Habitat Partitioning by Therevids at Sand Ridge State Forest

By Marianne Hartman, Michael Irwin, and Gail Kampmeier, Center for Economic Entomology

March 19, 1996

If you were a ground-dwelling insect larva, one of your greatest fears would be to meet up with a stiletto fly larva (family Therevidae). These small, snakelike predators swim through sandy soils in search of food. They devour any insect larva they encounter, including other therevids. Because of their voracious appetites, the stiletto fly larvae impact populations of many root-feeding arthropods in a major way. Their biodiversity may be useful as an ecological indicator because fluctuations in therevid populations coincide with fluctuations in the populations of the many arthropods on which they feed. They are also potential biological control agents of agricultural pests such as corn rootworms in sandy soils.

Marianne Hartman checks a malaise trap in the forest
INHS researcher Marianne Hartman checks a malaise trap at Sand Ridge State Forest.

We are studying the bioecology of therevids at Sand Ridge State Forest in Mason County, Illinois, where several species of adult therevids have been found. Nothing is known about how the larvae of different species coexist in such a habitat. Without some method of isolating therevid species, all of which are general predators, the better-adapted species would likely eliminate others by outcompeting them for food. The researchers on this project hypothesized that adult females of the various species lay their eggs in different microhabitats and the larvae develop at different times of the season, thus reducing potential interspecific competition.

More than 200 emergence traps were set out at Sand Ridge State Forest to test some aspects of this hypothesis. The emergence traps, made from upside-down buckets with an opening covered by a funnel and collecting cup, were placed randomly along 300 meters of a trail. Therevids in the sandy substrate beneath the traps were caught when they matured to adulthood and flew. Soil type, shadiness, ground cover, and vegetation were described around each emergence trap to depict the microhabitats of the traps. Two malaise traps (5 m long tent-like screens with poison collecting cylinders at either end) captured adult flies as they flew through the forest or along the trail. The malaise and emergence traps were checked for therevids once a week during the summers of 1992 and 1993.

The trail along which the traps were placed divided the area into two major habitats: natural and managed forest types. An undisturbed oak-hickory forest existed north of the trail, and a planted pine forest was located south of the trail. By looking for differences in therevid populations between the two forest types, we thought it would be possible to assess how replacing a natural with a managed forest affects populations of root-feeding arthropods.

The malaise traps captured nearly 400 therevids in 1992 and 100 in 1993. The emergence traps collected 75 and 150 therevids in 1992 and 1993, respectively. In total, some 6 species in 4 genera of therevids were trapped in malaise traps, while only 5 species in 3 genera were found in the emergence traps. The absence of 1 species from the emergence traps is an indication of habitat partitioning. This species apparently develops in a habitat out of the range of the emergence traps.

emergence trap
Example of an emergence trap used in the therevid study.

There was no obvious correlation between any therevid species and a microhabitat, and all 5 species of therevids from emergence traps occurred in each habitat type. In addition, populations of all species peaked at the same time; 2 species peaked twice and 3 species peaked once each season. All species were present in the first population peak, which occurred in late May to early June. The two most numerous species then peaked again in late July and early August. Separate population peaks would indicate that therevids were reducing competition by developing at different times. The data also suggest that some species may produce 2 generations per year. Until now, therevids have all been thought to produce a single generation per year. Perhaps these therevids overwinter as last instar larvae and emerge as adults in May. The eggs resulting from this generation would hatch and the larvae would develop rapidly during June and July when insect larvae are most abundant in the sandy substrate and when temperatures are warm.

The researchers found no apparent differences between the therevid populations in the natural and managed forests. It is unclear from this finding whether differences in species composition and population density occur in the root-feeding insect populations of the two forest types. Because therevids are generalist predators, they can survive on a variety of food resources.

We found that therevids have a dramatically skewed sex ratio. Approximately 85 percent of the therevids caught were females. This ratio was consistent for many of the more abundantly collected species. The males were caught slightly earlier than the females, indicating that they emerge earlier and mate when the females emerge. In this way they ensure that females will be mated, even though there are many more females than males.

There is much to learn about the bioecology of these flies. Their important roles in ecosystem dynamics and their potential as biological control agents strongly suggest that further investigations are warranted. The authors wish to thank John Taft and Rick Phillippe at the Illinois Natural History Survey’s Center for Biodiversity; the rangers at Sand Ridge State Forest, Dan Riggs and John Meredith; and the Illinois Department of Conservation for their assistance in this study. For more information about therevids, check out the Therevidae homepages on the Web.