Cotton Boll Weevil: An Evaluation of USDA Programs : a Report (1981)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2. COTTON INSECT CONTROL HISTORY OF INSECTICIDE USE ON COTTON Before the boll weevil invaded the United States from Mexico in about l892 there was relatively little damage to cotton by insect or spider mite pests. Arsenical insecticides—Paris green, London purple, and lead and calcium arsenates—were used to control occa- sional outbreaks of the bollworm and the cotton leafworm (Alabama argillacea). Nicotine sulfate dusts were sometimes employed to control the cotton aphid (Aphis gossypii) (Reynolds et al. l975). The cotton boll weevil, an invader almost devoid of natural enemies, rapidly changed this placid scene. By l909 it was reported to be causing at least $200 million in damage annually in the United States (Metcalf et al. l962) , and by l922 it had spread as far as the coastal regions of North Carolina and Virginia. The history of insecticide use on cotton insect pests can be divided into four periods, each of which is discussed below. The Arsenical Period The first insecticide recommended for control of the boll weevil was a spray utilizing Paris Green, London purple, or lead arsenate in combination with molasses (Townsend l895, Maley l902). The tech- niques for formulating and applying the spray were inadequate, however, and it was not until l923 that Coad and McNeil (l924) demon- strated the effectiveness of applying undiluted calcium arsenate dust by airplane in controlling the boll weevil. The effectiveness of this method for controlling another pest, the cotton leafworm, had been demonstrated the previous year. The success of dusting with calcium arsenate from airplanes was repeatedly demonstrated in Georgia and Texas in the period l925-l927 (Post l924, Thomas et al. l929), and aerial dusting became the prin- cipal method of applying insecticides to cotton until the early l950s. The use of calcium arsenate rose from about 3 million lbs. in l9l9 to approximately l5 million lbs. in l925. U.S. production 20

2l reached 33 million lbs. in l929, 43 million lbs. in l935, and a maximum of 84 million lbs. in l942. The total amount of calcium arsenate applied to domestic cotton fields from l9l9 to l948 has been estimated at about 850 million lbs. (Shepard l95l). This vast amount of calcium arsenate had substantial effects on the pest fauna of cotton fields and on the environment. Calcium arsenate was toxic to the natural arthropod enemies of the cotton aphid (ladybird beetles, syrphid flies, lacewing flies), and as a result the aphid became a serious threat, requiring the addition of nicotine sulfate to the arsenical dust. Annual production of nico- tine sulfate was about l million lbs. from l938 to l940, most of it used domestically on cotton; consumption reached l,460,000 lbs. in l943 (Shepard l95l). The bollworm also became a significant pest in this era because of the destruction of its natural enemies by calcium arsenate, but satisfactory control of this pest was maintained because bollworm larvae themselves are fairly susceptible to calcium arsenate. Calcium arsenate also had an impact on the natural enemies of the cotton flea hopper and of lygus bugs, particularly in the South- west and in newly irrigated areas of the West. Sulfur dusts were incorporated into control programs to deal with these pests as well as occasional outbreaks of red spider mites (Tetranychidae) (Reynolds et al. l975, NRC l975). During the arsenical period most of the insecticides were applied in boll weevil-infested areas. The Organochlorine Period In l945 the organochlorine insecticide DDT became available for domestic use. DDT brought about a revolution in cotton insect control. It was a persistent contact insecticide, and it was oil- soluble and could therefore be applied as a spray. Improved aircraft spray technology resulted in low volume sprays that almost totally replaced dust applications. Subsequently, benzene hexachloride (BHC) and toxaphene also became major cotton pest insecticides, and they were followed by aldrin, dieldrin, endrin, heptachlor, and ODD (TDE). U.S. production of DDT increased to l64,l80,000 lbs. by l960, benzene hexachloride to 84,599,000 lbs. by l956, and the aldrin-toxaphene group to 90,67l,000 lbs. by l960. Although statistics are not avail- able, from one-quarter to one-third of the organochlorine insecti- cides produced between l945 and l960 was probably applied to U.S. cotton. These insecticides had two important qualities: (l) high initial effectiveness against a wide variety of cotton pests; and (2) lengthy persistence, which made it possible to control newly emerging insects and insects migrating into treated areas. Spectacular increases in yields were obtained at high profit levels for many years (NRC l975, Reynolds et al. l975), and it appeared that complete control over arthropod pests of cotton had been achieved. The use of the organochlorine insecticides also decimated the parasites and predators of cotton pests, however, and often resulted in an increase in red spider mites. Grave problems of environmental pollution also resulted.

22 The Organophosphorus Period In l955 it was discovered that the boll weevil was beginning to become resistant to the organochlorine insecticides. The result of this discovery was a gradual but substantial shift to organophos- phorus insecticides—parathion, methyl parathion, azinphosmethyl, malathion, and EPN. These pesticides were effective against the boll weevil at relatively lower rates than the organochlorine insecticides but were not effective in controlling the bollworm and the tobacco budworm, which achieved the status of major pests as their natural parasites and predators were decimated. To control all of the major pests, growers then resorted to various mixtures of DDT, toxaphene, endrin, methyl parathion, azinphosmethyl, malathion, and EPN. These mixtures initially provided control of aphids, fleahoppers, plant bugs, leaf-feeding caterpillars, and spider mites as well as boll weevils, bollworms, and budworms. Growers demanded insecticidal mixtures that would produce cotton fields almost completely devoid of insects (Reynolds et al. l975, NRC l975). By the early l960s, however, the bollworm and the tobacco budworm had also developed a high degree of resistance to the organochlorine insecticides and the carbamate insecticide, carbaryl (see Table 2.l), and by the late l960s the tobacco budworm in the lower Rio Grande Valley of Texas and northeastern Mexico developed resistance to the organophosphorus insecticides as well. The use of methyl parathion was increased to l5 to l8 applications per season, but yield losses continued and in some areas the crop was almost totally destroyed. As a result of tobacco budworm resistance, many producers were forced out of business, and cotton production ceased on about 700,000 acres (Adkisson l97l). The organophosphorus-resistant tobacco budworm then spread to Louisiana and Arkansas, and from there to the cotton states of the Southeast. Tobacco budworm resistance reached such a high level that it became virtually impossible to control this pest with any insecti- cide. Meanwhile, however, a side effect of the greatly increased use of the organophosphorus insecticides was a dramatic increase in the number of cases of human poisoning from insecticides and the resur- gence of pests on other crops, such as citrus, following spray drift from cotton (Adkisson l97l). Then, in l973, the U.S. Environmental Protection Agency (EPA) banned the use of DDT to control cotton pests. The EPA action was a marked change in public policy toward insect control by chemicals. DDT plus toxaphene, often with methyl parathion added, had provided satisfactory control of the boll weevil, the bollworm, the cotton fleahopper, and plant bugs in the cotton-producing areas east of Texas, and the ban on DDT resulted in a shift to intensive use of the organophosphorus insecticides, often in combination with toxaphene, sometimes with endrin or chlordimeform (Reynolds et al. l975). Nonetheless, insect control had become ever more costly and ever less efficient (Reynolds et al. l975, NRC l975).

