HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Search for Related Content PubMed Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Agricola Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Related Collections Clover Crop Genetics

CROP BREEDING, GENETICS & CYTOLOGY

Recurrent Selection for Seedling Vigor in Kura Clover

Dep. of Agronomy and Plant Genetics, Univ. of Minnesota, 411 Borlaug Hall, 1991 U. Buford Circle, St. Paul, MN 55108

* Corresponding Author (ehlke001{at}umn.edu)

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Kura clover (Trifolium ambiguum M. Bieb.) is a rhizomatous perennial forage legume that has outstanding persistence once established. Low seedling vigor presents a challenge to kura clover establishment. Kura clover seedlings partition a significant portion of their dry matter to below ground growth. Therefore, seedling vigor could be improved by increasing dry matter allocation to the shoot. Our objective was to study the potential of recurrent phenotypic selection in the greenhouse for reduced root/shoot ratio to increase seedling vigor. Three cycles of divergent selection for root/shoot ratio and a control selection for large plant biomass (LP) were performed from a parent population consisting of ‘Rhizo’, ARS-2678, KZ1, and Erect Spreader populations. Selections for low root/shoot ratio (LoRS) and high root/shoot ratio (HiRS) were performed by independent culling after first selecting for large total biomass. Parent and selected populations were evaluated in the greenhouse and field. Three cycles of selection for LoRS and LP increased shoot yield in the field 42 d after planting (DAP) by 35 and 34%, respectively. Three cycles of selection for LoRS reduced root/shoot ratio in the field by 16%. Selection for HiRS did not affect seedling size in the field. Population mean shoot and total plant weights were correlated between greenhouse and field environments (average r = 0.89). Population mean seed weight was correlated with shoot weight in the field (r = 0.68, P < 0.05). The results indicate that phenotypic greenhouse selection for seedling size is an effective means of increasing seedling vigor in kura clover.

Abbreviations: ANOVA, analysis of variance • C₀, parent population • C₁, first selection cycle • C₂, second selection cycle • C₃, third selection cycle • HiRS, high root/shoot ratio • LoRS, low root/shoot ratio • LP, large plant biomass • DAP, days after planting

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
KURA CLOVER (Trifolium ambiguum M. Bieb.) is a rhizomatous perennial forage legume that has demonstrated outstanding persistence (Sheaffer and Marten, 1991). It has potential to enhance the productivity and profitability of pastures by functioning as a permanent legume component without need for reseeding. Unfortunately, low seedling vigor slows or prevents the establishment of kura clover (Hill and Mulcahy, 1995; Taylor and Smith, 1998; Townsend, 1970). Therefore, improving seedling vigor has become a necessary goal of our kura clover breeding program.

Research on birdsfoot trefoil (Lotus corniculatus L.) provides an excellent source of information on breeding for seedling vigor in a small seeded legume like kura clover (Twamley, 1970, 1972). Twamley (1967)(1974) proposed that selection for seed size might be a simple and inexpensive method to increase seedling vigor. Although seed size is often highly correlated with seedling vigor, further investigation showed that selection only for seed size is not likely to improve vigor (McLean and Nowak, 1997). Seed size may be a poor indicator of seedling vigor because of highly significant maternal effects (Twamley, 1967). Fresh shoot weights of birdsfoot trefoil seedlings grown in the greenhouse were highly correlated to fresh shoot weights in the field 7 wk after planting and may serve as a more precise indicator of seedling vigor than seed size (Twamley, 1967). Twamley (1972) found that selecting the most vigorous birdsfoot trefoil seedlings from the most vigorous lines (geno-phenotypic selection) proved to be the most effective method of increasing seedling vigor, although there was also a modest parent-offspring correlation (r = 0.27) for phenotypic selection only. After three cycles of geno-phenotypic selection, Twamley (1974) had increased seedling birdsfoot trefoil fresh weight in the greenhouse by 40% at 6 wk of age.

Smith (1995) used two cycles of phenotypic recurrent selection for seed size with six hexaploid kura clover plant introductions to increase seed size by 27%. In a field study, the large seeded population had an establishment rate two times that of the control population 6 wk after planting, but forage yield in the seeding year was similar for both populations.

Kura clover requires vernalization to flower, so a selection program involving genotypic selection will require three growing seasons to perform one cycle of selection. In contrast, phenotypic selection can be performed at a rate of one cycle per year. Phenotypic selection for seedling vigor in kura clover may be more effective than in birdsfoot trefoil if selection could be targeted at a specific cause for low vigor. Genrich et al. (1998) and Hill and Mulcahy (1995) reported that kura clover allocates a majority of its resources to root and rhizome development during the establishment phase, resulting in a root/shoot ratio of 2.0 to 3.0 after a year of growth. Spencer et al. (1975) found that the kura clover cultivar Summit had a root/shoot ratio of 2.4, whereas the white clover (Trifolium repens L.) cultivar Grassland Huia had a root/shoot ratio of only 0.2. They also found that white clover had ten times the foliar production of kura clover in the first growing season. Spencer and Hely (1982) reported that kura clover had up to 300% greater root yield than white clover after 1 yr, but that white clover had 30% greater shoot production than kura clover over a 5 yr period. After studying seedling development of kura clover, Genrich et al. (1998) concluded that since kura clover seedlings partition a significant amount of dry matter to root and rhizome growth, increased seedling vigor might be achieved by selecting for increased shoot growth.

Selecting for reduced root/shoot ratio is an obvious method to use for increasing dry matter partitioning to the shoot. However, the results of selection on a ratio are often not intuitive. Theory developed by Rowe (1995) can be used to predict the outcome of selection on a ratio. Assuming that shoot and root weight have similar genetic variances and are highly correlated, selection for reduced root/shoot ratio will rapidly reduce root weight while having little effect on shoot weight. Selecting for both low root/shoot ratio and large total plant weight would be a reasonable method to insure that selection would increase shoot weight while also reducing the root/shoot ratio.

Genrich (1995) initiated breeding for increased kura clover seedling vigor by completing one cycle of divergent selection for root/shoot ratio and selection for large plant biomass. When evaluated in the greenhouse, all three selected populations had significantly larger shoot mass than the parents, but no divergence was detected between the large and small root/shoot ratio populations. The favorable results obtained from a single cycle of selection indicated that shifting resource allocation of kura clover from the root to the shoot might be a reasonable mechanism of increasing seedling vigor in the species. Therefore, additional cycles of selection were warranted. The objective of this experiment was to study the potential of recurrent phenotypic selection in the greenhouse to increase total plant weight and reduce root/shoot ratio to increase kura clover seedling vigor.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Greenhouse Selection
Two additional cycles of recurrent selection (C₂ and C₃) were performed on the three populations created by Genrich (1995): low root/shoot ratio (LoRS), high root/shoot ratio (HiRS), and large plant biomass (LP). Genrich (1995) had selected equal numbers of plants for LoRS, HiRS, and LP from ‘Rhizo’, ARS-2678, KZ1, and Erect Spreader populations. Rhizo was developed by selection for vigorous rhizomatous growth, persistence, disease, and insect resistance (Henry and Taylor, 1989). Population ARS-2678 is a germplasm selected for drought tolerance, winter hardiness, rhizome production, forage and seed yield, nodulation, and N₂ fixation activity (Rumbaugh et al., 1991). The New Zealand breeding population KZ1 was selected from ‘Monaro’ (Anonymous, 1983) for uniform flowering. The breeding population Erect Spreader was developed by the University of Minnesota from plant introductions by selections for erect growth habit and rapid rhizomatous spreading.

We established 1080 seedlings of the LoRS, HiRS, and LP populations in 15 cm deep sand benches. Seeds were scarified with sandpaper and planted 1.3 cm deep and 2.5 cm apart in rows 90 cm long spaced 7.6 cm apart. Two to four seeds were placed at each 2.5 cm spacing, and the plants were randomly thinned to 1080 seedlings at the unifoliolate leaf stage by removing all but one seedling from each 2.5 cm spacing. Rhizobium leguminasarium biovar trifolii (LiphaTech, Milwaukee, WI) was surface applied after planting and the sand benches were thoroughly watered. Greenhouse temperature was maintained between 22 and 26°C. Supplemental lighting was provided for 16 h d^-1 at about 110 mmol s^-1 m^-2. Before planting, the benches were steamed and treated with 1.1 g m^-2 Terrachlor (pentachloronitrobenzene) fungicide. Before the C₂ selection, the sand was amended with 67 g m^-2 CaCO₃, 4.5 g m^-2 P, and 8.6 g m^-2 K. Prior to the C₃ selection, the sand was amended with 67 g m^-2 CaCO₃, 1.0 g m^-2 P, 27.7 g m^-2 K, and 900 g m^-2 Micromax slow release micronutrient fertilizer (Grace-Sierra Horticultural Products Company, Milpitas, CA), and the sand was flushed thoroughly with water before planting to reduce surface salinity. Other than the fertility treatments that were adjusted to prevent nutrient availability from limiting seedling growth, the protocol was identical for both cycles of selection.

Selection was performed using a form of recurrent restricted phenotypic selection (Burton 1974). Each population was divided into 12 plots consisting of 90 plants each (three rows) to reduce the effects of variable greenhouse lighting and temperature. Selections were made about 42 DAP. Seedlings were harvested and washed with water. The seedlings were clipped 0.5 cm above the crown, and the fresh shoot and root portions were weighed separately. Selection was performed by independent culling. Thirty plants with the largest total biomass were selected from each plot of 90 plants. Next, out of these 30, four plants were selected for high root/shoot ratio, low root/shoot ratio, or large total biomass (both root and shoot) in the HiRS, LoRS, and LP populations, respectively. Overall, the proportion of individuals selected was 4.4% for each of the three selection criteria.

Roots from the selected plants (four plants from each of 12 plots = 48 per population) were planted in 10-cm diameter pots and allowed to grow in the greenhouse until mid March when the plants were placed in a growth chamber for vernalization. The chamber was set at 4°C, and lighting was provided for 12 h d^-1 at 130 mmol s^-1 m^-2. In early May, the populations were transferred to isolated crossing blocks at St. Paul and Rosemount, MN. Seeds were harvested from the crossing blocks in August, threshed, and cleaned. A composite was constructed from each population by combining an equal quantity of seed from each plant by mass. Seed weight of each population was determined by weighing four replicates of 200 seeds from each population.

Greenhouse Evaluation
The first greenhouse evaluation was performed in 1996 using the same C₂ plants as were used in the third cycle of selection. Therefore, the protocol was the same as previously described in the third selection cycle. The four parent populations and the three C₂ populations were grown in a randomized complete block design with 12 replicates. A second greenhouse evaluation was performed in 1997 using the four populations making up the Cycle 0 (C₀) (Rhizo, ARS-2678, KZ1, and Erect Spreader) and the nine selected populations resulting from three selection cycles using each of the three selection criteria. The design was a randomized complete block with 15 replicates and 30 plants of each parent and selected population per replicate. Before planting, the sand was amended with 67 g m^-2 CaCO₃, 1.0 g m^-2 P, 13.9 g m^-2 K, and 90 g m^-2 Micromax micronutrient mix. All other procedures were as in the selection methods above. The four parent populations were used as an estimate for the C₀ because Genrich (1995) made the initial selections using equal numbers of plants from each of the four parent populations. Seeds from all populations were stored in a controlled environment at 4°C and less than 40% humidity to prevent loss of seedling vigor.

Field Evaluation
Four field evaluations were conducted during 1997 at St. Paul and Becker, MN. At each location, two separate experiments were conducted in which the four parent populations and the six C₁ and C₂ populations (two cycles by three selection criteria) were planted in randomized complete block designs with eight replications. One experiment at each location was space planted, and a second experiment was seeded in rows. In the space-planted experiments, seeds were planted to a 1 cm depth and seedlings were randomly thinned at the unifoliolate leaf stage to form a grid with 5 cm equal spacings and 56 plants plot^-1. The row-planted experiments were seeded at 0.15 g m^-1 in rows 178 cm long and 15 cm apart with each plot consisting of two rows. At St. Paul, both experiments were planted on May 21. At Becker, the row-planted experiment was planted on May 22 and the space-planted experiment was planted on July 15. All plots were inoculated with Rhizobium leguminosarum biovar trifolii and irrigated as needed. The soil at St. Paul was a Waukegan silt loam (fine-silty over sandy, mixed Typic Hapludoll) and at Becker the soil was a Hubbard loamy sand (sandy, mixed, Udorthentic Haploboroll). At both locations, pH, P, and K levels were greater than 6.5, 35 kg ha^-1, and 200 kg ha^-1. The silt loam soil was derived from loess parent material and has an organic matter content of about 30 g kg^-1 whereas the loamy sand was derived from glacial outwash and has an organic matter content of about 10 g kg^-1. Shoot and root fresh weight was determined 42 DAP by harvesting, washing, and weighing the seedlings. All seedlings in the space-planted plots were harvested, and in the row-planted plots seedlings were harvested from two randomly selected 30 cm lengths of row. Weeds were removed by hand from all plots except those in the St. Paul row-planted experiment. These plots had heavy weed pressure from Portulaca oleracea L. and Digitaria sanguinalis L. that reached a height of 15 cm by harvest.

Two additional field evaluations were performed at St. Paul in 1998 using the four parent populations and the nine populations from three cycles of selection for HiRS, LoRS, and LP in randomized complete block designs with 8 plants plot^-1. The first experiment was planted on June 23 and had six replications. The second experiment was planted on July 1 and had eight replications. All plots were inoculated with Rhizobium leguminosarum biovar trifolii and irrigated as needed. The soil characteristics were as in 1997 at St. Paul. Plots were seeded at a 1 cm depth and seedlings were randomly thinned at the unifoliolate leaf stage to form a grid with 5 cm equal spacings. Seedling shoot and root fresh weights from both experiments were determined after harvesting and washing all the seedlings on August 4, 1998.

Data Analysis
We determined changes in plant biomass due to selection by measuring seedling fresh weights. While fresh weight may be subject to variation due to changes in plant moisture status, it provided a rapid approach to measuring response to selection. We verified the association of fresh and dry weights in a preliminary experiment and found the correlation to be r = 0.98 (P < 0.001). Increased seedling size is needed to compete with other species during the establishment phase, and whether the increased size is due to higher water or dry matter content may be irrelevant to successful establishment of kura clover. Seedling fresh weights have been commonly used to evaluate seedling vigor in birdsfoot trefoil (McKersie and Tomes, 1982; Twamley, 1967, 1970, 1972, 1974).

Analyses of variance (ANOVA) were performed on plot means of fresh shoot weight, fresh root weight, root/shoot ratio, and total plant biomass from each experiment. For the preliminary greenhouse evaluation, nonorthogonal contrasts were used to compare the selected populations to the C₀. In the other experiments, orthogonal polynomial contrasts were used to evaluate the response to selection. Due to heterogeneity of error, results from the two locations in the 1997 field evaluation were analyzed separately, but the two planting methods within a location were analyzed together according to the technology adaptation experiment procedure as described by Gomez and Gomez (1984). For the 1998 field evaluation, the two planting dates were also analyzed together using the same procedure. A correlation analysis was conducted by calculating Spearman's rank-order correlation coefficients between the population means of greenhouse and field grown plants. Although the individual plant weights were often not normally distributed, all statistical calculations were performed using plot means, which were normally distributed. All statistical analyses were performed with SAS software (SAS Institute, 1990).

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Greenhouse Experiments
In the 1996 greenhouse evaluation, the C₂ LP population had shoot, root, and total biomass weights that exceeded those of the C₀ by 27, 20, and 24%, respectively (Table 1). Two cycles of selection for LoRS reduced the root/shoot ratio by 11% when compared to the C₀, and two cycles of selection for HiRS increased the root/shoot ratio by 11% when compared with the C₀. The C₂ LP population also had a 7% reduction in root/shoot ratio.

Field Experiments
The two soil types used in the 1997 field experiment produced large differences in plant growth. Shoot weight on the silt loam soil was more than two times larger than shoot weight on the loamy sand. These large differences resulted in heterogeneity of error that precluded a combined analysis across locations. Plants growing on the loamy sand lacked vigor and were somewhat chlorotic. These symptoms may be related to limited N₂ fixation and low N fertility of the soil. Nitrogen fertilization has been shown to increase the growth of kura clover seedlings (Seguin et al., 2000) because of ineffective symbiosis during the first 50 DAP.

Seeding method affected root, shoot, and total plant weight at both locations (Table 4). On the silt loam, total plant weights in the row-planted plots were 40% smaller than total plant weights in the space-planted plots. This reduction in plant growth was most likely due to the weed pressure in the row-planted plots. In contrast to the silt loam, total plant weights were 37% smaller in the space-planted plots than they were in the row-planted plots. Seeding date was probably the largest factor contributing to these differences in plant growth because the row-planted plots were seeded earlier in the season when soil N is more abundant.

On the silt loam, LoRS selection produced a linear increase in shoot and total plant weight and a linear reduction in root/shoot ratio (Table 4). The C₂ LoRS population had a shoot weight and total plant weight that were 11 and 10% greater than the C₀ and a root/shoot ratio that was reduced by 7% (Table 5). The LoRS selection did not result in any changes in seedling traits from the C₀ when evaluated on loamy sand. A probable explanation for this result is the inadequacy of N₂ fixation and low N availability of the loamy sand. Under conditions where fertility limits plant growth, a reduced root/shoot ratio is not likely to be an effective method for improving seedling vigor because maximum root production will be necessary for plant nutrition. In a nutrient limited environment, selection for both increased shoot and root weight would be the best mechanism for increasing seedling vigor, and this theory is consistent with the results of the field evaluation. Two cycles of HiRS selection did not produce a significant response in any of the measured seedling traits in the 1997 field evaluations (Table 4). This result is consistent with the 1997 greenhouse evaluation, in which no linear responses were produced by HiRS selection. Because HiRS selection failed to produce a consistent increase in root/shoot ratio, we conclude that kura clover may already be near the upper limits of biomass allocation to the root. If this is true, root mass is not likely to be increased without a corresponding increase in shoot mass.

The 1998 field experiments included the C₀ and populations from all three cycles of selection. Results were similar to those obtained in the 1997 greenhouse experiment, which included the same populations (Tables 2 and 6). The primary difference was that quadratic responses to selection were found in the greenhouse, but only linear responses were seen in the field. After three cycles of selection, LoRS and LP selection had increased shoot weight by 35 and 34%, respectively, and LoRS selection had reduced the root/shoot ratio by 16% (Table 3). Highly significant seeding date effects were due to the plants from the different seeding dates being harvested at different ages. Seeding date by population interactions were not significant, indicating that the populations performed consistently across seeding dates and growth stages.

Additional Cycles of Selection
Additional cycles of selection will be necessary to make agronomically valuable improvements in the seedling vigor of kura clover. Although we have demonstrated the effectiveness of LP and LoRS selection for increasing shoot weight of kura clover, three cycles of selection yielded populations with shoot weights similar to the most vigorous parent population. The LP selection method currently appears to be most promising because of its consistent performance under diverse environments. If shoot mass can be increased without sacrificing root mass, the plants may have a competitive advantage when nutrients are limiting, but LoRS selection may be beneficial in other environments. Therefore, the most conservative strategy would be to continue both selection methods until their strengths in different environments can be more fully determined.

Selecting simply for high shoot weight may be an efficient compromise between the LoRS and LP selection methods. Because the correlation between shoot weight and total plant weight is high (r = 0.97), selection for high shoot weight would resemble LP selection. But compared to LP selection, selection for large shoot weight would eliminate several plants with modest shoot weights and large root weights and include several plants with large shoots and low root/shoot ratios (Fig. 1). Selection for high shoot weight also would demand fewer resources by eliminating the need to carefully dig, wash, and weigh roots.

View larger version (26K): [in this window] [in a new window]   Fig. 1. Theoretical selections for low root/shoot ratio (LoRS), high root/shoot ratio (HiRS), large total plant weight (LP), and a proposed selection for large shoot weight in Kura clover indicated by the rectangle. The data presented are from greenhouse-grown C2 LP plants.

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Joint contribution of the Minnesota Agric. Exp. Stn. Minnesota Agric. Exp. Stn. Journal series paper 00-1-13-0157.

Received for publication August 8, 2000.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Anonymous. 1983. Register of Australian herbage plant cultivars. J. Aust. Inst. Agric. Sci. 49:243–244.
• Burton, G.W. 1974. Recurrent restricted phenotypic selection increases forage yields of Pensacola bahiagrass. Crop Sci. 14:831–835.[Abstract/Free Full Text]
• Genrich, K.C. 1995. Kura clover (Trifolium ambiguum Bieb.): growth and development during the seeding year and divergent selection for seedling vigor. M.S. thesis. Univ. of Minnesota, St. Paul.
• Genrich, K.C., C.C. Sheaffer, and N.J. Ehlke. 1998. Kura clover growth and development during the seeding year. Crop Sci. 38:735–741.[Abstract/Free Full Text]
• Gomez, K.A., and A.A. Gomez. 1984. Statistical procedures for agricultural research. 5th ed. Wiley, New York.
• Henry, D.S., and N.L. Taylor. 1989. Registration of ‘Rhizo’ kura clover. Crop Sci. 29:1572.[Free Full Text]
• Hill, M.J., and C. Mulcahy. 1995. Seedling vigour and rhizome development in Trifolium ambiguum M. Bieb. (Caucasian clover) as affected by density of companion grasses, fertility, drought, and defoliation in the first year. Aust. J. Agric. Res. 46:807–819.
• McKersie, B.D., and D.T. Tomes. 1982. A comparison of seed quality and seedling vigor in birdsfoot trefoil. Crop Sci. 22:1239–1241.[Abstract/Free Full Text]
• McLean, N.L., and J. Nowak. 1997. In vitro selection for improved seedling vigour in birdsfoot trefoil (Lotus corniculatus L.). Can. J. Plant Sci. 77:385–390.
• Rowe, D.E. 1995. Characteristics of elite population selected on ratio criterion: I. Traits with equal genetic variances. Crop Sci. 35:425–430.
• Rumbaugh, M.D., D.A. Johnson, and J.R. Carlson. 1991. Registration of ARS-2678 kura clover germplasm. Crop Sci. 31:497.[Free Full Text]
• SAS Institute. 1990. SAS user's guide: Statistics. 4th ed. SAS Inst., Cary, NC.
• Seguin, P., C.C. Sheaffer, N.J. Ehlke, M.P. Russelle, and P.H. Graham. 2000. Kura clover seeding year growth and N₂ fixation: effects of nitrogen fertilization and rhizobial inoculation. Agron. J. 92:1216– 1220.[Abstract/Free Full Text]
• Sheaffer, C.C., and G.C. Marten. 1991. Kura clover forage yield, quality, and stand dynamics. Can. J. Plant Sci. 71:1169–1172.
• Smith, R.R. 1995. Recurrent phenotypic selection for seed size in kura clover. p. 78. In 1995 Agronomy Abstracts. ASA, Madison, WI.
• Spencer, K., and F.W. Hely. 1982. Shoot and root responses to phosphorus by Trifolium ambiguum and Trifolium repens in a montane environment. N.Z. J. Agric. Res. 25:77–85.
• Spencer, K., F.W. Hely, A.G. Govaas, M. Zorin, and L.J. Hamilton. 1975. Adaptability of Trifolium ambiguum Bieb. to a Victorian montane environment. J. Aust. Inst. Agric. Sci. 41:268–270.
• Taylor, N.L., and R.R. Smith. 1998. Kura clover (Trifolium ambiguum M.B.) breeding, culture, and utilization. Adv. Agron. 63:153–178.
• Townsend, C.E. 1970. Phenotypic diversity for agronomic characters and frequency of self-compatible plants in Trifolium ambiguum. Can. J. Plant Sci. 50:331–338.
• Twamley, B.E. 1967. Seed size and seedling vigor in birdsfoot trefoil. Can. J. Plant Sci. 47:603–609.
• Twamley, B.E. 1970. Maternal influence in birdsfoot trefoil on the seedling vigor of the progeny. Can. J. Plant Sci. 50:103–106.
• Twamley, B.E. 1972. Genotypic, phenotypic, and geno-phenotypic methods of selection for seedling vigor in birdsfoot trefoil. Can. J. Plant Sci. 52: 227–232.
• Twamley, B.E. 1974. Recurrent selection for seedling vigor in birdsfoot trefoil. Crop Sci. 4:87–90.

This article has been cited by other articles:

				N. L. Taylor A Century of Clover Breeding Developments in the United States Crop Sci., January 16, 2008; 48(1): 1 - 13. [Abstract] [Full Text] [PDF]

				M. D. Casler and D. J. Undersander Selection for Establishment Capacity in Reed Canarygrass Crop Sci., April 25, 2006; 46(3): 1277 - 1285. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Search for Related Content PubMed Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Agricola Articles by DeHaan, L. R. Articles by Sheaffer, C. C. Related Collections Clover Crop Genetics