| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Plant Science, McGill Univ., Macdonald Campus, 21111 Lakeshore Rd., Sainte-Anne-de-Bellevue, QC H9X 3V9, Canada
b Dep. of Agronomy and Plant Genetics, Univ. of Minnesota, 1991 Buford Circle, St. Paul, MN 55108-6026
c Univ. of Minnesota West Central Res. and Outreach Center, State Hwy. 329, Morris, MN 56267
d Univ. of Minnesota North Central Res. and Outreach Center, 1861 Hwy. 169 E, Grand Rapids, MN 55744
* Corresponding author (philippe.seguin{at}mcgill.ca)
Received for publication March 11, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: DM, dry matter • GLYH, glyphosate high dose (3.3 kg a.i. ha–1) • GLYL, glyphosate low dose (0.8 kg a.i. ha–1) • KC, Kura clover • KCC, Kura clover cv. Cossack • KCE, Kura clover cv. Endura • MN02, Minnesota mixed grass site 2002 • MNT01, Minnesota tall grass site 2001 • MNS01, Minnesota short grass site 2001 • PAR, paraquat (0.9 kg a.i. ha–1) • QC01, Québec tall grass site 2001 • QC02, Québec tall grass site 2002 • RC, red clover • WC, white clover
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Kura clover is a rhizomatous perennial legume with exceptional persistence under intensive management (Taylor and Smith, 1998). It is also very winter-hardy, adapted to a wide range of edaphic conditions, and has high forage quality compared with other perennial legume species. Relative to alfalfa, at comparable stages of maturity, Kura clover has lower concentrations of neutral detergent fiber, acid detergent fiber, and acid detergent lignin, and higher crude protein and in vitro digestibility (Allinson et al., 1985; Seguin et al., 2002). Establishment problems have limited the widespread use of Kura clover. These are related in part to Kura clover's propensity to devote most of its fixed C to roots and rhizomes during establishment and to its slow nodulation and limited N2 fixation in the seeding year (Peterson et al., 1994; Seguin et al., 2001).
With its slow establishment but high persistence, Kura clover is best suited for permanent pastures. Permanent pastures are often located on marginal land with soils hard to till, and often contain at least some desirable perennial grass species. Legume sod-seeding is a preferred renovation technique for these pastures. In these situations, the resident vegetation can be suppressed by herbicide before no-till drilling to reduce competition and allow legume establishment. Glyphosate and paraquat have been most frequently used and generally provide good results (Seguin, 1998).
Kura clover was established successfully by sod-seeding in different locations in New Zealand without herbicide application (Moorhead et al., 1994) or following glyphosate application at 1.08 kg a.i. ha–1 (Woodman, 1999). In Minnesota, Cuomo et al. (2001) reported poorer establishment of Kura clover than red clover, birdsfoot trefoil, and alfalfa following sod-seeding with grass suppression using 0.62 kg a.i. ha–1 of glyphosate. However, by the fifth year, sod-seeded Kura clover spread and comprised a greater proportion of the sward than alfalfa, birdsfoot trefoil, and red clover (Cuomo et al., 2003).
Seguin et al. (2001) reported that addition of N to low fertility sandy soils helps Kura clover to establish by increasing the proportion of late nodulating plants that survive to later carry on N2 fixation. With other legume species (i.e., birdsfoot trefoil, alfalfa, and red clover) in Minnesota, applications of 30 to 60 kg N ha–1 before legume sod-seeding in grass-dominated swards increased legume contribution to total herbage yield (West et al., 1980). At 90 kg N ha–1, however, legume presence was reduced. There is no information on the effects of N fertilization on the establishment of sod-seeded Kura clover.