23 TABLE 2.l Measures of susceptibility and resistance of Heliothis boll- worm and budworm to various insecticides. The LD 50' measured in micrograms per gram larval weight, was determined 48 hours after topical application to 4th instar larva. LD50 DDT Susceptible Resistant H. virescens 132 (19€l-Florida)b 16,123 (1962-Tenas)b 16,510 (1965-Tenas)b iL. 2ea 26 (1960-Texas)d 5,680 (1962-Tenas)c 28 (1959-Tenas)d 14,150 (1962-Texas) c Endrin 30 (1962-Tenas) c H. virescens 26 (1970-Peru)1 3,980 (1970-Colombia) i 34 (1970-Menico) * 12,940 (1965-Tenas)9 58 (1961-Tenas)b H^ zea 12 (1960-Tenas)d 130 (1965-Tenas)9 20 (1962-Tenas)c 530 (1970-Nicaragua) i Methyl parathion 23 (1970-Mississippi) H. virescens 0.53 (1977-G«orgia)a 2,110 (1970-Tenas) i 0.57 (1970-Peru) i 3,580 (1969-Mexico) i 2.2 (1969-Mississippi)1 H. zea 2.2 (1970-Tenas)1 150 (1970-Menico) i 180 ( 1970-Nicaragua )i Carbaryl 310 (1970-Guatemala) i H. virescens 304 (1961-Tenas)b 54,570 (l965-Tenas)9 H^ zea 110 (1972-Tenas)c 540 (1965-Tenas)9 Perme thriii H. virescens 0.097 (1974-Tenas)f 1.64 (1977-Arizona) e 0.12 (1977-G«orgia)a 3.13 (1976-Tenas)h 0.28 (1978-Arizona)e 5.4 (1974-Tenas)f 0.29 (1976-Tenas)h H^ zea 0.47 (1976-Georgia) f 1.1 (1974-Tenas) f *A11 et al. <1977) ^razzel (1963) cBrazzel (1964) dBrazzel et al. (1961) *Crowder et al. (1979) £D*vis et al. (1975) ^Reynolds et al. (1975) Ntolfenbarger et al. (1977) iwolfenbarger et al. (1973)

24 The Current Period The situation during the past five years has been one of growing awareness that, as one author puts it, the entire Cotton Belt has been on an "insecticide treadmill" (Van den Bosch l978). Although synthetic pyrethroids capable of controlling the tobacco budworm appeared in l978, there have been other less positive developments. In l976 chlordimeform was withdrawn from active use because of its carcinogenic properties. In l978 it was once again permitted to be used, but only under very strict conditions. In l980 a ban was placed by EPA on the use of endrin east of the Mississippi River because of that chemical's very high toxicity to fish and other aquatic organisms. At the present time some 33 insecticides and acaricides are regularly used to control cotton insect pests, and another four, including the chitin synthesis inhibitor, difluben- zuron, have received conditional registration from EPA for specific uses (USDA l979). Overall Chemical Use It has been estimated several times that from 40 to 50 percent of all crop insecticides in the United States have been used to control cotton insect pests (Pimentel l973; USDA l965, l970, l974, l978). In l97l, for example, 73.4 million pounds of active insecti- cide ingredients were applied to 7.5 million acres of cotton. This was equivalent to 9.8 pounds of active ingredients per acre. In l976, 64.l million pounds were applied to 7.0 million acres. This was the equivalent of 9.2 pounds per acre (USDA l979). This use was also approximately 40 percent of the l62 million pounds of the active insecticide ingredients used by all U.S. farmers that year. Using 40 percent as an average figure, one researcher has calcu- lated that about 2.3 billion pounds of active insecticide ingredients have been applied to the U.S. cotton crop since l950. That is an average of more than 200 lbs. per acre (Metcalf l980). The estimated quantities of insecticides applied to cotton over the period l964-l976 are summarized in Table 2.2. The table also shows how insecticidal control of cotton insect pests has changed since accurate records became available in l964. Control efforts since then have been marked by a steadily increasing proliferation in the number of insecticides used and substantial changes in quanti- ties. From l964 to l976 the total amount of organochlorine insecti- cides decreased by half and the use of DDT came to an end, while toxaphene remained in large-scale use. During that same period the use of organophosphorus insecticides has approximately doubled, with methyl parathion becoming dominant and EPN very widely used. RESISTANCE OF COTTON PESTS TO INSECTICIDES The application of insecticides to cotton has demonstrated clearly that natural selection of resistant strains of cotton insect

25 TABLE 2.2 Use of insecticides on cotton in the United States, 1964-l976. Insecticide Active ingredient (pounds n 1000) 1964 1966 1971 1976 Area treated (acre* n 1000) 1964 1966 1971 1976 Inorganic CalciUM arsanate 2,518 69 57 2* Organic I. Orqanochlorinei 55,778 49,703 42,619 27,277 14,252 10,157 6,130 3,762 aldrin 17 123 - - - 16 • 161 . clordane • 3 - - - 6 . . DOT 23,558 19,213 13,158 . 6,901 4,767 2,383 _ ODD (TDE) 191 167 - _ 61 33 - _ dieldrin • 11 U • • 36 174 . endoaulfan - U - •77 • 56 • 325 endrin 1,865 510 1,068 311 1,194 403 262 325 lindane 540 163 - — 636 298 • - nathonychlor - 6 - - • • - - strobana - 2,016 216 - - 225 18 - tonaphene 26,915 27,345 28,112 26,289 5,016 3,881 3,275 3,112 other 2,660 ft - - 428 285 - - II. Orqanophoaphat«» 15,196 13,624 29,376 30,980 10,237 7,865 11,427 12,824 azinphoanathyl 250 200 288 229 641 222 119 - bidrin - 1,857 77* 251 - 1,416 1,797 378 demeton 47 - - • 322 - - 658 diazinon - - - 36 _ - - 51 dimetnoate - - - 87 _ . _ 237 dilulfoton -565 300 225 1,819 619 473 553 1,400 EPN - - - 6,140 - • - 1,496 athion . 73 • . . 26 30 - •alathion 1,811 559 670 43 213 245 273 55 methyl parathion 8,760 7,279 22,988 19,981 5,420 3,577 6,384 6,166 •onocrotophoa - - - 1,487 - - - 1,494 parathion 1.636 2,181 2,560 680 7S1 MO 682 561 phorata 10 - 100 158 35 _ l82 U5 trichlorfon - 963 144 - - 512 191 . other 2,117 212 1,617 69 2,236 534 1,216 213 III. Carbaaatea 4,524 1,571 1,291 1,445 1,002 4l5 294 1,137 aldicarb - - - 470 - . • 171 carbaryl 4,510 1,571 1,214 385 . . . 177 methomyl - - 40 590 _ - 84 789 other organic 6 - 2 - 102 - 24 - IV. Miscellaneous botanicala-biologicala chlordiaefon Total iniacticida used: 4,437 78,022 64,900 73,357 64,139 2,912 SOURCE: Data from USDA Agricultural Economic Report No. 131 (1965); No. l79 (l970); No. 252 (1974); No. 4l8 (l978); and NRC (l975).

