Towards an Integrated Pest Management for Sustainable Cocoa Production in Ghana

Beatrice Padi and G. K. Owusu Cocoa Research Institute of Ghana
P. O. Box 8
Tafo-Akim, Ghana

SUMMARY

Cocoa is perhaps the most important export crop for countries in the forest zone of West Africa and is the main source of foreign exchange for some of them, including Ghana. The cocoa industry in West Africa is dominated by a large number of peasant farmers who cultivate small farms of about 0.5 to 5.0 hectares and who lack the resources to expand or improve their farms. This and other factors combine to plague the industry with serious problems leading to dwindling production levels.

This paper highlights on the factors that limit cocoa production in West Africa, with special emphasis on Ghana, and reviews past and present efforts at solving them. The major constraint identified is the low standard of farm maintenance, especially the control of pests and diseases largely because of the low producer prices which do not allow the procurement of the necessary inputs. Other factors include the virtual absence of credit or loan facilities for farm improvement, the rapidly dwindling forests and deteriorating environmental conditions, and the generally poor social and economic circumstances of the rural communities in which farmers live. Future projections are made on integrated pest and disease management practices that will minimize the use of environmentally hazardous pesticides, and sound socio-economic considerations that could promote sustainable cocoa production in the West African sub-region.

INTRODUCTION

In the West African sub-region, cocoa is an important export crop in Ghana, Cote d'Ivoire, Nigeria, Cameroun, Togo and Sierra Leone. In 1992, 59.9 % of the world's cocoa was produced by Africa, 25.5% by Latin America and the Caribbean and the remaining 14.6% by Asia and Oceania (Mossu, 1992). With current annual production of over one million tonnes in Cote d'Ivoire and about 350,000 tonnes in Ghana, the two West African countries account for over 40% of world supply. In Ghana cocoa occupies a key position in terms of foreign exchange revenues and domestic incomes, as well as being the major source of revenue for the provision of socio-economic infra-structure (Anon, 1995).

The majority of cocoa farms in West Africa are small holdings owned by a large number of peasant farmers. For example, in Ghana, about 66% of farms are within the size range of 0-8 ha owned by 332,244 peasant farmers, with only 18.9% of the farms larger than 20 ha (Table 1). (Cocoa Services Division, unpublished data). For 66 years (1910 to 1977), Ghana retained world's leadership in cocoa production with market shares ranging from 30-40% of the world's total production (Bateman, 1988). Production peaked at 571,000 tonnes in 1964/65, but has since then fluctuated over the years between 150,000 and 350,000 tonnes per annum with Ghana losing her first position to Cote d'Ivoire (Gill and Duffus, 1989). The situation in Nigeria has been quite similar: yield in 1964/65 was 298,000 tonnes compared to the present annual yield of 110,000 - 150,000 tonnes (Are, 1970; Adeyimi, in press). The decline in production in Ghana in recent times has raised a great deal of concern and in 1995, a high-powered committee chaired by a Deputy Minister and comprising executives of the Ghana Cocoa Board, Ministries of Agriculture and Finance and other high-ranking government officials and farmers' representatives, was charged with conducting a rapid rural appraisal of the cocoa industry to identify the constraints to production and recommend measures for solving the problems with a view to arresting the decline in production (Anon. 1995). General and more localized studies have identified several technical factors which have contributed to the dwindling cocoa production levels in Ghana, Nigeria and other West Africa countries (Adeyimi, in press, Ollennu et al., 1989; Anon, 1990; Osei, 1993; Anon, 1995; Freud et al., 1996). Paramount among these are the ravages caused by cocoa capsids (Heteroptera: Miridae) and diseases such as swollen shoot caused by cocoa swollen shoot virus and black pod caused by the fungi Phytophthora palmivora and P. megakarya.

This paper highlights on the factors relating to poor pest and disease control and farm husbandry practices that limit cocoa production in West Africa, with special emphasis on Ghana, and reviews past and present efforts at solving them. Future projections are made on integrated pest management practices and sound socio-economic considerations for sustainable cocoa production in the sub-region.

FACTORS LIMITING COCOA PRODUCTION IN WEST AFRICA AND EFFORTS TO ADDRESS THEM.

1. Pest and disease problems:

a) Capsid and other pests:

Capsids, also known as mirids (Heteroptera: Miridae), are the most economically important insect pests of cocoa in West Africa. In Ghana, cocoa mirids have been known as serious pests since 1908 (Dungeon, 1910) and because of their devastating effects, the local farmers called them "Sankonuabe" which literally means "go back to the planting of oil palm", reflecting the situation before the introduction of cocoa (Johnson, 1962). Estimation of crop losses due to capsids are always complicated by the inadequacy of records and the complexity of losses from other causes such as fungal and viral diseases and drought. Crop losses in cocoa due to capsids and mealybugs have been estimated at 25-30% er annum (Wills, 1962). In 1957, crop loss in Ghana attributed to capsid damage alone was estimated at 60,000 to 80,000 tonnes of dry cocoa (i.e. about 25%) (Stapley and Hammond, 1959). Yield loss may be as high as 75% in cocoa farms attacked by capsids and left unattended for a period of over three years (Anon, 1951). Recent studies indicate that about 25-30 % of the national cocoa acreage in Ghana has significant mirid damage, with an annual crop loss of about 100,000 tonnes (Owusu-Manu, unpublished information).

Of the five mirid genera, viz. Sahlbergella, Distantiella, Bryocoropsis, Odoniiella and Helopeltis which attack cocoa in the Ethiopian region, the first four are endemic to and occur only in West and Central Africa. The most widespread and generally economically important mirid in Africa is Sahlbergella singularis Hagl. but in Ghana Distantiella theobroma (Dist.) is the more common and predominant species. S. singularis attacks cocoa from Sierra Leone in West Africa to the Congo Republic and Central African Republic in the east but probably does not extend beyond 90N, the approximate limit of the cocoa belt. It is also present in the island of Fernando Po, 32 km from the West African mainland. The second most important species, D. theobroma, formed only 5 to 15 per cent of the mirid population but this ratio is now probably lower as evidenced by the predominance of S. singularis since 1995 in several localities in Ghana where D. theobroma used to be the dominant species (Owusu-Manu, 1996; Padi in press). Bryocoropsis laticollis Schum and Helopeltis spp. feed mainly on pods but their feeding punches are superficial and these mirids are, therefore, considered as minor pests (Raw, 1959).

Cocoa mirids occur in low overall densities but often aggregate in "pockets" created by a break in the cocoa shade canopy followed by growth of vegetative chupons and fan tissue. Routine hand collection has given a maximum level of 1000 - 2000 capsids per hectare (Collingwood, 1971 a) and a mean of 6 capsids per 10 trees actually represents a high and damaging population level (Owusu-Manu, 1995). Extensive feeding by capsids on fan branches results in the degradation of the tree canopy, usually of more or less discrete groups of trees, up to around one hundred in number referred to as capsid "pockets". Where the damage covers a large area of exposed cocoa with more diffused damage in terms of fan shoots, death is easily seen in the dry season, a condition referred to as "capsid blast". If the more persistently attacked trees have the crown extensively reduced by die-back, the condition is described as "stag headed". Normal flowering and pod setting are severely affected as a result of the previous season's attack. In extreme cases, whole trees become unproductive and die. Seedlings may completely fail to become established and, if they are not killed, capsid attack delays their coming into bearing by several years. Mirids on pods feed largely in the parenchymatous husk tissue but the depth of stylet penetration is not known. Cherelles may wilt and pods less than three months old have very little chance of surviving, usually dying from mirid damage or from fungi entering the pods through the lesions (Gerard, 1968). In West Africa crop losses resulting from this are believed to be negligible because the tree compensates by shedding fewer fruits by cherrelle wilt. Mature pods seldom seem directly affected. A comparison between ripe pods that had been heavily attacked (more than half the surface blackened by feeding) by S. singularis in Nigeria and clean pods revealed no significant differences in dimensions, pod weight, number of beans or weight of peeled beans. Damage is most impressive after mirid peak numbers have passed (Gerard, 1968).

The female adults lay their eggs in the outer layer of pods and beneath the bark of young shoots. There are five nymphal stages, and the incubation period varies from 12-17 days and each nymphal stage may last from 3 to 6 days (Cotterell, 1926). The adults and all the five nymphal stages feed on cocoa pods, chupons and fans.

Feeding by cocoa mirids is characterized by dark markings known as "lesions", on both pods and shoots, which result from the collapse of plant tissue caused by the toxic saliva (Entwistle, 1965). Secondary damage characterized by canker and die-back occurs when the feeding lesions are invaded by parasitic fungi notably Calonectria rigidiuscula Berk. and Br. (Sacc.), the perfect stage of Fusarium decemcellulare, resulting in the death of the shoots and branches (Cotterell, 1926; Crowdy, 1947;). Estimates of the extent of infection of lesions by Calonectria in Ghana vary from 80 % to 1-15% (Usher, 1957) while in Nigeria an overall estimate of 95% has been given (Kay, 1957).

