1. Temporally constrained environments, such as habitats with short growth seasons or short hydroperiods, cause potentially strong selection on life histories. Depending on the predictability of these events and the extent of spatial and temporal heterogeneity, local populations could become adapted either via a fixed phenotype or via life history plasticity in response to these environmental cues.
2. We used a common garden experiment to investigate microgeographic variation in life history responses to combined changes in photoperiod (ambient/late) and hydroperiod (constant/drying) time constraint cues in an alpine caddisfly (Trichoptera). We compared six populations (three permanent/three temporary streams) originating from a small, alpine floodplain and which spanned an expected gradient in growth period duration (GPD) with distance from glaciers.
3. We made two main predictions in relation to locally varying selection pressures: (i) populations nearest glaciers (shorter GPD and strongest time constraints) should have the fastest development rates and (ii) populations from permanent streams should be less able to respond to drying hydroperiods than populations from temporary streams.
4. All populations and both sexes accelerated development in response to late season photoperiod cues. However, only permanent stream populations showed an increase in development time with increasing GPD, suggesting that other factors were influencing populations in temporary streams.
5. Permanent stream populations showed countergradient variation (genetic and environmental influences were in opposition) in development time, and under-compensation of growth rates resulted in a converse Bergmann cline in body size (smaller body size along gradients of declining season length). The extent of plasticity in response to hydroperiod, and the combined effects of both time constraints, differed between populations and sexes, but were not consistent among populations.
6. Taken together, our results suggest adaptive plasticity in response to season length. The lack of a predictable pattern in response to hydroperiod may be due to gene flow or weak selection. We conclude that spatially structured populations can strongly differ in phenotypic plasticity even at microgeographic scales.