|
|
||||||||
a USDA-ARS, Southern Plains Range Research Station, 2000 18th Street, Woodward, OK 73801
b USDA-ARS, Dale Bumpers Small Farms Research Center, Booneville, AR 72927
c Univ. of Arkansas, Dep. of Statistics, Fayetteville, AR 72701
* Corresponding author (tspringer{at}spa.ars.usda.gov)
| ABSTRACT |
|---|
|
|
|---|
0.05) of Illinois bundleflower (-1240 kg ha-1), roundhead lespedeza (-3460 kg ha-1), slender lespedeza (-3300 kg ha-1), and switchgrass (8370 kg ha-1). Specific combining ability (SCA) effects were found for DMY (P
0.1) of switchgrass-legume mixtures (1360 kg ha-1) and indiangrass-Illinois bundleflower mixtures (1230 kg ha-1). General combining ability and SCA effects were found for crude protein concentration (CPC) of all species and mixtures (P
0.1), respectively. General combining ability effects were found for in vitro dry matter digestibility (IVDMD) for switchgrass and the three legume species (P
0.05). The compatibility of these species could not be predicted solely by DMYs. Compatible mixtures, however, were identified with greater confidence when other variables, such as CPC, IVDMD, and visual observations, were taken into account. On the basis of total forage protein (DMY times CPC), the only compatible grass-legume mixture was indiangrass-Illinois bundleflower (SCA effect = 100 kg ha-1, P
0.05).
Abbreviations: CPC, crude protein concentration DM, dry matter DMY, dry matter yield GCA, general combining ability IVDMD, in vitro dry matter digestibility SCA, specific combining ability
| INTRODUCTION |
|---|
|
|
|---|
Grazing systems using forage legumes increase animal production (Fribourg et al., 1979; Jung et al., 1985; Rayburn et al., 1980; Stricker et al., 1979), and pastures with legumes have greater crude protein content, digestibility, and mineral composition for livestock diets, resulting in greater forage intake and animal performance (Marten, 1985). Kroth et al. (1982) reported the nitrogen (N) benefits from birdsfoot trefoil (Lotus corniculatus L.) and alfalfa (Medicago sativa L.), producing 115 and 200 kg N ha-1, respectively, annually. Residual N fixed by legumes increased subsequent forage growth of ryegrass (Lolium multiflorum Lam.) and was equivalent to a fertilization with 111 kg N ha-1 for arrowleaf clover (Trifolium vesiculosum Savi) and 121 kg N ha-1 for a mixture of arrowleaf and crimson clovers (Trifolium incarnatum L.; Lynd et al., 1984).
Legumes and grasses grown in mixtures can either be compatibleavoid competition with each other, competitivemake demands on the same resources, or allelopathicinteract with each other (Harper, 1977). These relationships are difficult to measure in traditional plot and grazing experiments because dominant species in mixtures have competitive advantages at the onset of measuring their compatibility and interactions. Combining abilities for species and for specific species mixtures can be estimated by a combining ability analysis of variance. This analysis has typically been used by plant and animal breeders to estimate combining abilities of breeding lines using variance components (Griffing, 1956). We used this approach to estimate combining abilities of binary mixtures of grasses and legumes (Springer et al., 1996).
Producers need information concerning combining abilities for cool- and warm-season species presently grown or having the potential of being grown for forage. In other regions, laboratory, greenhouse, and field studies have shown allelopathic and competitive effects of tall fescue (Festuca arundinacea Schreb.) and white clover (Trifolium repens L.) toward other species (McCloud and Mott, 1953; Peters, 1968; MacFarlane et al., 1982; Springer et al., 1996; Springer, 1996). Compatibility has been shown for tall fescue with either birdsfoot trefoil or white clover (Pederson and Brink, 1988; Beuselinck et al., 1992; Springer et al., 1996) and for switchgrass, indiangrass, or sideoats grama (Bouteloua curtipendula Michx.) mixed with either purple prairieclover [Petalostemon purpureum (Vent.) Rydb.], roundhead lespedeza, leadplant (Amorpha canescens Pursh), Illinois bundleflower, catclaw sensitive brier [Schrankia nuttallii (DC.) Standl.], or cicer milkvetch (Astragalus cicer L., Posler et al., 1993).
