My prior impression of our knowledge of patterns of ecological versatility in natural communities was that it was limited, incoherent, and inconsistent. This impression I can trace at least in part to several critical papers published in the 1970s and early 1980s, especially those of Schoener (1974b), Hurlbert (1978), Petraitis (1979), and Feinsinger et al. (1981). These workers argued strongly that the characterization of niche breadth is rather pointless if resource use is not linked to the relative availabilities of the resources. Perhaps Petraitis (1979: 709) said it most succinctly, by concluding: ‘Yet in order to understand the relationship of niche breadth and overlap to community structure, knowledge of available resources is imperative’. Thus, monitoring the availability of resources as well as their utilization (i.e., conformance) is seen as being the crucial part of estimating ecological versatility in the field.

However, given the semantics and operational problems outlined in Chapter 2, the issue of availability is just one of the difficulties that has to be contended with when one tries to quantify ecological versatility. We also need to consider, among other things, the difference between utilization and exploitation (i.e., fitness returns from utilizing resources), the proper identification, enumeration and restriction of resource categories (Real 1975), availability as opposed to abundance for each consumer taxon (Wiens 1989a), spatial and temporal scales of monitoring and variation (Chesson 1985, Powell and Richerson 1985), and differentiation within populations and its effect on versatility (Polis 1984, Wissinger 1992; see Chapters 5 and 6 for a detailed treatment). My impression was that, for all of these aspects, knowledge was even poorer than for the conformance issue.