Climatic factors, mainly light and humidity, are believed to influence the abundance of cocoa capsids. It has been suggested that humidity has a direct influence on population changes in West Africa and there is no doubt that predators and parasites together have some effect (Gibbs et. al., 1968). The preference of S. singularis fifth instar nymphs in laboratory experiments is for high humidity (90-95%) and a temperature (in saturated air) between 18.5 and 23.50C (Prins, 1965; Youdeowei, 1964). Complex choice experiments demonstrated the dominance of light response over that of humidity, dark situations being chosen irrespective of humidity levels (Madge & Gerard, 1967). In view of the susceptibility of mirid nymphs to desiccation, this is a striking result and may well indicate that concealment from predators is of more importance than protection from desiccation (Entwistle, 1972). Its bearing on aggregation of mirids on the tree is obvious but its significance, if any, to inter-tree distribution is uncertain. Mirids are particularly night-feeders, but in dull weather, they may be seen feeding during the day (Van Hall, 1932).

Although there have been previous studies to identify the indigenous natural enemies of cocoa capsids in West Africa as reviewed by Padi (1997 a), there have been no attempts at the biological control of capsids. Recorded predators include the ant Oecophylla longinoda, salticid spiders, reduviids, mantids, Grillidae and Pentatomidae. Parasitoids include Euphorus sahlbergellae Wlk., E. helopeltis, E. anates and Encyrtus cotterelli Watson Percentage parasitism has been generally lower in D. theobroma (up to 4%) than in S. singularis (20.0 - 59.5% ) Pathogen recorded include unidentified fungi inperfecta in Ghana and Nigeria and a white unidentified fungus which attack S. singularis. Beauvaria bassiana been reported to be lethal to Helopeltis theobromae in Malaysia but has not been recorded in West Africa. Other pathogens are Bacillus and Aspergillus species. Thus, there is a potential for the biological control of cocoa mirids in West Africa and studies have been initiated by CRIG, in collaboration with British and American institutions, to re-evaluate mirid natural enemies in Ghana for biological control purposes Padi, et al., in press).

S. singularis has seventeen indigenous alternative host plants all related to cocoa and in the order Malvales (Padi et al., 1996). These include six Cola and three Theobroma species (Table 2). For D. theobroma, only three alternative host plants, Adansonia digitata, Ceiba pentandra and Citrus spp., have been recorded. It is believed that these indigenous plants, some of which commonly grow beside cocoa, served and continue to serve as a source of mirid infestation onto cocoa which is presumably better preferred (Padi & Hollander, 1996).

In addition to capsids, termites have in recent times been of economic importance in some parts of Ghana (Ackonor, 1994). Sporadic attacks by other insect pests of minor importance also occur, causing varying degrees of damage. These insects include the pod feeders Bathycoelia thalassina (H.-S.) (Heteroptera: Pentatomidae) and Pseudotheraptus devastans (Dist.) (Heteroptera: Coreidae) (Owusu-Manu, 1971; Padi, et al., in press), the stem borer Eulophonotus mermeleon Fldr. (Lepidoptera: Cossidae) (Anon, 1995) and the pod borer Characoma stictograpta Hmps. (Lepidoptera: Noctuidae) (Akotoye, 1975). Anon (1995) observed that damage caused by termites and stem borers is becoming increasingly important in the whole country and recommended that CRIG must intensify research for a quick solution to these problems. Other insect pests which are particularly important on young cocoa include the Lepidopteran foliage pests Earias biplaga Wlk., Anomis leona Schaus., Orgyia basalis Wlk. and Prodenia litura F., the psyllid Tyora tessmanni (Aulm.), the aphid Toxoptera aurantii (Fonsc.), the thrip Selenothrips rubrocinctus (Giard.) the stem tip feeder Tragocephala spp. (Coleoptera: Cerambycidae) and Empoasca devastans Dist. (Homoptera: Fulgoroidae) which causes the tip burn disease (Collingwood & Marchart, 1971; Entwistle, 1972).

Damage caused by rodents (rats and squirrels) was estimated at 10% of the crop in Nigeria (Everard, 1968) whilst in Ghana an average loss of 2.27% was estimated (Taylor, 1961). In Sierra Leone, losses of 20 to 50% have been attributed to joint squirrel and monkey damage (Urquhart, 1955). Woodpeckers and parrots have been reported to attack cocoa in some parts of the world, e.g. Brazil (Bondar, 1939) and Jamaica (Anon, 1918) but there are no documented records of damage in Ghana although these birds occur in cocoa farms.

Since 1957 when Lindane was recommended, spraying with synthetic insecticides has been the only effective method for controlling capsids on cocoa in West Africa. Presently, spray treatment with Gammalin 20 (Lindane) at 280g a.i./ ha or 1.4 litres/ ha and Unden 20 (Propoxur) at 210g a.i./ ha or 1.1 litres/ha applied at monthly intervals from August to December, omitting November, is the only protection measure recommended in Ghana. The two insecticides are alternated every two years so as to extend their useful life time by reducing the possibility of development of resistance to either chemical. The treatment period coincides with the capsid population peak period which is linked with the peak cropping period (Collingwood, 1971 b).

New chemicals continue to be screened and, presently, Carbamult, a carbamate, and Actellic/ Talstar, a cocktail of the organophosphate Actellic with the pyrethroid Tarstar (Bifenthrin), have been passed by the Cocoa Research Institute of Ghana (CRIG) (Owusu-Manu, 1996 & 1997) and recommended to the Ghana Cocoa Board for use by farmers. Endosulfan (Thiodan 35 EC) has passed experimental tests but has not been passed by CRIG. Because cocoa capsids are cryptic and occur at very low population densities, they are best controlled by persistent and systemic insecticides. However, the adverse side effects of persistent chemicals, including the destruction of non-target beneficial insects and their long lasting residual effects in the environment, make them undesirable and there are moves to replace Lindane which has been used for mirid control in Ghana since 1957.

The standard method of pesticide application on mature cocoa in Ghana is by motorized knapsack mistblowers. Two methods of application are usually used. These are either to cover a tree thoroughly on one side from the trunk into the canopy (T1) or to cover both sides with the insecticide (T2) (Peterson et. al., 1963). In neighbouring countries of Nigeria, Cote d'lvoire and Cameroun, emphasis is placed more on the hydraulic compression sprayer and other smaller hand-operated types. In the Cameroun, a hand-held fogging device (the swing fog) is used as the standard method and good mirid control is claimed (Bruneau de Mire, 1966). This method of application has the advantage of covering large areas quickly but all treatments should be applied in the early morning when the smoke is carried up evenly and held in the canopy for a sufficient length of time to give an effective kill. For the last reason, it was not recommended in Ghana. Presently, another fogging technique, the FUMIVAP which has been recommended in Cote d'Ivoire is being tested in Ghana (Padi & Sarfo, in press).

In Ghana, the period between 1959 and 1962 witnessed mass spraying exercise by which cocoa farms throughout the country were sprayed free-of-charge by government Cocoa Extension Service. It is believed that the high production of over 751,000 tonnes achieved in the 1964/65 season was partly the result of the mass spraying exercise. However, it was not possible for government to sustain this exercise and this has resulted in fluctuations in production levels.

Presently, there are no specific recommended control measures for containing the minor pests on cocoa in Ghana although, in the past, treatment was by stem painting or spraying with systemic insecticides like Monocrotophos (Azodrin) and Dicrotophos (Bidrin) and persistent organochlorine insecticides like DDT once a month for the first three years of establishment (Collingwood & Marchart, 1970). These chemicals and are now banned universally. Young cocoa is, therefore currently subjected to the same capsid control treatment as mature cocoa, i.e. spraying with Lindane or Propoxur four times in a year, but using pneumatic instead of motorized knapsack sprayers. The general observation is that this treatment does not adequately control the insect pest complex that cause severe damage to young cocoa. The present insecticide treatment for termites (Ackonor, 1995) lacks a conclusively determined treatment schedule as it is based on limited field studies but has been recommended to farmers because of serious termite damage in some places (Ackonor, 1995). Basically, treatment is with 200 to 400 ml of 0.4% Dursban (Chlorpyrifos) solution poured slowly to the base of each transplanted seedling. This is equivalent to 480-960g a.i./ha of Chlorpyrifos. At the nursery level, prophylactic treatment against termites is by drenching the nursery ground with Dursban before placing the bagged nursed plants on the soaked ground (Ackonor, 1995). The environmental hazards and high cost associated with these methods of treatment are quite obvious and studies are in progress to improve upon the methods.

b) Cocoa Swollen Shoot Virus Disease

The control measures against cocoa swollen shoot virus disease (CSSVD) involve the eradication ("cutting out") of visibly infected trees and their symptomless immediate neighbours, followed by the replanting of areas with good soils. In Ghana, The eradication is done by trained staff of the Cocoa Extension Service. Details of the present control procedures have been given by Ampofo (1989) and Anon (1995). The eradication programme has encountered serious problems which largely relate to the expense and difficulty of operating routine survey and treatment operations on a massive scale and with the required efficiency (Ollennu et al., 1989). Other problems of the eradication programme have been discussed in detail in many reviews (e.g. Legg et al., 1981; Owusu, 1983; Ollennu et al., 1989). Since the eradication was introduced in 1946, about 200 million trees have been removed in Ghana since 1946 and millions more have been killed or seriously debilitated by the disease. Nonetheless, CSSV has not been controlled and is now nearly as prevalent in Ghana as before, particularly in the Eastern Region where the disease is endemic. The situation has, however, recently improved as a result of an on-going project, with external assistance, on the treatment and replanting of a large buffer zone ("cordon sanitaire") between the virus-endemic Eastern Region and the rest of the country. Moreover, the 1995 appraisal committee (Anon, 1995) recommended that in addition to the current CSSVD treatment and replanting in the Cordon Sanitaire which covers only 12% of the total cocoa-growing area of Ghana, a more extensive programme involving the rest of the country should be initiated.

