The declared areas of plantations range from about 1 ha to 12 ha, which was divided into three classes: [1 to 3 ha], [3 to 5 ha], and more than 5 ha. Overall, plantations with an area of [1 to 3 ha] represent 57.80% of the orchards. Plantations of [3 to 5 ha] and more than 5 ha, represent 22.02% and 20.18% respectively (Table 3). The average size of the orchards was 3.82 ha. The chi-square test showed no significant relationship (p = 0.9664) between the variable’s locality and orchard size class.

The age of the cocoa farms surveyed ranged from 10 to over 40 years. Orchards older than 40 years are the most numerous (68.93%), followed by orchards between 30 and 40 years old (26.21%) and those between 10 and 20 years old

(4.85%). No orchards of this last age class were recorded in the locality of Sérigbangan where 80% of the cocoa farms are old orchards of more than 40 years. In Yaodankro, plantations between 30 and 40 years old and those older than 40 years recorded 30.16% and 61.90% of the local orchard respectively (Table 3).

The survey revealed that three morphotypes (or morpho species) of insects are the main insect pests that are currently of concern to producers in Sérigbangan and Yaodankro. These are the capsids Sahlbergella singularis and Distantiella theobromae, the cocoa mosquito Helopeltis sp., and the green bug Bathycoelia thalassina (Figures 1-3). Regarding capsids and their damage, 43.64% of the producers said that they noticed a significant presence of this pest in their plantations, while 56.36% said that they did not observe too much damage from these insects, although they were present in their fields.

As for cocoa mosquitoes and their damage, they were observed much more by 54.55% of producers and less observed by 39.09%; and 6.36% of respondents indicated that they did not observe them in their plantations. Some farmers indicated that the cocoa mosquito is more damaging to the youngest fruits (cherries) and has no real impact on pods of a certain size.

Regarding the green bug, 67.27% of the producers consider this insect as one of the most observed and dangerous in their plantations, while for 32.73% of the farmers, the insect is less present in their orchards (Table 4). Chi-square tests showed that the level of capsid infestation in orchards was not related (p = 0.5968) to the production locality. For the green bug and the cocoa mosquito, the level of their populations in orchards was very significantly dependent on locality (p = 0.002, p = 3.826e−10, respectively for the insect types). These two insects were significantly more observed in Yaodankro cocoa farms. In addition to these insects, farmers reported signs of stem and pod borers as well as termites, which do not seem to be a real concern for them. The pathological disease of current concern to producers in the area is Swollen Shoot of the cocoa tree, the presence and extent of which did not vary by locality according to the chi-square test (p = 0.5792). At least 82.73% of the producers interviewed reported the presence of this viral disease in their orchards. They attributed the death of many cocoa trees to the action of the swollen shoot virus. Only 3.64% reported that they had not yet found the disease in their plantations. The fungal disease pod rot, whose presence or severity of infection is very highly dependent

on locality (p = 2.91e−12), seems to be no longer too much of a concern for most of the producers interviewed, especially those in Sérigbangan (Table 4).

Two types of weeding are carried out in the cocoa farms by the farmers interviewed.

These are manual weeding using machetes, which is carried out once to three times a year, and herbicide treatment using chemical herbicides, which is applied once a year. Among the farmers interviewed, 84.26% reported using chemical herbicides in addition to machete weeding. Those who carry out two manual weeding operations are the most numerous, with a rate of 66.35%, of which 54.81% use chemical weeding as a supplement to manual weeding. The proportion of those who carry out a single weeding operation with a machete is 22.12%, and all of them carry out a second weeding operation using herbicide products. And the rate of producers who carry out three manual weeding plus an application of herbicide is 6.7%.

Overall, 48.57% of producers do not perform any maintenance pruning of their plantations. Of the 51.43% of farmers who stated that they do this operation, 21.90% spend time on it. The latter generally carry out this operation between January and April, mostly in February-March, and 29.53% carry it out either during weeding, at the time of the pod harvest, or at times when the producer considers the presence of twigs and/or branches to be too cumbersome on certain cocoa trees. In Sérigbangan, 80.95% of cocoa farmers no longer spend time pruning, and only 4.76% do this activity, mainly between January and April. In Yaodankro, 28.13% of producers do not perform maintenance pruning, and 32.81% perform the operation between January and April. Regarding sanitary harvesting, 52.83% of producers stated that they clear the cocoa trees of mummified dry fruit and fruit with rot, and this is done during the harvesting of ripe pods. The chi-square test, with p = 3.01e−07 for pruning and p = 5.17e−08 for sanitary harvesting, shows that the performance of these activities depends on the cocoa localities. As for the management of these embarrassing fruits, producers leave them in place after they have fallen from the trees.

