Printer Friendly

Diallel analysis of carbon isotope discrimination and its association with forage yield among nine historically recognized alfalfa germplasms.

GENETIC IMPROVEMENT of WUE is an important goal or many plant breeding programs in the arid southwestern USA. Whether selection is based on yield per se, or on physiological traits, the ultimate goal is to increase a crop's harvest index per unit of water applied. Selection for low A has been proposed as a physiological approach to improve WUE in [C.sub.3] crop species. The association between WUE and A has been attributed to variation in the ratio of internal leaf C[O.sub.2] concentration ([C.sub.i]) to ambient C[O.sub.2] concentration ([C.sub.a]), where low [DELTA] values resulting from low [C.sub.i]/[C.sub.a] reflect a higher C[O.sub.2] assimilation to water transpiration ratio (i.e., gas exchange WUE; Farquhar et al., 1989; Hall et al., 1994). Johnson and Tieszen (1994) reported that A and shoot WUE were negatively correlated (r = -0.63 to -0.73) in alfalfa. This association reflects the fact that the C[O.sub.2] assimilation to stomatal conductance ratio is inversely related to [DELTA].

Variation for A among and within cultivated and exotic accessions suggests that opportunities exist to improve [DELTA] in alfalfa. Ranges of 0.6 to 1.4 [per thousand] have been reported for [DELTA] among diverse accessions of M. sativa subsp. sativa and nothosubsp. varia (Martyn) Arcang. (Johnson and Tieszen, 1994; Ray et al., 1998). A range of 0.8 [per thousand] was detected for [DELTA] among 78 winterhardy North American cultivars (Johnson and Rumbaugh, 1995), and a range of 0.9 [per thousand] among half-sib families of an elite alfalfa breeding population (Ray et al., 1999a, 1999b). Evidence clearly indicates that [DELTA] should be amenable to manipulation by traditional alfalfa breeding techniques. Narrow-sense heritabilities for [DELTA] under both irrigated and water-stressed conditions were 0.56 (Ray et al., 1999a, 1999b), while broad sense heritabilities based on individual plants exceeded 0.80 in alfalfa (Johnson and Rumbaugh, 1995). Johnson and Rumbaugh (1995) reported significant GCA effects, but not specific combining ability (SCA), for [DELTA] in a diallel among 14 clones from the NC-83-1 germplasm (Kehr et al., 1975). The lack of significance for SCA in that study may reflect the fact that NC-83-1 traced to only winterhardy accessions that were randomly intermated for two cycles. Opportunities to detect SCA and/ or heterosis effects for [DELTA] may be greater in hybrids derived from distinct populations adapted to a wider range of climatic conditions.

The effectiveness of selecting for physiological traits to improve WUE will depend on their correlation with a crop's harvest index under field conditions. Negative correlations between shoot biomass and [DELTA] were reported in four cool-season grasses (Johnson and Bassett, 1991). Dry matter yield and [DELTA], however, were positively correlated among nine alfalfa germplasms and 30 elite half-sib families grown in nonstressed and water-stressed field environments (Ray et al., 1998, 1999a, 1999b). Evaluation of nine alfalfa germplasms under irrigation indicated that populations with low [DELTA] tended to have slower growth and development rates than germplasms with high [DELTA] (Ray et al., 1998). However, elite half-sib families differing in [DELTA] had similar maturation rates (Ray et al., 1999a, 1999b). Evidence also indicates that the ranking of alfalfa accessions for [DELTA] may vary throughout the growing season (Johnson and Rumbaugh, 1995).

Nine germplasm sources commonly referred to as African, Chilean, Flemish, Indian, Ladak, M. falcata, M. varia, Peruvian, and Turkistan have been recognized as primary initial contributors to contemporary North American alfalfa cultivars (Barnes et al., 1977). Diallel analysis among these nine germplasms previously detected high parent heterosis for forage yield, ranging from -33 to 23%, as well as significant GCA and SCA effects (Segovia-Lerma et al. (2005). Our objective for this study was to analyze forage samples collected across multiple harvests during the forage yield diallel analysis to determine the influence of additive and nonadditive gene action on [DELTA], changes in [DELTA] affiliated with previously observed yield responses, and the behavior of A across harvests.