26 pests can occur rapidly, it has also been demonstrated that selec- tion can continue in a single species of pest so that it becomes resistant to several, often unrelated, insecticides. Since l947, as shown in Table 2.3, at least 2l species of cotton insects and mites have developed resistance to one or more insecticides. Of the 2l principal resistant species, l4 are resistant to at least two groups of insecticides, 6 are resistant to at least three groups, 5 (includ- ing the bollworm, the tobacco budworm, the cotton leaf perforator, the cabbage looper, and the beet armyworm) are resistant to four groups, and l, the tobacco budworm, is resistant to all five groups. Species resistant to some insecticides occur in localized areas of all the cotton-producing states (USDA l979). The boll weevil has shown resistance to some insecticides in l0 of the ll states where it occurs. DDT resistance developed in the boll weevil in l954 in Louisiana and Mississippi (Roussel and Clower l955), and the resistant strain of insect subsequently spread rapid- ly. By l960, all areas of the South and the Southeast infested by the boll weevil had reported the development of organochlorine- resistant weevils (Brazzel l96l). Bollworms and tobacco budworms have shown resistance to insecti- cides in all l2 of the major cotton-growing states. As shown in Table 2.l, the bollworm and particularly the tobacco budworm have developed enormously high resistance to DDT, endrin, methyl para- thion, and carbaryl, as well as to toxaphene-DDT mixtures. Rather surprisingly, as of l968 no cotton pest had been shown to have acquired resistance to calcium arsenate despite 25 years of heavy application (Newsome and Brazzel l968). This, however, may have been because scientific techniques for the study of resistance were not well-developed until after the introduction of the organo- chlorine insecticides, and also because resistance to insoluble stomach poisons like calcium arsenate is very difficult to measure. If anything, cotton insect pests appear to be developing resis- tance to new insecticides even faster than before. Two synthetic pyrethroids developed to control cotton insect pests, fenvalerate and permethrin, were given conditional registration by EPA in l979. Yet, as Table 2.l shows, there is already evidence that the tobacco bud- worm has developed resistance to permethrin. The history of chemical control of cotton insect pests during this century suggests that its future is doubtful. Cotton insect pests—particularly the various worms that feed on the leaves and bolls of the cotton plant—have shown resistance to insecticides in the rest of the world as well, and control is now obtained in some cases only by as many as 50 applications of insecticide per year. In Egypt, for example, more than 8ll million pounds of active insecti- cide ingredients were applied to cotton between l96l and l975, pri- marily to control bollworms and leafworms. The leafworm, however, exhibited resistance to virtually every available insecticide, and in l977 it was reported that no new insecticide used in Egypt had re- mained effective for more than 2 to 4 years (El-Sebae l977).

27 4{ 4J O • <a tn to a) 5- 0) W ••H 0) 'P 'P ^H -H 3 C S D •o > o t! « M 01 M 3 M HI 1 h W §Q 1 u 3 fl Qi O 0 0> 0 3 0 °'a 1 M - « 1 M 4) C C O a a -H §13 U 'H rH U M O K 1 o §• - £ O OQ.fi ,O , I i o 8 il M M M M M M M M M M M M MMMM MMMMMMMM XMXMM H I g U) 8 I 2 i i i * S •p •H • •H *H f 3 i •M 'H i i I I ag • 1 >, S 1 « x 3 "Q.MMH

28 Holistic Pest Control and Pesticide Use The tobacco budworm, Heliothis virescens, and the cotton boll- worm, Heliothis zea, have effective natural enemies, and as a result they are generally under adequate control over much of the Cotton Belt. The development of reliable techniques to assess the degree of biological control will make it possible to include entomophagous species in pest control plans (Hartstack et al. l975). The development of a strategy to preserve intact the parasite and predator species that help to suppress Heliothis is a key to successful pest management. Measures to control the boll weevil must be carefully refined, since the beneficial species holding Heliothis in check can be disrupted by chemical control measures aimed at the boll weevil. If a favorable ecological balance tending to suppress the Heliothis complex is destroyed by the application of insecticides for the boll weevil, there is usually no alternative but to continue applying insecticide until the crop is mature. Such a strategy is both very expensive and ecologically unsound. Far more than half of the cotton losses ascribed to insects and mites may be attributable to the Heliothis complex (DeBord l977). The crop losses and the increased costs of production caused by Heliothis, and the concomi- tant load of insecticide in the environment, are enormous. Careful management of boll weevil control programs can reduce the insecticide load, and careful timing of insecticide applications can avoid the destruction of the beneficial insects that control Heliothis. What this means is using insecticides toward the end of the growing season to minimize the number of boll weevil adults leaving cotton fields to overwinter (diapause control). This reduces the boll weevil population in the following growing season to the extent that insecticide control of the weevil is not needed until late in the growing season after Heliothis is no longer a problem. Various modifications of the diapause supression program have been used effectively in certain areas of the Cotton Belt since l964 (Cross l973). Entomologists do not anticipate that boll weevil management or eradication programs would remove Heliothis as a significant cotton pest. In the southeastern states particularly, continuing problems with Heliothis can be anticipated. Other plants, such as corn, will support Heliothis populations at levels sufficiently high to prevent beneficial insects from keeping them below the level at which they can cause serious economic loss. In certain years this could occur in most production areas as is the case with secondary pests. In- season applications of insecticides keep certain other pest species, such as the plant bugs, below damaging levels. RECENT DEVELOPMENTS IN COTTON INSECT CONTROL Plant Breeding and Cultural Management of Cotton Cotton cultivars from the United States are the main varieties grown in many countries. Much progress has been made in developing