There have been many previous attempts, both chemical (Mapother & Nicol, 1953; Armstrong, 1961; Hannah & Heatherington, 1957; Marchart, 1968; Firempong, 1984) and biological (Anon, 1951, 1952; Decker, 1953, 1955; Donald, 1953, 1955), at controlling the mealybug vectors of the virus, but these studies were discontinued for various reasons (Padi, 1997b). Recently, studies on the possible use of natural enemies have been revived (Ackonor, 1997).

c) Black pod disease

Losses resulting from the cocoa black pod disease in Ghana previously caused only by Phytophthora palmivora in Ghana was estimated at 4.9-19% (Blencove & Wharton, 1961; Dakwa, 1984). Recently, however, an outbreak of a more severe disease caused by P . megakarya first occurred in the north-western part of the cocoa belt (Dakwa, 1987) and since then losses due to black pod disease in this and other areas have increased dramatically to 60-100%. This virulent species of Phytopthhora is advancing steadily to the all-important southwest of the Ghana cocoa belt and near the border with Cote d'Ivoire (Opoku et al., 1997). A specific feature is that the original outbreak occurred in a marginal and drier cocoa area, a clear warning that this disease may cause enormous crop losses when it succeeds in spreading to southwest Ghana and Cote d'Ivoire where rainfall is high and moist conditions prevail.

Present recommendations for the control of P. megakarya involve the spraying to protect apparently healthy pods with copper-based fungicides, with or without metalaxyl, at four-weekly intervals, about 6-7 applications per year during the rainy season (Hislop & Park, 1960). Apart from the concern about the amount of the copper-based fungicides which may end up in the soil, the average Ghanaian cocoa farmer finds the repeated application of fungicides unaffordable and the majority of farmers either do not treat their farms at all, or do only one or two applications per year, thus incurring heavy crop losses every year (Henderson, et al., 1994; Opoku et al., 1997). Studies are, therefore, in progress aimed at reducing the frequency of fungicide application to four or less per year, in the hope of encouraging the farmers to adopt the recommendation.

In a study in Papua New Guinea aimed at determining the effects of various control measures on disease incidence, Konam et al.(In press) concluded that trunk injection with potassium phosphonate significantly reduced Phytophthora pod rot and increased pod yield and these effects were enhanced when combined with removal of pod husks and mummies. The application of potassium phosphonate or phosphorous acid carries little or no danger of residues or contamination of the environment, and the technology is being tested in Ghana with good prospects (Opoku et al., 1998).

Several invertebrates including ants such as Camponotus acvapimensis and Pheidole megacephala, beetles, the pod borer Characoma stictigrapta, as well as infected pods and pod husks have been implicated in the spread of black pod disease (Taylor & Griffin, 1981; Evans, 1973; Okaisabor, 1971), but their control on a large scale has not been attempted. It has been suggested that control of the ant species implicated in the spread of disease could considerably slow down the upward movement of the fungus from the ground or from pods near the ground, thus delaying the development of the epidemic (Okaisabor, 1974).

Cultural control, including frequent harvesting and removal of infected pods has also been recommended to control black pod disease (Asare-Nyako, 1969). However, cultural control is suitable for P. palmivora but not for the more virulent P. megakarya for which cultural should be combined with chemical control. The establishment of ground cover or removal of pod husks and mummies also significantly reduced disease incidence in the absence of ants.

Selection and breeding of cocoa varieties for resistance to black pod appears to be the most practicable approach to the problem. In Ghana, screening methods for the evaluation of cocoa types for resistance to P. palmivora are well documented (e.g. Amponsah and Asare-Nyako, 1973). There have been reports from many cocoa-growing countries of cocoa trees that have some measure of field resistance to P. palmivora (Rocha, 1965). Presently, there is a programme to assess the existing cocoa germplasm in Ghana for resistance or tolerance to P. megakarya (Abdul-Karimu & Bosompem, 1994).

d) Parasitic Mistletoes and Nematodes

In Ghana and elsewhere in Africa, six species of parasitic mistletoes of which Tapinanthus bangwensis is the most widespread, occur on cocoa. Preventive control is achieved by the maintenance of shade but cultural control by pruning is often necessary (Room, 1972; Phillips, 1977; Appiah & Owusu, 1997). The possibility of biological control has also been considered (de Bach, 1964).

Nematodes have long been encountered on cocoa (Whitehead, 1969; Asare-Nyako and Owusu, 1981).

2. Agronomic factors:

Good maintenance practices during the early establishment stage and sound management of the mature cocoa farm, together with socio-economic factors which favour the adoption of these recommended agronomic practices are of great relevance to sustainable cocoa production and have received a great deal of research attention.

a) Shade

The basic package of recommendations at the cocoa establishment phase include the planting of cocoa in association with various food crops, especially plantain, and herbaceous plants, shrubs, and trees such as Gliricidia sepium. These provide temporary or ground shade and also promote weed suppression and soil improvement. The other aspect is the provision of permanent shade and the management of shade levels. Weeding 3-4 times in the year is recommended during the establishment phase before the canopy closes. Weed control with herbicides such as Gramoxone and Roundup has also been recommended. For the efficient application of herbicides, it is recommended that cocoa is planted in lines (CRIG Farmer's Guide

No 3).

Traditionally in West Africa, cocoa shade relates to the density of forest trees left in the field after the initial clearing of the forest. Growing cocoa under shade stems from the belief that cocoa, being a second storey tree, thrives best under heavy forest shade. However, with the exploitation of forest trees for timber and other purposes, it has become necessary to plant alternative fast growing tree species to provide shade. Thus cocoa cultivation is of great importance for the conservation of the forest and associated fauna in Africa. As pointed out by Leston (1960) in relation to the snake fauna recorded in cocoa farms around Tafo, Ghana, the cocoa farm should be regarded ecologically as a type of forest. The earlier view that cocoa thrives best under heavy shade has, however, given way to a prescription of mild shade, for both yield levels and precocity. It has been shown that peaks of flushing and flowering in cocoa coincide with the period of maximum daily sunshine and maximum levels of sugar in cocoa trees (Owusu, et al., 1978, Adomako et al., 1990). Thus the recommendation in Ghana is to reduce overhead shade down to a maximum of 10 large or 15 medium trees per hectare (4 per acre) and there are intensive on-going studies to identify trees suitable for providing both temporary and permanent shade (Osei-Bonsu & Anim-Kwapong, 1997; Osei-Bonsu et al., in press). Permanent shade trees being screened include Terminalia ivoirensis, Ricinodendron leuclotii and Spathodea campanulata. In Cote d'Ivoire, the recommendation is to plant cocoa without shade. A study on the levels of permanent shade in cocoa farms in Ghana and Cote d'Ivoire (Freud, et al., 1996) showed that about 50% of the total cocoa area in both countries was under mild shade whilst an average of about 10% and 35% in Ghana and Cote d'Ivoire, respectively, was under no shade (Table 3). Thus, there is a gradual but sure move towards eliminating shade trees. This, combined with timber-related and other activities, is gradually causing the deterioration of the forest and its rich flora and fauna.

The fact that production levels in Cote d'Ivoire have risen within the past decade has been attributed mainly to the lower level of permanent shade under which cocoa is grown in that country, a more intensive maintenance and in particular higher levels of chemical control against capsids (Freud et al., 1996). While "mild shade" (about 4 large forest trees per acre) represents the norm (over half) the area in both Ghana and Cote d'Ivoire, heavy shade with virtually no fertilizer application still accounts for nearly 30 % of the total cultivation area in Ghana whilst, in Cote d'Ivoire, no shade with a few cases (about 10%) of fertilizer application comprises nearly 40% of the total area (Table 1). Yet another factor is the disproportionate impact of Cocoa Swollen Shoot Virus Disease (CSSVD) which, is lower in Cote d'Ivoire than in Ghana (Frued et al., 1996).

