Ambient temperature provides an adaptive explanation for seasonal reproduction in a tropical mammal

Abstract

Understanding how reproductive timing has evolved to reflect climatic conditions is increasingly important as the climate changes, influencing species persistence and ecosystem dynamics. Among endotherms, seasonal reproduction is often linked to natural selection for reproducing when biotic conditions (e.g. food availability) are most favourable. In contrast, we present evidence that direct effects of an abiotic factor have selected for seasonal reproduction in a tropical mammal. We have shown previously that the African wild dog (Lycaon pictus), a diurnally active coursing predator, has lower reproductive success at high ambient temperatures, when its hunting activity is constrained. We therefore hypothesized that natural selection would favour reproduction during cool weather conditions, in locations where such conditions occur predictably. We show that, as predicted, wild dogs reproduce seasonally at latitudes from 7 to 25°S, such that their pup‐rearing periods coincide with the coolest (but not the driest) weeks of the year. Wild dog reproduction is aseasonal at latitudes ≤2°. As the climate warms, some temperate‐zone species have tracked optimal conditions through altered phenology. However, in seasonally breeding wild dog populations, any alteration in phenology would require breeding in hotter weather, probably reducing reproductive success. This endangered species may thus have a limited ability to adapt to climate change. Seasonal reproduction might be a trait indicating climate change vulnerability among tropical species.

When to stay and when to leave? Proximate causes of dispersal in an endangered social carnivore

“…We calculated wild dog population density (individuals/km 2 ) by dividing the total number of known resident adults (aged ≥ 12 months) by the size of the study area. In northern Botswana, reproduction is seasonal and pups are usually born in June (McNutt, Groom, & Woodroffe, 2019). Therefore, we estimated annual population densities right before the onset of reproduction (i.e.…”

“…We calculated the sex ratio of all adult wild dogs in the study population (updated each June) as the ratio of resident males to resident females. For each sighting, we estimated the time (constrained between 0 and 12 months) since June, which corresponds to peak whelping time (McNutt et al., 2019), to account for seasonal effects in the timing of dispersal. To assess the influence of environmental predictors on dispersal rate, we calculated rainfall for each sighting as cumulative precipitation over the prior 3 months.…”

Dispersal behaviour of African wild dogs in Kenya

“…Wild dogs disappeared from the area in the 1980s, but recolonised naturally around 2000, increasing rapidly to densities similar to those observed in protected areas (Woodroffe, ). Local weather patterns are variable, and wild dog reproduction is not seasonal (McNutt, Groom, & Woodroffe, ).…”

“…We considered all other individuals subdominant, although we distinguished founder and natal animals, and littermate and nonlittermate siblings. In the absence of seasonal reproduction (McNutt et al, ), we defined a pack‐year as the time from leaving one den to leaving the next (mean 332 days, Woodroffe, ); other references to years indicate calendar years. We identified a pack territory as the 100% minimum convex polygon (MCP) encompassing all pack locations within a pack‐year.…”

Within‐ and between‐group dynamics in an obligate cooperative breeder

“…Wild dogs recolonized naturally in 1999 following an absence of approximately 20 years, quickly reaching densities similar to those reported from protected areas (Woodroffe, ). Rainfall patterns vary across the study site and are not consistently seasonal; wild dog reproduction is likewise not seasonal and is restricted almost entirely to a single alpha female in each pack (McNutt, Groom & Woodroffe, ).…”