Little information is available regarding the combining ability of warm-season grasses and legumes. Much of the combining ability literature is from laboratory and greenhouse experiments using cool-season forage species. Our objective was to estimate the combining ability effects for native, warm-season grasses and legumes grown in binary mixtures in the field.
| MATERIAL AND METHODS |
|---|
|
|
|---|
The species studied were Kaw big bluestem, Alamo switchgrass, Osage indiangrass, Sabine Illinois bundleflower, Kanoka roundhead lespedeza, and common slender lespedeza. Seeds of each species were germinated and grown in cavity seedling trays (196 cavities tray-1) and maintained in a greenhouse before field planting. Legume seeds were inoculated with specific Rhizobium before germination.
In April 1995, each species was transplanted by hand into field plots on 15-cm centers. Mixtures were planted at a 1:1 ratio alternating species within and between rows. Treatments consisted of the six pure stands and fifteen mixed stands. All combinations of species mixtures were used, i.e., grass with grass, grass with legume, and legume with legume. The field plot design was a randomized complete block replicated four times with individual plots 1.2 by 1.4 m in size. During the establishment year, plots were maintained weed-free by hoeing and dead plants were replaced to maintain plant populations.
In early March 1996 and 1997, before the initiation of new growth, residual dry-matter (DM) was removed from plots by burning. In mid-March each year, a broadcast fertilization of P and K was applied at a rate of 67 kg ha-1 to all plots. The source of P was triple superphosphate (0-46-0, N-P-K) and the source of K was potash (0-0-60, N-P-K). At the same time, grass-only plots received 67 kg ha-1 of N as ammonium nitrate (34-0-0, N-P-K).
Plots were harvested in the last week of June 1996 and 1997. Grasses were in the boot (R0) to early inflorescence emergence (R1) stage of growth at harvest (Moore et al., 1991). The forage DMY of each plot was determined by harvesting the entire plot to a stubble height of 10 cm. The plot was weighed fresh and a 250 to 300 g subsample of forage was collected for DM determination. The forage of each plot was then separated by hand into its respective species components. Each component was weighed fresh and a 250 to 300 g subsample of forage was collected for DM determination. Forage sub-samples were oven dried at 60°C. The DMY of each plot was calculated by multiplying the percentage DM of the oven dried subsample by the harvested green weight of the plot and converted to kilograms per hectare. The percentage of each species in mixture was calculated by dividing its DM weight by the total DM weight of the plot. Oven dried subsamples were ground to pass a 1-mm screen. Crude protein (N% x 6.25) was determined by the Micro-Kjeldahl procedure (AOAC, 1997) and IVDMD was determined by the Goering and Van Soest (1970) procedure modified to digest samples in an ANKOM Daisy II In Vitro Digester (ANKOM Technology Corp., Fairport, NY).
The lack of significant regrowth due to rainfall deficits after the June harvest each year precluded another harvest. The long-term average rainfall for the site is 10.6 cm in July and 7.7 cm in August. In 1996, rainfall was 4.0 cm above average in July and 4.0 cm below average in August. Although above average rainfall fell in July 1996, 36% was recorded in the second week of July and 59% in the fourth week of July. In 1997, rainfall was 8.3 cm below average in July and 2.0 cm below average during August.
Data for DMY, CPC, IVDMD, total forage protein (DMY times CPC), and total digestible forage (DMY times IVDMD) were analyzed as a split-plot in time analysis of variance with species mixture (treatment) as the main plot and year as the subplot (SAS Institute, 1988). Data were analyzed separately for each year when year effects or year x treatment interactions were present (P
0.05). Comparisons of means were made by Fisher's (protected) least significant difference (LSD) at P
0.05 (Steel and Torrie, 1980).