Table 4 presents the results of a regression analysis of agronomic factors affecting cocoa yield in Ghana and Cote d'Ivoire derived from studies by Freud et al. (1996) who extensively discussed the relative importance of the various agronomic practices. In both countries, the study confirmed the positive effects of reducing permanent shade. The polybag method of seedling transplant also appeared to have some positive effects on yield. In addition, weeded fallow, the bare earth seedling transplant, and the use of Amelonado seed type in Cote d'Ivoire had negative effects. The type of food crop used as temporary shade had no effect on yield in Cote d'Ivoire. In Ghana, however, weeded fallow appeared to have a negative effect whilst, contrary to the Ivorian situation, bare earth seedling transplant seemed to have some positive effect. Among the maintenance and management systems, anti-capsid treatments had a strong positive effect whilst weeding seemed to have some positive effect in both countries. The age of trees showed a strong quadratic curve in Cote d'Ivoire whilst it had no clear effects in Ghana. Concerning the agro-climatic zones under cultivation, there were negative effects associated with the Western and Central regions in Ghana whilst the Western region in Cote d'Ivoire had a positive effect. The practice of share cropping had positive effects in both countries.

The above results show that the cultivation of recommended hybrid cocoa should be intensified for higher yields. The results further point to the need to educate farmers on the importance of planting only those hybrids recommended by research. A survey conducted by the Cocoa Services Division in 1996 in Ghana showed that national averages for the distribution of the four types of cocoa commercially grown in the country were 42% Amelonado, 36% Amazon, 19% Hybrid and 3% mixed cocoa (Anon, 1995). Thus yields were found to be generally low, with as many as 64% of the farmers interviewed producing less than 256 kg/ha. It is also important that food crops and other plants to be used as inter-crops are selected with caution. The fact that locality had an effect on yield in the analysis of Freud et al., (1996) implies that there could be variable factors associated with different geographical regions which need to be identified and addressed. Possible factors are differences in soil fertility, the population dynamics of capsids and their damage, the intensity of black pod and swollen shoot diseases, and farmers' understanding and acceptance of recommended technologies (Henderson et al., 1995) all of which require further investigation.

b) Soil fertility:

Although most of the cocoa soils in Ghana have supported cocoa for up to 50 years or more, there have been little or no efforts to replenish their lost nutrients. The fertility of soils under good cocoa plantations with complete canopy formation can be maintained or sustained for a fairly long time due to the ability of cocoa to recycle nutrients back into the soil through leaf fall and decomposition of leaf litter. However, the continuous crop removal through harvested beans for a long time would result in the loss of nutrients from the soils (Ahenkorah & Akrofi, 1971; Appiah et al., 1996; Adeyemi, in press). Current on-going fertilizer verification trials on farms in Ghana which have generated a great deal of farmers' interest (Appiah et al., 1996) should, therefore, be considered as long overdue. The trials have demonstrated significant increases in yield and, thus, additional income for farmers. Consequently, farmers who took part in the trials now able and willing to buy other farm inputs like insecticides and fungicides for pest and disease control.

Results of studies in Nigeria which aimed at determining the effect of NPK fertilizer application and rehabilitation techniques on the growth and development of cocoa seedlings and chupons, showed that cocoa on fertilizer-applied plots performed better than that without fertilizer application, irrespective of the rehabilitation technique used (Adeyimi, in press). Earlier studies in Ghana had shown that it was economic to apply fertilizer to farms which yielded more than 560 dry cocoa/ha (Ahenkorah & Akrofi, 1971).

For environmental reasons, there have been criticisms against the use of inorganic fertilizers and there have been suggestions to promote the use of organic fertilizers like farm yard manure (Appiah, M. R. Personal communication). However, the use of organic fertilizers is uncertain for the following reasons:

i) The sources of organic fertilizers are limited and it will be difficult to get enough to meet farmers' demand.

ii) There is the problem of deposition of heavy metals like zinc, lead and mercury from the use of organic fertilizers.

iii) Organic fertilizers act by releasing inorganic products into the soil, and in this respect are not much different from inorganic fertilizers (Appiah, M. R, personal communication).

3) Socio-economic factors:

It is expected that earnings from cocoa will pay for all operations on the farm and leave enough surplus either for expansion or investment in other areas. Regrettably, however, cocoa prices in Ghana have not enabled the realization of this expectation (Anon, 1995). Identified farmers' constraints include:

i) Low producer price

ii) Income and expenditure pattern of farm households

iii) High prices of inputs and availability on a sustainable basis

iv) Farmer priorities, preferences and capacity to implement research recommendations

v) Pattern of land holdings, tenurial arrangements (e.g. *Abusa system), inheritance and fragmentation of farms.

vi) Lack of workable credit or loan facilities

vii) Poor social circumstances of farmers

These problems have caused some farmers to neglect their farms and to shift to other crops (Osei, 1993). The profit margin of about seven hundred and fifteen thousand cedis (¢714,903) per ha (Table 5), the equivalent of three hundred and ten US dollars ($310), for the Ghanaian cocoa farmer (Asante et al. , 1997) is grossly inadequate. The margin is even less under the share-cropping system.

The poor educational, health, communication and infra-structural facilities has also resulted in the drift of the youth from the rural areas to the urban areas. This has greatly affected the availability of farm labour, leading to high cost of labour and, consequently, the deterioration of farms.

* In the "Abusa system, the caretaker receives one-third share of the crop whilst the land owner takes two-third.

Moreover, the recommended methods for the control of capsids and black pod disease which involve mainly the use of conventional insecticides and fungicides are now considered by many to be environmentally unfriendly, posing a threat to both humans and non-target beneficial organisms.

Adoption of research recommendation by farmers

Due to the constraints discussed above, cocoa farmers in West Africa find research recommendations on pest and disease control and farm management practices unaffordable, and the rate of adoption has generally been very low (Ollennu et al., 1989; Anon, 1990; Donkor et al., 1991; Anon, 1995; Adeyimi, in press; Freud et. al., in press).

Freud et al. (in press) observed that only 0.3% and 1.8% of farmers in Ghana and Cote d'Ivoire, respectively, sprayed their farms against cocoa capsids four times a year as recommended. The majority of farmers spray their farms only once or twice a year whilst 49.8% and 33.8% in Ghana and Cote d'Ivoire, respectively, do not spray their farms at all. In Ghana, the situation has been worsened since 1996/97 when subsidies on cocoa pesticides and spraying machines were withdrawn, resulting in an increase of almost 600% in price of miridicides and almost 300% in the price of motorized knapsack spraying machines. After the price increases, many farmers have been unable to spray their farms because of their apparent inability to afford the new prices. and this is likely to lead to a sharp drop in production. On the other hand, the high government subsidies on pesticides in Ghana rendered them much cheaper than those in neighbouring countries with the result that pesticides from Ghana were smuggled into Nigeria and Cote d'Ivoire where they were sold at higher prices. For disease control, the picture has been similar: the majority of farmers either do not spray their farms al all, or do only one or two applications instead of the recommended 6-7 applications per year for black pod control, thus incurring heavy crop losses every year (Henderson et al., 1994; Opoku et al., 1997).

Due to the destructive nature of the eradication method for controlling cocoa swollen shoot disease, it has been of little interest to farmers and some even oppose it despite the payment of compensation for trees lost and grants for replanting.

For weed control, Anon (1995) observed that only one-third of farmers interviewed in a nation-wide survey in Ghana in 1991-1993 weeded their farms adequately (i.e. 3-4 times a year as recommended). Chromolaena odorata, locally called Acheampong weed, was reported to be a constraint in the establishment of farms, especially on secondary forest lands in all regions.

Concerning planting material, the majority of farmers (62%) used their own planting method. Moreover, only 27% used recommended research planting methods whilst 11% used both recommended and own methods.

Due to the low adoption rate of research recommendations by farmers, estimated yields in West Africa have remained well below the 2 to 3 tonnes/ha obtained on experimental stations. Moreover, some farmers have neglected their farms to shift to other crops (Osei, 1993). Thus there is an urgent need to develop more acceptable pest and disease control methods and management practices.

PROPOSED INTEGRATED MANAGEMENT PRACTICES FOR SUSTAINABLE COCOA PRODUCTION

The following integrated management practices and socio-economic policies are recommended for sustainable cocoa production in West Africa. The development of some of the methods involved has already been initiated and need to be intensified.

Promising low-cost biological approaches to pest management to be developed include the use of semiochemicals, bio-pesticides of plant origin, biocontrol agents, and resistant cocoa varieties. These methods, combined with sound cocoa agronomic practices, should create conditions unfavourable to pests and pathogens, without affecting beneficial insects and other organisms including pollinators and natural enemies.

