ESTIMATION OF INBREEDING EFFECTS IN AN ALFALFA BREEDING PROGRAM:
VIGOUR AND VARIATION IN HETEROZYGOSITY LEVEL BY RFLP MARKERS
Scotti C.1 ,Salvi S.1 ,Pupilli F.2 and S.Arcioni2
1 Istituto Sperimentale per le Colture Foraggere, V.le Piacenza 29, 26900 Lodi, Italy
2
Istituto per il Miglioramento Genetico delle Piante Foraggere del CNR, Via Madonna Alta 130,06128 Perugia, ItalyThe alfalfa breeding method set up at Lodi Institute provides for a phase of reduction of heterozygosity level through recurrent selfing and selection within selfed families. Some critical aspects of the selfing phase refer to: a)the knowledge of the starting point, that is the estimation of inbreeding coefficient of the parental non inbred plants; b) the effect of selection for vigour within selfed families. In fact, the additive value of genes and linkats implied in vigour and the average heterozygosity level both take part in phenotypic vigour expression (dry matter production in time). To estimate the variation of heterozygosity level during successive selfing generations as well as within each selfing generation, RFLP markers have been used.
Plant material is constituted by 6 S1 families (90 plants in total, randomly chosen within each family), 9 S2 families (166 plants in total, in two groups of opposite vigour within each progeny) derived from selection of plants in the previous generation, and 5 S3 families (39 plants in total) originated from selected S2 plants. Single plants were grown in 5 cm diameter PVC tubes (density of about 500 plants m-2), with not limiting water supply and cut every 30 days. Dry matter production of S1 and S2 families was compared with polycross families (S0 x S0) obtained from the corresponding parental plants; S3 families were compared with the corresponding S2 x S2 families. As for the estimation of heterozygosity level, plant DNA digested with Hind III was hybridised using 11 low copy number probes coming from a partial gene library of Medicago sativa (77 different fragments generated in total). The total number of fragments over the whole loci on individual plant basis was compared at every selfing level with the corresponding mother plants (MP) in the previous generations.
The total number of bands decreased from 41.17 (average of S0 MP) to 37.22 (-9.55%) in S1 and 34.38 (-15.68%) in the whole S2 families. Considering the progenies having a S3 generation, the decrease observed ranged from 42.75 (S0 MP) to 38.55 (-9.82%) in S1, 35.69 (-16.55%) in S2 and 34.78 (-19.46%) in S3. These results seem to be consistent with the reduction of 2nd and 3rd order allelic interactions in the first selfing generation and the increased frequency of diallelic genotypic structures in the subsequent ones. Inside each S2 group of opposite vigour, interplant variation for the total number of fragments is sufficient to allow the choice of high performing individuals with different estimated heterozygosity level.
No consistent correlations between estimated heterozygosity level and vigour were found. In fact, the two groups of opposite vigour within each S2 family did not differ significantly for the estimated heterozygosity level, indicating that different genotypic structures were randomly distributed in the two groups.