Printer Friendly

Correlated responses for forage traits.

Most breeding efforts in orchardgrass, as in nearly all perennial grasses, take place in spaced-plant nurseries, under relatively non-competitive conditions. A few cultivars have been developed from phenotypic recurrent selection [e.g., Boone (Buckner, 1966); Frode, Justus, and Latar (Alderson and Sharp, 1994); and AC Nordic (McElroy, 1993)], but most are developed by the more traditional approach described and diagrammed by Briggs and Knowles (1967). The essential elements of the traditional approach are evaluation of a large number of spaced plants in Stage I, polycrossing and progeny testing of a small number of clones in Stage II, and development of a synthetic cultivar from a subset of those clones.

A review of 58 orchardgrass cultivars from Canada, New Zealand, USA, and Europe [based on summaries written by Alderson and Sharp (1994) or Hanson (1972) or from registration articles (e.g., McElroy, 1993)] shows that emphasis in spaced-plant nurseries has been on agronomic traits such as maturity, disease resistance, vigor, seed fertility, and morphology. Of the 58 cultivar descriptions, 51 contained enough information to determine whether Stage II progeny test selection was employed. Twenty-six of 51 cultivars (51%), employed some form of progeny test to identify parental clones, with emphasis primarily on forage yield, usually based on sward plots. This percentage has remained relatively constant since progeny testing became widely used in the 1950s. Although selection intensities for Stage II are rarely reported, they are generally very mild, ranging from 0.26 to 0.67 (Alderson and Sharp, 1994; Bowley et el., 1994; Rumball, 1982a, b), likely because of the expense of polycross-progeny testing for forage yield. Thus, while half of the reported orchardgrass cultivars were developed by a combination of both phenotypic and genotypic selection methods, the greatest selection pressures were for spaced-plant traits.

There is currently no means of determining the importance of the phenotypic selection/spaced-planting stage relative to the genotypic selection/sward-plot stage of this selection protocol. There are cases where new orchardgrass cultivars have shown increased forage yield over check cultivars (Bowley et al., 1994; Hanna et al., 1977; McElroy, 1993, 1994; Rumball, 1982a, b; Taylor, 1976). In all cases, stages I and II were confounded in the cultivar development process, so that genetic progress for increased forage yield could not be attributed to either method or stage. There are no written reports in which genetic improvement in forage yield of orchardgrass has been documented relative to the base population from which either phenotypic or genotypic selection was initiated.

Measures of vigor or forage yield per se in spaced-plant nurseries are generally considered to be of poor predictive value for sward-plot forage yield in forage grasses (Carpenter and Casler, 1990; Casler and Hovin, 1985; Hayward and Vivero, 1975). Nevertheless, there are documented cases in which spaced-plant selection for vigor or forage yield have been responsible for increasing forage yield in sward plots of Italian ryegrass, Lolium multiflorum Lam. (Fujimoto and Suzuki, 1975); rye, Secale cereale L. (Bruckner et al., 1991); and Pensacola bahiagrass, Paspalum notatum var. saure Parodi (Burton, 1982). The objective of this study was to quantify genetic progress in sward-plot forage yield and related traits of four orchardgrass populations subjected to two cycles of recurrent phenotypic selection for agronomic traits and seed yield of spaced plants.

MATERIALS AND METHODS

Germplasm and Selection

Selection was practiced in four orchardgrass base populations: I79DT from Iowa (IA), MO2 from Missouri (MO), PLS4 from Pennsylvania (PA), and WO11 from Wisconsin (WI) (Casler et el., 1997). Local and convergent-divergent (C/D) selection methods were employed through two cycles. Local selection consisted of each breeder selecting in the local population (eg. I79DT at IA) without collaboration among the four locations. Convergent-divergent selection (Lonnquist et al., 1979) involved selection in each of the four populations at each of the four locations (Casler et el., 1997). In this manner, 16 subpopulations were created in each cycle. To begin the second cycle, equal amounts of seed were bulked from the three locations not in common with the Cycle-2 location. The procedure ensured that plants selected for superior performance in Cycle 2 had a maternal parent which was selected for superior performance at a different location.