The use of these IPM strategies should be based on a sound knowledge of the biology, behaviour and the population dynamics of the pests and pathogens in the different geographical regions of each producer country and should operate under favourable socio-economic conditions. Along these lines the under-mentioned studies have either been or are about to be initiated at the Cocoa Research Institute of Ghana (CRIG), alone or in collaboration with others:

1. Pest Management Strategies

a) Capsids

i) Integrated pest management in the Ghana cocoa industry:

CRIG, in collaboration with an American University, has initiated a project aimed at developing an Integrated Pest Management (IPM) package involving the use of capsid sex pheromones and other semiochemicals of capsid and plant origin, as well as natural enemies, especially pathogenic fungi, in combination with other available methods, for capsid control (Padi et al. in press). The presumption that sex pheromone might be a useful tool for capsid control is based on findings from an earlier study on capsid behaviour by King (1973) who demonstrated that D . theobroma produces a sex pheromone. The present study has so far identified five compounds two of which have been found to be attractive to both sexes of D. theobroma and S. singularis under laboratory conditions. These findings are yet to be confirmed by adequately replicated experiments using large numbers of laboratory-reared capsids. The study has also identified a number of potential natural enemies, including fungal pathogens, which are yet to evaluated in the laboratory.

ii) The use of botanical pesticides.

A project proposal has been prepared on the use of bio-pesticides of plant origin for capsid control. These include extraxts of Azadirachta indica (neem), Jatropha spp. and other local plants. The use of non-persistent plant extracts which have low mammalian toxicity will help reduce the environmental hazards and other objections associated with conventional insecticides.

It is of interest to note that a small group of Ghanaian farmers, the Traditional Organic Farmers Association (TOFA) have identified themselves with an international organization called the "International Federation of Organic Agriculture Movement" (IFOAM) whose main concern is to produce organic crops which are free from conventional insecticides. TOFA claims to produce organic cocoa, i.e. cocoa on which only pesticides of purely organic (natural) origin are used. The association also permits the use of non-natural products which are chemically synthesized but are identical to natural products such as pyrethroids and pheromones. The farmers in TOFA have been promised a high premium on organically produced cocoa. As at now, TOFA is not practicing any form of pest /disease control. It is, however, clear from field experience that, in West Africa, the continued production of cocoa without pest and disease control will create serious problems. It is, therefore, considered that until organic or other non-conventional pest and disease control substitutes have been identified, tested and recommended for use, the production of organic cocoa as advocated by TOFA will be unsustainable. Fortunately, TOFA has now approached CRIG to test "NEEMOL", an Azadirachtin extract from neem packaged in India, for its effectiveness against capsids. Recent laboratory and field studies conducted in Ghana have shown that capsid mortalities of 93% and 80% caused by a 20% neem seed water extract and 4.62% neem oil, respectively, fell below the acceptable values (Adu-Acheampong, 1997). Nevertheless, the neem derivatives have considerable potential for mirid control in small-scale farming systems, provided farmers will have a reasonably cheap source of neem and the promised premium on organic cocoa would compensate for the losses from capsid damage likely to be associated with the use of neem derivatives in lace of conventional insecticides. This presupposes that the use of neem will not pose any serious environmental problems and further studies are required to determine the effects of neem on non-targeted fauna and to accurately establish its use in an integrated pest management (IPM) system.

NEEMOL is also to be tested alongside extracts of other local plants such as Jatropha curcas and Piper guineense, which are known to have insecticidal or medicinal properties (Abbiw, 1990). Moreover, further studies would be required to investigate the possibility of farmers using such extracts and to incorporate the plants as inter- or barrier- crops in cocoa farms in order be readily available as and when required. Other organically certified products will also be tested both in the field and on cocoa beans in storage.

iii) Studies on the population dynamics of capsids and other insects on hybrid cocoa:

The existing capsid control regimen of spraying insecticides four times in the year between August and December was based on the occurrence of capsid peak population at that period which also coincided with the peak cropping of Amelonado cocoa. With the introduction of the Amazon hybrids which produce pods almost throughout the year, the temporal distribution of capsids and other insects appear to have changed. Moreover, in Ghana, the major cocoa-producing areas have shifted from the east to the West of the country and this change is also likely to affect insect population dynamics.

As a basis for the correct timing and frequency of application of all available control methods, a project proposal has been prepared by CRIG for a study on the population dynamics of capsids and other insect pests on hybrid cocoa in the different cocoa growing regions of Ghana.

iv) Cultural practices and minimum use of pesticides

Studies are in progress in Ghana aimed at reducing the frequency of application of pesticides for capsid and black pod disease control. Results so far on capsid control indicate that, on mature cocoa, it might be possible to reduce the frequency of application of miridicides from the recommended four times per year to two times, after an initial three years of treatment of four times per year leading to good canopy formation (Owusu-Manu & Siaw, 1996).

 Another possibility is the "spot spraying" treatment (Owusu-Manu et al., 1977; Owusu-Manu, 1990) which aims at spraying only the capsid pockets within the period March to June when capsid damage is mainly on vegetative tissue and is most critical to the survival of the trees. This, together with adequate overhead shade and a closed inter-lacing cocoa canopy, can reduce capsid numbers to levels below the economic threshold.

The "spot spraying" and the reduced frequency of application methods are yet to be recommended at the national level.

v) Resistant cocoa varieties:

The identification and development of cocoa varieties that are tolerant or resistant to capsid damage, although a long-term approach, is a potential tool that must be exploited. Information on cocoa types observed in the field as being tolerant to capsids has been reviewed by Padi (1997 a) and it is being suggested that the development of cocoa varieties tolerant/ resistant to capsid attack should be incorporated into on-going breeding programs for CSSV and black pod resistance. The use of such varieties will greatly reduce the need to apply insecticides to control capsids.

b) Termites and other pests.

There is the need to develop IPM methods for the control of the now frequent outbreaks of hitherto unimportant insect pests like Bathycoelia thalassina, Pseudotheraptus devastans, the stem borer Eulophonotus myrmeleon, termites and defoliating caterpillars. For termite control a method which is environmentally safer needs to be developed. The example of applying of Chlorpyrifos as dust into entrances of the soil-subterranean nests of leaf-cutting ant colonies in Latin America could be emulated by applying the dust into entrance holes of termite nests and mounts. The use of "baits" for trapping destructive ants practiced Latin America offers yet another opportunity of an environmentally friendly strategy for termite control.

2) Disease Management strategies

a) Swollen shoot disease.

It is important that swollen shoot disease control should be made less expensive and more acceptable to the farmers and that eventually, the disease should be controlled by measures other than eradication. In this regard, the revision of the eradication control programme in 1986 to incorporate recent research results and the reintroduction of monetary compensation and replanting grants (Ollennu et al., 1989), have resulted in significant improvement over the previous situation (Ampofo, 1997). In the medium to long term, studies are in progress to develop strategies that seek to eliminate or at least reduce the frequency of eradication. These strategies include the exploitation of available tolerance/resistance, the search for new/higher sources of resistance, the integrated management of the mealybug vectors (Padi, 1997b), cultural practices to reduce virus spread, mild virus strain protection (Owusu et al., 1996), and the development of non-conventional resistance.

Studies are in progress on the possible use of natural enemies (predators, parasitoids and pathogens) for the control of the mealybug vectors of CSSV (Asante and Ackonor, 1996, Ackonor, 1997) and these studies should continue. In addition, with the wide range of relatively safer insecticide groups now available, studies on chemical control of mealybugs started but abandoned several times (e.g. Mapother & Nicol, 1953; Hannah and Heatherington, 1957; Marchart, 1968; Firempong, 1976) should be revived.

b) Black Pod disease

For black pod disease, it has been shown that it is possible to reduce the frequency of fungicide application from eight times a year (3-weekly application) to five times a year (4-weekly application) and still achieve good control of P. megakarya (Opoku et al., 1998). This will help reduce the cost of black pod control and render the method more acceptable to farmers, in addition to reducing the amount of chemical used. The possibility of using phosphorous acid which apparently carries little or no danger of residues or contamination of the environment (Opoku et al., 1998) is even more encouraging. Furthermore, it is of interest to note that, in preliminary field trials in Nigeria, a mixture of herbal extracts locally called "Tiwantiwa" has been found to be as effective for black pod control as the two standard types of Bordeaux mixture used in that country (Olunloyo, 1997). Further field trials on "Tiwantiwa" is in progress to determine its side effects on the environment.

The control of ants and other insects implicated with the spread of pathogen has been suggested (Okaisabor, 1971) but this is controversial since the role of such insects, particularly of ants, in the tropical cocoa ecosystem is complex (Leston, 1970).

Another possibility for black pod control is the use of other pathogens such as Aspergillus and Trichoderma spp. which have been found to inhibit growth of P. palmivora under laboratory conditions in Ghana (Odamten & Clerk, 1984). In Nigeria, Okaisabor (1968) observed the suppression of P. palmivora by Botryodiplodia theobromae in mixed cultures on wounded pods and culture medium. These and other observations in Ghana (Attafuah, 1966) and elsewhere (Frais & Garcia, 1981; Odigie & Ikotum, 1982) suggest opportunities to be exploited for the biological control of Phtophthora spp. on cocoa.

