5 Directions for future research

Agroforestry entomology is an important area of research, but there is little information available at present about the kinetics of insects in agroforestry systems. Each agroforestry system has its own insect problems, which may be very different from those of other systems. Therefore, each system has to be studied separately.

• Identification of the insect pests in an agroforestry system would be the first step in initiating entomological studies. Of the many insects visiting the system, the injurious insects should then be identified after ascertaining the type and extent of injury inflicted by them. This exercise must be carried out for each plant species in each component of the system. The result will be a summary of the entire pest complex in the system. The same set of observations may be utilized to identify the host range of pests within the system and the common insect pests on component plants. These observations should be maintained at fixed intervals, preferably weekly, during all seasons for two years.
• The pest complex in an agroforestry system may change with the age of the system. Therefore, continuous observations should be maintained to record the changes, if any. This may be accomplished through frequent entomological surveys and through collaboration with national, regional and international organizations and networks in the countries where studies are proposed. The insects collected from each agroforestry system in different ecoregions may be maintained at the National Museums of Kenya, ICRAF Headquarters or at any other suitable place, for reference. All stages of larval instar should be displayed with the adult stage to facilitate identification. The collections should be identified by authentic identification services.
• The key species in a pest complex must be identified for further detailed studies. Studies of the life histories of these pests need to cover seasonal incidence, peak period of activity, mode of attack, life cycle, the vulnerable stages, predators and parasites on individual insect species with the stage attacked, etc. Such studies involve both field and laboratory research. Observations should be made on the predisposing factors affecting the population build-up of pests, for example, temperature, relative humidity and sunlight. Such studies would yield data to be used for pest forecasting.
• Estimations of losses caused by insect pests on individual plant species and on the system as a whole need to be made. Whereas the losses to crops can be estimated using the standard methods employed for agricultural crops, methods of estimating losses caused by insect pests on woody perennials may have to be devised. This is because the woody plants play a different role in agroforestry than in their natural stands. For example, in an alley-cropping system for soil-fertility enhancement, the leaves of woody plants are important as they contribute to soil enrichment. Under natural conditions, or in forestry, a small loss of leaves can be ignored if growth is not affected, but in an alley-cropping system even small amounts of leaf loss are significant. Each agroforestry system may attach a different value to its component plants and therefore the method of loss estimation may be different for each system. In addition to economic criteria, the sustainability aspect should also be taken into account in loss estimations.
• Information about agroforestry insect pests may be presented in the form of an identification manual containing figures or photographs of these pests together with information about their mode of injury, distribution, host range, life cycle, natural enemies, management practices, if developed, and other pertinent information.
• Agroforestry systems have a great potential for the establishment of biocontrol agents. After obtaining preliminary information about the natural enemies of the pests in the system, further studies on biological control may be taken up in association with CAB, the International Institute of Biological Control (IIBC) and other networks in the region. Identification of multipurpose trees and shrubs that favour the population build-up of natural enemies through provision of food (pollen, nectar, etc.) could be initiated at the same time. The constraints to the establishment of natural enemies also need to be identified. In Machakos, for example, the ant, Pheidole megacephaly, which attend coccaids, provides them protection from predators and parasites. Removal of the ants may result in population build-up of the natural enemies and thus in checking the coccoid. Microbial agents that check pest populations may be used to advantage in pest management. Once established, these agents provide cheap long-term protection against the host insects. Trials should be initiated to test the efficacy of the microbial agents against the major pests in agroforestry systems.
• The pollinators of the insect-pollinated plants in an agroforestry system need to be identified. This is essential so that the layperson can differentiate between useful and harmful insects. The knowledge of pollinator insects is especially important in the case of MPTS that are to be transported to distant sites from their native locations. Failure of fruit setting in such plants may be associated with a lack of pollinators in the new site.
• An important area of research in agroforestry entomology will be to study the effect of interactions among components on insect dynamics. A comparison of insect abundance on a plant species in monoculture and in agroforestry will provide first-hand information about the effect of co-cultivation on insect dynamics. The mechanisms that govern the level of pest populations in agroforestry are central to understanding the dynamics of insect activities in these systems. Therefore, the variables that determine insect abundance in agroforestry need to be studied. The combined effect of all variables acting independently or with other variables also needs to be determined. The results will indicate the advantages or otherwise of agroforestry practices in pest management.
• A woody plant species being considered for agroforestry must be studied for the insects associated with it in its natural habitat, especially if it is intended to introduce it in new areas. The knowledge of the pests and their natural enemies thus acquired will be of immense importance should some of the pests travel to the new area and establish there in the absence of natural enemies. In such circumstances, the natural enemies of the pest from the native areas may be introduced to the new location to establish themselves on the host insect. The case of the leucaena psyllid illustrates the need to carry out such studies in respect of each potential multipurpose tree and shrub species.
• A search should be initiated to identify MPTS associated with small pest complexes. Such plant species should be given priority consideration for agroforestry to reduce the risk of pest attack. Pest-tolerant provenances may be identified from among the tree and shrub species already used in agroforestry. These can also be used for tree-improvement programmes.
• Tree-improvement programmes should include insect resistance as one of their objectives. In the long term, high-yielding species or provenances may not provide as much output from the system if they are more insect prone than species with lower levels of productivity but higher insect resistance.
• There is a need to study various tree–crop combinations in different agroforestry systems for their insect-reduction capabilities. A comparison of the prevalence and extent of injury due to an insect pest on a particular host species when raised with other plant species will indicate the combinations that hinder the activity of that insect pest. Simultaneous observations should be made for all the major pests on a host species in different combinations. There could be combinations that favour some pests and inhibit others. Information from such studies would help in choosing optimal tree–crop combinations.
• The effect of different temporal and spatial arrangements of crops and woody perennials on the insect pests in an agroforestry system must also be studied. Thus, woody perennial–crop combinations and arrangements that support the minimum pest populations may be identified. Such designs should be tested at many locations before being approved for a particular agroforestry system.
• The plant species considered for agroforestry should be studied for their allelopathic characters and their ability to repel or attract major insect-pest species in the system. Inclusion of the species that repel the pests of other component species should be encouraged.
• Plant species that act as collateral or refuge hosts for the pests of crops or vectors of plant or human disease need to be identified. It may be desirable to replace such species with other suitable species.
• To develop effective pest-management programmes in agroforestry, agronomic and cultural practices that help reduce the major pest populations need to be developed. The practices that exacerbate pest injury also need to be studied so that these practices can be discouraged or discontinued, or measures can be developed to counter the ill effects of these practices.
• Integrated pest-management strategies must be devised. An ideal pest-management programme for agroforestry should entail no pestcides, or minimum use of them, but maintain pest populations below threshold levels (to be determined for each system) by integration of physical, cultural and biological control measures.
• The relevant information on insect pests and their natural enemies should be stored in a database and made available to practitioners of agroforestry. New information may be added to it periodically. Based on the knowledge thus acquired, it would then be possible to start a forecasting service on insect pests in agroforestry, in association with other agencies.