Heat tolerance predicts the importance of species interaction effects as the climate changes.

Lacy Chick,Clint A Penick,Aaron M Ellison,Robert R Dunn,Lauren M Nichols,Sara Helms Cahan,Nicholas J Gotelli,Nathan J Sanders,Sarah E Diamond

doi:10.1093/icb/icx008

Abstract

Few studies have quantified the relative importance of direct effects of climate change on communities versus indirect effects that are mediated thorough species interactions, and the limited evidence is conflicting. Trait-based approaches have been popular in studies of climate change, but can they be used to estimate direct versus indirect effects? At the species level, thermal tolerance is a trait that is often used to predict winners and losers under scenarios of climate change. But thermal tolerance might also inform when species interactions are likely to be important because only subsets of species will be able to exploit the available warmer climatic niche space, and competition may intensify in the remaining, compressed cooler climatic niche space. Here, we explore the relative roles of the direct effects of temperature change and indirect effects of species interactions on forest ant communities that were heated as part of a large-scale climate manipulation at high- and low-latitude sites in eastern North America. Overall, we found mixed support for the importance of negative species interactions (competition), but found that the magnitude of these interaction effects was predictable based on the heat tolerance of the focal species. Forager abundance and nest site occupancy of heat-intolerant species were more often influenced by negative interactions with other species than by direct effects of temperature. Our findings suggest that measures of species-specific heat tolerance may roughly predict when species interactions will influence responses to global climate change.

Highlights

Ecologists have widely acknowledged the importance of incorporating species interactions into forecasts of responses to climate change (Tylianakis et al 2008; Walther 2010), but there are few published studies that succeed in this because empirical data are rare and modeling tools are limited (Cahill et al 2012; Angert et al 2013)
Heat tolerance was a significant predictor of the importance of negative species interactions on forager abundance (F 1⁄4 8.72, P 1⁄4 0.00735, df 1⁄4 1), and there were no effects of site (F 1⁄4 0.0263, P 1⁄4 0.873, df 1⁄4 1) or an interaction between site and critical thermal maximum (CTmax) (F 1⁄4 0.105, P 1⁄4 0.749, df 1⁄4 1), indicating that this relationship was consistent between sites
In each of the four response-type-by-site datasets, the total number of observations of each species during the course of the warming experiment was uncorrelated with CTmax, which suggests that sampling bias was not responsible for the relationships between CTmax and negative species interaction effects that were observed (Duke Forest pitfall: r 1⁄4 0.113, P 1⁄4 0.665, df 1⁄4 15; Duke Forest nest box: r 1⁄4 0.382, P 1⁄4 0.618, df 1⁄4 2; Harvard Forest pitfall: r 1⁄4 À0.469, P 1⁄4 0.202, df 1⁄4 7; Harvard Forest nest box: r 1⁄4 À0.906, P 1⁄4 0.0939, df 1⁄4 2)

Summary

Introduction

Ecologists have widely acknowledged the importance of incorporating species interactions into forecasts of responses to climate change (Tylianakis et al 2008; Walther 2010), but there are few published studies that succeed in this because empirical data are rare and modeling tools are limited (Cahill et al 2012; Angert et al 2013). In grassland plant communities, the indirect effects of climate change, i.e., those mediated through species interactions, are strongest for species showing weak niche differentiation (Chu et al 2016) Such trait-based approaches have met with considerable success in forecasting the relative performance of species—the winners and losers—under the direct effects of climate change (see Pacifici et al 2017 for a recent data-driven review; see for trait-based forecasting frameworks: Suding et al 2008; Chown 2012; Foden et al 2013). In forest ant communities, greater heat tolerance is positively correlated with ant abundance and activity under experimental warming (Diamond et al 2012, 2013; Stuble et al 2013) and historical changes in abundance related to long-term patterns of climate change in the field (Resasco et al 2014) Again, these approaches lack an explicit consideration of indirect effects of temperature change via altered species interactions. A key question is whether heat tolerance can predict the importance of competitive species interactions under warming

Methods

Results

Conclusion