Gp. Closs et Ps. Lake, SPATIAL AND TEMPORAL VARIATION IN THE STRUCTURE OF AN INTERMITTENT-STREAM FOOD-WEB, Ecological monographs, 64(1), 1994, pp. 1-21
Food webs from the Lerderderg River, an intermittent stream in Victori
a, Australia, were compiled with the aim of examining changes in food
web structure in a highly variable habitat. Emphasis was placed on a h
igh degree of taxonomic precision. Spatial and temporal Variation in t
he food webs was assessed by partitioning the study area into three si
tes, located approximate to 1.5 km apart along the river. Sites differ
ed in overall stream width and the length of the low streamflow period
during summer. Three separate webs for each site were compiled for fo
ur different times of the year. Relatively little spatial variation in
community structure was observed. In contrast, temporal variation was
considerable, with species composition and the number of species in t
he community changing considerably over the year. The number of specie
s increased dramatically as the period of constant streamflow lengthen
ed. Detritivores dominated the community, both in terms of species and
individual numbers. The proportion of predators in the community incr
eased slightly by the end of the year, suggesting that recolonization
of the community by predators lags behind that of detritivores. The in
crease in the number of predator species also resulted in an increase
in the mean food chain length through the year. Patterns observed in t
he food webs tended to fall within the range of values reported from s
everal previous studies, suggesting that underlying constraints may st
ructure certain aspects of food webs. However, the constancy of certai
n food web statistics was attributable either to methodological decisi
ons made during compilation, or to an inherent property of the statist
ic itself. The potential sensitivity of several food web statistics to
the methodology used to compile a food web render between-web compari
sons difficult due to the confounding effects of methodology. This sug
gests that comparisons between food webs should be restricted to webs
derived from similar habitats using a comparable methodology.