Losos, J.B., and D. Spiller. 1999.  Differential colonization success and asymmetrical interactions
        between two lizard species. Ecology 80: 252-258.


         Small islands in the Bahamas often do not contain Anolis lizards.  We took an experimental approach to investigate whether absence of A. carolinensis and A. sagrei from these islands results from the inability of populations to become established and whether the presence of the other species has a negative effect on population density and probability of successful colonization.  We introduced propagules of five lizards of each species, either in allopatry or sympatry, on five blocks of three islands each.  Anolis sagrei populations thrived on almost all islands, whereas within three years, many A. carolinensis populations had gone extinct and most others were very small.  We conclude that an inability to successfully establish populations on small islands might explain the absence of A. carolinensis from these islands, but not the absence of A. sagreiAnolis carolinensis was not more likely to become extinct on islands containing A. sagrei, but population densities of A. carolinensis were lower on sympatric islands compared to allopatric islands.  In addition, A. carolinensis tended to perch higher, relative to available vegetation, on sympatric islands than they did on allopatric islands.  Conversely, the presence of A. carolinensis had little impact on A. sagrei populations.  However, inspection of population changes through time suggest that A. carolinensis may have initially had an effect on A. sagrei, but this effect decreased as A. carolinensis populations declined.  The observed absence of A. sagrei from these islands may indicate that its dispersal abilities are limited relative to the frequency of devastating hurricanes.

Non-Technical Summary

        Small islands in the Bahamas (islands so small--perhaps the size of a baseball diamond--that they officially are termed "rocks") often do not contain lizard populations.  When a lizard species is present on such islands, it is invariably Anolis sagrei, which is the most terrestrial of the four species of Anolis found throughout the Bahamas.

        Two hypotheses could explain these patterns.  First, the absence of lizards from most of these islands could result from lack of suitable habitat.  These islands have sparse and scrubby vegetation with few trees or other large plants.  This might explain the complete absence of A. carolinensis, which is a more arboreal species, usually found in trees at least a meter above the ground.  Alternatively, the absence of one or both species might result from inability to colonize the islands.  Although most of these small islands were connected to larger landmasses during the last ice age, when sea-levels were lower, the islands are small enough that populations might have perished since then due to the problems that plague small populations.  Alternatively, rare catastrophic events, such as hurricanes, might have extinguished such populations. Because small islands are low-lying, populations on these islands would be particularly vulnerable.  Whatever the cause of extinction, if colonization ability of these species is limited, then the absence of these species from small islands might result from the inability to recolonize islands by overwater dispersal.

        In addition, the absence of A. carolinensis might be a result of interspecific interactions.  If A. sagrei has a greater effect on  A. carolinensis than A. carolinensis has on A. sagrei, then the absence of A. carolinensis from islands suitable for lizard occupation might result.  Such effects could occur in many ways, including interspecific competition and predation.

        To test these hypotheses, we introduced propagules of the two species to small islands in the vicinity of Georgetown, Greater Exuma, Bahamas.  We located 15 islands lacking lizards and divided these into five groups of three similar islands.  Within each group, we randomly chose one island to receive A. carolinensis, one to receive A. sagrei, and one to receive both species.  All introductions comprised two adult males and three adult females; islands receiving both species thus received five lizards of each species.  Most, perhaps all females, were gravid at the time of introduction.

        A year after the introductions, all populations (with the exception of one population of A. carolinensis) had at least doubled in size to 10 lizards.  Several of the A. sagrei populations had increased enormously, one to an estimated 180 lizards!  Thus, in the short term, lizard populations had no problem persisting on these islands.  However, by three years after the initiation of the experiment, the story was quite different.  Nine of the 10 A. sagrei populations were thriving (one, for mysterious reasons, had crashed to one individual, a female; in subsequent years, this populations has recovered).  By contrast, the story for A. carolinensis was quite different.  Four populations had completely perished.  On another two islands, only a single, male lizard remained.  On an additional two islands, less than 10 individuals survived.  Only two islands had secure populations (one, however, was quite dense, with a population > 100 lizards).  Hence, the conclusion is that, for some reasons, most small islands are inhospitable for A. carolinensis.  What this reason is remains unclear; our guess is that it has to do with the high temperature and aridity of the islands during summer months.  Further biophysical research will be needed to understand this point.

        In addition, several lines of evidence suggest that the presence of A. sagrei has a negative effect on A. carolinensis.  First, in the first years of the experiment, when the populations of A. carolinensis were increasing, populations on islands without A. sagrei reached higher densities than populations on islands with A. sagrei.  Furthermore, the presence of A. sagrei led to a shift in habitat use by A. carolinensis.  Thus, A. carolinensis is in a bad situation on these small islands.  First, even on islands without A. sagrei, times are tough (three populations went extinct on such islands).  However, the presence of A. sagrei, which appears to thrive on these islands, further exacerbates the problem.  Hence, it is not surprising that A. carolinensis never occurs on these islands (however, we have no data on whether A. carolinensis actually is able to colonize these islands in the first place).

        By contrast, our date clearly indicate that A. sagrei can survive and prosper on these islands.  Thus, their absence from many small islands must result from factors causing extinction and a relatively low rate of recolonization.  Given the high population densities achieved on these islands, small-population related problems (e.g., inbreeding depression) are not likely to operate.  A more likely candidate are rare catastrophic events, such as hurricanes. Three years into this study, we unexpectedly had the opportunity to test this hypothesis.

Return to Losos Lab Home Page

Return to Curriculum Vitae