Humans are altering the global water cycle at an unprecedented rate, which will have far-reaching consequences for the ecology and evolution of entire ecosystems (e.g. changing aquatic-terrestrial linkages and altering genetic divergence). To understand how this may occur, we must first describe how genetic and phenotypic variation is distributed across landscapes. However, ecological niche models that describe distributions on the landscape often ignore intraspecific variation. With Populus angustifolia, a dominant riparian tree species that spans ~1700 km range in the Western U.S., we have a model system for incorporating genetics into ecological niche models. Our first goal was to describe the hydrological niche (gradients associated with the water cycle) of P. angustifolia in the western U.S. Our second goal was to determine if three major genetic provenances of the species are driven by the same environmental variables. We used environmental variables derived from WorldClim and NHDPlusV2, a collection of occurrence points from 17 populations spanning the tree range, and four Maxent models (entire species, three provenance models) to address these two goals. Our third goal was to determine if provenances are locally adapted by using common garden data on stomatal density and phenology.
Results/Conclusions
The results of the species model indicate that the hydrological niche of P. angustifolia is largely driven by precipitation seasonality, precipitation of the driest month of the year, and mean annual streamflow. Notably, mean annual temperature only contributed 5.1% to the model. The results of the three provenance models suggest that there is niche differentiation between the three genetic provenances of P. angustifolia. Provenance 3 is most similar to the species model, primarily driven by the same three variables, while Provenance 1 was explained most by temperature variables (mean temperature of the wettest quarter and mean temperature of the driest quarter). To address our final question, we found that genetic provenances show differences in the location (not total number) of stomata in both the field and common garden, suggesting local adaptation of this water-regulating trait. Additionally, streamflow, precipitation, and temperature interact to explain provenance differences in Spring phenology in the common garden. Our results support recent calls for ecological niche models to incorporate information on population genetic structure. Doing this should improve model predictions with climate change scenarios and allow us to make better-informed conservation decisions as populations diverge across landscapes.