Panicle seed weight was the only selection criterion uniformly applied at all four locations, with a 0.25 selection intensity. A 0.25 selection pressure was also applied to forage traits at each location, but these traits were allowed to vary among locations according to environmental conditions and breeders' preferences. They included resistance to stem rust (Puccinia graminis Pars.), crown rust (P. coronata Cda.), leaf streak (Scolecotrichurn graminis Fckl.), and purple leaf spot (Stagonospora arenaria Sacc.); winter survival; plant biomass; determinacy; and leafiness. All forage traits were determined visually and all selection was conducted in unreplicated spaced-plant nurseries. More details of the selection protocols are provided by Casler et el. (1997).

Evaluation of Selection Responses

The selection protocol resulted in 16 Cycle-1 subpopulations (four base populations each selected at four locations), four Cycle-1 composite populations (balanced bulks from the four locations), 16 Cycle-2 subpopulations, four Cycle-2 composite populations, and four Cycle-2 local populations. Seed of these 44 populations plus remnant seed of the four base populations was tested for purity and germination by the blowing method of Everson and Hotchkiss (1977). These 48 populations, along with eight orchardgrass cultivars, were planted in a randomized complete block design at each of the four locations (IA, MO, PA, and WI) in April or May 1989 (Casler et al., 1997). There were three replicates at MO and four replicates at the other locations. Plots consisted of five drilled rows and were 0.9 by 2.4 m at MO and 0.9 by 3.0 m at the other locations. The seeding rate was 13.4 kg pure live seed [ha.sup.-1]

Plots were clipped two or three times during the seeding year. Nitrogen fertilizer was applied in spring 1990 and 1991, and after harvests one and two in 1990 and 1991, at a rate of 90 kg N [ha.sup.-1] Applications of P and K fertilizer were made according to soil-test results. Plots were harvested three times per year in 1990 and 1991 with a flail-type harvester. A 300- to 500-g subsample of harvested forage was collected and dried at 60 [degrees] C in a forced-air dryer for dry-matter determination. Maturity, just prior to first harvest, was rated on each plot by the scale of Caster (1988). Ground cover of orchardgrass crown tissue was visually rated after the winter of 1990-1991 or after first harvest in 1991. Crown rust reaction was rated just prior to third harvest at WI and stem rust reaction was rated at the same time at IA in both years; symptoms were not observed at the other two locations. Plots were rated from 0 = no incidence of disease, to 5 = leaves completely diseased.

Dried forage samples from the second harvest were ground through a Wiley-type mill and reground to pass through a 1-mm screen of a cyclone mill. Forage samples from first and third harvests were not saved for laboratory analyses due to differences among entries in maturity and rust reaction, respectively. In vitro digestibility (NDSPCS) estimates were obtained by pretreating forage samples in neutral-detergent solution (NDS) prior to digestion with a prepared cellulase solution (PCS) (Bughrara and Sleper, 1986). A 0.2-g sample of dried herbage was placed in a 16- by 150-mm screw-capped culture tube to which 10 mL of hot (90-96 [degrees] C) NDS was added. Tubes were rotated for 1 h at 110 [degrees] C. The NDS was prepared according to Van Soest and Wine (1967), except that sodium sulfite and decahydronaphthalene were omitted. After incubation for 1 h at 110 [degrees] C, contents of the tubes were centrifuged while still hot. Supernatants were carefully aspirated, residues were resuspended in 10 mL of hot water, and they were returned to the incubator and rotated for an additional 30 min at 110 [degrees] C. Residues were washed three times with hot water. Finally, the residue was suspended in 10 mL of fresh PCS for 20 h at 50 [degrees] C prior to final filtering and weighing. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) concentrations were determined according to Van Soest et el. (1991). Relative feed value was computed from NDF and ADF according to Undersander et el. (1993). Crude protein (CP) concentration was computed as N x 6.25 (Nelson and Sommers, 1973).

Statistical Analyses

Forage yield, maturity, rust reaction, ground cover, crude protein concentration, relative feed value, and NDSPCS were analyzed by analysis of variance. The model was a split-plot-in-time, with populations and years as whole-plot factors (Steel and Torrie, 1980). Rust reaction was analyzed separately for each location due to different causal organisms. Residuals from these analyses were plotted against fitted values to identify potential heterogeneity of error variances and were tested for normality. Differences among population means were tested by contrasts, formulated to test linear and non-linear effects of selection cycles and their interactions with environments. Total selection responses for C/D selection were separated into contributions of each selection location according to Casler et al. (1997). Population and location effects were assumed to be fixed, while replicate and year effects were assumed to be random.

