifolia, restricted to Gavdos and Chrysi in Greece), and chamaephytes (notably the rare in Greece Helianthemum stipulatum), by high frequency/low cover of dry grassland species and by low participation of ammophilous species. The ecological profile, based on Ellenberg indicators, shows dominance of plants of extremely dry to dry sites (values 0–2), salt indifferent to medium halotolerant (values x, 0–2) and increased participation of plants of nutrient extremely poor to poor sites (values 2–4) or indifferent. A2: Malcolmia flexuosa-Nigella stricta Recorded only at Kedrodasos, at mostly grazed moderately deep dunes with moderately deep interdunes, on conglomerate. It is characterised by reduced participation of shrubs (including the resistant to grazing Verbascum spinosum), high frequency/high cover of annuals including few ammophilous species (notably the rare Aegean endemic Nigella stricta), and a large number of dry grassland and synanthropic vegetation species. The ecological profile, shows dominance of plants of extremely dry to dry sites (values 0–3) and increased participation of plants of intermediate to extremely nutrient-rich sites (values 5–9) and of halophobe to medium halotolerant (values 0–2) plants. The increased nutrient indicator values are mainly due to the increased participation of synanthropic species (e.g., Urospermum picroides, Sonchus oleraceus, Mandragora autumnalis). B1: Silene colorata-Ononis natrix This is the main Juniperus macrocarpa community type of moderately deep dunes on Gavdos and on Chrysi and at the deep dunes of Kedrodasos. Floristically characterized by the dominance of phanerophytes and chamaephytes (pines, sclerophyllous shrubs, phrygana), high frequency of annuals, and an array of dry grassland species. It often represents the innermost dune zone preceding pine forest or shrub on stable substrate. The ecological profile shows dominance of plants of extremely dry to dry sites (values 0–3) and slightly to very halotolerant (values 1–3). Plants of nutrient poor sites (values 2–4) dominate, with an increased frequency of plants of intermediate to nutrient rich sites (values 5–8). The increased nutrient indicator values are due to both ammophilous (e.g. Vulpia fasciculata) and synanthropic species (e.g. Sonchus oleraceus). B2: Silene succulenta-Cutandia maritima The main Juniperus macrocarpa community type of deep dunes at all sites (rare at Kedrodasos). Floristically characterized by a less prominent shrub layer and high frequency/high cover of annual and increased cover of perennial herbs including ammophilous species (notably Silene succulenta occurring only in the project sites in Greece) and fewer dry grassland species. The ecological profile is similar to community B1, but very halotolerant plants are prominent and there is increased participation of salt stressed plants (values 7–8) and nutrient rich sites (value 8). The increased nutrient indicator values are mainly due to ammophilous and ammonitrophilous drift line species (e.g. Cakile maritima, Salsola kali). B3: Elytrigia juncea-Medicago marina It occurs on foredunes and was found on deep dunes at Chrysi, and degraded, medium deep or flat dunes, at Falasarna. Floristically characterized by a less prominent shrub layer and by increased perennial versus annual herbs including mainly ammophilous species, especially sand binders (e.g. Limonium graecum, Sporobolus pungens) and few dry grassland species. The ecological profile is similar to community B2, but there is more increased frequency and cover of plants of salt stressed (values 6–8) and nutrient rich sites (value 8). This attribute is exemplified by the nitrophilous, halotolerant shrub Lycium schweinfurthii ecologia mediterranea – Vol. 41 (2) – 2015 51 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Figure 3 – C CA plot of 36 vegetation plots (123 species), with symmetric biplot scaling, axes I (eigenvalue 0.400, p = 0.005) and II (eigenvalue 0.342, p = 0.005). The two axes explain 18.9% of the variance of species data and 40.4% of the variance of species-environment relation. Triangles: sites; circles: species; diamonds: characteristic species of the plant community types; stars: most frequent species in all community types. Black arrows: variables with non significant contribution to the model; coloured arrows: variables with significant contribution to the model. Coloured species’points represent the species with a strong positive relation to the respective environmental variables (top 10 weighted averages of species with respect to the variable). Triplachne nitens and Lotus halophilous were the species with the highest frequency in all transect plots of all sites, except from Kedrodasos. The analysis further separated 27 subgroups (maximum dissimilarity 0.508) which were assigned to the 20 vegetation units (Figure 4, Table 4). These represent the spatial 52 succession of sand dune community types from the sea landwards, but since the transect plots may represent fragments or marginal zones of communities, the term “community type” is not used in the phytosociological sense. The juniper shrub vegetation units were assigned to the community types identified ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Figure 4 – H ierarchical classification of the transect plot data. Primary dune/dune grassland community types (habitat types 2110, 2230): C1, D1, D2, D3, D4, F1, and parts of F2 and F3; dune slack (habitat type 2190): D4; low shrub community types (habitat type 2260): C2, F2, F3; juniper shrub community types (habitat type 2250): A1, A2, B1, B2, B3, and part of F3. For community type abbreviations see text. by the plot data and named correspondingly. Based on both the vegetation and the transect plot data, the vegetation succession profiles in the study sites were constructed (Figures 4, 5, 6, and 7). Geomorphology, vegetation and dune development In this section we present the most characteristic relationships recorded between vegetation units and dune development while details are given in Table 4. In Chrysi (Figure 5) the dune system presents well developed primary, secondary, and tertiary vegetation zones. At the northwest part, a wide, moderately deep incipient dune zone is formed with high cover, typical ammophilous vegetation (mainly v.u. D2). Subsequently, a moderately high foredune with juniper shrub (v.u. B3.2) and a low and moderately high hind dune (v.u. A1) are formed, with moving sand (v.u. D2) and dune grassland openings (v.u. D3). At the northeast part the incipient dune zone is also wide, but ammophilous communities (v.u. D1, F2.2) form a high cover but narrow front zone followed by wide openings with scarce vegetation (v.u. D1, D2). At Kedrodasos (Figure 6) the primary vegetation zone is not well developed. Sparse, almost ecologia mediterranea – Vol. 41 (2) – 2015 flat incipient dunes with scattered ammophilous plants of low sand-binding ability of the vegetation unit (v.u. C1) (Centaurea pumilio, Pancratium maritimum, Pseudorlaya pumila) or shrubs (v.u. C2) are formed at the front zone of the east part and behind the rocky beach at the west part. At Sarakiniko-Gavdos (Figure 8), the incipient dune area is characterised by mostly low cover ammophilous vegetation (v.u. F3.2) followed by a low foredune with phrygana (v.u. F3.1) and then juniper shrub (v.u. B2.4). At the west part of Agios Ioannis, Gavdos (Figure 7), a front almost flat zone with scarce ammophilous plants (v.u. F1) is succeeded by rising, moderately deep foredune with juniper (v.u. B*, B2.1) and by deep dune areas with juniper (v.u. B2.1) and large openings with very low vegetation cover (v.u. F2.2, F3.2). At the extended dune system of Lavrakas, Gavdos (Figure 8) the incipient dune zone is not well developed and includes low cover ammophilous communities (v.u. F1) which are succeeded by low cover phrygana (v.u. F2.1) marking the transition to the moderately high foredune which is stabilised by the same phrygana (v.u. F2.1, F3.1) and by juniper shrub (v.u. B2.2, B2.4, B2.3). The hind dune develops on low dunes and shallow sand deposits with often thick juniper shrub (v.u. B*, B1). 53 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Figure 5 – S and dune vegetation profiles in Chrysi. 54 ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Figure 6 – S and dune profiles in Kedrodasos. Keystone and Indicator species On the basis of vegetation analysis we identified keystone species for the habitat and indicator species for the geomorphological characteristics and the quality of the habitat with regards to plot use and grazing. The indicator species were selected by CCA. The species with the top 10 positive weighted averages with respect to each variable were selected as showing positive relation and the species with the bottom 10 negative weighted averages were selected as showing negative relation. The general literature for each species was taken into account for the final selection (i.e., obviously generalist species were excluded). The criteria for the selection of keystone species were: a) species identified as characteristic (diagnostic and/or constant) of plant communities resulting from TWINSPAN; b) the most frequent species in habitat 2 250*, i.e. those occurring in more than 40% of all the plots; c) species that were identified as indicators of the geomorphological environmental ecologia mediterranea – Vol. 41 (2) – 2015 variables; d) species with high Ellenberg indicator value; e) the set of keystone species should contain representatives of all the functional attributes (see Appendix I). A set of 36 keystone species (31 for the habitat 2250) and a set of 79 indicator species were identified (Appendix IV). Both sets can be used for habitat quality assessment. Discussion The sand dune systems investigated in this study, represent an array of geomorphological conditions, a gradient of human influence and geographical isolation as reflected in the range of community types recorded. The plant species number of the juniper shrub community types in each site (Table 1), was comparable to that found in similar communities in Greece (unpublished relevé data of NATURA 2000 sites) and western Mediterranean (Acosta et 55 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Figure 7 – S and dune profiles in Gavdos-Ag. Ioannis. al. 2009). The community types of Kedrodasos which had the highest species numbers also had the highest number of synanthropic elements (sensu Mucina 1997; Acosta et al. 2006) (Appendix II). Floristically similarity is higher between Gavdos and Chrysi indicating stronger biogeographic affinities. The comparatively low floristic similarity of the juniper shrub 56 community types among islands can be attributed partly to geographic isolation and partly to the different environmental variables. Geographical vicariation in the Mediterranean Juniperus macrocarpa communities was also found by Géhu et al. (1990). In fact, the community types identified in the study sites seem to represent yet another geosynvicariant(s) of the Junipereta macrocarpae which cannot be ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Figure 8 – S and dune profiles in Gavdos-Lavrakas. classified with the existing associations in the South Aegean (Géhu et al. 1989). The life form and Ellenberg number spectra (Tables 2 and 3) differentiated the juniper shrub communities primarily according to the abiotic factors (foredune vs hind dune, dune depth, transition to stable substrate) and secondarily according to human impact (grazing). These two functional attributes can detect essential ecologia mediterranea – Vol. 41 (2) – 2015 qualitative changes in species composition of this priority habitat and can be used effectively in monitoring and assessment of its conservation status. The ecological gradient analysis of the juniper shrub communities was strongly influenced by the presence of species with narrow distribution (e.g. Limonium elaphonisicum and Limonium graecum) and this resulted 57 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis in biogeography masking to some extent the influence of the ecological factors tested. This supports the notion that random biogeographical events may affect community composition at a large scale (Forey et al. 2008). Also, the effects of natural disturbance and possible differences in precipitation and soil moisture among the sites, important factors for dune community composition (Forey et al. 2008; Miller et al. 2009), were not tested. Nevertheless, the analysis highlighted the well documented relationship between geomorphology and vegetation establishment in the dune systems (Provoost et al. 2011). Most importantly, the analysis showed that vegetation composition in the study sites is affected by anthropogenic disturbance and grazing. The effects of trampling on vegetation composition and structure in sand dunes are well documented (Santoro et al. 2012) as is the increased resilience of mobile and fixed dunes compared to semifixed ones (Lemauviel & Rozé 2003). Grazing, when practiced in a non-intensive manner, can be beneficial for the typical fixed dune communities (Halada et al. 2011) and juniper regeneration (Pihl et al. 2001). Although, this practice is advocated in central and northern Europe where the habitat’s natural successional trend is always towards climax woodland, it may not be beneficiary in the Mediterranean where stable dunes face severe exposure (Picchi 2008). In our study, grazing favored synanthropic species at the expense of dry grassland and ammophilous ones. Also, although these effects were not quantified, observations in the field revealed that grazing by goats may also change the architectural structure of the juniper trees, especially of Juniperus macrocarpa. Vegetation zonation was studied and profiles constructed for each site since front vegetation zones protection is crucial for the conservation of the fixed dunes where juniper develops (Hesp 2002). The plant community types identified exhibit similarities with sand dune communities described elsewhere in the Mediterranean (Mayer 1995; Sýkora et al. 2003) and show a typical linear succession (zonation) (e.g. Cassar & Stevens 2002). Of all three study areas only the dune systems in Chrysi, demonstrate well-developed primary, secondary, and tertiary vegetation zones. Vegetation zonation is better preserved in the western area which is less frequented by visitors compared to the eastern part, where there is degradation due to 58 trampling, sea-shell overcollection, and erosion. Extended drying and an unbalanced sex ratio have been observed in Juniperus macrocarpa trees mainly in the eastern part, probably due to extreme drought stress episodes (Thanos et al. 2010). It seems that vegetation has recovered naturally in the western part but that human disturbance has impeded this process in the east part. Aeolian sand deposition on a rugged terrain is more important for dune development in Gavdos and Kedrodasos. The typical primary dune vegetation with ammophilous species is not well developed while a shrub zone plays the main protective role for the juniper shrub dunes. It is only in Gavdos-Sarakiniko, the site most affected by human disturbance, that the abiotic conditions favor the development of typical primary dune vegetation, but this is degraded due to trampling and grazing. In Kedrodasos, the whole front vegetation is rather thin and this may be due partly to the rugged terrain and partly to excess trampling. Habitat Management Coastal sand dunes play an important role in providing a range of ecosystem functions and services (Barbier et al. 2011). The conservation of these habitats requires more detailed knowledge of their ecology and condition. To adhere to Habitats Directive (Council of European Community 1992) European member states should report every six years on the progress achieved in protected habitat conservation. Towards that end, the keystone and indicator species for the habitat identified in this study can be employed. In addition, the different processes and dune development stages that occur in the study sites require habitat management plans tailored to individual sites’needs. The conservation of the natural processes is a major target in dune system management (Van der Meulen & Salman 1996) due to their effect (Ciccarelli et al. 2012; Miller et al. 2009). Usually management interventions to tackle the above effects should be combined with relief from anthropogenic pressures (Picchi 2008). In this light and according to the results of our study, it seems that interventions including restoration of the front dune vegetation are warranted only in the east part of Chrysi, in Kedrodasos and in Gavdos-Sarakiniko where vegetation recovery is improbable or may take too long. ecologia mediterranea – Vol. 41 (2) – 2015 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) The processes affecting the dune systems of our study sites operate at different scales from local (e.g. anthropogenic disturbance) to catchment scale (e.g. land use changes) which affects sediment transfer and therefore habitat dynamics (Cassar & Stevens 2002). Gavdos and Chrysi islands belong to the first case where human impact is localized while Falasarna and Kedrodasos are additionally affected by agricultural and related road building activities at the catchment level. In Gavdos, grazing should be excluded from the priority habitat sites, while trampling should be controlled with the use of designated paths. Controlling of camping should be addressed by future management efforts while grazing should be restricted particularly in Falasarna and Kedrodasos. The proximity of the habitat in Falasarna and Kedrodasos to agricultural areas and grazing lands result in increased grazing pressure. In Chrysi, there are no grazing animals, while in Gavdos, grazing is not uniform across sites (i.e. intensive in Sarakiniko, localized in Ag Ioannis and absent in Lavrakas). Mitigation measures to address habitat pressures may include sand trapping, transplanted vegetation, fencing, dune walkovers and environmental education campaigns. Most of these actions were carried out as part of the JUNICOAST LIFE+ Project. While these activities at the site level are easier to implement, what still remains a challenge is controlling factors affecting the habitat at the catchment scale. Although a stakeholders group was established early on in the LIFE+ project the implementation of some of the actions is still faced by obstacles at the community level. This is yet again proven to be the major challenge which nature conservation faces in Europe, that is to persuade the public about the importance of Natura 2000 (Papageorgiou & Vogiatzakis 2006; Paloniemi et al. 2009). Acknowledgements This research was funded by the LIFE + Nature Programme “Actions for the conservation of coastal dunes with Juniperus spp. in Crete and the South Aegean (Greece)” (LIFE07NAT/GR/000296). We are grateful to Dr. Christina Fournaraki, curator at the Herbarium of MAICh, for her help in species identification. ecologia mediterranea – Vol. 41 (2) – 2015 References Acosta A., Carranza .M.L. & Izzi .C.F., 2009. Are there habitats that contribute best to plant species diversity in coastal dunes? Biodivers Conserv 18: 1087-1098. Acosta A., Ercole S., Stanisci A. & Blasi C., 2006. Sandy coastal ecosystems and effects of disturbance in Central Italy. JCoastal Res, SI 39: 985-989. Anon., 1992. Council Directive 92/43/EEC on the conservation of natural habitats and of wild fauna and flora. Official Journal European Commission 35. Barbier, E.B., Hacker, S.D., Kennedy, C., Koch, E.W., Stier, A.C. & Silliman B.R. 2011. The value of estuarine and coastal ecosystem services. Ecol Monogr 81: 169-193. Böhling N., Greuter W. & Raus T., 2002. Zeigerwerte der Gefäßpflanzen der Südägäis (Griechenland). Braun-Blanquetia. Review of Geobotanical Monographs 32: 1-108. Braun-Blanquet J., 1932. Plant sociology. McGrawHill, London. Brown AC &McLachlan A., 1990. Ecology of Sandy Shores. Elsevier, Amsterdam. Cassar L.F., & Stevens D.T., 2002. Coastal dunes under siege: a guide to conservation for environmental managers. International Environment Institute, University of Malta. Castillo S, Popma J, & Moreno-Casasola J. 1991. Coastal sand dune vegetation of Tabasco and Campeche, Mexico. J Veg. Sci 2: 73-88. Castillo JM, Rubio Casal AE, Luque CJ, Luque T, & Figueroa ME. 2002. Comparative field summer stress of three tree species co-occurring in Mediterranean coastal dunes. Photosynthetica 40: 49-56. Ciccarelli D., Bacaro G., & Chiarucci A. 2012. Coastline Dune Vegetation Dynamics: Evidence of No Stability. Folia Geobot. 47: 263-275. Chytrý M., Tichý L., Holt J., & Botta-Dukát Z. 2002. Determination of diagnostic species with statistical fidelity measures. J Veg. Sci 13: 79-90 Corre JJ. 1991. The sand dunes and their vegetation along the Mediterranean coast of France. Their likely response to climatic change. Landscape Ecol 6: 65-75 Council of the European Community, 1992. Council Directive 92/43 EEC of 21 May 1992 on the conservation of natural habitats and wild fauna and flora. Official Journal of the European Communities. Espejel I., Ahumada B., Cruz Y., & Heredia A., 2004. Coastal vegetation as indicators for conservation. In Coastal Dunes (pp. 297-318). Springer Berlin Heidelberg. Ellenberg, H. & D. Mueller-Dombois. 1967. A key to Raunkiaer plant life forms with revised subdivisions. Ber. Geobot. Inst. ETH Rübel, 37: 56-73. Forey, E., Chapelet, B., Vitasse, Y., Tilquin, M., Touzard, B. & Michalet, R. 2008. The relative importance of disturbance and environmental stress at local and regional scales in French coastal sand dunes. J Veg. Sci 19: 493-502. Garcia Novo F, Diaz Barradas MC, Zunzunegui M, Garcia Mora R, & Gallego Fernandez JB 2004. Plant Functional Types in Coastal Dune Habitats. Chapter Coastal Dunes, Springer, Series Ecological Studies 171: 155-169. Géhu JM, Costa M, & Biondi E. 1990. Les Junipereta macrocarpae sur sable. Acta Bot. Malacitana 15: 303- 309. 