1Centro Nacional de Technologia Agropecuaria y Forestal, Apdo. 885, San Salvador, El Salvador; 2CIMMYT, Lisboa 27, Apdo. Postal 6-641, C.P. 06600, Mexico, D.F. Mexico; 3Natural Resources Institute, University of Greenwich, Central Avenue, Chatham Maritime, Kent ME4 4TB, UK

Background and objectives
In most Latin American countries, maize is a major food and animal feed crop usually produced by small-scale subsistence farmers with little chemical input and low technology. An important disease contributing to low yields is corn stunt which is a complex of three pathogens, namely maize bushy stunt (MBS) phytoplasma, Spiroplasma kunkelii and maize rayado fino virus (MRFV). Economic losses due to the corn stunt complex (CSC) or its components have been enormous in Central America. Consequently, CSC is in some areas the most important target for maize germplasm improvement programs.

Germplasm evaluations have to date been carried out based on field observation of symptoms. However, symptoms induced by the CSC pathogens do not appear to be diagnostic of the presence of specific pathogens, with environmental factors, maize genotype and combination of pathogens influencing symptom expression. Knowledge of the distribution and prevalence of individual CSC pathogens in the screening nurseries of Central America will assist the breeding programmes. Highly sensitive, specific and robust techniques are needed in order to achieve this, and one of the objectives of this project was to evaluate existing, and develop new, diagnostic techniques such that they were appropriate for use in Mexico at CIMMYT.

Results and conclusions
The suitability of three ELISA formats, namely F(ab')2 protein A ELISA, protein A ELISA and plate trapped antigen (PTA)-ELISA, using specific antisera was assessed. Of the three assays, the F(ab') 2 protein A ELISA was found to be the most reliable.

For S. kunkelii, an oligonucleotide PCR primer pair was developed to amplify part of the spiralin gene. The PCR assay was specific for detecting this mollicute and was at least five times more sensitive than the F(ab')2 protein A ELISA. RT-PCR for detection of MRFV was also found to be specific for detecting MRFV in field and greenhouse samples. However, it was felt that because of the reliability of the F(ab')2 protein A ELISA for detection of MRFV, the extra cost and time in performing RT-PCR was not justified for routine use at CIMMYT.

For detection of MBS phytoplasma, attempts to produce a good polyclonal antisera were unsuccessful due to cross-reactions with some field-collected samples. Due to the lack of specificity of this experimental antisera, a specific PCR [1] was used for detecting this mollicute in maize samples. At CIMMYT, both the PCR for S. kunkelii and that for MBS phytoplasma were used to detect infections in greenhouse plants. The PCR for S. kunkelii was 97% efficient at detecting the pathogen in greenhouse plants, with the non-detected 3% being attributed to the presence of inhibitory compounds in some DNA extracts.

The diagnostic tests were used to identify the pathogens in maize samples collected at maize screening trials in Central America during 1995-96. Data collected represents the first report of the presence of S. kunkelii and MBS phytoplasma in El Salvador, Guatemala and Panama verified by diagnostic tests. The most prevalent pathogen was S. kunkelii either in single or mixed infections with the other CSC components. MBS phytoplasma was detected mostly in association with S. kunkelii. The presence of MRFV in the area, although low in relation to the other CSC pathogens, is significant because the losses induced by this virus can be high if susceptible germplasm is infected at an early stage of development. Symptoms were shown to be poor indicators of the presence of CSC pathogens.

1. Harrison NA, Richardson PA, Tsai JH, 1996. Plant Disease 80, 263-69.