Accessions representing each of nine alfalfa germplasm sources: African, Chilean, Flemish, Indian, Ladak, M. sativa subsp, falcata (hereafter referred to as M. falcata), M. sativa nothosp, varia (hereafter referred to as M. varia), Peruvian, and Turkistan (Barnes et al., 1977) were represented by 30 genotypes as described by Segovia-Lerma et al. (2005). The nine germplasms were intermated by hand (without emasculation) in a half-diallel mating design to produce 36 [F.sub.1] hybrid populations. The [F.sub.1] populations were generated by reciprocally crossing each plant within a germplasm to one other randomly selected plant from each of the other eight germplasms. Thus, 60 genotypes (30 from each population) contributed to each [F.sub.1] population. Parental populations were synthesized by randomly intercrossing all 30 genotypes within a given germplasm. An equal number of seed, within each reciprocal cross from each plant, was bulked to form balanced composite populations for each inter- and intracross population.

The 36 [F.sub.1] hybrid populations, the nine parents, and four check cultivars (Dona Ana, Wilson, Commercial 1, and Commercial 2) were planted during March 1996 using a randomized complete block design with three replications. Each population was planted in three-row plots, 1.5 m long, and seeded at a rate of 300 seed [plot.sup.-1]. Rows within plots were spaced 30 cm apart, and plots were spaced 60 cm apart. Control plots of the cultivar Dona Ana were established between the entry plots to minimize interplot interactions. Plots were sown on a Glendale sandy clay loam (fine-silty, mixed, superactive, calcareous, thermic Typic Torrifluvent, pH 8.0) at the Leyendecker Plant Science Research Center near Las Cruces, NM, USA. Before planting, plots were fertilized with 122 kg [ha.sup.-1] of phosphorous. No additional fertilizer was applied to the plots after establishment. Irrigation management followed that used by local commercial hay producers, where plots were flood irrigated with approximately 7 cm of water every 14 d from 15 April to 15 October during 1996 to 1998. No data were collected during the 1996 establishment year.

Carbon isotope discrimination and dry matter yield were determined on 30-d-old regrowth during the second week of May, June, and July in 1997 and 1998. Johnson and Rumbaugh (1995) reported that any alfalfa plant part could be sampled to determine [DELTA]. Consequently, [DELTA] was determined on 60 shoots (approximately 50 g DW) that were randomly sampled from each plot immediately before forage harvest and dried at 60[degrees]C for 48 h. The entire shoot sample was sequentially ground through a 1-mm screen with a shear mill (Model 4; Thomas-Wiley Corp., Philadelphia, PA) and an impact mill (Model SF; UDY Corp., Boulder, CO). Each sample was mixed extensively, and carbon isotopic composition values ([per thousand]) were determined using an isotope ratioing mass spectrometer (Biology Dep., Augustana College, Sioux Falls, SD) according to the procedures of Farquhar et al. (1989). Data were expressed as the ratio of [sup.13]C/[sup.12]C relative to the PeeDee belemnite standard and converted to A as described by Farquhar et al. (1989). Forage yield (kg [ha.sup.-1]) was determined by clipping and weighing foliage from each plot at a 5-cm height using a flail harvester. Forage yield data were adjusted to a dry matter basis by subsampling approximately 300 g of fresh forage from each plot and drying it at 60[degrees]C for 48 h.

Heterogeneity in soil texture resulted in spatial variation within replicates. The data were adjusted for field trend effects using nearest neighbor analysis via the "second-difference approach" (Besag and Kempton, 1986; Stroup et al., 1994), as provided by Agrobase Software (Agronomix Software, Inc., Portage la Prairie, NB, Canada). Adjusted A and biomass data were analyzed within and across years using Analysis III of Gardner and Eberhart (1966), as described by Murray et al. (2003). Analysis of variance across the 2 yr was initially conducted as a split-split-plot in time. Entries were considered as the whole-plot factor, the three harvest dates were the split plot, and years were the split-split plot. Significant entry x harvest interactions were detected for [DELTA]; therefore, additional analyses were conducted within each harvest, as a split-plot in time, where entries were the whole plot and years the split plot. To determine the performance of the experimental populations relative to the four checks, the data were first analyzed using a standard ANOVA that ignored the diallel arrangement. When considering the entry diallel arrangement, the check cultivars were excluded and entries were partitioned into parents/varieties and crosses. Crosses were partitioned into GCA and SCA according to Murray et al. (2003).