59 Pinelopi Delipetrou, Dany Ghosn, George Kazakis, Panagiotis Nyktas, Electra Remoundou, Ioannis N. Vogiatzakis Géhu JM. 1993. The coastal woodlands of Europe. In: Tekke R, Salman A. (eds.). Coastal Dune Woodlands along the Atlantic and North Sea Shores. EUCC Internal Report Series 4, Leiden, pp. 6-16. Géhu JM., Apostolidis N., Géhu-Franck J., & Arnold K. 1989. Premières données sur la végétation littorale des îles de Rodhos et de Karpathos. Colloques phytosociologiques XIX: 545-582. Halada L., Evans D., Romão C., & Petersen J. E., 2011. Which habitats of European importance depend on agricultural practices? Biodiver Conserv, 20: 23652378. Heslenfeld P., Jungerius P., & Klijn J., 2008. European coastal dunes: ecological values, threats, opportunities and policy development. In: Martínez, M. L.; Psuty, Norbert P. (eds.) Coastal Dunes, 335-351. Hesp P., 2002. Foredunes and blowouts: initiation, geomorphology and dynamics. Geomorphology 48(1): 245-268. Hill MO., 1979a. TWINSPAN - a FORTRAN Program for arranging Multivariate Data in Ordered Two way Table by Classification of the Individuals and the Attributes. Cornell University, Department of Ecology and Systematics, Ithaca, New York. Hill MO., 1979b. DECORANA - a FORTRAN Program for detrended correspondence analysis and reciprocal averaging. Cornell University, Department of Ecology and Systematics, Ithaca, New York. Jahn R. & Schönfelder P., 1995. Exkursionsflora für Kreta. Verlang Eugen Ulmer, Stuttgart (Hohenheim). Jongman R. H., ter Braak C. J. F., & Tongeren O. F. R., 1987. Data analysis in community and landscape ecology. Pudoc, Wageningen. 299 p. Lemauviel S., Gallet S., & Rozé F., 2003. Sustainable management of fixed dunes: example of a pilot site in Brittany (France). CRBiol 326: 183-191. Martinez ML & Psuty NP. (eds.), 2004. Coastal dunes: ecology and conservation. 386 pp. Berlin: SpringerVerlag. Ecological Studies 171. Maun M. A., 2009. The biology of coastal sand dunes. Oxford University Press. Mayer A., 1995. Comparative study of the coastal vegetation of Sardinia (Italy) and Crete (Greece) with respect to effects of human influence. IHW-Verlag, Munich. McCune B., & Mefford M. J., 1999. PC-ORD: multivariate analysis of ecological data; Version 4 for Windows; [User‘s Guide]. MjM software design. Miller TE., Gornish ES., Buckley HL., 2009. Climate and coastal vegetation: disturbance, recovery, and succession. Plant Ecol. 206: 97-104. Mueller-Dombois D. & Ellenberg H., 1974. Aims and methods of vegetation analysis. New York: Ed. J. Wiley & Sons, 547. Mucina L., 1997. Classification of vegetation: Past, present and future. J Veg Sci: 751-760. Musila WM., Kinyamario JI. & Jungerius PD., 2001. Vegetation dynamics of coastal sand dunes near Malindi, Kenya. Afr J Ecol 39: 170-177. Paloniemi R. & Vilja V., 2009. Changing ecological and cultural states and preferences of nature conservation policy: the case of nature values trade in SouthWestern Finland. J Rural Stud 25: 87-97. 60 Papageorgiou K. & Vogiatzakis I.N., 2006. Nature Protection in Greece: an appraisal of the factors shaping integrative conservation and policy effectiveness. Environ. SciPolicy 9: 476-486. Picchi S., 2008. Management of Natura 2000 habitats. 2250* Coastal dunes with Juniperus spp. European Commission. Pihl S., Ejrns R., Sgaard B., Aude E., Nielsen K.E., Dahl K., & Laursen J.S., 2001. Habitat and species covered by the EEC Habitats Directive – A Preliminary assessment of distribution and conservation status in Denmark. National Environmental Research Institute, Denmark. 121 pp. - NERI Technical Report no. 365. Available on: http://faglige-rapporter.dmu. dk Provoost S., Jones M. L. M., & Edmondson S. E., 2011. Changes in landscape and vegetation of coastal dunes in northwest Europe: a review. J Coastal Conserv 15: 207-226. Ranwell D. S., 1972. Ecology of Salt marshes and Sand dune. Chapman Hall, London. Botanische Jahrbucher Systematik, 75, 563-583. Rivas-Martínez S.,Díaz T.E., Fernández-González F., Izco J., Loidi J., Lousã M., & Penas A., 2002. Vascular plant communities of Spain and Portugal: Addenda to the syntaxonomical checklist of 2001. Itinera Geobotanica 15: 5-922. Santoro R., Jucker T., Prisco I., Carboni M., Battisti C., & Acosta A. T., 2012. Effects of trampling limitation on coastal dune plant communities. EnvironManage 49: 534-542. Sýkora K.V., Babalonas D. & Papastergiadou E.S., 2003. Strandline and sand-dune vegetation of coasts of Greece and of some other Aegean countries. Phytocoenologia 33: 409-446 ter Braak CJF., 1994. Canonical community ordination. Part 1. Basic theory and linear methods. Ecoscience 1: 127-140. ter Braak CJF. & Smilauer P., 1998. CANOCO Reference Manual and User’s Guide to Canoco for Windows: Software for Canonical Community Ordination (version 4). Microcomputer Power, Ithaca, New York. Thanos C.A., Kaltsis A., Koutsovoulou K., Skourti E., & Sarris D., 2010. Deliverable A.3: Report on the composition and structure of Juniperus macrocarpa subpopulations, Chania, Greece. The JUNICOAST project, Actions for the conservation of coastal dunes with Juniperus spp. in Crete and the South Aegean (www.junicoast.gr). Tichý L., 2002. JUICE, software for vegetation classification. JVeg Sci 13: 451-453. van der Meulen F. & Salman AHPM., 1996. Management of Mediterranean coastal dunes. Ocean Coast Manag 30: 177-195. Wikum D.A. & Shanholtzer G.F., 1978. Application of the Braun-Blanquet cover abundance scale for vegetation analysis in land development studies. Environ Manage 2: 323-329. Wolfe SA & Nickling WG., 1993. The protective role of sparse vegetation in wind erosion. ProgPhys Geog 17: 50-68. ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage Characterization of the Moroccan Alfa Steppe Functioning Using the Landscape Function Analysis Method Mchich DERAK1, Fernando T. MAESTRE2, José L. QUERO3, Victoria OCHOA3, Cristina ESCOLAR3, Santiago SOLIVERES3, Pablo GARCÍA-PALACIOS3 1. Direction Régionale des Eaux et Forêts et de la Lutte Contre la Désertification du Rif, Avenue Mohamed V, BP 722, 93000, Tétouan, Maroc. 2. Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Física y Química Inorgánica, Escuela Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, Calle Tulipán Sin Número, 28933 Móstoles, España. 3. Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Escuela Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, Calle Tulipán Sin Número, 28933 Móstoles, España. Auteur correspondant : [email protected] Received: 14 October, 2014; First decision: 9 January, 2015; Revised: 25 February, 2015; Second decision: 6 July, 2015; Revised: 13 July, 2015; Accepted: 1 September, 2015 Résumé Le suivi des processus écologiques est devenu un outil nécessaire dans les efforts de lutte contre la désertification. Les techniques de suivi disponibles sont généralement basées sur les indicateurs du fonctionnement de l’écosystème. Parmi ces techniques, la méthode de l’analyse fonctionnelle du paysage, connue sous le nom LFA (Landscape Function Analysis), permet d’évaluer le fonctionnement géochimique de l’écosystème de manière facile, précise, consistante et peu chère. Cette méthode n’a jamais été appliquée pour caractériser le fonctionnement des steppes d’alfa (Stipa tenacissima L.) marocaines. Dans la présente étude, nous avons appliqué la méthode LFA pour caractériser le fonctionnement des steppes d’alfa dans douze sites situés à l’est du Maroc. Nous avons testé la corrélation entre les indices LFA et les variables du fonctionnement du sol. Les résultats ont montré que les indices LFA peuvent être utilisés comme Mots clés : aménagement, alfa, Maroc, méthode LFA. ecologia mediterranea – Vol. 41 (2) – 2015 équivalents des variables du fonctionnement du sol comme le C organique, le N total, le pH et la salinité. Comparées aux steppes espagnoles, les steppes marocaines ont montré des niveaux fonctionnels bas justifiant des actions adéquates de conservation et de restauration écologique. L’application de la méthode LFA pour les steppes d’alfa est susceptible d’améliorer nos connaissances sur le fonctionnement de ces écosystèmes et d’aider les gestionnaires à mieux concevoir les plans d’aménagement y afférents. Abstract The monitoring of ecosystem processes has become an essential tool for combating desertification. Monitoring techniques are often based on indicators of ecosystem functioning. Among them, the Landscape Function Analysis (LFA) method allows the assessment of ecosystem geochemical functioning by means of an Keywords: Management, alfa, Morocco, LFA method. 61 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios easy, precise, consistent and cheap approach. This method has never been used to assess the functioning of the Moroccan alfa steppes (Stipa tenacissima L.). In the present study, we applied the LFA method to characterize the functioning of alfa steppes in 12 sites distributed throughout Eastern Morocco. We checked for correlation between the LFA indices and variables of soil functioning. Our results showed that LFA indices can be used as surrogates of soil properties related to soil functioning such as organic C, total N, pH and salinity. Moroccan steppes showed a low level of functionality compared to similar steppes in Spain, and require conservation and ecological restoration measures. The use of the LFA method in alfa steppes can be used to increase our understanding of their functioning, and may help managers to design more efficient management plans. Abridged English Version The monitoring of ecological processes has become a necessity for combating desertification (Herrick et al. 2005 ; Reynolds et al. 2007b). In the last decades, several monitoring techniques based on ecosystem functioning indicators have been developed. Among them, the Landscape Function Analysis (LFA) is based on the assessment of soil surface indicators, and provides useful information on ecosystem geochemical functioning. The method generates three numerical indices (Table 2) related to soil stability, water infiltration and nutrient cycling (Tongway 1995; Tongway & Hindley 2004). This simple, cheap and consistent technique has been widely applied for assessing the functioning of Spanish alfa (Stipa tenacissima L.) steppes (Maestre & Cortina 2004; 2006; Cortina et al. 2006; Maestre & Puche 2009; Mayor & Bautista 2012), but it has never been employed for the analysis of Moroccan alfa steppes. The main objective of this study is to assess the functioning of the Moroccan alfa steppes on the basis of the LFA indices. Our study was conducted in 12 semiarid sites in Eastern Morocco (Figures 1, 2; Table 1). We compared LFA indices values between bare soil and under alfa tussocks, and checked for correlation between the three LFA indices and variables associated to soil functioning (pH, salinity, organic C, Total N and available P). The values of LFA indices were lower in bare soil than under alfa tussocks (Figure 3, Table 4). This confirms the role of the spatial distribution of patches of vegetation in semiarid 62 ecosystems (Maestre et al., 2002). The LFA indices were correlated to the main soil functioning variables, except for available P (Table 5). This concurs with previous studies (Tongway & Hindley 2003; Ata Rezaei et al. 2006; Maestre & Puche 2009), and suggests that LFA indices can be used as surrogates of steppes functioning variables. The functional state of Moroccan steppes (expressed by LFA indices) was lower when compared with the state of Spanish steppes. This confirms previous field observations which had highlighted the advanced degradation level in Moroccan steppes (DREFLCD-O 2007). The LFA method represents a suitable tool for assessing the functional status of alfa steppes and thus, for improving their management. Moroccan steppes showed a low functioning level and may require urgent intervention to control erosive processes, restore nutrient cycling and also on controlling the human pressure on the landscape. Introduction La désertification est l’un des problèmes environnementaux les plus sérieux qui menacent la persistance des zones arides et leur capacité à fournir les biens et services écosystémiques (Évaluation des écosystèmes pour le millénaire : MEA 2005 ; Reynolds et al. 2007a). Le suivi des processus écologiques est devenu de plus en plus une composante nécessaire dans les plans de lutte contre la désertification en zones arides (Herrick et al. 2005 ; Reynolds et al. 2007b). Ce type de suivi permet d’évaluer l’état et l’évolution des processus écosystémiques et de prévenir des possibles risques de désertification, ce qui permettrait d’adopter des mesures de conservation et de gestion de plus en plus efficaces (Fernández et al. 2002). En outre, le suivi améliore nos connaissances sur la réponse de l’écosystème et du paysage aux actions de restauration et contribue par conséquent à la réduction de l’incertitude autour de ces actions (Navarro et al. 2009 ; Cortina et al. 2011). Dans les dernières décades, plusieurs techniques de suivi basées sur les indicateurs du fonctionnement de l’écosystème ont été mises en œuvre. Combinant les caractéristiques du sol et de la végétation, ces techniques permettent d’évaluer la résilience de l’écosystème face à l’érosion ainsi que sa capacité de ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage retenir et de recycler l’eau et les nutriments (Tongway 1995 ; Tongway & Hindley 2004 ; Maestre & Puche 2009). Parmi les techniques utilisées, on cite la méthode de l’analyse fonctionnelle du paysage, connue sous le nom méthode LFA (Landscape Function Analysis, Tongway 1995 ; Tongway & Hindley 2004). Cette méthode, basée sur des indicateurs mesurés au terrain, vise à évaluer le fonctionnement géochimique des écosystèmes à l’échelle du versant. Les écosystèmes ayant un état fonctionnel élevé ont tendance à conserver les ressources en sol, eau et nutriments. Par contre, ceux à état fonctionnel bas ont tendance à perdre les ressources existantes et à ne retenir qu’une faible fraction des précipitations incidentes (Tongway & Hindley 2004). La méthode LFA présente l’avantage d’être précise, consistante dans le temps et applicable à une large gamme d’écosystèmes. En outre, son application est simple, rapide et peu chère et elle peut être utilisée par les techniciens et les gestionnaires du terrain avec un effort moindre d’entraînement. La méthode LFA a été conçue et validée pour des écosystèmes semi-arides australiens (Tongway 1995 ; McR. Holm et al. 2002 ; Bartley et al. 2006), où elle a été fréquemment utilisée pour évaluer l’état fonctionnel du paysage et sa relation avec la structure de l’écosystème (p. ex. Ludwig et al. 2007 ; Kwock et al. 2010). Grâce à ses multiples caractéristiques et utilités, son usage s’est étendu à différentes régions du monde. À titre d’exemple, elle a été utilisée pour caractériser les pelouses en Iran (Ata Rezaei et al. 2006), pour étudier les effets des reboisements en Tunisie (Derbel et al. 2009 ; Jeddi et al. 2009), pour suivre le phénomène de désertification en Argentine (Oliva et al. 2010), et pour évaluer les effets des plantations d’Atriplex sur les fonctions du sol et du paysage au Maroc (Zucca et al. 2013), entre autres. L’application de la méthode LFA aux steppes de Stipa tenacissima L. a bien évolué durant ces dernières années en raison de l’étendue et de l’importance de ces écosystèmes à l’échelle du bassin méditerranéen et du risque permanent de dégradation d’origine humaine auquel ils sont soumis dans toute son aire de distribution (Cortina et al. 2009 ; Cortina et al. 2012). C’est en Espagne que cette méthode a été le plus appliquée pour évaluer la relation entre la structure et la fonction des steppes d’alfa (Maestre & Cortina 2004 ; 2006), pour étudier la relation entre la fonctionnalité du sol et la capacité de restauration de ces steppes ecologia mediterranea – Vol. 41 (2) – 2015 (Cortina et al. 2006), pour caractériser leur fonctionnement hydrologique à différentes échelles (Mayor & Bautista 2012), et pour étudier la relation entre les indices LFA et les différentes variables du sol qui caractérisent mieux le fonctionnement de l’écosystème (Maestre & Puche 2009). Au Maroc, l’un des pays d’Afrique du Nord où les steppes d’alfa s’étendent sur de très grandes superficies (Le Houérou 2001), l’étude du fonctionnement de ces écosystèmes est souvent réalisée à base de mesures qualitatives et d’opinions d’experts. Rares sont les études qui se sont focalisées sur la caractérisation des steppes d’alfa moyennant des techniques basées sur les indicateurs du fonctionnement de l’écosystème comme la méthode LFA. La présente étude a pour objectifs : (1) d’évaluer l’état fonctionnel des steppes marocaines à l’échelle du microsite moyennant la méthode LFA, (2) d’examiner la relation entre les indices LFA et les variables clés du cycle des nutriments des steppes d’alfa étudiées (pH, salinité, C organique, N total, P disponible), et (3) de proposer un cadre d’application de la méthode LFA dans les plans d’aménagements des steppes marocaines. Méthodes Zone d’étude Notre étude a porté sur 12 parcelles expérimentales situées à l’est du Maroc (figures 1 et 2). Le tableau 1 montre les valeurs des principales caractéristiques topographiques, climatiques, pédologiques et de végétation des parcelles échantillonnées. C’est dans cette région de l’Est que se concentre la majorité des 3,3 millions d’hectares de steppes marocaines (Benabid & Fennane 1999). L’alfa peut s’y trouver dans des ambiances semi-arides, arides ou sahariennes. L’aridité augmente en direction O-E et N-S. Le sol est généralement bien drainé et relativement rocheux (Benabid 2000). La principale vocation du terrain est le parcours suivi par l’agriculture. En général, la population locale souffre de difficultés économiques et exerce une forte pression sur les steppes qui se manifeste par le surpâturage, le labour et l’exploitation anarchique de l’alfa (Benabid 2000 ; El Rhazi 2003). L’industrie de l’alfa a connu un essor spectaculaire dans la période 1925-1975, mais à partir des années 1980, et à cause de la fermeture des usines 63 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios Figure 1 - L ocalisation des parcelles échantillonnées (étoiles) à l’est du Maroc. Les limites intérieures correspondent aux limites des régions marocaines. Les principales caractéristiques des 12 parcelles sont consignées dans le tableau 1. Figure 1 - Location of the sampling plots (stars) in eastern Morocco. Internal limits correspond to Moroccan regions’ limits. Main characteristics of the 12 plots are reported in table 1. Tableau 1 – V aleurs des principales caractéristiques des 12 parcelles échantillonnées. LAT = latitude (en décimal) ; LONG = longitude (en décimal) ; ALT = altitude (m) ; PENT = pente (°) ; TMA = température moyenne annuelle (°C) ; PMA = précipitation moyenne annuelle (mm) ; TSA = teneur en sable (%) ; TAR = teneur en argile (%) ; TLI = teneur en limons (%) ; CON = conductivité électrique (µS/cm) ; RS = richesse spécifique (N) ; CT = couvert total (%) ; pH = pH du sol ; COR = C organique (%) ; NTO = N total (%) ; PDI = P disponible (mg P g-1 sol). Table 1 – Main characteristic values of the 12 sampled plots. LAT = latitude (in decimal); LONG = longitude (in decimal); ALT = elevation (m); PENT = slope (°); TMA = annual mean temperature (°C); PMA = annual mean precipitation (mm); TSA = Sand content (%); TAR = clay content (%); TLI = silt content (%); CON = electrical conductivity (µS/cm); RS = specific richness (N); CT = total cover (%); pH = soil pH; COR = organic C (%); NTO = total N (%); PDI = available P (mg P g-1 soil). SITE LAT LONG 1 33.977 – 3.374 2 34.633 – 3.414 3 34.626 4 33.872 5 PENT TMA PMA TSA TAR TLI 776 18 16 265 57 8 36 756 20 16 339 56 7 38 – 3.465 922 17 15 385 43 16 41 – 3.634 1 001 7 15 307 66 4 29 33.933 – 3.559 726 11 16 289 73 5 6 34.442 – 3.593 1 028 14 15 399 54 7 34.473 – 3.639 867 4 16 401 8 34.431 – 2.193 946 6 15 321 9 34.310 – 1.999 1 144 7 14 10 34.159 – 2.373 1 002 5 11 33.052 – 2.423 1 339 12 33.068 – 2.729 1 427 64 ALT CON RS CT pH COR NTO PDI 99 7 17 8,3 0,7 0,088 0,005 114 16 28 8,5 1,4 0,156 0,005 108 11 11 8,3 2,1 0,182 0,005 99 10 33 8,4 1,8 0,157 0,005 22 67 9 26 8,5 0,6 0,068 0,004 5 41 120 14 46 8,3 1,6 0,167 0,006 43 5 52 121 7 37 8,4 0,8 0,113 0,003 53 6 41 90 11 37 8,3 2,2 0,207 0,007 377 52 5 44 127 18 19 8,4 2,1 0,269 0,010 15 294 59 4 37 86 11 23 8,3 2,1 0,216 0,006 2 15 283 82 4 15 75 7 28 8,5 0,5 0,043 0,008 6 15 310 68 3 29 90 13 18 8,4 0,7 0,071 0,005 ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage A B Figure 2 - Exemples des steppes d’alfa échantillonnés dans le site 4 (A) et le site 11 (B) à l’est du Maroc. Figure 2 – Examples of alfa steppes sampled in site 4 (A) and site 11 (B) in eastern Morocco. locales destinées à la confection de la fibre d’alfa et à l’arrêt de leurs exportations, l’alfa a perdu son rôle socio-économique localement. Actuellement, son exploitation se limite à l’usage quotidien comme combustible ou complément de fourrage, et à l’usage artisanal pour la confection des cordes, paniers, et autres ustensiles (Mhirit & Benchekroun 2006 ; El Rhazi 2003). Au cours des dernières années, le Haut-commissariat aux eaux et forêts et à la lutte contre la désertification marocaine a accordé un intérêt particulier à ces écosystèmes à travers l’adoption d’une série de mesures : réalisation d’un inventaire des ressources alfatières de la région de l’est du Maroc, élaboration des plans d’aménagement ecologia mediterranea – Vol. 41 (2) – 2015 et des steppes d’alfa pour la même région, et appui aux coopératives locales pour la restauration de l’usage traditionnel de l’alfa, entre autres (DREFLCD-O 2007). Estimation des indices du fonctionnement du paysage Les travaux de terrain ont été réalisés dans des parcelles de 30 × 30 m choisies de telle manière à couvrir une large gamme de conditions climatiques tout en réduisant la variabilité inter-site en termes de type de végétation, topographie et type du sol. Toutes les parcelles ont été établies dans des steppes d’alfa à 65 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios Tableau 2 – D escription et interprétation des caractéristiques de la superficie du sol utilisées pour le calcul des indices LFA. La valeur d’un indice LFA, exprimée en %, correspond à la somme des notes obtenues pour les différentes caractéristiques impliquées dans le calcul de cet indice (troisième colonne), divisée par la note maximale pouvant être attribuée à cet indice, soient 40, 57 et 43 respectivement pour les indices de stabilité, infiltration et recyclage des nutriments. Adaptation à partir de Tongway & Hindley (2004). Table 2 – Description and interpretation of soil surface characteristics used in the calculation of LFA indices. The value of an index, expressed in %, corresponds to the sum of the obtained marks for the different characteristics involved in the calculation of this index (third column), divided by the maximum mark assignable to this index, being 40, 57 and 43 for the stability, infiltration and nutrient cycling indices, respectively. Adaptation from Tongway & Hindley (2004). Caractéristique de la superficie du sol Description et interprétation Note maximale Couverture du sol Pourcentage de la couverture projetée de la végétation pérenne à une hauteur de 0,5 m, en addition aux roches et matériels boisés de plus de 2 et 1 cm de diamètre respectivement, ou tout autre objet longévive et immobile. Cet indicateur estime la vulnérabilité à la formation de la croûte physique. Couvert basal des herbacées pérennes et/ou couvert de la canopée des arbres et matorral. Cet indicateur permet d’évaluer la contribution de biomasse racinaire au processus du recyclage des nutriments. Couvert basal des herbacées pérennes et couvert de la canopée du matorral Couvert de litière et degré Quantité, origine et degré de décomposition de la litière. Estime la de décomposition disponibilité de la matière organique superficielle à la décomposition et au recyclage des nutriments. Couvert de la croûte Couvert des cryptogames (mousses, lichens et cyanobactéries) visibles biologique à la surface du sol. C’est un indicateur de la stabilité superficielle, de la résistance à l’érosion et de la disponibilité des nutriments. Degré de brisement Estime le degré de brisement de la croûte superficielle et la génération de la croûte du matériel édaphique vulnérable à l’érosion. Type et sévérité de l’érosion Évalue la nature et la sévérité du processus érosif actuel. Indice LFA concerné 5 Stabilité 4 Infiltration Recyclage des nutriments 30 Stabilité Infiltration Recyclage des nutriments Stabilité Recyclage des nutriments 4 4 Stabilité 4 Stabilité Matériel déposé Évalue la nature et la quantité des dépôts alluviaux. 