Skip to main content

Determinants of plant species richness on small Bahamian islands

Buy Article:

$51.00 plus tax (Refund Policy)



I investigated the determinants of plant species richness in two archipelagos, comparing the predictive power of different explanatory variables. I evaluated both conventional variables and alternative variables not commonly used in such analyses. I also investigated the effect of island location in relation to neighbouring landmasses on plant species richness and the predictive ability of regression models. Location

Archipelagos of small islands in the central Exumas and near the north-east coast of Andros, Bahamas. Methods

I surveyed plant species richness and obtained measures of seven predictor variables: total island area, the ratio of perimeter to total area, vegetated area, the ratio of vegetated area to total area, distance to the nearest large island, elevation and protection from surrounding islands. All seven predictor variables were evaluated as determinants of plant species richness in simple and stepwise multiple linear regressions. Analyses were conducted for each archipelago overall, and then separately for three categories of islands in the Exumas. Total area, elevation, and distance were evaluated as predictors of vegetation incidence in simple and stepwise multiple logistic regressions for both archipelagos. Results

Some expression of insular area was always the best single predictor of plant species richness in the linear regressions. Total area was a relatively poor predictor compared with other expressions of insular area. Distance, elevation, and protection explained relatively little of the overall variation in plant species number, although all variables were selected as significant in some models. A greater amount of variation in plant species richness was explained by the linear regression models in the Exumas (69.0%) compared with Andros (60.9%). Different variables were entered into the models for the three categories of islands in the Exumas, and adjusted coefficients of multiple determination ranged from 68.9% to 85.7%. In the logistic regressions, the model including total area and distance yielded almost 90% correct classification of vegetation incidence in the Exumas; no significant variables were selected for Andros. A group of exposed, outer islands supported many fewer species than more sheltered islands, on the basis of total island area or elevation. Main conclusions

The three variables commonly used in studies of determinants of insular species richness – total island area, distance, and elevation – were relatively poor predictors in most analyses. Alternative expressions of insular area – indicative of disturbance or shape in combination with area – were usually better predictors than total area and may more realistically reflect habitable area. Alternative predictors explained similar amounts of variation in plant species richness compared with commonly used predictors, and combinations of all variables into a single stepwise model resulted in increased predictive power. The predictive power of the models tended to be higher for groups of islands that were more sheltered by neighbouring islands. Exposed islands, although separated by relatively small distances from nearby protected islands, may be impacted by storms much more severely and possess many fewer species. The location of small islands relative to large landmasses, as well as their geological histories, should be taken into account in such analyses.

Keywords: Bahamas; disturbance; insular species richness; island biogeography; linear regression; logistic regression; predictor variable

Document Type: Research Article


Affiliations: School of Biological Sciences, Section of Integrative Biology and Brackenridge Field Laboratory, University of Texas, Austin, TX, USA

Publication date: 2002-07-01

  • Access Key
  • Free ContentFree content
  • Partial Free ContentPartial Free content
  • New ContentNew content
  • Open Access ContentOpen access content
  • Partial Open Access ContentPartial Open access content
  • Subscribed ContentSubscribed content
  • Partial Subscribed ContentPartial Subscribed content
  • Free Trial ContentFree trial content
Cookie Policy
Cookie Policy
Ingenta Connect website makes use of cookies so as to keep track of data that you have filled in. I am Happy with this Find out more