Comparison of two methods for determining tomato fruit size.
Walkof, C.
Fruit size of plant varieties and hybrid seedlings is an important character when breeding tomatoes in the north temperate zones. Small fruit size often associated with early fruit ripening is an undesirable combination. In an endeavor to breed for an early and large-fruited tomato, two methods of evaluating fruit size are available to the hybridizer. One that has been used by variety trial plot operators for many years is to count and weigh all the ripe fruits harvested. The total weight is divided by the number of fruits to get the individual fruit weight data. This method is tedious. It takes time to count the fruits and calculate individual fruit weights. It is also subject to error when inexperienced labor must be used for counting. The second method, suggested by Dr. T. M. Currence of St. Paul, is to weigh three representative fruits from the centre of each plant tested. Although some error may be anticipated when selecting the three representative fruits per plant, it has been observed that the accuracy obtained is greater than with the first method. In 1953 and 1954 the two methods were compared at Morden, Manitoba, using markedly different types of breeding material. A practically identical order of decreasing fruit size was obtained for the hybrids and varieties included in these tests. Table 1 shows a comparison of sixteen F1 hybrids and standard varieties. The individual fruit size data are quite similar for the two methods. This is supported by a highly significant correlation coefficient. Table 2 gives data relating to breeding material with markedly contrasting fruit size of the parent varieties. In spite of somewhat variable fruit size data between the two methods, the correlation coefficient is highly significant. This variability may be due to error in counting the large numbers of fruit. This emphasizes the perference for Method 2.
TABLE 1. Evaluation of fruit size of ripe tomatoes harvested from 16 F1 tomato hybrids and comparable standard varieties in 1954. _________________________________________________________ Hybrid Method 1 or No. of Mean individ- Rank in variety fruits ual fruit wt. relation to counted in grams fruit size __________________________________________________________ BB3^1 184 243 1 Bounty^1 144 182 2 H004 367 178 3 H001 254 172 4 D001 121 166 5 H003 321 165 6 H006 424 161 7 D003 153 157 8 Mustang 424 152 9 Meteor^1 235 150 10 Monarch 478 148 11 H005 412 139 12 H002 254 124 13 W024MD^1 232 115 14 Early Chatham^1 473 112 15 W024MP^1 193 102 16 __________________________________________________________ L.S.D. at P.05 26 _________________________________________________ Hybrid Method 2 or Mean individ- Rank in variety ual fruit wt. relation to in grams fruit size __________________________________________________ BB3^1 239 1 Bounty^1 175 2 H004 178 3 H001 174 4 D001 164 5 H003 157 6 H006 156 7 D003 153 8 Mustang 152 9 Meteor^1 149 10 Monarch 148 11 H005 143 12 H002 133 13 W024MD^1 129 14 Early Chatham^1 111 15 W024MP^1 110 16 __________________________________________________ L.S.D. at 19 Note - Method 1- consisted of counting, weighing the fruits and then dividing the weight by the number obtained. Method 2-consisted of weighing 3 representative fruits per plant of each hybrid and variety. Correlation coefficient for Individual Fruit Weights comparing Methods 1 and 2, r = .989** (1) - standard varieties. The hybrids and varieties were grown in 5 plant plots randomized in 6 replicated blocks. Data obtained in 1954. TABLE 2. Evaluation of fruit size of ripe tomatoes harvested from the parents, backcrosses, and backcross selections of a small-fruited x large-fruited tomato cross. ______________________________________________________________ Method 1 ___________________________________ Parent, cross Gener- No. of Mean Rank in or ation fruits individual relation selection counted fruit wt. to fruit in grams size ______________________________________________________________ Farthest North P1 4202 8 9 Early Jersey P2 157 114 1 Pi x P2 F1 2740 27 5 Fl x Pl Bc1 3213 18 8 Fl x P2 Bc1 759 51 2 Sel. 1 from F1xp1 S1 2596 21 7 Sel. 2 from F1xP1 S1 3009 22 6 Sel. 1 from F1xP2 S1 926 32 4 Sel. 2 from F1xP2 S1 332 36 3 ______________________________________________________________ L.S.D. at P.05 7 ___________________________________________________ Method 2 ________________________ Parent, cross Gener- Mean Rank in or ation individual relation selection fruit wt. to fruit in grams size ___________________________________________________ Farthest North P1 14 9 Early Jersey P2 118 1 Pi x P2 F1 41 5 Fl x Pl Bc1 29 8 Fl x P2 Bc1 75 2 Sel. 1 from F1xp1 S1 40 6 Sel. 2 from F1xP1 S1 37 7 Sel. 1 from F1xP2 S1 47 4 Sel. 2 from F1xP2 S1 51 3 __________________________________________________ L.S.D. at P.05 9 Note - Method 1 consisted of counting, weighing the fruits and then dividing the weight by the number obtained. Method 2 consisted of weighing three representative fruits per plant of each parent, cross and selection. Each parent, cross and selection was grown in a randomized block using 25 replicated single plant plots. Data obtained in 1953. Correlation coefficient for Individual Fruit Weights comparing Methods 1 and 2, r = .984**. After removing parent data so as to compare more homogeneous progeny data, r = .980**.