Background: Latitudinal clines in temperature and seasonality impose strong seasonal constraints on ectotherms. Studies of population differentiation in phenotypic plasticity of life history traits along latitudinal gradients are important for understanding how organisms have adapted to seasonal environments and predict how they respond to climate changes. Such studies have been scarce for species with a northern distribution.
Methodology/principle finding: Larvae of the northern damselfly Coenagrion johanssoni originating from semivoltine central, partivoltine northern, and partivoltine northernmost Swedish populations were reared in the laboratory. To investigate whether larvae use photoperiodic cues to induce compensatory growth along this latitudinal gradient, larvae were reared under two different photoperiods corresponding to a northern and southern latitude. In addition, field adult size was assessed to test the strength of possible compensatory growth mechanisms under natural conditions and hatchling size was measured to test for maternal effects. We hypothesized that populations originating from lower latitudes would be more time constrained than high-latitude populations because they have a shorter life cycle. The results showed that low-latitude populations had higher growth rates in summer/fall. In general northern photoperiods induced higher growth rates, but this plastic response to photoperiod was strongest in the southernmost populations and negligible in the northernmost population. During spring, central populations grew faster under the southern rather than the northern photoperiod. On the other hand, northern and northernmost populations did not differ between each other and grew faster in the northern rather than in the southern photoperiod. Field sampled adults did not differ in size across the studied regions.
Conclusion/significance: We found a significant differentiation in growth rate across latitudes and latitudinal difference in growth rate response to photoperiod. Importantly, growth responses measured at a single larval developmental stage in one season may not always generalize to other developmental stages or seasons.