1) Acquisition of products

The survey reveals that most of the producers (84.55%) use chemical insecticides to control insect pests, and 15.45% do not use them. Seventy-eight-point forty-nine percent (78.49%) of pesticide users indicated that they buy the products themselves from street vendors or local vendors. Of the remaining 21.51% who mentioned that they receive insecticides from their cooperatives, 15.05% indicated that they supplement it by buying other boxes (products) on the local market or from street vendors. Statistical analysis indicated that there is a very high significant relationship (p = 9.981e−08) between the variables “cocoa localities” and “use or not use of insecticides” against pests in the orchards. All the producers interviewed in Yaodankro use insecticides, while among those surveyed in Sérigbangan, 60.47% use them and 39.53% do not. It was noted that it is in the latter locality that producers are supplied with insecticides by cooperatives.

2) Insecticide products used

A total of 55 chemical insecticides with 11 active substances in 17 formulations, 03 fungicides (4 active substances in 03 formulations), and 01 bio stimulant were identified (quoted by the producers and/or empty containers found in the orchards). Of the insecticides identified, three do not appear on the list of pesticides registered in Côte d’Ivoire. These are CACAOEXTRA 36 SC, CAOBOSS SUPER 80 SC, and THODAN GRO 50 SC; all three of these products were found in Yaodankro. The active substance found in the most cited products, taken in isolation, is Bifenthrin (25.79%). It is followed by Imidacloprid (22.35%), Acetamiprid (15.47%), Lamb-da-cyhalothrin (13.18), and Thiamethoxam (12.61%) (Figure 4(a)). In the chemical formulation, 93.68% of the insecticides found are two active substances, 4.02% are three active substances, and 2.30% are single formulation insecticides (01 active substance). With a proportion of 36.21%, products combining Imidacloprid and Bifenthrin were the most cited. They are followed by insecticides based on Lamb-da-cyhalothrin + Thiamethoxam (16.67%), and Acetamiprid + Bifenthrin (11.49%) (Figure 4(b)). The active substances of the inventoried insecticides belong to 5 chemical families: Neonicotinoids, Pyrethroids, Avermectines, Pyriproxyfen, and Diamines. The

first two families are widely used in the Méagui area. They are slightly (Class III) or moderately (Class II) dangerous products. Almost all (96.91%) of the producers interviewed indicated that they had never used a biopesticide. The 3.09% (in Yaodankro) who did use biopesticides indicated that they had used them only once on a trial basis. The names of the biopesticides used could not be provided by users.

3) Application of insecticide products

Of the farmers who use insecticides, 68.89% indicated that they apply the treatments themselves. The remaining 31.11% reported using private individuals who are known to be experienced in the operation. The chi-square test (p = 1.615e−11) shows that using or not using an experienced applicator is dependent on locality. In Sérigbangan, 84.62% of producers use an applicator trained by the cooperative to which they belong.

Unlike those in Yaodankro, 90.63% of whom perform the operation themselves. The application device unanimously cited is the motorized backpack sprayer; the preparation of the product is done directly in the tank of the device. The survey also revealed that not all farmers wear full personal protective equipment (PPE) during treatments. However, 59.78% reported wearing a minimum of PPE, generally consisting of their usual field clothing, which consists of a long-sleeved shirt, pants, and a pair of boots, to which is sometimes added a cloth muffler and/or a hat for carrying out the activity. And the statistical test showed that the wearing or not of this minimum of PPE is done at significantly different proportions (p = 2.343e−05) from one locality to another. Approximately 96.15% of the applications in Sérigbangan versus 45.45% of those in Yaodankro were wearing PPE.

4) Frequency of insecticide treatments

The frequency of treatment of orchards varies from once a year to more than 6 times a year divided into three classes: [1 to 2 times], [3 to 5 times], and ≥6 times per year. The overall proportion of farmers who perform 1 - 2 treatments per year is 11.82%. That of those who make 3 - 5 applications per year is 29.09%, and 43.64% of respondents make 6 or more insecticide treatments during the year. Statistical analysis revealed that the frequency of insecticide application varied significantly (p = 4.899e−15) by locality. In Sérigbangan, the frequency of application was increased (30.23%) by the class of [1 to 2 times] and decreased (4.65%) by the class of 6 or more applications per year. On the other hand, in Yaodankro, 68.66% of cocoa farmers applied at least 6 times per year, and 31.34% applied between 3 and 5 times per year. The frequency of 1 to 2 applications per year was not mentioned by producers in the latter locality (Table 5). Growers applied an average of 1.06 ± 0.53 liters of insecticide product per hectare. Fungicide treatments for pod rot are used mainly in Yaodankro.

