Factors in the success and failure of microbial insecticides in vegetable crops

A search for patterns in the success and failure of microbial insecticides in vegetable crops was conducted through review of four case studies: the use of Bacillus thuringiensis (B.t.) var. tenebrionis for control of the Colorado potato beetle, the use of B.t. var. kurstaki for control of the diamondback moth, the use of various B.t.s for control of lepidopterous pests in tomatoes and celery, and the use of a granulosis virus for control of potato tuber moth. With success defined in terms of achievement of technical goals (efficacy), commercial goals (end-user and insecticide manufacturer satisfaction) and social, or public goals (environmental and health safety), only certain of the case studies could be judged a success. These successes shared a variety of features including: (1) use of the microbial insecticide as a component, rather than as the sole agent, in an integrated crop management program; (2) unavailability of conventional insecticides, due to insecticide resistance, lack of registered products or mandatory IPM programs, provided incentive for the use of microbial insecticides; (3) modification of the expectation that microbial insecticides will perform within the chemical paradigm – fast, lethal and on contact; (4) exploitation of all possible benefits of the microbial insecticide, including safety to natural enemies, as well as efficacy against the target insect, and (5) support from large private and public institutions in the form of research, grower education, scouting programs, subsidized production, and economic and legal incentives to the use of microbial insecticides. Introduction: defining success and failure There is no single, robust definition of success (or failure) that meets the varying criteria of the many groups that have a stake in the use of microbial insecticides. The example of products based on Bacillus thuringiensis var. tenebrionis (B.t.t.) first marketed for control of the Colorado potato beetle (Leptinotarsa decemlineata [Say] Coleoptera:Chrysomelidae) in 1988, is a case in point. From the researcher’s purely technical perspective, these products (which included Foil [Ecogen], M-One and M-Trak [Mycogen], and Novodor [Abbott],), when used properly, maintained insect populations consistently below the economic threshold level and controlled insects that were resistant to conventional insecticides; they were therefore judged to be a technical success (Zehnder & Gelernter 1989). Likewise, organic potato growers, who had in the past suffered from the lack of organicapproved Colorado potato beetle (CPB) products, welcomed these products with open arms. However, from the commercial perspective of the majority of (nonorganic) potato growers, B.t.t. products were judged to be of marginal value, especially once more effective and easier-to-use products such as the chloronicotinyl insecticide, imidacloprid, was introduced to the marketplace. Furthermore, from the insecticide manufacturer’s standpoint, B.t.t. products were never deemed a success, due to low consumer demand and low profitability; sale of B.t.t. based products was discontinued 302 W.D. Gelernter & J.T. Trumble Table 1. Microbial insecticides that satisfy technical (efficacy) goals, commercial goals (for end-users and microbial insecticide manufacturers) or the general public are indicated with a check ( √ ) while those that do not are indicated with an ‘X’ Technical End-user Insecticide manufacturer General public B.t.t. for Colorado potato beetle √ X X X B.t.k. for diamondback moth (1980–1990) √ √ √ X B.t.k. for diamondback moth (1990–present) X X X X B.t. in tomato IPM programs √ √ √ √ GV for potato tuber moth √ √ Uncertain To be determined by 1998 by almost all companies as a result. Finally, from the viewpoint of the general public, the benefits of these products have been negligible, since alternatives to the B.t.t. products such as imidacloprid are also relatively safe for humans and the environment. In attempting to describe the potential for success or failure of microbial insecticides, we therefore face a problem. If we focus on the criteria for just one of the four groups described above, we are bound to develop a skewed vision of which microbes are most likely to succeed and which are most likely to fail. Yet if we require agreement among all of the concerned groups, our standards become unreasonably high, and