Assessing patterns of senescence in Drosophila mojavensis reared on different host cacti.

Questions: Can senescence observed under laboratory conditions predict senescence under conditions thought to contribute to expression of intrinsic mortality rates in nature? Is senescence, under contrasting environmental conditions, described by alternative mortality models or by a single one with different parameter values? Organism: Drosophila mojavensis. Many populations of this cactophilic species use one of two principal hosts, pitaya agria cactus (Stenocereus gummosus) or organ pipe cactus (Stenocereus thurberi) to carry out their life cycles. Methods: Flies were grown on both cactus hosts and standard laboratory food over their entire life cycle. Adult mortality rates and mean longevity were calculated in all adult cohorts. We employed maximum likelihood procedures to determine which of four statistical models best described the mortality trajectories of these flies. Conclusions: Mortality rates of flies grown on cacti were best described by Gompertz and Gompertz-Makeham models, whereas flies grown on laboratory media were best described by Logistic and Logistic-Makeham models. Rates of mortality decelerated at older ages in individuals grown on laboratory media, but not in cactus-reared flies. Models commonly used in Drosophila laboratory studies may be inadequate to accurately assess the shape of natural mortality risk functions.