University of California, Department of Vegetable Crops, One Shields Avenue, Davis, California 95616-8746, USA

Background and objectives
Late blight, caused by Phytophthora infestans, and black mould, caused by Alternaria alternata, are two serious fungal diseases affecting cultivated tomato, Lycopersicon esculentum. Genetic resistance in cultivated tomato is limited, but wild Lycopersicon species contain a wealth of genetic diversity, including resistance to these two pathogens. Resistance to late blight (LB) and black mould (BLM) has been detected in wild L. hirsutum and L. cheesmanii, respectively. lnterspecific populations were created between cultivated tomato and L. cheesmanii (recombinant inbred) and L. hirsutum (reciprocal backcross) and used to conduct genetic inheritance and quantitative trait loci (QTL) mapping studies [1] on resistance to LB and BLM. Highly replicated experiments over time using several disease phenotype traits, types of disease assays, and multiple isolates were used to assay large segregating progeny populations and locate QTL most consistently associated with resistance. Selected QTL conferring resistance to LB and BLM have been targeted for introgression into cultivated tomato using marker-assisted selection (MAS) [1, 2] in conjunction with a modified backcross-single seed descent method in two separate populations. Selected backcross-selfed (BCxSx) lines representing different combinations of resistance QTL are being (BLM) or will be (LB) used to determine the effect of QTL on the disease resistance phenotype in a common genetic background, produce resistant breeding lines, and evaluate the efficacy of MAS for QTL introgression in tomato.

Results and conclusions
Resistance was found to be quantitatively inherited, with 11 and 14 QTL contributing to LB and BLM resistance, respectively. Heritability was low (BLM) to moderate (LB). Certain QTL were detected with two or more isolates, indicating QTL associated with broad-based (general) resistance to LB and BLM were identified. Resistance to both pathogens appeared to be associated with leaf surface (LB) or fruit surface (BLM) properties conferred by the wild parent. For example, inoculum droplet dispersal QTL were highly associated with LB resistance QTL, suggesting that this trait contributes to resistance. QTL on chromosomes 2, 3, 9 and 12 (a total of 5 BLM QTL) and 4, 5, 6 and 11 (a total of 4 LB QTL) have been selected for introgression into cultivated tomato. Selected first backcross (for LB) or F7 recombinant inbred lines (for BLM) were been used as pollen donors in two separate crossing programs to cultivated tomato to generate two backcross-single seed descent populations. RFLP markers associated with each selected QTL were converted to PCR-based SCAR markers and used on the BCx and BCxSl generations to select plants containing the desired combinations of resistance QTL for each disease in each of the two populations. In the BCxSl generation, different homozygous combinations of 4 (in the BC3S1 for LB) and 5 (in the BC2S1 for BLM) resistance QTL are selected and self-pollinated. Replicated field experiments will be conducted over multiple locations to assess disease resistance and horticultural traits of the BC2S2 (for BLM) in 1998 and BC3S2 (for LB) lines (projected for 2000) representing different homozygous QTL genotypic classes. The mean performance of lines in each QTL genotypic class will be compared to determine the effect of QTL, alone and in combination, on the disease resistance phenotype for each disease. Superior BC2S2 (for BLM) and BC3S2 (for LB) breeding lines that are resistant and horticulturally acceptable will be selected, and efficacy of MAS for QTL introgression will be assessed.

1. Lee M, 1995. Advances in Agronomy 55, 265-344.
2. Michelmore R, 1995. Annual Review of Phytopathology 15, 393-427.