we almost certainly condemn all microbial insecticides – and probably most insect control products, for that matter – to the dust heap as failures. This leaves us somewhere in between these extremes for a reasonable working definition of success. In this review, we will, somewhat arbitrarily, consider a microbial insecticide successful if it has met the criteria of at least two of the four groups described above: (1) the technical community; (2) end-users; (3) microbial insecticide manufacturers; and (4) the general public. In the example above, since the B.t.t. products only met the criteria of the technical community and of a minority of (primarily organic) growers, it would not be judged a success (Table 1). Promoting success by limiting alternatives: B. t. var. kurstaki and the diamondback moth As in the case of the Colorado potato beetle and B.t.t., growers of cole crops first became interested in microbial insecticides only when they were desperate – when all other control options had failed due to widespread development of diamondback moth (Plutella xylostella [Linnaeus] Lepidoptera:Plutellidae) resistance to chemical insecticides. Beginning in the early 1980s, the use of Figure 1. Potency, based on bioassay-generated LC50 values, of Bacillus thuringiensis delta endotoxins against larval insect species including the yellow fever mosquito (Aedes), the diamondback moth (DBM), the cupreous chafer (Anomala), the European corn borer (ECB: Ostrinia nubilalis [Hübner] Lepidoptera:Pyralidae), the Colorado potato beetle (CPB), the tobacco budworm (TBW: Helicoverpa virescens [Fabricius] Lepidoptera:Noctuidae), the beet armyworm (BAW) and the spruce budworm (SBW: Choristoneura fumiferana [Clemens] Lepidopetera:Tortricidae). Delta endotoxins assayed include protein mixtures isolated from B.t. var. israelensis (Bti), from B.t. var. kurstaki (strain HD-1) and from B.t. var. japonensis, strain BuiBui, as well as pure toxins such as CryIAc, CryIAb, CryIIIA. Potency values are averages of values found in the scientific literature from 1989–1995. products based on B.t. var. kurstaki (B.t.k.) such as Dipel (Abbott), Javelin (Novartis), MVP (Mycogen), and Condor (Ecogen) began to dramatically increase on cole crops. Unlike B.t.t., however, efficacy of B.t.k. products was regarded by growers as good to excellent, and relatively easy to use. The extremely high toxicity of B.t.k. delta endotoxins for diamondback moth (DBM) larvae, almost 100 times that of the toxicity of B.t.t. for CPB larvae (Figure 1) is the primary reason for this difference in end-user experiences. As a result, there was a short, happy window of time where it indeed seemed as though microbial insecticides had finally found a market where they consistently performed as well as, or even better than, chemical insecticides. But the honeymoon was short-lived. By 1990, the use of up to 50 applications per year of B.t.k. products for control of diamondback moth larvae led to the Factors in the success and failure of microbial insecticides in vegetable crops 303 first case in history of field resistance to a B.t. product (Tabashnik et al. 1990). Field resistance to B.t.k. has since been confirmed in Hawaii, Florida, Japan (greenhouses only), the Philippines, Honduras, Costa Rica and Guatemala (Perez & Shelton 1997). To add insult to injury, we are now beginning to see the first incidence of diamondback moth resistance to B.t. var. aizawai, a strain which was successfully (though briefly) used to control DBM populations that were resistant to B.t.k. products (Liu et al. 1996). In this example, we can therefore come to two very different conclusions about the success of microbial insecticides for DBM control, depending on the point in history that we choose to make our analysis. During the 1980s and early 1990s, the use of B.t.k. for control of DBM could have been judged as a success by almost any standards. The product was highly efficacious, endusers liked it, and private companies saw real growth in sales of B.t.k. products. However, by the mid-1990s, DBM resistance to B.t.k. in many key cole crop growing areas of the world made this product of minimal use to growers, and therefore to microbial insecticide manufacturers – a failure by almost any standard. Comparison of the cases of B.t.t. for the CPB and B.t.k. for the DBM leaves the distressing message that you just cannot win. Either the products are not effective enough, leading to end-user