Pest management in cotton

Cotton (primarily Gossypium hirsutum but also G. barbadense) is grown in over 70 countries and is the most important fibre crop grown worldwide (Gutierrez, 1995; Matthews, 1997b). Annual global production is approximately 19 X 106 t lint (Gillham et al., 1995) with China leading production followed by the USA, Central Asia, India, Pakistan, Brazil and Egypt. The need for irrigation and the dominant influence of pest problems in most systems greatly affects potential profits by increasing production costs and reducing yield (Luttrell et al., 1994). Insects are a major constraint and cotton was one of the first crops on which insecticide use reached unacceptable levels. The problems that subsequently arose had a dramatic impact on the cotton production industry and led to a concerted effort to find alternative systems of pest management.

In the USA misuse of chemical pesticides to control Lygus hesperus in the 1960s in the San Joaquin Valley induced other more serious pests, such as the cotton bollworm (Heliothis zea) and a number of defoliating noctuids (Spodoptera exigua, Trichoplusia ni and Estigmene acraea;

Gutierrez, 1995). In the Imperial Valley of California the pesticide induced outbreaks of whiteflies, mites and insecticide resistance to Heliothis virescens caused economic ruin in the 1980s and the industry effectively collapsed, falling from c. 44,000 ha of production to c. 6000 ha. In the Sudan, the use of aerial applications of chemicals so reduced the impact of natural enemies that the cotton suffered major outbreaks of whiteflies, Bemisia tabaci, the honeydew from which encouraged sooty moulds to the extent that the lint was downgraded (Eveleens, 1983). During the 1950s in Peru extensive aerial application including over 30 sprays per season of DDT, parathion and toxaphene caused the death of and poisoned thousands of people (Matthews, 1997b). Overall, the overuse and misuse of chemicals in cotton pest management represents an abject lesson in what not to do with a toxicant in an agricultural system, creating more problems than it solved.

Wherever cotton is grown in the world there exists an impressive array of pest insects that are associated with reductions in yield. There are pests from five different orders, Lepidoptera, Hemiptera, Coleoptera, Thysanoptera and Acarina, although a complex of Lepidoptera including the cotton bollworms Heliothis and Helicoverpa, spiny bollworms Earias spp. and red boll-worm Diparopsis spp. are most well known. The cotton boll weevil Anthomonas grandis, the tobacco bud-worm Heliothis virescens and plant bugs such as jassids, whiteflies and aphids (particularly Bemisia tabacci and Aphis gossypii) can cause excessive fruit loss. The Lygus bugs attack very small fruits hence the plant has time to compensate for this damage unless the rate of loss is high (Gutierrez, 1995). However, boll weevils and bollworms attack more mature fruits making it more unlikely that the plants can compensate, hence damage tends to be greater from these pests. The pink boll-worm Pectinophora gossypiella similarly causes most damage when it attacks mature bolls and yield losses accrue through destruction of seed and reduced lint quality (Gutierrez et al., 1977; Stone and Gutierrez, 1986a,b).

Mites, particularly the red spider mite Tetranychus urticae, kill cells in photosyn-thetically active leaves and hence have an impact on yield. Insect defoliators such as Spodoptera exigua and Trichoplusia ni also reduce availability of photosyntheti-cally active area but only have a significant effect on yield when they attack young leaves (Gutierrez, 1995). Yield may also be reduced by stem borers such as Eutinobrothrus brasiliensis that kill whole plants and stunt the growth of others; this can be prevented by application of seed treatments (Dos Santos et al., 1989).

Chemcial insecticides remain the major tactic for insect pest control in cotton. Most cotton is routinely examined for insect pest damage with private consultants providing scouting services in the USA and Australia (Fitt, 1994; Luttrell et al., 1994). In Australia, scouts examine up to 60 plants per 100 ha, 2-3 times per week whereas in the USA most fields are scouted 1-2 times per week and samples are generally based on examination of 100 terminal buds and fruiting forms per 20-40 ha. The use of scouting in Andhra Pradesh cotton by both state and federal agencies has reduced the number of pesticide applications from as many as 20 to 3-6 (Raheja, 1995). Cotton consumes 50% of the insecticides used annually in India even though it occupies only 5% of the cultivated area; 80% of synthetic pyrethroid consumption is confined to cotton alone. Hillocks (1995) lists the insecticides used in cotton in Africa including: Endosulphan for Lygus, American bollworm, spiny bollworm and cotton leafworm control, Pirimicarb for Aphis gossypii control, Amitrax for red spider mite and whitefly control and Dimethoate for cotton stainers.

Although economic threshold levels for pesticide application against key pests have been known since the 1950s in countries such as Sudan they were not modified over decades which resulted in the application of insecticides at lower densities than was actually necessary (Zethner, 1995). A wide variation in threshold values exists across and within different cotton systems for specific pests. Thresholds for Heliothis species in Brazil range from 10 to 40% of the squares infested by larvae relative to crop phenological development (Table 10.3). Assuming a density of 10 plants m~2 and 5 fruits/plant, thresholds are 4-20 larvae per 100 plants, 5-25 larvae per 100 plants, 10-20 larvae per 100 plants and 50-200 larvae per 100 plants in the USA, the Commonwealth of Independent States (former USSR), Australia and Brazil respectively.