The combining abilities of species and species mixtures were determined by a combining ability analysis of variance (Griffing, 1956) Method 4, Model 1 procedure and PROC GLM (SAS Institute, 1988). By definition, the general combining ability (GCA) is the mean performance of a species when expressed as a deviation from the overall mean of all species combinations. The specific combining ability (SCA) is the deviation of the "expected" value (the overall mean plus the sum of the GCAs of the two species in mixture) from the mean value of the two species in mixture. Similar to the concept of "Relative Yield Total" as summarized by Harper (1977), we defined an SCA effect = 0 to indicate a competition between species, e.g., the species make similar demands on resources. When the SCA effect = 0, each species contribution to the mixture is equal to its expected share. We defined an SCA effect > 0 to denote a compatibility between species, e.g., the species avoided competition by making different demands on resources. When SCA effect > 0, each species contribution to the mixture is greater than its expected share. An SCA effect < 0 suggested an incompatibility between species, e.g., the species interact with each other. When SCA effect < 0, each species contribution to the mixture is less than its expected share.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
0.05). Dry-matter yields of grass mixtures were generally three to four times greater than legume mixtures and grass-legume mixtures were generally equal to or greater than legume mixtures (Table 1). In addition, DMYs of grass-legume mixtures were equal to or slightly greater than their corresponding grasses grown in pure stands (P
0.05). This last result differs from Posler et al. (1993), who reported substantially greater forage yields for 17 of 18 grass-legume mixtures compared to the corresponding grasses grown alone. Four of nine species used in their study were used in our study. The difference between our experiments is possibly due to soil fertility. The soil of their experiment site was low in available NO3-N. The soil of our site was amended with NO3-N before planting and a maintenance fertilization of NO3-N was added to grass plots each year.
|
In vitro digestion averaged 542 g kg-1 for legume pure stands, 575 g kg-1 for legume mixtures, 539 g kg-1 for grass-legume mixtures, 529 g kg-1 for grass mixtures, and 548 g kg-1 for grass pure stands (Table 1). In vitro digestion averaged 480 g kg-1 for grass-legume mixtures and 475 g kg-1 for grass pure stands in the study by Posler et al. (1996). Although our overall data are 50 to 100 g kg-1 greater in IVDMD, neither study found appreciable differences between the digestibilities of grass-legume mixtures and grasses in pure stands.
A year x GCA interaction for DMY is explained by the increase in GCA for Illinois bundleflower from -2,400 to -70 kg ha-1 (P
0.01). The proportion of Illinois bundleflower in mixtures increased from 1996 to 1997 (Table 2). This increase was from increases in plant size as well as seedling recruitment. The harvest height of 10 cm allowed for some flowering and seed production on the lowest branches of Illinois bundleflower. The year x GCA interaction for CPC resulted from magnitude differences among species (P
0.01). The trends from 1996 to 1997 were the same, where grass species had positive GCAs for DMY and negative GCAs for forage quality and legumes had negative GCAs for DMY and positive GCAs for forage quality (Table 3). The year x GCA interaction for IVDMD was probably caused by the decrease in IVDMD for Illinois bundleflower and increases in IVDMD for roundhead lespedeza and slender lespedeza in 1996 compared to 1997 (P
0.01, see Table 1 under "Pure stands").
|
|
0.1) and indiangrass-Illinois bundleflower mixtures (P
0.1, Table 4). Posler et al. (1993) observed that Illinois bundleflower was quite competitive with switchgrass and indiangrass. The data in our experiment, analyzed with the combining ability analysis, suggested that Illinois bundleflower was compatible with both grass species (P
0.1). The earlier discussion on differences in plant populations between the two experiments and the fact that Illinois bundleflower has excellent seedling vigor, which may give it a competitive advantage when direct seeded with these grasses, may explain the different conclusions.
|
0.01 for all species combination except indiangrass-Illinois bundleflower where it was P
0.09) and were positive for mixtures of grasses or legumes and negative for grass-legume mixtures (Table 4). The negative SCAs for the grass-legume mixtures can be explained by unequal contributions of forage in the samples. In grass-legume mixtures, the forage was dominated by the grass component. The mixture of switchgrass-Illinois bundleflower averaged 76% grass and 24% legume (Table 2). This resulted in lower CPCs than would be expected if the sample were equally represented by both components in the mixture. For indiangrass-Illinois bundleflower, the components were more equally represented in the mixture, averaging 59% indiangrass versus about 41% Illinois bundleflower (Table 2).
