used to detect possible recombinants that arose from strain crossings. Little has been done with the isozymes of microsporidia. Except for a brief description of the isozymes of several species by Joslyn et al. ( 1979), this tool has not been exploited to analyze the variations that might exist between strains or geographical isolates.
We have identified several areas of research that need to be addressed before microsporidia can be genetically manipulated and reliably used in the biological control of European corn borer.
Microsporidian transmission in European corn borer is accomplished through both horizontal and vertical infection. Vertical transmission of N. pyrausta from overwintering infected corn borers to their progeny plays a major role in the infection of first-generation larvae, but is less important in transmitting N. pyrausta to the second generation. Horizontal infection occurs mainly during the second generation when larval population densities are high and feeding chambers are inhabited by both infected and uninfected larvae. Laboratory studies using both colonized and feral O. nubilalis infected with N. pyrausta indicate that mortality is greatest in transovarially infected larvae subjected to environmental stress, such as overwintering. Whether the epizootiological situation is similar in N. furnacalis needs to be determined, but it is certain to differ from a pathogen that is transmitted purely horizontally, as is V. necatrix. The interaction between introduced microsporidia and other biological control agents, such as parasitoids, has been examined only to a limited extent, in part because of the difficulties in identifying microsporidia species unequivocally. It is important that natural pathogen populations not be displaced by the artificial introduction of other biocontrol agents. This means that exotic species should be scrutinized not only with respect to their potential host range, but also for their impact on indigenous pathogens. The epizootiological models related to microsporidia need additional refinement and will serve to guide continued research in this area. As Onstad and Carruthers ( 1990) have pointed out, "empirical research provides underlying data for the development and validation of epizootio- logical models. Models are useful in guiding empirical research." This relationship must not be reversed.
A major problem is to obtain a definitive identification of the microsporidia being employed for biological control and to avoid working with a mixed infection. Several species may actually be present in what appears to be an epizootic or infection caused by one species. Most of the original species descriptions were made using light microscopy and ultrastructural or molecular details are often not available. The species descriptions have to be expanded to include molecular ( Vossbrinck et al. 1993) or serological markers. Because the necessary equipment and reagents may not be available to all researchers active in microsporidiology, the establishment of national reference centers, able to provide identification or certified strains, would be a great advantage.