While much progress has been made toward identifying the spectrum of organisms’ responses to climate change, which include distributional, phenological, demographic, behavioural and genetic changes (reviewed in Parmesan 2006), our ability to accurately forecast the magnitude and direction of these changes requires a more mechanistic understanding of the interactions among the abiotic and biotic drivers that determine organismal fitness and performance. More specifically, future work should combine observational, experimental and modelling approaches to: I, quantify the net fitness effects of multiple abiotic variables; II, determine which species’ traits and/or geographic locations correspond to increased sensitivity; and III, examine the effects of biotic interactions in mediating the responses of organisms and shaping future ecological communities under climate change.
For many organisms, particularly insects and other ectotherms, changes in temperature have pronounced effects on life-history traits and fitness (Parmesan, Root & Willig 2000; Bale et al. 2002). In one of the first studies to quantify the net effects of changes in both temperature means and extremes, Buckley & Kingsolver (2012) examine the demographic impacts of climate change on two Colias species (sulphur butterflies) in alpine habitats, one of the biomes most vulnerable to biodiversity losses under rapid climate change (Parry & IPCC 2007). In this issue of Functional Ecology, Buckley & Kingsolver (2012) incorporate climatic data and demography into a biophysical model to determine whether fitness and activity responses to recent changes in temperature means and extremes correspond to current C. meadii and C. eriphyle lower and upper elevational limits in the Rocky Mountains. Overall, the projections of their biophysical models reveal opposing effects of temperature changes on flight and fecundity and do correspond with the current elevational limits of these two butterfly species, demonstrating that changes in both mean and extreme temperatures set species’ distributional boundaries and therefore need to be considered when forecasting species’ responses to climate change.
Another novelty of this study is that Buckley & Kingsolver (2012) compare two closely related species that differ in wing melanism, thoracic fur thickness and elevational limits. By contrasting a high-elevation species that is able to achieve higher body temperatures and therefore fly, mate and oviposit under cooler conditions than an otherwise similar (with regard to host plant specificity and other important ecological traits), low-elevation species, the authors are able to effectively tailor their projections on the impacts of changes in temperature means and extremes to different phenotypes at various elevations rather than generalizing across species (and phenotypes) and geographic location. This approach helps to identify the functional drivers, thermoregulatory traits in this case, of individualistic responses of species to climate change.
Other work examining the thermal physiology of species has led to projections that, despite the greater magnitude of warming expected at higher latitudes, organisms living at lower latitudes, where they live closer to their upper thermal limits and experience less variation in thermal conditions, will be most sensitive to rapid changes in mean annual temperatures (Deutsch et al. 2008; Tewksbury, Huey & Deutsch 2008). Huey et al. (2009) considered these physiological limits as well as the potential for behavioural and physiological compensation in lizards and projected that, with changes in the mean maximum temperatures during the warmest part of the year, tropical forest lizards are likely to be replaced by warm-adapted, open-habitat competitors and predators. For ants, projections based on thermal tolerance, climatic, ecological and phylogenetic data suggest that ants that live in tropical forest canopies at low elevations are those most at risk of declines under warming (Diamond et al. 2012). Yet, other organisms such as intertidal mussels are expected to experience localized extinctions that do not vary monotonically with latitude because of the geographic mosaic of thermal environments they encounter across their ranges (Helmuth et al. 2002). Altogether, these findings suggest that organisms are likely to respond in idiosyncratic ways to climate change and that these responses will differ across space.
The sensitivity of plants, insects and fish to rapid climatic changes also has been found to vary with habitat and resource specialization, dispersal ability, dependence on other environmental cues or interspecific interactions, and other life-history traits (e.g. Jiguet et al. 2007; Angert et al. 2011; Mattila et al. 2011). More specifically, Lenoir et al. (2008) found that plants that are restricted to mountain habitats and have rapid life cycles are those that shifted their distributions the most under recent warming. Similarly, work on marine fishes also has found the greatest climate-driven distribution shifts in species with faster life cycles and smaller body sizes (Perry et al. 2005). Lavergne, Molina & Debussche (2006) found that the greatest abundance reductions for rare plants during the last century have been for aquatic, water-dispersed and annual plants, while the greatest increases have been in herbaceous perennial plants, tree and wind-dispersed plants. For butterflies, Diamond et al. (2011) found greater phenological advances for butterfly species with small range sizes, advanced overwintering stages and higher host specificity. More work is needed to determine whether species’ traits that correspond to increased sensitivity to climatic changes are consistent across taxa.
Buckley and Kingsolver’s integration of observational and modelling approaches provides greater predictive power for forecasting species’ responses to climate change (Buckley & Kingsolver 2012). Future work on Colias and other butterflies should build on these methods by incorporating the effects of climate change on other life stages (i.e. larval development). For butterflies that specialize on a limited number of host plants, biotic constraints such as host plant distribution and demography, particularly because they may respond differently to changes in temperature, precipitation and CO2, also should be important components of future modelling efforts. Finally, experimental manipulations that examine the relationships between temperature and fitness simulated by biophysical models and identify the factors that set distributional limits could reduce uncertainty and further improve the predictive power of future models. Once we gain a more mechanistic understanding of the functional drivers and net effects of abiotic change at the population and species levels, we can forecast how communities will respond to climate change across large geographic areas.