2018 ASHS Annual Conference

The germplasm of many crops including tomato is narrow due to constraints imposed during domestication, thus increasing the importance of wild relatives as a source of genetic variation. However, segregation distortion, linkage drag and reproductive barriers impede the transfer of desirable traits from the wild to cultivated species. This project was undertaken to understand segregation distortion and reproductive barriers in the F₂ and backcross (BC) populations from the interspecific cross between tomato, S. lycopersicum, and the wild species, S. pennellii.

The three populations derived from crosses between S. lycopersicum M82 and S. pennellii LA716 and their interspecific F₁ were evaluated for % pollen stainability and % seed germination as well as if plants produced flower, fruit or seed. Percent stainable pollen varied greatly for the parents: 99.22% M82, 16.31% LA716 and 85.7% of the interspecific hybrid, F₁. Both backcross populations had similar means, 57.9% BCLyc and 48.9% for BCPen, while the F₂ mean was 31.7% but varied from zero to 99.2%. The F₂, as expected from previous research had only 27.75% of the population producing fruit, of which, only 9.43% produced seed. As expected, all of the BCLyc flowered and 98% produced fruit with germinable seed. Conversely, the majority of the BCPen produced flowers, but only 8.7% of the plants produced self-fruit and only 3.26% produced seed. Only two populations, F₂ and BCLyc exhibited any abnormal seed germination (18.8% and 1.1% respectively), with all of the germinated seed from the F₂ being abnormal.

Of the 1000 SNP markers, 181 had no data and an additional 184 had no variation or were incorrect, leaving 635 markers used for analysis. The F₂ population had 43% of the markers deviate from the expected 1:2:1 ratio. The BCLyc population had over 63% of their markers deviating from the 1:1 ratio with excess tomato alleles and all of the markers on two chromosomes skewed. The BCPen population, only 21.9% displayed distorted segregation and one chromosome did not skew from the expected ratio. In addition, most of the skewing BCPen markers favored homozygous LA716 combination, but a few skewed towards the heterozygous combination on chromosomes 5 and 9. Each population was examined for the percent of S. pennellii genome present with 50% expected for the F1 and 54.4% observed. The two backcross populations were projected to have 25% (BCLyc) and 75% (BCPen), our analysis found an average of 17.8% for BCLyc and 77.6% for BCPen respectively.