29 cultivars that consistently achieve high yields under good management (Bridge et al. l97l). Some high-yielding cotton varieties have the ability to adapt to environmental stress, including substantial insect damage. The traits that provide such resilience have been discovered in rare and isolated types of cotton and utilized to confer host plant resistance (HPR) against the major cotton insect pests, including the boll weevil, bollworm, tobacco budworm, pink bollworm, and Lygus species. The development of host plant resistance is approached from a holis- tic viewpoint, i.e., reducing the overall vulnerability of the cotton plant to the entire insect complex. The traits desired for breeding strains of cotton that are resistant to insects are generally found in otherwise poorly adapted cultivars or in wild relatives of cultivated cotton. As a result, certain adverse effects also occur when traits that improve resis- tance are transferred into well-adapted cultivars. These agronomi- cally inferior but resistant strains of cotton are useful only under severe or chronic infestations when resistance is more important than yield potential. Cultivars with a specified characteristic that improves resistance often do not yield as well as nonresistant culti- vars in the absence of the pest species, in order for any HPR trait to be valuable, the long-term average protection afforded by the characteristic must be greater than the mean yield reduction due to the negative effects of breeding. Carefully designed experiments have given indications of the reduction in insect damage that various traits can provide, singly or in combination. It does not appear possible to calculate the effects directly, however, because the interactions among pest-resistant plants, the insects themselves, and beneficial species are too complex. Only insect-resistant cultivars that provide yields comparable to those of standard cultivars in a pest-free environment gain commercial acceptance. After two decades of increasingly intensive efforts to develop host plant resistant cultivars of cotton, only a few are generally accepted. They include Stoneville 825, Coker 4l3, and Tamcot SP-2lS. Each of these incorporates a single trait that gives measurable protection against Heliothis. The absence of nectary glands, nectarilessness, in Stoneville 825 provides some additional protection against plant bugs. Cotton hybrids may offer new opportunities for improving cotton cultivars. It may be more feasible to make an F^ hybrid from a resistant parent and a high-yielding parent than to try to combine resistance and agronomic performance in a single pure line (Milam et al. l980). Davis (l979) reported that certain interspecific hybrids have significantly higher yield than the standard commercial varie- ties. Thus, it may be possible to breed insect-resistant hybrids that are also high-yielding. Preliminary data indicate the possibil- ity of combining high yield and high bollworm resistance in an inter- specific hybrid of cotton (Call and Weaver l980) , and other types of resistant hybrids are being sought.

30 Breeding Strategies for Host Plant Resistance (HPR) As used here, the word "ambivalent" refers to a trait or charac- teristic that increases a plant's resistance to one insect species but increases its vulnerability to another. Most of the known resis- tance characteristics in the various kinds of cotton are ambivalent. According to Lukefahr (l977), red plant color is the only major HPR trait that shows no ambivalence. Because of ambivalence, breeders often seek combinations of HPR traits, a goal called compensated ambivalence. Ambivalence does not have to be compensated for if the insect species to which vulnerability is increased is not a serious pest in a particular region, but there is always the possibility that a species that does no harm to normal cotton may attack modified cotton (Murray et al. l965). Furthermore, natural selection among cotton insect pests may negate host plant resistance. It took many years for most biologists to realize the ecological impact of insec- ticides, and it may take a similar period of time to observe the full effects of widespread use of HPR varieties of cotton. A particularly effective technique for growing cotton is to intersperse a few rows of a susceptible strain at wide intervals in a field primarily planted in an HPR strain. The target insect pest tends to avoid the HPR plants and concentrate on the susceptible plants, which can then be treated with insecticide. This is called "trap cropping," and a number of ways in which it can be used to manage weevil populations have been described (Namken et al. l98l; Jones et al. l978a, l978b). Short-Season Varieties of Cotton Researchers in Texas have developed a technique for avoiding boll weevil damage by utilizing short-season varieties of cotton. The life span of the boll weevil in southern Texas is such that cotton will escape damage from the first generation of boll weevils to emerge each growing season if overwintered populations are less than 22 weevils per acre. Newly developed short-season strains set fruit rapidly and reduce the amount of crop damage from the second generation of weevils that develop later in that growing season (Walker and Niles l97l). In one experiment, two applications of insecticide at an early flowering stage in cotton (pinhead square stage) reduced boll weevil population levels below the economic threshold for 59 days, allowing most of the cotton bolls to mature. Use of the short-season tech- nique has reduced in-season insecticide applications by half and has avoided late-season increases in tobacco budworm populations (Heilman et al. l977). This early application of insecticides does temporar- ily disrupt the suppression of Heliothis, and as a result damage to pinhead squares by Heliothis may rise to 25 percent or more. But heavy damage can be endured at the early squaring stage, and if no more insecticide is used the natural enemies of Heliothis will recover in time to protect the crop through the main fruiting period

3l (Walker et al. l978). Since the size of the boll weevil population is related to the number of generations in a growing season, the elimination of a single generation can greatly reduce economic loss and the need for additional control measures. Namken et al. (l98l) point out the superiority of the early blooming rate of certain new cultivars over the standard full season cultivar, Stoneville 2l3. They show that a higher number of blooms per acre in the first 20 days of blooming lead to significantly earlier maturity and, in some cases, higher yields. For full-season varieties the duration of the fruiting period varies and is normally terminated by low temperatures in the fall (Gipson and Joham l968a, l968b). This means that in years with long warm fall seasons the use of the short-season varieties involves a deliberate sacrifice of yield (Fisher and Cannon l98l). Reduced yield, however, has proven to be an acceptable tradeoff for reduced vulnerability to insect attack in parts of Texas. Short-season cultivars also reduce the costs of water, labor, and machinery, and the savings in insecticide costs may be highly significant when the short-season technique is coupled with careful selection and timing in the use of insecticides (Walker et al. l978) . Host Plant Resistance Against the Boll Weevil Reduced oviposition (egg-laying) by the boll weevil has been found in a number of strains of cotton. Frego Bract. In the l960s a large number of trials demonstrated that the modified bract type called frego bract was attacked much less severely by boll weevils than normal cotton (Jones et al. l977). The boll weevil populations in frego bract fields were one third as large as the populations in cotton fields of other types when no diapause program was applied to either the frego or non-frego fields and in-season treatments for weevils were applied as needed (Jenkins and Parrott l97l). Weevil suppression through the use of frego was variable, depending in part on the size of overwintering populations and on in-season and diapause insecticide applications. Insecticide was not needed in frego fields until 4 weeks after non-frego fields received their first treatment. The resistance shown by frego bract cotton may therefore make it possible to postpone insecticide appli- cations that would otherwise disrupt the beneficial species which hold Heliothis in check. Four small test plantings of paired frego and non-frego cotton showed that the boll weevil populations on the frego were between 66 and 94 percent less than on the non-frego cotton. Resistance was attributed to the "upsetting of normal patterns of behavior" in the weevil. If natural selection then resulted in altered weevils which preferred frego, normal varieties might then show some resistance (Jenkins and Parrott l97l). The structure of frego bract cotton allows a much larger amount of insecticide to penetrate to the flower bud. There was signifi- cantly higher mortality of boll weevils on frego than on non-frego

