62 W. J. Streever 



River. However, during periods of high river levels, water from the 

 Suwannee River overcomes the head of pressure generated by aquifer 

 water. Flow in the surface channel reverses, and Suwannee River 

 water enters the cave system. Following a flow reversal in February 

 1991, virtually all of the animals in accessible portions of the Peacock 

 Springs cave system were killed. Animals affected by the kill include 

 crayfish, catfish, and Asiatic clams (Streever 19926). The cause of 

 the kill is unknown. 



This paper addresses two issues. First, I present 1994 P. pallidus 

 census data and subjective observations of other taxa, and second 

 I discuss a possible mechanism for the increase in crayfish population 

 density following the 1991 kill. 



METHODS 



In 1990, divers established eight belt transects (100 x 4 m) in 

 association with permanent guidelines in the Peacock Springs cave 

 system. Transects were numbered 1-8. Both before and after the fauna 

 kill in February 1991, crayfish were visually censused along transects 

 by divers carrying 50-watt lights. Because cave crayfish roam in the 

 open on the cave floor, visual censusing provides a simple method of 

 estimating population density. After the kill, two additional belt tran- 

 sects, called 2-a (240 m long) and 8-a (350 m long), were established 

 in a tunnel north of the Peacock I opening and north of the Peacock 

 III opening, respectively. Transect 2-a incorporated transect 2, and 

 transect 8-a incorporated transects 7 and 8. All crayfish that were 

 visible from the cave's permanent guidelines were included in censuses 

 along transects 2-a and 8-a, making their width dependent on water 

 clarity and size of the cave passage. In general, the width of transects 

 2-a and 8-a varied between 4 and 10 m. The large size of the new 

 transects was intended to allow censusing of the smaller crayfish 

 densities occurring after the 1991 kill. A map of the cave showing 

 transect locations was provided by Streever (19926). 



The 1994 transect censuses were made between 9 January 1994 

 and 25 November 1994. During the 1994 censuses, crayfish were 

 recorded as belonging to a size class of less than or greater than 1.5 

 cm in overall length, estimated from the anterior edge of the rostrum 

 to the posterior edge of the telson. All censuses along each of the 

 original transects were averaged to yield values for each transect 

 before the kill and in 1994. The null hypothesis of no difference 

 between censuses taken before the 1991 kill and those taken in 1994 

 was tested by a paired Mest, with averages of before and after censuses 

 along individual transects comprising pairs (Zar 1984). 



