Warming tolerance across insect ontogeny: influence of joint shifts in microclimates and thermal limits

National audienceNot all species will experience global warmingequally. First, organisms liv e in microcli mates, thetemperatures of which can deviate substantially frommacroclimate temperatures (Gates 1980, Angilletta2009). Second, the physiological performance andtolerance of organisms filters microclimate conditions,often nonlinearly (Huey and Kingsolver 1993, Huey etal. 2012). Consequently, the impact of warming on thepersistence and distribution of species cannot be inferredsolely from macroclimatic observations. Instead, what isrequired is a mechanistic understanding of bothmicroclimate and the physiological tolerance of species(Williams et al. 2008, Helmuth et al. 2010). For instance,tropical ectotherms were predicted to suffer the mostfrom global warming because they already live close totheir ther mal li mits, wh ereas temperat e ectother mswould benefit from a higher warming tolerance: a largerdifference between macroclimate temperature and ther-mal limits (Deutsch et al. 2008, Tewksbury et al. 2008).Recently, a study showed that equilibrium bodytemperatures of most ectotherms at all latitudes arehigher than both their thermal limits and ambient airtemperature (inferred as the highest monthly mean ofdaily maximum air temperature), suggesting the need tothermo regulat e for the mai ntenanc e of popula tions(Sunday et al. 2014). However, the importance ofmicroclimates is often neglected in studies on warmingtolerance of ectotherms, as we will detail (Potter et al.2013, Scheffers et al. 2013a).Warming tolerance (WT) is calculated as the devia-tion between the thermal limit—e.g., the criticalmaximal temperature, CTmax, or the uppe r letha ltemperature, ULT—and the habitat temperature, whichusually is taken as a monthly or annual average ofmacroscale air temperature (Deutsch et al. 2008). Onthat basis, a dire ct r elatio nshi p betwe en warmingtolerance and absolute latitude is expected in ectothermsbecause macroclimate temperature increases from thepoles to the equator and there is no clear pattern ofCTmax or ULT over latitudes (Addo-Bediako et al.2000, Deutsch et al. 2008, Sunday et al. 2011, Diamondet al. 2012, Arau´ jo et al. 2013). Frequently, thislatitudinal pattern of warming tolerance is used toforecast potential distribution shifts in ectothermscaused by global warming (Sunday et al. 2012). Macro-climate temperature, however, can be uninformativeregarding the actual relationship between physiologicaltolerance and latitude, because the pertinent variableoften is the microclimate temperature (Fig. 1; see Hueyet al. 2012, Sunday et al. 2014). Two lines of evidencesuggestthatlackofknowledgeonmicroclimate