Evaluating Biological Controls

In the very first agricultural entomology course that I took as an undergraduate student (was it really a hundred years ago?), the professor stated that the most important characteristic of any approach to pest management was that it had to work. It made sense a hundred years ago and it makes sense today. We can extol the virtues of biological control all we want on the basis of it being a natural and safe practice, or that there are thousands of technical research articles about it, but people won't adopt it unless it is effective. That longago professor listed some other factors affecting the adoption of pest control practicesnamely, they had to be cost-effective and they had to be relatively easy to use. Finally, if a control practice met all the above criteria, it would be helpful if it didn't have negative impacts on the agroecosystem (i.e., the crop) or the larger environment.

The Cooperative Extension Service has a long and noble history regarding the evaluation of pest control practices, at least, to a point. And these countless research and demonstration trials over the last several decades have done a lot to help pest managers make decisions about pest control options. But some types of pest management evaluation is easier than others, and some types have been done more often than others. One thing that we have done really well is evaluate the efficacy of pesticides. To simplify "spray and count" work, you count bugs or weeds or disease lesions in your plots (hopefully, replicated plots). Then in one set of plots you apply the pesticide, then go back and count the bugs or weeds or disease lesions in all the plots. It's pretty easy to say "We killed all the bugs" or "99% fewer weeds germinated in the herbicide plot" or "We got a 22% increase in yield." We've also been pretty good at evaluating the impacts of certain other types of pest control practices, such as a tractor-drawn cultivator or a new disease-resistant crop variety.

But how many Extension workshops or field days have you participated in where the actual impact of a biological control was evaluated in a quantitative sense? I suspect, in most states, relatively few. There are some reasons to explain this situation. Much biological control is provided by nature, and it's hard to keep it out of some plots (or fields) while letting it be effective in others. Many biological control organisms are very mobile, and small plots just don't give good data, and the small-business based biological control industry is virtually without the resources to fund the needed large-scale research plots. And there aren't very many state and federal research $$ to fund this type of research or demonstration work.

So, how can we evaluate the effectiveness of biological controls? As there are many types of natural enemies, and many ways of using them, there must also be a variety of ways to evaluate the job that they are doing. Of course, different people have differing needs as to evaluation. For example, extension personnel may wish to conduct applied research or demonstration programs, in which case plots should be carefully designed and properly replicated. A grower, on the other hand, simply wants to know whether or not a particular biological control is effectively reducing pest damage; replicated plots and statistical design are usually not a consideration (although they certainly can be). Hopefully, the comments below will address both types of needs, but please consider the following as general guidelines; you will have to develop specific evaluation protocols based upon the pest, the natural enemy and how it is used, the crop, and the environmental factors that influence everything else.

General considerations for evaluation of natural enemies. The proper evaluation of natural enemies necessitates an understanding of the life cycles of the target pest and the natural enemy, including the ability to recognize various stages in the life cycle. For some natural enemies, this is easier than for others. For example, if you are releasing lady beetles, do you know how to distinguish the species you are releasing from naturally-occuring species in your area (there are lots of types of orange and black lady beetles!). Do you know how to recognize egg, larval, and pupal stages? If you are releasing whitefly or aphid parasites, do you know what the parasitized whiteflies or aphid mummies look like? If evaluating the biological control of a leafminer, can you distinguish the pest from its natural enemies? These types of issues are also important in dealing with phytophagous insects used in weed biological control; can you identify egg, larval, pupal and adult stages, as well as recognize, with a fair degree of certainty, the damage symptoms attributed to the weed control agent?

Microbial pesticides. Generally, the evaluation of microbial pesticides is similar to the evaluation of conventional pesticides, especially for insect and weed control. But because many microbial pesticides are effective from the action of living microorganisms, care must be taken to assure that the pathogens (or insect parasitic nematodes) are fully viable at the time of application. In research work, extension people should assure that they are using fresh product that has been stored under acceptable conditions. Some products, notably some species of insect parasitic nematode, have a relatively short shelf life. For many products it is difficult to evaluate the viability of the organisms and there has to be some reliance that the manufacturer is supplying a viable product. Keep some of the material properly stored in case there is a perceived failure, at which time the company will be able to assess the quality of the product. The viability of insect-parasitic nematodes can be checked rather easily after they are mixed with water simply by viewing them with a hand lens (they should wiggle). Remember that microbial pesticides often take longer to show effectiveness than conventional pesticides. Insecticides based on Bacillus thuringiensis (Bt) cause gut paralysis and cessation of feeding within a fairly short period after ingestion. Although damage is stopped at this point, actual death may not occur until up to 2-3 days after application. Diseased individuals begin to discolor after 12-36 hours. Therefore, it is best to conduct the post-application evaluation 2-3 days after application. In many cases, insect pathogens affect the cuticle, which breaks down rapidly under environmental conditions; therefore evaluation should occur in a timely fashion after death is expected. In some cases fungal pathogens, such as Beauveria bassiana, may be a bit difficult to evaluate. If atmospheric moisture is sufficiently high, noticeable sporulation from the cadaver may be seen, but this is not always the case. With very small insects such as aphids and leafhoppers, the fungal mycelium may so distort the cadaver that it is difficult to even identify the remains as the target insect. In cases of less mobile targets, such as aphids, it is helpful to flag or otherwise mark infested leaves before the application, and count individual colonies, to determine the ultimate outcome. Mycoherbicides need to be evaluated in the context of the normal symptoms that would be expected from the pathogen, such as lesions or fruiting bodies. Total weed biomass should also be compared between treated weeds and check plots, as well as reproductive success of the weeds (i.e., production of viable seeds).

