Natural and Randomly-Assembled Lichen Communities Compared Using the Species-Area Curve

Abstract

Previous studies have indicated that similar species-area curves can be expected in undisturbed saxicolous lichen communities despite differences in species composition and location. However, significant differences in the regression parameters of these curves, especially the slope of the log S/log A line, were found to be caused by disturbance. In an attempt to investigate the possible reasons for this difference, random communities were assembled using species composition data from three natural saxicolous lichen communities differing in the level ofdisturbance. These random communities had the same percent coverage of species and bare area as their natural counterparts; however, the spatial distribution of species was randomized. For the undisturbed locations, the random communities had significantly steeper log S/log A regression lines than natural communities. However, no significant difference between random and natural species-area curves was observed for the community simplified by pollution. These results suggest that simplification results in alterations of community composition that make it more random. They also suggest that undisturbed natural saxicolous lichen communities are assemblages regulated not by random processes, but rather by interactions that limit the number of species that can co-occur in the habitat. The species-area relationship has been investigated for nearly 75 years, and a number of theoretical models have been published (reviews by Conner & McCoy 1979; Kilburn 1966). The most popular of these is the power function model first proposed by Arrhenius (1921) but most closely associated with MacArthur and Wilson (1967) and Preston (1960, 1962): S = CAz in which S = species number, A = area, C is the intercept and z is the power constant. This model is frequently approximated by the log S/log A transformation (log S = log C + log A*z), and assumes a dynamic equilibrium between immigration and extinction rates on habitat islands that vary in size and distance away from colonizing sources. Although there is still much discussion about the biological meaning of the relationship between species number and habitat area, the power function model has been used to explain species-area patterns in hundreds of studies involving a variety of organisms and island situations (Conner & McCoy 1979). A recent study of the species-area relationship in saxicolous lichen communities (Lawrey 1991b), which reviewed original and previously-published (Armesto & Contreras 1981; Orwin 1970, 1972) data, yielded two interesting results: 1) Species-area curves of undisturbed communities do not differ significantly despite obvious differences in species richness, composition and location; however, 2) communities simplified by pollution exhibit significantly steeper curves, suggesting an alteration of the processes that regulate community development. These results demonstrate the utility of the species-area curve as an indicator of disturbance, but also suggest profitable methods of studying the dynamics of saxicolous lichen communities. In attempting to explain why simplified saxicolous lichen communities have steeper species-area 0007-2745/92/137-141$0.65/0 This content downloaded from 157.55.39.83 on Sun, 09 Oct 2016 04:14:46 UTC All use subject to http://about.jstor.org/terms 138 THE BRYOLOGIST [VOL. 95 curves, one must consider factors that accentuate differences in species number between rocks of large and small size. Two potential factors are: 1) A lower diversity of potential colonizers caused by the elimination of sensitive species and a reduced growth of tolerant species; and 2) a reduced level of competition within and between species, which would allow higher species numbers in the largest habitats containing the most numerous colonists. Taken together, these factors might account for the observed differences in species-area curves of disturbed and undisturbed saxicolous lichen communities. Furthermore, if true, these hypotheses indicate the importance of biotic factors (colonizing and competitive abilities of species) in directing development of natural saxicolous lichen communities, and suggest an absence of such factors in pollution-simplified communities. Difficult to test directly, these hypotheses are nevertheless testable if one assumes that the absence of species interactions in communities will tend to randomize species composition. This assumption is reasonable given the rarity of observed random spatial patterns in all but the least developed of natural plant communities. As a general rule, pioneer communities exhibit little regulation by biotic interactions, but tend to become less random as community development continues and species interactions intensify (discussed by Lawrey 1991 a). If pollutioninduced community simplification tends to eliminate biotic interactions and randomize community composition, it should be possible to compare natural and randomly-assembled communities from polluted and unpolluted locations using the speciesarea curve. The present paper presents results of such a comparison.