4 Stabilité Rugosité de la surface du sol Évalue la rugosité de la surface du sol en fonction de sa capacité à capturer et à retenir les ressources mobiles telles que l’eau, les propagules, les sédiments et la matière organique. 5 Infiltration Recyclage des nutriments Résistance de la surface aux perturbations Test d’humectation Mesure la facilité avec laquelle le sol peut être mécaniquement altéré pour générer un matériel vulnérable à l’érosion hydrique ou éolienne. Évalue la stabilité des fragments naturels du sol à l’humectation rapide. 5 4 Stabilité Infiltration Stabilité, Infiltration Texture du sol Classifie la texture de la surface du sol. 4 Infiltration orientation SE-SO et en versant de pente inférieure à 30°. Pour chaque parcelle, la méthode LFA a été appliquée moyennant dix points quadrats de 50 × 50 cm distribués au hasard, cinq au sol nu et cinq sous les touffes d’alfa. Pour chaque point quadrat, nous avons évalué onze variables de la superficie du sol qui définissent son fonctionnement (tableau 2). L’évaluation a été réalisée sur une échelle semi-quantitative en suivant les indications de Tongway & Hindley (2004). En utilisant le modèle Excel développé par David Tongway (http://www.csiro.au/services/EcosystemFunctionAnalysis), les onze variables ont été combinées pour obtenir les trois indices LFA : stabilité, infiltration et recyclage des nutriments. L’indice de stabilité est lié à la résistance du sol à l’érosion et à sa capacité de récupération après perturbation. L’indice d’infiltration renseigne sur la faculté du sol à répartir l’eau de pluie en eau disponible pour les plantes et en eau de ruissellement 66 perdue par le système. L’indice de recyclage des nutriments estime l’efficience de recycler la matière organique et de la retourner au sol en nutriments assimilables par les plantes. Les trois indices se présentent sous forme de pourcentage et leur valeur est inversement proportionnelle à l’état de détérioration d’une fonction de l’écosystème (Tongway & Hindley 2004). Estimation des variables associées au fonctionnement du sol Pour caractériser le fonctionnement du sol, nous avons utilisé cinq variables : pH, salinité (estimée par la conductivité électrique), C organique, N total et P disponible. Ces variables sont fortement corrélées avec les processus de recyclage des nutriments (Forster 1995 ; Chapin III et al. 2002) et déterminent en grande mesure le fonctionnement des écosystèmes arides et semi-arides ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage (Whitford 2002). Ces variables ont été évaluées dans chacun des points quadrats de 50 × 50 cm utilisés pour l’estimation des indices LFA. Ainsi, pour chaque point quadrat, nous avons extrait approximativement 250 g du sol (0-7,5 cm de profondeur) en utilisant un tube cylindrique. Une fois extrait, le sol a été tamisé puis laissé à l’air libre pour sécher. Les analyses de C organique et P disponible ont été réalisées dans le laboratoire du département de biologie et géologie de l’université Rey Juan Carlos (Móstoles, Espagne), et celles du N total dans l’université de Jaén (Jaén, Espagne). Le C organique a été déterminé par colorimétrie après oxydation avec le mélange du bichromate de potassium et de l’acide sulfurique (Anderson 1993). Le N total a été obtenu à l’aide d’un analyseur CN (Leco CHN628 Series, Leco Corporation, St Joseph, MI, USA) et le contenu en P disponible a été déterminé selon une extraction 0,5 M NaHCO3 (pH : 8,5) (Olsen & Sommers 1982). Les valeurs des cinq variables obtenues pour chaque parcelle sont consignées dans le tableau 1. Tableau 3 – Statistiques descriptives des trois indices de LFA calculés à l’échelle des microsites. Table 3 – Descriptive statistics for the three LFA indices calculated at the microsite scale. Indice LFA (%) Moyenne ± ES (n = 12) Min. Max. Stabilité 63 ± 1,0 40 78 Infiltration 37 ± 1,5 12 60 Recyclage nutriments 29 ± 1,6 9 52 Analyse statistique Pour évaluer la signification statistique des différences entre les valeurs des indices LFA obtenus pour le sol nu et sous les touffes d’alfa, nous avons procédé à une analyse de variance (ANOVA) à deux facteurs, en considérant la parcelle comme facteur aléatoire et le type du microsite (sol nu et sous touffes d’alfa) comme facteur fixe. Les cinq variables associées au fonctionnement du sol ont montré une distribution non normale (Test de Kolmogorov-Smirnov, p < 0,05), et par conséquent leur corrélation avec les indices LFA a été mesurée à l’aide du coefficient de corrélation de Spearman. Résultats Les valeurs moyennes des indices LFA calculées à l’échelle du microsite ont été de 63, 37 et 29 respectivement pour les indices de stabilité, infiltration, et recyclage de nutriment (tableau 3). En comparant les valeurs des indices LFA obtenues pour les deux types de microsites, on remarque que les sols nus ont montré des valeurs significativement inférieures à celles en dessous des touffes d’alfa (figure 3). La différence était bien saillante ecologia mediterranea – Vol. 41 (2) – 2015 Figure 3 - Représentation des trois indices LFA en % obtenus à l’échelle des microsites. Figure 3 – Representation of the three microsite LFA indices in %. 67 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios Tableau 4 - R ésultats de l’ANOVA à deux critères pour les trois indices de LFA comparés entre sol nu et sous les touffes d’Alfa. Table 4 – Two ways ANOVA results for the three LFA indices compared between bare-ground areas and Alfa tussocks. Indice LFA (%) Stabilité Infiltration Facteur Microsite F p-value 424,278 < 0,0001 Parcelle 5,926 0,003 Microsite* Parcelle 3,041 0,002 315,883 < 0,0001 0,604 0,792 Microsite* Parcelle 13,309 < 0,0001 Microsite 554,479 < 0,0001 Parcelle 0,300 0,971 Microsite* Parcelle 13,274 < 0,0001 Microsite Parcelle Recyclage nutriments Statistiques Tableau 5 – C oefficients de corrélation de Spearman à l’échelle de microsites entre les trois indices LFA et variables du sol évaluées. Les valeurs de p et de n sont montrées au-dessous de chaque coefficient. Les corrélations significatives sont montrées en caractère gras. pH = pH du sol ; Corg = carbone organique (%) ; Ntot = nitrogène total (%) et Pdis = phosphate disponible (mg P g-1 sol). Table 5 – Spearman correlation coefficients obtained at the microsite scale between the three LFA indices and the assessed soil variables. The p values and n are showed below each coefficient. Significant correlations are highlighted with bold character. pH = soil pH; Corg = organic carbon (%); Ntot = total nitrogen (%) et Pdis = available phosphate (mg P g-1 soil). Indices LFA Stabilité Infiltration Recyclage nutriment – 0,252 0,005 120 – 0,261 0,004 120 – 0,258 0,004 120 Cond 0,271 0,003 118 0,328 < 0,001 118 0,361 < 0,001 118 Corg 0,265 0,003 120 0,350 < 0,001 120 0,326 < 0,001 120 Ntot 0,201 0,028 120 0,216 0,018 120 0,250 0,006 120 Pdis – 0,097 0,290 120 0,063 0,492 120 – 0,018 0,846 120 Variables du sol pH surtout pour les indices d’infiltration et du recyclage des nutriments. Cette différence s’est montrée significative en comparant tous les microsites étudiés entre eux (facteur Microsite) et aussi en comparant les microsites à l’intérieur de chaque parcelle (facteur Microsite*Parcelle) (tableau 4). 68 Les trois indices LFA ont montré une certaine relation avec les variables du fonctionnement du sol. Ils se sont montrés négativement corrélés avec le pH, positivement corrélés avec la salinité, le C organique et le N total, et sans corrélation avec le P disponible (tableau 5). En analysant ces relations pour chacun des deux types de microsites, nous remarquons que les corrélations ont changé de force et de direction, en devenant non significatives pour le pH, faibles pour la salinité, le C organique et le N total et relativement significatives avec le P disponible (tableau 6). Discussion Les sols nus ont montré des niveaux fonctionnels inférieurs à ceux en dessous des touffes d’alfa. Cette différence était significative en comparant les deux types de microsites dans des conditions topographiques, lithologiques et de végétation hétérogènes (toutes les parcelles confondues) et aussi en effectuant la même comparaison pour des conditions homogènes (à l’intérieur de chaque parcelle). Ce résultat concorde avec les résultats d’autres études menées sur les steppes d’alfa en Espagne (Maestre & Puche 2009 ; Mayor & Bautista 2012) et en Tunisie (Derbel et al. 2009). Cette différence est due essentiellement au fait que le sol nu agit comme source des ressources en eaux et sédiments, et les touffes d’alfa comme collecteur de ces ressources (Puigdefábregas et al. 1999). Ceci confirme l’importance de la distribution spatiale en tâches de la végétation dans les écosystèmes semi-arides (Maestre et al. 2002). Les trois indices LFA se sont montrés corrélés aux variables du fonctionnement du sol étudiées, sauf au P disponible. Ceci confirme les résultats des études réalisées dans des milieux semi-arides de l’Australie (Tongway & Hindley 2003), de l’Iran (Ata Rezaei et al. 2006) et de l’Espagne (Maestre & Puche 2009), et montre que les indices LFA peuvent être utilisés comme équivalents aux variables estimatives directes du fonctionnement des steppes d’alfa des zones semi-arides. Ces indices renseignent sur les principaux processus fonctionnels de l’écosystème en relation avec le recyclage des nutriments, la fertilité du sol, l’activité microbienne, etc. Cependant, la capacité des indices LFA à caractériser le fonctionnement des steppes est plus évidente quand on considère une gamme ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage Tableau 6 – C oefficients de corrélation de Spearman à l’échelle du sol nu et sous les touffes d’alfa entre les trois indices LFA et les variables du sol évaluées. Les valeurs de p et de n sont montrées audessous de chaque coefficient. Les corrélations significatives sont montrées en caractère gras. Pour les abréviations, voir tableau 5. Table 6 – Spearman correlation coefficients obtained at the bare soil and under alfa between the three LFA indices and the assessed soil variables. The p values and n are showed below each coefficient. Significant correlations are highlighted with bold character. For abbreviations, see table 5. Variables du sol Sol nu Sous alfa Indices LFA Indices LFA Stabilité Infiltration Recyclage nutriment Stabilité Infiltration Recyclage nutriment pH – 0,092 0,484 60 – 0,009 0,948 60 – 0,090 0,496 60 – 0,108 0,412 60 – 0,088 0,501 60 0,011 0,933 60 Cond – 0,043 0,746 59 – 0,194 0,141 59 0,003 0,984 59 0,064 0,628 59 0,352 < 0,006 59 0,259 < 0,047 59 Corg – 0,072 0,583 60 – 0,067 0,610 60 0,025 0,852 60 0,040 0,764 60 0,292 0,024 60 0,104 0,429 60 Ntot 0,017 0,898 60 – 0,149 0,256 60 0,089 0,501 60 0,165 0,209 60 0,288 0,026 60 0,162 0,217 60 Pdis – 0,336 0,009 60 0,262 0,043 60 0,099 0,453 60 – 0,064 0,627 60 – 0,118 0,369 60 – 0,383 0,003 60 variée et hétérogène des microsites. En effet, quand les microsites des sols nus et des touffes d’alfa sont considérés séparément, on obtient certains changements dans la signification des corrélations entre les indices LFA et les variables du fonctionnement du sol. Des changements similaires ont été observés par Mayor (2008) et Maestre & Puche (2009), ce qui montre que les indices LFA permettent de détecter des variations fonctionnelles entre des microsites de différents types et microconditions mais pas autant à l’intérieur de chaque type de microsites. La méthode LFA a été amplement utilisée et validée pour caractériser le fonctionnement des steppes d’alfa en Espagne. En considérant comme références les valeurs des indices LFA obtenues pour les steppes espagnoles, nous pouvons approcher l’état du fonctionnement des steppes marocaines étudiées. De cette manière, en comparant les valeurs moyennes des indices de stabilité, infiltration et recyclage des nutriments obtenues à l’échelle du microsite pour les steppes marocaines (63, 37 et 29, respectivement, tableau 3) et celles obtenues par Maestre & Puche (2009) pour les steppes espagnoles (70, 44 et 35, respectivement), nous pouvons considérer que les steppes d’alfa marocaines se caractérisent par ecologia mediterranea – Vol. 41 (2) – 2015 un état fonctionnel bas. Ceci est probablement dû aux conditions climatiques qui sont plus sévères au Maroc qu’en Espagne (tableau 7) et qui constituent un facteur limitant au fonctionnement des steppes marocaines (Le Houérou 1992, 1995 ; El Rhazi, 2003) et aussi à la grande pression humaine à laquelle ces dernières sont soumises continuellement. L’état fonctionnel bas constaté pour les steppes marocaines confirme les observations faites sur le terrain par les chercheurs et gestionnaires marocains qui font état d’un état de Tableau 7 – Moyennes ± ES des principales caractéristiques climatiques des 12 sites d’alfa marocaines échantillonnés dans cette étude et les steppes d’alfa espagnoles étudiées par Maestre & Puche (2009). Table 7 – A verages ± SE of main climatic characteristics of the 12 Moroccan alfa sites sampled in this study and of Spanish alfa steppes studied by Maestre & Puche (2009). Pays TMA PMA IA Maroc 15,1 ± 0,2 330,8 ± 13,9 0,26 ± 0,01 Espagne 14,11 ± 0,2 418,4 ± 5,4 0,36 ± 0,01 TMA = température moyenne annuelle (°C) ; PMA = précipitation moyenne annuelle (mm) ; IA : indice d’aridité. TMA = annual mean temperature (°C); PMA = annual mean precipitation (mm); IA: aridity index. 69 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios dégradation avancé de ces steppes justifiant des mesures urgentes pour leur conservation et restauration écologique (DREFLCD-O 2007). En raison de l’intérêt que présente la méthode LFA, nous pensons que son application peut être de grande utilité dans les plans d’aménagement et de gestion des steppes d’alfa marocaines. Dans ces plans, la tendance générale est d’établir des parcelles d’échantillonnage où l’on collecte des données quantitatives en relation avec les caractéristiques de la végétation comme la couverture, la biomasse et l’état de régénération, tout en mesurant certains paramètres des touffes comme la hauteur, le diamètre, la densité et le poids. Dans ce sens, nous recommandons de compléter ces données par l’estimation des indices LFA dans les mêmes parcelles afin de mieux comprendre l’état du fonctionnement des steppes et d’obtenir une information pertinente pour suivre les processus de dégradation de l’écosystème avec le temps. Cette information peut être très utile pour définir des mesures de gestion et de restauration encore plus précises et justifiées et pour établir l’ordre de priorité aux zones objet d’intervention. Dans de futures études, et afin de mieux comprendre le fonctionnement des steppes marocaines, nous suggérons d’établir un nombre suffisamment grand de parcelles d’échantillonnage reflétant la grande variabilité des conditions écologiques et socio-économiques existantes. En outre, nous suggérons que, dans les mêmes parcelles d’échantillonnage, les mesures portent simultanément sur les facteurs abiotiques, la structure et la composition de la végétation, le fonctionnement de l’écosystème (à travers les indices LFA), et la pression humaine, et ce dans un contexte de changements globaux qui pèsent sur les écosystèmes terrestres. Conclusion La méthode LFA s’est montrée comme un outil adéquat pour caractériser le fonctionnement des steppes d’alfa en milieux arides. L’intégration de cette technique dans les plans d’aménagement des steppes d’alfa marocaines est susceptible d’améliorer leur mode de gestion. En comparaison avec les steppes espagnoles, les steppes marocaines se caractérisent par un état fonctionnel plus 70 bas, et nécessitent des interventions urgentes qui doivent être focalisées sur la stabilité du sol, les processus érosifs, le recyclage des nutriments et sur le contrôle des pressions humaines sur le milieu naturel. Remerciements Cette recherche a été financée par le Conseil Européen de la Recherche à travers le Septième Programme-Cadre de la Communauté Européenne (FP7/2007-2013)/ERC Grant agreement 242658 (BIOCOM). Nous adressons nos vifs remerciements au Haut-commissariat aux eaux et forêts et la lutte contre la désertification du Maroc et à ses Directions Régionales du nord-est à Taza et de l’est à Oujda pour leur inestimable appui logistique et technique, et à MM. Mokaram Jabir, Kanafi Aziz, Mouflih Boujemàa et Cortina Jordi pour leur précieuse collaboration dans cette étude. Références bibliographiques Anderson J.M. & Ingramm J.S.I., 1993. Tropical Soil Biology and Fertility: A Handbook of Methods. Wallingford, UK: CABI. Ata Rezaei S., Arzani H. & Tongway D., 2006. Assessing rangeland capability in Iran using landscape function indices based on soil surface attributes. J. Arid. Environ. 65: 460-473. Bartley R., Roth C.H., Ludwig J.A., McJannet D., Liedloff A., Corfield J., Hawdon A. & Abbott B., 2006. Runoff and erosion from Australia’s tropical semiarid rangelands: influence of ground cover for differing space and time scales. Hidrol. Process. 20: 3317-3333. Benabid A. & Fennane M., 1999. Principales formations forestières au Maroc. In: Le Grand Livre de la forêt marocaine, Mardaga, Sprimont. Benabid A., 2000. Flore et écosystèmes du Maroc : Évaluation et préservation de la biodiversité. Ibis Press, Paris. Chapin III F.S., Matson P. & Money H.A., 2002. Principles of Terrestrial Ecosystem Ecology. Springer, New York. Cortina J., Maestre F.T., Vallejo R., Baeza M.J., Valdecantos A. & Pérez-Devesa M., 2006. Ecosystem structure, function, and restoration success: are they related?, J. Nat. Conserv. 14: 152-160. Cortina J., Maestre F.T. & Ramírez D., 2009. Innovations in semiarid restoration. The case of Stipa tenacissima L. grass steppes. In: Bautista S, Aronson J, Vallejo R (eds.), Land Restoration to Combat Desertification: Innovative Approaches, Quality Control and Project Evaluation. Fundación CEAM, Valencia. Cortina J., Amat B., Castillo V., Fuentes D., Maestre F.T., Padilla F. & Rojo L., 2011. The restoration of ecologia mediterranea – Vol. 41 (2) – 2015 Caractérisation du fonctionnement des steppes d’alfa marocaines par la méthode de l’analyse fonctionnelle du paysage vegetation cover in the Iberian southeast. J. Arid. Environ. 75: 1377-1384. Cortina J., Ruiz-Mirazo J., Amat B., Amghar F., Bautista S., Chirino E., Derak M., Fuentes D., Maestre F.T., Valdecantos A. & Vilagrosa A., 2012. Bases para la restauración ecológica de espartales. Gland, Suiza y Málaga, España, UICN. Derbel S., Cortina J. & Chaieb M., 2009. Acacia saligna plantation impact on soil surface properties and vascular plant species composition in central Tunisia. Arid Land Res. Manag. 23: 28-46. DREFLCD-O : Direction régionale des eaux et forêts et de la lutte contre la désertification de l’Oriental, 2007. Étude de la nappe alfatière des communes rurales de Maatarka, Tendrara et Béni Guil. Renseignements Généraux. Rapport interne, 35 p. El Rhazi M., 2003. Phytoécologie, production et valorisation de l’alfa (Stipa tenacissima) de la région de l’Oriental. Thèse de doctorat, université d’Oujda. Fernández R.J., Archer E.R.M., Ash A.J., Dowlatabadi H., Hiernaux P.H.Y., Reynolds J.F., Vogel C.H., Walker B.H. & Wiegand T., 2002. Degradation and recovery in socio-ecological systems: a view from the household/farm level. In: Reynolds J.F., Stafford Smith M. (eds.), Do Humans Create Deserts? The Ecological, Meteorological, and Human Dimensions of Global Desertification. Dahlem University Press, Berlin. Forster J.C., 1995. Soil physical analysis. In: Alef K, Nannipieri P (eds.), Methods in Applied Soil Microbiology and Biochemistry. Academic Press Inc, San Diego. Herrick J.E., Van Zee J.W., Havstad K.M. & Whitford W.G., 2005. Monitoring Manual for Grassland, Shrubland, and Savanna Ecosystems. Volume II: Design Supplementary Methods and Interpretation. USDA-ARS, Las Cruces. Jeddi K., Cortina J. & Chaieb M., 2009. Acacia salicina, Pinus halepensis and Eucalyptus occidentalis improve soil surface conditions in arid southern Tunisia. J. Arid. Environ. 73: 1005-1013. Kwok A.B.C., Eldridge D.J. & Oliver I.A.N., 2010. Do landscape health indices reflect arthropod biodiversity status in the eucalypt woodlands of eastern Australia ? Aust. Ecol. 36: 800-813. Le Houérou H.N., 1992. Relations entre la variabilité des précipitations et celles des productions primaire et secondaire en zone arides. Dans Le Floc’h E, Gouzis A, Cornet A, Bille JC, (eds.), L’aridité, une contrainte au développement. ORSTOM, Paris. Le Houérou H.N., 1995. Bioclimatologie et Biogéographie des steppes arides du Nord de l’Afrique. Diversité biologique, développement durable et désertisation. Dans Le Houérou HN (ed.), Bioclimatologie et biogéographie des steppes arides du Nord de l’Afrique : diversité biologique, développement durable et désertisation. CIHEAM, Montpellier. Le Houérou H.N., 2001. Biogeography of the arid steppe land north of the Sahara. J. Arid. Environ. 48: 103128. Ludwig J.A., Bastin G.N. & Wallace J.F., Mcvicar TR, 2007. Assessing landscape health by scaling with remote sensing : when is it not enough ? Landscape Ecol. 22: 163-169. Maestre F.T., Huesca M.T., Zaady E., Bautista S. & Cortina J., 2002. Infiltration, penetration resistance and microphytic crust composition in contrasted microsites within a Mediterranean semi-arid steppe. Soil Biol. Biochem. 