CÓRDOBA NATIONAL UNIVERSITY
College of Agriculture, P.O. Box 509, 5000 Córdoba, Argentina.
H.A. Guidobaldi and R.H. Maich.
Fertility in triticale (number of seeds/spikelet) is a characteristic with a direct response to selection and a low heritability. The degree of fertility has been observed as an indirect response to previously determined ontogenic stages, such as the number and weight of seeds per spike, biological yield, harvest index, and seeds and spikes per sq. m. In wheat, changes in the number of seeds and reproductive structures of the spike have been studied widely. However, some evidence indicates that the progress in not homogeneous. Different genetic progress has been observed in the middle of the spike, especially with respect to the number of seeds per spikelet. We analyzed families of wheat with progressive microevolutionary differences. We found differences in the number of fertile florets in the middle and lower thirds of the spike and in the lower one-third for the primordial florets. Meiotic index (percentage of tetrads without micronucleus) is another triticale characteristic that has been studied in relation to its effect on the spike and spikelet fertility. A direct response to selection and an indirect response to the number of seeds per spike and spikelet were shown. The aim of our triticale research was to establish if disruptive selection for spikelet fertility and meiotic index caused an indirect response in previously determined ontogenic stages (number of primordia and fertile florets).
We evaluated 60 F2-derived families (F2:5) from a cross between two hexaploid triticales, Tatú and Don Frank, released in 1995. From the F2 plant population, we measured meiotic index and spikelet fertility. Using these as selection criteria, two groups were generated, one composed of 30 plants divergently selected with low and high meiotic index values and another 30 plants selected with low and high spikelet fertility. During 1999, the F2:5 families were grown at the Ferreyra Experimental Farm of the College of Agriculture (Córdoba National University) Córdoba, Argentina (31°29' S; 64°00' W). One-row plots, 5-m long and 0.2 m apart, with a planting rate of 250 seeds/sq m were used to evaluate populations. A complete randomized design was used, with fifteen families belonging to the same group as a replication. Five spikes per family were collected during flowering and maturity. From the first harvested sample, the number of primordia and fertile florets per spikelet were measured. From the second sample, the number of seeds and their weight per spikelet were determined. The spikes were split into lower (spikelets 4-5), middle (spikelets 14-15), and upper (subterminal) thirds. An ANOVA was performed and Duncan's Multiple Range test was used to compare treatment means.
We found significant differences between all thirds for all analyzed characters. The middle one-third showed the highest number (primordia, flowers, florets, and seeds) and weight (seeds) of the spikes. The lower one-third of the spike was second most significant. These differences could be due to the physiological and anatomical features of the spike. A significant direct response to selection was observed for the number of seeds per spikelet. Differences between the high- and low-fertility groups of 24.9 % and 24.5 % for the lower and middle thirds, respectively, were observed. A significant indirect response to spikelet fertility after disruptive selection were observed with respect to primordia florets in the lower (20.6 %) and middle (9.2 %) thirds. The high-fertility group differed from the low-fertility group in the number of fertile florets for both the lower (15.8 %) and middle (12.2 %) thirds. Between the low- and high-fertility group means, no significant differences were found in the upper one-third.
Using meiotic index as a selection criteria, significant differences in the primordia florets were observed between high- and low-fertility groups for the lower (12.9 %) and middle (8.3 %) thirds. Families in the low meiotic index group had higher mean values than the other divergent group. A similar tendency was observed for fertile florets in the lower (18.4 %) and middle (10.6 %) thirds.
Significant, positive correlations between the primordia and fertile florets in all thirds and for both selection criteria were obtained. In the middle third of the high-fertility group and in all thirds of the group with a high meiotic index, fertile florets and the number of seeds per spikelet had significant positive correlations.
In summary, a direct response to selection in the number of seeds per spikelet was not the same along the spike. At the same time, the number of primordia and fertile florets showed significant, correlated responses. The number of seeds per spikelet, which was determined in previous ontogenic stages (primordial and fertile florets) showed an inverse relationship with the meiotic index but no influence on spikelet fertility. A compensatory effect between these cytological disorders and the early spike development stages might be operating.
S.P. Gil, M.M. Cerana, L.I. Mas, and R.H. Maich.
