Annual production was calculated for 3 dominant larval odonate populations (Ladona deplanata, Epitheca spp. and Celithemis fasciata) coexisting in the littoral zone of an abandoned farm pond. Dragonfly populations and their prey were collected simultaneously with an Ekman grab at 2 wk to 1 mo. intervals. Production for each dragonfly species was calculated using both the Allen curve method and the removal-summation method. Annual production for the 3 dominant species was .apprxeq. 6 g/m2 (dry wt) and total odonate production was estimated as 8 g/m2. Weekly odonate production was consistently > 0.3 g/m2 from July-Sept. With a mean annual standing stock of .apprxeq. 2 g/m2, the odonate annual turnover ratio (TR) is .apprxeq. 4. Weekly TR were highest during summer months, generally declining as individual size increased. From back calculations of the amount of food required to feed the odonates, it was estimated that a minimal value of annual prey production was 18 g/m2 and a minimal estimate of weekly prey production in the summer months was almost 1 g/m2. With an observed prey standing stock of 0.6 g/m2, this would necessitate an annual prey TR of 30 and a weekly summer TR > 1. Since such an indirect method of calculating prey production does not account for other sources of prey mortality such as fish predation and emergence, the TR are considered minimum values, perhaps only half of actual values. These relatively high TR values agree quite well with estimates based upon the amount of food required to sustain fish predators. The significance of these estimates made using indirect methods is that they are considereably higher than estimates made using more traditional direct methods. An attempt is made to resolve this discrepancy by examining sources of error associated with both approaches. Furthermore, since odonates have a standing stock frequently 2-3 times that of their prey and a large daily consumption capacity, prey are probably saved from annihilation because they can find sufficient refuges.