34: 895-898. ecologia mediterranea – Vol. 41 (2) – 2015 Maestre F.T. & Cortina J., 2004. Insights into ecosystem composition and function in a sequence of degraded semiarid steppes. Restor. Ecol. 12: 494-502. Maestre F.T. & Cortina J., 2006. Ecosystem structure and soil surface conditions drive the variability in foliar d13C and d15N of Stipa tenacissima in semiarid Mediterranean steppes. Ecol. Res. 21: 44-53. Maestre F.T. & Puche M.D., 2009. Indices based on surface indicators predict soil functioning in Mediterranean semi-arid steppes. Appl. Soil. Ecol. 41: 342-350. Mayor A.G., 2008. El papel de la dinámica fuente-sumidero en la respuesta hidrológica, a varias escalas, de una zona mediterránea semiárida. Ph.D. Thesis, University of Alicante. Mayor A.G. & Bautista S., 2012. Multi-scale evaluation of soil functional indicators for the assessment of water and soil retention in Mediterranean semiarid landscapes. Ecol. Indic. 20: 332-336. McR. Holm A., Bennet L.T., Loneragan W.A. & Adams M.A., 2002. Relationships between empirical and nominal indices of landscape function in the arid shrubland of Western Australia. J. Arid. Environ. 50: 1-21. MEA: Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-Being: Current State and Trends. Island Press, Washington DC. Mhirit O. & Benchekroun F., 2006. Les écosystèmes forestiers et périforestiers : situation, enjeux et perspectives pour 2025. Disponible sur : www.rdh50. ma/fr/pdf/contributions/GT8-7.pdf. Navarro R.M., Guzman J.R., Herrera R., Lara P.A., Torres M., Ceacero C., Del Campo A. & Bautista S., 2009. Monitoring Guidelines for the Implementation of Forest Restoration Projects in Mediterranean Regions. In: Bautista S, Aronson J, Vallejo R (eds.), Land Restoration to Combat Desertification: Innovative Approaches, Quality Control and Project Evaluation. Fundación CEAM, Valencia. Oliva G., Gaitán J., Bran D., Nakamatsu V., Salomone J., Buono G., Escobar J., Frank F., Ferrante D., Humano G., Ciari G., Suarez D., Opazo W. & Adema E., 2010. Manual para la instalación y lecturas de Monitores MARAS (Monitoreo Ambiental para Regiones Áridas y Semiáridas). INTA, Proyecto PNUD GEF 07/G35. Olsen S.R., Sommers L.E., 1982. Phosphorus. In: Page A.L., Miller R.H., Keeney D.R. eds. Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. Madison, WI, American Society of Agronomy and Soil Science Society of America. Puigdefábregas J., Solé-Benet A., Gutiérrez L., Del Barrio G. & Boer M., 1999. Scales and processes of water and sediment redistribution in drylands: results from the Rambla Honda field site in Southeast Spain. Earth-Sci. Rev. 48: 39-70. Reynolds J.F., Stafford Smith D.M., Lambin E.F., Turner II B.L., Mortimore M., Batterbury S.P.J., Downing T.E., Dowlatabadi H., Fernández R.J., Herrick J.E., Huber-Sannvald E., Leemans R., Lynam T., Maestre F.T., Ayarza M. & Walker B., 2007a. Global desertification: building a science for dryland development. Science 316: 847-851. Reynolds J.F., Maestre F.T., Kemp P.R., Stafford-Smith D.M. & Lambin E., 2007b. Natural and human dimensions of land degradation in drylands: causes and consequences. In: Canadell J., Pataki D., Pitelka L.F. (eds.), Terrestrial Ecosystems in a Changing World. Springer-Verlag, Berlin. 71 Mchich Derak, Fernando T. Maestre, José L. Quero, Victoria Ochoa, Cristina Escolar, Santiago Soliveres, Pablo García-Palacios Tongway D.J., 1995. Monitoring soil productive potential. Environ. Monit. Assess. 37: 303-318. Tongway D.J. & Hindley N., 2003. Indicators of ecosystem rehabilitation success. Unpublished report. Available from http://www.cse.csiro. au/publications/2003/IndicatorsOfMinesiteRehabilitationSuccessStage2-100703.pdf. Tongway D.J. & Hindley N., 2004. Landscape Function Analysis: procedures for monitoring and assessing landscapes. CSIRO Sustainable Ecosystems, Camberra. 72 Whitford W.G., 2002. Ecology of Desert Systems. Academic Press, London. Zucca C., Pulido-Fernández M, Fava F., Dessena L. & Mulas M., 2013. Effects of restoration actions on soil and landscape functions: Atriplex nummularia L. plantations in Ouled Dlim (Central Morocco). Soil Til. Res.133: 101-110. ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) Organisation de la diversité floristique dans la forêt de Beni-Haoua (Chlef, Algérie) Adda ABABOU1,*, Mohammed CHOUIEB2, Abdelkader BOUTHIBA3, Djamel SAIDI1, Khalladi MEDERBAL4 1. Department of Biology, Faculty of Sciences, University Hassiba Ben Bouali, Chlef, Algeria. 2. Department of Agronomy, Faculty of Sciences and Engineering, University Abd El Hamid Ibn Badis, Mostaganem, Algeria. 3. Departement of Hydraulic, Institute of Agronomical Sciences, University Hassiba Ben Bouali, Chlef, Algeria. 4. Department of Biology, Faculty of natural and life sciences, University Ibn Khaldoun Tiaret, Algeria. * Corresponding author: [email protected] Received: 9 January, 2015; First decision: 11 February, 2015; Revised: 28 February, 2015; Second decision: 7 April, 2015; Revised: 13 April, 2015; Accepted: 8 May, 2015 Abstract Résumé The purpose of this study was to provide an inventory and an analysis of plant species occupying Beni-Haoua, a mountainous coastal ecosystem, with a rarely studied well-developed forest. As a result, 87 species were recorded in 7 sites, the Jaccard classification resulted in 4 groups of sites with significantly different diversities. According to the ϕ-coefficient of association, which can be used as a measure of fidelity, among the 87 species, 34 diagnostic species were distributed over four plant communities, with a fidelity value ranging from 55 to 100%, 28 differential species, among which 16 species were common to 2 plant communities and 12 common to 3 plant communities. The redundancy analysis (RDA) showed that among the studied environmental variables, altitude and pH were the most important ones. Indeed, according to the detrended correspondence analysis (DCA), plant species occurrence and distribution in the study area were affected by a strong altitudinal gradient. La région de Beni-Haoua est un écosystème littoral, montagneux, avec un patrimoine forestier très développé, qui n’a fait l’objet que de très rares études. Le principal objectif de cette étude est l’inventaire des espèces végétales ainsi que l’identification des principaux groupements végétaux que comporte cet écosystème par le biais des méthodes statistiques comme le coefficient de fidélité ϕ et l’analyse canonique des redondances (RDA). Ainsi, le coefficient de Jaccard a permis l’identification de 4 groupes de relevés (sites) avec des diversités floristiques significativement différentes. Selon, le coefficient de fidélité ϕ, parmi les 87 espèces recensées, 34 espèces sont réparties sur 4 unités de végétations avec un degré de fidélité allant de 55 % à 100 % et considérées comme espèces caractéristiques, 28 espèces différentielles parmi lesquelles 16 espèces communes à 2 unités de végétation et 12 espèces communes à 3 unités de végétations. La RDA a montré une très forte corrélation entre les espèces végétales et les paramètres environnementaux : l’altitude et le pH sont les paramètres environnementaux les plus importants conditionnant la distribution de la végétation. En effet la DCA a montré que l’occurrence et la distribution de la végétation dans la région de Beni-Haoua sont influencées par un fort gradient altitudinal. Keywords: Beni-Haoua, RDA, Phi coefficient, plant communities, diversity. Mots clés : Beni-Haoua, RDA, coefficient Phi, communautés végétales, diversité. ecologia mediterranea – Vol. 41 (2) – 2015 73 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Introduction Plant communities are characterized by a definite floristic composition, presenting physiognomic and structural uniformity (Flahault & Schroter 1910), with a highly complex and variable composition throughout ecosystems (Tozer 2003, Buckley et al. 2004, Sherman et al. 2008). Since, ecosystems are often described by their plant communities and environmental conditions (Wisheu & Keddy 1989, Kingston & Waldren 2003), the environmental control model is considered to be the primary factor affecting species occurrence and communities structure and composition (Collins et al. 1989, Cornwell & Ackerly 2009, Flinn et al. 2010). In this context, understanding the occurrence and the spatial patterns of species in connection with environmental conditions, is a major goal in ecological studies (Legendre 1993). The achievement of this goal relies largely upon the right choice of statistical methods, including classification methods, statistical fidelity measures, modelling methods (McCullagh & Nelder 1989) and multivariate analyses dedicated to the analysis of data sets with more than one variable. Among a large set of multivariate techniques, constrained ordination analysis has been widely used by ecologists to predict vegetation composition and response to environmental conditions (Collins et al. 1989, Zhu et al. 2005, Ababou et al. 2009, 2010), and as mentioned by Ramette (2007) plant studies rank first with insect studies for their use of constrained ordination. The aim of this study was to investigate the area of Beni-Haoua, a coastal ecosystem with a well-developed forest, rarely surveyed for its flora, in order to: (1) inventory the existing plant species within this ecosystem; (2) identify for each species the most influential environmental gradient; (3) identify the plant communities and species showing high fidelity to each community; (4) investigate the environmental gradients influencing these communities. Material and methods Study area Covering approximately 400 km2, the study area is a mountainous, coastal ecosystem with a highly-developed forest, located in the eastern part of Dahra mountains and comprised between 36°19’35”- 36°33’16” N of latitude and 1°22’50”-1°41’59” E of longitude (Figure 1). The area is characterized by a very rugged relief, steep slopes and an average height of 500 meters, the highest point stands over 1,000 m at Bissa Mountain. It is a Figure 1 – Location of the study area. 74 ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) typical Mediterranean area in terms of landscape structure, composition and climate, distinguished by hot, dry summers and relatively rainy winters with an annual rainfall ranging from 400 to 600 mm, an average temperature of 6 °C during winter and 30 °C during summer. In terms of vegetation, the landscape is covered with natural sclerophyllous vegetation alternating with bare soils, dominated in the maritime face by Pinus halepensis and Pinus pinaster, but above a certain altitude Quercus suber and Quercus ilex are the most significant species, whereas, Juniperus oxycedrus and Pistacia lentiscus are less imposing. Within this area, Bissa is the densest forest, including Quercus suber, Quercus ilex, Pinus halepensis, holly and wild fruit trees, such as Prunus avium, Arbutus unedo and Rubus fruticosus, whereas the mattoral part of this forest is dominated by Calluna vulgaris, Cistus monspeliensis and Lavandula dentata within a very dense Pistacia lentiscus cover. Soil and vegetation sampling The vegetation inventory was carried out during spring 2013 (March - May) in 7 sites (Bissa, Briera, Kadaian, Bouriach, Boucheral, Souhalia and Souamer) by exploring most of the area accessible in each site and considering only the presence/absence. A total of 87 species was recorded (Appendix), among which 25 species common to all sites were excluded from the analysis; species identification was done based on Quezel & Santa (1962, 1963) and botanical experts. Also, in order to study the effects of environmental variables on species composition and occurrence, for each of the 7 sites, we recorded the altitude (using a topographical map and a GPS), as well as the electrical conductivity (EC) (conductivity meter), the percentage of organic matter (OM) (Walkley and Black method, 1934) and the pH (pH meter) of the soil. Data analysis Initially, a co-linearity test was performed between environmental variables; as a result the variance inflation factors (Montgomery & Runger 2006, O’brien 2007) were less than 10 for the 4 selected variables, indicating no significant correlation. In order to minimize data-normality problems, the Shapiro-Wilk’s normality test showed that, with the exception of altitude, which was log-transformed, the ecologia mediterranea – Vol. 41 (2) – 2015 remaining variables were normally distributed. The most important floristic gradients were assessed using a Detrended Correspondence Analysis (DCA) (Hill & Gauch 1980) which is an ordination technique that reveals independent gradients with maximum species dispersion (Schmidtlein & Sassin 2004). In order to relate the inventoried vegetation to the measured environmental variables we opted for direct gradient analysis (constrained ordination), where species are directly related to measured environmental factors. In this context, the two most commonly used constrained ordination are either the Redundancy Analysis (RDA) (ter Braak 1994, Zuur et al. 2007) or Canonical Correspondence Analysis (CCA) (ter Braak 1986), to that end, DCA showed a gradient length almost equal to 3 (3.38) suggesting that RDA is more appropriate than CCA (Jongman et al. 1996). However, RDA are strongly affected by double zeros (Zuur et al. 2007), thus, an alternative is to apply the Hellinger transformation (Rao, 1995); according to Legendre & Gallagher (2001) this approach is less sensitive to double zeros. In connection with RDA, a linear regression (Zar 1999, Montgomery 2007) was also applied to evaluate the most effective environmental variable on each individual species. To cluster the plant species listed in the study area the Jaccard coefficient of similarity (Magurran 1988) was firstly applied, followed by the ϕ-coefficient of association. This coefficient is a statistical measure of association which can be used as a measure of fidelity; it can be calculated as follows (Bruelheide 2000): ϕ Where, N = total number of sites in the data set Np = number of sites in a particular group of sites obtained through Jaccard classification; n = number of occurrences of the species in the data set; np = number of occurrences of the species in a particular group of sites. Finally, we performed a Shannon-Wiener diversity index (Shannon index) followed by a t-test to evaluate biodiversity among sites and an analysis of similarity test (ANOSIM) (Clarke 1993) to examine the differences among the communities identified through ϕ-coefficient. 75 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Results Species assemblage and environmental factors Gradient analysis of vegetation assemblage (DCA) The first 4 axes of the RDA explained 92.5% of the cumulative variance of species data and 97.1% of the cumulative variance of speciesenvironment relationship, the Monte Carlo permutation test showed that all canonical axes of RDA were significant (p < 0.05), meaning that species composition was significantly related to the measured variables (Figure 3). The highest cumulative percentage of variance (68.1%) was explained by the first 2 canonical axes, the first axis with 40.4% of variance explained was significantly (p < 0.05) related to pH (r = - 0.82) and altitude (r = 0.81), the variance explained by each of these 2 variables according to forward selection was respectively 33.2% and 32%, the combined variance explained by pH and altitude was 56.6%. The second axis (27.7%) was negatively related to diversity (r = - 0.85), EC (r = - 0.68), OM (r = - 0.61) and positively related to pH (r = 0.46) (Table 1). According to the marginal effects (λ1), the eigenvalues expressed by each variable used individually in RDA (Table 2) indicated that the best explanatory variables were pH (0.33) and altitude (0.32). As indicated by the conditional effects (λA) (Table 2), these 2 variables showed the most significant (p < 0.05) increases in the total sum of eigenvalues during the forward selection, whereas the contribution of OM and EC was not significant. The first two ordination axes of the DCA explained 43.7% variability of vegetation assemblages. With an explained variance of 38.1%, the first axis symbolizing the longest gradient (3.38) in species composition, represented the low-high altitudinal gradient, so species dispersion along this axis reflected an altitudinal gradient of species assemblage. The second axis with only 5.6% of explained variance, reflected the variability caused by the remaining environmental gradient, mainly, OM and EC (Figure 2). ϕ-coefficient of association Figure 2 – D etrended correspondence analysis (DCA) ordination diagram of the plant species in Benin-Haoua. Species names are indicated in Table 4 and Table 5. 76 Initially, the classification of sites based on Jaccard’s similarity coefficient and cluster analysis resulted in 4 different groups (Group A = Bissa and Souamer; Group B = Briera and Kadian; Group C = Boucheral and Souahlia; Group D = Bouriach) (Figure 4), the highest level of diversity according to Shannon index were shown by groups A (4.38) and C (4.11) whereas group D showed the lowest diversity (3.5), when comparing the 4 groups. The t-test (Table 3) showed significant differences in diversity between groups A and B, A and D, B and D and C and D, but no significant differences between A and C and B and C. Subsequently, the ϕ-coefficient of association related to each species was calculated in accordance with the Jaccard classification. The ϕ-coefficient resulted in 4 vegetation units A, B, C and D differentiated successively by 22, 10, 1 and 1 diagnostic species (species ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) Figure 3 – R edundancy analysis (RDA) ordination diagram of 62 species, 4 environmental variables and 7 sites. Species names are indicated in Table 4 and Table 5 column code. Table 1 – Eigenvalues and percentage of variance explained by RDA, with Pearson correlations (r) between environmental variables and the 4 canonical axes. Axes Axis 1 Axis 2 Axis 3 Axis 4 Eigenvalues 0.384 0.265 0.172 0.104 Species-environment correlations 0.99 0.989 0.991 0.929 Species data 38.4 64.9 82.1 92.5 Species-environment relation 40.4 68.1 86.3 97.1 – 0.823 ** 0.455 0.279 – 0.186 EC 0.056 – 0.684 * – 0.434 0.523 OM – 0.172 – 0.606 * 0.468 – 0.198 0.814 ** 0.238 – 0.528 0.039 0.251 – 0.851 ** 0.194 – 0.408 F P 3.985 0.034 Cumulative percentage variance of: Environmental variables pH Altitude Shannon index Monte Carlo test (999 permutations) Test of significance of all canonical axes ecologia mediterranea – Vol. 41 (2) – 2015 77 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Table 2 – M arginal and conditional effects of each environmental variable obtained through forward selection in the RDA ordination. Marginal Effects Conditional Effects λ1 λA P F pH 0.33 0.33 0.036 2.49 Altitude 0.32 0.26 0.048 2.44 Shannon index 0.24 0.18 0.202 2.31 EC 0.19 0.14 0.070 3.26 OM 0.16 0.04 0.560 0.86 Variable Table 3 – V alue of Shannon index (H) for each group (A, B, C, D) of sites, with value of t-test (higher part of the matrix) comparing between groups diversity and significance level (lower part of the matrix) (*** p < 0.001; ** p < 0.01; * p < 0.05), with, Group A = Bissa and Souamer; Group B = Briera and Kadian; Group C = Boucheral and Souahlia; Group D = Bouriach. (H) Group A 4.38 Group A Group B 3.99 Group C 4.11 Group D 3.50 P-value t-test Group A Group B Group C Group D 2.66 1.84 4.20 0.74 2.15 Group B ** Group C NS NS Group D *** * occurring in a single vegetation unit), each with the highest fidelity value to the vegetation unit to which it was assigned (Table 4). It was also remarkable that many differential species (species occurring in a few vegetation units) were shared between the 4 vegetation units, 6 species between A and B, 5 species between A and C, 5 species between B and C and 12 species between A, B and C (Table 5). According to the results of regression analysis (Table 4), vegetation unit A showed two major subgroups of diagnostic species significantly (r2 = 0.9; p < 0.01) related to environmental conditions. The first subgroup was negatively related to pH and the second was positively related to altitude. Vegetation unit B was mostly positively (r2 = 0.8; p < 0.05) related to pH and EC. Vegetation unit C was negatively related to pH and EC. Vegetation unit D was positively related to pH and altitude, and both vegetation unit C and D were not significantly related to the studied environmental factors. 2.76 ** Figure 4 – D endrogram showing the classification of 7 reference sites in Beni-Haoua based on binary data according to the Jaccard’s similarity coefficient. 78 ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) Table 4 – Synoptic table of 34 diagnostic species distributed over 4 vegetation units (A, B, C, D), based on Jaccard classification and Φ-coefficient of association. In addition the results of linear regression analysis showing the most influential environmental factors on each species with r2 (determination coefficient) and (P-value) (significance level) (** p < 0.01; * p < 0.05). Φ-coefficient Species Code Φ value Urginea maritima (L.) Baker U.Mar 100 Rubia peregrina L. R.Per Erica cinerea L. E.Cin Hedysarum spinosissimum L. Vegetation Unit Linear regression pH EC A – ** 100 A – ** 100 A – ** OM Altitude r2 P – 0.9 0.01 – 0.9 0.01 – 0.9 0.01 ** H.Spi 100 A – – 0.9 0.01 M.Com 100 A – ** – 0.9 0.01 Centella asiatica L. C.Asi 100 A – – 0.9 0.01 Quercus ilex L. Q.Ile 73 A – 0.6 0.14 Anagallis arvensis L. A.Arv 73 A 0.5 0.21 Alnus glutinosa (L.) Gaertn. A.Glu 73 A – 0.6 0.14 Cistus monspeliensis L. C.Mon 73 A – 0.9 0.01 E.Arb 73 A – 0.6 0.14 – Muscari comosum (L.) Mill. Erica arborea L. – – * – – – – Linum corymbiferum Desf. L.Cor 73 A 0.6 0.14 Quercus suber L. Q.Sub 64.6 A + + ** 0.9 0.01 Salvia verbenaca L. S.Ver 64.6 A + + ** 0.9 0.01 Asperula arvensis L. A.Ars 64.6 A + + ** 0.9 0.01 Crataegus laevigata Poir. C.Lae 64.6 A + + 0.9 0.01 Schinus molle L. S.Mol 64.6 A 0.6 0.15 Anthyllis tetraphylla L. A.Tet 64.6 A Bromus madritensis L. B.Mad 64.6 A – Marrubium multibracteatum H&M. – – – – ** – – ** 0.6 0.14 – 0.6 0.15 – 0.6 0.15 + ** 0.9 0.01 * M.Mul 64.6 A Moehringia trinervia L. M.Tri 54.8 A Quercus coccifera L. Q.Coc 54.8 A – – 0.6 0.17 Artemisia absinthium L. A.Abs 100 B +* +* 0.8 0.05 Althaea officinalis L. A.Off 100 B + + * 0.8 0.05 Gladiolus byzantinus Mill. G.Byz 100 B +* +* 0.8 0.05 Helichrysum arenarium (L.) Moench H.Are 100 B + + * 0.8 0.05 Helichrysum stoechas (L.) Moench H.Sto 100 B +* +* 0.8 0.05 Jacobaea maritima (L.) Pel.Mei. + + ** * * J.Mar 100 B + 0.8 0.05 Chamaerops humilis L. C.Hum 73 B –* –* 0.8 0.06 Chamaelirium luteum L. C.Lut 73 B – – * 0.8 0.06 Daucus carota L. D.Car 73 B –* –* 0.8 0.06 Picris echioides L. P.Ech 73 B 0.6 0.19 Opuntia ficus–indica (L.) Mill. O.Ind 73 C – 0.5 0.21 Lavandula angustifolia Mill. L.Ang 64.6 D + 0.3 0.46 * * * + + – + Analysis of similarity (ANOSIM) The ANOSIM showed a highly significant difference (p < 0.001) in taxonomic composition between vegetation unit A and B, significant differences (p < 0.05) between vegetation unit A and D, B and D and C and D: as indicated by the R value close to 1 there was a clear ecologia mediterranea – Vol. 41 (2) – 2015 difference in taxonomic composition between these vegetation units, whereas no significant differences were observed between A and C and B and C (Table 6), for which the R values almost equal to 0, indicative of an important overlap. 