Two types of stability analysis were performed on the forage yield data. Lin and Binns (1991) described Type 1 and Type 4 stability as measures of a genotype's homeostatic property, Type 1 across locations and Type 4 among years within locations. In their work and that of others, variance statistics are used to measure these properties. Absolute deviations are an alternative measure of variation and can be used in an analysis of variance to test whether mean absolute deviation (ADM) values differ among populations (Levene's test, Snedecor and Cochran, 1980). Values of Type-4 ADM were computed as [absolute value of [X.sub.ijkl] - [M.sub.ijk.]], where [X.sub.ijkl] = the total forage yield of the ijkth plot in Year 1 and [M.sub.ijk]. = the 2-yr mean forage yield of the ijkth plot (i = 1, ..., 56 populations; j = 1, ..., 4 locations; k = 1, ..., 4 replicates; and l = 1,2 years). Values of Type-1 ADM were computed as [absolute value of [X.sub.ij..] - [M.sub.i...]], where [X.sub.ij..] = the mean total forage yield of the ith population at the jth location and [M.sub.i...] = the mean forage yield of the ith population. These values were subjected to the same analysis of variance and contrast model as described above.

RESULTS AND DISCUSSION

Population and population x location interaction effects were significant (P [less than] 0.01) for all variables evaluated. Population x year and/or population x location x year interactions were significant (P [less than] 0.05) only for forage yield and rust reaction. Contrasts for selection effects were tested for homogeneity between years and only 2.2% were found to be heterogeneous. Therefore, population x year and population x location x year interactions, although generally significant, did not influence the magnitude or significance of selection responses. Therefore, all selection responses were presented as 2-yr means.

Two cycles of C/D selection led to increased forage yield, by 7% for MO2 and 8% for WO11 [ILLUSTRATION FOR FIGURE 1a OMITTED]. While the MO2 response appeared to be due primarily to Cycle-1 efforts, deviation from linearity was not significant for either response (P [greater than] 0.05). Local selection had no effect on forage yield of these two populations. Single-degree-of-freedom contrasts for linear and nonlinear selection effects generally did not interact with locations. Therefore, neither selection protocol led to populations differentially adapted to one or more locations. All forage yield responses were computed as means over evaluation locations.

Neither selection method affected forage yield of I79DT [ILLUSTRATION FOR FIGURE 1a OMITTED], the population with the longest field-oriented selection history. The high yield of I79DT-C0 suggested that it may have already had a high frequency of favorable alleles affecting forage yield and the indirect selection pressures applied in this experiment may have been insufficient to increase this frequency. Conversely, both selection methods led to dramatic reductions in forage yield of PLS4: 7% for local selection and 8% for C/D selection. Because the selection protocols were applied uniformly to all populations and local and C/D selection had similar effects on PLS4, these results indicate that the genetic control of forage yield is dramatically different in PLS4 than the other four populations. These responses suggest that selection pressure per se led to reduced forage yield by reducing the frequency of favorable additive alleles for forage yield, or that PLS4 was characterized by a high frequency of nonadditive intralocus and/or interlocus associations. The former is not likely, because this response was unique to PLS4 and the selection protocol was specifically directed to traits related to plant vigor. Non-additive associations, previously shown to be important in controlling forage yield of orchardgrass (Christie and Krakar, 1980a, b; Knight, 1971), would be lost rapidly during two cycles of phenotypic selection.

The value of the C/D selection system is illustrated in Table 1. In MO2 and WO11, the significant forage yield increases were due to the accumulation of efforts at two and four locations, respectively (P [less than] 0.05). In the case of MO2, even though efforts at IA and MO did not appear to contribute to increased forage yield, the contributions of PA and WI were large enough to result in significant overall progress. For WO11, the significant progress contributed by each selection location likely was due to the population's extreme heterogeneity and the ease with which inferior plants were discarded. Despite the apparent ease with which WO11 [TABULAR DATA FOR TABLE 1 OMITTED] was improved at each location (Table 1), local selection did not affect forage yield of WO11 [ILLUSTRATION FOR FIGURE 1a OMITTED], suggesting that the contributions of each location can only be utilized in the context of C/D selection in which selections from different locations were allowed to recombine with each other during Cycle 2.

