Of mites and men : agro-ecological factors affecting the neotropical predatory mite "Typhlodromalus aripo" DeLeon and its potential to control the cassava green mite in the mid-altitudes of Cameroon

Cassava (Manihot esculenta Crantz) is a main staple in many parts of Africa. In the processed form of gari1, it constitutes an essential income source for the rural women in certain areas. This holds certainly true for the target area of this project, the North-West Province (NWP) of Cameroon. There, one of the most important biotic constraints to cassava cropping is the cassava green mite Mononychellus tanajoa (Bondar, 1938) (Acari: Tetranychidae). To reduce cassava yield losses caused by M. tanajoa, the neotropical predatory mite Typhlodromalus aripo DeLeon, 1967 (Acari: Phytoseiidae) was first introduced into Africa in 1993. At present, T. aripo is established in 20 countries of sub-Saharan Africa. But the predator has been slow in colonizing and establishing in mid-altitude regions. This is particularly the case in latitudes above 4° North and South where temperatures are cooler and relative humidity in the dry season is lower than in the low altitudes and in areas closer to the equator. The objective of this thesis was to develop biocontrol strategies against M. tanajoa in the mid-altitudes of the NWP of Cameroon by enhancing the establishment of T. aripo.
In a field release study in the years 2002 to 2004, the population dynamics and establishment of a strain of T. aripo from the mid-altitudes (Bam strain from Minas Gerais, Brazil) was examined and compared to a lowland strain (Pir strain from Bahia, Brazil) used earlier. Both strains were released in cassava fields in the lower (600 to 850 meters above sea level) and the higher (1100 to 1300 meters above sea level) range of our mid-altitude study area in northwest Cameroon. Additionally, we conducted a screenhouse population-level experiment with the objective to determine the short-term dynamics of the Bam and the Pir strains of T. aripo and their effects on M. tanajoa populations.
In view of establishing dry season predator reservoirs, the population dynamics of predator and pest were studied in three habitat types: (a) on dry grassland hill sites with low soil fertility where cassava predominates; (b) on multiple cropping sites, which are more humid and more fertile and have a highly diverse cropping system; and (c) on riparian forest sites, which are humid and shady. Earlier studies had shown that T. aripo abundance is generally higher on cassava cultivars with pubescent apices, where T. aripo seeks refuge during the day. We introduced two cassava cultivars, one with hairy apices and one with
semi-hairy apices, and compared them in all three habitat types with a glabrous local cultivar in terms of their suitability to the predators.
In the above-mentioned studies we saw that T. aripo disappeared from the cassava apex – where it usually stays – during the dry seasons and reappears after the onset of the rains. We conducted a field enclosure experiment of cassava plants with the objectives to determine if (i) T. aripo recolonizes the cassava plant from the surrounding vegetation, if (ii) it survives in the soil or leaf litter below the cassava plant, and if (iii) T. aripo survives at very low densities in the apex.
The studies outlined before had shown that T. aripo may not be an efficient option to control M. tanajoa. To make M. tanajoa resistant or tolerant cassava cultivars available to farmers seemed to be a promising alternative approached. By means of a formal on-farm variety trial, of farmer-designed variety trials, and an assessment of farmers’ ability to differentiate between new varieties, we explored to which extent decentralized and participatory cassava variety selection are useful, and how much we can build on farmers’ own experimentation.
We found that, unlike in other regions of sub-Saharan Africa, M. tanajoa populations peaked at the end of the dry season (or shortly after the beginning of the rainy season) and not shortly after the onset of the dry season. Concurrently predator abundance dropped to very low levels in the dry season and recovered only four to eight weeks after the beginning of the rainy season. Despite the asynchrony with its prey, T. aripo was able to persist in both lower and higher altitudes for more than one year. However, in contrast to the lower altitude range, the predators were not able to persist beyond one cropping cycle in the higher altitudes, because they did not spread to neighbouring fields. The introduction of a new T. aripo strain (Bam) – which was supposed to be better adapted to the mid-altitude temperatures than the strain used earlier (Pir) – had no effect on the long-term establishment of the predators. Unexpectedly, though, the Pir strain had an advantage over the Bam strain in the first three months after release, which coincided with the dry season. This result was consistent with the result of the strain evaluation in the screenhouse. Possibly, the Pir strain, which originates from a dry area in Brazil, was better adapted to the dry season months than the Bam strain.
Typhlodromalus aripo disappeared earlier and came back later in grassland hill and riparian forest sites, as compared to the multiple cropping sites. The predator’s disappearance in the dry season was related to low ambient relative humidity in the habitat, and to plant parameters indicating low plant vigour, low apex retention, small apex diameter and poor
apex hairiness. For the predators’ return in the rainy season, plant parameters indicating strong host-plant vigour and large apex diameter were the most important factors. In the multiple cropping habitat type, the hairy cultivar was able to host the predators longer and on a higher population level than the semi-hairy and the glabrous cultivar. T. aripo dynamics in the riparian forest sites was particular, and we suspect an interaction with local phytoseiids in this habitat type.
The timing of the predator’s reappearance in the cassava apex of the different treatments of the field enclosure experiment suggests that T. aripo survives the dry season in very low densities in the cassava apex. This result is supported by additional studies: (a) In the course of a vegetation survey, T. aripo was not found on any other plant species than cassava. Instead, two new indigenous phytoseiids were discovered. (b) An assessment of the efficiency of non-destructive visual in-field apex inspections proved that about 10 % of the cassava apices that had T. aripo were not recognised as such. (c) A plant material transfer trial showed that predator-free plants could neither be infested with stems of previously infested plants, nor with leaf litter or top soil collected from the surroundings of the previously infested plants. However, there are also two studies that point to a possible role of the ground in T. aripo dry season survival: (d) A screenhouse experiment on vertical migration revealed that T. aripo sometimes does leave the cassava plant and walk over bare ground. (e) Micro-climate measurements in various cassava plant parts proved that the cassava apex and the cassava stem base are the locations with the highest relative humidity during the dry season – which makes the stem base a potentially interesting refuge.
The formal on-farm variety trial showed no differences between varieties despite the use of 11 replications. This let us suspect strong interactions between varieties and fields. Due to the high heterogeneity between farmers’ fields, decentralized selection is a prerequisite to give farmers access to the information they need to take decisions concerning new varieties. The trial also showed that heterogeneity in terms of cassava variety performance was more pronounced between fields with low mean yields (< 8.6 Mg ha-1), as compared to fields with high mean yields (> 8.6 Mg ha-1). Farmers set up their own experiments in a systematic way that allowed comparisons between the new varieties. To farmers, cassava variety testing is a long-term process. In each cropping cycle, emphasis is given to specific objectives. Selection starts earliest in the second cropping cycle. We propose a cassava variety selection scheme that largely builds on farmers’ own experience in cassava variety testing, and on their proven ability to distinguish between varieties that are new to them. The cultural habit to exchange
planting material freely with neighbours and relatives grants a fast and effective (though not systematic) dissemination of new genetic material.
In summary, this thesis shows that the potential of T. aripo in suppressing M. tanajoa populations in the mid-altitudes of northwest Cameroon is limited because of the asynchronous population cycle of predator and prey and because of the slow dispersal of the predatory mite. Cassava cultivars resistant to M. tanajoa may represent a better option to reduce pest mite damage. To introduce new cultivars to the area, a cassava variety selection and dissemination scheme is proposed that builds on farmers’ own experimentation and expertise.