The potential impact of generalist predators on armyworm mortality was evaluated in the field, through a predator removal study, and in the laboratory, through controlled feeding trials. The most common potential armyworm predators removed from the field included ground beetles (Carabidae), rove beetles (Staphylinidae), ants (Formicidae), and spiders (Araneae). Armyworm damage to corn plants was significantly greater where generalist predator populations were reduced, through the use of pitfall traps and exclusion arenas, than in the control where predator populations were unaltered. The differences in the proportion of damaged plants and the degree of damage between the predator removal treatment and the control were statistically significant. Generalist predator consumption rates of live armyworm larvae in the laboratory were variable, however most predators did feed on the larvae. Large carabid beetles, including Pterostichus chalcites Say, Pterostichus lucublandus Say, and Scarites subterraneus F., exhibited the highest consumption rates.

Generalist predators were sampled in four reduced-tillage corn systems which differed in the degree of soil disturbance and quantity and structure of the surface mulch due to tillage and cover crop management practices. The two sampling methods which were used, pitfall trapping and vacuum sampling, showed similar trends in predator abundance. The treatment with the highest degree of mulch ground cover had the highest overall predator abundance while the treatment which was disked and had no surface mulch had the lowest. Although several species tended to prefer the system with the least amount of ground cover, most of the common species preferred the treatment with the most groundcover. Pitfall trap catches over a three-day period indicated that predator activity was significantly higher during the day than night in all treatments. However, a laboratory study provided evidence that night activity in the field may have been reduced due to unusually low temperatures. Catch data from pitfall traps, unbaited and baited with live armyworm larvae, indicated that long-distance chemical detection is not an important cue for generalist predators in finding armyworm as prey.

A removal sampling technique was used and evaluated for estimating the absolute densities of ground beetles (Carabidae). Removal sampling is a method of absolute density estimation based on the decline in successive catch numbers as individuals are removed from a population. Field arenas were used to isolate sampling areas in a no-till corn field and barrier pitfall traps were installed within the arenas to remove the carabids. Three three-week sampling trials were conducted in 1991 providing data for estimates on 5 June, 26 June, and 17 July. A single six-week sampling period was conducted in 1992 providing data for estimates on 2 June based on three, four, five, and six weeks of sampling. A linear regression method was used for calculating the estimate from the removal sampling data. The technique's practicality and agreement with the assumptions of removal sampling are discussed.