Our objectives were (i) to evaluate the establishment of two cultivars of Kura clover compared with red clover and white clover when sod-seeded into grass-dominated swards, (ii) to compare glyphosate and paraquat as sod suppression treatments, and (iii) to investigate the effect of N fertilization on the establishment of sod-seeded Kura clover.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
Herbicide applications were made 1 to 7 d before sod-seeding using a hand-held sprayer in a solution applied at 200 L ha–1, with a pressure of 241 kPa. Seeds were inoculated with appropriate rhizobial peat-based inoculant (Urbana-Labs, St. Joseph, MI) and were drilled directly in the resident grass vegetation using a Tye no-till drill in Minnesota (The Tye Co., Lockney, TX) and a disc-drill no-till seeder in Québec (Fabro, Swift Current, SK, Canada) at rates of 13, 9, and 3 kg ha–1 for KC, RC, and WC, respectively, for equivalent seed numbers per area. Seeding occurred in early May at all sites. In plots receiving N, ammonium nitrate was broadcast at 66 kg N ha–1 applied 1 wk after seeding and 44 kg N ha–1 6 wk later.
In July of the seeding year, populations of established legume seedlings were determined by counting plants within two or three 0.2-m2 quadrats in each plot. Plots were harvested at each site according to forage accumulation, dictated largely by precipitation patterns (Table 1). Number of harvests thus varied among sites with two harvests in QC02; three in QC01, MNT01, and MNS01; and four in MN02. Paraquat treated plots also had one additional harvest in the spring at all sites. The contribution of sod-seeded clovers, grasses, and weeds were determined visually as a percentage of the total biomass at each harvest. A 0.6 by 4.4 m area was cut in the center of plots at each harvest to a 7-cm stubble height using a flail forage harvester in Québec (Swift Machine & Welding, Swift Current, SK, Canada). A 0.9 by 6.0 m area was similarly harvested in Minnesota (Carter Manufacturing Corp., Brookston, IN). Representative 500-g samples of harvested forage were obtained from each plot, dried in a forced-air oven at 60°C for 48 h, and weighted to determine DM content.
All data were subjected to analysis of variance (ANOVA) using PROC GLM of the SAS software (SAS Inst., 1985) to identify significant (P < 0.05) treatment effects and interactions. Data were first analyzed in a combined analysis (McIntosh, 1983) regrouping sites, species/cultivars seeded, herbicide, and N fertilization treatments in a combined split-split-plot design (data not shown). Data from each site were then reanalyzed separately as a split-split plot design using PROC GLM of the SAS software because of the presence of numerous and complex site x treatment interactions. Appropriate LSD values (P < 0.05) were used for mean comparisons based on Gomez and Gomez (1984), when F tests were significant at P < 0.05.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Sod-Seeded Clover Densities
Sod Suppression
Herbicide sod suppression was the factor with the greatest impact on sod-seeded clover densities through main effects and interactions with N fertilization and/or clover species. A positive association between suppression intensity and clover density was observed at all sites (Fig. 1). Clover densities averaged 180% greater in GLYH and 170% greater in GLYL plots than when PAR was used in 38 and 28 of 40 site–species–N treatment combinations, respectively. Clover densities were also greater in GLYH than GLYL plots in 21 of 40 site–species–N treatment combinations; however, the increase was smaller, averaging 44%. Mueller-Warrant and Koch (1983) also reported lower plant densities of sod-seeded alfalfa with paraquat compared with glyphosate. Paraquat, a contact herbicide, exerts short-term control over vegetation, whereas glyphosate, a translocated herbicide, extends its control over a longer period and greater establishment often occurs (Bélanger and Winch, 1985).
|
|
|
No consistent difference was observed between plant densities of KCE and KCC across sites. Despite generally low plant densities of KC compared with RC and WC, densities observed with glyphosate sod suppression should be sufficient to ensure significant KC contributions to forage yields in post-seeding years. Kura clover plants spread via rhizomes and its contribution to yield increase in post-seeding years (Cuomo et al., 2001, 2003).
Nitrogen Fertilization
Nitrogen fertilization influenced clover densities through interactions with herbicides and/or species at all sites except at QC01, where N had no effect. With PAR, N fertilization reduced clover plant densities in 9 of 20 site–herbicide–species treatment combinations. In the 11 other cases, densities of all clover species were extremely low independent of N applications (Fig. 1). This reduction in clover densities is attributable to an increase in the competitive vigor of resident grasses from N fertilization. With PAR, N fertilization reduced RC densities by 60% in four of five sites. It also reduced plant densities of KCC by 75%, of KCE by 35%, and of WC by 55% on average at all sites except QC02. In GLYL- and GLYH-treated plots, there was no consistent effect of N fertilization on plant densities. Nitrogen fertilization, thus, did not increase the number of surviving Kura clover plants, a phenomenon observed in conventionally seeded stands (Seguin et al., 2001).
