Variation in host plant quality: influences on the mating system of a desert grasshopper

Previous field work on the grasshopper Ligurotettix coquilletti revealed that females were not evenly distributed among male mating territories, Larrea tridentata (creosote) bushes, but were clustered at particular locations. These sites generally harboured several signalling males simultaneously and also possessed foliage preferred by the insects as a food source, this preference being based on the relative absence of various extra-foliar compounds. The clustering of females, therefore, could result from a preference for specific bushes because of the resources (i.e. food) available there and/or an orientation to groups of males per se. Here, a field experiment designed to test the former alternative is described. For this experiment 24 bushes with known histories of prior occupancy were selected and epigamic effects (i.e. male signalling) were equalized among the bushes over a 7-week period by maintaining one singing male per bush. Monitoring the distributions of both naturally occurring and released females showed that female settlement during the experiment resembled that seen under natural conditions, i.e. females still accumulated on those bushes with histories of high occupancy. When females whose chemosensory organs had been removed were released in the study plot, however, settlement occurred indiscriminately. It is concluded that, while female orientation to groups of males may occur, variability among bushes in their foliar chemistry alone appears sufficient to promote a skewed distribution of females.