Taxonomic sufficiency of epibenthic species in scientific bottom

Taxonomic sufficiency of epibenthic species in scientific bottom trawl surveys A. Brind’Amour, P. Laffargue, J. Morin, M-L. Cochard, S. Vaz Département Ressources Biologiques et Environnement, Ifremer 1 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Context « Identifying organism to the level of taxonomic resolution sufficient to satisfy the objectives of a study » (Ellis 1985) • Use higher-taxon level as surrogates for species • Historically used in « pollution » studies (Somerfield & Clarke 1995, Bowman & Bailey 1997) 2 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Context • Marine Strategy Framework Directive (2008/56/EC) : Achieve Good Environmental Status by 2020 Descriptors 1, 2, 4, 6 (macroinvertebrates) • European and institutional fish bottom trawl surveys routinely identify macro and mega-invertebrates epibenthic species (Brind’Amour et al. 2009, Laffargue & Martin 2010) Some taxa not identified at the species level (experts ) Misidentification of morphologically similar species Inachus dorsettensis O. albida O. ophiura Anapagurus hyndmanni Pagurus bernhardus Macropodia rostrata 3 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Objectives 1° Assess the taxonomic level to which epibenthic mega and macro-fauna can be studied without losing ecological information (community structure) 2° Compare the results from fish bottom trawl surveys conducted at different spatial scales (coastal bays to shelf) 3° Elaborate some recommandations (e. g. list of species) for which identification can be done at higher-taxon level 4 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Case studies North Bay of Seine (Nour. Seine) Coast Shelf Bay of Biscay (EVHOE) Bay of Vilaine (Nur. Vil) South Ecosystem vulnerability • Biarritz • October 18 -21 2011 © Ifremer North Sea (IBTS) European & French scientifc surveys 5

~ Methodology L. holsatus L. depurator *Done at the five levels: Species, Genus, Family, Order, Morphospecies Data* (density) + HCA (Ward) Assess the similarity betwen the classifications for the five levels Rand Index Ref. Tax (Aphia. ID) NMDS (Bray-Curtis) Total • 223 species in the 4 surveys • 42 morphospecies (i. e. species potentially misidentified) + individual species Ecosystem vulnerability • Biarritz • October 18 -21 2011 Community structure and the spatial distribution of the taxa 6

Results Community structure (n. MDS) Sediment and depth gradient 7 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Results Community structure (n. MDS) Sediment and depth gradient 8 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Results Comparaison (Rand Index) R= Bay of Biscay Ecosystem vulnerability • Biarritz • October 18 -21 2011 a+b+c+d 9

Results Comparaison (Rand Index) R= 33 27 24 14 24 58 51 39 20 38 130 105 79 37 82 Bay of Seine Bay of Vilaine Bay of Biscay a+b+c+d 151 122 98 43 91 North Sea 10 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Conclusions • Taxonomic sufficiency seems case-dependent (small vs large scale surveys): degrading the taxonomic information to genus/family less of a problem in species-dominant community (Seine) than in species-rich and evenly composed community (North Sea) • In large-scale surveys, misidentification of community structure can reach differences up to 70% (range 10 -70%) when taxa are studied at the genus level • Identification of morphospecies may be a “compromise” in largescale surveys encompassing multiple habitats • Desperate needs for benthic taxonomists (!) • Ongoing work Increase the number of datasets and different years to test the effect of species richness and evenness 11 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Thank you ! A special thank’s to all the scientists who participated to data collection and species identification (IBTS / EVHOE / Nour. Seine / Nur. Vil) 12 Ecosystem vulnerability • Biarritz • October 18 -21 2011

Results Low species richness and high dominance gen /sp fam/ ord/ sp sp gen fam/ ord/ High species richness and high eveness /sp sp sp Ecosystem vulnerability • Biarritz • October 18 -21 2011 13

~ Methodology L. holsatus L. depurator *Done at the five levels: Species, Genus, Family, Order, Morphospecies Data* (density) + Community structure and the spatial distribution of the taxa NMDS (Bray-Curtis) Ref. Tax (Aphia. ID) HCA (Ward) Total • 223 species in the 4 surveys • 42 morphospecies (i. e. species potentially misidentified) + individual species Assess the similarity betwen the classifications for the five levels Rand Index Informs on the taxa within each group and Ind. Val discriminant taxa 14 Ecosystem vulnerability • Biarritz • October 18 -21 2011