124 
John R. Holsinger and David C. Culver 
Karst Areas as Islands 
Several authors (e.g., Barr 1968, Culver et al. 1973) have explored 
the potential analogy between caves and islands. Of special interest is 
whether the number of species in a cave or a karst region is determined 
by an equilibrium of immigration and extinction rates when applied to 
individual caves or parts of caves. The time scale is ecological in the 
sense that populations rather than species are becoming extinct. Craw- 
ford (1981) and Culver (1982) have critically reviewed the validity of the 
cave-island analogy in ecological time. On a larger geographic scale, the 
number of species in a karst region may be determined by a balance 
between the rate of isolation of species in caves and the rate of extinc- 
tion of cave species. The time scale for this process is evolutionary 
rather than ecological. 
As Simberloff (1976) points out, there has been an uncritical accep- 
tance of island biogeography theory, and attempts to test the hypothesis 
are few. The best tests of the hypothesis involve direct observations of 
immigrations and extinctions, but such verification is clearly not possi- 
ble for evolutionary time scales. Therefore we must fall back on an 
analysis of area effect. It is often assumed that there is a one-to-one 
correspondence between island biogeography theory and a value of z in 
the following equation: 
S=CA Z (1) 
where S is species numbers, A is area, and C and z are fitted constants. 
Although processes other than an equilibrium between immigration and 
extinction can result in a z-value near 0.26 (Connor and McCoy 1979), 
the validity of the equilibrium model does not require a z-value of 0.26 
(Culver 1982). 
Analysis of area effect can provide some useful clues about the pro- 
cesses that determine species numbers. First, the absence of an area 
effect would indicate that area was incorrectly measured or that 
some other variable and some other process is more important. For 
example, terrestrial cave species numbers might be determined by avail- 
ability of suitable epigean ancestors for which elevation might be more 
important than area. Second, if the island analogy holds, then equation 
(1), sometimes called the power function, should be a better fit than the 
untransformed linear model: 
S=C + zA (2) 
Equation (2) represents a model for area effect where species numbers 
are controlled by passive sampling from the species pool (Connor and 
McCoy 1979), and does not involve a balance between immigration and 
extinction. Third, if the power function is the best fit, then the larger the 
exponent z, the longer the time required to reach equilibrium (Culver et 
al. 1973). A large z-value indicates that it is unlikely the system is in 
equilibrium. What follows is a preliminary analysis, with extensive 
analysis in preparation by the authors. 
