No abstract

The mojavensis cluster of the repleta species group of Drosophila (Drosophilidae: Diptera) consists of three species. One is newly described as D. navojoa. A second species, described here as D. arizonae, replaces D. arizonensis, which has become a junior subjective synonym for D. mojavensis, the third species in the cluster. A phylogeny of the three species is presented, based on chromosomal inversions, morphology, and the ability to produce hybrids. Breakage points are assigned for all inversions, and male genitalia are figured; 186 crosses were made from 225 possible combinations among 15 geographic strains from the southwestern United States, Mexico, and Guatemala. It is confirmed that D. mojavensis and D. arizonae are very closely related and shown that D. navojoa is more distantly related in regard to all criteria. This relationship is supported by the geographical positions of the ancestral gene sequences in each species, which show a sequential northwest movement (D. navojoa----D. arizonae----D. mojavensis) from southern Mexico to southern California and northern Arizona. The relationship is also supported by the fact that D. navojoa breeds in Opuntia cactus, an ancestral behavior, whereas the other two species breed chiefly in Stenocereus cacti, a derived behavior. The possible role of this host plant shift in speciation is discussed.

No abstract

Three types of tests were conducted to determine the causal factors for the narrow larval niche breadth and the low amount of larval niche overlap observed among common species of Drosophila breeding in rotting cereus cacti in the Sonoran Desert. (1) Field test and observations show that host plant discrimination for feeding adults is very high but not absolute. (2) Substrate substitution experiments conducted individually with four species of cactiphilic Drosophila on six kinds of artificially rotted cacti show that every species except D. mojavensis reproduces bests on the cactus (or cacti) on which it is resident in nature. Even so D. mojavensis and D. arizonensis, both of which are polyphagic, emerge in large numbers from five kinds of cacti. (3) Laboratory competition tests with mixed species established in three cases that the resident species was unaffected, as compared with monoculture controls, while the alien species usually emerged in significantly fewer numbers. In two other competition situations neither the alien nor resident species was significantly affected. Inability of the polyphagic species to invade certain non—host cacti in nature could be due to competition in the first three cases. In the latter two cases either the physical characteristics of the substrate or the climate affecting distribution (and/or density) appears to be the principal restricting factor. Narrow host plant specificity is accompanied by specialized nutritional requirements and more continuously available rot pockets in the host plants.

No abstract