Diallel effect estimates and the difference of the estimates from zero were obtained by using PROC GLM (SAS Institute, 1989). The standard error of the diallel-effect estimates and their significance (i.e., different from zero) were obtained using a t statistic generated by PROC MIXED (SAS Institute, 1992). The effects of varieties (parents), average heterosis, GCA, and SCA were estimated based on contrasts between parent and cross means via ESTIMATE statements in GLM or MIXED. Each contrast (C) was constructed as a linear function of the plot observations:

C = [[SIGMA].sup.t.sub.j=1] [c.sub.j] [[bar.y].sub.j],

where c represents the contrast coefficients subject to [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] represents the sample mean of entries (parents or crosses), and t denotes the number of entries involved in the contrast. Simple linear regression models were used to determine the influence of GCA and parent effects per se (as independent variables) on hybrid performance (dependent variables) for and yield. Results from these regressions were reported as correlation coefficients.


Our initial goal for this diallel study was to monitor forage yield potential and heterosis among these populations under commercial management conditions (Segovia-Lerma et al., 2005). Thus, irrigation scheduling was managed to prevent severe moisture stress among the plots. Secondarily, we attempted to determine changes in A that accompanied hybrid yield response. The expense associated with determining [DELTA] precluded our ability to analyze samples from all harvests. Thus, we focused on the first three harvests of each year, which typically contribute about 60% of the total seasonal biomass in a seven-harvest system commonly practiced in southern New Mexico. Higher yields and forage quality also accompany the first two harvests (compared with later harvests), providing growers with the greatest economic returns. Environmental conditions during the three harvests of this study are summarized in Table 1. Temperatures were similar within each harvest across years. Most rainfall events were <2 mm in both years, indicating that most precipitation did not effectively contribute to soil moisture. The largest precipitation event in each year was 8 and 11 mm, which occurred on 6 June 1997 and 10 June 1998, respectively. Mild water-stress symptoms (e.g., slight wilting of the youngest terminal internodes) was observed late in the day during the 2 or 3 d before the June and July harvest in each year. Such responses are not uncommon at this time of day and year in southern New Mexico because high evapotranspiration and temperatures can facilitate acute but temporary moisture stress on a daily basis. Plots were able to recover full turgidity overnight.

Performance of entries for [DELTA] was similar across years within each of the three harvests (data not shown). However, analyses of variance and rank correlations for [DELTA] across years and harvests indicated that entries performed differently (P < 0.001) in the May harvest, as compared with the June and July harvests. Entries performed similarly in June and July. Significant differences were detected among parents and their crosses for [DELTA] in each harvest (Table 2). Average heterosis (varieties vs. crosses) was not significant for this trait in any harvest. Parents differentially transmitted genetic effects influencing [DELTA] to their progeny, as demonstrated by the significant GCA effect across all harvests. The SCA effect was significant in the May harvest only. Evaluation of the data according to Analysis II of Gardner and Eberhart (Murray et al., 2003) indicated that variation for parental/variety heterosis and midparent heterosis was not significant (data not shown). These results confirm that additive effects are of major importance in determining [DELTA]. The significance of interactions between main effects and years in Analysis III varied between harvests (Table 2). The higher residual mean square encountered in July reflected higher variances for A in M. falcata, Ladak, and M. varia, and their respective hybrids in 1998 (data not shown).

Values for [DELTA] from parents, their hybrids, and four check cultivars are presented in Table 3 based on the May harvest and the average of the June and July harvests. In these harvests, the check cultivar Wilson, previously developed for improved production under deficit levels of irrigation (Melton et al., 1989), possessed the highest [DELTA] value among the checks. The [DELTA] value for Wilson was also similar to that of the diallel hybrids possessing the highest [DELTA] in each harvest. The range for [DELTA] among the nine parents was greatest in the July harvest (1.4 [per thousand]), and was similar to that reported by Johnson and Tieszen (1994) among alfalfa accessions collected from geographically diverse regions. The range for A among the 36 hybrids was also greatest in July (0.8 [per thousand]) and similar to that reported among 78 winterhardy U.S. populations (Johnson and Rumbaugh, 1995).

Forage yield of parents, their hybrids, and four checks are presented in Table 4. Phenotypic associations indicated that higher-yielding populations tended to have higher [DELTA] (r = -0.45, not significant; r = 0.96, [alpha] = 0.01; and r = 0.96, [alpha] < 0.01 for parents; and r = 0.37, [alpha] = 0.05; r = 0.66, [alpha] = 0.01; and r = 0.74, [alpha] = 0.01 for hybrids in May, June, and July, respectively). A higher [DELTA] among higher-yielding entries is reasonable if increased yields reflect greater carbon fixation rates, and hence, greater stomatal conductance. Internal leaf carboxylation efficiency, working with conductance, may also be an important factor causing variation in [DELTA] among entries. It is noteworthy that [DELTA] for the two overall highest yielding hybrids in our study (African x Peruvian and Chilean x Peruvian) did not differ significantly from either those entries with highest [DELTA], or for those of several M. falcata hybrids.

