Walkof, C.
Double backcross breeding was used in an attempt to develop early-ripening, large-fruited rains of tomatoes which may have value as parents for F1 hybrids and pedigree crosses. The Farthest North and Early Jersey varieties were used for the initial cross. One backcross line was developed by crossing the F1 with the extra early-ripening and small-fruited Farthest North. The plants with largest fruits were selected. A second backcross line was obtained by crossing the F1, with the late, large-fruited Early Jersey. The progeny with the earliest ripe fruits were selected.
he effectiveness of the double backcross method was tested by comparing the time-of-ripening and fruit-size characters of the parents, backcrosses and selections. Time-of-ripening was measured by the heat units required above a base temperature of 59 deg. F from transplanting until the first natural ripe color showed on the fruit. Fruit size was based on weight of three representative fruits from the centre of the plant.
imitations imposed by the environment on the expression of the two characters was evident. It was difficult to evaluate the latter adequately in single plant segregates. This suggested the desirability of selecting as many plants as could be tested conveniently. It also indicated the need of accompanying selection with progeny testing.
electing for increased fruit size in the Farthest North backcross progeny afforded slight favorable advantages. A small gain over the fruit size of the recurrent parent was obtained at odds of 16 to 1. No marked improvement in early ripening was obtained in the selections from the Early Jersey backcross.
here seemed to be an association between the time-of-ripening and fruit-size characters. Selection for greater earliness appeared to be accompanied by reduced fruit size and conversely selection for larger fruit size by later ripening. This situation is thought to be due to linkage between many of the genes which govern these characters. The data of most of the selections tended to be positively correlated. However, one selection (D) from F1 x Early Jersey yielded data with a negative correlation and which approached significance. Similar tendencies were obtained when selection D was crossed with selections from F1 x Farthest North. These results suggest that selection D may have been a crossover type plant. Generally, the change in association of the two characters during the crossing procedure was small. This indicates tight linkage relations between genes for time-of-ripening and fruit-size in the varieties crossed. A small crossover percentage in the 12 linkage groups of the tomato could limit the chances of obtaining desirable recombinations of genes responsible for the two characters. Large backcross populations may help in selecting for the desired character combinations.
It was the practice to select for one character at a time in the backcross progeny. Some evidence obtained suggested the advantage of selecting both characters, in each of the backcross lines. The chances of finding crossover type plants seem greater when this is done.