Forage Dry Matter Yield and Species Composition
Sod Suppression
As for plant densities, herbicide treatment markedly influenced seeding year total forage yield and botanical composition of renovated swards through main effects and interactions with N fertilization and/or species at all sites.
Clover yields in the renovation year increased as intensity of grass suppression increased (Fig. 2 and 3). Herbicide x species interactions were observed in each of the five sites and were driven by variation in the magnitude of differences among species in GLYH- and GLYL-treated plots compared with PAR-treated ones. Differences were driven by numerically small but proportionately large changes occurring where PAR was used. Across sites, DM production of all clovers in PAR-treated plots remained low, being lower than 1 Mg ha–1, except in MN02, where rainfall was significantly above the 30-yr average. Kura clover, WC, and RC contribution to total yield were on average 3, 9, and 17%, respectively, with PAR. With GLYL, KC, WC, and RC contributions were greater, representing 15, 20, and 40% of the total yield, respectively. Finally, with GLYH, KC, WC, and RC content was greater still, representing 25, 35, and 65% of the total DM yield, respectively.
Others have also reported greater legume establishment and DM production when glyphosate was used instead of paraquat for sod suppression (e.g., Koch et al., 1987). The high legume proportion we observed with RC in GLYH-treated plots could, however, be considered too high for pasture swards. Establishment should optimally result in no more than 50% succulent legumes in the sward to avoid the incidence of bloat in ruminant animals (Howart et al., 1991). Sheaffer and Swanson (1982) also obtained high RC proportion (>85%) when sod-seeded following applications of glyphosate at high rates (1.7 kg a.i. ha–1).
Grass DM production decreased as suppression intensity increased. It generally remained higher in plots treated with paraquat, a contact herbicide, than when grasses were suppressed with glyphosate (Fig. 2 and 3). Grass production following PAR application was 130% greater than following GLYL in 30 of 32 sites–species–N treatment combinations across all sites except MNT01. The MNT01 site differed from other environments as no differences were observed in grass DM production between PAR and GLYL in seven of eight species–N treatment combinations. Paraquat grass suppression was exceptionally successful at MNT01 and grasses grew little following herbicide application. Grass DM production was on average nearly sixfold higher with PAR than GLYH in all of 40 site–species–N treatment combinations, and 2.5-fold greater on average in GLYL- than GLYH-treated plots in 37 of 40 site–species–N treatment combinations. The lower dose of glyphosate thus allowed greater grass recuperation following application. Across sites, grass production represented 14, 43, and 78% of total DM following GLYH, GLYL, and PAR applications, respectively. This is concordant with observations from Koch et al. (1987) in New Hampshire, who reported that smooth bromegrass still dominated stands in which sod was treated with paraquat (0.56 kg a.i. ha–1), whereas glyphosate at high dose (2.24 kg a.i. ha–1) killed 90% or more of the initial stand.
Weed infestation was positively associated with level of sod suppression (Fig. 2 and 3). On average, across sites, PAR, GLYL, and GLYH applications resulted in weed content representing 15, 35, and 50% of total DM, respectively. Weed DM production was on average 340% greater in GLYH than PAR and 80% greater than GLYL-treated plots in 35 and 27 of 40 site–species–N treatment combinations, respectively. Significantly greater weed DM production was also observed with GLYL compared with PAR in 31 of 40 site–species–N treatment combinations. Main weeds in QC01 and QC02 were milkweed (Asclepias syriaca L.), Canada thistle [Cirsium arvense (L.) Scop.], dandelion (Taraxacum officinale Wiggers), and tutfted vetch (Vicia cracca L.). In Minnesota they were dandelion, Canada thistle, Pennsylvania smartweed (Polygonum pensylvanicum L.), barnyardgrass (Echinochloa crusgalii L.), and foxtail species (Setaria sp.). Weed infestation was especially high in MN02 and MNS01; weed biomass represented 50% of total DM on average in GLYH- and GLYL-treated plots. The summer was exceptionally humid in MN02 and allowed high weed production, while at MNS01 short-grasses dominating the sward were more severely affected by the herbicidal treatments than at the other sites; weed encroachment followed. Weed biomass was concentrated in the first harvest and generally decreased in subsequent harvests (data not presented). Rioux (1994) also reported severe weed infestation when alfalfa sod-seeding followed broadcast application of glyphosate at high rate (2.5 kg a.i. ha–1) in a smooth bromegrass–dominated sward; dandelion represented as much as 45% of total biomass in the second harvest of the seeding year.
