Correlation of Single-plant Yield with Multiple-harvest yield in Pickling Cucumber

admin

Cucurbit Genetics Cooperative Report 5:14-15 (article 7) 1982

Todd C. Wehner
North Carolina State University, Raleigh, NC 27650

There are 2 basic methods for evaluating breeding material for yield: selection based on single plants, and selection based on rows (usually progeny of single plants). Single-plant selection permits the evaluation of more genotypes with a given amount of resources than progeny row selection does, and has been successful for yield improvement of maize (1). The objective of this experiment was to determine the correlation of multiple-harvest yield with single-plant yield for plants grown at 4 densities, and harvested at 2 stages. Several densities were tested because low densities aid in the separation of plants, while higher densities permit more plants to be screened in a given area.

Fifteen cultivars and lines were evaluated in a yield with 2 replications and 5 harvests made between June 15 and 29. Plots were 9 m long and were planted on 1.5 m centers at 64,400 plants/ha. The same 15 cultivars and lines were also planted in small plots at 10,600; 21,400; 64,400; and 128,900 plants/ha. Single plants were harvested at processing (10% oversize fruit) and at mature (seed harvest) stages. Multiple-harvest yield was then correlated with single-plant yield using Pearson produce-moment and Spearman rank correlation procedures. The Pearson and Spearman correlations were similar, so only Pearson correlations are presented.

There was no correlation between multiple-harvest and single-plant yield (Table1). The correlations were best for single plants harvested at processing stage and grown at 64,400 plants/ha (the same conditions under which the multiple-harvest trial was grown and harvested). However, only 2 of the 5 replications for that density and harvest stage had significant correlations, and the average correlation was not significant. Also, single-plant yield was less effective in separating the lines. While there were fairly large differences among lines for multiple-harvest yield, the differences for single-plant fruit number were smaller and the coefficients of variation (CV) were larger (Table 2).

Table 1. Correlation of multiple-harvest yield with single-plant yield at processing stage – 10% oversize fruit, and at the mature fruit stage – seed harvest, for single plants grown at 4 densities.

Density (plants/ha)

Replication

128,900

64,400

21,400

10,600
1 .48++ (.06) .40+ (.08) .30 (.55*) -.11 (-.48++)
2 .04 (-.20) .56* (-.14) .01 (.30) .11 (-09)
3 .37 (.32) .11 (-.47++) .22 (.17) .49** (-.19)
4 -.31 .10 (-.16) -.11 (.27) -.03 (.28)
5 -.20 (.04) .13 (.03) -.22 (-.07) -.09 (.07)
Mean .16 (.02) .26 (-.13) .04 (.24) .07 (-08)

**, *, ++, + Significant correlation at the 1, 5, 10, and 15% levels, respectively.

Table 2. Mean yield of 15 lines of pickling cucumber tested at Clinton, NC in 1981.

Cultivar or line

Seed source

Multiple harvest yield ($/ha)

Fruit number
Processing stage (no./plant)
Mature stage (no./plant)
Castlehy 2012 Castle 3282 2.4 3.1
Greenpak Harris 3268 3.0 3.3
G 56 D NCSU 3095 2.4 3.0
Tamor Asgrow 3008 2.9 3.4
Regal Harris 2920 2.4 3.2
Multipik Petoseed 2680 2.8 3.9
PSR 1479 Petoseed 2616 2.4 3.6
Calico Petoseed 2522 2.3 3.0
Triplemech Petoseed 2430 2.6 3.0
G 76 NCSU 2282 3.0 3.7
Tempo Harris 2159 2.7 3.7
Calypso Harris 2100 2.7 3.1
Explorer Petoseed 2040 2.7 2.7
Score Asgrow 2013 2.2 3.2
Sampson Petoseed 1793 1.9 3.2
LSD (5%) 1099 0.6 1.0
CV (%) 20 37 46

Considering the lack of correlation between multiple-harvest and single-plant yield, it appears that selection for yield should not be based on single plants. The selection unit should probably be at least a single progeny row.

Literature Cited

  1. Mareck, J. H. and C. O. Gardner. 1979. Responses to mass selection in maize and stability of resulting populations. Crop Sci. 19:779-783.