Overall, we failed to reject the null hypothesis, meaning there was no relationship found between the burrowing of frosted elfin caterpillars and temperature or soil moisture. However, our results corroborate previous work and contribute to the overall knowledge of burrowing depth for this species. The proportion of individuals that pupated below the surface in our study (15%) is comparable to the 19.64% of preceding research. Based on data provided by Thom et al. (2015) and Meyer et al. (2023) it is unlikely C. irus can survive fire unless they burrow much deeper beneath the surface. Only two caterpillars out of 60 (~ 3%) burrowed at a level deep enough (~ 1.5 cm) to survive the expected lethal temperatures of a typical prescribed fire in north Florida. If our findings mimic natural propensities to burrow, high mortality can be expected when prescribed fires are applied to management units containing elfin pupae. Meyer et al. (2023) observed a higher rate of natural burrowing with four out of 10 burrowing to depths of 1, 1.5, 2.5, and 5 cm. While the sample was small, data suggest chances of survival could be as high as 30%. Future laboratory investigations on the rate of burrowing in frosted elfins could test more extreme variables using more common Lycaenid species that exhibit comparable burrowing behavior.
While there was no significant difference in pupae burrowing across temperature or soil moisture conditions, there are other potential influences on burrowing behavior. We believe that the caterpillars may dig opportunistically when they encounter a micro-pocket of low compaction during wandering. This may have been a factor with two individuals that dug near the stick and one that used a space between the sand and cup that originated when the stick was placed. It is possible that the energetic cost of digging does not outweigh the probability of being burned. The variation in pupation behavior observed in this species may be controlled by inherited traits while burrowing likelihood is further influenced by energy, conditions, or the ability to find a suitable location. For example, genetics, environmental factors, and sex influence larval behavioral phenotypes which are linked and heritable in individuals of Drosophila melanogaster. Phenotypes that travel further to forage for food pupated higher or burrowed deeper in containers (Sokolowski and Hansell 1983; Riedl et al. 2007).
Additionally, C. irus may burrow to escape predation. For example, insect predators can have different thermal tolerances than their prey, which may limit interactions (Huey and Kingsolver 1989; Chown and Nicolson 2004). Formica lasioides (Hymenoptera: Formicidae) predation of Platyprepia virginalis (Boisduval) (Lepidoptera: Arctiidae) decreased when larvae took refuge in cool, wet leaf litter (Karban et al. 2015). Potentially, C. irus’ burrowing behavior protects individuals from their most common predators. Callophrys irus may be less likely to burrow in the absence of predators, which it could detect from olfactory stimuli from the environment and was not present in our study. Regardless of the evolutionary influences on this behavior, the frequency of burrowing could be critical information for the successful stewardship of vulnerable populations.
Historically, fires in southeastern pine savannas primarily occurred during the lightning season in May to mid-July and were low-intensity, fast-moving fires (White and Harley 2016). When considering the tradeoff between burrowing and conserving energy, spatial and temporal heterogeneity of prescribed fires (i.e. ‘pyrodiversity’) in this habitat may have made energy conservation more favorable for the frosted elfin over evolutionary time. For example, in north Florida sandhill pine ecosystems, researchers recommend burning in 2–4 year intervals. In this climate, burning before February and after April protects frosted elfins during their most vulnerable life stages as eggs or caterpillars (McElveen pers. comm.; Jue et al. 2022). Burning outside this recommended time frame may also affect the successional stages of the host plant and the ability of adults to recolonize a hostplant patch from neighboring populations within their flight time of 2–5 weeks (Thom 2013). Fire affects vegetation structure across 40% of the world’s land surface and across many different ecosystems (Chapin et al. 2011). The best time to burn for C. irus is specific to the site and habitat type being managed. Seasonality along with local coevolution with fire should be considered.
Our results support previous recommendations of refugia and rotational burn schedules throughout occupied C. irus habitat. Pyrodiversity has been recommended for the benefit of avian, invertebrate, and plant species of longleaf pine habitats (Hanula and Wade 2003; Knight and Holt 2005; Chitwood et al. 2017; Robertson et al. 2019; Weber et al. 2022). Land managers should assess fire return intervals, ignition techniques, and weather conditions when applying fires to small, isolated populations of this imperiled butterfly. Importantly, artificial intervention may be required to ensure population survival post-burn.