Scaling in stream communities

Proc Biol Sci. 2002 Dec 22;269(1509):2587-94. doi: 10.1098/rspb.2002.2195.


Scaling relationships between population density (N) and body size (W), and of their underlying size distributions, can contribute to an understanding of how species use resources as a function of size. In an attempt to resolve the controversy over the form of scaling relationships, an extensive dataset, comprising 602 invertebrate species, was obtained from two geographically separate stream communities (Seebach in Austria and Mynach in Wales). We analysed the temporal consistency of the N-W relationship, which was subjected to ordinary least squares (OLS), bisector (OLS(BIS)) and quantile regressions, and species-size spectra with seasonally collated data. Slopes of seasonal OLS(BIS) regressions did not depart from -1 in either community, indicating a seasonally convergent scaling relationship, which is not energetically constrained. Species-size spectra may scale with habitat complexity, providing an alternative explanation for the observed body-size scaling. In contrast to the right-skewed species-size frequency distributions of single-species assemblages, the size spectra of these benthic communities exhibited 'central tendencies', reflecting their phyletic constitution. The shape of species body-mass spectra differed between the two communities, with a bimodal and seasonally convergent pattern in the Seebach community and a seasonally shifting unimodality in the Mynach community. The body-size spectra of large, mostly insect, species (greater than or equal to 1 mm) scaled to seasonal variations in habitat complexity (i.e. fractal D), suggesting that habitat structure constrains the community organization of stream benthos.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Body Constitution
  • Ecosystem*
  • Fresh Water*
  • Invertebrates / anatomy & histology*
  • Invertebrates / classification
  • Invertebrates / physiology*
  • Models, Biological
  • Population Density
  • Regression Analysis