CABI Bioscience, Silwood Park, Ascot, Berks. SL5 7TA, UK

Background and objectives
Cacao (Theobroma cacao L.) is the source of cacao powder and butter and a major economic crop for many tropical countries. The crop can be seriously constrained by pests, causing severe losses to the harvest and hence to income. Two of the major disease problems in South America are frosty pod rot caused by Moniliophthora roreri Evans et al and witches' broom caused by Crinipellis perniciosa (Stahel) Singer. As the world's consumption of cacao products increases control of these diseases becomes more imperative, and development of a biological control strategy for their management would be advantageous, particularly in view of increasing concern about pesticide residues in cacao beans [1].

Screening of microbial biological control agents on whole diseased plants is potentially a complex and time consuming exercise. A long period is required for pod development, and the size and number of plants necessary makes it impractical in non-cacao growing countries. Plate screening can give results which may not be relevant to whole plants or indeed to the field situation. This is especially true for C. perniciosa where the in vitro mycelium represents the saprophytic and not the invasive parasitic phase of the life cycle [2]. Undifferentiated cacao callus has been used previously to confirm the dual mycelium theory for C. perniciosa [2], and recent advances in tissue culture techniques offered the opportunity to study the development of both C. perniciosa and M. roreri on differentiated cacao tissues in an in vitro system.

Materials and methods
Callus and embryos were produced according to the somatic embryogenesis method (M. Guiltinan, The Pennsylvania State University, USA. pers. comm.). Spores of M. roreri were inoculated onto the callus or the embryo using a fine needle. Spores were inoculated onto plates without callus as a control. All plates were sealed and kept in the dark for the remainder of the test.

Results and conclusions
In several examples, the conidia of M. roreri germinated and the resultant mycelium overgrew the callus in a dikaryotic state and spread onto the media. However, growth and sporulation of the fungus were significantly enhanced compared to that on the control plates. In addition, the morphology of fungal growth appeared to be related to the cacao type from which the callus was developed. In other samples, the fungus grew much more slowly and the mycelial colonisation was internal rather than external. This is considered to represent the monokaryotic parasitic phase [2]. Inoculation with basidiospores of C. perniciosa led to similar results. These results could lead to the possibility of developing a rapid screen for testing resistance of cacao germplasm collections.

1. Woods GAR, Lass RA, 1985. Cocoa. Longman Scientific and Technical.
2. Evans HC, 1980. Transactions of the British Mycological Society 74(3), 515-523.