5) Producers’ level of knowledge on recommended timing and frequency of insecticide treatments

Of the producers surveyed, about 63% answered “NO” to the question of whether they had knowledge of recommended timing and frequency of insecticide treatments. Of the 37% who answered “YES” to the question by indicating 2 to 4 applications per year, 27% were able to mention at least one of the months corresponding to the periods during which treatments are recommended in Côte d’Ivoire, i.e. December-February and July-August. The chi-square test revealed that the level of education of producers on the recommended periods (p = 0.5059), as well as on the frequency (p = 0.4654) of treatments, is similar in the two study locations (Table 6).

6) Compliance with pre-harvest interval (PHI)

Regarding compliance with the PHI, 75.90% of respondents reported that they usually apply pesticides just after picking and shelling activities. The delay between the application of products and the harvesting of ripe pods can, therefore, extend over 2 to 3 weeks. On the other hand, 24.10% said they did not pay particular attention to the time between treatments and harvesting (Table 6). The attitude of producers concerning the respect of the PHI is not a function of locality according to the chi-2 test (p = 0.3287).

All the producers surveyed harvested the pods using machetes and pruners when they noticed that there was enough ripe fruit on the trees. Most (93.64%) cocoa farmers harvested only once a month, and only 6.36% reported harvesting twice a month. For some farmers, it was found that as they harvest, they or some of their family members put the pods in small piles, while for others, the pods are first picked from the whole field before being put in small piles. Subsequently, the pods are re-piled in three or four places in the plantation (depending on the area) for shelling. The time interval from the first day of harvest to the day of shelling varied from 3 days to more than 7 days with an overall average of 6.77 ±

PHI: pre-harvest interval; Not Att: producer not paying attention to the PHI after insecticide application.

2.24 days. About 32.73% of the producers reported that they perform harvest to shelling activities generally within 3 - 5 days, 40% perform them within 6 - 7 days, and 27.27% go beyond 7 days before shelling. The chi-square test of independence (p = 0.042) shows that this time frame varies from one locality to another (Table 7).

A mini machete called a “shelling knife” is unanimously mentioned by the producers as the equipment used to open the pods. During the interview, they stated that they never use any other tool than the shelling knife to perform this operation. The producers do not detach the beans from the placentas during shelling. Regarding the separation of healthy pods from those in poor condition (fresh pods blackened or in a state of blackening and/or pods broken during harvesting), the farmers said that shelling is done without distinguishing the condition of the piled pods. However, they specify that during the operation, if the contents of a pod are not at all good (the case of mummified pods, beans that have germinated too much, or pods that have been badly attacked by insects causing total or almost total abortion of the beans), then the latter is thrown away with its contents. In Sérigbangan, during the operation, the beans are first placed in small baskets and then spilled onto a black plastic tarp after these containers are filled. In contrast, in Yaodankro, almost all producers use empty fertilizer bags (50 kg) as containers for the beans during shelling and for fermentation of the beans (Figure 5).

Four techniques for fermenting the beans are overall applied by the cocoa farmers in the localities surveyed. These are fermentation on a bed of banana leaves, used by only 6.36% of producers; fermentation on a bed of black plastic tarp, used by 32.72% of respondents; fermentation in bags only, used by 19.09% of producers; and fermentation in bags and then on a bed of black plastic sheeting. Regarding this last method of fermentation, 41.82% of farmers interviewed said that after shelling, the beans remain in bags, and the day before or two days before they are put in the sun, they are tipped in a heap on the tarp and covered by it. Only in Yaodankro was fermentation in the bags noted, and the use of banana leaves was noted only in Sérigbangan (Table 8). For the duration of fermentation, 65.45% of the producers let 3 to 5 days pass before putting the beans in the sun. Twenty-four-point fifty-five percent (24.55%) reported 5 to 6 days of fermentation, and 10% reported 2 to 3 days of fermentation. The statistical test revealed that this fermentation time depends significantly (p = 0.0003091) on the locations (Table 9). No stirring is done by producers when the beans are fermenting, and prior removal of placentas is not done before fermentation but rather on the first and/or second day of drying.

The only drying method used by producers is sun drying, which varies from 4 days to more than 7 days depending on the time of year. Nearly half (50.91%) of the respondents stated that they usually dry cocoa for 5 to 6 days; 21.82% do so for 4 to 5 days and 27.27% indicated that they do 6 to 7 days of drying. However, farmers indicated that during periods of low sunlight, cocoa is exposed to the sun for more than 7 days. The chi-square test indicated that the proportion of

farmers observing these times varied significantly (p = 0.0003595) across localities (Table 9). As a drying medium, almost all (91.82%) of the producers interviewed dry cocoa on black plastic tarps installed on the ground (Figure 6); 3.64% do so on cemented areas and 4.55% use trailer tarps. At sunset (and in case of rain) during the drying days, the beans are piled up and covered with black tarps.

On the last day of drying, the dried beans are put in bags and transferred to the store by the tracker or stored for a few days at the producer’s place waiting for the arrival of the buyer (tracker or cooperative).