Cool as a caterpillar: Understanding the responses of butterflies to temperature across the life cycle

Anthropogenic effects, including land-use and climate change, have had dramatic and wide-ranging impacts on the natural world. The impacts of these changes have already been detected in many systems, with global biodiversity and abundance of many taxa declining across spatial scales. The loss of biodiversity has knock on impacts to human society, which are only just beginning to be understood. Only with a detailed understanding of how species will respond to environmental changes, and with effective dissemination of this knowledge, can we hope to slow and halt biodiversity loss. In this thesis, I use butterflies as a model taxon to understand the responses of insects to changing habitat and microclimatic conditions. Insects are a diverse and globally distributed group of organisms that play fundamental roles in ecosystem processes, many of which are critical to human societies. Despite this, there is growing evidence that insects are declining at rapid rates around the world, and many of these declines have been attributed to anthropogenic change. Butterflies show detectable responses to environmental change, including spatial responses, such as range shifts, and temporal responses, such as changes to timing of life cycle events. They are diverse, wide-spread, and ecologically sensitive, with complex life cycles that differ in morphology, behaviour, and ecology, and therefore differ in their sensitivities and requirements. In my first data chapter, I use a systematic mapping method to screen, quantify and discuss current knowledge on butterfly responses to temperature. I found that Nymphalidae were the most studied family, likely owing to their high number of species, including large, charismatic and common species of cultural importance. In contrast, Riodinidae were rarely studied, likely due to their elusive nature and being concentrated in the tropics. The tropics were less studied than temperate regions, and more studies reported the responses of butterflies at the adult life stage compared to all other life stages combined. I found that of the responses recorded, behavioural responses were the least common. I found that *in situ* studies were more common than *ex situ*. Taken together, there is an incomplete understanding of butterfly responses to temperature, which may lead to ill-informed decisions. I make suggestions for how to resolve these knowledge gaps, including calling for an increased focus on the tropics, the establishment of more butterfly monitoring schemes, particularly in the global south and South America, incorporating non-adult life stages into butterfly monitoring schemes, and conducting more studies on species from under-represented families. In my second data chapter, I investigated the habitat associations of 11 species of day-flying Lepidoptera as larvae and their foodplants from four nature reserves in Bedfordshire, UK. These study sites were the focus of fieldwork for three data chapters (Chapters 3, 4, and 5). These associations were tested across two spatial scales relevant to both Lepidoptera and land-managers; the reserve-scale and the foodplant patch-scale. I also assessed whether these associations were related to ecological traits. At the reserve-scale, I found substantial variation across species, with a tendency for species that overwinter at non-adult life stages to have stronger habitat associations. The majority of species shared similar habitat associations as their foodplants, indicating that management that benefits foodplants will also benefit these Lepidoptera. However, there were notable exceptions to this, with some species (*Erynnis tages*, *Cupido minimus*, *Polyommatus coridon*, *Aglais urticae*) having conflicting habitat associations with their foodplants, and therefore requiring focused management. At the foodplant scale, four species were associated with specific foodplant characteristics; two with taller foodplants (*C. minimus*, *Anthocharis cardamines*), and two with dense foodplant patches (*Aglais io*, *A. urticae*). These habitat associations can be used to manage for these species and their foodplants. In my third data chapter, I use a single species approach to highlight how citizen science data from butterfly monitoring schemes can be used alongside habitat use data to provide a more complete picture of how species change over time, focusing on the small blue (*C. minimus*). I used 26 years of data from the national UK Butterfly Monitoring Scheme and four years of targeted egg surveys across 14 years to investigate the effects of local temperature on small blue emergence date and total abundance, and whether foodplant characteristics predicted oviposition behaviour. I found that adult small blues were emerging earlier over time, which correlated with higher maximum temperatures in February. In contrast, total abundance was not related to temperature, or abundance in the previous year. Oviposition behaviour was broadly consistent across time, with females selecting foodplants that were tall and apparent, surrounded by taller vegetation, and in low density patches. These results imply that management for greater availability of tall foodplants surrounded by tall vegetation would encourage oviposition across a greater number of flowers, reducing competition and improving larval survival in this rare and declining species. In the fourth data chapter, I investigated the capacity of 14 species of day-flying Lepidoptera to thermoregulate, whether this was influenced by morphological or ecological traits, and whether this capacity differed between adults and larvae. I also investigated what mechanisms species used to thermoregulate, and whether this differed between life stages. I found that larvae were worse at thermoregulating than adults, and that thermoregulatory capacity differed between families, species, and with body length, whereby Pieridae were better at thermoregulating than Nymphalidae, and large larvae were better at thermoregulating than small larvae. I found that adults relied on behavioural thermoregulation, whereas larvae relied more on microclimate selection. This implies that larvae are more dependent on their immediate area to thermoregulate than adults, and that management should maintain or protect vegetation surrounding butterfly foodplants to allow larvae to thermoregulate effectively under climate change. Finally, in the last data chapter, I investigated the impacts of temperature on tropical butterflies. I collected field data on 54 butterfly species in Panama, and also conducted heat knockdown assays on a subset of these species (24) in the lab, to determine whether ecological traits influenced the ability of tropical butterflies to thermoregulate, whether similar ecological traits also influenced thermal tolerance, and whether there was an interaction between the capacity to thermoregulate and thermal tolerance. Thermoregulation and thermal tolerance were influenced by family, wing length, and wing colour, with Pieridae, and butterflies that were large or dark having the strongest ability to thermoregulate, but Hesperiidae, small and dark butterflies tolerating the highest temperatures. There was also an interaction between capacity to thermoregulate and thermal tolerance, whereby species better at thermoregulating had lower thermal tolerance, and vice versa. This implies that species with more stable body temperatures in the field may be more vulnerable to increases in ambient temperatures than previously thought, particularly extreme temperatures such as heatwaves. Butterflies make a valuable taxon to investigate responses to environmental change. Though there are gaps in our understanding, I have started to address these, and made suggestions for future research directions. I have identified species, life stages, and ecological traits that make some butterflies more vulnerable to change than others. I have demonstrated that sensitivity to change differs across the life cycle, and these differences will require different management strategies. Crucially, this highlights that butterfly responses to environmental change can be predictable, and therefore can be managed for to improve butterfly conservation.

Cambridge Conservation Initiative (CCI) Evolution Education Trust (EET)