ARC-Grain Crops Institute, Private Bag X1251, Potchefstroom, 2520, Republic of South Africa

Background and objectives
Ciaviceps africana, the causal organism of sorghum ergot, colonizes unfertilized ovaries, replacing host tissues with a fungal sphacelium. The typical symptom is the conidium-bearing sticky exudate which forms at the tips of infected florets. The latter produce no grain. The severity of ergot is dependent on weather during two critical phases of plant development. Low temperature 23-27 ;days preflowering reduces pollen viability. Since sorghum florets are susceptible to infection only from flower opening to fertilization, reduced pollen viability extends the susceptible period, giving the pathogen the competitive advantage for the ovary [1]. Cool, wet weather (20-28C) during days 1-4 post anthesis promotes infection and pathogen development. The extent of predisposition and rate of pollination and subsequent fertilization under disease-favourable conditions in genotypes differs, resulting in various degrees of ergot 'escape resistance'. In this study, the relationship between expected ergot severity of a flowering date (ergot potential sensu McLaren [2]) and observed ergot incidence in 60 experimental sorghum hybrids was determined and the point of escape resistance breakdown (ERB) was calculated. Pollen viability under high (>16C) to low (<13C) preflowering temperatures was also compared.

Materials and methods
Ergot severity in 60 experimental hybrids was evaluated at Potchefstroom and Bethlehem, South Africa, using the methodology described by McLaren [2]. Climatic data associated with each flowering date were collected and the respective ergot potentials were calculated according to the model of McLaren and Fiett [3]. Nonlinear regression analysis (Y=AXb) was applied to calculate the ERB point and rate of resistance breakdown in each hybrid. On four flowering dates (corresponding to higher and lower preflowering temperatures) pollen was collected from random heads and pollen viability was determined using iodine-starch analysis [1,3].

Results and conclusions
Susceptible hybrids were characterized by low ERB points and rapid rates of resistance breakdown. Pollen viability in these hybrids was reduced to less than 40% at 12.7C preflowering temperature. Four hybrids showed inherent poor pollen viability by yielding <50% viable pollen, despite a preflowering mean temperature of 16.8C (population mean=75.6%). A further four hybrids had low ERBs despite a high percentage pollen viability at low temperature. This could be attributed to floral characteristics, including gaping and male-female compatibility, which promoted infection and pathogenesis. PAN 31 was the most resistant hybrid, requiring an ergot potential of 38.35% to induce ERB. Similarly, a slower rate of ERB in this hybrid (r=0.43 per ergot potential unit) with increasing ergot potential, reduced disease severity. Pollen viability at 12.7C was 87.3% (population mean=43.2%). These variables resulted in a mean ergot severity of 9.1% in this hybrid, compared with a population mean of 34.8% over the trial period. Results emphasize the need for taking pre- and post-flowering weather into account in ergot evaluations, as well as pollen viability and other host floral characteristics. Selection of hybrids with high pollen viability, despite low preflowering temperatures, high ERBs and an ability to pollinate and fertilize rapidly, can reduce the risk of crop losses from ergot.

1. McLaren NW, Wehner FC, 1992. Journal of Phytopathology 135, 328-334.
2. McLaren NW, 1992. Plant Disease 76, 986-988.
3. McLaren NW, Flett BC, 1998. Plant Disease 82, 26-29.