Results of the regression analyses are summarized in Table 5 as the correlation coefficient between actual and predicted values for A and forage yield. Regression models based on parental GCA effects were superior in their predictive capacity for both hybrid traits compared with models based on variety (parental per se) effects. The [DELTA] of parents in May was not a good predictor for [DELTA] of their respective hybrids, indicating that progeny testing would be required during this harvest. In June and July, both GCA and per se effects for [DELTA] were useful predictors of hybrid [DELTA] and forage yield. Predicted hybrid forage yield based on parental per se A in May was negatively correlated with actual hybrid yield values. This association was positive in both the June and July harvests.

Variety effects for [DELTA] in Ladak, M. varia, and particularly those for M. falcata, became increasingly negative as the growing season progressed (Table 6). Variety effects for [DELTA] of populations containing primarily subspecies sativa germplasm increased across harvests, particularly those for Chilean, Indian, and Peruvian. Actual [DELTA] values for all hybrids and parents increased across harvests, with the exception of M. falcata, where [DELTA] decreased across harvests.

The greatest increase observed for GCA effects across harvests occurred in the nondormant populations African, Indian, and Peruvian. The magnitude of the decrease observed for GCA effects across harvests was similar for Ladak, M. falcata, and M. varia. In June and July, only those germplasms which contained significant M. falcata contribution demonstrated negative [DELTA] GCA. Populations possessing essentially subspecies sativa parentage possessed positive GCA effects during the same period.


Cumulative data from the current study, and those of Johnson and Rumbaugh (1995) and Ray et al. (1998) likely provide reasonable estimates for potential [DELTA] variability available among many U.S. cultivars. Our results suggest that fundamental differences in stomatal conductance may exist between subspecies sativa and subspecies falcata types. Given that most North American populations contain predominately subspecies sativa germplasm, with moderate contributions from M. varia and minor contributions from Ladak and M. falcata, it appears that modest opportunities exist to improve [DELTA]. Of the nine parents, M. falcata appeared to offer the greatest potential to reduce [DELTA] in arid southwestern production environments. However, yield penalties associated with the hybrids from this parent may limit its immediate usefulness. Plant breeders are not necessarily restricted to using M. falcata types if they wish to reduce [DELTA]. Johnson and Tieszen (1994) reported a range for A among accessions of subspecies sativa from the National Plant Germplasm System, that was as great as what we detected between subspecies sativa and subspecies falcata accessions.

The merit of expanding [DELTA] evaluation beyond North-American-based populations was emphasized by the positive association between [DELTA] and yield in the current study, and in previous research (Ray et al., 1998, 1999a, 1999b). We recognize, however, that these studies utilized relatively few germplasms. Thus, the reported relationships between these two traits may not be representative of M. sativa as a whole. For example, of 18 subspecies sativa accessions evaluated for [DELTA] by Johnson and Tieszen (1994), 14 were independently evaluated for the crop descriptor, "fall regrowth height," as provided by the Germplasm Resource Information Network of the National Plant Germplasm System (www.; verified 15 June 2004). With data provided by these two resources, we detected no association between [DELTA] and fall regrowth (r = 0.31; P > 0.28). If fall regrowth height provides a crude estimate of yield potential for a given environment (McKenzie et al., 1988), these results suggest that yield penalties may not be as severe for low [DELTA] populations as published reports may indicate. Two of the 18 accessions (PI 434600 and PI 430636), which possessed low [DELTA] values in the study of Johnson and Tieszen (1994), were also evaluated for forage yield during 3 yr in southern New Mexico. They ranked first and third, respectively, for yield among 88 alfalfa core collection accessions and two checks (Vernal and Spredor II) that possessed limited fall regrowth height (Ray, 2000, unpublished data). In another diallel study, the fall-dormant PI 434600 ('Fortin Pergamino' from Argentina) possessed the highest GCA for yield among nine parent populations selected from the alfalfa core collection. These nine parents were comprised of three populations that were selected for high per se yield from within each of three general fall dormancy classes, dormant, semidormant, and nondormant (Ray, 2003, unpublished data). Some PI 434600 hybrids were the highest yielding in the study, and equaled or exceeded the performance of five elite check cultivars.

