Temperature is a major limiting factor in the success of entomopathogenic nematodes as biological pest control agents in the greenhouses. We compared the effectiveness of the cold-adapted Steinernema feltiae and Heterorhabditis marelatus with an intermediate-adapted S. carpocapsae and warm-adapted H. bacteriophora, H. indica and H. zealandica strains at 20/28 +/- 1 degrees C (night/day) against the fungus gnat, Bradysia difformis, infesting poinsettia (Euphorbia pulcherrima) in the greenhouse. Nematodes were applied either at the recommended rate of 1.25 x 10(5) infective juveniles/m(2) in a single application or 2.5 x 10(5) infective juveniles/m(2) in four split applications. The efficacy of nematodes against fungus gnats differed with potting medium (Nursery-mix or Pro-mix) and species. In the Nursery-mix, a single application of H. bacteriophora significantly reduced fungus gnat emergence relative to the control throughout the experimental period of 63 days but S. feltiae reduced fungus gnats only up to 7 days after treatment (DAT). H. indica significantly reduced fungus gnat emergence compared to the control at 3, 7, 28, 49 and 63 DAT. In the Pro-mix, a single application of H. indica significantly reduced fungus gnat emergence compared to the control from 14 up to 63 DAT whereas H. marelatus and H. zealandica were effective against fungus gnats only up to 3 DAT. In the growth chamber at 20 +/- 1 degrees C (nighttime) and 28 +/- 1 degrees C (daytime), a single application of H. indica also showed a similar trend against fungus gnat population throughout the experimental period but the differences were statistically insignificant. When applied in four split applications (each containing 0.625 x 10(9) infective juveniles/ha) in the Pro-mix, only H. bacteriophora and H. indica significantly reduced fungus gnat emergence compared to S. feltiae and the control. We conclude that only H. bacteriophora (52%) and H. indica (54%) performed better than S. feltiae (38%) against fungus gnats at temperatures above 25 T and the split application produced no advantage over a single application. (C) 2007 Elsevier Inc. All rights reserved.