79 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Table 5 – Synoptic table of 28 differential species common to 2 or 3 vegetation units. In addition to the results of linear regression analysis showing the most influential environmental factors on each species with r2 (determination coefficient) and (P-value) (significance level) (** p < 0.01; * p < 0.05). Φ-coefficient Species Code Φ value Lavandula dentata L. L.Den 54.8 Pallenis spinosa (L.) Casso Chrysanthemum myconis L. Ceratonia siliqua L. Vegetation Unit Linear regression pH EC OM A, B +* * Altitude r2 P + 0.80 0.03 P.Spi 54.8 A, B + + 0.80 0.03 C.Myc 54.8 A, B +* + 0.80 0.03 + + 0.80 0.03 0.30 0.46 C.Sil 54.8 A, B G.Rob 40.0 A, B Quercus robur L. Q.Rob 25.8 A, B 0.50 0.26 Asplenium adiantum-nigrum L. A.Adn 54.8 A, C –* –* 0.70 0.07 Asplenium ceterach L. A.Cet 54.8 A, C – – * 0.70 0.07 Astragalus monspessulanus L. A.Mon 54.8 A, C –* –* 0.70 0.07 Thymus vulgaris L. T.Vul 40.0 A, C – – Cistus albidus L. C.Alb 25.8 A, C Capsella bursa-pastoris (L.) Medik C.Bup 40.0 B, C Cynoglossum creticum Mill. C.Cre 40.0 B, C Melica minuta L. M.Min 40.0 Raphanus raphanistrum L. R.Rap 40.0 Geranium robertianum L. Tetragonolobus biflorus Desr. * – – + * – 0.80 0.05 + 0.73 0.07 + – 0.80 0.03 + – 0.80 0.03 B, C + – 0.80 0.03 B, C + – 0.80 0.03 + * * –* T.Bif 40.0 B, C + 0.10 0.83 E.Bon 25.8 A, B, C – + 0.40 0.40 Echium plantagineum L. E.Pla 25.8 A, B, C – + 0.40 0.40 Avena sterilis L. A.Ste 25.8 A, B, C – + 0.40 0.40 Arisarum vulgare Targ.Tozz. A.Vul 25.8 A, B, C – + 0.40 0.40 Bellis annua L. B.Ann 25.8 A, B, C – + 0.40 0.40 Borago officinalis L. B.Off 25.8 A, B, C – + 0.40 0.40 Oxalis corniculata L. O.Cor 25.8 A, B, C – + 0.40 0.40 Papaver rhoeas L. P.Rho 25.8 A, B, C – + 0.40 0.40 Scorpiurus muricatus L. S.Mur 25.8 A, B, C – + 0.40 0.40 Taraxacum officinale F.H. T.Off 25.8 A, B, C – + 0.40 0.40 Tanacetum parthenium (L.) Sch. Bip. T.Par 25.8 A, B, C – + 0.40 0.40 Valeriana tuberosa L. V.Tub 25.8 A, B, C – + 0.40 0.40 Erigeron bonariensis L. Table 6 – ESimilarity analysis (ANOSIM) between the 4 vegetation units (Veg. Unit) with P-value (higher part of the matrix) and R-value (lower part of the matrix). P values (*** p < 0.001; * p < 0.05) Veg. Unit A Veg. Unit A R-Values 80 Veg. Unit B Veg. Unit C Veg. Unit D 0.0002 *** 0.169 0.019 * 0.060 0.028 * Veg. Unit B 0.156 Veg. Unit C 0.034 0.062 Veg. Unit D 0.781 0.978 0.041 * 0.977 ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) Discussion The goal of this study was to inventory and examine the plant diversity in relation to the environmental conditions in the coastal Mediterranean forest of Beni-Haoua. Overall, as indicated by the DCA, the study area was characterized by a strong altitudinal gradient. Indeed, the forward selection in the RDA ordination showed that plant species distributions were strongly affected by altitude and pH. The effect of these two variables on plant species distributions was consistent with RDA ordination results, the highest proportion of variation in species data (38.4%) was explained by the first canonical axis, correlation analysis indicated that this axis was highly positively related to altitude and negatively related to pH, so this axis was interpreted as altitudinal gradient axis, on which a change in species composition from low to high altitude was observed, the amount of variability in species-environment relation along this axis was 40.4%, suggesting that the influence exerted on the distribution of vegetation in the study area by all the remaining unmeasured biotic and abiotic factors was less important in comparison to altitude, these findings were highlighted by several authors in previous studies (Reeder & Riechert 1975, Gutierrez et al. 1998, Erschbamer et al. 2006). The proportion of variation in species data and species-environment relation explained by the second canonical axis was less important, this axis was positively correlated with pH and negatively correlated to OM, EC and Shannon index, meaning that the diversity in the study area was influenced by OM in a positive way (R = 0.7), and pH in a negative way (R = – 0.46), the latter observation was as also reported by Isermann (2005). The 87 species recorded as binary data (presence/absence) and used in the Jaccard classification enabled us to identify 4 different groups of sites, significantly distinct in species richness and diversity according to t-test. Through the ϕ-coefficient of association, these 4 groups of sites have been successfully differentiated by 4 vegetation communities. The first vegetation community (vegetation unit A) encompasses 22 diagnostic species closely related to high-altitude, this community was distinguished by three species of the genus Quercus (Quercus ilex L., Quercus suber L. and Quercus coccifera L.), one of the most important woody angiosperms in the area, positively correlated with altitude according ecologia mediterranea – Vol. 41 (2) – 2015 to the results of this study, also mentioned by Barbero at al. (1992) & Ducousso et al. (1996), the second vegetation community (vegetation unit B) observed at low-altitude, includes 10 diagnostic species, mostly significantly correlated with pH and EC, the third and fourth community less rich in diagnostic species have not shown significant correlations with the studied environmental factors. The results of regression analysis suggested in general that altitude and pH were the most important factors in species occurrence and community composition in the coastal region of Beni-Haoua. In this study, ANOSIM test demonstrated a clear difference in species composition among the vegetation communities, especially between the community related to high-altitude (vegetation unit A) and the community related to low-altitude (vegetation unit B) (p < 0.001) implying that changes in flora were connected to the change of altitude, vegetation is therefore an important ecological indicator of the local topography (Pinder et al. 1997, Qiu et al. 2012). Conclusion Through the 7 sites selected for this study, we inventoried 87 plant species among which 62 species closely related to the selected environmental factors, nevertheless, the statistical analysis revealed that species occurrence and community composition were mainly affected by an altitudinal gradient, so that the composition of plant communities related to high altitude was highly different than that related to low altitude. Finally, among very few studies carried out in Beni-Haoua forest our study improved understanding species distributions and occurrence in a southern Mediterranean forest. References Ababou A., Chouieb M., Khader M., Mederbal K., Saïdi D. & Bentayeb Z., 2009. Multivariate analysis of vegetation in the salted lower-Cheliff plain, Algeria. B. Soc. Bot. Mex. 85: 59-69. Ababou A., Chouieb M., Khader M., Mederbal K. & Saïdi D., 2010. Using vegetation units as salinity predictors in the lower-Cheliff, Algeria. Turk. J. Bot. 34: 1-10. Barbero M., Loisel R., & Quezel P., 1992. Biogeography, ecology and history of Mediterranean Quercus ilex ecosystems. Vegetatio 100: 19-34. 81 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Bruelheide H., 2000. A new measure of fidelity and its application to defining species groups. J. Veg. Sci. 11: 167-178. Buckley H.L., Burns J.H., Kneitel J.M., Walters E.L., Munguia P. & Miller T.E., 2004. Small-scale patterns in community structure of Sarracenia purpurea inquilines. Community Ecol. 5: 181-188. Clarke K.R., 1993. Non-parametric multivariate analysis of changes in community structure. Aust. J. Ecol. 18: 117-143. Collins S.L., Mitchell G.S. & Klahr S.C., 1989. Vegetation-Environment Relationships in a Rock Outcrop Community in Southern Oklahoma. Am. Midl. Nat. 122: 339-348. Cornwell W.K. & Ackerly D.D., 2009. Community assembly and shifts in plant trait distributions across an environmental gradient in coastal California. Ecol. Monogr. 79: 109-126. Ducousso A., Guyon J.P. & Kremer A., 1996. Latitudinal and altitudinal variation of bud burst in western populations of sessile oak Quercus petraea (Matt Liebl). Ann. Sc. Forest. 53: 775-782. Erschbamer B., Mallaun M. & Unterluggaue P., 2006. Plant diversity along altitudinal gradients in the Southern and Central Alps of South Tyrol and Trentino (Italy). Gredleriana 6: 47-68 Flahault C. & Schroter C., 1910. Report on phytogeographical nomenclature. Proceedings of the 3rd International Botanical Congress, Brussels. 1: 131164. Flinn K.M., Gouhier T.C., Lechowicz M.J. & Waterway M.J., 2010. The role of dispersal in shaping plant community composition of wetlands within an oldgrowth forest. J. Ecol. 98: 1292–1299. Gutierrez J.R., Lopez-Cortes F. & Marquet P.A,. 1998. Vegetation in an altitudinal gradient along the Rıo Loa in the Atacama Desert of northern Chile. J. Arid. Environ. 40: 383-399. Hill, M.O. and H.G. Gauch. 1980. Detrended correspondence analysis, an improved ordination technique. Vegetatio 42: 47-58. Isermann M., 2005. Soil pH and species diversity in coastal dunes. Plant Ecol. 178: 111-120. Jongman R.H.G., Ter Braak C.J.F. & van Tongeren O.F.R., 1996. Data analysis in community and landscape ecology. Cambridge Univ. Press, Cambridge. Kingston N. & Waldren S., 2003. The Plant Communities and Environmental Gradients of Pitcairn Island: The Significance of Invasive Species and the Need for Conservation Management. Ann. Bot. 92: 31-40. Legendre P. & Gallagher E., 2001. Ecologically meaningful transformation for ordination of species data. Oecologia 129: 271-280. Legendre P., 1993. Spatial autocorrelation: trouble or new paradigm. Ecology 74: 1659-1673. Magurran A.E., 1988. Ecological diversity and its measurement. Princeton Univ. Press, Princeton. McCullagh P. & Nelder J.A., 1989. Generalized Linear Models. 2nd ed. Chapman and Hall, London. Montgomery D.C., Peck E.A. & Vining G.G., 2006. Introduction to Linear Regression Analysis. 4th ed. Wiley-Interscience, New Jersey. 82 Montgomery D.C. & Runger G.C., 2006. Applied Statistics and Probability for Engineers. 4th ed. John Wiley & Sons, New Jersey. O’Brien R.M., 2007. A caution regarding rules of thumb for variance inflation factors. Qual. Quant. 41: 673690. Pinder J.E., Kroh G.C., White J.D. & Basham May A.M., 1997. The Relationships between Vegetation Type and Topography in Lassen Volcanic National Park. Plant. Ecol. 131:17-29. Quezel P. & Santa S. 1962. New flora of Algeria and southern desert regions. Volume 1. ed. CNRS, Paris. Quezel P. & Santa S. 1963. New flora of Algeria and southern desert regions. Volume 2. ed. CNRS, Paris. Qiu B., Zhong M., Zeng C., Tang Z. & Chen C., 2012. Effect of topography and accessibility on vegetation dynamic pattern in Mountain-hill Region. J. Mt. Sci. 9: 879-890. Ramette A., 2007. Multivariate analyses in microbial ecology. FEMS Microbiol. Ecol. 62:142-160. Rao C.R., 1995. A review of canonical coordinates and an alternative to correspondence analysis using Hellinger distance. Questiio 19: 23-63. Reeder W.G. & Riechert S.E., 1975. Vegetation change along an altitudinal gradient, Santa Cruz Island, Galapagos. Biotropica 7: 162-175. Schmidtlein S. & Sassin J., 2004. Mapping of continuous floristic gradients in grasslands using hyperspectral imagery. Remote Sens. Environ. 92: 126-138. Sherman R., Mullen R., Haomin L., Zhendong F. & Yi W., 2008. Spatial patterns of plant diversity and communities in Alpine ecosystems of the Hengduan Mountains, northwest Yunnan, China. J. Plant. Ecol. 1: 117-136. Ter Braak C.J.F., 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179. Ter Braak C.J.F., 1994. Canonical community ordination. Part I: Basic theory and linear methods. Ecoscience 1: 127-140. Tozer M., 2003. The native vegetation of the Cumberland Plain, western Sydney: systematic classification and field identification of communities. Cunninghamia 8: 1-75. Walkley A. & Black I.A., 1934. An examination of Degtjareff method for determining soil organic matter and a proposed modification of the Chromic acid titration method. Soil Sci. 37: 29-37. Wisheu I.C. & Keddy P.A., 1989. Species richnessstanding crop relationship along four lakeshore gradients: constraints on the general model. Can. J. Botany 67: 1609-1617. Zar J.H., 1999. Biostatistical analysis. 4th ed. Prentice Hall, New Jersey. Zhu M., Hastie T.J. & Walther G., 2005. Constrained ordination analysis with flexible response functions. Ecol. Model. 187: 524-536. Zuur A.K., Ieno E.N. & Smith G.M., 2007. Analysing Ecological Data. Springer, New York. ecologia mediterranea – Vol. 41 (2) – 2015 Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) Appendix – L ist of species recorded in the Mediterranean region of Beni-Haoua during spring (March May) 2013. In bold are the species used in the analysis. Class Family Species Cycle of life Life-form Anacardiaceae Schinus molle L. Perennial Phanerophyte Pistacia lentiscus L. Perennial Nanophanerophyte Centella asiatica L. Perennial Hemicryptophyte Daucus carota L. Annual Therophyte Eryngium triquetrum Vahl. Biennial Hemicryptophyte Foeniculum vulgare (Mill) Gaertn. Perennial Hemicryptophyte Torilis nodosa Gaertn. Annual Therophyte Apocynaceae Nerium oIeander L. Perennial Phanerophyte Asteraceae Anacyclus pyrethrum (L.) Casso. Annual Therophyte Artemisia absinthium L. Perennial Chamaephyte Bellis annua L. Annual Therophyte Calendula arvensis L. Annual Therophyte Centaurea melitensis L. Annual Therophyte Chrysanthemum myconis L. Annual Therophyte Erigeron bonariensis L. Annual Therophyte Galactites tomentosa (L.) Moench. Annual Therophyte Helichrysum arenarium (L.) Moench. Perennial Hemicryptophyte Helichrysum stoechas (L.) Moench. Perennial Chamaephyte Hypochaeris glabra L. Annual Therophyte Jacobaea maritima (L.) Pel.Mei. Perennial Chamaephyte Pallenis spinosa (L.) Casso. Annual Therophyte Phagnalon saxatile (L.) Casso. Perennial Chamaephyte Picris echioides L. Annual Therophyte Scolymus maculatus L. Annual Therophyte Silybum marianum (L.) Gaertn. Biennial Hemicryptophyte Tanacetum parthenium (L.) Sch. Bip. Perennial Hemicryptophyte Taraxacum officinale F.H. Perennial Hemicryptophyte Betulaceae Alnus glutinosa (L.) Gaertn. Perennial Phanerophyte Boraginaceae Borago officinalis L. Annual Therophyte Cynoglossum creticum Mill. Biennial Hemicryptophyte Echium plantagineum L. Biennial Hemicryptophyte Brassica nigra (L.) Koch. Annual Therophyte Capsella bursa-pastoris (L.) Medik Annual Therophyte Magnoliopsida Apiaceae Brassicaceae Raphanus raphanistrum L. Annual Therophyte Cactaceae Opuntia ficus-indica (L.) Mill. Perennial Phanerophyte Caryophyllaceae Moehringia trinervia L. Annual Therophyte Cistaceae Cistus albidus L. Perennial Chamaephyte Cistus monspeliensis L. Perennial Chamaephyte Convolvulaceae Convolvulus althaeoides L. Perennial Hemicryptophyte Ericaceae Erica arborea L. Perennial Nanophanerophyte Erica cinerea L. Perennial Nanophanerophyte Anthyllis tetraphylla L. Annual Therophyte Astragalus monspessulanus L. Perennial Hemicryptophyte Calycotome spinosa (L.) Lamk Perennial Nanophanerophyte Ceratonia siliqua L. Perennial Phanerophyte Hedysarum spinosissimum L. Annual Therophyte Scorpiurus muricatus L. Annual Therophyte Tetragonolobus biflorus Desr. Annual Therophyte Fabaceae ecologia mediterranea – Vol. 41 (2) – 2015 83 Adda Ababou, Mohammed Chouieb, Abdelkader Bouthiba, Djamel Saidi, Khalladi Mederbal Class Family Species Cycle of life Life-form Fagaceae Quercus coccifera L. Perennial Phanerophyte Quercus ilex L. Perennial Phanerophyte Quercus robur L. Perennial Phanerophyte Quercus suber L. Perennial Phanerophyte Erodium cicutarium L’Her. Annual Therophyte Geranium robertianum L. Annual Therophyte Lavandula angustifolia Mill. Perennial Chamaephyte Lavandula dentata L. Perennial Chamaephyte Marrubium multibracteatum H&M. Perennial Hemicryptophyte Salvia verbenaca L. Perennial Hemicryptophyte Geraniaceae Pinopsida Liliopsida Magnoliopsida Lamiaceae Filicopsida 84 Thymus vulgaris L. Perennial Chamaephyte Malvaceae Althaea officinalis L. Perennial Hemicryptophyte Oleaceae Olea europea L. Perennial Phanerophyte Oxalidaceae Oxalis corniculata L. Perennial Hemicryptophyte Papaveraceae Papaver rhoeas L. Annual Therophyte Plantaginaceae Plantago coronopus L. Annual Therophyte Primulaceae Anagallis arvensis L. Annual Therophyte Rosaceae Crataegus laevigata Poir. Perennial Mesophanérophyte Rubiaceae Asperula arvensis L. Annual Therophyte Rubia peregrina L. Perennial Hemicryptophyte Valerianaceae Valeriana tuberosa L. Perennial Cryptophyte Agavaceae Agave americana L. Perennial Hemicryptophyte Araceae Arisarum vulgare Targ.Tozz. Perennial Cryptophyte Arecaceae Chamaerops humilis L. Perennial Nanophanerophyte IIridaceae Gladiolus byzantinus Mill. Perennial Cryptophyte Liliaceae Chamaelirium luteum L. Perennial Cryptophyte Muscari comosum (L.) Mill. Perennial Cryptophyte Urginea maritima (L.) Baker Perennial Cryptophyte Linaceae Linum corymbiferum Desf. Annual Therophyte Poaceae Ampelodesma mauritanicum (Poir.) Perennial Hemicryptophyte Avena sterilis L. Annual Therophyte Bromus madritensis L. Annual Therophyte Melica minuta L. Perennial Hemicryptophyte Juniperus oxycedrus L. Perennial Phanerophyte Tetraclinis articulata (Vahl) Perennial Phanerophyte Pinus halepensis Mill. Perennial Phanerophyte Pinus pinaster Soland. Perennial Phanerophyte Asplenium adiantum-nigrum L. Perennial Hemicryptophyte Asplenium ceterach L. Perennial Hemicryptophyte Cupressaceae Pinaceae Aspleniaceae ecologia mediterranea – Vol. 41 (2) – 2015 Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) Aperçu du régime alimentaire de la loutre d’Europe, Lutra lutra, dans l’est de l’Algérie (Parc national d’El-Kala) Roland LIBOIS1, *, Rachida GHALMI2, Amina BRAHIMI1, 3 1. Labo de zoogéographie, ULG, Bâtiment B22, Chemin de la vallée, 4, 4000 Sart Tilman, Belgique 2. École nationale supérieure des sciences de la mer et de l’aménagement du littoral ; Campus universitaire de Dely-Ibrahim Bois des Cars, BP19, 16320 Alger, Algérie 3. Département des sciences agronomiques, Université Mohammed Kheider, 7000 Biskra, Algérie * Corresponding author: [email protected] Received: 1 Avril, 2015; First decision: 20 October, 2015; Revised: 29 November, 2015; Accepted: 30 November, 2015 Abstract In 1997, faeces samples (spraints) (n = 175) from the European otter (Lutra lutra) were taken in seven localities of the El-Kala region. This is a restricted eco-geographic complex in which freshwater hydrographical systems, comprising rivers, ponds and wet coastal zones (brackish lagoons, shipping channels), are highly interwoven. The frequency of occurrence and relative abundance of consumed taxons were calculated on the basis of 493 identified prey. Fish made up more than 88% of the otter’s catch (relative abundance), with a strong predominance of Pseudorasbora parva, Luciobarbus callensis and undetermined Cyprinidae. Anuran amphibians made up 9% of the prey. The remainder were represented, in order of significance, by mullets, eels (Anguilla anguilla), bleaks (Alburnus alburnus), perciformes, gobies, insects, birds, crustaceans, etc. However, Cyprinidae, the eel and the barbel dominated in terms of ingested biomass. The otter’s diet varies with local conditions: lagoons and channels have marine or migratory fish; rivers are dominated by barbels; ponds by Cyprinidae. Finally, small-sized Keywords: otter, diet, National Park El Kala, Algeria, Pseudorasbora. ecologia mediterranea – Vol. 41 (2) – 2015 fish (topmouth gudgeon and barbel) dominated the diet in terms of numbers: 62% are smaller than 12.5 cm. The introduction of the topmouth gudgeon into the region could have catastrophic consequences for endemic fish, such as Pseudophoxinus callensis. Résumé En 1997, des échantillons de fèces (épreintes) (n = 175) de la loutre d’Europe (Lutra lutra) ont été récoltés dans sept localités dans la région d’El-Kala. C’est un complexe écogéographique restreint où les systèmes hydrographiques dulcicoles, constitués de rivières, d’étangs et de zones humides littorales (lagunes saumâtres, chenaux maritimes), sont assez imbriqués. Les fréquences d’occurrence et l’abondance relatives des taxons consommés ont été calculées à partir des 493 proies identifiées. Les poissons constituent plus de 88 % des prises (abondance relative) avec une très large prédominance de Pseudorasbora parva, Luciobarbus callensis et des cyprinidés indéterminés. Les amphibiens anoures constituent 9 % des proies. Le reste Mots clés : loutre, régime alimentaire, Parc national d’El Kala, Algérie, Pseudorasbora. 