The Effect of Host Specificity on the Interspecific Associations of Bark Lichens

Abstract

Ten corticolous macrolichen species occurring in eleven upland forest stands in north central Oklahoma were tested for interspecific associations in the forest as a whole and on each of the two major oak species. Results in- dicate that association values for lichen pairs in the forest as a whole are gen- erally not useful since they are composites of the association values on the two tree species. The two tree species are not equally inhabited by all lichens; hence, associations between lichens vary--somtimes radically--from one tree to another. The fact that many bark-inhabiting lichen species tend to be host specific is by now well documented (e.g., Barkman, 1958; Hale, 1967; Adams & Risser, 1971). In attempting to delineate bark communities with the aid of Cole's (1949) index, a means of measuring the degree of association between two species, Hale (1955) found that host specificity presented a problem. As he pointed out, two cryptogams with high preference for the same tree species may appear highly positively associated, when in fact they are distributed independently of each other. Hale demonstrated that this effect of host specificity can be avoided by testing for cryptogamic associations on one tree species at a time, or on groups of similar tree species. This approach is more expensive in terms of computation time and may be impractical when more than a few tree species are involved. Hale, who was working with an unwieldy amount of data and seven major tree species, limited his analysis to a single pair of similar tree species. Our investigation was conducted in north central Oklahoma, where only two tree species comprise more than 90% of the total basal area in upland forests. With forests of such simple composition it is relatively easy to test for lichen associations on each of the important tree species. Measurements of association in the forest as a whole can be compared with measurements on each of the two major tree species, and, al- though causal relationships remain obscure, variations in associations from one host species to another can be detected.