Ricci, M., Bio Integrated Technology srl, Pantalla, Perugia, I-06050, Italy
Glazer, I., Department of Nematology, ARO, Volcani Center, PO Box 6, Bet Dagan 50250, Israel
Campbell, J.F., Department of Entomology, Rutgers University, New Brunswick, NJ 08903, United States, Department of Nematology, University of California, Davis, CA 96516, United States
Gaugler, R., Department of Entomology, Rutgers University, New Brunswick, NJ 08903, United States

Five bioassays were compared for their usefulness to determine the virulence of four nematode strains. The objective of this study was to develop standard assays for particular nematode species. In all assays, the nematodes Steinernema feltiae (strain UK), S. riobravis, S. scapterisci Argentina and Heterorhabditis bacteriophora HP88 were exposed to Galleria mellonella larvae. All bioassays except the sand column assay were conducted in multi-well plastic dishes. In the penetratoin rate assay, the number of individual nematodes invading the insect was determined after a 48-h exposure to 200 infective juveniles (IJs). In the one-on-one assay, each larva was exposed to an individual nematode for 72 h before insect mortality was recorded. In the exposure time assay, insect mortality was recorded after exposure to 200 IJs for variable time periods. The dose-response assay involved exposing larvae to different nematode concentrations over the range 1-200 IJs/insect and recording mortality every 24 h for a 96-h period. In the sand columns assay, insects were placed in the bottom of a plastic cylinder filled with sand. Nematodes were applied on top of the sand and insect mortality was determined after IJs had migrated through the cylinder. The highest mortality level in the sand column assay was obtained with IJs of s. feltiae followed by H. bacteriophora; treatments with S. riobravis and S. scapterisci produced lowe levels of insect mortality. In the other four assays, S riobravis was the most virulent, followed by S. feltiae, H. bacteriophora and S. scapterisci. In the exposure time assay, rapid mortality was achieved when the insects were exposed to S. feltiae and S. riobravis. For these nematode species, a gradual increase in the number of individuals which penetrated into cadavers was recorded. Conversely, the number of nematodes in the cadavers of insects infected by H. bacteriophora and S. scapterisci remained low during the entire exposure period. In this assay, exposing the insects to these nematodes resulted in a gradual increase in mortality. In the dose-response assay, complete separation among nematode species was obtained only after 48 h of incubation at a concentration of 15 IJs/insect. LD50 and LD90 values were calculated from dose-response assay data. However, these values did not indicate differences among the different nematode species. The present study demonstrated the variation in entomopathogenic nematode performance in different bioassays and supports the notion that one common bioassay cannot be used as a universal measure of virulence for all species and strains because nematodes differ in their behaviour. Furthermore, particular assays should be used for different purposes. To select a specific population for use against a particular insect, assays that are more laborious but which simulate natural environmental conditions (e.g. the sand column assay) or invasion by the nematode (e.g. the penetration rate assay) should be considered. In cases were commercial production batches of the same nematode strains are compared, simple and fast assays are needed (e.g. the one-on-one and exposure time assays). Further studies are needed to determine the relationships between data obtained in each assay and nematode efficacy in the field.