Abstract
<jats:p>Studies estimating the effect of temperature on wild animal populations typically focus on survival as a proxy for population health. However, high and low temperatures can also negatively influence animal reproduction, and recent work suggests that the thermal sensitivity of reproduction is more important for population persistence that the thermal sensitivity of survival in some species groups. If this is the case, studies focusing on survival may be underestimating the impacts of warming on population persistence. However, no study has yet compared the thermal sensitivity of reproduction and survival of the same species on a large taxonomic scale. We here take advantage of a systematic database of studies comparing the effect of temperature on the reproductive output (fecundity) and survival/lifespan of adults of 306 ectothermic invertebrate species. We used meta-analysis to quantify the thermal sensitivity of all three life history traits in relation to temperature conditions assumed to be optimal for population growth rate for each species (‘reference temperatures’). We found that reproductive output and adult survival were maximised close to the reference temperatures, whereas adult lifespan was maximised 12 to 18℃ below reference temperatures. All three life-history traits were negatively affected by warming, but a multivariate meta-analysis showed that reproductive output declined more than twice as steeply with warming than adult lifespan. We suggest that a reduction in reproductive capacity may be a key driver of wild population declines due to warming, and that studies focusing on survival are underestimating the impacts of warming on population persistence.</jats:p>