Development time and host-parasitoid stability: an experimental test

Background/Question/Methods

Recent theoretical studies have demonstrated that parasitoid-host stability is strongly enhanced by either a decrease in the duration or an increase in the variability of the duration of the vulnerable host stage. To date, there have never been unambiguous experimental tests of these theoretical predictions. The cowpea weevil Callosobruchus maculatus and its parasitoid Anisopteromalus calandrae were used as a model predator-prey system. In Experiment 1, the duration of the invulnerable juvenile host stage was experimentally lengthened by 60% (long-duration treatment) or reduced by 60% (short-duration treatment) relative to unmanipulated controls. Based on theoretical models, an increase in the duration of the invulnerable host stage (T_H1) should induce generation cycles in the host. In Experiment 2, the development time of the vulnerable host stage (T_H2) was experimentally made to be more variable (high-variance treatment) or to have approximately normal variability (normal-variance treatment). Increased variability in T_H2 should be strongly stabilizing to the system. 

Experimental microcosms consisted of large petri dishes containing cowpea weevils, parasitoids and moth beans (Vigna acontifolia) as hosts. Six replicates of each experimental treatment and controls were used and the experiments were run for approximately two years (> 30 generations).

Results/Conclusions

Unmanipulated control populations showed clear evidence of long-period cycles (ca. 100 days or 3 generations) similar in length to those predicted by our models for the cowpea weevil and parasitoids.

Microcosms containing only the cowpea weevil showed no evidence of cycles, of any period, suggesting that the host-parasitoid interaction is responsible for cycling in host populations. For Experiment 1, manipulating T_H1 strongly affected host-parasitoid population dynamics. In the short-duration treatment, host densities were higher and parasitoid densities were lower, and for both species, variability in densities were lower than for control microcosms (i.e., relatively stable host-parasitoid dynamics). In the long-duration treatment, host-parasitoid dynamics were much less stable, with shorter cycles and higher amplitude fluctuations in both host and parasitoid populations. For Experiment 2, the data fit model predictions. Although this experiment will not be complete for a few more months, results suggest reduced amplitude fluctuations in the high- as compared to the normal-variance treatments. Together, these results strongly support existing theory that age-structure is a key factor in promoting complex predator-prey/parasitoid-host population dynamics.