The effects of connectivity, management and climate scenarios on river fish communities - SEINARIOS

Swann Felin defended his doctoral thesis on 27/03/2025. 

In the face of global change, one of the major challenges for contemporary societies will be the preservation of biodiversity. In this context, it is necessary to identify the most vulnerable areas, especially within drainage basins and their river networks, very sensitive to these pressures. In order to inform public policy and help prioritize decision-making, it has become essential to better understand the consequences of management decisions and climate change scenarios on future biodiversity, which necessitates the study of projections (i.e. simulations) based on species models.

Summary 

River ecosystems are among the most species-diverse in the world. Global change, coupled with anthropization of watersheds and fragmentation of riverscapes, jeopardize river fish communities and put their future conservation in peril. There is an urgent need for knowledge to inform the design of the operations necessary to the conservation of freshwater ecosystems and fish species. However, the influence of watershed management, river fragmentation and land use on fish distribution and conservation is still poorly understood. This thesis aimed to address this knowledge gap by investigating the role of watershed management and fragmentation on fish species distribution in a warming 21st century, in order to provide information in support of river management and produce tools to better approach this issue. We first studied longitudinal connectivity and its influence on river fish communities, then we built watershed management scenarios which we integrated into species distribution models to study the influence of climate change and watershed management on potential future fish communities, before studying the spread of uncertainty between these two components.
The Seine-Normandie basin, a highly anthropized watershed in Northern France, was selected for this study. A dataset comprising over 2,000 study sites was compiled for the analysis of longitudinal connectivity. Novel connectivity indices were developed, based on functional traits of fish and river fragmentation, some indices accounting for local context (e.g. river size and river centrality within the basin). We were able to assess the relative importance of connectivity in fish distribution via species distribution modelling, indices which included local context performing especially well. We found that dams play a pivotal role in shaping fish distribution.
In order to model future fish distribution, four climate projections were selected across two time horizons, 2050 and 2100. In addition, we constructed four management scenarios to describe different future political orientations with respect to riverscape management. We were able to produce several prospects of species distribution through ensemble modeling. Most species were found to exhibit a decline in range over time due to climate change, management scenarios producing more complex outcomes. Scenarios linked with limited environmental restoration led to high rates of extirpation while scenarios with more ambitious prospects often resulted in enhanced survival rate. These responses exhibited wide variation across species, hindering the identification of a single scenario fitting the whole community.
Finally, we highlighted the relative importance of uncertainty linked to climate change and territorial management across eight species. We found that in most cases climate change carried more uncertainty than watershed management. However the portion of uncertainty carried by territorial management was still important, ranging between 6 % and 30 % on average depending on species. Some species (the bleak, roach) tolerant for warm water or environmental stressors associated with climate change were more impacted by territorial management in specific sites.
Through this PhD we provided insights into the role of connectivity on the composition of river fish communities as well as the influence of management and climate change on future river ecosystems. We were also able to provide new practical clues for future management and new methodologies for assessing connectivity in river networks and integrating it into modeling approaches. We produced plausible futures for river fish distribution, allowing us to assess the importance of climate change and territorial management on fish species conservation. We believe many of these approaches can be applied in similar contexts and contribute to the creation of knowledge necessary to implement the urgent measures required to safeguard river ecosystems in future decades.