Temperature-size responses during ontogeny are independent of progenitors' thermal environments

Warming generally induces faster developmental and growth rates, resulting in smaller asymptotic sizes of adults in warmer environments (a pattern known as the temperature-size rule). However, whether temperature-size responses are affected across generations, especially when thermal environments differ from one generation to the next, is unclear. Here, we tested temperature-size responses at different ontogenetic stages and in two consecutive generations using two soil-living Collembola species from the family Isotomidae: Folsomia candida (asexual) and Proisotoma minuta (sexually reproducing).

We used individuals (progenitors; F0) from cultures maintained during several generations at 15 °C or 20 °C, and exposed their offspring in cohorts (F1) to various thermal environments (15 °C, 20 °C, 25 °C and 30 °C) during their ontogenetic development (from egg laying to first reproduction; i.e., maturity). We measured development and size traits in the cohorts (egg diameter and body length at maturity), as well as the egg diameters of their progeny (F2). We predicted that temperature-size responses would be predominantly determined by within-generation plasticity, given the quick responsiveness of growth and developmental rates to changing thermal environments. However, we also expected that mismatches in thermal environments across generations would constrain temperature-size responses in offspring, possibly due to transgenerational plasticity.

We found that temperature-size responses were generally weak in the two Collembola species, both for within- and transgenerational plasticity. However, egg and juvenile development were especially responsive at higher temperatures and were slightly affected by transgenerational plasticity. Interestingly, plastic responses among traits varied non-consistently in both Collembola species, with some traits showing plastic responses in one species but not in the other and vice versa. Therefore, our results do not support the view that the mode of reproduction can be used to explain the degree of phenotypic plasticity at the species level, at least between the two Collembola species used in our study. Our findings provide evidence for a general reset of temperature-size responses at the start of each generation and highlight the importance of measuring multiple traits across ontogenetic stages to fully understand species' thermal responses.