Seasonal and habitat variability in the fungal pathogens, neozygites cf. floridana and Hirsutella thompsonii, associated with cassava mites in Benin

A survey of the pathogenic fungi associated with mites on cassava in Benin, West Africa, revealed both geographical and seasonal variation in the presence of Neozygites cf. floridana (Weiser and Muma) and Hirsutella thompsonii Fisher on Mononychellus tanajoa (Bondar) and Oligonychus gossypii (Zacher). Few dead and infected mites were found during the dry season, regardless of vegetation zone. In three of 30 surveyed sites, N. floridana was found infecting 1% of the dead M. tanajoa and 2% of the dead O. gossypii, while H. thompsonii was observed infecting 20% of the dead M. tanajoa in a single site. The frequency of sites having infected mites during the wet season was 3.5 times greater than that seen during the dry season. N. floridana infected 10% of the dead M. tanajoa and 19% of the dead O. gossypii on young leaves. Mites infected with N. floridana were found either in the coastal Southern Forest Mosaic (SFM) or in the Northern Guinea Savanna vegetation zones. N. floridana was rare in the low mite densities associated with mature leaves. H. thompsonii was found on 19% and 29% of the dead M. tanajoa on young and mature leaves respectively. All M. tanajoa infected with H. thompsonii on young leaves and mature leaves (75%) were found in the SFM. A single M. tanajoa was the only infected mite found in the Southern Guinea Savanna. Relatively few O. gossypii were infected with H. thompsonii. N. floridana and H. thompsonii were found together in three sites, but never on the same host. Phytoseiids were never found infected with either pathogen. In a regression analysis, the number of dead mites was significantly estimated from the total number of mites for both species, regardless of leaf age. The numbers of dead M. tanajoa on mature leaves were also estimated from the proportion infected with H. thompsonii. The numbers of infected mites on young leaves were estimated from their association with the SFM for M. tanajoa infected with H. thompsonii, and from total mites for O. gossypii infected with N. floridana. On mature leaves, infected mite numbers were estimated from the numbers of dead M. tanajoa infected with H. thompsonii. The merit of introducing more virulent or better adapted isolates of N. floridana to control M. tanajoa in Africa is discussed.