32 cotton when both were sprayed with azinphosmethyl (Parrott et al. l973) . Frego bract varieties, however, have not yet become commercially feasible because of the extreme susceptibility of frego bract varie- ties to plant bugs. Damage to frego bract strains from plant bugs (Lygus species) may be as much as twice as great as damage to normal bract strains (Jones l972), and for this reason Meredith (l980) projects that commercial use of frego bract varieties will not occur in the l980s. J. E. Jones (Louisiana State University, Baton Rouge, personal communication, l98l) is confident, however, that the suscep- tibility of frego bract to plant bugs can eventually be overcome by combining frego with a compensating trait, such as nectarilessness that confers some resistance to Lygus. Red Plant Color. There are two HPR strains with red color that exhibit a valuable weevil-resistant trait. A gene that imparts intense red color (Rjj to the entire plant elicits as strong a negative reaction from the boll weevil as frego bract (Jones et al. l978a, l978b) but gives a lower yield than a red stem (R2) type. Cotton plants with the red stem trait are competitive in yield with their normal green counterparts under all but the most favorable conditions (Jones et al. l977). Host Plant Resistance Against Other Key Pests Nectarilessness. A trait that makes cotton partially resistant to attack from both Lygus and Heliothis became available when the genetic factors causing the absence of leaf, bract, and involucral nectary glands were bred into upland cotton from the wild species G. tomentosum. The trait is controlled by two unlinked recessive genes (Meyer and Meyer l96l) and poses no great difficulty in breeding. Nectariless types have been backcrossed into three major varie- ties and had no significant effects on yield or fiber properties (Meredith et al. l973). Significantly reduced Lygus populations have been reported on nectariless varieties (Schuster and Maxwell l974). Genetic modification of normal cotton into nectariless cotton has also been reported to result in reduced bollworm egg-laying by several investigators (Davis et al. l973, Lukefahr et al. l965, Schuster and Maxwell l974). This reduction may be close to 50 per- cent, but there was high variability between trials (Schuster and Maxwell l974, Davis et al. l973). Part of the variability may be due to the fact that nectarilessness also suppresses populations of beneficial insects that attack the bollworm (Schuster and Maxwell l974; J. Ellington, New Mexico State University, Las Cruces, unpub- lished personal communication, l98l). Leaf Smoothness. Another modification that significantly affects Heliothis behavior is leaf smoothness. This trait has been transferred into upland cotton from the wild G_^ armourianum (Meyer l957). The genes responsible for the trait have also been found in Central American "dooryard" accessions, and in a commercial variety of American upland cotton (Lee l97l). Combinations of two or more of

33 the genes that account for smoothness can produce glabrous or "super smooth" cotton. The smooth-leaf characteristic is highly ambivalent, however. Glabrousness gave cotton plants (except for North Carolina Smooth) resistance to Heliothis, the cotton fleahopper, and the pink boll- worm, but resulted in greater numbers of cabbage loopers and leaf- hoppers (Lukefahr l977). The tarnished plant bug (Lygus lineolaris) caused a significantly greater reduction in the number of flower buds and in the lint yield of smooth leaf cotton as compared to pubescent cottons (Meredith and Schuster l979). Jones et al. (l977) confirmed the increased susceptibility of smooth-leaf types to plant bugs and leafhoppers. The primary value of leaf smoothness is the protection it provides against Heliothis. Significantly fewer Heliothis eggs were laid on Deltapine smooth leaf cotton than on normally pubescent Stoneville 2l3. The suppres- sion effect of smoothness on bollworm egg-laying is confirmed by Lukefahr et al. (l965). Crossing glabrous and frego bract cotton with okra-leaf cotton partially reduced their susceptibility to plant bugs (Jones et al. l978a) . Okra leaf counters potential plant bug damage by enhancing the fruiting rate. High resistance to whitefly was also associated with okra leaf and super okra leaf. Near-glabrousness gave a mod- erate degree of resistance (Jones et al. l976). The smooth-leaf types (except for 31113) have been reported to have a low percentage of lint and erratic yield (Meredith l980). Glabrous isolines were slightly lower in yield and significantly later in reaching maturity than their hairy counterparts. The late- ness of these glabrous types was associated with susceptibility to an "early season pest complex" involving plant bugs and leafhoppers (Jones et al. l977) . High Gossypol. There are naturally occurring plant pigments in cotton, notably gossypol, that are toxic to some insects at high concentrations (Lukefahr and Martin l966). The gene that results in high concentrations of floral bud gossypol confers resistance to Heliothis and Lygus and may suppress leafhoppers, but it has the ambivalent property of leading to severe attack by thrips and white- flies. Nonetheless, the excellent protective effect of high gossypol against Heliothis has resulted in intensive efforts to incorporate this trait into commercial cotton (Sappenfield and Dilday l980). Parrott et al. (l98l) reported that certain high-gossypol strains of cotton showed no yield loss when infested with tobacco budworm. Artificial infestation (Parrott et al. l98l) and the withholding of insecticide protection (Bailey et al. l978) have been used to demon- strate the protective value of high gossypol concentrations. Breeding strains of cotton with high gossypol content, however, is a lengthy and difficult procedure (Sappenfield et al. l974). Furthermore, high gossypol has a negative effect on yield (Meredith l980, Sappenfield and Dilday l980). Gossypol content has also been reported to be negatively correlated with boll size and the ratio of lint to seed (Wilson and Lee l976, Dilday and Shaver l980).

