Feature Article

Host Specificity of Insect Pathogens:
Evaluation for Biological Control Programs

Predators, parasites and pathogens are important components in the natural control of insect populations. These natural enemies cycle with their host populations, preventing explosive population growth and dampening outbreak cycles. Most insect species do not become pests because various environmental and biological factors, including natural enemies, act on the populations to keep them below economic levels. Nevertheless, some insects may undergo outbreaks and cause economic damage of crops. The use of natural enemies to prevent pest populations from increasing to an economic damage threshold has several advantages over the use of chemical insecticides, including the possibility of long-term or even permanent control and high host specificity, which means that they pose no significant risk to other organisms.

Pathogens are important natural enemies of insects, and most of the thousands of species of bacteria, viruses, fungi, nematodes, and protozoans that infect insects are highly host specific and are effective in reducing population levels of their hosts. Host specificity is an especially important consideration for pathogens that are manipulated for biological control purposes. Before a pathogen is introduced into a new region for biological control of a pest species, it is imperative to know the potential host range and possible effects on other species of insects.

Susceptibility of a particular insect species to a pathogen defines the host range of that pathogen. The number of host species that can be infected by a single pathogen species varies considerably. Pathogens that are very host specific have a narrow host range and infect only one or a few closely related host species. Those with a wide host range are considered to be generalists and, therefore, are not very host specific. Beauvaria bassiana has been considered to be in the latter category; however, even this well known fungal pathogen of several crop pests may actually exist as a species or subspecies complex, members of which are identical in form to other members of the complex but are individually fairly specific to a particular host or small group of host species.

The issue of the host specificity of insect pathogens has been raised by researchers, environmentalists and regulators more often in recent years as pathogens have gained importance in the biological control arena. There are several reasons for concern. Generalist pathogens, introduced for biological control of a pest species, could theoretically become epidemic in nontarget species of insects, if not in other animal groups. Parasites and predators of a pest species could be affected, either directly by the pathogen or indirectly as the food source (the target pest) declines in numbers. Indeed, laboratory studies conducted in the past have shown that, although infections rarely were initiated in organisms outside the arthropod group, many insect pathogens have broad physiological host ranges when force-fed or injected into nontarget insects under optimal conditions. For example, one microsporidian (Protozoa), Vairimorpha lymantriae, is a pathogen of European populations of the gypsy moth but is not found in North American populations. When tested in the laboratory as a possible biological control agent in the U.S., infections occurred in many of the 35 nontarget moth species to which it was fed but closer examination of these infections revealed some unexpected results. Most of the infections appeared to be different from infections seen in the gypsy moth. In some instances, the pathogen killed the host insect, even at very low dosages. While this seems to support the assumption that this is a generalist species, in actuality it is probably an ecological "dead end". Death always occurred before the pathogen could produce the infectious forms necessary to infect another host.

When native moth species were exposed to V. lymantriae and to four other gypsy moth microsporidia, other responses included: a very low percentage of a nontarget host species becoming infected, the production of abnormal forms of the pathogen, strong short-term immune responses in the host tissues, and low levels of infection in which very few infectious propagules were formed. Further investigations showed that when infected natural host insects were placed with uninfected natural hosts in small cages that maximized contact between larvae, most of the 11 microsporidian species tested were transmitted to the uninfected hosts, probably via the feces or silk. In a few cases, the natural host could also transmit a microsporidian to a nontarget host. However, when nontarget hosts infected in the laboratory were placed with uninfected nontarget hosts of the same species, no transmission occurred. In other words, most of the microsporidia tested could not be transmitted from one nontarget host to another nontarget host of the same species, even in the most favorable conditions where contact was unavoidable. Given the ecological complexity of field situations, transmission of most microsporidian species within a nontarget population would be highly unlikely to occur.

Unusual responses by nontarget hosts to a pathogen have also been reported in fungi and appear to be similar to those seen in microsporidian infections. Although it is more difficult to detect unusual responses in viral infections, it is acknowledged that sublethal or atypical infections may occur in nontarget hosts. Unlike the other major pathogen groups, most bacteria that infect insects, including naturally occurring Bacillus thuringiensis or Bt, are usually opportunistic pathogens, infecting only hosts that are stressed or injured. (A notable exception is Bacillus popillae, a pathogen of the Japanese beetle.) Although most appear to have a broad host range, recent studies of Bt have shown that a complex of subspecies exist, each of which have different, though often overlapping, host ranges.

The use of insect pathogens as biological control agents has many advantages, among which are safety to humans and other vertebrate animals and the ability to propagate in the pest population so that multiple, innundative releases are not usually necessary (Bt does not fit into this category; it rarely cycles in insect populations and is used as a microbial insecticide because of its virulent toxin). In some situations, it is desirable to have absolute specificity to the natural host. In most cases, though, a narrow but not strict host range is desirable. The control agent should target more than one pest (usually 2 or 3 closely related insects) to warrant the cost of production, but should not harm nontarget organisms or other natural enemies of the target pests.

As interest is renewed in the use of pathogens as biological control agents of pest insects in field crops, orchards, urban areas, and forests, host specificity testing will be a major factor in determining the safety and efficacy of pathogens. Although laboratory data is useful, and is often the only method available for estimating the ecological host range of pathogens prior to release of exotic organisms, care must be taken in the interpretation of these data. New laboratory tests and methods of evaluation will help to determine more accurately the ecological host specificity of pathogens being considered for field release in biological control programs.

- Lee F. Solter and Joe V. Maddox, Illinois Natural History Survey, Champaign, Illinois.

Return to Contents Menu Vol. II  No. 9

Go To Index