dissatisfaction and private company withdrawal from the marketplace (as in the case of B.t.t.), or the products are so efficacious that they are overused, causing development of resistance (as in the case of B.t.k. and the DBM). Is it possible that this ‘no-win’ situation is the outcome of the unreasonable expectation that microbial insecticides can perform like chemical insecticides? Can our successes be longer-lived, more sustainable, if we change those expectations? Promoting sustainable success: B.t. var. kurstaki in integrated tomato and celery programs Avoiding the chemical paradigm In most cases, the efficacy of microbial insecticides is inferior to the chemical insecticides that they might replace due to lack of acute toxicity, lack of residual activity, specificity, and lack of contact activity (in the case of B.t. and baculoviruses). By focusing exclusively on microbial insecticides as ‘silver bullets’ that will solve specific pest control problems, and by looking at only one pest at a time (as in the CPB and DBM examples above), are we overlooking some of the less tangible, longer-term benefits of the use of microbials? The widespread adoption of tomato and celery pest management programs that rely on B.t.k. products as much for their lack of disruption of natural enemy complexes, as for their ability to control lepidopterous pests, would suggest that this is the case. Tomato IPM in Mexico, California and Ohio Prior to 1986, lepidopterous pests on tomatoes (tomato pinworm, Keiferia lycopersicella [Walsingham] Lepidoptera:Gelechiidae; tomato fruitworm, Helicoverpa zea [Boddie] Lepidoptera:Noctuidae; beet armyworm, Spodoptera exigua [Hübner]; yellow striped armyworm (Spodoptera ornithogalli [Guenée] Lepidoptera:Noctuidae) were controlled with up to 40 applications per crop of chemical insecticides such as methomyl and permethrin. In addition to the expense and environmental problems caused by these applications, destruction of beneficial insect complexes resulted in upset infestations of leafminers (Liriomyza spp.) and whiteflies (Bemisia spp.), which in turn required additional insecticide applications for control (Bolkan & Reinert 1994). To address these problems, an integrated program, which relies on various combinations of the components listed below, has been implemented in various forms on large hectarages of fresh and processing tomatoes in Mexico, California and Ohio (Bolkan & Reinert 1994, Trumble et al. 1994): • pheromones for tomato pinworm mating disruption; • B.t.k., B.t. var. aizawai or abamectin (Avid, produced by Novartis) based products for fruitworm and armyworm control; release of the parasitoid wasp Trichogramma pretiosum for further fruitworm control; • insecticidal soaps for whitefly control; • development of economic thresholds for all pests; • weekly scouting programs. Because it was possible to see significant reductions, or the complete elimination of hard chemicals from tomato pest management programs, leafminer and whitefly populations were reduced via the resurgence of natural regulation by parasitoids and predators. In addition, replacement of methomyl sprays with B.t.k. applications for control of soybean loopers (Pseudoplusia includens) in nearby soybean fields also contributed to the build up of natural enemy complexes for control of whiteflies. Economic analyses indicate that growers on the IPM programs were usually more 304 W.D. Gelernter & J.T. Trumble profitable ($64–$1000 per hectare) than growers using conventional insecticides, due to reduced frequency of pesticide applications and/or higher yields in the IPM fields. (Bolkan & Reinert 1994, Trumble et al. 1994). Celery IPM in California In celery, an IPM program targeted two key pests, the beet armyworm and Liriomyza leafminers, and produced similar results. The substitution of B.t. var. aizawai (Xentari, produced by Abbott Labs) for methomyl applications for control of beet armyworm larvae allowed natural enemies of leafminers to survive, resulting in a reduced need for insecticide applications targeted against leafminers. When necessary, applications of abamectin (Avid, produced by Novartis), which has minimal activity against parasitoids and predators, were made against leafminers. As for tomatoes, the celery IPM program has demonstrated increased profitability (up to $410 per hectare in grower validation trials) when compared to the higher input standard chemical program (Trumble et al. 1997).