Climate change, hydrology and fish morphology: Predictions using phenotype-environment associations

Background/Question/Methods

In response to environmental change, a population may: 1) disperse to new areas containing suitable habitat conditions, 2) adapt to the new environment, or 3) suffer population declines and risk local extirpation.  While many models have been introduced that examine the new habitat conditions to which a species could disperse in the face of environmental change (e.g., species distribution models), fewer models consider the ability of populations to adapt to these changing conditions.  Such models are especially critical for species that are dispersal limited.

Species distribution models examine the relationship between species occupancy and environmental variables and then use this relationship to determine areas which contain suitable environmental conditions to which species can disperse.  We extend this framework to predict the ability of populations to adapt to environmental change by examining the relationship between phenotype and the environment.  These phenotype-environment associations are then used to determine the phenotypes the population must attain in order to persist within a habitat with altered environmental conditions.  We apply this framework to examine the adaptability of stream fish populations in the Midwestern United States to future changes in flow rates caused by global climate change. 

Results/Conclusions

Our analyses indicate that the mean body shape of populations of two species, the Blackside Darter (Percina maculata) and the Stonecat (Noturus flavus), are significantly associated with flow, as individuals found in high-flow habitats have a more streamlined body shape than individuals from low-flow habitats.  Hydrologic models based on future temperature and precipitation data from regional climate models generally predict subbasin-level declines in flow rate in the years 2051 – 2060 relative to current flow rates.  Thus, stream fish should generally evolve less-streamlined body shapes in the future. 

To examine if fish populations are able to adapt to these hydrologic changes, we simulated trait evolution of body shape using a quantitative genetics model of linear reaction norms.  Using these simulations, we determined the combination of selection strength and phenotypic plasticity necessary for each population to reach the expected body shape.  These simulations revealed that the adaptability of some populations to future changes in flow rate requires strong and potentially unrealistic levels of selection and plasticity.  Consequently, these species may incur substantial demographic costs across populations based on future streamflow regimes.