Use of Bacillus thuringiensis in Pest Management of the Tomato Ecosystem in Trinidad Mona T. Jones Entomologist Ministry of Agriculture, Lands and Food Production Central Experiment Station, Centeno Arima, Trinidad Major pests of tomatoes in Trinidad include: the pin worm, Keiferia lycopersicella; the army worm, Spodoptera eridania; the fruit worm, Heliothis virescens; leaf miners, Liromyza spp.; fruit piercing bugs, Nezera viridula, and Phthia picta; the mole cricket, Scapteriscus vicinus. Commercial producers have resorted to excessive application of chemical pesticides with consequent increase in leaf miner infestations, serious outbreaks of mites and developing resistances in mole cricket populations. Too frequent use of pesticides is also a deterrent to implementation of classical biological control. Utilization of the relatively specific bacterial insecticide, Bacillus thuringien- sis, seems encouraging as a pest control strategy for the tomato ecosystem. Despite the tact that tomato is currently the most widely cultivated vegetable crop in Trinidad, large quantities of the pro- cessed fruit are still being imported into the country. The de- mand for both fresh and processed fruit is quite encouraging, but the farming community is faced with the problems of irrigation, labour, storage, transport and marketing as well as pest and disease outbreaks. All of these have been deterrents to the establishment of a viable tomato industry. The market is geared for production of fresh table varieties so that an attractive looking fruit with high consumer appeal is always desirable. Thus, fruit which has been blemished by pest attack is unmarketable. This paper presents an overview of the problem of pest infestation during 1981 to 1984 and suggests possible strategies for in- tegrated management utilizing Bacillus thuringiensis Berliner. Pest Composition and Degree of Damage in the Tomato Ecosystem In Trinidad, the tomato plant is associated with a number of insect and mite pests of which the lepidopterans, Spodoptera eridania, Heliothis virescens, Keiferia lycopersicella, and the fruit-piercing bugs, Nezara viridula and Phthiapicta are the most serious. Other associated species include Spodoptera sunia and furgiperda, Agrotis repleta, Manduca quinquemaculata, Trichoplusia ni, Arvelius albopunctatus, Scapteriscus vicinus, Liriomyza sativae, various Chrysomelid beetles, membracids and mites. This is quite a range of pest species. Nonetheless, data col- lected at Piarco in 1982 suggests that approximately 6% of fruits may be lost through direct feeding by the entire pest complex, most species being responsible for less than 1% fruit damage. In 1981, much higher losses were observed in the Piarco area, in- festation being caused by both the army worm Spodoptera eridania and the fruit worm Heliothis virescens, in a situation in which the pin worm Keiferia lycopersicella was almost non- existent. Such high infestation levels of Heliothis had not other- wise been observed in either commercial farms or experimental plots during the period 1980 to 1984. Heavy pinworm infesta- tions on commercial farms had been reported during the mid-1970's, but outbreaks seem sporadic and documentation of data which confirms such high infestation levels is wanting. Data collected during 1982 to 1984 have demonstrated low pin worm fruit infestation levels. 176 Preliminary defoliation studies conducted under greenhouse conditions at Centeno (1983) suggest that yields may not be significantly affected by defoliation if all plant requirements are optimally supplied. Up to 50% defoliation produced no signifi- cant yield reductions. Similar observations have been made in the U.K. (Stacey, 1983). Perhaps this is due to the fact that the tomato plant produces a tremendous amount of foliage so that the uninfested leaves may be capable of manufacturing sufficient food for fruit production, or as Stacey suggests, there may even be some degree of over-compensation. The negative effects on yield which occur under conditions of disease are far more drastic, perhaps because in addition to reducing chlorophyllic leaf tissue, diseases such as Pseudomonas solanacearum cause collapse of xylem bundles with consequent interruption of food supply to the inflorescenses. In the southern desert valleys of California, Heliothis and Spodoptera damage occurred in 21% and 8.5% of fruits in un- treated fields for 1979 and 1980 respectively (Steenwyk, 1983). This is considerably higher than any recorded observations for Trinidad. Of the more serious pests, H. virescens is the only true fruit feeder. In 1981, approximately 6,800 Calypso fruits /wk/ha were fruitworm infested. In addition, many stems and calyces were tunnelled, causing flower loss. At that level of infestation, both green and orange phases were present. During 1982-1984's sampling, such tunnelling was almost nonexistent and the green phase of the worm was absent. Calypso fruit infestation/wk/ha was 656, 4,350, and 1,875 for 1982, 1983 and 1984 dry season trials respectively. B. thuringiensis (16,000 IUP per mg) in the form of Dipel at 0.55 kg AI per ha, reduced infestation levels in the aggregate of five varieties of tomato fruits by 54.5% and 57.6% respectively in dry and wet season trials of 1982. At an ap- plication rate of 1.1kg AI/ha using a high volume sprayer, reduc- tions were 76.2% and 78.8% respectively for dry season trials of 1983 and 1984. Weekly applications of B. thuringiensis in the form of Dipel at 0.56 kg AI/ha, reduced injury from 7.1% to 1.6% in Southern California when a boom sprayer was used (Oatman et al., 1983). Infestation levels seem slightly higher during the wet than dry season and peaked at about the same time as did fruit produc- tion. This may have been due to the fact that oviposition is related to flower production (Zalom et al., 1983). Polistes canadensis is an extremely efficient predator of this pest and low population levels observed at Corn/Soyal Piarco from 1982 to PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY-VOL. XX