85 Roland Libois, Rachida Ghalmi, Amina Brahimi est représenté, selon l’importance, respectivement par des mugilidés, des anguilles (Anguilla anguilla), l’ablette (Alburnus alburnus) des perciformes, des gobiidés, des insectes, des oiseaux, des crustacés... Cependant, les cyprinidés, l’anguille et le barbeau dominent pour la biomasse ingérée. Le régime alimentaire est fonction des conditions locales : les lagunes et chenaux avec des poissons marins ou amphihalins ; les rivières où dominent les barbeaux ; les étangs avec les cyprinidés. Enfin, les poissons de petite taille (pseudorasbora et barbeau) dominent le régime en nombre : 62 % font moins de 12,5 cm. L’introduction de pseudorasbora dans la région pourrait être catastrophique pour des poissons endémiques, comme Pseudophoxinus callensis. Introduction Since the beginning of the 1980’s, whether it is related to fresh, brackish or marine waters, the diet of the Eurasian otter (Lutra lutra L.) has been described in numerous publications. Broyer & Erome (1982) and then Mason & Macdonald (1986) and Kruuk (2006) summarised the findings of most of these publications. In its category of specialised, semiaquatic predators, the Eurasian otter has been shown to be relatively opportunistic and euryphagous, feeding mainly on fish, but also on various animals found in aquatic environments, such as cyclostomes, crustaceans and amphibians. Only two studies have been published on the diet of otters in North Africa, on the basis of their spraints: the first of these was carried out in Morocco (Broyer et al. 1988) and deals with 389 droppings collected mainly in Saharan river wadis, and wadis of the High Atlas and Middle Atlas. This data is characterised by strong variations in the class of prey, depending on seasons and localities. Although fish traditionally play a major dietary role, the analysis made by these authors was limited to the identification of specimens up to the level of zoological classes (fish, amphibians, etc.). The second study was carried out in the Beth river wadi, which is a typical river of the Middle Atlas (Morocco) (Libois et al. 2015). It was based on 760 droppings, methodologically collected at six stations, over two annual cycles, derived from one collection campaign per season (2,442 identified prey). Fish made up more than 70% of the catches (relative abundance) with a very strong predominance of barbels (Luciobarbus labiosa [Pellegrin], Labeobarbus fritschii [Günther] 86 and Labeobarbus paytoni [Boulenger]). The remaining prey corresponded to (in decreasing order): anuran amphibians, insects, ophidians, Mediterranean tortoises (Mauremys leprosa Schweigger), birds, crustaceans and small mammals. Over time, the otter’s diet has evolved for various reasons, of both climatic and anthropogenic origins: major floods and their consequences profoundly modified the composition of the fish population and leading to the local extinction of the small barbels; then, in the autumn of 2010, fishermen introduced Cichlidae into the river. In addition, fish of small size dominated their diet in terms of numbers: 80% are less than 10 cm in size. The present study deals with the dietary habits of the Eurasian otter, in the rivers and wet brackish zones of El-Kala in Algeria. It is proposed to provide an overview of the otter’s choice of prey in a restricted eco-geographic complex, in which freshwater hydrographical systems, comprising rivers, ponds and wet coastal zones (brackish lagoons, shipping channels) are highly interwoven. Study area In the spring of 1997, in the region of ElKala, a series of spraints was collected around the Mellah lagoon (860 ha; 36.9°N 8.34°E) (n = 18), the Oubeira (2,174 ha; 36.8°N 8.4°E) (n = 19) and Tonga (2,392 ha; 36.9°N 8.5°E) (n = 17) ponds, in the Messida channel (36.88°N 8.53°E) (n = 26), as well as in the El-Kébir (36.7°N 8.4°E) (n = 20), Bougous (36.66°N 8.39°E) (n = 74) and ElAreug (36.86°N 8.33°E) (n = 7) river wadis (Ghalmi 1997). The brackish Mellah lagoon collects the waters of a small number of rivers that flow directly into the sea, via a relatively narrow channel. The lagoon is lined with a coastal strip colonised by French Tamarisk (Tamarix gallica L.), where the otter leaves its droppings on large branches at a height of approximately 1.5 m above ground. The vegetation in the Oubeira pond is dominated by helophytes (Phragmites australis [Cav.]) and hydrophytes (Potamogeton sp. and Myriophyllum sp.). In order to restrict the development of aquatic plants, some species of fish have been introduced, grass carp in particular (Aristichthys nobilis [Richardson], Ctenophayngodon idella [Valenciennes], Hypophtalmichthys molitrix [Valenciennes]). The Tonga pond is a vast P. australis and Scirpus lacustris L. reed bed, which is becoming ecologia mediterranea – Vol. 41 (2) – 2015 Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) silted and communicates with the sea via the Messida channel, a backwater blocked by fallen trees, in which the banks are covered by a dense shrubby vegetation. The tailrace of the El-Kebir wadi, slightly downstream from the Mexena dam, is lined with dense riverside vegetation comprising ash (Fraxinus angustifolia Vahl), alder (Alnus glutinosa [L.]) and oleander (Nerium oleander L.). Its course and flow rate are highly irregular, with violent flooding in winter. Faecal matter was collected in a zone that is strongly disturbed by the dam site. The Bougous wadi, a tributary of the El-Kebir wadi, is a torrential river whose bed comprises large pebbles or rocky slabs. Its banks are covered by a degraded forest of cork oaks (Quercus suber L.) and zeen oaks (Quercus canariensis Willd.), accompanied by oleander (N. oleander) and tamarisk (T. gallica and Tamarix africana [Poir.]). The El-Areug wadi flows through a narrow ravine, with a width not exceeding one meter. This river, with its torrential regime, is situated in a highly degraded cork oak forest and feeds the Mellah lagoon. Methods In the laboratory, the analysis of the spraint contents involved the identification of undigested remains of various prey. For this, a standardised method for dropping treatment was followed (Libois et al. 1987a). Teleosts were determined through the recognition of characteristic bone fragments, using reference collections and previous studies: Libois et al. (1987a, b); Libois & Hallet-Libois (1988). Finally, reference collections were prepared for barbels from the El-Kala region, ophidians and amphibians. The feathers were identified by Dr. R. Rosoux, on the basis of the ornithological reference collection of the natural sciences museum in Orléans, France. To optimise data processing, three methods were used: occurrence, abundance and relative biomass, in accordance with the recommendations of Libois et al. (1987a, 1991); Libois & Rosoux (1989, 1991) and Libois (1995, 1997). The characteristic bone fragments were methodically measured, and the size of the biomass of the consumed fish was thereafter estimated on the basis of the studies of Wise (1980) for vertebrae, and the studies of Libois & Hallet (1988) for cephalic fragments. In the case of amphibians, their ecologia mediterranea – Vol. 41 (2) – 2015 mass was estimated to be 10, 15, 20, 30 and 40 g, depending on the estimated size. We estimated the weighted mass of ophidians to be 100 g and that of birds, depending on the species: the respective values were consulted in the “Handbook of the birds of the world” (Anatidae: Carboneras 1992; Rallidae: Taylor 1996). Results A total of 493 prey are identified for a total of 294 occurrences (Tables 1, 2). In general, the El-Kala otters have a “classical” diet with diversified prey: this can be broken down into fish, amphibians, reptiles, birds, arthropods and even molluscs. In terms of occurrence and abundance, fish are by far the main type of prey: barbels (Luciobarbus callensis [Valenciennes]), undetermined Cyprinidae and an introduced fish: the topmouth gudgeons (Pseudorasbora parva [Schlegel]). The biomass represented by the undetermined Cyprinidae, barbels and eels (Anguilla anguilla [L.]) make up two thirds of the prey found in the otter’s droppings (Figure 1). The dietary intake is specific to the seven sites. Indeed, in the surroundings of the Mellah lagoon, mainly small mullets (< 225 mm), eels, and also other diadromous fish: gobies (Gobiidae), big-scale sand smelts (Atherina boyeri Risso) and some Sparidae. Near to the Oubeira pond, large Cyprinidae dominate, although the species could not be identified. On the other hand, on the Tonga pond, the dominant prey are P. parva and anuran amphibians. In terms of biomass, birds predominate, with purple swamp hens (Porphyrio porphyrio [L.]) and two other rallidae, followed by eels (5 individuals more than 45 cm in length). The specific prey richness is greater in the Messida channel: 12 taxons were counted, including the P. parva, which makes up the main component of the food (70%). However, as a consequence of the small size of this invasive species, the corresponding food intake is low. The biomass is dominated by large eels, perciformes and mullets. As the channel feeds into the sea, marine prey (chitons, crustaceans, decapods) or diadromous fish (mullets and perciformes) can be found there. In the rivers (El-Kebir and Bougous), 87 Roland Libois, Rachida Ghalmi, Amina Brahimi Table 1 – Prey occurrence of otter’s spraints by locality in the El-Kala National Park Total Mellah Oubeira Tonga Messida El Kebir Bougous El Areug Polyplacophora 1 1 Crustacea 3 1 1 1 Insecta 4 2 1 1 Anguilla anguilla 25 11 5 5 4 Cyprinidae ind. 57 Alburnus alburnus 7 1 Carassius sp. 3 1 Ctenopharyngodon idella 1 1 Cyprinus carpio 1 Luciobarbus callensis 70 Pseudorasbora parva 43 Atherina boyeri 2 2 Gambusia sp. 2 1 1 Mugilidae 20 11 1 Perciformes 8 3 Gobiidae 2 2 Teleostei ind. 5 2 1 Anoura 36 3 5 Ophidia 1 1 Aves 3 2 1 294 37 27 34 52 27 102 15 Bougous El Areug 15 2 12 26 2 5 1 2 1 1 3 11 8 20 59 2 9 1 6 2 4 2 9 9 5 3 2 Table 2 – Prey abundance of otter’s spraints by locality in the El-Kala National Park Total Mellah Polyplacophora 1 Crustacea 3 1 Insecta 5 2 Anguilla anguilla 25 11 5 Cyprinidae ind. 65 Alburnus alburnus 12 1 Carassius sp. 3 1 Ctenopharyngodon idella 1 1 Tonga 16 Messida El Kebir 1 1 1 1 2 5 4 2 14 30 3 9 2 2 1 Luciobarbus callensis 112 Pseudorasbora parva 162 Atherina boyeri 2 2 Gambusia sp. 2 1 1 Mugilidae 30 21 1 Perciformes 9 4 Gobiidae 5 5 Teleostei ind. 5 2 1 Anoura 45 3 7 Ophidia 1 1 Aves 4 3 1 493 51 39 81 101 31 168 22 Cyprinus carpio 88 Oubeira 1 1 12 54 10 69 99 2 27 1 6 2 4 2 12 9 5 3 6 ecologia mediterranea – Vol. 41 (2) – 2015 Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) Figure 1 – Prey relative biomass of otter’s spraints (El-Kala National Park) (21.5 kg). barbels play a key role in the otters’diet. In the El-Areug wadi, close to the Mellah lagoon, the brackish waters receive prey, such as mullets from the marine environment, and prey such as barbels from the freshwater environment. The frequency distribution of size classes has been studied for the fish that are most abundantly consumed by the otter: L. callensis, P. parva, A. anguilla, mullets and other Cyprinidae (Figure 2). A very high proportion of the fish are small in size: 62% are smaller than 125 mm. Figure 2 – Fish distribution size in the otter’s spraints (El-Kala National Park). ecologia mediterranea – Vol. 41 (2) – 2015 89 Roland Libois, Rachida Ghalmi, Amina Brahimi Discussion In the El-Kala region, the otter’s diet basically resembles that found in studies carried out in Europe (Broyer & Erome 1982; Mason & Macdonald 1986; Libois 1995) and in Morocco (Broyer et al. 1988; Libois et al. 2015). In fact, it is made up essentially from fish, but also amphibians, reptiles, birds, mammals, crustaceans and insects, most of which are aquatic or semi-aquatic animals, or animals that are temporarily related to water in their biological cycle, as in the case of the artificial studies carried out by Heptner & Naumov 1974; Mason & Macdonald 1986; Libois 1995; Kruuk 2006. Although aquatic habitats can be highly diverse, ranging from small rivers to ponds and lagoons, the otter demonstrates considerable trophic plasticity, as well as a remarkable adaptive capacity, depending on the various environments in which it lives, which provide a large variety of prey. It demonstrates a clear pattern of opportunistic behaviour, since the proportions of prey classes remain the same, whether the results be expressed in terms of presence or relative abundance. This confirms the conclusions of similar studies carried out in Europe: Erlinge (1967); Fairley & Wilson (1972); Webb (1975); Callejo-Rey et al. (1979); Jenkins et al. (1979); Chanin (1981); Wise et al. (1981); Gormally & Fairley (1982); Green et al. (1984); Bouchardy (1986); Delibes & Adrian (1987); Libois et al. (1987a); Callejo (1988); Libois & Rosoux (1989); Libois (1995, 1997). In the El-Kala region, we observed strong variations at different sites, like for example in Scotland (Kruuk & Moorhouse 1990) or in the French Massif Central (Libois 1997). On the other hand, we are not able to draw clear conclusions concerning the otter’s possible preference for one particular class of size. We have noticed that most of the fish it catches, barbels and P. parva in particular, are small in size (less than 10 cm), an observation that had already been made in other regions by various authors (Webb 1975: stickleback, Gasterosteus gymnurus [Cuvier], loach, Barbatula barbatula [L.], and sculpin, Cottus gobio L.; Jenkins & Harper 1980: pike, Esox lucius L. and perch, Perca fluviatilis L.; Jenkins et al. 1979: pike, perch, salmonidae; Chanin 1981; Green et al. 1984; Libois 1997; Libois et al. 2015: barbels, Luciobarbus labiosa and Labeobarbus fritschii). In 90 hydrosystems, small-sized fish generally dominate in terms of numbers. Nevertheless, it is necessary to compare the frequency distributions of the size of a prey-species in a predator’s food intake and in its habitat (capture using electro-fishing, traps or fishways), which was not possible to implement in the context of the present study. The introduction of P. parva appears to have had significant repercussions within the ElKala fish community. Although Pseudophoxinus callensis (Guichenot) lives in the Oubeira pond (Paris: MNHN-16-2000-5725, det. Daget), its absence in the otter droppings could be indicative of competition between the two species, to the advantage of P. parva. The latter species is zooplanctivorous, but consumes the eggs and fry of other fish and at the international level it is thus considered to be noxious (Keith et al. 2011; Witkowski 2011). For the El-Kala region and the Kroumirie, where Pseudophoxinus are endemic, this poses a serious problem for the conservation of biodiversity. Acknowledgement We wish to thank Mr. K. Djeffel, Director of the El-Kala National Park, together with his entire team, and in particular Mr. K. Bousentouh and A. Boukrabouza for their assistance in the field. Mr. René Rosoux, Scientific Director of the Orléans natural science Museum, provided us with the identification of feathers in spraints, and corrections to the first draft of this manuscript. References Bouchardy C., 1986. La loutre. Sang de la terre, Paris. 174 p. Broyer J. & Erome G., 1982. Eléments d’écologie de la loutre Lutra lutra (L). Premières données bibliographiques. Bièvre 4 : 33-58. Broyer J., Aulagnier S. & Destre R., 1988. La loutre, Lutra lutra angustifrons Lataste, 1885, Maroc. Mammalia 52: 361-370. Callejo A., 1988. Le choix des proies par la loutre (Lutra lutra) dans le nord-ouest de l’Espagne, en rapport avec les facteurs de l’environnement. Mammalia 52: 11-20. Callejo-Rey A., Rivera J.G., Bas-Lopez S., SanchezCanals J.L. & De Castro-Lorenzo A., 1979. Primeros datos sobre la dieta de la Nutria, Lutra lutra (L.), en aguas continentales de Galicia. Donãna Acta Vert. 6: 191-202. ecologia mediterranea – Vol. 41 (2) – 2015 Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) Carboneras C., 1992. Family Anatidae. In: del Hoyo J., Elliott A. & Sargatal J. (eds.), Handbook of the Birds of the world, vol. 1. Lynx Edicions, Barcelona, 536-628. Chanin P., 1981. The diet of the Otter and its relations with the feral Mink in two areas of Southwest England. Acta Theriol. 26: 83-95. Delibes M. & Adrian I., 1987. Effects of crayfish introduction on otters, Lutra lutra, food in the Donãna National Park, SW. Spain. Biol. Conserv. 42: 153159. Erlinge S., 1967. Food habits of the fish otter, Lutra lutra, in South Swedish habitats. Viltrevy 4: 371443. Fairley J.S. & Wilson S.C., 1972. Autumn food of otter (Lutra lutra) on the Agivery river, Co. Londonderry, Northern Ireland. J. Zool. 166: 468-469. Ghalmi R., 1997. Étude préliminaire du régime alimentaire de la loutre (Lutra lutra) dans le nord-est algérien (Parc national d’El-Kala). DES Sciences Naturelles, Université de Liège, Liège. 57 + 13 p. Gormally M.J. & Fairley J.S., 1982. Food of otters, Lutra lutra, in freshwater lough and an adjacent brackish lough in the West of Ireland. J. Zool. 197: 313-321. Green J., Green R. & Jefferies D.J., 1984. A radiotracking survey of otters, Lutra lutra, on a Pertshire river system. Lutra 27: 85-145. Heptner V.G. & Naumov N.P., 1974. Die Säugetiere der Sowjetunion. Band II. Fischer Verlag, Iena. 1009 p. Jenkins D. & Harper R.J., 1980. Ecology of otters in Northern Scotland. II. Analyses of otter and mink faeces from Deeside (N.E. Scotland) in 1977-1978. J. Anim. Ecol. 49: 737-754. Jenkins D., Walker J.G.K. & Mccowan D., 1979. Analyses of otter (Lutra lutra) faeces from Deeside, NE Scotland. J. Zool. 187: 235-244. Keith P., Persat H., Feunteun E. & Allardi J., 2011. Les poissons d’eau douce de France. Biotope et Muséum national d’histoire naturelle, Mèze et Paris. 552 p. Kruuk H., 2006. Otters: ecology, behaviour and conservation. Oxford University Press, Oxford. 265 p. Kruuk H. & Moorhouse A., 1990. Seasonal and spatial differences in food selection by otters (Lutra lutra) in Shetland. J. Zool. 221: 621-637. Libois R.M., 1995. Régime et tactique alimentaire de la loutre (Lutra lutra) en France : synthèse. Cahiers Ethol. 15: 251-274. Libois R., 1997. Régime et tactique alimentaire de la loutre (Lutra lutra) dans le Massif central. Vie et Milieu 47: 33-45. Libois R.M. & Hallet-Libois C., 1988. Eléments pour l’identification des restes crâniens des poissons ecologia mediterranea – Vol. 41 (2) – 2015 dulçaquicoles de Belgique et du nord de la France. II. Cypriniformes. Fiches d’ostéologie animale pour l’archéologie. Sér. A. n° 4. CRA-CNRS. Valbonne, 24 p. Libois, R.M. & Rosoux R., 1989. Écologie de la loutre (Lutra lutra) dans le marais poitevin. I. Étude de la consommation d’anguilles (Anguilla anguilla). Vie et Milieu 39: 191-197. Libois R.M. & Rosoux R., 1991. Ecologie de la loutre (Lutra lutra) dans le marais poitevin. II. Aperçu général du régime alimentaire. Mammalia 55: 35-47. Libois R.M., Hallet-Libois C. & Lafontaine L., 1987a. Le régime alimentaire de la loutre (Lutra lutra) en Bretagne inférieure. Rev. Ecol. (Terre & Vie) 42: 135-144. Libois R.M., Hallet-Libois C. & Rosoux R., 1987b. Eléments pour l’identification des restes crâniens des poissons dulçaquicoles de Belgique et du nord de la France. I. Anguilliformes, Gastérostéiformes, Cyprinodontiformes, Perciformes. Fiches d’ostéologie animale pour l’archéologie. Sér. A. n° 3. CRACNRS. Valbonne, 15 p. Libois R.M., Rosoux R. & Delooz E., 1991. Écologie de la loutre (Lutra lutra) dans le marais poitevin. III. Variation du régime et tactique alimentaire. Cahiers Ethol. 11: 31-50. Libois R., Fareh M., Brahimi A. & Rosoux R., 2015. Régime alimentaire et stratégie trophique saisonnière de la Loutre d’Europe, Lutra lutra, dans le moyen-Atlas (Maroc). Rev. Ecol. (Terre & Vie) 70: 314-327. Mason C.F. & Macdonald S.M., 1986. Otters: ecology and conservation. Cambridge University Press, Cambridge. 236 p. Taylor P.B., 1996. Family Rallidae. In: del Hoyo J., A. Elliott A. & Sargatal J. (eds.), Handbook of the Birds of the world, vol. 3. Lynx Edicions, Barcelona: 108209. Webb J.B., 1975. Food of the Otter (Lutra lutra) on the Somerset levels. J. Zool. 117: 486-491. Wise M.H., 1980. The use of fish vertebrae in scats for estimating prey size of otters and mink. J. Zool. 192: 25-31. Wise M.H., Linn I.J. & Kennedy C. R., 1981. A comparaison of feeding biology of Mink, Mustela vison and Otter, Lutra lutra. J. Zool. 195: 181-213. Witkowski A. 2011: NOBANIS – Invasive alien species fact sheet – Pseudorasbora parva. – From: Online Database of the European Network on invasive alien species – NOBANIS www.nobanis.org. Date of access 1/2/2015. 