Given that [DELTA] was influenced primarily in an additive fashion in our study, and appears to be moderately heritable, it would be worthwhile to evaluate recurrent selection strategies as a means to develop elite breeding populations with reduced [DELTA]. The most direct approach would be to practice selection for improved yield within populations already possessing relatively low [DELTA], but good yield potential (e.g., PI 434600 or perhaps Peruvian, which both demonstrated high GCA for forage yield but relatively low per se [DELTA]). Monitoring [DELTA] in each breeding cycle would determine if gross shifts in [DELTA] accompany selection for improved yield.

Alternatively, the positive association observed between [DELTA] and hybrid yield indicated that selection for high [DELTA] may be warranted in flood-irrigated production environments of the arid southwest. Values for [DELTA] affiliated with the overall two highest-yielding hybrids suggest that opportunities to develop high-yielding hybrids with moderate [DELTA] may also be possible. Parental GCA effects consistently provided the best prediction of hybrid [DELTA]. However, in harvests experiencing conditions conducive to high evapotranspiration, variety/parental effects also provided reasonable estimates of hybrid [DELTA]. The apparent sensitivity of [DELTA] to harvest environments indicated that this trait should be monitored throughout the growing season, particularly in those harvests that possess the greatest differences in environmental growth conditions (e.g., early, mid-, and late-season harvests).

Abbreviations: [DELTA], carbon isotope discrimination; GCA, general combining ability; SCA, specific combining ability; WUE, water-use efficiency.
Table 1. Means for carbon isotope discrimination across nine
alfalfa parents and 36 diallel hybrids, and associated
environmental variables, during three harvests and 2 yr.

                            Mean daily ambient

          Carbon isotope    1997        1998        Precipitation

Harvest    1997    1998   Max.  Min.  Max.  Min.    1997    1998

               %                   [degrees]C           mm

May       19.11   19.46     27     9    26     6      12       1
June      19.84   19.81     32    14    32    11      21      11
July      19.80   20.48     35    15    36    16       3       8

Table 2. Mean squares for carbon isotope discrimination (%o)
from a diallel analysis among nine alfalfa germplasms during
May, June, and July harvests across 2 yr near Las Cruces, NM.

                                         Mean square

Source                       df  May        June        July

Blocks (B)                    2  0.067      0.080      0.032
Entries (E)                  44  0.174 *    0.182 **   0.534 **
  Parents/varieties (V)       8  0.164 *    0.395 **   1.217 *
  Varieties vs. crosses (h)   1  0.100      0.051      0.082
  Crosses (C)                35  0.179 **   0.137 **   0.391 **
    GCA                       8  0.367 **   0.309 *    1.139 **
    SCA                      27  0.123 *    0.086      0.169
E x B                        88  0.076 **   0.057 *    0.135
Years (Y)                     1  7.857 **   0.058      31.577 **
E x Y                        44  0.047 *    0.061 **   0.138
  V x Y                       8  0.070 *    0.064 *    0.254
  [bar.h] x Y                 1  0.017      0.392 **   0.005
  C x Y                      35  0.042      0.050 *    0.115
  GCA x Y                     8  0.070 *    0.072 *    0.152
  SCA x Y                    27  0.034      0.044 *    0.104
Residual                     90  0.031      0.028      0.159
CV, %                            0.92       0.84       1.90

* Significant at [alpha] = 0.05.

** Significant at [alpha] = 0.01.

([dagger]) Diallel conducted according to Analysis III of Gardner
and Eberhart (1966), as reported by Murray et al. (2003). GCA,
general combining ability; SCA, specific combining ability.

Table 3. Carbon isotope discrimination of nine alfalfa parents
(diagonal) and their diallel hybrids (off diagonal). Data are
presented as that obtained from the May harvest (above diagonal,
and diagonal underlined) and the average of the June and July
harvests (below diagonal, and diagonal in parentheses) across 2 yr.

