Issue
Agron. Sustain. Dev.
Volume 28, Number 1, January-March 2008
Page(s) 113 - 137
DOI http://dx.doi.org/10.1051/agro:2007042
Published online 18 December 2007
Agron. Sustain. Dev. 28 (2008) 113-137
DOI: 10.1051/agro:2007042

Sustainable pest management for cotton production. A review

Jean-Philippe Deguine1, Pierre Ferron2 and Derek Russell3

1  CIRAD, UMR PVBMT, Saint-Pierre, 97410 La Réunion, France
2  INRA (retired), Centre de recherche de Montpellier, 2 place Viala, 34060 Montpellier Cedex 1, France
3  Natural Resources Institute, University of Greenwich, ME44TB, UK

(Accepted 4 September 2007; published online 18 December 2007)

Abstract - Cotton cultivation, often highlighted for its excessive consumption of plant protection products, is taken as a model to illustrate the development of the ideas and practices of crop protection over the last 50 years. Cotton is grown in 69 countries on 30-35 million hectares and the production exceeded 20 million tones of lint in recent years. Despite the continual improvement in the performance of chemical control strategies, harvest losses remain very high, of about 30%. The largest consumer of pesticides in the world, the cotton production system has the advantage of having been an experimental model for many crop protection programmes under various agronomic conditions and in the presence of diverse pest complexes. Without attempting an exhaustive bibliography, this review explores how and why the ideas underlying crop protection have significantly evolved since the advent of synthetic pesticides. After a spectacular demonstration of yield growth through the application of chemical control, cotton production was rapidly confronted by the secondary effects of this control. These included the appearance of evolved insecticide resistance and the appearance of new damage caused by pests considered up to then as of only secondary importance. In extreme cases, the economic viability of the production systems themselves have been compromised following increases in the application rate and frequency of insecticidal treatments. In general, harvest losses have remained high despite the constantly improving technical performance of pest control chemicals. Two models of the future of crop protection can be drawn: total pest management which involves the eradication of pests, and integrated pest management (IPM), which aims at the management of pest populations below economic thresholds by a mixture of chemical control and a suite of alternative control measures. The first method, total pest management is limited in agricultural systems to particular cases in which the pest in question has no significant alternate hosts in the vicinity of the crop system. On the other hand, the application of IPM is constrained both by the difficulties in exploiting the concept of an `intervention threshold' and by the limitations of many of the specific non-chemical techniques proposed, but does have the advantage of taking into consideration the full pest complex in a cropping system. In practice, it has been a calendar schedule, largely of insecticidal treatments, established on the basis of earlier local observations which has been most widely adopted by growers. This strategy has produced significant improvements in production in the cotton producing countries of francophone Africa and elsewhere. This has led to area-wide integrated pest management which takes into account the potential for natural factors to regulate populations in a specific region. In cotton production, biological control by introduction and acclimation of beneficial arthropods has not been notably successful because of the difficulty of developing a suite of beneficial organisms capable of responding effectively to the diversity of pests in the system, the annual nature of the crop, and the disrupting effects of chemical control measures directed against the remaining pests. Only inundative biological control has had significant success and then in particular cases where the pressure of chemical insecticides has been reduced. More benefit is to be obtained from the active conservation of the indigenous fauna of beneficial organisms. In spite of an increased general environmental awareness, in practice it has been the growth of evolved resistance to pesticides which has had the dominant role in constraining the growers to a more rational use of control strategies. These can be illustrated by the development of window strategies for control measures across the growing season, initially in Australia. The reduction in chemical control treatments made possible by the efficacy of genetically modified cotton has shown the positive role that indigenous natural enemies can play. At the same time, however, there has been a growth in the importance of pest species which are unaffected by Bt toxins. For example, the sucking pests are progressively coming to displace the vegetative and fruit feeding caterpillars as key pests of Bt cotton. Taking into account the spatio-temporal dimension of natural population regulatory factors has led to changes in agricultural practices and production systems. In cotton, for example, production systems maintaining permanent ground cover, are having increasing success. Intercropping and trap cropping have been favourable to the maintenance of beneficial arthropod complexes and unfavourable to the growth of pest populations. This new design context for crop protection in general and for cotton in particular, in applying the principles of agroecology, moves towards the concept of a truly sustainable agriculture. This implies a change of strategy towards a total systems approach to sustainable pest management, characterised by a movement from a paradigm of pest control field-by-field, through farm-by-farm and agroecosystem-by-agroecosystem, to a landscape by landscape approach.


Key words: agroecology / area-wide pest management / biotech cotton / chemical control / conservation biological control / crop losses / eradication  / farmscaping / integrated pest management / pesticide resistance / total pest management

Corresponding author: jean-philippe.deguine@cirad.fr

© INRA, EDP Sciences 2007