Occupancy patterns of viperid and Colubrid snakes on islands in the Sea of Cortez

Background/Question/Methods

Based on expectations from number of mainland source species, islands in the Sea of Cortes are disproportionately occupied by populations of viperid snakes (rattlesnakes) when compared with colubrid snakes. Rattlesnakes are viviparous, produce toxic venoms for consumption of large prey, and have lower mass-specific metabolic rates—traits that may be advantageous for overwater dispersal to, and establishment on, islands with harsh and stochastic environments. However, theses same traits may also hinder population persistence in such environments over ecological time-scales because rattlesnakes must divert more resources to reproduction, their venoms may be energetically costly to produce, and they have greater absolute metabolic rates because of greater absolute mass. To test whether rattlesnakes disproportionately occupy islands because of more frequent colonization (higher immigration rates) or because of longer persistence on islands (lower extinction rates), we modeled logistic probability functions of species richness as a function of island area based on empirical distribution data, and interpreted our models in the context of predictions from Island Biogeography Theory.

Results/Conclusions

The island area threshold where the probability of one species present exceeds the probability of no species present is greater for rattlesnakes than it is for colubrid snakes (2.29 km² vs. 1.25 km²). For rattlesnakes the probability of one species present on an island peaks at 7.08 km², and for two species present at 398.1 km² (for colubrids, the probability peaks are at 3.16 km² and 20.0 km², respectively). In contrast, the cumulative probability of one rattlesnake species occurring on a very small island (< 1 km²) is greater than it is for one colubrid species (0.26 versus 0.18). Thus, although greater area is required to support populations of rattlesnake species, rattlesnakes are also more likely to be found on very small islands. Furthermore, the probability function for colubrid presence on islands is much steeper than it is for rattlesnakes, indicating that colubrid faunas have relaxed to equilibrium. Collectively, our results support the premise that rattlesnakes are disproportionately represented on islands because they are better overwater dispersers rather than because they have a longer time to extinction.