Populational buffering within the C/D bulk population was possibly a factor in the improvement of forage yield of MO2 and WO11. The MO2 and WO11 C/D bulk populations showed greater gains than did the mean of their subpopulations derived from individual selection locations, evidence of populational buffering among the subpopulations derived by selection at different locations (Table 1).

There was a large range in mean forage yield of the eight check cultivars, from 9.57 to 10.65 Mg [ha.sup.-1] (LS[D.sub.0.05] = 0.55 Mg [ha.sup.-1]). The four C0 base populations had mean forage yields encompassing most of this range, from 9.54 to 10.23 Mg [ha.sup.-1]. The C/D selection protocol doubled this range equally on both ends of the distribution; the range among C1 and C2 sub-population means was 9.20 to 10.65 Mg [ha.sup.-1]. Whereas, only I79DT-C0 (10.23 Mg [ha.sup.-1]) was within one LSD unit (P [less than] 0.05) of the highest-ranked cultivar, I79DT-C2 (10.24 Mg [ha.sup.-1]), MO2-C2 (10.19 Mg [ha.sup.-1]), and WO11-C2 (10.44 Mg [ha.sup.-1]) were all within this range. Among the C2 populations derived by local selection, only I79DT[C2.sub.local] was within this range, and it represented no change from I79DT-C0.

Previous studies have shown either no association (Kalton et al, 1955) or a positive association (Leudtke, 1984) between forage yield and seed yield of orchardgrass. The above selection responses for forage yield were parallel to those observed for seed yield (Barker et al., 1997), most notably the increase due to C/D selection in MO2 and WO11, the lack of response in I79DT, and the decrease due to both selection methods in PLS4 for both forage [ILLUSTRATION FOR FIGURE 1 OMITTED] and seed (Barker et al, 1997) yield. These studies support the results of Leudtke (1984), suggesting a positive genetic correlation exists between forage and seed yield of orchardgrass within some populations.

Neither C/D nor local selection had a strong or meaningful effect on Type 1 or Type 4 stability (data not shown). Only 13% of the contrasts among populations for Type 1 or Type 4 ADM were significant (P [less than] 0.05). Of these, 10 of 33 contrasts indicated increases in mean ADM due to selection, with magnitudes and frequencies similar for both Type 1 and Type 4. Most of the significant selection effects for Type 1 ADM were observed at the MO and PA evaluation locations, while most of the responses for Type 4 ADM were observed at the PA and WI evaluation locations. There were no other patterns to suggest that Type 1 or Type 4 ADM showed any consistent response to either C/D or local selection. Therefore, while C/D selection increased mean forage yield in MO2 and WO11, their overall stability, among locations or years, did not show consistent or meaningful changes.

Selection responses for maturity score were highly variable across evaluation locations, possibly due to the inherent difficulty of scoring a plot made up of plants in various stages of development and to differential average maturity at the time of first harvest: 6.1 (pre-anthesis) at IA, 3.7 (mid-emergence) at MO, 6.6 (beginning anthesis) at PA, and 5.5 (panicle opening) at WI. A detailed analysis of genotype x location interaction for maturity score did not reveal any biologically meaningful tendencies. Nearly all interactions were due to a differential magnitude of selection responses among locations, not to a differential direction of selection responses. Furthermore, despite these interactions, 10 of 16 responses in mean maturity score, averaged over locations, were significant (P [less than] 0.01; Table 2). Thus, because many factors were confounded among locations (geographic and edaphic characteristics, harvest management, and person conducting the maturity ratings) and because of the lack of biologically meaningful interaction, the genotype x location interactions were ignored. Instead, mean maturity responses across the four evaluation locations were reported.