Frequent sanitation field surveys to remove infected pods will greatly help in reducing/eliminating tree cankers, the other basic source of inoculum besides the soil reservoir. Also as suggested by Fulton (undated), successful Black Pod management depends on the soil habitat originally selected for water drainage qualities and continuous modification to fit "supressive" soil specifications (i.e. liming to maintain a soil pH range of 6.0-6.5) which will help keep Phytophthora inoculum potentials at a low level.

Studies on the development of cocoa varieties resistant to black pod disease should be intensified.

3. Agronomic and Socio-Economic problems:

a) Agronomic considerations.

i) Intercropping and shade

Studies on the incorporation of suitable food crops and other fast growing plants as inter-crops on a more scientific basis during the establishment phase of cocoa to provide shade and to serve as physical barriers to capsids and other insect pests should be intensified. Preferably, non-host plants of pests and diseases which will also be useful and will generate additional income for farmers should be selected. The provision of good shade will help to suppress weeds and lead to the early formation of a closed canopy and thereby create conditions unfavourable to cocoa capsids. On the other hand, in countries like Ghana where most farms appear overshaded, the thinning of shade to the recommended level of 10 -15 trees per hectare, with the necessary fertilizer application, should be encouraged for sustainable yield (Ahenkorah and Akrofi, 1971)

There is the additional need to identify fast growing evergreen trees which are preferably leguminous or beneficial in other ways, as permanent shade. The use of these fast growing trees will help to restore and sustain the degenerating forest. Selected trees should not be hosts to cocoa insect pests or diseases. In this respect, the use of Gliricidia sepium as a permanent shade tree in Ghana should be discontinued since it sheds its leaves under dry conditions, unless properly managed, and is also a host plant for Planococcoides njalensis, the most important mealybug vector of CSSV (Padi & Hollander, 1996). The elimination of G. sepium as a shade tree in cocoa plantations has previously been recommended based on the fact that it is a host plant for mealybugs, thrips and aphids (Fulton, undated), but a suitable replacement has not been found.

The adoption of sound cultural practices such as regular mistletoe removal, pruning and mulching, and the use of leguminous plants as cover crops for weed control and for replenishing soil nutrients should be encouraged to reduce maintenance costs.

ii) Soil nutrition

The use of inorganic fertilizers in areas of depleted soil nutrients. The on-farm fertilizer verification trials now in progress in selected locations of Ghana (Appiah et al., 1996) is in the right direction and should be extended to cover other areas. Further expansion should include the use of organic sources such as cocoa pod husk and farmyard manure, to reduce cost.

b) Socio-economic considerations.

It necessary for the producer price of cocoa to be increased to a level comparable to, if not more than those of other crops. A high producer price will be an incentive for farmers to improve their farms, if they know they will obtain fair returns on their investment (Anon, 1995). Moreover, a remunerative producer price is necessary for attracting the youth into cocoa farming and to sustain farmers' interest. As recommended by Anon (1995), there is the need to constantly review the producer price, taking the economic situation into account, including the following factors:

i) Inflation

ii) World market price

iii) Production cost

iv) Income from competing crops

v) Adequate profit margin

vi) Announcement of producer price in good time to enable farmers take adequate advantage of the from price increase.

In addition, farm inputs such as recommended planting material, insecticides, fungicides, spraying machines and standard pruners should be made available to farmers at the right time. In Ghana, the cocoa extension services which is responsible for farmer education should be provided with adequate facilities and support. This is the best way to ensure farmers' adoption of recommendations for farm improvement and maintenance.

Incentive-oriented policies such as bonuses and credit facilities with terms of payment which take into consideration the cash flow of cocoa farmers would be essential to sustain cocoa cultivation. In Ghana, these activities should be channelled through the five existing farmers' societies/ associations since past experience has shown farmers are capable of running viable credit programs and that credit programs controlled by government ended in failure. In addition, good educational, health, communication and other infrastructural facilities should be made available in the rural areas where farmers live, to improve their social conditions. This will also prevent the migration of the youth to the towns and cities encourage them to enter into cocoa farming to replace the old farmers, many of whom are above the age of 50 years in Ghana.

Furthermore, there is the need for intensive education to enhance farmers' awareness that investment in their farms can be recouped with reasonable profit if research and extension recommendations are followed. This is the best way farmers can reap the full benefits of their labour and help to make cocoa production sustainable in West Africa, dependent as it is on numerous small-holder peasant farmers.

Table 1. Cocoa production in Ghana; farm size, number of producers, and total area cultivated.

Farm size Number of Areas cultivated

Source: Cocoa Services Division (unpublished data)

Table 2. Alternative host plants of Sahlbergella singularis. Recorded in West Africa (From Entwistle, 1972).

Host Plant Order: Family

From Padi et al., 1996

TABLE 3: LEVELS OF PERMANENT SHADE IN COCOA PLANTATIONS IN GHANA AND COTE D'IVOIRE ( % of total cocoa acreage actually harvested).

 TABLE 4: MULTIPLE REGRESSION ANALYSIS OF FACTORS AFFECTING COCOA YIELD

SUMMARY OF RESULTS.

Key: "-" coefficient insignificantly different from zero, "*" significant at 90%, "**" significant at 95%, "***" significant at 99%, in two-tailed test; "/" separates estimates for 1992/93 and 1993/94 seasons:

Table 5: Contribution of various farm operations and inpts to cocoa production costs at price/ha (1995).

Farm operation/input Cost (¢) % of total cost

LABOUR*

CAPITAL INPUTS

DEPRECIATION

PROFITABILITY

REFERENCES:

Abbiw, D. K. (1990). Useful plants of Ghana: West African uses of wild and cultivated plants. Intermediate Technology Publications Ltd and Royal Botanical Gardens, U.K.: 337.

Abdul-Karimu, A. and S. A. Bosompem (1994). Akomadan P. megakarya resistance selection programme. Rep. Cocoa Res. Inst. Ghana, 1992-93: 58.

Ackonor, J. B. (1995). Guide to termite control in cocoa and coffee farms in Ghana. Farmers' Guide no. 8., Cocoa Research Institute of Ghana, Tafo.

Ackonor, J. B. (1977). Preliminary studies on breeding and predation on Scymnus (Pullus) sp. and Hyperaspis egregia Mader on Planococcoides njalensis (Laing): Proc. 1st Cocoa Pests and Diseases Seminar, Accra, Ghana, 6-10 November, 1995: 238-241.

Adeyimi, A. A. (in press). NPK fertilizer in rehabilitated cocoa: Effects on the growth and development of cocoa under different methods. Proc. 12th Int. Cocoa Res. Conf. Salvador, Bahia, Brazil, Nov., 1996.

Adomako, D., Frimpong, E. B., Ampofo, S. T. and Brew, K. M. (1990). Observations on cocoa flushing in Ghana under different shade regimes, with particular reference to the distribution of natural light. The Planter, Kuala Lumpur, 66: 562-579.

Adu-Acheampong, R. (1997). Laboratory and field evaluation of neem (Azadirachta indica A. Juss) for the management of cocoa mirids (Heteroptera: Miridae). M.Phil thesis submitted to the University of Ghana in October, 1997.

Adu-Ampomah, Y. (1994). Alternative hybrids programme. Rep. Cocoa. Res. Inst. Ghana, 1990/91: 46-49.

Ahenkorah, Y. and Akrofi, G. S. (1971). Recent results of fertilizer experiments on shaded cocoa (Theobroma cacao L.) in Ghana. Proc. 3rd Int. Cocoa Res. Conf., 23-29 November, 1969, Accra, Ghana: 65-78.

Akotoye, N. A. K. The biology of Characoma stictograpta Hampson (Lepidoptera: Noctuidae), a pod feeding pest of cocoa in Ghana: MSc thesis. Zoology Dept., Univ. of Ghana, Legon: 274 pp.

Akrofi, A. Y. (in preparation). A review of Phytophthora pod disease situation in Ghana: 15 pp.

Akrofi, A. Y., Opoku, I. Y. and Appiah, A. A (in press). Extended interval of application of fungicides for the control of black pod disease in Ghana. Proc. 2nd Int. Cocoa Pests and Diseases Seminar., 6-10 Nov., 1995, Yammousukro, Cote d'Ivoire.

Ampofo, S. T. (1989). Swollen shoot control: review of past and present work. Proc. Cocoa Swollen Shoot Workshop, 25-26 July, 1989: 4-12.

Amponsah, J. D. and Asare-Nyako, A. (1973). Glasshouse method for screening seedlings of cocoa (Theobroma cacao L.) for resistance to black pod caused by Phytophthora palmivora (Butl.). Tropical Agriculture (Trinidad). 50, No. 2: 143- 152. Anon (1918). Woodpeckers and cocoa. J. Jamaica Agric. Soc., 22 (2), 65-9, and (3), 102-7.

Anon (1951). Capsid Research: Chemical control. Ann. Rep. W. Afr. Cocoa Res. Inst. 1949-50: 40-42.

Anon (1990). Analysis of constraints on adoption of CRIG recommendations in Offinso District: Results and discussion. (C.P. Henderson and A. P. Jones ed.): 15pp. Cocoa Research Institute, Tafo, Ghana.

Anon (1995). Report on causes of recent decline in cocoa production in Ghana and measures to revamp the industry. Office of the President, Accra: 103 pp.

Appiah, M. R., Ofori-Frimpong, K., and Afrifa, A. A. (1996). Farmer fertilizer verification trials. Rep. Cocoa Res. Inst. Ghana, 1995/95: 52.

Appiah, A. A. and Owusu, G. K. (1997). Cocoa mistletoes - a review. Proc. 1st Int. Cocoa Pests and Diseases Seminar, Accra, Ghana, 6-10 November: 272-279.

Are, L. A. (1970). Cocoa rehabilitation in Nigeria. Annual Report of the Cocoa Research Institute of Nigeria for 1968-69: 20-29.

Armstrong, K.B. (1961). Systemic insecticides. Rep W. Afr. Cocoa Res. Inst., 1959-60: 39-40.

Asare-Nyako, A. (1969). The black pod disease of cacao, a review of the situation in Ghana. Proc. 3rd Int. Cocoa Res. Conf., Accra, Ghana: 370-374.

Asare-Nyako, A. and K. Owusu, G. K. (1981). Meliodogyne incognita infection of cocoa seedlings. Proc. 7th Int. Cocoa Res. Conf. 4-17 November, 1979, Douala, Cameroun : 457-461.

Asante, E. G. (1997). The economic relevance of plant disease and pest management in the Ghana cocoa industry. Proc. 1st Int. Cocoa Pests and Diseases Seminar., 6-10 Nov, 1995, Accra, Ghana: 288-299. 288-299.

Asante, S.K. and Ackonor, J. B. (1996). Natural enemies of cocoa mealybugs. Rep. Cocoa res. Inst. Ghana, 1994/96: 158.

Attafuah, A. (1996). Black pod disease: biological control. Rep. Cocoa Res. Inst. Ghana, 1963-65,: 29-30.

Bach de, P. (1964). Biological control of insect pests and weeds. (Chapman and Hall, London: 844 pp.

Bateman, M.J. (1988). Ghana Cocoa Pricing Policy study. Commodity information Inc. 1981: 77pp.

Bondar, G. (1939). Insecto daninhos e parasitas do cacau na Bahia. Bolm. tec. Inst. Cacau Bahia, 5: 112 pp.

Bruneau de Mire, P. (1966). Comparation entre deux modes des traitement anti-mirids du cacaoyer: la thermonebulisation et l'atomisation. Proc. Cong Int. Rech. Agron. Cacao, Abidjan: 154-159.

Collingwood, C.A. (1969) Assessment methods for cocoa mirid trials. Cafe', Cacao, The' 13 (3): 206-208.

Collingwood, C.A. (1971). A comparsion of assessment methods in cocoa mirid count trials. Proc. 3rd International Cocoa Res. Conference, Accra, Ghana: 161-168.

Collingwood, C. A. and Marchart, H. (1970). Chemical control of capsids and other insect pests in cocoa rehabilitation. Proc. 3rd International Cocoa Res. Conf., Accra, Ghana, 1969:

Collingwood, C.A. and Marchart, H. 1971. Varietal difference in capsid suceptibility. Rep. Cocoa Res. Inst., Ghana, 1969-70: 96-98.

Cotterell, G. S. (1926). Preliminary studies on the life history and habits of Sahlbergella singularis Hagl. and S. theobroma Dist., attacking cocoa in the Gold Coast with suggestions on control measures,. Bull. Dept. Agric. Gold Coast., 3: 26 pp.

Crowdy, S. H. (1947). Observations on the pathogenicity of Calonectria rigidiuscula (Berk and Br.) Sacc. on Theobroma cacao L., Ann. Appl. Biol. 34: 45-59.

Dakwa, J. T. (1987). A serious outbreak of black pod disease in marginal area of Ghana. Proc. 10th Int. Cocoa Res. Conf., Santo Domingo: 447-451.

Donkor, M. A., Henderson, C. P. and Jones, A. P. (1991). Survey to quantify adoption of CRIG recommendations. FSU Research Paper, 3, Cocoa Research Institute, Tafo, Ghana.

Dungeon, G. C. (1910). Notes on two West African Hemiptera injurious to cocoa. Bull. Ent. Res. 1: 59-61.

Entwistle, P. F. (1965). Entomology. Rep. Cocoa Res. Inst., Nigeria, 1963-64: 66-72.

Entwistle, P. F. (1972). Pests of cocoa. Longman Group Ltd.: 779 pp.

Evans, H. C. (1973). New developments in black pod epidemiology. Cocoa Growers Bulletin 20: 10-16.

Everard, C. O. R. (1968). A Report on the rodent and other vertebrate pests of cocoa in Western Nigeria., Res. Divis., Min. Agric., Ibadan, Nigeria: 123 pp.

Firempong, S. (1984). Screening of insecticides for mealybug control Rep. Cocoa Res. Inst., Ghana, 1976/77-1978/79: 72.

Frais, T. G and Garcia, E. R. (1981). Effectiveness of some micro-organisms antagonistic to Phytophthora palmivora (Butl.) in controlling blackrot of cocoa pods. Revista Mexicana de Fitopatologa 1 (3): 16-20.

Freud, E. H., Petithuguenin, P. and Jaques Richard (in press). The diffusion and impact of cocoa research themes in Ghana and Cote d'Ivoire. Proc. 12th Int. Cocoa Res. Conf., Salvador, Bahia, Brazil, 17-23 Nov., 1996.

Gerard, B. M. (1968). A note on mirid damage to mature cocoa pods. Nigerian Ent. Magazine, 1, 59-60.

Gibbs, D. G. and Leston, D. 1970. Insect phenology in a forest cocoa farm locality in West Africa. J. appl. Ecol., 7: 519 - 48.

Gibbs, D. G., Pickett, A. and Leston, D. 1968. Seasonal population changes in cocoa capsids (Hemiptera: miridae), in Ghana. Bull. Ent. Res., 58: 279-293.

Gill and Duffus (1989). Cocoa Statistics. Gill and Duffus Gp. Ltd.: 29 pp.

Hammond P. S. (1957). Notes on the progress of pest and disease control in Ghana. Proceedings of the Cocoa Conference, London, 1957: 110-118.

Hannah, A. D. and Heatherington, W. (1957). Arrest of the swollen shoot virus disease of cocoa in the Gold Coast by controlling the mealybug vectors with the systemic insecticide, Dimefox. Ann. Appl. Biol., 45: 473-480.

Henderson, C. P., Asante, E. G., Donkor, M. A., Ameyaw, K. O., Luterbacher, M. C., Akrofi, A. Y. and Boakye, P. D. (1994). Farmer managed fungicide trials (Volta and Brong Ahafo Regions). Rep. Cocoa Res. Inst., Ghana, 1991/92: 70-76

Hislop, E. C. and Park, P. O. (1960). Fungicide research. Rep. West Afr. Cocoa Res. Inst., 1960-61: 96-106.

Johnson, C. G. (1962). Capsids: A review of current knowledge. In Agriculture and land use in Ghana (Wills, J. B. ed.). Oxford University Press: 504 pp.

Kay, D. (1957). Quart. Rep. W. Afr. Cocoa Res. Inst., 48: 26

King, A. S. B. (1973). Sex attraction of Distantiella theobroma (Dist.). Ent. Exp. Appl., 16,: 234-254.

Konam, J. K., Jason, Dennis, J. Saul, J. Flood and D. Guest. Integrated management of Phytophthora diseases of cocoa in Papua New Guinea. Proc. 12th Int. Cocoa Res. Conf., Salvador, Bahia, Brazil, 17-23 Nov., 1996: 10 pp.

Legg, J. T., Owusu, G. K., Ollennu, L. A. A. and Lovi, N. K. (1981). The problems of controlling cocoa swollen shoot disease in Ghana. Proc. 6th Internat. Cocoa Res. Conf., Caracas, Venezuela,: 267-278.

Leston, D. (1960). The snakes of Tafo, a forest cocoa farm locality in Ghana. Bull. de l'Institute Food a mental d'Afrique Noir, Series A. Jone, 30 No 2: 737-770.

Leston, D. (1970). Entomology of the cocoa farm. Ann. Rev. Entomol., 15: 273-294.

Leston, D. (1971). Ants, capsids and swollen shoot in Ghana: interactions and the implications for pest control. Proc. 3rd Internat. Cocoa Res. Conf., Accra, Ghana, 1969:

Madge D. S. and Gerard, B. M. (1967). Physiology of cocoa mirids. Rep. Cocoa Res. Inst. Nigeria, 1965/66: 37-8.

Marchart, H. (1969). Uptake of systemic Insecticides. Rep. Cocoa Res. Inst. Ghana, 1967/68: 40.

Mopother, H. R. and Nicol, J. (1953). Mealybug studies. Rep. W. Afr. Cocoa Res. Inst., 1950-51: 40.

