INVESTIGATION OF POTENTIAL GENE FLOW LIMITATION OF BEHAVIORAL ADAPTATION IN AN ARIDLANDS SPIDER

This study investigates the possibility that gene flow underlies the a pparently maladaptive behavior of a riparian woodland population of th e desert spider Agelenopsis aperta with respect to territorial, foragi ng, and antipredatory behaviors. I found that other local populations of A. aperta in the vicinity of the riparian woodland habitat are prey -limited and exhibit an '' aridlands '' phenotype (high aggressiveness in competitive interactions over energy-based territories and a lack of discrimination among potential prey types). The riparian woodland p opulation deviates from surrounding populations in the area in that pr ey are abundant and this population shows a mixture of ''aridlands'' a nd ''riparian'' (low aggressiveness towards conspecifics and discrimin ation of prey profitability) phenotypes. Electrophoretic analyses of p opulation subdivision in the area indicate that significant levels of gene flow have occurred, at least, sometime in the past. Drift fence a nalyses of spider movement futher indicate that there is marked unidir ectional movement of spiders each year from the more arid habitats int o the riparian woodland. Experimental manipulation of gene flow and pr edation pressure demonstrates that gene flow restricts adaptation in t his habitat: one generation of predation pressure in the absence of ge ne flow is sufficient to cause a marked shift in spider behavior towar ds the expected ''riparian'' phenotype.