34 Therefore, the best way to achieve high yield plants with a high gossypol content may be through interspecific hybrids. Singh and Weaver (l972) reported an interspecific cross whose gossypol content was closer to that of the Pima cotton plant parent with high gossypol than to that of the XG-l5 upland parent with a lower gossypol content. The Prospects for Host Plant Resistance (HPR) Boll weevil-resistant varieties of cotton will not become a reality until breeders are able to combine traits that help cotton plants resist the boll weevil with genetic backgrounds that insure at least a normal level of resistance to other cotton insect pests. Red stem varieties of cotton with a minimum of negative traits will probably be commercially important in the near future, but frego bract's resistance to the boll weevil cannot be exploited until the variety's increased susceptibility to Lygus is overcome. Most of the available HPR traits have been known and used by cotton plant breeders for more than a decade. This is about the length of time needed to breed out the agronomic defects that come from introducing traits from exotic plant varieties. The future of breeding for host plant resistance in cotton holds promise of signif- icant breakthroughs. In addition, integrated pest management pro- grams may accelerate the development of trap cropping systems in which both resistant and susceptible varieties of cotton play a useful role. Pheromones of Cotton Insect Pests Much of the fundamental behavior of insects in searching for food, sexual partners, and egg-laying sites is controlled by the release of specific chemical signals, called semiochemicals, produced in the insect environment. Semiochemicals that act interspecifically are called allomones if they favor the insect that produces them and kairomones if they favor the insect that receives them (Brown et al. l970) . Semiochemicals that act intraspecifically between individuals of the same species are called pheromones (Karlson and Butenandt l959). In the two decades since the identification of the sex pheromone of the silkworm Bombyx mori as trans-l0-cis-l2-hexadecadien-l-ol, or bombycol (Butenandt et al. l959), intensive study has demonstrated the almost ubiquitous presence of these chemical messengers in insect species and their essential role in reproduction. The sex pheromones of many insect pests, including several pests of cotton, have been identified and are now available as synthetic chemicals. Much pro- gress has been made in using these sex pheromones to monitor the environment for the presence of the pest, to trap and destroy large numbers of insects seeking mates, and to suppress mating and repro- duction by confusing and disrupting natural pheromone signals. Captures in traps baited with pheromones have been studied as a way

35 of predicting the need for insecticide applications to combat the pink bollworm (Toscano et al. l974) and the boll weevil (Rununel et al. l980). Current knowledge about the chemical identity of pheromones of important cotton insect pests is shown in Table 2.4. The sex phero- mones of most Lepidoptera are blends rather than simple chemical compounds, and "fine tuning" of the pheromone blend is essential to elicit maximum insect response. The pheromone of the cabbage looper is apparently an exception to the preceding statement, since it consists of only a single component. Even with Heliothis, however, significant communication response between male and female moths has resulted from employment of a single parapheromone, cis-9-tetra- decenyl formate (Mitchell et al. l975). Little effort has been made to use pheromones to control either Heliothis or the various kinds of armyworms, but communication between cabbage loopers has been disrupted by using l00 evaporative sources per 0.l hectare plot (Gaston et al. l967). Although the sex pheromone blend of Heliothis is used in survey and detection, much research is still needed to improve the use of this "tool." Boll Weevil The identification and synthesis of the components of grandlure, the boll weevil sex pheromone (Tumlinson et al. l97l) , affords new opportunities for cotton insect pest management. Unlike the sex pheromones of a majority of the lepidoptera, which are produced by females, grandlure is elaborated by the male boll weevil in fecal pellets. Grandlure apparently functions as an aggregating pheromone during early spring and again during the fall, when boll weevil populations migrate. During the cotton plant's fruiting period grandlure functions as a male sex pheromone with a relatively short range. Grandlure is a combination of two terpenoid alcohols and a cis-trans mixture of aldehydes (Table 2.4). It has been especially useful in monitoring boll weevil infestations. Dispensers containing 25 mg of grandlure are effective for about 4 weeks, and more than l million of the dispensers were used in monitoring experiments between l973 and l977. A number of attempts have been made to use traps baited with grandlure to suppress the spring population of boll weevils as it emerges from overwintering sites. In fields of 35 and 73 acres in Mississippi the use of l0 pheromone traps per acre was estimated to have captured 75 percent of the overwintering population (Mitchell et al. l976). The efficiency of such pheromone traps is inversely related to pest density; hence, trapping is a feasible control measure only when boll weevil populations are already at low levels (Mitchell and Hardee l974). Knipling (l979) has explored many of the theoretical problems involved in determining the optimum employment of grandlure.

36 S € & M 8 X o 1 M pi £• i h 1 H a 1 I *5 i i « l~ 10 — 4-> 4J SS a ,H JH II « H S- H § S I •& | 1 1 H 1 if _ s « 01 *H r* T3 iV 5 . 3 • 9 •g s 5 S 2 • ! S • §U 4J O 01 >* 01 Id E Pf Jl • 1 * - u 2 g S n I, : s M & £S S5 3 ^555 P55^, H x II -H ss •rt *. g «^ H 41 O 35 01 -H O O O 8' is-Ai-nenaaecaa etradecenyl ace rans-12 tetrade etradecenyl ace rans-12-tetraca 4J 1 >• X 1 5" •-! O O <J <U >* 1 1 1 1) O E <J < •-> •H i- «J *H •H 1 < fH rH X ' ? i O a 1 I >, j — i — i i id i) : « « E E x a • &A p< j? S S S E IBB 01 01 1 ) 01 01 O 0 ,H . 1 U U 01 01 iH ) 01 01 *O fO iH 1 > C O >i S 4J fH 9 I- X X Oj 0 01 01 1 •O 01 01 1 >* 4J a 4-1 4J E E X 0 01 01 .3 «j >o *a <H < 1 T3 T3 iH Id «J « » • iillii s g ^5557 Bill ; Oi i *O *O m •• f. oi 01 id •• j 01 X X E 1 :uoi«xtio M - ?« IN 1 1 I i*> m » IH 1 £ £ O <M O rH 1 1 0) 0 - X X O 1 4J 0) 4>J 1 1 1 O iH 1 -O 1 u u *> YV » n fi •H d p* *o ^ 4 o* r*- ^I FH ff* i d p* I I i 9 i i h i i i i W^ O"i 7 |VTTS 0) -H « » 1 Siii. i S 8. > » 0 X • CO JJ • r 1,1 iia 5 °olT! "o £ 3 1 n • n ) ol O R OI O *J O u ol o •^1 *H ol E «- O 0 4J ol o a t 1 s en s Q B • | g • a $ • X «* i i 4-' H

