Fruit flies (Diptera: Tephritidae) are reported to be responsible for infesting around 40% of the mango production in Africa, at great economic cost. Weaver ant colonies (Oecophylla sp.) are commonly found in mango trees and their presence is reported to have a deterrent effect on fruit flies in Australia and West Africa.
In this study, small scale farmers did not think weaver ants protected their mangoes from fruit flies. Observational studies confirmed the farmers’ views. No volatile compounds, likely to be responsible for the weaver ants’ deterrent effect, were identified.
This study focused on small scale farmers in Tanzania who have extensive experience of natural populations of weaver ants in their ‘Dodo’ mango trees. Interviews conducted with farmers and pickers were used to determine their perception of the impact of weaver ants on fruit fly infestation in their crop. Independent observation studies were made to verify the experiences expressed by the interviewees. Mango traders and consumers were also interviewed to better understand the impact of fruit fly infestation on the quality and quantity of fruits in the supply chain.
Eighty percent of the farmers, and all pickers and traders thought weaver ants have no positive effect on the fruit fly infestation. With a few exceptions, all traders were satisfied with the mango quality they bought, although they all lost some fruits due to fruit flies. The consumers were also satisfied with the mangoes they bought. The farmers did not view fruit flies as their largest production problem, citing premature flower and fruit drop as being of greater significance.
Not surprisingly, when questioned, the farmers could see the benefits of adopting weaver ants as biological control if more profit was to be made. However, the local supply chain that they feed into targets a high proportion of income limited consumers, in turn limiting the scope for increased prices, and the growers themselves are resource limited. Consequently, a biological control system involving managed weaver ants could be an economical challenge.
The fruit fly infestation level was examined in mangoes bought at the village market, from farmers and from a mango picker. The level of fruit fly infestation was compared in mangoes from trees with and without weaver ants. Mangoes from trees with and without weaver ants were also placed below mango trees and fruit flies’ attempts to land and oviposit on the mangoes was video filmed.
The mangoes bought at the village market had a higher infestation level compared to the mangoes bought from the farmers. The infestation level in the mangoes bought from the mango picker varied depending on the postharvest treatment; the mangoes which had been left below a mango tree for one week had a very high infestation level while the mangoes which had been packed directly after harvest were not infested.
The mangoes harvested from trees with weaver ants were infested with fruit flies, while the mangoes harvested from trees without weaver ants were not infested. The filmed fruit flies did not show any signs of being deterred by fruits from trees with weaver ants and there was no statistical difference in the number of flies landing on mangoes from trees with and without weaver ants. The number of pupae emerging from the filmed mangoes varied a lot with zero infestation in some fruits and more than 100 pupae emerging
from other fruits, indicating that other factors than the presence of weaver ants affect the fruit flies’ decision on where to oviposit.
It was not uncommon for farmers to place newly harvested mangoes below mango trees for a length of time before packing them. The effect, of this practise, on the fruit fly infestation level was studied by placing covered and uncovered mangoes below the same mango tree. Uncovered mangoes turned out to have a much higher infestation level compared to covered mangoes, regardless of whether the mangoes came from trees with small or big weaver ant populations.
In the study area, pre-maturely and maturely harvested mangoes were routinely treated with smoke from a fire to induce ripening and to reduce the fruit fly infestation level. The smoke’s effect on fruit fly infestation was tested. The number of infested fruits was very low in smoked and unsmoked fruits and there was no difference in the infestation level. However, heat produced by the ripening mangoes could not escape from the baskets in which the fruits were packed and over a few days the temperature increased to approximately 40oC which has been reported to be sufficient to kill fruit fly eggs and larvae. Thus the traditional way of packing mangoes helped to kill fruit fly eggs and larvae at an early stage so visible damage was avoided.
As fruit flies were not a major problem to the farmers, and weaver ants were not shown to be effectively deterring fruit flies, there is no great motivation for farmers to adopt weaver ants. Assuming the weaver ants could be managed in a way that made weaver ants deter fruit flies effectively there are still some economic aspects which should be studied further. It is necessary to study the farmers’ investments in relation to their expected economic benefits. The mangoes are sold domestically to traders and consumers who in many cases may not be willing to pay a higher price for higher quality mangoes. If the number of mangoes available in the market increases more than the demand it can lead to price drops and the farmers may not earn a higher income compared to the current situation.
In addition to the studies described above, volatiles emitted by weaver ant infested plants were compared with plants not infested with weaver ants. A number of known deterrent compounds were identified, but none of the compounds seemed to be related to the presence of weaver ants as they were present in similar quantities in samples with and without weaver ants. However, three compounds, which attract fruit flies, trans-beta-ocimene, cis-beta-ocimene and 2-butanol, seemed to be related to weaver ants. It is possible that hormesis is involved and that the effect of these three compounds could be deterrent under certain conditions.
In this study weaver ants were not effective in deterring fruit flies. However, in studies carried out under different circumstances, for example in Benin and Australia, weaver ants were shown to be very effective in deterring fruit flies. Further investigation is therefore needed to identify which factors are responsible for the differences in the ants’ deterring effects. Some factors which could be interesting to study are the different mango genotypes and the ant genotypes as the genes are responsible for which volatile compounds are emitted. The ant density in relation to the canopy size and mango trees chemical responses to weaver ants and scale insects would also be interesting to study.