Both the length of the season that is favourable for insect growth and reproduction and thenumber of generations emerging per season (voltinism) increase with decreasing latitude. Thus,time constraints on reproduction and juvenile development decrease with decreasing latitude,except where voltinism changes and time constraints suddenly increase as the season must beshared with one more generation.
I studied the evolution of insect life histories in relation to time constraints from twoperspectives: polymorphism and clinal variation. Life history polymorphism in seasonalenvironments was studied with the butterfly Pieris napi that has discrete life history strategies, andpolymorphic natural populations. Experimental studies showed that asymmetric intraspecificlarval competition and divergent timing of reproduction between the strategies may promote themaintenance of polymorphism. A simulation model showed that the divergent timing ofreproduction between the strategies is sufficient to maintain polymorphism even in the absence ofintraspecific competition.
Clinal variation was studied empirically with four geometrid moths (Cabera exanthemata,Cabera pusaria, Chiasmia clathrata and Lomaspilis marginata) and generally in theory. Due tolatitudinal saw-tooth variation in time constraints, traditional theory predicts a saw-tooth cline inbody size and development time. A common garden experiment with the four geometrid mothsdid not support the traditional theory even when a saw-tooth cline in body size was found,suggesting that the theory is based on unrealistic assumptions. A theoretical analysis showed thatreproductive effort should be high in populations under intense time constraints and low inpopulations experiencing no time constraints, resulting in a saw-tooth cline. In the four geometridmoths, support for these predictions was found, although the observed clinal variation deviatedfrom the predicted pattern.
The results imply that clinal variation is expected in almost any continuous life history trait,whereas polymorphism of different strategies may emerge when life histories fall into discretecategories. When inferring the evolution of a single trait, complex interdependencies amongseveral traits should be considered, as well as the possibility that the time constraints are notsimilar for each generation in multivoltine populations.