91 Résumés de thèses Coralie CALVET 2015 Analyse de l’utilisation de la compensation écologique dans les politiques publiques comme outil de conciliation des intérêts économiques et des objectifs de conservation de la biodiversité Analysis of the use of biodiversity offsetting in public policies as a balancing tool of economic interests and biodiversity conservation objectives. Thèse de doctorat en sciences économiques soutenue le 17 décembre 2015 à l’université d’Avignon. Jury - Valérie Boisvert (Pr, université de Lausanne, Suisse), Sophie Thoyer (Pr, Supagro, Montpellier, France), rapporteurs ; John Thompson (directeur de recherches Umr Cefe-Cnrs, Montpellier, France), François Mesléard (Pr, université d’Avignon, France) examinateurs ; Philippe Puydarrieux (chef de bureau, Ministère de l’écologie, Paris, France) membre invité ; Claude Napoléone (ingénieur de recherches, Ecodev-Inra, Avignon, France), Thierry Dutoit (directeur de recherches, Umr Imbe-Cnrs, Avignon, France) codirecteurs. Mots-clés : compensation écologique, conservation de la biodiversité, politiques publiques, économie néo-institutionnelle. Keywords: biodiversity offsetting, biodiversity conservation, public policies, new-institutional economics. Les gouvernements se sont récemment engagés à enrayer l’érosion de la biodiversité. Dans ce contexte, la compensation écologique est apparue comme une réponse politique en permettant, en principe, de répondre à l’exigence de conciliation de deux intérêts souvent antagonistes : le développement économique et la conservation de la biodiversité. L’objectif de ce travail de thèse est d’analyser si la compensation écologique peut accomplir cette promesse. J’envisage cette problématique avec trois angles d’analyse complémentaires et de façon interdisciplinaire en mobilisant les apports de l’économie et de l’écologie. Premièrement, dans une approche théorique, je pose la question de la com92 patibilité du principe de la compensation avec son objectif de conservation de la biodiversité. Je pose ensuite la question de la possibilité d’atteindre l’objectif d’absence de perte nette de biodiversité dans la mise en œuvre de la compensation. Pour cela, j’étudie empiriquement deux mécanismes de compensation au travers de deux cas d’étude français : une banque de compensation, et la contractualisation agro-environnementale. J’utilise principalement les outils de l’économie néo-institutionnelle pour analyser l’efficacité de ces mécanismes pour la réalisation des objectifs écologiques de la compensation. Au travers d’une approche épistémologique, ma troisième interrogation porte sur le rôle des dynamiques politiques dans la diffusion et dans la promotion de la compensation écologique dans la communauté scientifique. L’analyse théorique met en évidence des limites intrinsèques au principe de la compensation pour atteindre ses objectifs de conservation de la biodiversité, notamment au regard de l’impossibilité d’adopter une approche écologique complexe de la biodiversité dans le processus de la compensation. L’étude empirique montre que les modes d’organisation de la compensation comportent également des limites qui obligent à des compromis susceptibles de remettre en cause l’atteinte des objectifs écologiques de la compensation. Ces résultats mettent en évidence le rôle et l’importance des institutions dans la mise en œuvre des compensations, notamment pour limiter l’apparition de comportements opportunistes, responsables des principaux problèmes d’efficacité identifiés. Enfin, l’analyse épistémologique révèle que le développement et la promotion de la compensation écologique répondent à un agenda politique principalement porté par les politiques anglo-saxonnes et certains acteurs de conservation. Ainsi, la compensation écologique n’est pas un objet neutre car elle sert à la diffusion d’une certaine idéologie sur la pratique de la conservation de la biodiversité dans le sillage du développement durable et de l’économie verte. Pour conclure, ce travail permet de souligner que la conciliation des intérêts économiques et écologiques constitue une problématique complexe dont la voie du consensus ne semble pas permettre de répondre aux enjeux d’érosion de la biodiversité. La compensation offre en somme une occasion de penser les conditions de possibilités et d’impossibilités de la protection de la nature aujourd’hui. Governments have recently pledged to halt the loss of biodiversity. In this context, biodiversity offsetting (BO) appeared as a political response by allowing, in principle, to reconcile two often conflicting interests: economic development and biodiversity conservation. The objective of this work is to analyse whether BO can fulfill that promise. I am conecologia mediterranea – Vol. 41 (1) – 2015 Résumés de thèses sidering this issue from an interdisciplinary perspective by mobilizing the contributions of economy and ecology and using three complementary approaches. First, in a theoretical perspective, I ask the question of the compatibility of the offsetting principle with its biodiversity conservation goal. Second, in an empirical approach, I investigate the strengths and limits of using BO mechanisms in conservation policies, particularly two specific mechanisms analysed through two French case-studies: habitat bank and agri-environmental contracts. I mainly use the new institutional economics framework to analyse the effectiveness of these mechanisms for achieving the environmental objectives of BO. Finally, I adopt an epistemological approach to question the role of political dynamics in the diffusion and the promotion of BO in the scientific community. The results of the theoretical analysis highlights the inherent limits to the BO principle to achieve its conservation goals, especially with regard to the impossibility to adopt a complex ecological approach to biodiversity in the process of BO. My empirical study shows that the mechanisms of BO also have limitations that force compromises likely to jeopardize the achievement of environmental objectives of BO. These outcomes highlight the role and the importance of institutions in the implementation of BO, especially to provide clear and precise rules in order to limit the emergence of opportunistic behaviors, responsible for the major problems identified. Finally, my epistemological analysis shows that the development and the promotion of BO actually respond to a political agenda driven by the Anglo-Saxon policies and some conservation stakeholders. Overall my work emphasizes that BO is not a neutral object as it serves to spread a certain ideology on the practice of biodiversity conservation through the lens of sustainable development and green economy. Finally, this work allows stressing that reconciling economic and conservation interests is a complex problem that cannot be solved by using idealized consensus. Rather, the concept of BO offers an opportunity to think about the conditions of possibilities and impossibilities of the protection of nature today. Thibaut FRÉJAVILLE 2015 Vulnérabilité des forêts de montagne des Alpes occidentales au changement de régime d’incendie Vulnerability of mountain forests to changing fire regime in the western Alps Thèse de doctorat en sciences de l’environnement soutenue le 25 juin 2015 à l’université d’Aix-Marseille, École doctorale 251. ecologia mediterranea – Vol. 41 (1) – 2015 Encadrement - Thomas Curt (directeur de recherches, Irstea, Aixen-Provence, France), Christopher Carcaillet (Directeur d’études, école pratique des hautes études) codirecteurs. Mots clés : Alpes, biogéographie, incendies, inflammabilité, résistance, traits fonctionnels. Key-words: Alps, biogeography, fire, flammability, functional traits, resistance to fire. Les forêts d’altitude connaissent une émergence croissante des feux. Dans le sud de l’Europe, les effets du réchauffement du climat sont accrus par les changements d’usage qui génèrent une accumulation forte du combustible végétal et une reconnexion des massifs forestiers de montagne. Via une analyse rétrospective des feux et du climat des dernières décennies dans le sud-est de la France, cette thèse montre d’abord que l’augmentation de l’activité des feux a été largement restreinte aux zones de montagne, en particulier dans les Alpes du sud, les politiques de lutte contraignant les feux au sein des paysages méditerranéens. Pour quantifier la vulnérabilité au feu des écosystèmes et des espèces d’arbre de montagne, j’ai ensuite croisé des données de climat et de végétation des Alpes occidentales afin de caractériser les attributs d’inflammabilité et de résistance du végétal qui gouvernent ses effets et sa réponse au feu. Ces informations ont été utilisées pour simuler le comportement des feux de surface et leurs dommages (mortalités des arbres) au sein des niches environnementales des espèces dominantes. Les résultats obtenus montrent que l’abondance et les propriétés d’inflammabilité des litières, des strates herbacées et arbustives varient au sein des Alpes en fonction de l’ouverture de canopée, sa hauteur et la saisonnalité des précipitations. Ces gradients biotiques et abiotiques gouvernent ainsi les niches d’inflammabilité des espèces, depuis les forêts denses et humides peu inflammables des Alpes externes du nord aux forêts ouvertes inflammables périméditerranéennes des Alpes du sud et subalpines des Alpes internes. Par ailleurs, les traits de résistance au feu des arbres varient entre espèces en lien avec l’inflammabilité des communautés, selon un compromis entre tolérance (écorce épaisse en climat sec) et évitement (houppier haut en climat humide), suggérant une coévolution entre espèces et feux. De ces interrelations entre climat, composition et structure de la canopée résultent des patrons biogéographiques de vulnérabilité des écosystèmes et des espèces au feu. Le chêne, le pin sylvestre et le pin noir connaîtraient des taux de mortalité élevés. Leur vulnérabilité est néanmoins réduite au sein des forêts fermées et plus humides du montagnard, constituant, ainsi, une marge favorable. À l’inverse, les espèces qui dominent ces marges favorables (le sapin et l’épicéa) sont peu vulnérables (excepté le hêtre et son écorce fine), bien que les taux de mortalité s’accroissent rapidement au sein de leurs marges sèches, à basse (forêts supraméditerranéennes) ou haute altitude (forêts subalpines des Alpes internes). Enfin, l’absence d’attributs de résistance pour le pin à crochet et en particulier le pin cembro définit ces espèces comme les plus vulnérables aux changements globaux. En effet, leur niche est restreinte aux forêts subalpines les plus alticoles et 93 Résumés de thèses particulièrement inflammables en conditions extrêmes, alors que l’écorce épaisse du mélèze qui codomine ces forêts lui permettrait d’outrepasser les feux, comme cela a été observé par le passé. Ainsi, du fait des propriétés de leur niche et de leurs traits d’histoire de vie, les arbres Alpins ne sont pas tous égaux face à l’émergence des feux, mettant en lumière l’importance des résultats de cette thèse dans l’anticipation possible des dommages des feux futurs au sein de la diversité des conditions de végétation et de climat des montagnes du sud de l’Europe. Mountain forests are experiencing an increase in fire occurrence. In southern Europe, land use changes enhance the effects of global warming by promoting a fuel build-up and a greater connectivity of forests. Through a retrospective analysis over the past decades of fire and climate in southeastern France, this dissertation firstly shows that Mediterranean ecosystems have experienced a strong decrease in fire activity, emphasizing the strong efficiency of fire suppression policies, while fires have been more frequent and large in mountains like the southern Alps. I used vegetation and climate data to characterize the flammability and fire resistance traits of mountain forests of the western Alps that determine their effects and responses to fire. Then, assessing the vulnerability to fire of mountain ecosystems and tree species was carried out by simulating the behaviour of surface fires and their impacts (tree mortality) within the environmental niches of dominant tree species. Results show that tree cover and height, and the seasonality of precipitation largely drive the amount and flammability properties of surface fuels (litter, grass and shrubs). Therefore, these environmental gradients govern the flammability niches of tree species, from the few flammable dense and moist forests of the northern Alps, to the open and highly flammable sub-Mediterranean and subalpine forests of the southern and inner Alps, respectively. Otherwise, fire resistance traits of trees vary according to community flammability through a trade-off between tolerance (thick bark in dry climates) and avoidance strategies (high crown in moist climates), suggesting a coevolution between species and surface fires. These interrelationships between climate, composition and canopy structure highlight biogeographic patterns of ecosystem and tree species vulnerability to fire in the western Alps. Fire simulations indicate that Quercus pubescens, Pinus sylvestris and P. nigra should suffer high mortality rates under extreme fire weather conditions. However, their vulnerability is strongly reduced within the close and moist forests of the montane belt, which therefore constitute favourable margins for these species. On the contrary, tree species that dominate these low flammable margins (Abies alba, Picea abies) are less vulnerable to fire (except Fagus sylvatica having a too thin bark), although they should suffer high mortality rates towards their dry margins at low (sub-Mediterranean forests) and high elevation (subalpine forests). Finally, having thin barks and low crowns, the subalpine pines (P. uncinata, P. cembra) seem the most vulnerable to global changes. Indeed, their narrow niches are mostly restricted in the most elevated forests that are highly flammable under extreme fire weather. In contrast, their co-dominant species Larix decidua displays a thick bark and the lowest vulnerability, highlighting that this species may resist to surface fires as shown in the last millennia. To conclude, the niche properties and the life history traits of Alpine trees make them unequally exposed to an increasing fire risk, highlighting the importance of the findings of this thesis to anticipate the impacts of future fires across the diversity of climate and vegetation of southern European mountains. Jacques GAMISANS, un botaniste catalan en Corse (1944-2015) « En abordant en 1966, l’étude de la végétation des montagnes corses, je pensais, étant donné la liste impressionnante des travaux de mes prédécesseurs, avoir affaire à un ensemble floristique parfaitement connu [...]. L’entrée dans le vif du sujet m’a montré qu’il n’en était rien [...]. Certains secteurs montagneux n’avaient encore jamais reçu de visite de botanistes et la prospection systématique des massifs corses pendant sept années (1966-1972) m’a permis de découvrir quelques taxons inédits ou nouveaux pour l’île. » Jacques Gamisans, 1976. Phytocoenologia, 3 (4). C très fier de ses origines, Jacques Gamisans est né en 1944. Après ses études secondaires, il acheva la première phase de son cursus universitaire par la soutenance à l’université de Toulouse de son diplôme d’études supérieures (Des) portant sur les Caractères anatomiques des Calamagrostis de la flore française. Jacques Gamisans a été enseignant-chercheur (maître de conférences) à l’université d’Aix-Marseille III (faculté des sciences et techniques de Marseille-Saint-Jérôme) jusqu’en 1995, puis à l’université Paul-Sabatier de Toulouse où il termina sa carrière universitaire en 2004. C’est lors d’un premier voyage d’agrément à l’invitation d’un ami d’Oletta, en 1962, qu’il découvrit cette remarquable île-montagne qu’est la Corse. Dès lors, son orientation était toute trouvée, il allait s’attacher à étudier la flore et la végétation orophiles de cette île car les domaines alticoles étaient encore peu connus. À partir de 1966, il commença à arpenter assidûment la région de Vizzavona et à en étudier la végétation alticole dans le cadre du projet de création du parc national de Corse, projet qui ne verra pas le jour mais qui fut l’une des prémices à la création du parc naturel régional de Corse. Cette recherche le conduisit à soutenir en 1968 sa thèse de spécialité « Étude phytosociologique de la zone montagneuse correspondant au projet de Parc national de Corse », à l’université d’Aix-Marseille sous la direction du professeur Pierre Quézel, responsable du laboratoire de botanique et d’écologie méditerranéenne. Intégré à ce laboratoire comme assistant puis maître-assistant, Jacques Gamisans poursuivit ses recherches dans le cadre de son doctorat consacré à la végétation de la haute montagne corse, en développant des approches en floristique, phytosociologie et biogéographie, toujours sous la houlette de P. Quézel. atalan ecologia mediterranea – Vol. 41 (2) – 2015 En 1975, il soutint sa thèse de doctorat d’État ès sciences naturelles, « La végétation des montagnes corses », devant un jury prestigieux composé de MM. P. Quézel, H. Ellenberg, M. Grison, P. Ozenda, A. Pons, Ch. Sauvage et R. Tomaselli. Cette thèse fut publiée in extenso dans la revue internationale Phytocoenologia. Ce travail conduisit à la description et à l’analyse détaillées de 42 associations de végétations de l’étage supraméditerranéen à l’étage alpin, dont 30 associations inédites, et à la mise en évidence de 15 taxons nouveaux pour la science et 15 nouveaux pour la Corse. La soutenance, le 4 mars 1975, fut d’ailleurs un jour marquant pour la connaissance de la végétation corse et de sa dynamique, puisque son collègue et ami Maurice Reille, un paléoécologue spécialisé dans l’étude des pollens fossiles, soutint aussi, l’après-midi même, son doctorat d’État sur la végétation tardiglaciaire et holocène de Corse ! Leurs terrains communs et les résultats complémentaires obtenus ont conduit à des conceptions novatrices à l’échelle méditerranéenne, avec une mise en perspective de la dynamique et de la biogéographie du peuplement végétal de ces hautes montagnes méditerranéennes encore marquées par la présence de taxons de souche européenne, notamment le fameux élément artico-alpin. Jacques Gamisans continuait en parallèle ses travaux de botanique systématique et il publia, entre 1970 et 1985, dix « Contributions à l’étude de la flore corse » dans la revue suisse Candollea. Plusieurs taxons nouveaux pour la science furent décrits (Adenostyles briquetii, Erigeron paolii, Seseli djianeae, Trisetum conradiae, etc.) et de nombreuses espèces inconnues en Corse furent découvertes. Ces herborisations lui permirent de réaliser en 1985 une première mouture du Catalogue des plantes vasculaires de la Corse. Cette synthèse sera ensuite approfondie et complétée dans une seconde édition publiée en 95 Hommage à Jacques GAMISANS 1993 avec son complice Daniel Jeanmonod, conservateur aux Conservatoire et Jardin botaniques de la ville de Genève. À partir de 1986, il présida le comité scientifique « Flore corse », qui permit d’achever le monumental Prodrome de la flore corse initié par John Briquet, grâce à la publication de plusieurs monographies de familles botaniques. Sa collaboration fructueuse avec D. Jeanmonod aboutit à la réalisation du Flora Corsica, la première flore complète moderne de la dition publiée en 2007 et mise à jour dans une seconde édition parue en 2013, un ouvrage remarquable par sa clarté et sa concision. Ses activités floristiques n’ont pas faibli comme le montre la description en 2011 avec Laetitia Hugot, directrice du Conservatoire botanique national de Corse, d’un remarquable endémique du massif du Cintu, Hippocrepis conradiae. Au cours des années 1980 et 1990, Jacques Gamisans étendit ses études phytoécologiques corses aux zones de basse et moyenne altitudes, avec plusieurs travaux détaillés de cartographie de la végétation effectués dans le Niolu, le Haut-Venacais, les réserves naturelles de la presqu’île de Scandola et de l’étang de Biguglia, les îles Lavezzi et Cerbicales, etc. Plusieurs travaux furent aussi consacrés à l’étude de la distribution et de la dynamique des divers types de forêts ou de ligneux caractéristiques des paysages végétaux de Corse (Pin laricio, Aulnes, Genévrier thurifère, etc.). Tout ce corpus approfondi de relevés détaillés facilita la rédaction de la première monographie phytoécologique complète de l’île, La Végétation de la Corse, publiée en 1991 et rééditée en 1999. Si ses recherches se sont presque exclusivement focalisées en Corse, Jacques Gamisans a participé à d’autres travaux en Méditerranée, notamment dans le cadre de programmes conduits avec des collègues du laboratoire de botanique et d’écologie méditerraMission dans la haute montagne corse en juillet 1971, avec Maurice Reille. En haut, au sommet du Monte Rotondo ; néenne de l’université d’Aix-Marseille. À l’étranger, en bas, dans les pozzines de Vaccaja. (Clichés DR) il aida ainsi, durant les années 1970, Gilles Bonin dans la réalisation de son doctorat d’État portant sur la végétation forestière de l’Italie méridionale, et avec Jean-Pierre Hébrard il étudia les forêts du nord de la Grèce corse. En 1981, tous deux publient un article intitulé « À pro(Épire, Macédoine et Thrace) lors de deux missions (1976- pos de certaines espèces de la flore corse menacées de dis1977) financées par le Cnrs. Mais c’est en France méridionale parition », prémices d’actions plus importantes initiées par le (Provence, Drôme, Cévennes) et dans les Pyrénées, notamment parc naturel régional de Corse, le Conservatoire botanique de avec son collègue et ami Michel Gruber, lui aussi maître de Porquerolles et l’Agenc, dans le cadre du programme européen conférences à l’université d’Aix-Marseille, que Jacques Gami- Medspa en 1989-1993, « Inventaire permanent et protection des sans fut le plus actif, hors de sa terre de prédilection. Enfin, il plantes menacées, rares ou endémiques de la Corse » ; Jacques participa à plusieurs études phytochimiques sur les activités Gamisans participa activement à ce projet pilote et novateur à antioxydantes et sur les corps gras de divers végétaux. l’échelle de la Méditerranée qui permit d’élaborer les premières Cette activité soutenue s’est traduite par la publication listes rouges de la flore menacée et diverses actions de conserde 170 articles scientifiques dont 8 ouvrages principaux, et de vation in situ ou ex situ des végétaux les plus vulnérables. nombreux rapports d’études. Lors de la création en 2007 du Conservatoire botanique Sans doute au contact de Marcelle Conrad, qu’il avait national de Corse, Jacques Gamisans devint tout naturellement rencontrée dès 1966 à Vizzavona, Jacques Gamisans s’est président du conseil scientifique qu’il présida jusqu’en 2014. Il rapidement attaché à la préservation du patrimoine végétal siégea aussi de longues années au conseil scientifique régional 96 ecologia mediterranea – Vol. 41 (2) – 2015 Hommage à Jacques GAMISANS Les principaux ouvrages de Jacques Gamisans BRUN B., BRUN L., CONRAD M. & GAMISANS J., 1975. La Nature en France : Corse. Horizons de France, Paris : 224 p. GAMISANS J., 1985. Catalogue des plantes vasculaires de la Corse. Précédé de données statistiques et d’un exposé synthétique sur l’origine de cette flore et son organisation en ensembles de végétation. Parc naturel régional de la Corse, Ajaccio : 231 p. GAMISANS J., 1991. La végétation de la Corse. In : Compléments au Prodrome de la flore corse. Annexe 2. Éd. Conservatoire & Jardin botaniques de Genève : 391 p. (seconde édition publiée en 1999 parÉdisud, Aix-en-Provence). Jacques Gamisans lors de l’excursion du colloque international « Connaissance et conservation de la flore des îles de la Méditerranée », octobre 1993. (Cliché F. Médail) du patrimoine naturel (Csrpn) de Corse et au Conseil scientifique du parc naturel régional de Corse. Par ses écrits et ses conférences qu’il donnait régulièrement en été, il a su sensibiliser la population locale, les touristes, les administrations et les politiques à la préservation de ce patrimoine naturel unique, n’hésitant pas à fustiger certaines pratiques délétères : « Dans un pays méditerranéen montagneux comme la Corse, aux fortes pentes, la gestion par le feu est actuellement une aberration écologique qui conduit à une érosion irréversible des sols, à la dominance des pyrophytes et à une notable perte de valeur pastorale et de biodiversité » (Gamisans, 2010, p. 327). Jacques Gamisans avait à cœur de fournir au grand public une information botanique de qualité et il publia plusieurs ouvrages qui connaissent de grands succès éditoriaux : La Flore endémique de la Corse (1996, 2014), Le Paysage végétal de la Corse, publié en 2010 et la Flore des maquis de Corse et des végétations associées, parue en 2014. Durant près de cinquante années d’activités ininterrompues en Corse, Jacques Gamisans a réalisé une œuvre considérable qui a permis d’asseoir définitivement les connaissances sur la flore vasculaire et la typologie de la végétation de l’île. Sa modestie et sa discrétion étaient au service de sa passion pour la montagne méditerranéenne et pour ce fier pays catalan dont il suivait avec bonheur les évolutions identitaires. GAMISANS J. & JEANMONOD D., 1993. Catalogue des plantes vasculaires de la Corse (ed. 2). In : Compléments au Prodrome de la flore corse, Annexe 3. Éd. Conservatoire & Jardin botaniques de Genève : 258 p. GAMISANS J. & MARZOCCHI J.