For calculated total forage protein (DMY times CPC), the only significant SCA effect was indiangrass-Illinois bundleflower (SCA effect = 100 kg ha-1 crude protein, P
0.05). As stated earlier, those species combinations with significantly high SCA were considered compatible with each other. The SCA effect for this mixture could be attributed to its higher CPC and higher than expected DMY. On the basis of partial correlation coefficients, CPC was more closely associated with total forage protein of indiangrass-Illinois bundleflower (partial R2 = 0.67, P < 0.01) than was DMY (partial R2 = 0.32, P < 0.01).
Year x SCA effects (P
0.02) were found only for IVDMD. Several factors could be involved in this interaction. The most obvious are the differences in digestibilities among the pure stands of legumes as discussed earlier for year x GCA interactions. Overall, no trend was observed for SCA effects for IVDMD and the only significant grass-legume mixture was indiangrass and Illinois bundleflower (-24 g kg-1, Table 4). However, four grass-legume combinations show significant SCA effects for calculated total digestible forage (DMY times IVDMD, Table 4). We could attribute the SCA effects for these mixtures to higher than expected DMYs. On the basis of partial correlation coefficients, DMY was more closely associated with total digestible forage of the four grass-legume mixtures (partial R2 = 0.97, P < 0.01) than was IVDMD (partial R2 = 0.02, P < 0.01).
| CONCLUSIONS |
|---|
|
|
|---|
Cultivar differences were not addressed in this experiment. In a similar experiment, Springer et al. (1996) found differences in combining ability between tall fescue cultivars and legume species. However, cultivars with similar gross morphology and growth habits to these species should give similar results.
Selective grazing by livestock was not addressed in this experiment. For example, preferential grazing of legumes may reduce the legume component in mixtures with grasses. Unless the legume is tolerant of grazing, it may not persist in mixtures.
In our experiment, the compatibility of these species could not be predicted solely on DMYs. Compatible mixtures, however, were identified with greater confidence when other variables, such as crude protein concentration, in vitro digestibility, and visual observations, were taken into account. Under the conditions of our experiment, a mixture of indiangrass-Illinois bundleflower was compatible for total forage protein. This species combination should persist well under closely managed grazing because the legume has good seed set and seedling recruitment. Although legume mixtures with Alamo switchgrass did show some promise in the experiment, visual observations suggested that over time Alamo switchgrass may shade out these legume species. Switchgrass accounted for 84% of the DM for switchgrass-legume mixtures.
Producers need information about the ability of cool- and warm-season species to coexist and thrive together. In this experiment, a mixture of indiangrass-Illinois bundleflower was the most compatible grass-legume mixture as determined on the basis of total forage protein (SCA effect = 100, P
0.05), while a mixture of big bluestem-Illinois bundleflower was incompatible (SCA effect = -70, P
0.15). Providing information on specific combining abilities for species mixtures would allow producers to choose species for their forage program.
| NOTES |
|---|
|
|
|---|
Received for publication June 13, 2000.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. L. Springer, R. L. Gillen, and R. W. McNew Combining Ability of Binary Mixtures of Introduced, Cool- and Warm-Season Grasses and Legumes Crop Sci., November 7, 2007; 47(6): 2540 - 2546. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Fischbach, P. R. Peterson, C. C. Sheaffer, N. J. Ehlke, J. Byun, and D. L. Wyse Illinois Bundleflower Forage Potential in the Upper Midwestern USA: I. Yield, Regrowth, and Persistence Agron. J., May 13, 2005; 97(3): 886 - 894. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Byun, C. C. Sheaffer, M. P. Russelle, N. J. Ehlke, D. L. Wyse, and P. H. Graham Dinitrogen Fixation in Illinois Bundleflower Crop Sci., March 1, 2004; 44(2): 493 - 500. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||