Island Diversity
In lieu of an abstract, here is a brief excerpt of the content:

25 Island Diversity The flora of all the islands under consideration includes 388 taxa (species , subspecies, and varieties) in 251 genera and 78 families (appendix A). The majority of these are found on Isla Tiburón. A statistical breakdown of the island flora is presented in table 1.1, in the order that the islands are described in this chapter. Asteraceae, Fabaceae, and Poaceae are the largest families on the islands. It is also of note that Cactaceae is the fifth most species-rich family in the total flora of the islands. These findings compare well with those for the flora of the Sonoran Desert as a whole (Felger 2000a; Van Devender et al. 2010; Wiggins 1964). EachislandintheMidriffRegionisauniquedesertmicrocosm.Thegeologic histories and varied topographies of the islands set the stage for their botanical diversity, and a suite of biotic and abiotic factors shapes the community of plants present on the islands. The principal factors that control the floristic diversity on the Gulf Islands are topography and island size, isolation (or lack thereof), soil conditions, aridity, and the presence or absence of herbivores. Distance and isolation among Gulf of California islands are not nearly as extreme as in oceanic archipelagos like Hawaii and the Galapagos; however, the intermediate level of isolation has allowed radiation to occur in certain taxa of flora and fauna. An expected relationship of increased diversity of plant species is seen with increasing island area and topographic heterogeneity (including elevation). Tiburón, the largest island in Mexico, is correspondingly the most diverse island in the Gulf, and the rugged Sierra Kunkaak is the most diverse area of the island. To assess the area–diversity relationship for the Sonoran Islands, we used the typical species–area model employed in island biogeography theory: s = kAz, where s is the number of species, A is island area, z is the exponent, and k is a scale coefficient (MacArthur & Wilson 1967). An estimation of z was made by calculating the slope of the log–log function when island area and species diversity are graphed (fig. 1.19). By regressing the species richness of each island against the exponent-transformed area (Az), k was estimated. Knowing k and z, the expected number of species for each island was calculated . Using Pearson residual analysis (Duffy 1990), we were able to identify 26  The Islands and Their Vegetation Figure 1.19.  Species–area relationships of vascular plants on the Sonoran Islands on log-transformed axes. The slope of the line (z) is 0.304 and r2 = 0.71. The majority of the Sonoran Islands fall very close to the regression line, but with three outliers. Two islands have significantly fewer species than expected (Patos, and Mártir; open circles) and one island has significantly more species than expected (Dátil; line filled circle). Graph by Exequiel Ezcurra. the islands that have significantly more or significantly fewer species than predicted by McArthur and Wilson’s model of island biogeography (table 1.2). The islands identified as having depressed levels of diversity—San Pedro Mártir and Patos—are guano islands with high levels of nitrogen and phosphorus, which undoubtedly exerts strong chemical control on the flora (Felger & Lowe 1976; Wilder & Felger 2010). Dátil, which is remarkably rich in species, is a topographically diverse land-bridge island closely associated with Tiburón. Comparing the similar-sized islands of Nolasco and Dátil, which have skewed diversities (58 and 101 species, respectively ), shows the importance of the geologic legacy of an island. While Dátil is a land-bridge island, Nolasco has been geologically isolated for several million years, and its diversity is consistent with the expected number of species predicted from the above model (for more detail, see Felger et al. 2011). It is also interesting to note that the flora of Cholludo, both a guano and land-bridge island, while relatively rich in species, also does not significantly differ from the prediction. It seems that the close proximity and past connection of Cholludo to Tibur ón outweigh the filtering of species imposed by the guano soil conditions. This may provide evidence that the geologic history of an island is the most important factor controlling its observed diversity. Inter-island distance is not sufficient to be a primary controlling factor on the flora of the islands due to the short distances (certainly the case for Tiburón) and vagrant species, such as seabirds, which create connections between the islands and the mainland (Rose & Polis Table 1.1  Statistical Summary...