Determining the best recombinant parent for yield using male-sterile testers in barley.
 M.C. Therrien and S.K. Plett
Agriculture and AgriFood Canada
Box 1000A, R.R. # 3, Brandon, Mb.
R7A 5Y3

Generally, in deciding which parents to use in a cross, a plant breeder will select parents from elite material with characteristics that are well understood. In the case of qualitative genes, the desired traits are known to be fixed and likely well-mapped genetically. Using these traits in a cross becomes relatively straightforward. In the case of quantitative traits, including yield, the breeder generally relies on the elite material's "track record". For a cultivar, that 'track record' may be replete and its' utility as a parent may be well known. However, for newly-created elite lines, and often for established cultivars, the potential of the genotype for enhancing a quantitative trait, such as yield, is not known. Obtaining an estimate of the yield potential of a given parent would be quite useful in a breeding program.

Male sterility genes in barley allow for making crosses without tedious emasculation. male sterile lines can be 'top-crossed', which can produce relatively large quantities of F1 hybrid seed with little effort. We exploited this capability and created a partial diallel cross whereby we determined General Combining Ability (GCA) and Specific Combining Ability (SCA), using parent-offspring regression, as is traditionally done in corn. Knowing these two genetic parameters, we could choose the 'best' parent for recombination of yield potential.

We originally chose six established cultivars, with a wide range of yield potential, as our 'unknown' putative parents. As our 'testers', we chose seven male-sterile lines from our male-sterile breeding effort. These lines are elite material of widely diverse genetic background and having near-isogenic male fertile sibs. We topcrossed the six cultivars onto the seven testers in a partial diallel array (reciprocals not possible). We then planted the selfed parent seed, the F1, and near-isogenics in a 4-replicate completely randomized design in the field. Each plot consisted of a single 3 m row, or 1.5 sq. m. of harvestable area. Each row was harvested for grain yield.

Results of yield analysis are summarized in Table 1, along with GCA and SCA analysis. Of the original 6 cultivars, two of these (Bedford and A286-8) did not provide sufficient seed for testing. Also, all but one tester (EX345-1) crossed almost exclusively with either the six-row cultivars (Lacombe and Conquest) or the two-row cultivars (Manley and Deuce). This suggests an incompatibility mechanism, although some seed was produced in the 'incompatible' crosses (<10% of 'compatible'). Therefore, it would seem prudent to use a relatively large array of testers. Using simple means, as well as GCA and SCA, the cultivar Lacombe was consistently identified as the parent with the best combining ability for yield. This is in agreement with yield data and experience relating to Lacombe and cultivars derived from this high-yielding six-row feed barley. This method does lend itself to utility in a breeding program, being low on labour and land requirements. It is also possible to recover useful germplasm directly from the diallel. The main drawback is that some effort would be required to establish elite testers from male-sterile germplasm. However, once established, the testers could be regenerated in perpetuo.

TABLE 1. YIELDS (T/HA), GCA, AND SCA FOR DIALLEL MALE-STERILE 
         CROSSES 1995

      PARENT:  6.0    5.6     6.4     4.3     5.6     5.0     4.4
PARENT       EX340-4 EX344-2 EX345-1 EX350-1 EX355-2 EX356-4 EX396-1
 5.6 LACOMBE   ---    ---     6.9     ---      7.0     6.4    7.6
 4.2 CONQUEST  ---    ---     6.7     ---      4.5     6.2    4.8
 4.9 MANLEY    5.5    6.3     6.2     8.1      ---     ---    ---
 5.0 DEUCE     6.5    6.2     7.3     5.9      ---     ---    ---
     MEAN      6.0    6.2     6.8     7.0      5.8     6.3    6.2
     SD        1.6    2.0     2.0     1.1      0.9     0.8    1.5

               TWO SIX
               ROW ROW
OVERALL MEAN = 5.6 5.8
OVERALL S.D. = 1.3 1.5


GCA
           EX340-4 EX344-2 EX345-1 EX350-1 EX355-2 EX356-4 EX396-1
LACOMBE      ---     ---     22.0    ---      8.5     7.1    19.4
CONQUEST     ---     ---     18.0    ---      9.9     9.6    11.4

MANLEY       14.2    22.3    14.1    18.1     ---     ---    ----
DEUCE        17.4    17.5    23.7     3.7     ---     ---    ----


SCA

TESTER:   EX340-4 EX344-2 EX345-1 EX350-1 EX355-2 EX356-4 EX396-1
SIX-ROW:    ---    ----     1.1    ----    14.0     1.1    20.2
TWO-ROW:    5.6     0.6     7.3    12.3     ---     ---    ----