Cucurbit Genetics Cooperative
Other Crop Genetics Cooperatives
Home About Membership Reports Gene Lists Conferences Links Search NCSU
Cucurbit Genetics Cooperative Report 17:125-127 (article 37) 1994

Heritability of Seed Size in Hull-less Seeded Strains of Cucurbita pepo L.

Carle, R. Bruce and J. Brent Loy

Department of Plant Biology, University of New Hampshire, Durham, NH 03824

A breeding project to develop high seed yielding strains of hull-less (naked) seeded pumpkins has been ongoing for the past twelve years at the University of hew Hampshire.

Seed size is an important trait in a hull-less seeded pumpkin, both from the standpoint of seed yields and for consumer acceptance. Because earlier investigations indicated that the highest seed yields could be obtained in small-fruited strains, we were interested in the relationship of fruit size to seed size. A recently developed hybrid out of our program, 'Snackjack', has reasonably large hull-less seed (av. 13 g/100) and relatively small fruit (0.5 to 1.0 kg). Two large-fruited, hull-less seeded plant introduction lines (PI 264241) and PI 285611) have very large seeds (23 to 26 g/100 seed). Initial attempts to transfer large seed from PI264241 into small-fruited genotypes met with failure. In the present study crosses were made between PI 285611 and a small-fruited, hull-less seeded strain (NH29-13-5, F6) with moderately large seed (av. 14 g/100) to study heritability of fruit and seed traits and the relationship between fruit and seed size.

PI 285611 was crossed to NH29-13-5 in the greenhouse, spring, 1992. F1 plants were grown in the field during summer 1994, and three selfed fruits were obtained to provide F2 seed for the genetic study. During the summer of 1993, a population of 450 F2 plants were grown in a single plot at the Woodman Research Farm in Durham,, NH. In an adjacent plot 10 plants of PI 285611 and 25 of NJ29-13-5 were grown; the additional plants of NJ29-13-5 were needed for other breeding purposes. We attempted to obtain as many selfs as possible on F2 plants so that most selections could be carried to the F3 generation. For the heritability study we were able to obtains seed samples from fruits of 250 F2 plants; environmental variances were computed from data on 5 fruits and seeds of PI 285611 and 9 fruits and seeds of NJ29-13-5. Data were taken on fruit weight and diameter. Additional seed samples were obtained from small-fruited, large-seeded selections that were not included in the heritability study because they were not randomly selected. Seeds were dried thoroughly on screens at ca. 30 ˚ C in a forced air dryer, and the following measurements were taken on 5 randomly sampled seeds per seed lot per fruit: seed weight/100, seed thickness, seed width and seed length, using a digital caliper. The data presented here represent only the first 100 fruits sampled to date.

Heritability of fruit weight was relatively low (61.7%). This was not unexpected given the fact that even within a plant there is considerably variation in fruit size, depending upon the fruit load the vigor of the plant. Heritabilities for seed weight, length and thickness were 76.4, 78.7, and 85.2%, respectively, A lower heritability for seed width (68.1%) can probably be ascribed to the effect of different degrees of expression of the hull-less trait on seedcoat development along the seed margins.

Seed thickness and fruit weight were not correlated (R2 = 0.05); both parents had fairly thick seeds. SEed weight and fruit weight (R2 = 0.38) and seed length and fruit weight (R2 = 0.46) were both positively correlated, but to a moderate degree. Therefore, it was possible to obtain small-fruited selections with fairly large seed. However, only one selection out of 450 F2 plants had both small (1.2 kg) fruit and a seed weight (25.4g/100) similar to PI 285611. All seed size traits showed transgressive segregation (Table 1); whereas, no small-fruited genotypes were recovered in the F2 generation approaching the size of NH29-13-5, the small-fruited parent. This was unexpected because fruits and seeds of F1 plants grown in 1992 were about as large as PI 285611. This indicates the existence of heterotic effects for fruit and seed size in the above cross. It would be interesting to know if heterosis for seed size would be displayed in crosses of two small-fruited strains, one with large and the other with small seeds.

The prospects for selecting small-fruited strains of pumpkin with very large seeds appear good. Heritability of seed size appears to be high and is possibly underestimated in this study because of the small sample size of parents and likely greater variance displayed for the environmental component. It should be noted, however, that we attempted to eliminate fruit likely to have poor seed fill by not including fruit from plants that looked diseased or from fruit not deemed to be fully developed. A major question to be answered is whether small fruit with large seed will show lower total seed yields because of greater early seed abortion. Larger seeds are larger nutrient sinks; therefore, carbohydrates and other resources translocated to a small developing fruit may be limiting for allocation to the large number of fertilized ovules (300 to 600) within a fruit.

One final phenomenon worth mentioning is that seed size is almost totally determined by maternal factors, and most probably by genetically determined differences in size of the seed coat. Histochemical and biochemical studies of normal and hill-less seed coats during embryogenesis indicate that the seed coat is a transient storage organ for reserve materials that appear to be later utilized by the developing embryo (1, 2). In addition, size of the seed coat may mechanically limit expansion of the embryo within it. Whatever the mechanism, seed size differences are not evident in F2 seed taken from F1 plants, but rather the differences occur in F3 seed harvested from F2 plants.

Table 1. Fruit and seed size attributes for PI285611 and NH29-13-5 pumpkin cultigens and an F2 population derived from them.

Range
Seed and fruit variables1
Mean
Variance
High
Low
Fruit weight (kg)

PI.1285611 (5 fr)

5.17
1.86
7.50
4.00

NH29-13-5 (9 fr)

0.61
0.01
0.75
0.45

F2

3.44
2.43
8.39
1.13
Seed Weight (g/100)

PI1285611

23.62
3.73
25.94
20.18

NH29-13-5

14.27
2.57
16.74
11.56

F2

17.88
13.37
27.76
10.38
Seed length (mm)

PI285611

18.56
0.32
19.35
17.76

NH29-13-5

14.93
0.26
15.76
14.32

F2

16.05
1.38
18.98
13.52
Seed Width (mm)

PI1285611

10.14
0.33
10.59
9.10

NH29-13-5

8.96
0.15
9.59
8.28

F2

9.14
0.77
12.84
7.57
Seed Thickness (mm)

PI285611

2.32
0.02
2.45
2.06

NH29-13-5

2.28
0.01
2.38
2.14

F2

2.44
0.09
3.03
1.81

 1 Values for seed parameters represent average measurements of 5 seeds randomlysampled from seed lots obtained from individual fruits.

Literature Cited

  1. Stuart, S.G. and J.B. Loy. 1983. Comparison of testa development in normal and hull-less seeded strainsof Cucurbita pepo L. Bot. Gaz. 144:491-500.
  2. Stuart, S.G. and J.B. Loy. 1988. Changes in tests composition during seed development in cucurbita pepo L. Plant Physiol. (Life Sci. Adv.) 7:191-195.
Home About Membership Reports Gene Lists Conferences Links Search NCSU
Department of Horticultural Science Box 7609North Carolina State UniversityRaleigh, NC 27695-7609919-515-5363
Page citation: Wehner, T.C., Cucurbit Genetics Cooperative;
Created by T.C. Wehner and T. Ng, 1 June 2005; design by C.T. Glenn;
send questions to T.C. Wehner; last revised on 11 December, 2009