Northern Arizona University Flagstaff, Arizona, United States
Background/Question/Methods In order to understand the impacts of climate change on natural ecosystems, we must be able to predict species’ responses to shifts in their environment. Phenotypic responses include shifts in mean traits as well as the direction and magnitude of trait plasticity, which can cascade to affect numerous dependent species within complex community networks. These impacts are especially strong in species of high ecosystem consequence such as foundation species. Here, we evaluate differences in trait means and plasticity in a foundation riparian species of the southwestern US, Populus fremontii (Fremont cottonwood), to combined biotic and abiotic stresses. We then seek to understand whether these responses can be predicted from the past climate to which these populations have evolved. Using a replicated common garden design combined with a simulated herbivory treatment, we were able to disentangle genetic (G), environmental (E), and G x E effects of climate transfers and herbivory damage on morphological, phenological, and phytochemical traits among 16 populations of Fremont cottonwoods collected throughout Arizona.
Results/Conclusions We found that: 1) There were strong genetic effects for most traits within the three common gardens as well as for phenotypic plasticity across gardens. 2) These trait differences led to reciprocal differences in growth and survival, indicating local adaptation to provenance climates. 3) The magnitude of plasticity depended on the trait measured and a population’s source climate. Whereas some traits showed higher plasticity in populations from hot climates, others showed the opposite pattern, 4) Tradeoffs between growth and defense may explain differences in plastic responses to abiotic vs. biotic stress. 5) Traits and trait plasticities show evidence of climate-driven divergent selection based on QST-FST analysis and climate-phenotype regressions. Overall, these common garden studies demonstrate large genetic, environmental, and GxE interactions for many important traits. In the Southwest, where warming temperatures and increasingly variable precipitation patterns have already impacted riparian forests, predicting changes in tree traits under climate change is essential for successful management into the future. Although genetic divergence in both mean traits and plastic responses across a steep climate gradient provide some predictive power, additional selection pressures from insect attacks and changing streamflow require a multivariate approach to understanding the links between past and future performance.