Both local and C/D selection led to later mean maturity as measured in sward plots [ILLUSTRATION FOR FIGURE 1b OMITTED]. Mean responses ranged from 0.3 to 0.9 for local selection and 0.2 to 0.6 for C/D selection. Three of the responses (MO2 local, PLS4 local, and PLS4 C/D) were non-linear, with greater change observed in Cycle 1 than in Cycle 2. At least two of the four selection locations had significant (P [less than] 0.05) contributions toward later maturity for each population in C/D selection (Table 2). On average, these selection responses were similar in overall magnitude and direction to those observed for the spaced-plant evaluation, but showed some differences. Population WO11 showed a response in maturity in sward plots, but not for heading date of spaced plants, while I79DT responded to both selection methods when evaluated as spaced plants, but only local selection when evaluated in sward plots. These discrepancies indicate that selection responses for heading date or maturity score cannot be unilaterally extrapolated from spaced plants to sward plots or vice versa. These genetic changes in mean maturity score were small enough that they should not significantly affect the use and management of these orchardgrass populations.

Two populations, PLS4 and WO11, showed significant [TABULAR DATA FOR TABLE 2 OMITTED] reductions in stem rust reaction measured in Iowa, resulting from either local or C/D selection [ILLUSTRATION FOR FIGURE 2 OMITTED]. Conversely, stem rust reaction of I79DT and MO2 showed no response to either selection protocol. The latter two populations have been previously selected for resistance to stem rust and had the lowest mean stem rust reaction of the four base populations. Thus, I79DT and MO2 may not have shown a response in stem rust reaction due to low additive genetic variance associated with a relatively high frequency of stem rust resistance alleles. Crown rust reaction in Wisconsin was reduced by both selection methods in WO11 and by C/D selection in MO2, but was not affected by selection in I79DT or PLS4.

All four locations contributed to the reduction in stem rust reaction observed in WO11 during C/D selection, while only two locations contributed to the reduction in stem rust reaction of PLS4 (Table 3). The largest contribution for both populations was from MO, where stem rust reaction was one of the primary selection criteria (Casler et al., 1997). There were no differences between the mean of the individual-location contributions and overall C/D selection, indicating that populational buffering did not contribute to the observation of reduced stem rust in Cycle 2 of C/D selection. Thus, reduced stem rust reaction was most likely due to selection for stem rust resistance at the MO location and, for WO11, elimination of low-vigor plants at all four locations. For crown rust reaction at Wisconsin, the MO contribution was significant for three of four populations (Table 3), suggesting the possibility of an association between selection for stem rust resistance at MO and crown rust reaction at WI. This is further substantiated by the relatively high phenotypic correlation coefficient between stem rust reaction at Iowa and crown rust reaction at Wisconsin (r = 0.68, 54 dr, P [less than] 0.01). Further research will be necessary to determine if this association has biological significance. Because of the similar magnitude of the individual-location contributions, the observed reduction in crown rust reaction for WO11 seems to have been due principally to selection for vigor.

Both C/D and local selection significantly (P [less than] 0.01) increased persistence of WO11, as measured by ground cover, with means of 92, 95, and 98% for WO11-C0, WO11-C2, and WO11-[C2.sub.local], respectively. These responses were consistent across all four evaluation locations. Improvement of WO11 persistence was likely due to elimination of low-vigor plants. No other population showed a significant selection response for ground cover.

There were no consistent selection responses for CP, NDSPCS, RFV, NDF, or ADF (data not shown). There were a small number of significant (P [less than] 0.05) selection responses for each of the above variables, but they were generally due to measured responses at a single evaluation location. The lack of response for these variables was interesting, given the reported strong negative correlation between forage yield and several measures of nutritional value in another orchardgrass population (Stratton et al., 1979). The genetic changes in forage yield in MO2 and WO11, without correlated response for any measure of forage nutritive value, suggest that [TABULAR DATA FOR TABLE 3 OMITTED] total forage yield and mid-summer-forage nutritive value are only loosely linked in orchardgrass and that selection responses for either forage yield or a nutritive value trait need not be accompanied by negative responses for the other trait. This is similar to conclusions drawn from selection studies conducted on smooth bromegrass, Bromus inermis Leyss. (Carpenter and Casler, 1990) and switchgrass, Panicum virgatum L. (Vogel et al., 1981). This conclusion may be different for forage nutritive value traits determined on reproductive growth, which are strongly influenced by relative maturity.