Germplasm                   African     Chilean     Flemish


African                       19.36       19.16       19.14
Chilean                       20.14       19.17       19.49
Flemish                       20.07       20.21       19.20
Indian                        20.01       20.17       20.26
Ladak                         20.00       19.97       19.88
M. falcata                    19.94       19.83       19.89
M. varia                      19.97       20.19       19.79
Peruvian                      20.03       20.01       20.17
Turkistan                     20.02       20.13       19.90
Mean [DELTA] of crosses       20.02       20.08       20.02

Mean [DELTA] of four checks in
 June in July                             20.17
Range of checks                           20.10-20.32
LSD (0.05)                                0.29

Germplasm                    Indian      Ladak     M. falcata


African                       18.99       19.38       19.18
Chilean                       19.43       19.47       19.29
Flemish                       19.45       19.26       19.31
Indian                        19.12       18.99       19.04
Ladak                         19.87       19.26       19.21
M. falcata                    19.61       19.63       19.42
M. varia                      19.92       19.77       19.64
Peruvian                      20.18       20.13       19.84
Turkistan                     20.12       19.92       19.59
Mean [DELTA] of crosses       20.01       19.89       19.74

Germplasm                   M. varia    Peruvian


African                       19.21       19.31
Chilean                       19.62       19.29
Flemish                       19.23       19.41
Indian                        19.16       19.17
Ladak                         19.62       19.53
M. falcata                    19.34       18.98
M. varia                      19.38       19.25
Peruvian                      19.96       18.90
Turkistan                     20.00       20.32
Mean [DELTA] of crosses       19.90       20.08

Germplasm                  Turkistan   of crosses

African                       19.33       19.21
Chilean                       19.24       19.37
Flemish                       19.26       19.31
Indian                        19.15       19.17
Ladak                         19.45       19.36
M. falcata                    19.11       19.18
M. varia                      19.59       19.37
Peruvian                      19.38       19.29
Turkistan                     19.36       19.31
Mean [DELTA] of crosses       20.00

Mean [DELTA] of four
 checks in May                19.30
Range of checks               19.10-19.44
LSD (0.05)                     0.32

Table 4. Dry matter yield of nine alfalfa parents (diagonal) and their
diallel hybrids (off diagonal). Data are presented as that obtained
from the May harvest (above diagonal, and diagonal underlined) and
the average of the June and July harvests (below diagonal, and diagonal
in parentheses) across 2 yr.

Germplasm               African     Chilean     Flemish

                                   kg [ha.sup.-1]

African                     2224        2263        2032
Chilean                     1774        2471        2775
Flemish                     1760        2188        1809
Indian                      1823        1886        1834

Ladak                       1825        1878        1191

M. falcata                  1673        1500        1088

M. varia                    1799        1927        1553

Peruvian                    2328        2228        1915

Turkistan                   1858        2021        1902

Mean yield of crosses       1855        1925        1679

Mean yield of four checks in            2148
 June and July                          2039-2436
Range of checks                          279
LSD (0.05)

Germplasm                Indian      Ladak     M. falcata

                                   kg [ha.sup.-1]

African                     2032        2211        2076

Chilean                     2563        2716        2387

Flemish                     2300        1713        1116

Indian                      2225        1655        1811
Ladak                       1316        1795        1722
M. falcata                  1179         919         254
M. varia                    1451        1180         916

Peruvian                    1990        1739        1251

Turkistan                   1736        1508        1256

Mean yield of crosses       1652        1444        1223

                                                           Mean yield
Germplasm               M. varia    Peruvian   Turkistan   of crosses

                                         kg [ha.sup.-1]

African                     2058        2699        2304        2209

Chilean                     2681        2973        2511        2609

Flemish                     2030        2566        2386        2115

Indian                      2067        2353        2379        2145

Ladak                       1869        2632        2239        2095

M. falcata                  1231        2136        1583        1758

M. varia                    1942        2377        2128        2055
Peruvian                    1754        2170        2631        2546
Turkistan                   1614        2016        2126        2270
Mean yield of crosses       1524        1902        1739

Mean yield of four checks in May                    2597
Range of checks                                     2422-2931
LSD (0.05)                                           450

Table 5. Correlation coefficients between actual and predicted
values for carbon isotope discrimination (DELTA) and forage yield
of 36 diallel hybrids (dependent variable), as determined by
linear regression analyses based on general combining ability
(GCA) and per se performance of parents (independent variables).

                                       Parental GCA

                                Carbon isotope
Hybrid trait                    discrimination   Forage yield

                                        May harvest

Carbon isotope discrimination   0.68 **          0.27
Forage yield                    0.35 *           0.88 **

                                       June harvest

Carbon isotope discrimination   0.72 **          0.68 **
Forage yield                    0.89 **          0.93 **

                                       July harvest

Carbon isotope discrimination   0.82 **          0.77 **
Forage yield                    0.85 **          0.90 **

                              Parental per se performance

                                Carbon isotope   Forage
Hybrid trait                    discrimination   yield

                                        May harvest

Carbon isotope discrimination   -0.04            0.26
Forage yield                    -0.59 **         0.70 **

                                       June harvest

Carbon isotope discrimination   0.60 **          0.64 **
Forage yield                    0.72 **          0.79 **

                                       July harvest

Carbon isotope discrimination   0.72 **          0.65 **
Forage yield                    0.78 **          0.72 **

* Coefficients are significantly different from zero at [alpha] = 0.05.

