PNL Volume 14 1982

RESEARCH REPORTS

UTILIZATION OF THE HALF-SEED TECHNIC IN THE BREEDING OF HIGH PROTEIN

VARIETIES OF PEA. I. GENETIC ANALYSIS OF LINES AND THEIR HYBRIDS WITH

RESPECT TO CHOOSING CROSS COMBINATIONS FOR BREEDING

Kaczmarek, Z. and M. Surma

Swiecicki, W. K.

Institute of Plant Genetics, Poznan

Poland

Plant Experiment Station, Wiatrow, Poland

Earlier investigations showed the potential of using the half seed

method of estimating protein content (PNL 13:54-56). We intend to use

this technic in genetic-breeding work towards developing lines of pea

with high protein content. In practice it will involve selection of

single seeds on the basis of the analysis of protein content in a half

of a seed without the embryo.

On the basis of the earlier studies the following four lines were

chosen for further investigation (with the average protein content):

Mt 4042 - 29.2%, Wt 3527 - 24.4%, Wt 3026 - 29.2%, and L.X. - 25.6%.

These lines were crossed in the following combinations: L.4042 x L.3527,

L.4042 x L. 3026 , L. 4042 x L.X, L.3527 x L. 3026 , L.3527 x L.X,

L.3026 x L.X. For each cross combination, the protein content was

analyzed (micro-Kjeldahl) in halves of seed without the embryo in the

parental F1, F2, and backoross populations.

With such information it is possible to get general genetic-

breeding information concerning the whole set of lines studied and their

hybrids (diallel analysis of the F1 and generations) and more

detailed genetic information concerning each of the cross combinations

(analysis of the generations P1, P2, F1, F2, B1, B2). The estimation of

general combining ability, obtained on the basis of the diallel

analysis, were similar for both generations, F1 and F2. The high

protein lines L.4042 and L.3026 had significant positive effects whereas

lines L.3527 and L.X. showed significant negative effects.

The diallel analysis revealed an additive-dominant model of gene

action for both F1 and F2. Thus, it was possible to obtain an estima-

tion of genetic parameters according to Mather, i.e. D,F, h², H₁, H₂, and

to estimate dominance and broad and narrow sense heritabi Fities

(Table 1).

PNL Volume 14

1982

RESEARCH REPORTS

The diallel analysis showed a significant dominance for low protein

content, but the degree of dominance was different in different cross

combinations. It varied from 1.06 (L.4042 x L.X.) to 8.47

(L.4042 x L.3026). The highest value was obtained in crosses involving

two high protein lines, L.4042 and L.3026, indicating a genetic dif-

ference between them in spite of the similarity in protein content in

the lines theemselves. On the basis of the results of diallel analysis

and the analysis of particular cross combinations two hybrids,

L.4042 x L.3026 and L.3026 x L.3527, were chosen for further breeding.