We studied the effect of four cycles of recurrent selection for grain yield on the number of primordia and fertile florets in bread wheat. All families evaluated in the study were grown at the Ferreyra Experimental Farm of the College of Agriculture (Córdoba, Argentina, 31°29' S and 64°00' W), during two consecutive years 1998 and 1999. The families were derived from S1 or S2 plants selected in 1996. The cycles compared were C0 (initial); C1, C2, and C3 (intermediate); and C4 (most evolved). Five, main-shoot spikes from each family (12/cycle) were divided into thirds (the lower third, spikelets 4 and 5; the middle third, spikelets 9 and 10; and the upper third, subterminal) in order to determine the number of primordia and fertile florets. Results were analyzed by ANOVA and Duncan's Multiple Range Test. Statistical differences between cycles were observed in the number of fertile florets per spikelet in the middle and upper thirds, although the floret primordia differed significantly only in the middle one-third. The results of the 2-year analysis, regardless of the degree of significance, showed that the mean values of the families derived from C4 were higher than those recorded for the other cycles analyzed.
M.E. Dubois, R. Maich, and Z.A. Gaido.
Economically, bread wheat breeders focus their attention on grain yield. However, quality aspects must be considered in order to promote product differentiation in the market. Studies in continental climates have shown that there is no relationship between grain quality and yield. This is the case of the Argentine, central semiarid region, where we would like to improve seed production while maintaining protein content and bread-making quality. The objective of our work was to study the effect of three cycles of recurrent selection for grain yield on the bread-making quality in bread wheats cultivated under semiarid conditions in Argentina. Forty-eight, S-derived families, 12 from each of the analyzed cycles of recurrent selection (C0, C1, C2, and C3), were evaluated in the field at the Universidad Nacional de Córdoba for 3 years. The ANOVAs had two sources of primary variation (cycle and year) and one secondary (interaction). The values of each family from one self-cycle was used as the experimental error. Although the yield had a genetic progress of 11.4 % after three cycles of recurrent selection, significant differences were not observed in protein content, gluten, and mixogram among cycles. In general, the mixograms corresponded to those for strong wheats, with fairly short mixing times, and are considered to have good loaf volumes. More than 50 % of the same wheats have good mixing times (4-6), the appropriate relationship between dough elasticity and extensibility, and good mixing tolerance. In conclusion, recurrent selection and recombination was shown to be efficient generating an increase in grain yield after three cycles of recurrent selection without having a negative effect on the bread-making quality.
R.H. Maich, G.A. Manera, and M.E. Dubois.
Fifteen years ago, the College of Agriculture initiated a recurrent selection program for breeding bread wheat. Sixteen commercial varieties were crossed with the main objective of increasing grain yield, with six cycles of recurrent selection (2 years/cycle). Genetic progress was demonstrated after analyzing two and three cycles of recurrent selection after at least 3 years of evaluation. The present work evaluated experimental and commercial lines derived from recurrent selection for their grain yield and its components.
During 1999, two families per cycle (C0 to C4) and three commercial varieties were evaluated at Ferreyra. All plots were five rows, 5 m in length, and 0.20 m apart. Planting was at a rate of 250 seeds/m2. A randomized block design with three replications was used to study grain yield and spike number/m2, grain yield and seed numbers per spike, and kernel weight. All traits measured showed a significant effect of the material. Although, none of the experimental lines out-yielded the commercial varieties, suggesting that there was no consistent change, our data showed that the two C3-derived families had the highest values for number and kernel weight per spike. A possible explanation may be found in another factor of high yield, the number of spikes/m2, where one of the commercial varieties had the highest value for this characteristic. Several reports indicate that there is a negative relationship between these yield components. Taking account that one of the C3-derived families had a significantly higher harvest index than all the controls, we remain cautiously optimistic about the use of an recombinant selection program to improve grain yield.
R.H. Maich, G. Manera, and C. Bainotti (INTA, Marcos Juárez).