The use of thresholds has not prevented the need for insecticide resistance management strategies in some countries to extend the useful life of some chemical insecticides. The most successful has been the pyrethroid resistance management strategy adopted in Australia (see Case Study in Chapter 4, p. 110) which confined the use of pyrethroids to a defined period each year on all crops (Forrester et al., 1993) in order to reduce the selection pressure on the major pest Helicoverpa armigera. While this strategy enabled farmers to continue growing cotton, an alternative IPM programme is urgently needed and research on trans-genic cotton and other tactics is currently being developed (Matthews, 1997b).

The transgenic cotton Bollgard® containing a Bt gene was first commercialized in the USA in 1996 on 729,000 ha which then increased to just over one million hectares a year later, followed in 1998 by planting of around 2 million hectares (Merritt, 1998). During 1998 China planted 53,00 ha, Mexico 40,500 ha, Australia 81,000 ha and South Africa 12,000 ha. In a three year study in Mississippi, USA where Bt cotton was compared with conventional cotton, Bt cotton received on average 6.7 sprays compared with 11.7 sprays on conventional fields (Stewart et al., 1998). The Bt cotton costs, at $61.48 per acre (which includes a $32 technology), were lower than the insecticide costs of $68.15 per acre in conventional cotton. In addition, in some Bt cotton crops higher levels of beneficial insects were found than in conventional crops (Stewart et al., 1998) while in others no differences have been demonstrated (Van Tol and Lentz, 1998).

Natural enemy populations have also been shown to be enhanced when mating disruption is used for the control of P. gossypiella (Campion et al., 1989). Today, almost all pink bollworm management is achieved by using the synthetic pheromone Gossyplure (Howse et al., 1998). With the development of Gossyplure combined with appropriate slow release formulations (Campion et al., 1989; Critchley et al., 1989) cotton growing countries affected by pink bollworm seized eagerly on the new mating disruption strategy. The Egyptian Ministry of Agriculture, more than any other, has made a significant contribution to the use of pheromones for pink

Table 10.3. Economic thresholds for cotton pests in Brazil (from Ramalho, 1994).

Species

Critical period (days)

Economic threshold

Sampling site

Aphis gossypii 10-60 71% plants infested All of plant

Frankliniella sp. 10-25 6 nymphs and/or adults per plant All of plant

Anthomonasgrandis 40-120 10% infested squares Medium squares - upper i of plant

Heliothis virescens 50-60 40% infested squares by larvae 1/3 grown squares from upper i of plant

Heliothis virescens 80-120 10% infested squares by larvae 1/3 grown squares from upper i of plant

Pectinophora gossypiella

75-120 1 1% hard bolls damaged

First hard bolls from top to plant bottom bollworm control with 300,000 acres of cotton (36% of the country's cotton) treated in 1994.

The use of mating disruption as a control technique tended also to reduce the severity of outbreaks of secondary pests, probably due to the conservation of natural enemy populations. However, it is unlikely that natural enemy populations can control all secondary pests; this is certainly thought to be the case of Helicoverpa spp. (Titmarsh, 1992). Augmentative control of Helicoverpa using the egg parasitoid Trichogramma spp. suggests that the approach is not economically viable in US cotton systems (King and Coleman, 1989). However, in Usbekistan where there are now over 700 factories producing Trichogramma pinoi and to a lesser extent Bracon herbetos, the approach is considered a highly effective means of controlling H. armigera and Agrotis moths (Matthews 1997b). Other natural enemies such as the entomopathogens, Bt and NPVs have been used against H. virescens and H. zea and although they are not widely applied, interest is increasing (Luttrell, 1994).

A variety of other control techniques including lure and kill for the boll weevil A. grandis (McKibben et al, 1990; Smith et al., 1994), host plant resistance obtained through various morphological and biochemical traits (Fisbie et al., 1989; King and Phillips, 1983) and cultural controls such as destruction of stubble to reduce over-wintering pupal populations of H. armigera (Fitt and Daly, 1990) are used in different parts of the world. The impacts of a variety of control measures have been incorporated into complex management systems utilizing simulation models and expert systems (Luttrell, 1994) but in most situations simple decision rules have been devised to integrate measures (Fig. 10.2). In Asia, a farmer participatory approach to cotton IPM has been tested (Anon., 1999). An IPM approach evaluated at Wanshong, China included a range of control measures (Box 10.1) that provide cotton yields comparable with conventional farmer practice and reduced the number of insecticide applications necessary by 29%. This reduced total input costs leading to profits 7.5% higher per hectare than was achieved

Weeks 0

0 0

Post a comment