Existing Aridity Gradient Mirrors Future Fish Assemblage Projections Under Climate Change Scenarios

Authors: Joshuah Perkin, Lindsey Elkins, Rebecca Mangold, Mariana Rocha, Astrid Schwalb, Ben Schwartz, Weston Nowlin, Karl Cottenie, Christina Saltus, Richard Johansen, David Smith

Year: 2022

Abstract

Anthropogenically-driven climate change combined with existing ecosystem degradation is projected to cause future losses of global freshwater biodiversity, particularly in arid and semi-arid areas within temperate climate regions. It is therefore necessary to understand both contemporary drivers of freshwater biodiversity loss as well as how future climatic conditions might affect humans and nature. For example, within aquatic ecosystems, streamflow and thermal regimes are fundamental regulators of ecosystem properties, but both are expected to change in response to climate change. Analysis of existing environmental gradients has the potential to aid in projecting climatic influences on a variety of organisms; however, few analyses have focused on freshwater fish assemblages. We present a multi-scale, spatiotemporal approach to predictive ecological modelling that ultimately demonstrates that an existing aridity gradient is a suitable proxy for freshwater fish assemblage response to climate change projections for the same region. We conducted our study using fish collections from 100 sampling reaches distributed across the central Colorado River basin of Texas. We combined fish assemblage surveys, local habitat characteristics, and remotely sensed geospatial riverscape data to (1) analyze spatial variation in fish-environment relationships under current conditions, and (2) used climate change projections for air temperature and precipitation to predict shifts in fish assemblage structure under multiple emissions scenarios. Our results revealed that spatial shifts in fish assemblage structure along an existing aridity gradient mirrored the modelled assemblage-level shifts under climate change projections for the emissions scenarios (or “Representative Concentration Pathways”) 4.5 and 8.5 projected through 2100. The model predicts a general shift towards invasive, warm-water assemblages and the potential loss of endemic, cool-water-dependent species. Our framework underscores the importance of multiscale, spatiotemporal modelling approaches that consider multiple dimensions of the total environment while assessing patterns and predictors of ecological change.