Mossu, G. (1992). The Tropical Agriculture: Cocoa CTA. Ed. René Coste, Macmillan, pp 103.

Nicol, J. (1953). Progress of research on systemic insecticides at WACRI. Proc. 4th Cocoa Conf., London: 98.

Odamten, G. T. & Clerk, G. C. (1984). Studies on the possible use of two soil fungi, Aspergillus niger and Trichoderma viride as biological control agents of Phytophthora palmivora and the effect of their metabolites on cocoa seedlings. Proc. 9th Int. Cocoa Res. Conf., Lome, Togo: 293-301.

Odigie, E. E. and Ikotum, T. (1982). In-vitro and in-vivo inhibition of growth of Phytophthora palmivora (Butl.) by antagonistic microorganisms. Fitopatologia Brasileira, 7 (2): 57-169.

Okaisabor, E. K. (1968). An active fungal parasite of Phytophthora palmivora (Butl). PANS (B) 14: 353-60.

Okaisabor, E. K. (1971). The mechanism of initiation of Phytophthora pod rot epiphytotics. Proc. 3rd Int. Cocoa Res. Conf., Accra, Ghana, 1969: 398-404.

Ollennu, L. A. A., Owusu, G. K. and Thresh, J. M. (1989). The control of cocoa swollen shoot disease in Ghana. Cocoa Growers' Bulletin No. 42, december, 1989: 25-35.

Olunloyo, A. O. (1997). Efficacy of herbal plant substitutes in the control of black pod disease of cocoa - Preliminary results. Proc. 1st Int. Cocoa Pests and Diseases Seminar, 6-10 Nov, 1995, Accra, Ghana: 119-125.

Opoku, I. Y., Akrofi A. Y. and Appiah, A. A. (1996). The effect of Phosphorous acid on black pod disease incidence and yield. Rep. Cocoa Res. Inst., Ghana, 1994/95: 93-95.

Opoku, I. Y., Akrofi, A. Y. and Afrifa, A. A. (1997). The spread of Phytophthora megakarya on cocoa in Ghana: Proc. 1st Int. Cocoa Pests and Diseases Seminar, 6-10 Nov, 1995, Accra, Ghana: 98-107.

Opoku, I. Y., Akrofi, A. Y., Appiah. A. A. & Luterbavher, M. C. (1998). Trunk injection of Potassium Phosphonate for the control of black pod disease of cocoa. Tropical Science (in press).

Osei-Bonsu, K. and Anim-Kwapong, G. J. (1997). Preliminary evaluation of some forest trees for cocoa cultivation. Proc. 20th Biennial Conf. Ghana Sci. Assoc., UST., Kumasi, Ghana., 4-8 August, 1997.

Osei-Bonsu, K., Anim-Kwapong, G. J., Oppong, F. K, Amoah, F. M. and Frimpong, E. B (in press). Evaluation of cassava as temporal shade for establishing cocoa. Rep. Cocoa Res. Inst., Ghana, 1996/97.

Osei, J.K. 1993. Construction of a low density linkage map of Theobrma cacao L. using random amplified DNA markers. Ph.D. Thesis. Univ. of Pensylvania.

Owusu-Manu, E. (1996). Capsid population studies. Rep. Cocoa Res. Inst., Ghana, 1994/95a: 160.

Owusu-Manu, E. (1971). Bathycoelia thalassina - another serious pest of cocoa in Ghana. in CMB Newsletter, Vol. 47: 12-14.

Owusu-Manu E. and Siaw J. (1996 b). Frequency of timing of insecticide application. Rep. Cocoa Res. Inst., Ghana., 1994/95: 172-173.

Owusu-Manu, E; Somuah, J. M and Padi, B. (1977). Studies on seasonal movements of mirids, Distantiella theobroma Dist. and Sahlbergella singularis (Hagl.) (Hemiptera: Miridae) in an infested area. Proc. 6th Int. Cocoa Res. Conf., Caracas, Venezuela, November, 1977: 401-407.

Owusu-manu, E. (1990). Chemical control of the cocoa mirids Distantiella theobroma (Dist.) and Sahlbergella singularis (Hagl.) in relation to their seasonal movement. Integrated pest management in tropical and subtropical Cropping Systems, 8th-15th February, 1989, Bad Durkheim Germany, No. 2: 363 -373.

Owusu-Manu, E. (1995). The need for chemical control of cocoa mirids in Ghana. In: Cocoa Pests and Diseases Seminar, Accra, Ghana, 6-10 November, 1995: 1 (Abstract).

Owusu, G.K. (1983). The cocoa swollen shoot disease problem in Ghana. In Plant virus Epidemiology, edited by R.T. Plumb and J.M. Thresh, Oxford: Blackwell Scientific: 19.

Owusu, G. K., Ollennu, L. A. A. and Dzahini-Obiatey, H. (1996). The prospects for mild strain cross-protectio to control cocoa swollen shoot disease in Ghana. Proc. 12th Intern. Cocoa Res. Conf., Salvador, Bahia, Brazil, 1996: (in press). Owusu, J. K., Adomako, D. and Hutcheon, W. V. (1978). Seasoal changes in total free sugar content of cocoa plants. Physiol. Plant, 44: 43-47.

Padi, B. and J. den Hollander (1996). Morphological variations in Planococcoides njalensis (Laing.) on cocoa in Ghana. Entomologia et Applicata, 79: 317-328.

Padi, B., Owusu, G. K. and Kumah, N. K. (1996). A record of Desplatsia dewevrei (De Wild & Th. Dur) (Tiliales: Tiliaceae) as an alternative and potential breeding host plant for the cocoa mirid Sahlbergella singularis Hagl. Proc. 12th Intern. Cocoa Research Conf., November 1996, Salvador, Bahia, Brazil: (In press).

Padi, B. (1997a). Prospects for the control of cocoa capsids - Alternatives to chemical control. Proc 1st Int. Cocoa Pests and Diseases Seminar, Accra, Ghana., 6-10 Nov, 1995: 28-36,

Padi, B. (1997b). Prospects for the control of cocoa mealybugs. Proc. 1st Int. Cocoa Pests and Diseases Seminar., 6-10 Nov, 1995, Accra, Ghana: 249-263.

Padi, B., Lowor, S. Ackonor, J. B and Kudekor, P (in press) Rep. Cocoa Res. Inst., Ghana, 1995/96.

Padi, B. and Sarfo, J. E (in press) FUMIVAP trials. Rep. Cocoa Res. Inst. Ghana, 1995/96.

Phillips, E. J. (1977). Mistletoe on cocoa in Ghana. Ghana Journal of Agric. Science.

Prins, G. (1965). A laboratory rearing method for the cocoa mirid Distantiella theobroma (Dist.) (Hemiptera: Miridae). Bull. Ent. Res., 55: 615-616.

Raw, F. 1959. An insectary method for rearing cocoa mirids, Distantiella theobroma (Dist.) and Sahlbergella singularis Hagl. Bull. Ent. Res., 50: 11-12.

Rocha, H. M. (1965). Cocoa varieties resistant to Phytophthora palmivora (Butler). A literature review. Cocoa Costa Rica), 10 (1): 1-10.

Room, P. M. (1972). Mistletoe on West African cocoa. Cocoa Growers Bulletin, 18: 14-18.

Stapely, A nd Hammond, P. S. (1959). Capsid control on mature cocoa. New Gold Coast Farmer, 2: 109-115.

Taylor, K. D. (1961). An investigation of damage to West African cocoa by vertebrate pests. Cyclostyled report, West Afrcan Cocoa Research Institute, Tafo, Ghana.

Taylor, B. and Griffin, M. J. (1981). The role and relative importance of different ant species in the dissemination of black pod disease of cocoa- In Epidemiology of Phytophthora on cocoa in Nigeria: 114-131.

Urquhart, D. H. (1955). Report on the Cocoa Industry in Sierra Leone, with notes on the cocoa industry in the Gold Coast. Cadbury Brothers, Bournville, 43 pp.

Usher, G. (1957). Capsid studies: fungi associated with capsid lesions. Rep. W. Afr. Cocoa Res. Inst., 1955-56: 56-9.

Warburg, O. 1902. Eine Rindenwanze als kakaoschadlingim kamerun, Tropenpflanzer, 6: 638 - 640.

Whitehead, A. G. (1969). Nematodes attacking coffee, tea and cocoa and their control. In Nematodes of Tropical Crops, ed. J. E. Peachey, pp. 236-50. Tech. Commun. no. 40, Commonw. Bureau Helminth.

Wills, J. B. (1962). Agriculture and Land use in Ghana. Oxford University Press, London.

Youdeowei, A. (1964). Progress in the laboratory rearing of cocoa mirids in Nigeria. Proc. Conf. on Mirids and Other Pests of Cacao, Ibadan, Nigeria, 1964, :98-100.

Youdeowei, A. (1974). The cocoa mirid problem in Nigeria. Proc. 4th Conf. West African Cocoa Entomologists, 9th-13th December, 1974, Tafo, Accra, Ghana: 29-34.