Canadian butterfly climate debt is significant and correlated with range size

Abstract

Climate change is causing rapid shifts in species’ range limits, leading to poleward expansions and range losses toward the equator. However, ‘climate debt’, the gap between required and realized range shifts under changing climates, can accumulate when species are unable to track shifting conditions sufficiently rapidly to keep pace with climate changes. Currently, we do not know the rate at which species will keep pace via dispersal to track their climate envelopes, yet understanding potential differences in climate debt is central to estimating how climate change will influence extinction risk. Here, we use historical observations of 155 butterfly species found in Canada to construct climate‐based environmental niche models for each species and then compare projections with observed modern distributions to quantify climate debts. This approach suggests that high levels of climate debt are accumulating within the vast majority of these species. Such failure to track changing climates may arise from some combination of interspecific interactions such as particular food availability for specialists, abiotic barriers such as mountain ranges, or species’ intrinsic dispersal capacities. Our linear models relating climate debt to a variety of biological predictors suggest that the debts we documented are accumulating independently of dispersal ability, diet breadth, and phylogeny. A proxy for range size is the only significant predictor of climate debt, with species with narrower ranges accumulating more debt: this suggests that species with narrow ranges may be at risk from both a reduction of suitable habitat in their current range and the failure to colonize newly available habitat. Identifying the factors, whether intrinsic or imposed by local environmental conditions, that limit species’ capacities to colonize areas beyond their historical limits is vital to conservation planning.

Inclusion of trophic interactions increases the vulnerability of an alpine butterfly species to climate change

“…The direct effects of climate change on our tested butterflies followed expectations that HS in lower latitudes would decline and in high latitudes would gain. We again caution against attributing the gains observed as fact because other studies have identified that species typically lag behind poleward shifts in climate (Kerr et al, 2015; Lewthwaite et al, 2018; Menéndez et al, 2007; Parmesan et al, 1999). The maps of predicted future suitability for butterfly species are also averages across all GCMs, and thus are neglecting nuances associated with specific climatic patterns, such as drier conditions projecting declines in P. smintheus and P. occidentalis .…”

“…Although the direct effects of climate on P. smintheus are well understood, the indirect effects are unknown. Previous studies exploring the indirect effects of climate change on butterflies typically are missing the mechanism that drives the direct effects climate on the population dynamics of their target species (Lewthwaite et al, 2018; e.g. Menéndez et al, 2007).…”

Long‐term demographic surveys reveal a consistent relationship between average occupancy and abundance within local populations of a butterfly metapopulation

“…In environments where dispersal is not a limiting factor and there is no strong spatial correlation in habitat quality, such simple relationships can be reasonable approximations, but in fragmented landscapes where occupancy is a function of both colonizations and local population dynamics, we should not let this assumption stand untested. An extreme example are bioclimatic envelope models lacking any population dynamic component that are used to predict range shifts under climate change (Araújo et al 2005, Lewthwaite et al 2018, in spite of the availability dynamic alternatives (Keith et al 2008, Buckley et al 2010, Leroux et al 2013. These models, which are fit to records of presence and absence of the study species, often assume a simple relationship between carrying capacity and patch characteristic, such as area and habitat quality, and use this relationship as the basis of extinction probabilities.…”

“…Such deficiencies are most critical when predicting under novel environmental conditions and outside a species' current range. An extreme example are bioclimatic envelope models lacking any population dynamic component that are used to predict range shifts under climate change (Araújo et al 2005, Lewthwaite et al 2018, in spite of the availability dynamic alternatives (Keith et al 2008, Buckley et al 2010, Leroux et al 2013.…”

Gene flow and climate‐associated genetic variation in a vagile habitat specialist

“…Within Alberta and Saskatchewan, changes in mean annual temperature, growing degree‐days, and vegetation composition are expected to be most pronounced in central and southern regions, including the present‐day range of P. m. dodi (Barber, Nielsen, & Hamann, 2016; Schneider, 2013; Zhang, Nielsen, Stolar, Chen, & Thuiller, 2015). There is some evidence that vagile North American butterfly species may track their climatic niches poleward as temperatures warm; however, more often than not, these range expansions are not sufficient to offset contractions toward the equator (Lewthwaite et al., 2018). Indeed, there exist few opportunities for P. m. dodi to track its climatic niche northward as temperatures rise.…”