1Crop Research Institute, Kumasi, Ghana; 2University College, Gower Street, London WC1E 6BT, UK

Background and objectives
Cassava is the most important source of carbohydrate for approximately 500 million people living in tropical Africa. It is subject to many diseases, of which that caused by African cassava mosaic virus is probably the most devastating. In 1989, a new disease was reported from Zaire and given the name stem-tip dieback (STDB), but attempts to define the causal agent were unsuccessful [1]. Observations in Ghana suggested that the disease was widespread. The objectives of the investigation were therefore (i) to identify the causal agent and prove its involvement in the disease; (ii) to determine the incidence of the disease in Ghana; and (iii) to describe attributes of the pathogen that may contribute to its virulence.

Materials and methods
Stems of cassava plants with dieback symptoms typical of the disease were collected from the field, cut into sections and surface sterilized before plating on potato dextrose agar. The fungus that grew from the sections was subcultured and examined under a light microscope. To extract DNA, mycelium and spores were scraped from the surface of agar cultures and used as inoculum for shake cultures on a medium consisting of Czapek-Dox nutrients supplemented with V8 juice. After incubation for 48 h, mycelium was collected, frozen in liquid nitrogen and ground to a powder. DNA was extracted from the powder using the CTAB method. Domain 2 of the 28S subunit of rDNA was polymerized using the primers Pn 9 (5' CTTAAGCATATCAATAAGCGGAGG 3') and Pn2 (5' GTTCACCATCTTTCGGGTCG 3') and standard techniques. The product was sequenced using the primer Pn4 (5' CCTTGGTCCGTGTTCAAGACGGG 3'), and the sequences compared with those of other species of Colletotrichum, using the programme MEGA [2].

A survey of the disease was conducted on 25 farms in each of the cassava-growing regions of Ghana: Ashanti, Brong Ahafo, Central and Western, by scoring 100 plants randomly selected for presence or absence of the disease and rating the severity on a 1-5 scale.

Serial two-fold dilutions of culture filtrates of the fungus, grown on Czapek-Dox medium supplemented with cassava tuber extract or citrus pectin, were tested for their toxicity to cells isolated from cassava leaves using fluorescein diacetate as a vital stain. Culture filtrates were fractionated by acetone precipitation and partitioning of the supernatant against ethyl acetate after removal of the acetone. Ethyl acetate preparations were separated on silica gel by flash chromatography. Culture filtrates were also tested for their pectate lyase activity.

Results and conclusions
Spores, and appressoria which developed from them on germination, corresponded in morphology to those of the genus Colletotrichum. Koch's postulates were completed with the organism, proving that it was the cause of STDB. Comparison of the sequence of 193 base pairs of domain 2 of the 28S ribosomal subunit with the same region from other species of Colletotrichum showed that the causal agent of STDB was C. gloeosporioides, since it differed from other authentic isolates of this organism by only 1 base pair.

Disease incidence varied in the four cassava-growing regions of Ghana from 24-61%, and severity from 2.0-3.4 on the 1-5 scale.

Culture filtrates of the fungus were toxic to cells of cassava. On acetone precipitation, approximately equal amounts of activity were found in both precipitate and supernatant but their sum exceeded that of the original by 70%. About one-third of the activity of the supernatant partitioned into ethyl acetate and this was separated by flash chromatography into three active compounds. Electron-impact mass spectrometry showed that these shared a prominent peak at m/z 355 and several others, suggesting that they were a family of similar compounds.

1. Muimba-Kankolongo A, Simba L, Singh TP, Muyolo G, 1989. Agriculture, Ecosystems and Environment 25, 151-164.
2. Sherriff C, Whelan MJ, Arnold GM et al., 1994. Experimental Mycology 18, 121-138.