Notes on Niche Relationships and Community Structure

Clearly the various characteristics of the niche, the limits to similarity and overlap between ‘adjacent’ niches, the phenomenon of niche separation, all have profound consequences in the structuring of communities.

An understanding of the fundamentals of niche structure may allow us to draw a number of conclusions about the way in which communities may be organised. This indeed is one of the main hopes stimulating all the recent work in the niche.

Schoener (1974), as we have already noted, has suggested that restrictions of niche structure are such that at least along a single resource dimension organisms should be regularly spaced within the community; from different premises DeVita (1979) draws the same conclusion.

Pianka (1981) points out that such over dispersion of species might be expected (from considerations of potential competition) even in multidimensional resource space, with each species minimising its interactions with all others.

Other constraints of niche composition or limits to overlap and separation have further implications for community structure: and our understanding of the fundamental principles of niche relationships is potentially very powerful in offering an analysis of community design.

But, how does our theoretical conception of niche structure relate to the real world? To what extent is this regular spacing of organisms in resource space a real phenomenon? How many, and which, niche dimensions are actually important in separating species? Are individual niche even the basic unit involved in community structure, or are organisms associated into clusters of related niches of functionally similar species (guilds’), with organization of resource space undertaken with respect to these guilds as the basic unit rather than individual niches?

There is certainly considerable evidence that such clustering of species does occur. But in such case are these guilds merely a result of built-in design constraints upon consumer species, does the apparent clustering simply reflect natural gaps in resource space, or can guild structure evolve even when resources are continuously distributed as a means of reducing diffuse competition? And what factors are involved in the spacing of individual species within the guide? (Pianka, 1980). If brief, our theoretical knowledge of community design is not yet strong enough; our understanding is not as yet sufficiently complete for this approach, to be used in a practical way in community analysis.

There are still many problems unresolved. Perhaps, the biggest problem, apart from that of translation of the theory itself into parameters which can be measured in real communities, is the converse: that of measuring the appropriate characteristics or real communities which an be used to advance the theory, and answer some of those questions we have catalogued here.

Attempts to define some of the ‘rules’ of community design have concentrated on studies of the separation or overlap of niches within real communities (e.g.,) MacArthur, 1970; May, 1974; Yoshiyama and Roughgarden, 1979; Rappoldt and Hogeweg, 1980), extending this to an analysis of what is referred to as ‘species packing’ within communities.

But it is extremely difficult to synthesize the conclusions of all this work, and the implications for community design: the literature itself is confused and conflicting. Different authors appear to work from different premises, and have worked in markedly different communities which may well not obey the same ‘rules’. (Thus, in an analysis of changes in niche relationships with increasing species diversity in communities of desert lizards, Pianka (1973) found evidence of decreased niche overlap with increasing species number.

Cody (1974) using methods similar to those of Pianka, in an analysis of bird communities in scrub habitats, came to the diametrically opposite conclusions: finding that average overlap between adjacent niches’ within a community increased with increasing diversity.)

In addition, ‘packing’ is variously defined by different authors in terms of the number of species that can be accommodated per unit volume of resource space (or in a single dimension, per unit distance of a resource continuum) (e.g., Roughgarden, 1974; Rappoldt and Hogeweg, 1980), as the closeness of packing of resource utilization distributions for non-overlapping species (Pianka, 1975), the extent of overlap observed in resource utilization within the community or between adjacent pairs of species within the community (MacArthur, 1970; Roughgarden and Feldman, 1975).

While all these do indeed represent facets of the way species are packed into a community it is true, all equally represent rather different facets.

The number of species which can be accommodated is a measure of community fill; the pattern in which these species are arranged within niche space-their actual spatial relationship to one another-is an aspect of packing design: while the closeness of packing of nonoverlapping species, or the degree of overlap within the community are derived functions of the extent of community-fill, the packing design and the type of organism involved. Nor are these separate elements necessarily related to each other in any simple way.

Failure to distinguish these different aspects, or assumption that they show necessary correlation, leads to a great deal of confusion. Thus extent of resource overlap for example (the measure most commonly referred to as ‘species packing’ (e.g., Roughgarden and Feldman, 1975) is not as has sometimes been implied a valid index of community fill at all, indeed it is not necessarily related to community fill in any way.

For while it is the case that increased ‘fill’ of a community may will be reflected by an increase in niche overlap in many instance (as Cody, 1974), an increase in the number of species within a given resource space in other circumstances may equally be accommodated without such increased overlap but by restriction of niche widths. (Thorman (1982) has shown a decrease in niche widths amongst guilds of estuarine fish with increasing species number.) Yet again, a greater number of species may be accommodated merely by expansion of the actual resource spread exploited (as Pianka, 1973: this is why increased species number was accompanied by decreased niche overlap in this study of desert lizards).

Measurement of the extent of niche-overlap within the community is no more and no less than a measurement of niche overlap. Likewise closeness of packing of non-overlapping niches is precisely that and no more. Extension beyond this to rules of community organization is fraught with complexity.