Species–area curves

Abstract

Summary1. The species‐area curve has been studied now for more than one hundred years. In this time four hypotheses have been proposed to account for this pattern ‐ the Random Placement (or Passive Sampling) Hypothesis, the Habitat Diversity Hypothesis, the Equilibrium Theory (or Area Per Se Hypothesis), and the Disturbance Hypothesis. The Random Placement Hypothesis was the first of these and is the simplest, proposing that nothing other than a random placement of species and individuals over area is occurring. This should be considered the Null Hypothesis for species‐area studies and must be tested and rejected before any other hypothesis can be considered viable.2. The Habitat Diversity Hypothesis explains the species‐area curve via the addition of new habitats with increasing area. It is supposed to result in a Power Function curve. It has been invoked a number of times but has not yet been shown experimentally to give rise to a species‐area curve.3. The Equilibrium Theory proposes that species become extinct faster on small islands as a result of the lower population sizes on such islands. This is probably the hypothesis most frequently invoked to account for the species–area curve. It is also said to result in a species‐area curve of the Power Function form. Many of the tests of this hypothesis have, however, been inadequate and on only two occasions has the null hypothesis been tested.4. The Disturbance Hypothesis also proposes that species become extinct faster on small islands, but in this case it is supposed to occur because disturbances are more frequent and more intense on these islands. This hypothesis has only rarely been considered yet it is consistent with many of the observations used to support the Equilibrium Theory. Furthermore, it has been shown experimentally, in one situation at least, to give rise to a species‐area curve.5. Two methods exist by which the null hypothesis of Random Placement can be tested. One is based on probability calculations; the other on sampling in the field. Both methods have been successfully used to test the null hypothesis, and to demonstrate biologically interesting and meaningful patterns.6. The existence of a Power Function curve has been taken to indicate that the Equilibrium Theory is correct. Likewise, an Exponential curve has been taken by some as indicating that Random Placement is occurring. These views can be shown, theoretically and empirically, to be invalid. The fitting of a particular type of curve does not test any of the hypotheses described.7. Species‐area curves can be meaningfully used to measure the relative species diversity of a community, and biologically interesting patterns can be found by comparing communities in this way.8. More intensive studies, testing the null hypothesis and performing manipulative experiments, are necessary if the processes underlying the species–area curve are to be understood.