Pimm and Lawton (1980) proposed that food webs should be compartmentalized between, but not within, habitats. Whereas the relatively distinct communities representing disturbed versus undisturbed patches within an ecosystem represent compartmentalization, the communities within habitat patches should not be compartmentalized. This property largely follows from the constant connectivity hypothesis (i.e., compartmentalization is inconsistent with equal linkage among species).
The vague definition of habitat complicates assessment of compartmentaliza-tion. For example, does soil/litter constitute a habitat or a subunit of the site habitat? Soil/litter subcommunities tend to be distinct from plant-based above-ground subcommunities.
Nevertheless, compartmentalization can be identified within recognized habitats. j. Moore and Hunt (1988), Polis (1991b), and Reagan et al. (1996) found distinct compartmentalization within the community of a single patch when arthropod species or "kinds" were distinguished (see Table 9.2). Distinct com-partmentation of arthropod assemblages has been shown among plant species (e.g., Fig. 9.7) and even between trees and sapling of the same species (Basset 2001). Compartmentalization reflects the development of component communities composed of specialists feeding on particular resources and the resulting channels of energy and material transfer. Host specificity appears to occur more frequently and at a finer spatial scale among herbivorous and detritivorous arthropods, based on their small size, short life spans, and intricate biochemical interactions (see Chapter 3) that facilitate rapid adaptation for utilization of particular resources, even within individual leaves (e.g., Mopper and Strauss 1998, K. Parsons and de la Cruz 1980). Many parasitoids also are host specific, so that compartmentalization is maintained at higher trophic levels among arthropods. Of course, generalists at all trophic levels connect compartments and maintain the web of interactions. J. Moore and Hunt (1988) found that compartmentalized models of food webs were more stable than were noncompartmen-talized webs.
Was this article helpful?