37 Pink Bo11worm Both the natural pheromone gossyplure, or cis-7-cis-ll-hexadeca- dienyl acetate, and hexalure, a less active synthetic parapheromone (cis-7-hexadecenyl acetate) have been studied for use in control programs. The pink bollworm population in the newly infested area of the San Joaquin Valley, California, has been monitored with about l00,000 "delta" traps, each containing about l,000 micrograms of gossyplure, as part of a male sterilization program (Foster et al. l977) . The use of gossyplure to control the pink bollworm in cotton fields seems to have been evaluated most thoroughly in the People's Republic of China. It has been reported (NAS l977) that in a 27- hectare cotton field in China baited with 30 traps per hectare containing gossyplure, 290,000 male pink bollworm moths were cap- tured. This was estimated to be about 25 percent of those present. The disruption of communication between male and female pink bollworms using hexalure also is a practical possibility. In one experiment, the release of about 330 g of hexalure per hectare during the l6-week growing season caused most of the females to remain unmated and thus produced a reduction of 75 percent in the larval population of the next generation (Shorey et al. l974). It has been estimated that by using the best techniques of microencapsulation and impregnation in hollow microfibers, satisfactory control of the pink bollworm could be obtained by releasing about l5 g of gossyplure per hectare per season. Male Sterilization The use of sterile males to control insect pest populations is a relatively new technique. A Russian geneticist named Serebrovsky originally outlined this approach in l940 (Proverbs l969, Whitten and Foster l955). The first significant application of the technique, in which male insects are exposed to radiation, was the eradication of the screwworm fly (Cochliomyia hominivorax) from the southeastern United States (Knipling l960). The technique is now being used to maintain eradication through the annual release of 8 to l0 billion sterilized flies in the southwestern United States and the adjacent area of Mexico (CEQ l978). Successful suppression of the codling moth, using both sterile males and sterile males plus sterile females, has also been demon- strated. A release rate of 40 fully sterile insects to one wild insect was sufficient to rapidly reduce the natural population. The cost, however, was considerably higher than the cost of conventional control methods (Proverbs l970). Programs of less demonstrable success have also utilized sterile males. Approximately l00 million sterile pink bollworm moths have been released annually in the San Joaquin Valley to prevent the establishment of this pest. In a similar program, sterile Mexican fruit flies have been released annually in southern California since l964, the purpose being to prevent the establishment of flies dispersing from Mexico.

38 Numerous experiments with male sterilization in other pest species, such as the oriental fruit moth, the melon fly, the Carib- bean fruit fly, the cotton bollworm, the horn fly, the stable fly, and several mosquito species have shown favorable results. Diptera species appear to be the most amenable to effective sterilization. Lepidoptera and Coleoptera require larger and more debilitating doses of radiation to effect sterilization and present a complex problem of determining the competitiveness of sterilized insects with unsteril- ized ones. Continual improvements in equipment, diet, and technique for the mass rearing of the boll weevil have occurred over the past quarter century (Griffin et al. l98l; J.E. Wright, Mississippi State Univer- sity, Mississippi State, MS, personal communication, l98l). In l980, six million adults could be delivered per week to the North Carolina eradication project at a cost of $3 to $4 per thousand (T.B. Davich, Mississippi State University, Mississippi State, MS, personal commun- ication, l98l). The principal problem in obtaining sterile boll weevils has been the development of a radiation treatment that would achieve full sterility of both sexes. Gamma irradiation was tried initially but had to be rejected because the dose inducing permanent sterility was rapidly lethal (Davich and Lindquist l962). During the following fifteen years a variety of chemosterilants was tried, but all had at least one major drawback. Either the sterilizing dose was debilitat- ing or fatal, or the dose failed to induce complete sterility, or it proved impossible to sterilize females at the dosage capable of sterilizing males (Wright and Villavaso l98l). Attempts to discover a new method of sterilization were intensi- fied after the inconclusive Pilot Boll Weevil Eradication Experiment (PBWEE) in l97l to l973. The methods used in PBWEE sterilized males only, and the two sexes of the mass-reared insects had to be sepa- rated by hand, an obvious and very costly defect in the procedure (Lloyd et al. l976; Davich l976). Recent tests indicate the best method of sterilization now available is to feed weevils for 5 days with diflubenzuron followed by a gamma irradiation treatment in a nitrogen atmosphere (J.E. Wright, Mississippi State University, Mississippi State, MS, personal communication l98l). Reproduction by adults treated by this method is apparently zero, since treated weevils failed to establish a detectable population in a weevil-free area (Mitchell et al. l980). The longevity of male weevils sterilized by the gamma irradia- tion-diflubenzuron process, however, is severely affected. Mortality at day eight was 96 percent in a sterilized weevil population, as compared to 40 percent for unsterilized males (Boll Weevil Research Laboratory l98la). Studies of the attraction of females to traps baited with sterile males have indicated that the attraction of females to both sterile males and grandlure is greatly reduced by the presence of normal males (Boll Weevil Research Laboratory l98la) . The most recent studies place the field competitiveness of weevils treated with irradiation and diflubenzuron at 23 percent for the first four days after release. The apparent loss of vigor after

39 day four is supported by day seven mortality figures of 58 percent for sterile males and 6 percent for normal males (Boll Weevil Re- search Laboratory l98lb). Attempts to increase the longevity of sterile males are continuing (T.B. Davich, Mississippi State Univer- sity, Mississippi State, MS, personal communication, l98l), and a new technique for the aerial dispersal of weevils demonstrated good dispersal throughout the target field for the first time (Boll Weevil Research Laboratory l98lc). Unfortunately, there have been no experimental demonstrations of population suppression using fully sterile weevils produced by cur- rent technology. In l980 an extensive test involving two methods for the dispersal of sterile weevils in a test and control area contain- ing comparable weevil populations in North Carolina provided no clear findings (J.E. Wright, Mississippi State University, Mississippi State, MS, personal communication l98l; Boll Weevil Research Labora- tory l98lc). Inadequate preliminary monitoring and the seasonal influx of boll weevils from more heavily infested areas are believed to have contributed to lack of success in this experiment. A large-scale test was also conducted in the Mississippi Delta in l980, utilizing a 300-acre test site adjacent to a 500 acre con- trol site. This test also failed to give a clear cut indication of suppression. Reduction in egg hatch in the infested squares was considered to be the best criterion, and there was only a slight (3 percent) reduction in egg hatch due to the release of sterile boll weevils. This test was performed under severe climatic stress condi- tions not favorable to weevil survival and reproduction (T.B. Davich, Mississippi State University, Mississippi State, MS, personal commun- ication l98l). High temperatures on the soil surface are believed to have contributed heavily to mortality among the sterile weevils. Tests scheduled for l980 in Nebraska to provide more information were apparently scrapped for lack of funds. In summary, the use of sterile male boll weevils for eliminating natural boll weevil populations suffers from inconclusive data. Total suppression has been attempted in only a few field experiments, and these experiments have failed in one way or another. Diflubenzuron Diflubenzuron, or N-(4-chlorophenyl)-(2,6-difluorobenzoyl)- urea*, was originally described (Wellinga et al. l973) as an insect growth regulator with a novel mode of action—a specific inhibitor of the synthesis of the N-glucosamine polymer, chitin, a critical compo- nent of the insect exoskeleton. This specific biochemical inhibition is a common property of an extensive series of substituted benzoyl ureas (Wellinga et al. l973). Two related compounds, penfluron, or N-(4-trifluoromethylphenyl)-(2,6-difluorobenzoyl)-urea (Olivier et al. l977), and trifluron, or N-4-(trifluoromethylphenyl)- *also named N-(4-chlorophenyl)-aminocarbamyl-2,6-diflurobenzamide.

