Supersaturated and Relaxing Island Faunas: A Critique of the Species-Age Relationship

Abstract

We critically examine Wilcox's recent study which purports to prove dynamic equilibrium theory by demonstrating a species diversity-age relationship for lizard faunas on seventeen Baja California 'supersaturated' islands. We show both conceptual and technical problems with his analysis and propose an alternative explanation for his results. The proposition that species diversity results from balanced rates of species immigrations and extinctions was popularized by Preston (1962) and MacArthur & Wilson (1963, 1967). The experiments of Simberloff & Wilson (1970) and subsequently many others (see references in Simberloff, 1974) have been construed as conclusive support of this hypothesis. For several years, the hypothesis received little criticism. This 'dynamic equilibrium theory' states that species number remains constant because of observable but balanced species extinctions and immigrations, while species composition constantly changes. The existence of immigration, extinction, and 'turnover' (changes in species composition in the absence of changes in species number) is crucial to this theory. Lynch & Johnson (1974) criticized Diamond's (1969) conclusion that species 'turnover' is significant. Subsequently, Smith (1975) suggested that the high apparent rate of turnover was a problem of scale, and that as 'extinction' and 'immigration' has been defined, the equilibrium hypothesis was trivial. Simberloff (1976) recalculated his estimates of turnover for mangrove island arthropods and concluded that although turnover consisting of population extinctions occurred, far more of the original estimates represented either transient intra-population movement or death of individuals which never produced populations in the first place. Thus, the very experiments cited as the foundation for the dynamic equilibrium hypothesis were found not to evidence much turnover. Recently, Wilcox (1978) has contended that if 'non-equilibrium, supersaturated' island faunas are 'relaxing' (decreasing in species numbers), this can be construed as support for dynamic equilibrium theory. Were he to demonstrate such a relationship, it could be taken as support for Diamond's (1972) and Terborgh's (1975) ideas of long relaxation times. However, it has no bearing on dynamic equilibrium theory, since, as mentioned above, the critical requirement for dynamic equilibrium is observable, balanced rates of species immigrations and extinctions with species composition changes in the absence of changes in species number. Not only are Wilcox's (1978) conclusions irrelevant to the controversy involving dynamic equilibrium theory, but his analysis of the island agediversity relationship for lizard faunas on the islands in the Baja California region is beset with conceptual and technical problems that undermine its implication of long relaxation times. Given the omnipresent species-area relationship (Connor & McCoy, 1979) one would expect changes in island area resulting from sea level fluctuations to affect island 0305-0270/79/1200-0311 $02.00 ? 1979 Blackwell Scientific Publications 311 20 This content downloaded from 207.46.13.134 on Sun, 25 Sep 2016 05:00:36 UTC All use subject to http://about.jstor.org/terms 312 Stanley H. Faeth and Edward F. Connor species numbers. Diamond (1972), Terborgh (1975), Simpson (1974), and now Wilcox (1978) suggest that changes in species numbers resulting from such events do not occur instantaneously, but lag several thousand years behind changes in an island's physiographic characteristics, particularly area. Relaxation curves showing this trajectory of species loss over time (Terborgh, 1975) are usually a virtual mirror image of the familiar colonization curve (MacArthur & Wilson, 1963; Simberloff & Wilson, 1970). However, relaxation times are said to be extremely protracted, unlike those for colonization. Relaxing islands are thus supersaturated with species and are losing species in an inexorable return to the 'proper' number of species for a given island's area, and in Wilcox's example (1978), latitude. Wilcox presumes the islands in the Baja California region to be supersaturated, since when the islands were connected to the mainland the resident numbers of lizard species would have been some larger subset of the large mainland pool. Now isolated, the islands presumably retain more species than they 'should' have for their size and latitude. Specifically, Wilcox relates lizard species diversity to island age on seventeen Baja California land bridge islands, as well as to island area, elevation, isolation, distance from mainland, and latitude. He reasons that the observation of an inverse relationship between island age and diversity implies that young islands are still supersaturated and relaxing. He explores this relationship in three ways: (1) simple correlation between island area and diversity, (2) stepwise multiple regression techniques using species diversity as the dependent variable with island age and other physical parameters as explanatory variables, and (3) a type of partial correlation analysis where, after the effects of area and latitude are 'factored out,' residual species diversity is plotted against age. He then construes the plot of residual species diversity versus age as proof that species diversity is 'relaxing' toward an equilibrium value consistent with the present size and latitude of the island.