CONCLUSIONS

The results of this study, combined with the previous two reports (Casler et al., 1997; Barker et al., 1997), have led us to develop both biological and programmatic conclusions. Biologically, we have shown that forage and seed yield of orchardgrass can be simultaneously improved, along with genetic shifts toward later heading and anthesis, in some germplasms. Selection progress resulted from both the accumulation of favorable alleles for these traits and populational buffering of subpopulations derived from divergent selection locations. Single-location selection was ineffective for increasing either forage or seed yield. Many correlated selection responses were observed to conflict with prior predictions from correlation coefficients, suggesting that there is no substitute for selection experiments. Furthermore, the dramatic variation in selection response among base populations underscores the need to replicate critical selection experiments on multiple germplasm sources.

Programmatically, these studies suggest that interregional collaboration among forage breeders can, in some cases, lead to progress unattainable by any individual breeder. A long-term commitment to team goals, an ability and willingness to compromise individual goals and methods, and long-term flexible funding were essential elements for participation in this experiment. Conversely, an individual breeder with access to a range of diverse test sites may be able to successfully employ multiple-location selection, although the only report of this is for one commercial organization (Winsett et al., 1991). Our results suggest that maintenance of individual non-collaborating orchardgrass breeding programs may limit the progress attainable by a more regional approach. While the NE-144 Regional Research Committee was an effective vehicle for organizing this experiment, its ability to sustain future regional efforts on a number of forage species is limited by lack of funding.

Abbreviations: ADF, acid detergent fiber; ADM, absolute deviation from the mean; C/D, convergent-divergent; CP, crude protein; NDF, neutral detergent fiber; NDS, neutral detergent solution; NDSPCS, in vitro digestibility; PCS, prepared cellulase solution; RFV, relative feed value.

REFERENCES

Alderson, J., and W.C. Sharp. 1994. Grass varieties in the United States. USDA Agric. Handb. No. 170 (Revised). U.S. Gov. Print. Office, Washington, DC.

Barker, R.E., M.D. Casler, I.T. Carlson, C.C. Berg, D.A. Sleper, and W.C. Young III. 1997. Convergent-divergent selection for seed production and forage traits in orchardgrass. II: Seed yield response in Oregon. Crop Sci. 37:1054-1059 (this issue).

Bowley, S.R., D. Hancock, and D. Wood. 1994. Jay orchardgrass. Can. J. Plant Sci. 74:339-340.

Briggs, F.N., and P.F. Knowles. 1967. Introduction to plant breeding. Reinhold Publ. Corp.

Bruckner, P.L., P.L. Raymer, and G.W. Burton. 1991. Recurrent phenotypic selection for forage yield in rye. Euphytica 54:11-17.

Buckner, R.C. 1966. Registration of Boone orchardgrass. Crop Sci. 6:304.

Bughrara, S.S., and D.A. Sleper. 1986. Digestion of several temperate forage species by a prepared cellulase solution. Agron. J. 78:94-98.

Burton, G.W. 1982. Improved recurrent restricted phenotypic selection increases bahiagrass forage yields. Crop Sci. 22:1058-1061.

Carpenter, J.A., and M.D. Casler. 1990. Divergent phenotypic selection response in smooth bromegrass for forage yield and nutritive value. Crop Sci. 30:17-22.

Casler, M.D. 1988. Performance of orchardgrass, smooth bromegrass, and ryegrass in binary mixtures with alfalfa. Agron. J. 80:509-514.

Casler, M.D., I.T. Carlson, C.C. Berg, D.A. Sleper, and R.E. Barker. 1997. Convergent-divergent selection for seed production and forage traits in orchardgrass. I. Direct selection responses. Crop Sci. 37:1047-1053 (this issue).

Casler, M.D., and A.W. Hovin. 1985. Predicting forage yield from morphological traits in reed canarygrass. Crop Sci. 25:783-787.

Christie, B.R., and P.J. Krakar. 1980a. Performance of advanced generation hybrids of orchardgrass (Dactylis glomerata L.). Can. J. Plant Sci. 60:479-483.

Christie, B.R., and P.J. Krakar. 1980b. Performance of double cross hybrids and single cross mixtures of orchardgrass (Dactylis glomerata L.). Can. J. Plant Sci. 60:485-490.

Everson, L.E., and D.K. Hotchkiss. 1977. A comparison of the blowing and hand methods for the purity analysis of Dactylis glomerata seed. Seed Sci. Tech. 5:451-462.

Fujimoto, F., and S. Suzuki. 1975. Studies on variation and selection in Italian ryegrass populations. III. Response to selection for higher dry matter weight. Japan J. Breed. 25:323-333.