Three populations (A, B, and UNC) were evaluated for grain yield/m2. We measured a 1 % difference between the mean values of lines derived from visual plant selections performed by two technicians. Positive differences between the visual and random selections of approximately 8.6 % for A, 2.9 % for B, and 6.5 % for the UNC populations were observed. The mean values of lines derived from a random sampling showed the following differences: B versus A, 40 %; B versus UNC, 14.2 %, and UNC versus A, 22.6 %. The frequency of superior lines respect to the random sample mean value of the B population were 0 of 62 lines for the A population, 7 of 65 for the B population, and 4 of 68 for the UNC population.
The two technician-selected samples had similar agronomic performance. The effectiveness of visual plant selection was higher in the population with lowest yield potential (population A). A higher frequency of superior recombinant lines was obtained from the population with the highest yield potential (population B).
INSTITUTO DE RECURSOS BIOLÓGICOS, CNIA-INTA
CC 25 (1712), Castelar, Pcia. de Buenos Aires, Argentina.
A. Acevedo*, C.A. Kamlofski, L. Pflüger**, and E.Y. Suárez.
* Departamento de Ciencia y Tecnología, Universidad Nacional de Quilmes, Roque Sáenz Peña 180, B1876BXD Bernal, Buenos Aires, Argentina and ** Departimento di Agrobiologia ed Agrochimica, Universita degli Studi della Tuscia. Via 5. Camilo de Lellis s/n. 01100 Viterbo, Italy.
A chlorina mutant was isolated from after mutagenic treatment of seed of the Argentine bread wheat cultivar Leones INTA. Seed was soaked overnight in a 2 % solution of ethyl methanesulfonate (EMS) and grown in the field. Four spikes per M1 plant were individually harvested and seed of each spike was sown in individual rows in order to maintain lineage. Several M2 plants with a chlorina phenotype were detected in the progeny of one out of four spikes of an M1 plant. Chlorina plants had much lighter green leaves than the mother line and might be useful as genetic markers and for photosynthesis studies.
To determine the inheritance of the mutation, two-way crosses were performed between the mutant and several wheat cultivars, including the mother line. F1-hybrid plants exhibited chlorina phenotype, and a 3:1 ratio of chlorina to normal plants was observed in the F2 progeny. F1-hybrid plants also were backcrossed to Leones INTA, and the progeny revealed a 1:1 ratio of chlorina to normal plants. Taken together, these results demonstrate that EMS induced a simple inherited mutation that affects leaf color in Leones INTA. This data agree with genetic studies of other chlorina mutations, which also indicated a monogenic inheritance (Pettigrew and Driscoll 1970, Sears and Sears 1968, Williams et al. 1983).
Electrophoretic analysis of the glutenin proteins in the mutant corresponded to those of Leones INTA, confirming the isogenic nature of the mutant. Root-tip mitotic analyses of the mutant and the mother line showed that they were euploid (2n = 42) plants with no obvious chromosome abnormalities.
References.
A. Acevedo*, C.A. Kamlofski, and E.Y. Suárez.
* Departamento de Ciencia y Tecnología, Universidad Nacional de Quilmes, Roque Sáenz Peña 180, B1876BXD Bernal, Buenos Aires, Argentina.
A mutagenic treatment was made in the Argentine bread wheat cultivar Leones INTA. Seeds were soaked overnight in a 2 % solution of ethyl methanesulfonate (EMS) and grown in the field. Four spikes for each M1 plant were individually collected and seeds of each spike were planted in individual rows to maintain the lineage. M2 plants with a brachiate phenotype were observed in the progeny of one out of four spikes of an M1 plant. Because brachiate plants remained short, had tiny leaves, and were lethal, normal plants from the same progeny were crossed with Leones INTA for progeny testing. Although one-third of the crosses showed no segregation, two-thirds of the crosses had a 1:1 ratio of normal:brachiate plants in the F2 progeny. These results suggest that the normal-appearing plants used as parents in crosses that segregated carry the mutation in the heterozygous condition.
The normal-appearing plants used as parents in the aforementioned crosses also were selfed and evaluated. As expected, plants with one allele of the mutated gene had a 3:1 ratio of normal:brachiate plants. Interestingly, plant height appears to be associated with dosage of the allele at the mutated locus. The tallest plants did not have an allele of the mutated gene, the shortest plants had one allele, and brachiate plants had two alleles (homozygous recessive) of the mutated gene.
These data demonstrate that EMS induced a simple inherited mutation that affects plant height in wheat.