Total DM production responded differently to herbicides in Québec and Minnesota. In QC01 and QC02; total DM production was significantly lower (35% lower on avg.) in GLYL than in PAR-treated plots in 13 of 16 site–species–N treatment combinations. It further decreased by 30% in GLYH-treated swards compared with GLYL in 7 of 16 site–species–N treatment combinations (Fig. 2). The midsummer droughts encountered in both years in Québec did not allow forage regrowth to compensate for loss of grass forage caused by herbicides; total production therefore decreased with increased level of suppression. Severe grass suppression often excessively reduces total forage yield in the seeding year (Seguin, 1998). A pasture in West Virginia renovated with red clover and birdsfoot trefoil produced 30% less forage in the seeding year than nonrenovated swards following sod suppression with paraquat (0.84 kg a.i. ha–1) (Bryan, 1985). In Minnesota the trend was reversed with greater (37% greater on avg.) total DM production in 11 of 24 site–species–N treatment combinations in GLYL- than in PAR-treated plots (Fig. 3). It further increased by 50% in 14 of 24 site–species–N treatment combinations from GLYL to GLYH. Increase in total biomass with increasing suppression in Minnesota is attributable to increased clover and weed DM production that compensated for the grass DM production lost to herbicides.
Sod-Seeded Clover Species
Clover species had small, inconsistent effects on total, grass, and weed DM production across sites; but, as expected, it exerted a strong influence on legume yield at all sites through highly significant main effects, but also through interactions with N fertilization and/or herbicides.
Plots seeded with RC had greater legume DM production than other species in 21 of 30 site–herbicide–N treatment combinations. Red clover was the only legume to produce biologically significant legume yields in PAR plots, its contribution to total DM averaging 16%. Kura clover and WC produced <140 kg clover DM ha–1 in PAR-treated plots, except in the unusually humid season of MN02 (Fig. 2 and 3). In GLYL and GLYH treatments, RC contribution to total DM averaged 40 and 60%, respectively. Red clover superiority is attributable to its rapid and aggressive development and early foliage growth, making it particularly well suited for sod-seeding situations (Belzile, 1988).
There were no consistent differences in clover DM production between the two cultivars of KC across sites. Others have reported differences in suitability for sod-seeding between cultivars of a species (Kunelius and Campbell, 1984). In four of five sites, production of sod-seeded Kura clover remained <1 Mg DM ha–1 (avg. 10% of total DM) in the seeding year within all treatments. Only the exceptional weather conditions encountered in Minnesota in 2002 allowed KCC and KCE to achieve greater initial production, >1.5 Mg DM ha–1 (avg. 30% of total DM) in GLYL and GLYH treatments. Kura clover is known to establish slowly and to have lower initial yields when compared with other legumes. Seguin et al. (1999), in Minnesota, reported Kura clover seeding year yields of <500 kg DM ha–1 when solo-seeded with or without herbicides. In another study conducted in Minnesota, sod-seeded Kura clover occupied only 10% of the sward in its first year of growth (Cuomo et al., 2001).
Legume DM production in WC plots was also minimal, being lower than 0.5 Mg DM ha–1 at both sites in Québec. White clover is particularly sensitive to soil water deficits (Spencer et al., 1975). The drought conditions encountered locally in both years greatly affected WC establishment and growth. At the three sites in Minnesota, WC production was, however, more than twofold greater than KC in 12 of 18 site–herbicide–N treatment combinations. Others have also reported greater first-year production of WC when compared with KC (Spencer et al., 1975; Moss et al., 1996).
