Species-area relationships were studied for anoline lizards on 147 islands in the Caribbean. The relative importance of ecological and evolutionary factors in determining species number varied with island size. On small islands, only ecological factors affect the species-area relationship. Further, the importance of different ecological factors such as colonization, competition, and extinction varied among different types of islands. On landbridge islands, differential extinction as a function of island area appears to play a key role in producing a species-area relationship. By contrast, limited colonization success generally prevents ocean islands from accumulating more than two species. Among the larger islands, evolutionary factors are the primary determinant of species number. Detailed examination of several components of evolutionary diversification indicates that the species-area relationship among the Greater Antilles primarily results because larger islands have increased number of habitat niches occupied and a greater number of closely-related species that are ecologically similar and allopatrically distributed; increased subdivision of certain habitats plays a lesser role.
 

         The species-area relationship is as close as the field of ecology has to a universal law.  I investigated whether a relationship exists among Caribbean island area and number of species of Anolis lizards.  I used a data set of 145 islands published by Stan Rand in 1970 in Breviora, supplemented with data from several other islands.  Island area encompasses six orders of magnitude; species number ranges from 1 to 56 (Cuba).  No data were available for anole-less islands, but almost every island of any size in the Caribbean has at least one species of anole.

         Overall, a significant relationship exists; larger islands tend to have more species than smaller islands.  However, the relationship is not strong (r²=0.29).  Moreover, a linear regression does not provide a good fit, as the line appears to curve upward among the largest islands in the Caribbean, the Greater Antilles.

         Detailed examination of the islands reveals that the overall species-area relationship is actually a composite of different types of islands, each displaying its own pattern and underlying cause.  These classes of islands are:

1. Land-bridge islands.  These are islands that used to be connected to the larger islands of the Greater Antilles, but whose connection was severed by rising sealevels.  Examples include the many islands of the Virgin Islands and a variety of islands around Hispaniola and Cuba.  A strong species-area relationship exists among these islands.  All species on these islands are also found on the "mainland."  Hence, this pattern is caused by area-dependent extinction (i.e., "faunal relaxation").

2. Great Bahamas Bank--10,000 years ago, the landmass of the Great Bahamas Bank was entirely emergent, forming an island rivalling Cuba in size.  Rising sealevels fragmented this bank into thousands of islands.  One to four anole species are found on these islands and species area is strongly linked to island area.  As with category #1, this is a clear example of faunal relaxation.

3. Oceanic islands--These are islands that are separated by deep-water and have never been connected to larger landmasses.  Species on these islands are often endemic.  Among these islands, no species-area relationship exists and no more than two species can be found. Phylogenetic evidence indicates no evidence for within-island speciation (i.e., if two species occur on an island, they are not closely related).  Three categories of oceanic islands exist, but they show no notable differences:

 a. Other Bahamian Banks (e.g., Little Bahama Bank, San Salvador)
 b. Greater Antillean oceanic islands (e.g., Navassa, St. Croix, Grand Cayman)
 c. Lesser Antilles

        It is interesting that no within-island speciation has occurred anywhere in the Lesser Antilles despite the large size and rugged topography of some islands, such as Guadeloupe and Martinique.  Moreover, some of these islands also have a number of small, offshore islets that in theory could provide opportunities for peripatric speciation.  Molecular analyses suggest that anoles have occupied these islands for long periods; thus, recency of arrival is not an explanation for lack of speciation.  Rather, there appears to be some threshold island size below which speciation does not occur.

4. Greater Antilles--On these islands, specation runs rampant.  Based on existing phylogenetic information (subsequently confirmed by more recent molecular studies), relatively little inter-island exchange has occurred.  Hence, most of the diversity on islands has resulted from within-island diversification.  This finding explains why the slope of the species-area relationship increases at large island sizes: evolutionary factors are kicking in, supplementing the ecological factors responsible for patterns on smaller islands.

         Among the four islands of the Greater Antilles (Cuba, Hispaniola, Jamaica, and Puerto Rico), a strong species-area relationship also exists.  Examination of the lineages on these islands suggests the basis for this relationship.  The extent of allopatric speciation, as measured by the number of "allospecies" (i.e., closely related, ecologically-similar species with allopatric distributions) is a function of island area.  This is not surprising, as the opportunity for isolation probably increases as a function of island area (particularly given the topography of Cuba and Hispaniola, both of which are dissected by several mountain ranges). 

        In addition, the number of habitats occupied, as indexed by morphology, is also a function of island area.  Why this should be is less clear, because even the smallest island, Jamaica, is topographically and climatically quite diverse.  It is not immediately obvious that there are habitats present on, for example, Cuba, that are absent on Jamaica.  For example, Cuba has species specialized for using habitats such as cave walls and river streams and Hispaniola has a species adapted to life in the leaf litter.  Yet, all of these habitats (or "niches")  are surely present on Jamaica and Puerto Rico, but no species has adapted to use them on these islands.  Why larger islands have more habitats occupied is a question worth pursuing.