UC Davis

Climate change is expected to decrease precipitation in arid California, limiting tomato production in the state. Cultivated tomato, Solanum lycopersicum, has a limited genetic base for improvement. S. habrochaites, a tomato wild relative, is water-stress tolerant and may serve as a genetic source of abiotic stress-tolerance traits. With the advent of inexpensive and rapid genotyping platforms, high-throughput genomic methods have increasingly been developed to improve the speed and efficiency of plant breeding programs. These techniques are limited due to the lack of high-quality phenotype data to accompany the abundant genomic data available. High-throughput phenotyping (HTP) technologies are being developed to effectively harness high-throughput genomic methods to accelerate plant breeding for tolerance to water-limited environments. Three experiments were performed in the field to evaluate introgression lines derived from S. habrochaites for their potential for improving cultivated tomato, and to determine the effectiveness of a HTP multispectral imaging robot for use in tomato breeding. Each of the three experiments utilized a split-plot experimental design, with main-plots each assigned to one of two water treatments, full crop evapotranspiration (ETc) or 40% ETc applied post-fruit set, and subplots assigned to different tomato genotypes. The first chapter evaluated a set of 24 introgression lines (ILs) derived from S. habrochaites and their recurrent parent and the data were used to perform bin mapping for a set of 15 traits. The second chapter used 5 ILs from the first chapter, plus 3 inbred lines to obtain a set of 15 F1 hybrids using a North Carolina Design II mating scheme. These parents and their hybrids were evaluated for combining ability for a set of 11 traits. The third chapter used two sets of tomato genotypes, a training set and a validation set, which included ILs and modern hybrid cultivars to evaluate the effectiveness of an HTP robot at collecting accurate phenotype data for a set of 11 traits. Breeding lines derived from S. habrochaites and HTP technologies have the potential to improve cultivated tomato. A total of 268 trait-genomic region associations (TGRAs) were identified among 22 of 24 ILs included in the first chapter. TGRAs were identified for each of the 15 traits. Horticulturally desirable TGRAs were identified for soluble solids content, fruit weight, degree of fruit sunburn, canopy cover, and maturity. IL LA3933 possessing an introgression on chromosome 4 from S. habrochaites may be suitable for use in developing inbred lines for a hybrid breeding program. GCA estimates of -52.55 to 75.21 were obtained for each of the 5 IL parents and 3 inbred line parents for each of the traits. Red pixel number data were successfully extracted from images collected by the HTP robot on a subset of experimental plots. A Spearman rank correlation of r = 0.76 was identified between manually collected ripe yield data and red pixel number, indicating the presence of a strong correlation. Additional data processing and analysis, possibly including the use of big data methodologies would be required to fully determine the potential of the HTP multispectral-imaging robot for tomato breeding.