Introduction
Large-scale ecogeographical gradients can explain variation in diverse traits, from body size (Ashton 2002) to colour (Friedman & Remeš 2017). Several ecogeographical patterns have been formalised into rules; yet there is persistent debate regarding underlying causes and the taxa to which they apply (Gaston et al. 2008; Chown & Gaston 2010). This problem is epitomised by Gloger’s rule and Bogert’s rule, which both describe ecogeographical patterns of melanisation. Gloger’s rule describes the tendency for heavily pigmented (darker) forms to be found in hotter and more humid regions (Delhey 2017, 2019). This relationship may be driven by one or more factors including camouflage in low light environments (Zink & Remsen Jr 1986; Cheng et al. 2018), protection from ultraviolet light or parasites (Burtt Jr & Ichida 2004; Chaplin 2004), or pleiotropic effects of genes regulating both climatic adaptations and melanin-based coloration (Ducrest et al. 2008). Bogert’s rule (also termed the thermal melanin hypothesis) describes the tendency for darker animals to occur in colder regions because darker colours absorb more solar radiation, thus providing thermal benefits (Bogert 1949). Traditionally, Bogert’s rule has been applied to ectotherms and Gloger’s rule to endotherms; however, the accumulated evidence suggests that both rules may apply broadly to ectotherms or endotherms (Trullas et al. 2007; Delhey 2018; Galván et al. 2018; Delhey et al. 2019). Efforts to reconcile the seemingly opposing effects of these rules have so far been hampered by the difficulty of disentangling the underlying drivers.
Most of the evidence that supports either Gloger’s or Bogert’s rule relates to visible colour (all or part of the wavelengths range from 300–700 nm); however, the spectrum of direct sunlight extends well beyond this range. Wavelengths from 700–1400 nm (near-infrared, NIR) include approximately 50% of solar energy (Stuart-Fox et al.2017) and can therefore strongly affect heat gain. By contrast, NIR does not directly affect camouflage because little or no NIR light can be seen by animals (Stuart-Fox et al. 2017). Examining ecogeographical gradients in NIR reflectance can therefore help to distinguish underlying drivers of ecogeographical patterns of animal coloration (Cuthill et al. 2017; Stuart-Fox et al. 2017; Ruxton et al. 2018). To date, ecogeographical patterns of NIR reflectance have only been examined in Australian birds and butterflies (Medina et al. 2018; Munro et al. 2019), which both show thermally adaptive variation. However, Australia is a hot and dry continent so it remains unclear whether climatic gradients in NIR reflectance exist for other taxa or climates.
Butterflies are a model group to investigate ecogeographical patterns of light manipulation due to their thermal biology and extraordinary diversity in coloration. They are primarily ectothermic like many insects and regulate their body temperature through both physical and behavioural traits (Clench 1966). Physical properties of the thorax and basal wings (i.e. parts of the wings that are close to thorax) directly affect the temperature of flight muscles through heat conduction (Heinrich 1974). The wings beyond the basal region may have less impact on thermoregulation because there is less haemolymph circulation and fewer vascular extensions that can carry significant quantities of heat to the thorax (Arnold 1964; Kammer & Bracchi 1973; Wasserthal 1983; Kingsolver 1987). However, wings can overheat quickly under direct sunlight due to their low thermal capacity, and butterflies have evolved sophisticated wing scale structures to control wing temperature through radiative cooling (Tsai et al. 2020). Butterflies also regulate their temperature through various behavioural mechanisms including dorsal and lateral basking (opening/folding wings to expose thorax and wing surface), ground-contact, orientating themselves in relation to the position of the sun, and shivering (Clench 1966). The reflectance of butterflies plays a crucial role during behavioural thermoregulation such as dorsal (wings open) and lateral (wings closed) basking because the efficacy of these behaviours depends on how much light their body and wings absorb (Kingsolver 1987, 1988). The reflectance of both (especially dorsal) regions seems to play a key role in warming up during basking while those of ventral regions seem to be additionally related to preventing overheating (Kingsolver 1987). Thus, thermal pressures may act differently on dorsal and ventral surfaces which consequently result in the evolution of reflectance differences between dorsal and ventral surfaces. Specifically, higher dorsal-ventral contrast is predicted to evolve in species that are found in heat-stress environments where the benefit for high ventral reflectance for cooling is greater. However, this prediction has yet to be formally tested.
In this study, we tested whether climate predicts the reflectance of both dorsal and ventral regions of 343 European butterfly species using full-spectrum photography of museum specimens. We compiled climatic niche characteristics of each species and tested multiple hypotheses regarding ecogeographical patterns of butterfly reflectance after controlling for phylogeny and phylogenetic uncertainty (Schweigeret al. 2014). We specifically addressed four questions regarding ecogeographical variation in butterfly reflectance: (1) does butterfly reflectance follow the patterns predicted by Bogert’s rule in both visible and near-infrared wavebands? (2) does NIR reflectance show thermally adaptive patterns independent of visible reflectance? (3) does butterfly reflectance follow the patterns predicted by Gloger’s rule? (4) does climate predict ventral-dorsal contrast in butterflies?