Molecular methods for detecting the coconut lethal disease (LD) phytoplasma in Tanzania
Lethal disease (LD), a phytoplasma lethal yellowing-type disease of coconut palms, is the major threat to coconut cultivation in the coastal areas of Tanzania. Two molecular approaches have been developed for early and accurate disease diagnosis. Random fragments of LD phytoplasma DNA were generated as probes for pathogen detection. LD DNA extracted from infected coconut tissue was randomly fragmented and cloned into pUC 18. Selected recombinants were labelled with DIG-dUTP and used as probes in dot-hybridizations with total DNA from LD infected palms. The probes hybridized strongly to DNA from infected palms, but there was also a significant level of background hybridization to DNA from healthy palms. The second technique used oligonucleotide primers for conserved regions of the 16S rRNA gene and variable spacer regions between 16S and 23S rRNA genes in the polymerase chain reaction (PCR). Amplification of phytoplasma rDNA was primed from LD-infected palms in Tanzania, Kenya and Mozambique, and no amplification products were obtained from healthy coconut tissue. By use of these techniques infection could be reliably detected in the spear leaves and root tips of affected palms thereby avoiding destructive palm sampling. The pathogen was found in all meristematic tissues with highest concentrations of phytoplasmas in the petioles of young unemerged leaves, the area below the growing point and the root tips in palms with moderately advanced diseases symptoms. Root tips proved reliable for sampling when compared to spears, and are now recommendet do be sampled together with the spears in routine, non-destructive sampling. Phytoplasmas could be detected in symptomless palms one month before the onset of disease symptoms by use of DNA probes and two months before by PCR, when spear leaves were sampled monthly from 180 randomly selected palms for a year. Of the 24 palms which subsequently developed disease LD was detectable in 25% prior to the onset of disease and in 46% at the time disease symptoms were visible. In 29% of these palms, phytoplasmas were not detected at all. No phytoplasmas were detected in any of the palms which remained healthy. The genetic relatedness of the LD phytoplasma to twelve different non-coconut infecting phytoplasmas, two spiroplasmas and phytoplasmas causing LYD in Kenya, Mozambique, Ghana, Florida and Jamaica were investigated. The LD DNA probe did not hybridize to any of the non-coconut infecting phytoplasmas and spiroplasmas. However, the probes detected a strong genetic relationship to all the LYD phytoplasmas. By use of PCR analyses, the phytoplasma causing LYD in Kenya was not found to differ from LD, but the pathogen causing LYD in Mozambique was found to be different. This appeared to be more closely related to the LYD phytoplasmas in West Africa. Studies on auchenorrhinchous insects in LD infected coconut fields revealed a strong relationship between seasons and insect flight into the fields. They also showed that local environmental conditions have a strong influence on vector populations, and may be indirectly responsible for the differences in disease incidence observed in different regions of the country. The flight pattern of auchenorynchous insects in general and of Diartrombus mkurangai in particular , coincided with the pattern of disease spread implying that this species is the most probable vector of LD. A good correlation obtained when the disease incidence data was regressed on the numbers of Diastrombus mkurangai and Meenoplus spp, but not on the total number of trapped auchenorynchous insects provided additional evidence or implicating these species as vectors of LD. More than 5000 individual insects were analysed by PCR in attempts to identify the insect vector or vectors for LD. PCR products of the right size were amplified from a few individuals of the species Diastrombus mkurangai and Meenoplus spp, and were shown to be LD phytoplasma by RFLP analysis. The techniques have provided a quicker and more reliable means of detecting LD phytoplasmas in coconut tissues and in putative insect vectors than the conventional methods. Possible improvements on the techniques are suggested and the prospects of utilising them to find a sustainable method of disease control discussed.