Members of the insect order Odonata are excellent examples of organisms that demonstrate complex life histories. Both the larval and adult stages must be studied to understand the dynamics of such species, A population of the dragonfly Pachydiplax longipennis (Odonata: Libellulidae) was studied at a small fishless pond in north central Illinois in 2008 and 2009. Additionally, a dynamic population model of the species was developed using the graphical modeling software, STELLA, to further understand the life history dynamics of P. longipennis.
The larval dragonfly community in the pond was composed of nine species, all of which were also present as adults. The adult dragonfly community contained an additional four species, for a total of 13. Although, the maximum larval density of Pachydiplax longipennis, which occurred in the middle of the summer, was ∼15 m -2 in 2008 and 2009, mean density was higher in 2009. Based on this maximum density, it was estimated the maximum larval population size for the pond was ∼170,000. Head capsule width and total length of larvae were used to identify 14 larval instar classes for the species. Changes in head capsule width between adjacent in star classes generally conformed to Dyar's Ratio, with the exception of the changes between the first and last two instars. Skipping of instar classes was common among larvae reared in the lab. Mean maximum P. longipennis adult abundance occurred in July in both 2008 and 2009. It was higher in 2008 than that observed in 2009, ∼12 per 10 m sector versus 8 per 10 m sector. The estimated adult population size in 2009 based on mark-recapture data using Craig's estimation method was 2,000. Average clutch size, determined from six captured, mated females, was 1,238±431 eggs per clutch. Average clutch survivorship was 27.51%±16.38.
A density-ceiling model generated a stable population of Pachydiplax longipennis larvae and adults that cycled in 54 week intervals. Short term (2 years) results predicted an early instar larval population of ∼175,000 individuals, a late instar larval population of ∼40,000, and an adult population of ∼4,000. Long term (20 years) results predict early instar larval population of ∼300,000 individuals, a late instar larval population of ∼75,000, and an adult population of ∼6,000. Long term estimates were comparable to those predicted by larval and adult sampling. Sensitivity analysis of varying mortality rates found that changing early instar larval mortality rate had a significant impact on observed abundances in all modeled life stages, while changes in breeding adult mortality had little effect. Simulations of ten different survivorship scenarios of larval and adult mortality resulted in three specific categories of response in terms of larval and adult abundances: one or both reached carrying capacity, both went extinct, or either or both stabilized at an intermediate abundance. Scenario results also suggested a greater importance of larval stage mortality rates, similar to the results of the sensitivity analysis. A density-dependent model generated unrealistic results in both short term and long term simulations.