Edinburgh Napier University

  At least ten different static patterns have been suggested by the analysis of food webs. However, the existence of many of these patterns has been questioned in recent years. One pattern that has remained immune from such criticism is intervality. In this paper, we develop a continuous measure of niche overlap, by calculating the number of non-triangulated quads (the simplest structure that confounds intervality) in a web. In contrast to intervality, this measure is applicable to webs of any size. We then explore the implications for niche overlap of including parasite - host links in a real food web, that of the Ythan estuary, Aberdeenshire. Increasing the number of parasite-host links increases the number of non-triangulated quads in a web. This increase is greater than that predicted by the cascade model. One explanation for the high incidence of intervality in real food webs is that species are ordered into a hierarchy dependent on body size. Parasites should obviate this ordering, since they are smaller than their hosts. Splitting the parasite species into 'trophospecies' according to their separate life history stages reduces the number of non-triangulated quads. This indicates that parasites do not reduce the chances of intervality (or increase the number of non-triangulated quads)