REPRODUCTIVE TRAITS, HABITAT USE AND TEMPLET THEORY - A SYNTHESIS OF WORLD-WIDE DATA ON AQUATIC INSECTS

1. Using world-wide data on the reproductive biology of 131 species (i n eight orders) of aquatic insects, we used multivariate analyses to e xamine: (i) relationships among reproductive traits determining life c ycle, fecundity, morphology, behaviour and physiology; (ii) relationsh ips among traits determining spatial and temporal habitat characterist ics at different scales; and (iii) the relationship between reproducti ve and habitat-use traits. This provided a test of predictions of the habitat templet concept on trends of species traits along gradients of habitat heterogeneity. 2. The major trends observed in the relationsh ips among reproductive traits were that larger females had larger eggs , which were more vulnerable to perturbations such as droughts and oft en laid in cocoons. In addition, they laid the eggs in larger numbers of smaller clutches than smaller females. Other traits (e.g. egg numbe r or incubation time) did not show clear trends. 3. Females that depos ited eggs at sites of low local temporal heterogeneity (within plants) used, at the same time, gross habitats of high temporal heterogeneity (temporary waters). In contrast, traits in habitat use hardly differe d on well-known gradients of temporal heterogeneity along running wate rs (from source to estuary). The number of habitat units used by ovipo siting females generally increased with the spatial scale considered, most species oviposited in a single small habitat unit but in several gross habitats. 4. A significant (P < 0.01) relationship between trait s in reproduction and habitat use demonstrated that habitat acted as a templet for reproductive strategies. This relationship was dominated by larger females having larger, unattached eggs which were more vulne rable to droughts and were oviposited in temporally more stable small- scale habitats (within wood or macrophytes, or within cocoons spun by the female) but more unstable large-scale habitats (primarily temporar y waters). Thus, only on the small habitat scale did some of our obser vations correspond to the predictions of the habitat templet concept ( e.g. larger size or higher vulnerability in more stable habitats). How ever, many species had traits in reproduction that did not show trends as predicted by the concept. 5. This and other recent studies of the relationships between traits of freshwater organisms and the heterogen eity of their habitats have shown that habitat acts as a templet for s pecies life history traits. However, many of the details observed in t hese studies did not correspond to predictions of the templet concept because of trade-offs among the traits and scale problems in the descr iption of habitat heterogeneity. Therefore, future studies should focu s on groups of organisms which are as similar as possible in the trade -offs among their species traits and on the potential relationships of habitat heterogeneity across multiple scales.