** Coefficients are significantly different from zero at [alpha] = 0.01.

Table 6. Estimates of diallel effects for varieties/parents (V),
general combining ability (GCA), mean of varieties ([[mu].sub.v]),
and mean of crosses ([[mu].sub.c]), and their respective standard
errors (in parentheses) for carbon isotope discrimination in each
of three harvests across 2 yr.

                         May harvest

Effect         Estimate        Effect        Estimate

                  %o                            %o

[V.sub.1]       0.12          GC[A.sub.1]    -0.09 **
[V.sub.2]      -0.07          GC[A.sub.2]     0.10 **
[V.sub.3]      -0.04          GC[A.sub.3]     0.03
[V.sub.4]      -0.12          GC[A.sub.4]    -0.13 **
[V.sub.5]       0.02          GC[A.sub.5]     0.09 **
[V.sub.6]       0.18 **       GC[A.sub.6]    -0.12 **
[V.sub.7]       0.14 **       GC[A.sub.7]     0.10 **
[V.sub.8]      -0.34 **       GC[A.sub.8]     0.00
[V.sub.9]       0.12          GC[A.sub.9]     0.03
SE             (0.07)                        (0.03)
[[mu].sub.v]   19.25          [[mu].sub.c]   19.29

                        June harvest

Effect         Estimate        Effect        Estimate

                  %o                            %o

[V.sub.1]       0.23 **       GC[A.sub.1]     0.03
[V.sub.2]       0.10          GC[A.sub.2]     0.08 **
[V.sub.3]       0.06          GC[A.sub.3]    -0.01
[V.sub.4]       0.18 **       GC[A.sub.4]     0.06 *
[V.sub.5]      -0.10          GC[A.sub.5]    -0.06 *
[V.sub.6]      -0.60 **       GC[A.sub.6]    -0.18 **
[V.sub.7]      -0.10          GC[A.sub.7]    -0.05
[V.sub.8]      -0.07          GC[A.sub.8]     0.08 **
[V.sub.9]       0.22 **       GC[A.sub.9]     0.06 *
               (0.06)                        (0.02)
[[mu].sub.v]   19.85          [[mu].sub.c]   19.81

                        July harvest

Effect         Estimate        Effect        Estimate

                  %o                            %o

[V.sub.1]       0.37 *        GC[A.sub.1]     0.08
[V.sub.2]       0.34 *        GC[A.sub.2]     0.16 **
[V.sub.3]       0.14          GC[A.sub.3]     0.12 *
[V.sub.4]       0.26          GC[A.sub.4]     0.04
[V.sub.5]      -0.33 *        GC[A.sub.5]    -0.12 *
[V.sub.6]      -1.05 **       GC[A.sub.6]    -0.34 **
[V.sub.7]      -0.05          GC[A.sub.7]    -0.11
[V.sub.8]       0.08          GC[A.sub.8]     0.16 **
[V.sub.9]       0.25          GC[A.sub.9]     0.00
               (0.15)                        (0.06)
[[mu].sub.v]   20.10          [[mu].sub.c]   20.15

* Effects are significantly different from zero at a = 0.05.

** Effects are significantly different from zero at a = 0.01.

([dagger]) Diallel effects of parent germplasms designated as
1 = African, 2 = Chilean, 3 = Flemish, 4 = Indian, 5 = Ladak,
6 = M. falcata, 7 = M. varia, 8 = Peruvian, and 9 = Turkistan.


Research supported by a United States Department of Agriculture grant (#99-34186-7496) to the Southwest Consortium on Plant Genetics and Water Resources, and the New Mexico Agriculture Experiment Station. We also thank the anonymous reviewers of this manuscript for their useful comments and perspectives.


Barnes, D.K., E.T. Bingham, R.P. Murphy, O.J. Hunt, D.F. Beard, W.H. Skrdla, and L.R. Teuber. 1977. Alfalfa germplasm in the United States: Genetic vulnerability, use, improvement, and maintenance. USDA-ARS Tech. Bull. 1571. U.S. Gov. Print. Office, Washington, DC.

Besag, J., and R. Kempton. 1986. Statistical analysis of field experiments using neighboring plots. Biometrics 42:231-251.