-F., 1996. La Flore endémique de la Corse. Édisud, Aix-en-Provence : 207 p. (seconde édition publiée en 2014 par Édisud, Saint Rémy-deProvence). JEANMONOD D. & GAMISANS J., 2007. Flora Corsica. Édisud, Aixen-Provence : 1 008 p. (seconde édition publiée en 2013 dans le Bull. Soc. Bot. Centre-Ouest, numéro spécial 39 : 1 072 p.). GAMISANS J., 2010. Le Paysage végétal de la Corse. Albiana, Ajaccio : 343 p. GAMISANS J., 2014. Flore des maquis et des végétations associées de Corse. Albiana, Ajaccio : 303 p. La liste complète des publications de Jacques Gamisans figure sur son site internet : http://jacquesgamisans.blogspot.fr Frédéric Médail, 29 septembre 2015. ecologia mediterranea – Vol. 41 (2) – 2015 97 Fabrication : 04250 Turriers www.transfaire.com www.naturalia-publications.com Achevé d’imprimer : février 2016 00-ecol-med-vol41(2)-couv-corTC_Mise en page 1 27/01/16 08:33 Page2 ecologia mediterranea Éditrices en chef : Dr Élise Buisson et Dr Brigitte Talon UMR CNRS IRD IMBE Université d’Avignon, IUT Site Agroparc, BP 1207 84911 Avignon cedex 09 France Instructions aux auteurs ecologia mediterranea publie des articles de recherche originaux sur des sujets se rapportant à l’écologie fondamentale ou appliquée des régions méditerranéennes. La revue exclut les articles purement descriptifs ou de systématique. ecologia mediterranea privilégie les domaines scientifiques suivants : bioclimatologie, biogéographie, biologie de la conservation, biologie marine, biologie des populations, écologie des communautés, écologie forestière, écologie génétique, écologie marine, écologie microbienne, écologie du paysage, écologie de la restauration, écologie végétale et animale, écophysiologie, paléoclimatologie, paléoécologie. La revue accepte également des articles de synthèse, des notes/communications courtes, des comptes rendus d’ouvrages, des résumés de thèses, ainsi que des commentaires sur les articles récemment parus dans ecologia mediterranea. La revue publie aussi des actes de colloques faisant l’objet d’un numéro spécial. Dans ce cas, prendre contact avec les éditrices. Comité éditorial Dr Pierre CHEVALDONNÉ, CNRS, Université Aix-Marseille, Marseille, France Dr Marc CHEYLAN, EPHESS, Montpellier, France Dr Cécile CLARET, Université AixMarseille, Marseille, France Dr Bruno FADY, INRA, Avignon, France Pr Thierry GAUQUELIN, Université AixMarseille, Marseille, France Dr Grant WARDELL-JOHNSON, Université Western, Australie Dr Raphaël GROS, Université AixMarseille, Marseille, France Dr Frédéric GUITER, Université AixMarseille, Marseille, France Pr Serge KREITER, SupAgro, Montpellier, France Dr Audrey MARCO, École nationale supérieure du paysage, Marseille, France Pr Frédéric MÉDAIL, Université AixMarseille, Marseille, France Pr François MESLÉARD, Université Avignon-Tour du Valat, France Dr Tom PARKER, San Francisco State University, États-Unis Dr Philippe PONEL, CNRS, Université AixMarseille, Marseille, France Dr Roger PRODON, EPHE, Montpellier, France Dr Sandra SAURA-MAS, Autonomous University of Barcelona, Espagne Dr Isabelle SCHWOB, Université AixMarseille, Marseille, France Dr Thekla K. TSITSONI, Aristotle University of Thessaloniki, Grèce Dr Éric VIDAL, IRD, France Dr Mercedes VIVAS, Universidad of Concepción, Chili Dr Ioannis VOGIATZAKIS, Open University of Cyprus, Chypre Les manuscrits sont soumis à des lecteurs spécialistes du sujet. La décision finale d’accepter ou de refuser un article relève des éditrices. L’article proposé doit être envoyé en version électronique à [email protected] (version doc(x) ou rtf). Pour la mise en forme du document, voir les instructions qui suivent. Une fois leur article accepté, les auteurs devront tenir compte des remarques des lecteurs, puis ils renverront leur texte corrigé sous deux mois toujours sous format électronique (doc(x) ou rtf). Passé ce délai, la seconde version sera considérée comme une nouvelle proposition. TYPES DE MANUSCRIT À préciser sur la première page lors de la soumission d’un manuscrit. Article de recherche : contribution inédite découlant d’une étude complète. Ce type d’article fait typiquement une vingtaine de pages et environ 6 000 à 8 000 mots. Note/communication courte : observation nouvelle ou rapport d’expérience dans un contexte pertinent avec les sujets visés par la revue. Ce type d’article fait typiquement une dizaine de pages et environ 3 000 à 4 000 mots. Article de synthèse : revue critique et originale de sujets spécifiques d’actualité ou d’un champ de recherche de pointe dans le domaine de l’écologie méditerranéenne. Ce type d’article fait typiquement une vingtaine de pages et environ 6 000 à 8 000 mots. Commentaire : avis sur des sujets déjà publiés dans ecologia mediterranea ou réflexion critique sur des problèmes d’intérêt général en écologie méditerranéenne. Ce type d’article fait typiquement une à cinq pages et environ 1 000 à 3 000 mots. Compte rendu d’ouvrage : revue critique d’ouvrages (livres, monographies, manuels, etc.) dans le domaine de l’écologie méditerranéenne. Les auteurs d’ouvrages souhaitant voir un compte rendu publié dans ecologia mediterranea doivent envoyer un exemplaire de l’ouvrage en question aux éditrices en chef. Résumé de thèse : résumé d’une thèse soutenue récemment dans le domaine de l’écologie méditerranéenne. Auteur, année, titre, spécialité et université (e.g. Thèse de doctorat en écologie soutenue le 3 avril 2012 à l’université de Padoue, Italie, laboratoire xxx), composition du jury, mots clés, résumé de 1 000 mots maximum. TEXTE Les articles (dactylographiés en Times 12, double interligne, format A4) doivent être rédigés en anglais ou en français. Si l’article soumis n’est pas rédigé en anglais, il est demandé (en plus des résumés) une version anglaise abrégée ainsi qu’une traduction en anglais des titres des figures et tableaux. L’article doit être complet : type de manuscrit, titres anglais et français, auteur(s) et adresse(s), résumés en anglais et en français (au minimum), version anglaise abrégée (si le texte n’est pas en anglais), mots clés anglais et français, texte, puis remerciements, bibliographie, liste des titres des figures et tableaux puis les figures et tableaux (un(e)/page). Ainsi, pour la soumission du manuscrit, les illustrations seront intégrées au document et non envoyées ISSN 0153-8756 http://ecologia-mediterranea. univ-avignon.fr ABONNEMENT (contact : [email protected]) France Europe Monde Abonnement à adresser à : SARL Transfaire Immeuble Wanad F-04250 TURRIERS Frais de port Total 60 € 60 € 60 € 6€ 12 € 16 € 66 € 72 € 76 € Code banque 19106 Code guichet 00839 numéro de compte 13995626000 Domiciliation : CA SISTERON IBAN : FR76 1910 6008 3913 9956 2600 062 BIC : AGRIFRPP891 PREMIÈRE PAGE La première page contient : 1) le type de manuscrit visé (article de recherche, communication courte, etc.) ; 2) le titre de l’article ; 3) le nom et prénom des auteurs ; 4) l’adresse de chaque auteur sera indiquée avec le courriel de l’auteur pour la correspondance. Dans le cas où la publication est le fait de plusieurs auteurs, il doit être précisé lors du premier envoi la personne à qui doit être retourné l’article après lecture ; 5) le nombre total de mots de l’introduction à la bibliographie. RÉSUMÉS, MOTS CLÉS ET VERSION ABRÉGÉE Les résumés doivent comporter 300 mots au maximum et la version abrégée (français si le manuscrit est en anglais et anglais si le manuscrit est en français) 1 000 mots (environ une page). Le nombre de mots clés est limité à six, dans la langue des résumés ; ils ne doivent généralement pas figurer dans le titre. BIBLIOGRAPHIE La bibliographie regroupera toutes les références citées et elles seules. Les références seront rangées dans l’ordre alphabétique des auteurs et de façon chronologique. Les abréviations internationales des titres des revues doivent être utilisées (ISI Journal Abbreviations Index). Vérifier attentivement le manuscrit pour s’assurer que toutes les références citées dans le texte apparaissent bien en bibliographie et inversement. Article Andow D.A., Karieva P., Levin S.A. & Okubo A., 1990. Spread of invading organisms. J. Ecol. 4: 177-188. Ouvrage Harper J.L., 1977. Population biology of plants. Academic Press, London, 300 p. Chapitre d’ouvrage May R.M., 1989. Levels of organisation in ecology. In: Cherret J.M. (ed.), Ecological concepts. Blackwell Scientific Public, Oxford: 339-363. Acte de conférence Grootaert P., 1984. Biodiversity in insects, speciation and behaviour in Diptera. In: Hoffmann M. & Van der Veken P. (eds.), Proceedings of the symposium on “Biodiversity: study, exploration, conservation”. Ghent, 18 November 1992: 121-141. Rapport et thèse Jaouadi W., 2011. Écologie et dynamique de régénération de l’Acacia tortilis (Forsk.) Hayne subsp. raddiana (Savi) Brenan var. raddiana dans le parc national de Bouhedma (Tunisie). Thèse de doctorat de l’Institut national agronomique de Tunisie, 180 p. Editors-in-Chief: Dr Élise Buisson & Dr Brigitte Talon UMR CNRS IRD IMBE Université d’Avignon, IUT Site Agroparc, BP 1207 84911 Avignon cedex 09 France ecologia mediterranea publishes original research reports and syntheses in the fields of fundamental and applied ecology of Mediterranean areas, except for descriptive articles or articles about systematic. The editors of ecologia mediterranea invite original contributions in the fields of: bioclimatology, biogeography, conservation biology, marine biology, population biology, community ecology, forest ecology, marine ecology, genetic ecology, landscape ecology, microbial ecology, restoration ecology, plant and animal ecology, ecophysiology, palaeoecology, palaeoclimatology. The journal also publishes reviews, short communications, book reviews, Ph. D. thesis abstracts and comments on papers recently published in the journal. ecologia mediterranea invite conference organizers to get in touch with the editors for special issues as part of conference/symposium proceedings. Manuscripts are peer-reviewed by appropriate referees. The final decision to accept or reject the manuscript is made by the editors. To submit a paper, please send an electronic version of your paper to [email protected] (doc(x) or rtf). Please read the following guidelines to prepare your manuscript. When the article is accepted, the authors should take reviewer’s comments into consideration. They will send back to the journal Editorial Office, within 2 months. After this deadline, the manuscript will be considered as a new submission. Editorial Board Dr Pierre CHEVALDONNÉ, CNRS, Université Aix-Marseille, Marseille, France Dr Marc CHEYLAN, EPHESS, Montpellier, France Dr Cécile CLARET, Université AixMarseille, Marseille, France Dr Bruno FADY, INRA, Avignon, France Pr Thierry GAUQUELIN, Université AixMarseille, Marseille, France Dr Grant WARDELL-JOHNSON, Université Western, Australie Dr Raphaël GROS, Université AixMarseille, Marseille, France Dr Frédéric GUITER, Université AixMarseille, Marseille, France Pr Serge KREITER, SupAgro, Montpellier, France Dr Audrey MARCO, École nationale supérieure du paysage, Marseille, France Pr Frédéric MÉDAIL, Université AixMarseille, Marseille, France Pr François MESLÉARD, Université Avignon-Tour du Valat, France Dr Tom PARKER, San Francisco State University, États-Unis Dr Philippe PONEL, CNRS, Université AixMarseille, Marseille, France Dr Roger PRODON, EPHE, Montpellier, France Dr Sandra SAURA-MAS, Autonomous University of Barcelona, Espagne Dr Isabelle SCHWOB, Université AixMarseille, Marseille, France Dr Thekla K. TSITSONI, Aristotle University of Thessaloniki, Grèce Dr Éric VIDAL, IRD, France Dr Mercedes VIVAS, Universidad of Concepción, Chili Dr Ioannis VOGIATZAKIS, Open University of Cyprus, Chypre ISSN 0153-8756 Les mots « figures » et « tableaux » annoncés dans le texte sont écrits en toutes lettres et en minuscules. Indiquer le nom d’auteur et l’année de publication (mais indiquer tous les auteurs dans la bibliographie). Exemples : “Since Dupont (1962) has shown that...”, or “This is in agreement with previous results (Durand et al. 1990; Dupond & Dupont 1997) ...”. Le numéro de page de la citation n’est mentionné que dans le cas où elle est entre guillemets. Si la publication est écrite par plus de deux auteurs, le nom du premier doit être suivi par et al. http://ecologia-mediterranea. univ-avignon.fr L’usage d’un acronyme ou d’une abréviation technique doit être précédé de sa signification lors de sa première apparition. Les codes de nomenclature doivent être respectés selon les conventions internationales. Les mots latins doivent être mis en italiques (et al., a priori, etc.), et en particulier les noms de plantes ou d’animaux. Lors de la première apparition du nom d’une espèce, il est demandé d’y faire figurer le nom d’auteur (exemple : Olea europaea L.). FIGURES ET TABLEAUX clé RIB 62 Guidelines for authors CITATIONS ET RENVOIS APPELÉS DANS LE TEXTE ABRÉVIATIONS, NOMENCLATURE ET MOTS LATINS 1 an = 2 numéros Abonnement séparément. Le texte des articles de recherche doit comporter quatre parties non numérotées : introduction, méthodes, résultats, discussion. Les sous-titres ne sont pas numérotés. Par exemple : Introduction Méthodes Site d’étude Plan expérimental Analyses des données Résultats Discussion L’emploi de mots soulignés est à proscrire. Les noms d’auteurs cités figureront en minuscules dans le texte comme dans la bibliographie. En français, n’utilisez les majuscules que pour les noms propres, sauf exception justifiée. Les ponctuations doubles ( : ; ? ! ) sont précédées d’un espace, contrairement aux ponctuations simples ( , . ). En revanche, toutes les ponctuations sont suivies d’un espace. La mise en forme définitive du texte sera assurée par la revue. ecologia mediterranea Pour la soumission du manuscrit, les illustrations seront intégrées au document, à la suite des références bibliographiques (voir la partie « Texte »). Une fois l’article accepté, les illustrations doivent être envoyées séparément du texte, prêtes à l’impression (résolution 300 dpi). Tous les documents devant être insérés doivent être annoncés dans le texte, numérotés dans l’ordre d’apparition et légendés. TIRÉS À PART Il n’est plus fourni de tirés à part mais un pdf par article. TYPES OF MANUSCRIPTS Please, mention the type of manuscript you are submitting on the first page of your submission. Research article: research report of broad scope that is an original contribution to science. The typical length of research papers is about 6,000 to 8,000 words. Short communication: brief report of new observations or of important findings that deserve publication before broader studies are completed. The typical length of short communications is about 3,000 to 4,000 words. Reviews: critical appraisal of broad areas of investigation or research in Mediterranean ecology. The typical length of reviews is about 6,000 to 8,000 words. Commentaries: opinion on topics recently published in ecologia mediterranea or essays on topics of general interest in Mediterranean ecology. The typical length of short communications is about 1,000 to 3,000 words. Book review: critical appraisal of a Book interesting to the readers of ecologia mediterranea. Editors who wish to see their book reviewed in ecologia mediterranea should get in touch with the editors of the journal. Ph. D. thesis abstracts: abstract of the Ph. D. thesis recently defended in the fields covered by ecologia mediterranea. Author, Year, Title, University (e.g. Ph. D. thesis defended on April, 4th 2014 at University of California, Irwin, USA, Lab xxx), Composition of board, Keywords, Abstract (1,000 words max.). MANUSCRIPT PREPARATION Manuscripts (typewritten Times 12, with double line spacing) must be written in English or in French. If the language is not English, you should join an English short version and English titles of figures and tables. The manuscript must be complete: e.g. title in English and French, author(s) and address(es), abstract in English and French, an English short version (if English is not the language used in the article), keywords in English and French, text, acknowledgements, references, figures and tables (one / page). For the submission, figures and tables are integrated in the SUBSCRIPTION (contact: [email protected]) 1 year = 2 issues word document and not sent separately. For research papers, the text should normally consist of 4 sections: introduction, methods, results, discussion which are not numbered. Subtitles are not be numbered either: Introduction Methods Study site Experimental design Statistical analyses Results Discussion Use lower-case letter type for names in the text and in the reference section. Do not underline any word. In English, there is one blank after any punctuation, never before. Copy editing of manuscripts is performed by the journal. FIRST PAGE The first page contains: 1) the type of manuscript (research paper, short communication, etc.); 2) the title of the article; 3) the names of the authors; 4) the address of each author + an email at least of the corresponding author; 5) the total number of words from the introduction to the references included. ABSTRACTS, KEYWORDS AND SHORT VERSION Abstracts should be no longer than 300 words. The English short version should not exceed one page long (1,000 words). Do not use more than six keywords. Keywords should not be present in the title. REFERENCES All publications cited in the text should be presented in a list of references following the text of the manuscript. The list of references should be arranged alphabetically on author’s names, and chronologically for each author. You should abbreviate the titles of periodicals mentioned in the list of references (except if you are not sure of it; ISI Journal Abbreviations Index). Check the manuscript to make sure that all references are cited and that all citations in the text are included in the references. Use following system to write the references: Journal article Andow D.A., Karieva P., Levin S.A. & Okubo A., 1990. Spread of invading organisms. J. Ecol. 4: 177-188. Book Harper J.L., 1977. Population biology of plants. Academic Press, London, 300 p. Book chapters May R.M., 1989. Levels of organisation in ecology. In: Cherret J.M. (ed.), Ecological concepts. Blackwell Scientific Public., Oxford: 339-363. Conference proceedings Grootaert P., 1984. Biodiversity in insects, speciation and behaviour in Diptera. In: Hoffmann M. & Van der Veken P. (eds.), Proceedings of the symposium on “Biodiversity: study, exploration, conservation”. Ghent, 18 November 1992: 121-141. IN-TEXT CITATIONS The words “figure” and “table” announced in-text should be written in extenso and with lower-case letter type. In the text refer to the author’s name and year of publication (followed by pages only if it is a quotation). If a publication is written by more than two authors, the name of the first author should be used followed by “et al.” (this indication, however, should never be used in the list of references: first author and co-authors should be mentioned in it). Examples: “Since Dupont (1962) has shown that…”, or “This is in agreement with previous results (Durand et al. 1990; Dupond & Dupont 1997)…”. ABBREVIATES, NOMENCLATURE AND LATIN WORDS Subscription Postage Total 60 € 60 € 60 € 6€ 12 € 16 € 66 € 72 € 76 € France Europe World Explanation of a technical abbreviate is required when the first use. International convention codes for nomenclature should be used. Latin words should be in italic (et al., a priori, etc.), particularly for plants or animals’ denomination (the first time, please precise author’s name: for example, Olea europaea L.). FIGURES AND TABLES Mail this subscription to: SARL Transfaire Immeuble Wanad F-04250 TURRIERS Bank: CA SISTERON IBAN: FR76 1910 6008 3913 9956 2600 062 BIC: AGRIFRPP891 All illustrations should be included in the word document after the section “References”. Once the paper is accepted for publication, illustrations should be sent separately (3,000 dpi). All the illustrations should be cited and should have a legend. REPRINTS A pdf version will be supplied free of charge for each paper. 6 mm (96 p) ecologia mediterranea Vol. 41 (2) – 2015 00-ecol-med-vol41(2)-couv-corTC_Mise en page 1 27/01/16 08:33 Page1 Vol. 41 (2) – 2015 Sommaire – Contents Éditorial – Editorial .......................................................................................... 3 MEDECOS special issue Seed Bank Divergence Between Arctostaphylos Adans. (Ericaceae) and Ceanothus L. (Rhamnaceae) Suggests Different Seed Predator Interactions Developing Allometric Volume-Biomass Equations to Support Fuel Characterization in North-Eastern Spain V. THOMAS PARKER ............................................................................................... Studying Shoot and Root Architecture and Growth of Quercus ithaburensis subsp. macrolepis Seedlings; a Key Factor for Successful Restoration of Mediterranean Ecosystems T. TSITSONI, N. GOUNARIS, A. B. KONTOGIANNI, V. XANTHOPOULOU-TSITSONI 33 ....................................................... Creation of an Integrated System Model for Governance in Urban MTEs (Mediterranean-Type Ecosystems) and for Adapting Cities to Climate Change – Preliminary Results T. TSITSONI, M. TSAKALDIMI, M. GOUSIOPOULOU ......................................................... ....................... Vol. 41 (2) – 2015 Revue internationale d’écologie méditerranéenne International Journal of Mediterranean Ecology 5 15 B. D. PEDRA, J. GODOY PUERTAS, L. FUENTES LOPEZ ecologia mediterranea 25 Vegetation Dynamics of Coastal Dunes with Juniperus spp. in Crete, Gavdos and Chrysi Islands (Greece) Caractérisation du fonctionnement des steppes d’Alfa marocaines par la méthode de l’analyse fonctionnelle du paysage ............ 45 M. DERAK, F. T. MAESTRE, J. L. QUERO, V. OCHOA, C. ESCOLAR, S. SOLIVERES, P. GARCÍA-PALACIOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Floristic Diversity Patterns in the Beni-Haoua Forest (Chlef, Algeria) 61 ...................................... Insight into the Dietary Habits of the Eurasian Otter, Lutra lutra, in the East of Algeria (El-Kala National Park) 73 ............................................................................ Résumés de thèses – Ph. D summaries 85 .......................................................................................... Hommage à Jacques Gamisans 92 .......................................................................... 95 P.DELIPETROU, D. GHOSN, G. KAZAKIS, P. NYKTAS, E. REMOUNDOU, I.N. VOGIATZAKIS A. ABABOU, M. CHOUIEB, A. BOUTHIBA, D. SAIDI, K. MEDERBAL R. LIBOIS, R. GHALMI, A. BRAHIMI C. CALVET, T. FRÉJAVILLE Revue indexée dans Pascal-CNRS et Biosis ISSN 0153-8756 ecologia mediterranea RESEARCH PAPERS Editors-in-Chief: Dr Élise Buisson & Dr Brigitte Talon Institut méditerranéen de biodiversité et écologie (IMBE) Mediterranean Institute of Biodiversity and Ecology Naturalia Publications