To protect their cotton crops, producers in Mali primarily use chemical insecticides, which they generally apply five or six times at regular intervals (every fortnight) from 45 days after emergence (DAE). However, some producers adopt other approaches - staggered targeted control (since 1994) or threshold-based interventions (since 2001) - that enable them to make insecticide savings of 36% and 72% respectively for the two approaches (Renou et al., 2012). Reducing insecticide use on cotton crops in Mali will therefore centre on the dissemination of these new approaches and more particularly that of threshold-based interventions. Progress is currently slow, since it requires prior training and a commitment on the part of producers. It is therefore necessary to simplify and improve the rules relating to threshold-based interventions, without losing efficacy and cost-effectiveness, to allow producers to understand the approach better, and to speed up training programmes.

Project: PASE II

Line 2: sustainable intensification of cropping systems

Activity 5: integrated cotton crop protection strategies

Researchers in Mali have already developed cotton crop protection strategies that do not involve chemical insecticides (Renou et al., 2009; 2010 and 2011). Those strategies, which also boost production and revenues, centre on a high planting density (double the rate currently recommended) and on topping cotton plants, which has been proved to reduce bollworm numbers (Renou et al., 2011). However, to be on the safe side, particularly during the period before cotton plant topping, these protection strategies have always included the possibility of threshold-based interventions against this pest, which is of great importance in Mali (Cabanilla et al., 2005), although this has never proved necessary. Given the current reluctance on the part of growers to practise high planting densities, any improvements to cotton protection techniques in Mali will probably have to be based on topping cotton plants and on threshold-based interventions against bollworms. However, topping every plant in a plot is time-consuming, and growers may not adopt the practice, although they have sometimes done so in the past. Luckily, since 2009, topping has been shown to have new advantages (Téréta, personal communication) that may lead to its adoption by prducers. It has been shown that the effects of topping certain plants extend to neighbouring plants that are not topped, probably by triggering the emission of volatile compounds (Marchand, 2012).

Experiments at Farako station

The overall aim of the studies done at the Farako station in 2014 was to determine, the implementation conditions for threshold-based interventions and cotton plant topping, so as to ensure optimum efficacy against bollworms and cost-effectiveness. The specific aims set for each study were to:

• determine the minimum number of plants to be topped within a plot without losing efficacy against bollworms
• determine how, in practice, to implement an 80% reduction in the numbers of plants to be topped
• assess the possible influence of the cultivar on the effects of topping on both topped plants and their neighbours
• determine at what time effects are transmitted from a topped plant to its non-topped neighbour
• assess the merits of synchronous threshold-based interventions as opposed to the asynchronous interventions practised by growers.

Main results:

Concerning topping cotton plants, the studies: (i) confirmed the possibility of reducing the number of plants to be topped in a plot by 80%, by proceeding (ii) in a regular fashion in all the rows rather than one row in five, with the prospect of (iii) rendering four calendar-based insecticide treatments unnecessary, (iv) revealed a varietal influence on transmission of the effects of topping to neighbouring plants that (v) nevertheless does not compromise the future of topping, (vi) showed that transmission of the effects of topping to neighbouring plants occurred during the first two days after topping, although it was apparently (vii) incomplete, and (viii) confirmed the reduction in plant-eating caterpillar numbers after topping, particularly those of Haritalodes derogata, which is not a moth although it behaves nocturnally. As regards threshold-based  interventions against bollworms, (i) the studies did not reveal any significant difference between synchronous and asynchronous interventions, and (ii) simpler rules based on a percentage of plants infested by the pest and a percentage of plants with pest damage were drafted.

On-farm trials: villages of Benguéné, Ziguéna and Nafégué

It would be wise to steer the practises of many cotton producers towards these new approaches that protect the environment and human health and are also more sustainable. The main aim of these on-farm trials is not solely to check the merits of these new protection approaches but also to assess them comparatively in conjunction with producers, to ensure that they adopt them either partly or in full, with or without particular adjustments. The secondary aim was to provide producers with knowledge they may not previously have had, to allow them to change their current protection practices.

Main results:
Adapting plots and fertilization levels had almost no significant effect on bollworm, jassid and whitefly populations. Compared to calendar-based insecticide treatments, threshold-based interventions against bollworms showed the same biological weaknesses as at research stations, albeit much less frequently. Not respecting decision-making rules was probably to blame, but it reduced the cost-effectiveness of this protection programme, which is one of its main advantages on research stations and farms. The favourable welcome from producers warrants continuing the study, albeit with improvements (type of observations and decision-making rules). Topping cotton plants confirmed its ability to reduce bollworm populations, but a little less so than on research stations, due to insecticide protection. The extension of those effects to neighbouring, non-topped cotton plants and the favourable welcome from producers warrant continuing the study of this practice, making the most of the progress made on research stations as regards implementation. Lastly, treatment practices among producers are generally satisfactory, although (i) improvements are required (eg managing pyrethrinoid-resistant H. armigera populations), (ii) certain aspects should be explored further (eg control of the quantities applied) and (iii) understanding of certain practices could be improved (eg the non-constant interval between treatments and the very early start to protection, particularly on late-sowed crops).