Clover species effects on total DM production were inconsistent across sites. There were significant interactions with herbicides and N in QC01 and MNS01, respectively. Clover species did not influence total DM production at QC02, MNT01, or MN02. In QC01 and QC02, low precipitation in midsummer did not allow legumes to sustain DM production. In Minnesota, heavy weed infestations increased total DM production and masked the effects of species on total production.
Clover species did not influence weed DM production in MNS01 and MNT01 but did at other sites through main effects or interactions with herbicide. In the heavily weed-infested plots of MN02, there was significantly greater weed biomass with both cultivars of KC than with WC or RC, for both GLYL and GLYH treatments (Fig. 3). In previous work in Minnesota, high weed infestation occurred in solo-seeded KC plots due to its slow establishment and poor competitive ability (Seguin et al., 1999). Grass production was mainly influenced by herbicide and N treatments. Effects of clover species introduced on grass DM production were weak and inconsistent.
Nitrogen Fertilization
Seeding year N fertilization had limited effects on clover DM production of all species across sites. Legume DM production was not influenced by N in MNT01 and MN02, but was at the other sites through interactions with species and/or herbicides. There were either no effects or positive effects of N on KC production in 19 and 9 of 30 site–species–herbicide treatment combinations, respectively. This contrasts with Seguin et al. (2001), who reported increased KC establishment and seeding year production following N fertilization when conventionally seeded. Differences between studies may be explained by differences in seeding methods. Nitrogen fertilization likely promotes competition against the establishing legume when sod-seeded, which could in turn reduce or cancel its direct positive effects on clover growth. West et al. (1980) reported that application of N at moderate rates during legume sod-seeding increased legume content, but reduced legume content when N was applied at >90 kg N ha–1. Fertilization also had little or no effects across sites on WC, and effects on RC were inconsistent. Overall, there was a significant negative effect of N on yield of all clovers in only 7 of 60 site–species–herbicide treatment combinations, mainly in PAR-treated plots. This suggests that N fertilization could be used with glyphosate to increase total forage production without negatively affecting clover yield in the renovation year, though potential carryover effects in postseeding years must be assessed.
Nitrogen x herbicide interactions influenced grass DM production at all sites. With GLYH, grass yield was not influenced by N fertilization in 19 of 20 site–species combinations. The lack of response to fertilization in GLYH-treated plots is attributable to the excessive suppression resulting in death of the grass resident vegetation. In contrast, a positive grass response to N fertilization was observed in 19 and 13 of 20 site–species treatment combinations in PAR- (avg. increase of 64%) and GLYL-treated plots (avg. increase of 69%), respectively. The proportion of grass in total DM was, however, unaffected by N fertilization, averaging across N treatments 77 and 44% in PAR- and GLYL-treated plots, respectively.
Nitrogen x herbicide interactions were also observed for weed biomass at all sites in Minnesota. In QC01 and QC02, N had no effect on weed biomass. In Minnesota, N fertilization never increased weed biomass in PAR-treated plots, which were dominated by grasses. However, N doubled weed biomass in GLYL- and GLYH-treated plots in 9 and 7 of 12 site–species treatment combinations, respectively. Nitrogen fertilization, however, did not increase weed proportion in the sward, which represented across sites an average of 35 and 50% of total DM for GLYL- and GLYH-treated plots, respectively.
Finally, N had a significant effect on total DM production at all sites, either through main effects or interactions with herbicide, N increasing total DM production in 43 of 60 site–herbicide–species treatment combinations. Interactions with herbicides were caused at QC02 by a lack of response to N fertilization on total production in GLYL- and GLYH-treated plot while there was a significant effect in PAR-treated plots. Overall, N fertilization increased total seeding year DM production in PAR- (40% increase on avg.), GLYL- (43% increase on avg.), and GLYH- (35% increase on avg.) treated plots in 14, 17, and 12 of 20 site–species–herbicide treatment combinations, respectively.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Laberge, P. Seguin, P. R. Peterson, C. C. Sheaffer, and N. J. Ehlke Forage Yield and Species Composition in Years following Kura Clover Sod-Seeding into Grass Swards Agron. J., August 17, 2005; 97(5): 1352 - 1360. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | 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 | |||
