1 resultado para Linear mixed models

em Archimer: Archive de l'Institut francais de recherche pour l'exploitation de la mer


Relevância:

100.00% 100.00%

Publicador:

Resumo:

For decades, global climate change has directly and indirectly affected the structure and function of ecosystems. Abrupt changes in biodiversity have been observed in response to linear or sudden modifications to the environment. These abrupt shifts can cause long-term reorganizations within ecosystems, with communities exhibiting new functional responses to environmental factors. Over the last 3 decades, the Gironde estuary in southwest France has experienced 2 abrupt shifts in both the physical and chemical environments and the pelagic community. Rather than describing these shifts and their origins, we focused on the 3 inter-shift periods, describing the structure of the fish community and its relationship with the environment during these periods. We described fish biodiversity using a limited set of descriptors, taking into account both species composition and relative species abundances. Inter-shift ecosystem states were defined based on the relationship between this description and the hydro-physico-chemical variables and climatic indices defining the main features of the environment. This relationship was described using generalized linear mixed models on the entire time series and for each inter-shift period. Our results indicate that (1) the fish community structure has been significantly modified, (2) environmental drivers influencing fish diversity have changed during these 3 periods, and (3) the fish-environment relationships have been modified over time. From this, we conclude a regime shift has occurred in the Gironde estuary. We also highlight that anthropogenic influences have increased, which re-emphasizes the importance of local management in maintaining fish diversity and associated goods and services within the context of climate change.