Hanna, M.R., S. Smoliak, and D.B. Wilson. 1977. Chinook orchard grass. Can. J. Plant Sci. 57:615-616.

Hanson, A.A. 1972. Grass varieties in the United States. USDA Agr. Handb. No. 170. U.S. Gov. Print. Office, Washington, DC.

Hayward, M.D., and J.L. Vivero. 1984. Selection for yield in Lolium perenne. II. Performance of spaced plant selections under competitive conditions. Euphytica 33:787-800.

Kalton, R.R., R.C. Leffel, C.E. Wassom, and M.G. Weiss. 1955. Evaluation of combining ability in Dactylis glomerata L. I. Clonal and outcross progeny performance. Iowa State Col. J. Sci. 29:631-657.

Knight, R. 1971. A multiple regression analysis of hybrid vigour in single crosses of Dactylis glomerata L. Theor. Appl. Genet. 41: 306-311.

Leudtke, R. 1984. Breeding orchardgrass for satisfactory Oregon seed yield combined with high Iowa forage yield and quality. Ph.D. diss. (Diss. Abstr. 85-05841). Iowa State Univ., Ames.

Lin, C.S., and M.R. Binns. 1991. Genetic properties of four types of stability parameters. Theor. Appl. Genet. 82:505-509.

Lonnquist, J.H., W.A. Compton, J.L. Geadelmann, F.A. Loeffel, B. Shank, and F.A. Troyer. 1979. Convergent-divergent selection for area improvement in maize. Crop Sci. 19:602-604.

McElroy, A.R. 1993. AC Nordic orchardgrass (Dactylis glomerata L.). Can. J. Plant Sci. 73:845-846.

McElroy, A.R. 1994. AC Splendor orchardgrass (Dactylis glomerata L.). Can. J. Plant Sci. 74:553.

Nelson, D.W., and L.E. Sommers. 1973. Determination of total nitrogen in plant material. Agron. J. 65:109-112.

Rumball, W. 1982a. 'Grasslands Kara' cocksfoot (Dactylis glomerata L.). N.Z. J. Exp. Agric. 10:49-50.

Rumball, W. 1982b. 'Grasslands Wana' cocksfoot (Dactylis glomerata L.). N.Z. J. Exp. Agric. 10:51-52.

Snedecor, G.W., and W.G. Cochran. 1980. Statistical methods. 7th ed. Iowa State Univ. Press, Ames.

Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics. 2nd ed. McGraw-Hill Book Co., New York.

Stratton, S.D., D.A. Sleper, and A.G. Matches. 1979. Genetic variation and interrelationships of in vitro dry matter disappearance and fiber content in orchardgrass herbage. Crop Sci. 19:329-333.

Taylor, D.K. 1976. Sumas orchard grass. Can J. Plant Sci. 56:403-404.

Undersander, D.J., W.T. Howard, and R.D. Shaver. 1993. Milk per acre spreadsheet for combining yield and quality into a single term. J. Prod. Agric. 6:231-235.

Van Soest, P.J., J.B. Robertson, and B.A. Lewis. 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccarides in relation to animal nutrition. J. Dairy Sci. 74:3583-3597.

Van Soest, P.J., and R.H. Wine. 1967. Use of detergents in the analysis of fibrous feeds: IV. Determination of plant cell-wall constituents. J. Assoc. Off. Anal. Chem. 50:50-55.

Vogel, K.P., F.A. Haskins, and H.J. Gorz. 1981. Divergent selection for in vitro dry matter digestibility in switchgrass. Crop Sci. 21:39-41.

Winsett, B.L., S.D. Stratton, and S.J. Baluch. 1991. Registration of 'Benchmark' orchardgrass. Crop Sci. 31:492.
COPYRIGHT 1997 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Convergent-divergent Selection for Seed Production and Forage Traits in Orchardgrass, part 3
Author:Casler, M.D.; Berg, C.C.; Carlson, I.T.; Sleper, D.A.
Publication:Crop Science
Date:Jul 1, 1997
Words:5062
Previous Article:Seed yield response in Oregon.
Next Article:Striga hermonthica seed inoculum rate effects on maize hybrid tolerance and susceptibility expression.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters