Published online 3 October 2006
Published in Agron J 98:1410-1415 (2006)
DOI: 10.2134/agronj2006.0084
© 2006 American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
Production Papers
Crimson Clover Seed Production and Volunteer Reseeding at Various Grazing Termination Dates
Gerald W. Evers* and
Gerald R. Smith
Texas Agric. Exp. Stn., TAMU Agric. Res. & Ext. Center, P.O. Box 200, Overton, TX 75684
* Corresponding author (g-evers{at}tamu.edu)
Received for publication March 22, 2006.
 |
ABSTRACT
|
|---|
Crimson clover (Trifolium incarnatum L.) is a widely used cool-season annual clover in the southeastern USA, but is not considered a dependable reseeder. Grazing termination dates (GTD) of 2-wk intervals from 1 April to 15 May for 3 yr were used on small plots of seven crimson cultivars. Seed production characteristics, seed germination after seed harvest, remaining hard seed percentage after 90 d heat treatment, and volunteer reseeding were determined. There was a general decline in flower density, seed weight flower–1 and seed yield as GTD was delayed. There was variability among cultivars for flower density and seed weight flower–1, but in only 1 out of 3 yr for seed production. Percentage germination, soft seed, and hard seed after seed harvest were not influenced by GTD. Cultivars did differ for percentage germination and hard seed in 2 of 3 yr. Cultivar ranking was not consistent among years except for Columbus that always had high germination and low hard seed percentages. ‘Auburn’ and ‘Flame’ maintained the best hard seed percentage after the 90 d heat treatment. Volunteer seedling densities of 100 m–2 or higher occurred if grazing was terminated by mid-April except for Columbus. The poor reseeding of crimson clover is due to the initial low hard seed production and decline of hard seed percentage during the summer.
Abbreviations: GTD, grazing termination date
|
INTRODUCTION
|
|---|
CRIMSON CLOVER is one of the predominant cool-season annual legumes grown in the southeastern USA (Knight, 1985). It has better seedling vigor and earlier forage production (Evers, 1999), and is not as sensitive to soil pH as most annual clover species (Evers, 2003). Cold tolerance of crimson clover is excellent (Brandsaeter et al., 2002) with a reported N fixation rate of 155 kg N ha–1 (Brink, 1990). It is primarily used for overseeding warm-season perennial grasses to extend the grazing season and reduce the need for nitrogen fertilizer (Knight, 1970; Dunavin, 1982). Because it is an early maturing annual clover species, it is less competitive than most other clover species to the recovery of warm-season grasses in the spring. Early maturing cultivars are also used as winter cover crops in no-tillage crop production (Myers and Wagger, 1991). A major limitation of crimson clover is poor volunteer reseeding potential.
If crimson clover could be managed to reseed each autumn, the annual costs of seed and planting could be minimized. Other advantages of volunteer stands are that the seed are already present when the first significant rainfall occurs in early autumn and stands may be thicker because of more seed per hectare than the recommended seeding rate. A disadvantage of managing cool-season annuals for reseeding is the loss of grazing in the spring when the stocking rate is reduced or livestock are removed to allow for seed production.
Successful reseeding of annual clovers is dependent on the production of an adequate seed crop with a high percentage of hard seed. Increasing daylength in the spring initiates flowering in crimson clover (Butler et al., 2002). Therefore, flowering date will vary with latitude. Hard seed refers to seeds with an impermeable seed coat that prevents absorption of moisture to initiate germination (Kendall and Stringer, 1985). Production of hard seed can be reduced by rainfall during seed maturation in some species (Smith, 1988). Hard seed percentage of crimson clover declined to less than 10% by the following autumn at five locations in California (Williams and Elliott, 1960).
There have been numerous studies on the effect of mechanical defoliation of crimson clover on seed yield. In Oregon (44° N lat) Rampton (1969) reported that clipping ‘Dixie’ crimson clover to a 4.4-cm height when it reached a minimum height of 20.4 to 23 cm, increased seed yields, total germination (germination + hard seed) percentage, and hard seed percentage but decreased 1000-seed weight compared to unclipped crimson clover. In New Zealand (40° S lat), a midspring defoliation reduced flower density from 550 to 90 flowers m–2 and seed yields from 1340 to 440 kg ha–1 compared to an early spring defoliation on a Grasslands selection of crimson clover (Mueller-Warrant et al., 1996). In southern Italy (41° N lat) moisture stress during seed development of ‘Campano’ crimson clover reduced seed yield and number of seed per flower in 1 of 2 yr (Iannucci et al., 1996). Seed weight was reduced slightly but seed germination was not affected.
In northern Mississippi (34° N lat) Knight (1971) conducted a spring clipping trial on ‘Autauga’, ‘Chief’, and ‘Frontier’ crimson clovers overseeded on ‘Coastal’ bermudagrass [Cynodon dactylon (L.) Pers.]. Average volunteer seedling densities were 561, 21, and 621 seedlings m–2 for the 1 April, 25 April, and 25 May clipping dates, respectively. Crimson clover was at full bloom at the 25 April date that reduced seed production. Knight (1971) stated that careful grazing management to allow seed production would be required after 1 April if reseeding was desired.
Rising fuel and commercial nitrogen fertilizer costs make managing annual clovers for reseeding a desirable option. Crimson clover is primarily used for grazing, but previous studies used mechanical defoliation to study seed production and volunteer reseeding. Recent cultivar releases have not been evaluated or compared to older cultivars for reseeding. The objective of the study was to compare seed production and reseeding of seven crimson clover cultivars when grazing was terminated at 2-wk intervals from 1 April to 15 May.
|
MATERIALS AND METHODS
|
|---|
The study was conducted on a different site each year within a 2-ha common bermudagrass pasture at the Kilgore College Demonstration Farm near Overton, TX (32°17' N, 94°58' W). The study was on a Sacul fine sandy loam (clayey, mixed, thermic Aquic Hapludults) for the 1993–1994 and 1994–1995 growing seasons and on a Iuka fine sandy loam (coarse-loamy, siliceous, active, acid, thermic Aquic Udifluvents) in the 1996–1997 growing season. Crimson clover cultivars were overseeded into a lightly disked sod on 29 Sept. 1993, 18 Oct. 1994, and 25 Oct. 1996 at a rate of 18 kg ha–1 in 1.5- by 4.5-m plots with 18-cm row spacing. The same seed lots were used for the first 2 yr but different seed lots were used the third year.
‘AU Robin’, Chief, Dixie, Flame, and ‘Tibbee’ crimson clover were used all 3 yr and Auburn and Columbus crimson clovers were only used 2 of 3 yr because seed was not commercially available. Three-year averages to reach 50% flower near this location, reported as days after AU Robin, the earliest flowering, were: Auburn 2 d, Flame 6 d, Tibbee 9 d, Dixie 9 d, Chief 9 d, and Columbus 37 d (G.W. Evers, unpublished data, 1992–1993, 1993–1994, 1994–1995). Seventy kilograms of P per hectare and from 70 to 100 kg K2O ha–1 were applied each year at planting according to soil test recommendations and 1 kg B ha–1 was applied the first 2 yr. Treatments were arranged in a split-plot design with four replications. Main plots were GTD of approximately 31 March, 14 April, 28 April, and 12 May while subplots were crimson clover cultivars. Insufficient number of crimson clover plants persisted until the last GTD in 1993–1994 and 1994–1995 to collect reliable seed production data.
Four beef cows (Bos taurus) and their autumn-born calves began grazing the pasture beginning in early March each year. Plots were not grazed shorter than 5 cm during the grazing period. Cattle were excluded from the GTD main plots on the appropriate dates by electric fence. Between mid-May and mid-June, a 1-m row was hand-harvested at random from each subplot after the crimson clover matured in each GTD treatment. Flowers were counted and removed by hand. Seeds were thrashed by hand using rubber-covered rubbing surfaces and cleaned to determine seed yield and calculate seed weight per flower.
Four replications of 100 seeds from each subplot were placed on a 7.5- by 7.5-cm indented germination pad in a 9- by 5-cm Petri dish. Petri dishes were placed in a Model 818 Dual Program Illuminated Incubator (Precision Scientific, Chicago, IL) set at 25°C day and 15°C night temperatures with 12-h days. Germination pads were kept moist with deionized water during the study. After 10 d the initial percentage of germinated, soft, and hard seed was recorded. A seed was considered soft if it absorbed moisture and was swollen but did not germinate. If no sign of moisture penetration was visible, a seed was considered hard. A seed was counted as germinated if both the shoot and radicle were present. In the 1993–1994 and 1994–1995 growing seasons, four replications of 100 seeds not used for the initial germination test were placed in the incubator at high diurnal alternating temperatures of 40°C for 12 h and 25°C for 12 h for 90 d. This treatment simulates summer soil surface temperatures and is used to measure the breakdown of hard seed during the summer (Smith, 1988). After 90 d of heat treatment, another germination test identical to the initial test was conducted to determine the percentage of hard seed remaining.
After the plots were sampled for seed in May and June, beef cattle were allowed to graze the study area during the summer. In mid-September the cattle were removed and the study site was mowed to a 5-cm height. On 20 Oct. 1994, 10 Nov. 1995, and 28 Oct. 1997, volunteer crimson clover seedling density was determined by counting seedlings in a 0.09-m2 quadrate at two random locations within each subplot. Rainfall was recorded at the Texas A&M University Agricultural Research & Extension Center, approximately 2 km from the study site.
Data were analyzed by year because the number of GTD's varied among years and two cultivars were represented in only 2 of 3 yr. Data were analyzed as a split-plot within years by GLM using PC-SAS (SAS Institute, 1987). When a GTD x cultivar interaction occurred, analysis of variance was conducted within GTD. An arcsin transformation was performed on germination percentages before statistical analysis (Gomez and Gomez, 1983). Untransformed data are reported. Differences among means were determined by the Fishers Protected LSD Test at the P = 0.05 level of significance.
|
RESULTS AND DISCUSSION
|
|---|
For data reporting and discussion, a growing season is defined from September through November of the following year that includes autumn establishment, spring seed production, summer germination of soft seed, and volunteer reseeding the following autumn. Monthly rainfall for the three growing seasons and long-term average are reported in Fig. 1
. In each growing season there were months with less than 50% and over 300% of the long-term monthly average. Spring rainfall influenced seed production, summer rainfall influenced germination of soft seed and percentage hard seed, while autumn rainfall influenced volunteer reseeding.
View larger version (53K):
[in this window]
[in a new window]
|
Fig. 1. Monthly rainfall during the 1993–1994, 1994–1995, and 1996–1997 growing seasons, and long-term average.
|
|
Seed Production
Flower Density
There were no interactions between GTD's and cultivars for flower density (Table 1). The trend was for flower density to decrease as GTD was delayed. However, flower density at 28 April GTD was only significantly lower than the 14 April GTD in 1993–1994 and the 31 March GTD in 1994–1995. In 1996–1997 there was a significant decrease in flower density with each succeeding GTD.
View this table:
[in this window]
[in a new window]
|
Table 1. Flower density, seed per flower, and seed yield as influenced by GTD, cultivar, and the effects of their interaction.
|
|
Flower density differences among cultivars were small in 1993–1994 and 1994–1995. Chief had one of the highest flower densities and AU Robin had one of the lowest flower densities both years. In 1996–1997 there were greater differences among cultivars ranging from 720 flowers m–2 for Columbus, which was significantly higher than the other cultivars, down to 271 flowers m–2 for AU Robin. At this location Columbus flowers in April, which is 4 to 5 wk later than the other cultivars. Low April rainfall resulted in low flower density in 1993–1994 and good April rainfall in 1996–1997 caused Columbus crimson to have the high flower density. The other crimson cultivars initiate flower production in March that had near normal rainfall all 3 yr.
Seed Weight Per Flower
Seed weight per flower was not affected by GTD in 1993–1994 but tended to decrease with advancing GTD's in 1994–1995 and 1996–1997 (Table 1). There were significant differences among cultivars each growing season. In 1993–1994 Tibbee and Columbus produced over 60 mg seed flower–1. In 1994–1995 the only difference among cultivars was between AU Robin and Chief. There was a GTD x cultivar interaction for seed weight flower–1 in 1996–1997. The general trend was for seed weight per flower to decrease with delayed GTD's up to the 28 April GTD except for the late-maturing Columbus that increased with delayed GTD's (Table 2). Tibbee had one of the higher seed weight flower–1 at the three earliest GTD's and one of the lower seed weight flower–1 at the latest GTD.
Seed Yield
Seed yields followed the same trend as flower density and generally decreased as GTD was delayed (Table 1). Tibbee crimson had higher seed yields than the other cultivars in 1993–1994 and there were no significant differences among cultivars in 1994–1995. There was a GTD x cultivar interaction in 1996–1997 that was the only growing season to have four GTD's. The interaction was mainly due to the late-maturing Columbus. It had one of the lowest seed yields at the earliest GTD and the highest seed yield at the latest GTD (Table 3). Except for the earliest GTD, Auburn had one of the lowest seed yields. Seed yields of the medium-maturity Tibbee, Dixie, and Chief were good at the first three GTD's and intermediate between the early cultivars and Columbus at the late GTD. Seed yields in this study, with continuous grazing until the specified GTD, are substantially less than those reported from clipping studies (Knight, 1971; Mueller-Warrant et al., 1996).
Seed yields of crimson clover can vary widely depending on management and climatic conditions during flowering and seed maturation (Myers and Wagger, 1991; Mueller-Warrant et al., 1996). Extending the grazing season beyond 15 April eliminated seed production 2 out of 3 yr. Considering all growing seasons, Tibbee was consistently one of the best seed producers. Recommended crimson clover seeding rates range from 18 to 22 kg ha–1 (1.8–2.2 g m–2). From a seed production standpoint, sufficient seed was produced in all GTD x cultivar scenarios to exceed that seeding rate.
Seed Germination
Grazing termination dates did not influence germination, soft seed, or hard seed percentages after seed harvest in any year except for soft seed in 1994–1995 when soft seed percentage at the 14 April GTD was significantly lower than the other GTD's (Table 4). This difference is probably not of biological importance since soft seed percentage was less than 2.5% for all GTD's. There were cultivar differences for germination in 2 of 3 yr. Columbus had significantly higher germination than Auburn in 1993–1994 and Dixie in 1996–1997. There was a GTD x cultivar interaction for soft seed in 1996–1997. Columbus and Dixie had lower soft seed percentages than some of the other cultivars the first three GTD's and the highest in the latest GTD (data not shown). There were cultivar differences for initial hard seed percentage in 2 of 3 yr. In 1993–1994 Chief and Auburn were the only cultivars to produce over 40% hard seed. In 1996–1997 Columbus had a significantly lower hard seed percentage than the other cultivars and Auburn had a significantly higher hard seed percentage than Chief.
View this table:
[in this window]
[in a new window]
|
Table 4. Initial percentage germination, soft seed, and hard seed as influenced by GTD, crimson clover cultivar, and their interaction.
|
|
Auburn always had one of the lower germination percentages and higher hard seed percentages indicating it would have the best potential for autumn self-reseeding. Columbus always had one of the higher germination rates and lower hard seed percentages indicating it would have the poorest reseeding potential. Neither Auburn nor Columbus were in the 1994–1995 study and may be the reason there were no differences among cultivars for germination, soft seed, or hard seed that year. Initial hard seed percentages after seed maturation for all crimson cultivars were low compared to more dependable reseeding species like arrowleaf clover (Trifolium vesiculosum Savi.), ball clover (Trifolium nigrescens Viv.), and rose clover (Trifolium hirtum All.) with over 60% hard seed (Hoveland and Evers, 1995).
The 90 d, high temperature treatment was conducted only on seed collected from the 1993–1994 and 1994–1995 growing seasons because the seed from the 1996–1997 season was lost. There were no GTD x cultivar interactions. In 1993–1994 hard seed percentage after 90 d heat treatment was lower at the 28 April GTD than the 14 April GTD (Table 5). Although there were no differences in initial hard seed percentage (Table 4) the above-normal May rainfall following the 28 April GTD in 1993–1994 probably caused the more rapid breakdown of hard seed (Smith, 1988). Hard-seed percentages were two to three times higher in 1994–1995 than 1993–1994 but there were no differences due to GTD's. Cultivar differences for hard seed after the 90 d heat treatment occurred both years. In 1993–1994 Auburn retained the highest hard seed percentage and Columbus and Tibbee had the lowest.
View this table:
[in this window]
[in a new window]
|
Table 5. Hard seed percentage after 90 d heat treatment as effected by GTD, crimson clover cultivar, and their interaction.
|
|
The initial hard-seed percentage of Chief and Tibbee in 1993–1994 was also high but dropped to less than 13% after the heat treatment. Columbus had the greatest breakdown of hard seed after heat treatment. The decrease of hard-seed percentage during the summer in addition to the low initial hard seed levels are major limitations to dependable reseeding of crimson clover. Auburn and Columbus were not in the 1994–1995 study. That year Flame retained the highest hard seed percentage followed by AU Robin and Chief. Of the cultivars represented in both years AU Robin and Flame retained the highest hard seed percentage and Tibbee retained the lowest hard seed percentage.
Volunteer Reseeding
A volunteer seedling density of 100 m–2 was considered a good crimson clover stand. Although the trend for volunteer seedling density was to decrease as GTD was delayed, there were significant differences only in 1996–1997 (Table 6). The lower seedling densities in 1993–1994 than 1994–1995 were probably due to the low hard-seed percentages at the end of summer (Table 5). The hard-seed percentage after the 90 d heat treatment on the 1993–1994 seed was half or less of the 1994–1995 percentage. The very high seedling density at the 31 March GTD in 1996–1997 and the decrease with delayed GTD's are directly related to seed production (Table 1). Acceptable volunteer seedling densities were obtained all 3 yr if grazing was terminated by mid-April. The good seedling density at the 28 April GTD in 1994–1995 is due to the high hard-seed percentage maintained after the 90 d heat treatment (Table 5).
View this table:
[in this window]
[in a new window]
|
Table 6. Influence of GTD, crimson clover cultivar, and their interaction on volunteer crimson clover seedling density.
|
|
There were significant differences among cultivars for volunteer seedling density each year. Auburn and Flame had 150 or more volunteer seedlings m–2 and Columbus had only 43 in 1993–1994. In 1994–1995 Flame had significantly more volunteer seedlings than the other cultivars. Auburn and Flame had the highest volunteer seedling densities that were due to good spring seed production and maintaining a good hard seed percentage after the 90 d heat treatment (Table 5). Seedling densities were less than 50 m–2 for Columbus due to the combination of relatively low seed yields and low hard-seed percentages initially and after the 90 d heat treatment. There was a GTD x cultivar interaction in 1996–1997 because of significant cultivar differences at the 31 March and 14 April GTD and none at the 28 April GTD (Table 7). All cultivars except Columbus and Tibbee had more than 100 volunteer seedlings m–2 through the mid-April GTD. All cultivars had less than 80 seedlings m–2 at the 28 April and 12 May GTD's.
|
CONCLUSION
|
|---|
The poor reseeding potential of crimson clover cultivars is due to initial low hard seed production and decline of hard seed percentage during the summer. Grazing of crimson clover should terminate by mid-April at latitude 32° N lat to have sufficient seed production for minimum volunteer reseeding stands of 100 seedlings m–2. Volunteer reseeding potential did vary among cultivars. Auburn and Flame had the best reseeding and Columbus the poorest. This variability among cultivars indicates selection for improved reseeding is possible.
|
REFERENCES
|
|---|
- Brandsaeter, L.O., A. Olsmo, A.M. Tronsmo, and H. Fykse. 2002. Freeze resistance of winter annual and biennial legumes at different development stages. Crop Sci. 42:437–443.[Abstract/Free Full Text]
- Brink, G.E. 1990. Seasonal day matter, nitrogen, and dinitrogen fixation patterns of crimson and subterranean clovers. Crop Sci. 30:1115–1118.[Abstract/Free Full Text]
- Butler, T.J., G.W. Evers, M.A. Hussey, and L.J. Ringer. 2002. Flowering in crimson clover as affected by planting date. Crop Sci. 42:242–247.[Abstract/Free Full Text]
- Dunavin, L.S. 1982. Vetch and clover overseeded on a bahiagrass sod. Agron. J. 74:793–796.[Abstract/Free Full Text]
- Evers, G.W. 1999. Seedling growth comparison of arrowleaf, crimson, rose, and subterranean clovers. Crop Sci. 39:433–440.[Abstract/Free Full Text]
- Evers, G.W. 2003. Emergence and seedling growth of seven cool-season annual clovers as influenced by soil pH. J. Sustain. Agric. 23:89–107.
- Gomez, K.A., and A.A. Gomez. 1983. Statistical procedures for agricultural research. 2nd ed. John Wiley & Sons, New York.
- Hoveland, C.S., and G.W. Evers. 1995. Arrowleaf, crimson, and other annual clovers. p. 249–260. In R.F. Barnes et al. (ed.) Forages: The science of grassland agriculture. Vol. 1. 5th ed. Iowa State Univ. Press, Ames.
- Iannucci, A., G. Ronga, N. DiFonzo, and P. Martiniello. 1996. Effects of moisture stress on seed yield and quality in four annual clovers. J. Appl. Seed Prod. 14:25–29.
- Kendall, W.A., and W.C. Stringer. 1985. Physiological aspects of clover. p. 111–159. In N.L. Taylor (ed.) Clover science and technology. Agron. Monogr. 25. ASA, CSSA, and SSSA, Madison, WI.
- Knight, W.E. 1970. Productivity of crimson and arrowleaf clovers grown in Coastal bermudagrass sod. Agron. J. 62:773–775.[Abstract/Free Full Text]
- Knight, W.E. 1971. Influence of spring mowing on reseeding and productivity of selected annual clovers in a grass sod. Agron. J. 63:418–420.[Abstract/Free Full Text]
- Knight, W.E. 1985. Crimson clover. p. 491–502. In N.L. Taylor (ed.) Clover science and technology. Agron. Monogr. 25. ASA, CSSA, and SSSA, Madison, WI.
- Mueller-Warrant, G.W., M.D. Hare, G.B. Douglas, and A.G. Foote. 1996. Effect of sowing date and defoliation timing on seed production of hairy vetch, crimson clover, and serradella in New Zealand. J. Appl. Seed Prod. 14:1–10.
- Myers, J.L., and M.G. Wagger. 1991. Reseeding potential of crimson clover as a cover crop for no-tillage corn. Agron. J. 83:985–991.[Abstract/Free Full Text]
- Rampton, H.H. 1969. Influence of planting rates and mowing on yield and quality of crimson clover seed. Agron. J. 61:92–95.[Abstract/Free Full Text]
- SAS Institute. 1987. SAS/STAT guide for personal computers. Version 6.0. SAS Inst., Cary, NC.
- Smith, G.R. 1988. Screening subterranean clover for persistent hard seed. Crop Sci. 28:998–1000.[Abstract/Free Full Text]
- Williams, W.A., and J.R. Elliott. 1960. Ecological significance of seed coat impermeability to moisture in crimson, subterranean, and rose clovers in a Mediterranean type climate. Ecology 41:785–790.[CrossRef][ISI]
TOP
ABSTRACT
INTRODUCTION
