Changes of macrofauna stable isotope compositions in a very inconstant seagrass detritic habitat: actual diet modification or baseline shift?

Decayed leaves of the Neptune grass Posidonia oceanica, detached and then exported during storms, constitute an important compartment in terms of organic matter transfer from the seagrass bed to the other habitats, particularly coastal habitats. These exported litter accumulations (ELA) support a diverse (more than 130 species) and abundant (up to 4900 id.m-2) vagile macrofauna (invertebrates > 500µm) assemblage which may play a key role in the degradation, enrichment and carbon transfer from P. oceanica dead material to coastal food chains. Indeed, preliminary results of vagile invertebrates gut content observations show that even if only a few of these species ingest a large proportion of P.oceanica dead leaves fragments, most of the others ingest a small but non-negligible part, suggesting a potential role of the whole community in the mechanical fragmentation of the dead leaves.

ELA are very dynamic habitats with highly variable food availability, quality, and composition. Such an inconstant habitat may result in drastic modifications of the invertebrate community but also of its trophic structure and ecology. To test this hypothesis of influence of pulsed availability, quality and composition of food sources on the vagile macrofauna diet, we took seasonal samples in Calvi Bay (Corsica, 8°45’E; 42°35’N), at two sites between August 2011 and May 2012. Stable isotopes analysis (C&N) were conducted at an individual level on dominant macrofauna species and mixing and isotopic niche model packages in R were used. Bayesian inference “SIBER” package highlighted significant seasonal and spatial differences of diet at the community, specific and even intraspecific level. Data confirm the potential transfer of seagrass material to animal tissues but in various proportions depending of the species and the season. But one question remained: are these variations reflection a true diet change, or only a spatiotemporal baseline variation of the food sources isotopic composition? “SIAR” Bayesian mixing model showed that it depends on the species and that the two responses co-occur. We emphasized the need to work at the specific level but also with an adequate temporal resolution for sampling to encompass diet and baseline variability.