1CIMMYT South Asia Regional Program, PO Box 5186, Kathmandu, Nepal; 2Centro de Investigación Agricola Tropical, Ejército Nacional 131, Casilla 247, Santa Cruz, Bolivia; 3CIMMYT, Lisboa 27, AP 6-641, 06600 Mexico D.F., Mexico

Background and objectives
Spot blotch of wheat caused by Bipolaris sorokiniana (Sacc. in Sorok.) Shoem., is a major disease constraint in the humid subtropics [1], in particular in the rice-wheat system that covers more than 12 million ha in the Indian subcontinent. Yield losses up to 20% are observed in commercial fields. Sources of genetic resistance to spot blotch are limited, and under conditions of high temperature and relative humidity favouring rapid and severe epidemics, small amounts of quantitative resistance are not easily detected. For germplasm screening, the assessment of differences in disease progress may be preferred to one disease rating in the season or to the highest disease score. Another key problem with selection for spot blotch resistance is the negative correlation that exists between heading date, plant height, and disease severity [2]. As a result, it is difficult to select early maturing resistant genotypes. Late maturing wheat is not desirable in warm areas of South Asia due to the risk to be exposed to heat stress during grain development [2]. A study was conducted to assess the use of area under disease progress curve (AUDPC) estimates for scoring entries grouped by maturity and plant height. The method may allow a more efficient screening of genotypes resistant to spot blotch with a desirable maturity.

Materials and methods
Twenty-six wheat genotypes classified into four groups according to maturity (early-late) and plant height (tall-short) were studied in Poza Rica, Mexico (Veracruz, 21°N, 60 masl). In this location, warm and humid conditions cause severe natural spot blotch development every year. A randomized complete block design (4 replicates) was used with maturity/height groups as main plots and genotypes as sub-plots (3x0.75 m). Five tillers were marked in each plot and percentage diseased leaf area (%DLA) was assessed on the flag (F) and F-1 leaf at three dates after most plants had flowered; AUDPC was calculated. A 00-99 scale was also used to assess disease severity. Yield loss and thousand grain weight (TGW) loss were studied after comparison with yield data obtained in an adjacent field with the same experiment but treated with tebuconazole every 10 days after tillering.

Results and conclusions
Spot blotch severity was higher in early genotypes, the group with early and short entries being more severely damaged. Also, in all genotypes the disease was more severe on F-1 than F leaf showing the disease progression from lower to upper parts of the plants. High disease severity was largely associated with early maturity and to a lesser extent with low plant height. In all groups except one (short and early) genotype comparisons based on %DLA showed significant differences (P<0.05) between entries. However, using AUDPC estimates, significant differences between genotypes were detected within all groups. In the group of early maturing short entries, genotypes Batuira and Chil/Chumai appeared significantly more resistant than NL 297 and Sonalika, two popular early wheat varieties grown in South Asia. In none of the four groups, the slope of the logit line calculated from %DLA scores was significantly different between entries. Yield losses and TGW losses were significantly (P<0.01) correlated with AUDPC and highest disease score (%DLA or 00-99 scale) but regression coefficients were very low suggesting that other factors than spot blotch determined yield. Results indicated that using AUDPC after grouping genotypes according to maturity and height helps in selecting early genotypes with higher spot blotch resistance.

1. Dubin HJ, Rajaram S, 1996. Annual Review of Phytopathology 34, 503-526.
2. Sharma RC, Dubin HJ, Devkota RN, Bhatta MR, 1997. Plant Breeding 116, 64-68.