Trade-offs between reproductive and somatic (storage) investments in animals: a comparative test of the Van Noordwijk and De Jong model

Abstract

Classical life-history theory predicts ‘trade-offs’ between reproductive and somatic investments. However, empirical studies have shown that intraspecific phenotypic correlations between these two resource investments are often positive or nonsignificant, rather than negative as predicted. The model of Van Noordwijk and De Jong (1986) was proposed to explain these unexpected results. According to their model, positive correlations between reproductive and somatic investments will result if individual variation in resource acquisition exceeds that of resource allocation, whereas negative correlations will result if individual variation in resource allocation exceeds that of resource acquisition. To test this model, I used body storage/condition as an index of somatic investment because it is usually strongly related to level of resource acquisition. I predicted that laboratory studies should more often show negative correlations between reproductive and somatic investments than field studies, because individual variation in resource acquisition is expected to be lower in controlled laboratory environments than in variable natural environments. A literature review revealed that correlations between somatic (storage) investment and reproductive investment (estimated as clutch/litter mass, number of offspring per clutch/litter, or number of clutches/litters) among conspecific breeding female animals are more often positive (15 species) or nonsignificant (17 species) than negative (6 species). Moreover, as expected, five of six negative correlations were observed in laboratory studies, whereas 13 of 15 positive correlations were observed in field studies. It is concluded that future empirical and theoretical work on life histories should consider individual variation in both resource acquisition and allocation and the interaction between the two.