Augmentative biological control of insects and mites. Certain augmentative biological control programs have been pretty carefully evaluated for effectiveness. Examples include various pests of greenhouse crops, natural enemies (mostly parasitoids) of filth flies in large-animal and poultry establishments, and natural enemies of stored grain pests. However, the effects of augmentation biological control may vary from location to location due to the nature of pest/natural enemy relationships and differences in environmental conditions. Variability is less likely under the relatively constant conditions of greenhouses and grain storage facilities as compared with field, forest, or garden situations. As with microbial pesticides, an important consideration is the quality of the natural enemies to be released. Part of the evaluation process should be an assessment of quality when the shipment is received and as it is prepared for release.

Evaluating the impact of released natural enemies can be a bit of a challenge because they are fairly mobile critters. Adult, flying insects have different capacities to disperse from release sites based upon their size, flight ability, and normal searching behavior. In designing true research (including demonstration research) plots, these characteristics must be kept in mind for the organisms being evaluated. Plots must be large enough so that natural dispersal doesn't dilute the recommended quantity released. Also, plots must be far enough apart so that dispersing natural enemies do not have a major impact in the comparison control plots (where the natural enemies are not released). Unwinged predators (such as lacewing larvae or predatory mites) are generally less dispersive, but even these can move relatively great distances while searching for prey.

Evaluation of released natural enemies for actual control, such as on a farm or in a home garden, tends to be more subjective than quantitative, because usually we don't set aside untreated "control" areas where the natural enemies are not released (although this is certainly possible in farming operations, and is recommended, especially when first getting accustomed to a new biological control program). As mentioned above, it is very useful to be able to recognize all stages of the natural enemy. If releasing a native natural enemy (such as convergent lady beetles or green lacewings) it is helpful to take a count of the naturally-existing population before making releases. This can be done on a unit-area basis or a timed-sample. For example, if you are releasing lacewings in vegetables, walk a certain length of row (maybe 10 ft) and count the number of lacewing larvae and adults found. Repeat this in three or four locations. While counting the lacewings, also take a census of the target pest(s). Keep track of these numbers. If releasing lacewings eggs, allow a few days for the eggs to hatch and the larvae to get big enough to be visible, and then take a similar census as before. Do you see an appreciable increase in lacewing larvae? Has there been a reduction in the pest population? A second follow-up sample a few days later will help confirm the results of your first sample; as the lacewing larvae grow, the pest consumption rate should increase and the pest population level should decrease.

Predators are free-living, even as larvae, and generally fairly easy to see (although a hand lens will be helpful with small species such as minute pirate bugs and predatory mites). Many parasitoids, on the other hand, spend their larval stage inside their host insect (the pest). The impact of these is more difficult to evaluate. One way is to collect some of the target pests a few days after the parasites have been released, and raise them in a cage or container until they complete their life cycle; if parasitized, they should eventually be killed (for some parasites this may not occur until the end of the host's life cycle) and parasite pupae and/or adults should be found in the cage. If sufficient specimens are collected, and set up properly, you can gain data on percent parasitism, which is one indicator of success. In some types of situations, if you know what you are looking for, you don't have to rear the pests to determine degree of success. For example, a few years ago one of my graduate students, Paul Whitaker, and I were conducting a biological control field day for fresh market vegetable growers. Our host was an organic farmer who had a small population of imported cabbageworm on his crucifers. Paul started picking off larvae one at a time and pulling them apart with his fingers to reveal the multiple larvae of the gregarious parasite Cotesia within. At first, people were a bit dismayed, but soon everyone was picking off caterpillars and pulling them apart. Each reported how many were so examined, and how many were parasitized, and, as a group, in a few minutes we were able to determine that this population was very heavily parasitized.

In conclusion I must reiterate that, for biological control to be used, it has to be demonstrated to be effective. With the exception of the use of some microbial pesticides or other commercially available natural enemies, biological control is often a bit more difficult to evaluate than practices that have an immediate response, such as the use of a broad spectrum insecticide, or tillage for weed control. By its very nature, biological control should be a more natural, and therefore long term process. The ideal biological control is one that continues on, successfully, generation after generation, without us even having to think about it. But pest populations vary through time, for lots of reasons. It is simply not good enough to conclude that biological control works because you see few bugs, or, conversely, that biological control doesn't work because there is lots of damage of some sort. Therefore, when we first embark on a new biological control program, it is essential that we be able to evaluate the actual impact of the natural enemies. Even after we become comfortable with a successful program, it is essential to continue monitoring the populations of pests and their natural enemies. Because of biological and environmental differences from year to year, crop to crop, and location to location, the specifics of evaluation have to be developed for each specific program. Common to all, however, is the need to be able to recognize the stages of natural enemies and the signs of impact on the pests and their damage. Finally, if your first attempt at biological control doesn't work "as advertised" (whether it be natural biological control or purchased natural enemies), please don't give up. Just as there are nuances to using a pesticide (water chemistry, nozzle spacing, tank pressure, ground speed, etc.) there are also complicating factors with biological control. Consult with others who have experience or expertise to determine how best to improve your results.

- Dan Mahr, University of Wisconsin - Madison

Return to Contents Menu Vol. VI  No. 12

  Go To Index