An Ecological Time Series of Pacific Oyster (C. gigas) Growth and Survival: data curation and analysis of 24 years of monitoring along the French coast
|Author(s)||Normand Julien1, Dubroca Laurent1, Fleury Elodie1|
|Meeting||2019 Future Oceans2 IMBeR Open Science Conference, 17-21 June 2019, Brest, France|
As global change becomes a major concern, the need to contextualize discrete observations to highlight long-term evolutions also becomes more relevant. In response to this challenge, Marine Historical Ecology (MHE) has recently emerged as a new discipline within the marine sciences, dedicated to studying the relationship between long-term evolution of marine ecosystems and temporal variations of external predictors such as meteorological parameters, or descriptors of human activities. In recent years, therefore, considerable effort has been devoted to study long-term time series on fisheries and environmental data. This approach has increased our knowledge about natural variability and functioning of marine environments and has helped us to characterize breaking points and baseline shifts in these systems.
However, MHE studies were rarely planned when data acquisition has begun, and some information about the critical components of ecosystem functioning, although helpful, could be missing without turning back. In particular, very few long-term datasets on mollusk bivalves have been released, although these species are cultivated worldwide, could account for a large part of the biomass and also act as ecosystem engineers. From an ecological point of view, oysters also have the advantage to exhibit great phenotypic plasticity in response to environmental variability and to be sessile, making them perfect sentinels of the environment.
In this context, we have compiled dataset from three monitoring networks of oyster production coordinated by IFREMER, named REMORA, RESCO, and ECOSCOPA. These networks have monitored growth and mortalities rates for spat and adult Crassostrea gigas oysters reared in 13 locations along the French coastline during the 1993-2017 period. We modeled the evolution of mean individual weights and mortality rates as a function of time to cope with changes in data frequency acquisition during annual monitoring campaigns. We have thus produced standardized indicators associated with a detailed metadata file, in order to share them in an open data depository.
Oyster growth and survival appeared indeed very variable across years and sites. Annual variations explain a significant part of spatial mortality variations and site x year matrices depicted waves of mortalities coinciding with the massive-virus-associated-epizooty that hit spat oyster until 2008. On the contrary, adult oyster growth appeared to be more driven by site-specific environmental conditions. As a first approach to investigate the causal determinants of these variations, we used classification methods to rely on them with seasonal changes of phytoplankton and hydrological parameters, considering another restricted dataset (only nine years and 11 localizations) issued from the Ifremer-REPHY-hydrobiological monitoring network. This approach revealed unexpected relationships between environmental parameters and oyster traits variations, which strongly encouraged further explorations with a broader dataset of environmental descriptors.