40 (2-chlorophenyl)-urea, have been reported to be more effective than diflubenzuron, in some cases, and less effective in others, depending on the insect species. Only diflubenzuron, however, has conditional EPA registration for use on cotton against the cotton boll weevil and on forests for control of the gypsy raoth, Lymantria dispar. Because of the complexities of the registration process, it seems unlikely that competing products will rapidly displace diflubenzuron. The chemical structure of diflubenzuron is closely related both to those of the persistent herbicidal ureas—e.g., monuron, or N-(4-chlorophenyl) N, N-dimethylurea—and to that of the herbicide dichlorbenil, or 2,6-dichlorobenzonitrile. All of these herbicides have been widely used in the United States for three decades (Herbicide Handbook l974). Degradative Pathways Studies of the degradation of diflubenzuron have been made with three different radiolabeled moieties (Metcalf et al. l975). It has been shown that the parent molecule cleaves at both C(O)-N bonds in the -C(O)NHC(O)NH- bridge to form the primary degradation products 2,6-diflurobenzamide and p_-chlorophenylurea. The 2,6-difluoro- benzamide is subsequently converted to 2,6-difluorobenzoic acid and the p_-chlorophenylurea to p_-chloroaniline. None of these degradation products was found to be biomagnified extensively in the organisms of the laboratory model ecosystem (Metcalf et al. l975). Persistence Diflubenzuron has a low water solubility of 0.3 to 0.l0 ppm, and it has an octanol/water partition coefficient of about 3500. Extensive laboratory studies with three different radiolabeled preparations of diflubenzuron showed that it does not become highly bioconcentrated. The bioconcentration or ecological magnification factors for the parent compound were alga, l8 to 83 times; snail, 86 to l35 times; mosquito larva, 596 to 779 times; fish (Gambusia), l4 to l9 times (Metcalf et al. l975). These factors indicate a decrease in residue concentration through food chains. The levels of bioconcentration are miniscule compared to those for DDT and other organochlorine insecticides under similar test conditions (Metcalf and Sanborn l975) . The relatively high values in mosquito larva reflect the affinity of diflubenzuron for the insect cuticle. Diflubenzuron has moderate persistence in biological systems. In a laboratory model ecosystem treated with diflubenzuron, the percentages of parent diflubenzuron found in the organisms after 30 days of exposure were alga, 46 to 6l; snail, 73 to 90; mosquito larva, 84 to 98; and fish, 5.2 to 6.7. The range of values shows the results with two separate radiolabeled preparations, one labeled in the difluorobenzoyl and the other in the £-chloroaniline moieties (Metcalf et al. l975).

4l Persistence in Water. Diflubenzuron may persist in water for considerable periods. After 33 days in the water phase of a laboratory model ecosystem, two radiolabeled preparations of diflubenzuron had 24 to 3l percent of the total extractable radioactivity present as the parent compound (Metcalf et al. l975). The primary hydrolysis products are 2,6-difluorobenzoic acid and £-chlorophenylurea, and their production is photochemically catalyzed through conversion to 2,6-difluorobenzamide and p_-chlorophenyl isocyanate intermediates (Metcalf et al. l975). The half-life of diflubenzuron in water is strongly pH-dependent, ranging from about l day under alkaline conditions to 2 weeks or more under neutral or acid conditions (EPA l979). Persistence in Soil. As expected from its similarity to monuron, diflubenzuron can be highly persistent in soil. The residues of two radiolabeled preparations placed in the soil of a laboratory model ecosystem showed only about l percent decomposition after 4 weeks at 27°C (Metcalf et al. l975). Verloop and Ferrell (l977) demonstrated that soil persistence of diflubenzuron was related to particle size, with 2 micron-sized particles having a half-life of 0.5 to l week and l0 micron-sized particles having a half-life of 8 to l6 weeks. The effect of particle size on the degradation rate of diflubenzuron residues on cotton plants that are shredded and cultivated into the soil after harvesting is not clear. One experiment showed that such diflubenzuron residues showed no appreciable degradation 9 months after being placed in the soil. The content of the residue recovered by solvent extraction was 95 percent intact diflubenzuron (Bull and Ivie l978). Diflubenzuron was found to bind tightly to soil particles and did not readily leach away. Toxicology The action of diflubenzuron in inhibiting chitin synthetase (Verloop and Ferrell l977, Hajjar and Casida l978) suggests a high degree of specificity of this insecticide for insects and related arthropods and for Crustacea. The concentrations of diflubenzuron that are lethal to aquatic invertebrates are very small, with acute LC5Q values in ppm for water flea, Daphnia, 0.00l5; sand flea, Gammarus, ca 0.040; clam shrimp, Eulimnadia, 0.000l5; mysid shrimp, Mysidopsis, 0.002; and brine shrimp, Artemia, 0.002 (LC50 is the concentration in water that is lethal to 50 percent of the aquatic organisms in the test population). Chronic reproductive effects over 6 to 30 days were observed in blue crab at 0.0005 ppm and March crab at 0.00l ppm (EPA l979). Vertebrate animals are much less susceptible. The UC$Q value for guppies is l00 ppm, and phytotoxic effects were not observed at levels up to l0,000 ppm. The lethal oral dose (LD50) for laboratory rats and mice is greater than l0,000 rag/kg. Diflubenzuron per se does not appear to cause tumors, but its primary degradative pathway to £-chloroaniline, which is related to known human bladder carcinogens, has raised questions about diflubenzuron's overall tumor-causing potentiality. P-chloroaniline is mutagenic in the Ames Salmonella assay (EPA l979).