Condon, A.G., and R.A. Richards. 1993. Exploiting genetic variation in transpiration efficiency in wheat: An agronomic view. p. 435-450. In J.R. Ehleringer et al. (ed.) Stable isotopes and plant carbonwater relations. Academic Press, San Diego, CA.

Farquhar, G.D., H.R. Ehleringer, and K.T. Hubick. 1989. Carbon isotope discrimination and photosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40:503-537.

Gardner, C.O., and S.A. Eberhart. 1966. Analysis and interpretation of the variety cross diallel and related populations. Biometrics 22: 439452.

Hall, A.E., R.A. Richards, G.D. Wright, and G.D. Farquhar. 1994. Carbon isotope discrimination and plant breeding. Plant Breed. Rev. 12:81-113.

Johnson, D.A., and M.D. Rumbaugh. 1995. Genetic variation and inheritance characteristics for carbon isotope discrimination in alfalfa. J. Range Manage. 48:126-131.

Johnson, R.C., and L.B. Bassen. 1991. Carbon isotope discrimination and water use efficiency in four cool season grasses. Crop Sci. 31:157 162.

Johnson, R.C., and L.L. Tieszen. 1994. Variation for water-use efficiency in alfalfa germplasm. Crop Sci. 34:452-458.

Kehr, W.R., D.K. Barnes, E.L. Sorensen, W.H. Skrdla, C.H. Hanson, D.A. Miller, T.E. Thompson, I.T. Carlson, L.J. Elling, R.L. Taylorl, M.D. Rumbaugh, E.T. Bingham, D.E. Brown, and M.K. Miller. 1975. Registration of alfalfa germplasm pools NC-83-1 and NC-83-2. Crop Sci. 15:604-605.

McKenzie, J.S., R. Paquin, and S.H. Duke. 1988. Cold and heat tolerance. p. 259-302. In A.A. Hanson et al. (ed.) Alfalfa and alfalfa improvement. Agron. Monogr. 29. ASA, CSSA, and SSSA, Madison, WI.

Melton, B., M. Wilson, and C. Currier. 1989. Registration of 'Wilson' alfalfa. Crop Sci. 29:485-486.

Murray, L.W., I.M. Ray, H. Dong, and A. Segovia-Lerma. 2003. Clarification and reevaluation of population-based diallel analyses: Gardner and Eberhart Analyses II and III revisited. Crop Sci. 43: 1930-1937.

Ray, I.M., M.S. Townsend, and J.A. Henning. 1998. Variation for yield, water-use efficiency, and canopy morphology among nine alfalfa germplasms. Crop Sci. 38:1386-1390.

Ray, I.M., M.S. Townsend, and C.K. Muncy. 1999a. Heritabilities of water-use efficiency traits and correlations with agronomic traits in water-stressed alfalfa. Crop Sci. 39:494-498.

Ray, I.M., M.S. Townsend, and C.K. Muncy. 1999b. Heritabilities and interrelationships of water-use efficiency and agronomic traits in water-stressed alfalfa. Crop Sci. 39:1088-1092.

SAS Institute. 1989. SAS/STAT user's guide, v. 6, 4th ed., Vol. 2. SAS Inst., Cary, NC.

SAS Institute. 1992. SAS/STAT software: Changes and enhancements. Release 6.07. Tech. Rep. P-229. SAS Inst., Cary, NC.

Segovia-Lerma, A., L.W. Murray, M.S. Townsend, and I.M. Ray. 2005. Population-based diallel analysis among nine historically recognized alfalfa germplasms. Theor. Appl. Genet. (in press).

Stroup, W.W., P.S. Baenziger, and D.K. Multize. 1994. Removing spatial variation from wheat yield trials: A comparison of methods. Crop Sci. 34:62-66.

I. M. Ray, * A. Segovia-Lerma, and L. W. Murray

I.M. Ray and A. Segovia-Lerma, Dep. of Agronomy and Horticulture, and L.W. Murray, University Statistics Center, New Mexico State Univ., Las Cruces, NM 88003. Received 5 Nov. 2003. * Corresponding author (
COPYRIGHT 2004 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Crop Breeding, Genetics & Cytology
Author:Ray, I.M.; Segovia-Lerma, A.; Murray, L.W.
Publication:Crop Science
Date:Nov 1, 2004
Previous Article:Moving beyond the winter hardiness plateau in U.S. oat germplasm.
Next Article:Agronomic performance of hard red spring wheat isolines sensitive and insensitive to photoperiod.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters