Crop Science 42:45-50 (2002)
© 2002 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Effect of Cultivar and Environment on Seed Yield in Alfalfa
Eduardo-Daniel Bolaños-Aguilara,
Christian Huyghe*,a,
Christian Ecallea,
Jacques Hacquetb and
Bernadette Juliera
a INRA, Unité de Génétique et d'Amélioration des Plantes Fourragères, 86600 Lusignan, France
b Fédération Nationale des Agriculteurs Multiplicateurs de Semences, 86600 Lusignan, France
* Corresponding author (huyghe{at}lusignan.inra.fr)
 |
ABSTRACT
|
|---|
Seed yield of alfalfa (Medicago sativa L.) is important in determining the effective distribution of new cultivars to farmers. Many genetic and environmental factors affect seed yield. This study was conducted to explain seed yield variation induced by either environmental conditions or cultivars. We analyzed seed yield, aboveground phytomass, harvest index, and seed yield components for a set of 12 cultivars at four locations across France in each of three years. Each location x year combination was considered an environment. Seed weight, number of pods per inflorescence, number of seeds per pod, and mean seed weight were measured. Mean seed yield was 801 kg ha-1. Large variation in seed yield was found among cultivars and environments. The cultivar x environment interaction was significant. Among environments, seed yield was highly correlated with aboveground phytomass at harvest (r = 0.94) as the lowest seed yields were obtained in the seeding year. The cultivars most adapted to grazing showed the lowest seed yields. Seed yield was genetically correlated with lodging resistance (r = -0.89) and harvest index (r = 0.99). The mean harvest index was 12.7%. The seed weight per inflorescence showed a high broad-sense heritability (0.58) and a high genetic correlation with seed yield (r = 0.91) and with harvest index (r = 0.96). Variation in seed weight per inflorescence was associated with variation in the number of seeds per pod and number of pods per inflorescence. Seed weight per inflorescence appears to have a strong genetic association with seed yield in alfalfa. Environments with high aboveground phytomass potential also have high seed yield potential.
Abbreviations: Y0, seeding year • Y1, first year of regrowth • Y2, second year of regrowth
|
INTRODUCTION
|
|---|
THE SEED YIELD OF ALFALFA cultivars is not recognized as having agronomic value to producers. However, the ability of a cultivar to give a high seed yield determines competitive selling price, a key factor for its effective distribution to farmers (Falcinelli, 2000). Seed yield also determines the economic viability of seed producers. The mean seed yield of alfalfa in France is low (250–500 kg ha-1) in view of its large potential for phytomass production and large numbers of flowers present in seed crops (Lorenzetti, 1981).
Genetic diversity for seed yield and seed yield components in alfalfa was described between and within populations by Bolaños-Aguilar et al. (2000). Similarly, genetic variation was identified for pollen fertility (Viands et al., 1988), ovule fertility (Rosellini et al., 1998), and ease of tripping (Knapp and Teuber, 1994), traits that influence seed yield. However, the actual seed yield improvement achieved in breeding has been limited, as these characters are not easy to measure on large numbers of plants and are not the only limiting factors of seed yield in dense canopies.
Alfalfa seed yield also depends to a great extent upon environmental conditions and agronomic practices. Abu-Shakra et al. (1977) showed that seed yield was significantly affected by the number of forage harvests prior to seed production, and this effect was mainly associated with variation in the number of fertile stems per plant and pods per raceme. Abu-Shakra et al. (1969) and Askarian et al. (1995) reported similar effects as a consequence of different row spacings. In dry environments, Abu-Shakra et al. (1969) and Taylor and Marble (1986) have shown that frequent irrigation was beneficial to seed yield, as it resulted in more pods per inflorescence. However, Hutmacher et al. (1991) pointed out that an excessive water supply may lead to excessive vegetative growth and therefore be detrimental to seed production, possibly through a reduction in the degree of tripping (Goldman and Dovrat, 1980). Steiner et al. (1992b) established that water replacement of 70% of accumulated evapotranspiration was optimal for seed production. Increasing amounts of water increased the number of racemes, but decreased the number of pods per raceme.
Boçsa and Buglos (1983) suggested that a high number of seeds per pod combined with high self-compatibility were the most important selection criteria for high seed yield. Hacquet (1990) found a positive correlation between seed yield and number of seeds per pod across 46 experimental conditions combining cultivars and environments. Taylor and Marble (1986) reported a positive correlation between seed yield and seeds per pod among environments differentiated primarily by water supply. In contrast, Askarian et al. (1995) and Kowithayakorn and Hill (1982) found that the number of seeds per pod was an unimportant yield component, when the principal source for yield variation was plant density. Genter et al. (1997) showed that the number of seeds per pod was significantly reduced by plant defoliation at flowering, though the effect was small (4.1 seeds vs. 4.5 for the nondefoliated control plants).
The objective of this study was to identify characteristics induced either by environmental or genetic factors that explain seed yield variation in alfalfa.
|
MATERIALS AND METHODS
|
|---|
Cultivars and Environments
Twelve registered or experimental cultivars with a fall dormancy rating of 3 to 5 were used for this study (Table 1). Three cultivars (Coussouls, Magali, Medalfa) were Provence types, while the others were Flemish types. Two cultivars (Luzelle and Coussouls) were selected for their adaptation to grazing and more prostrate growth habit. The cultivar Radius was provided by Professor Z. Staszewski (Instytu Hodowli i Aklimatyzacji Roslin, Radzikow, Poland) and was specifically selected for pod setting under cool temperatures in Poland. The experimental cultivar Lp2507, provided by B. Tharel (Barenbrug-Tourneur-Recherches, France), possesses the lp mutation that confers long inflorescences (Bodzon et al., 1998).
View this table:
[in this window]
[in a new window]
|
Table 1. Description and average seed yield of 12 alfalfa cultivars evaluated at four locations in France across 3 yr.
|
|
This material was cultivated at four experimental locations representing the main areas of alfalfa seed production in France: Lusignan (46°26' N, 0°08' E), Condom (43°58' N, 0°23' E), Marans (46°19' N, 1°01' W), and Etoile (44°49' N, 4°53' E). The first location was provided by the Institut National de la Recherche Agronomique, Unité de Génétique et d'Amélioration des Plantes Fourragères, while the other three locations were provided by the Fédération Nationale des Agriculteurs Multiplicateurs de Semences. The soil characteristics (structure, pH) at the different locations are shown in Table 2. Weather data for the different locations across the 3 yr are shown in Table 3.
View this table:
[in this window]
[in a new window]
|
Table 3. Monthly means of minimum (Min) and maximum (Max) air temperatures, rainfall, and evapotranspiration (ET) during the 3 yr at four locations in France.
|
|
All trials were established during spring 1997 (Table 4) in randomized complete blocks with three replicates at each location. Each plot contained three rows 5 m long and 0.42 m apart, with a space of 0.84 m between adjacent plots. The drilling density was 2.4 kg ha-1 for all cultivars and locations. As alfalfa was previously grown at all four locations, no Rhizobium inoculation was used, but the plants were nodulated.
View this table:
[in this window]
[in a new window]
|
Table 4. Dates of sowing, clipping, flowering, and harvest, and average seed yield of 12 alfalfa cultivars grown at four locations in France.
|
|
The trials were conducted across 3 yr (1997, 1998, and 1999), that is, the seeding year (Y0) and the first and second years of regrowth (Y1 and Y2, respectively). Pest control practices were in accordance with those recommended for each location. No irrigation was provided. The plots were not clipped in Y0. In Y1 and Y2, they were clipped at all locations, except at Etoile, where clipping was considered detrimental to crop growth because of the regular summer droughts. The clipping dates are shown in Table 4. Large populations of pollinators were observed at all locations and considered sufficient for optimal pollination of alfalfa as all flowers were tripped in the different trials. Two days before harvest, all trials were defoliated with Diquat (dihydro-6,7 dipyrido[1,2-a:1,2'-c] pyrazidiinium) (600 g a.i. ha-1).
Characters Scored
Flowering date was scored on each plot when 50% of the stems had one flowering inflorescence. Lodging was scored during the flowering period on a scale from 1 (no lodging) to 5 (fully lodged).
Prior to harvest, 30 well-podded inflorescences from 30 main stems were sampled from each plot, and the number of pods per inflorescence was counted. The inflorescences were dried at 40 °C until their weight was stable, or 
72 h. The first two pods from these inflorescences were threshed and the seeds counted and weighed. From these data, the mean seed weight and the number of seeds per pod was calculated. The remaining inflorescences were threshed and the seeds weighed. By recombining with the seed weight measured on the first sample, total seed weight per inflorescence was calculated.
After harvesting the remainder of the plot with a combine harvester, the total vegetative phytomass was collected in each plot and its dry matter content estimated on a subsample of 500 g dried at 80 °C for 48 h. The seeds were dried, cleaned, sieved, and weighed. The total aboveground phytomass was calculated from the dry matter collected at the back of the combine, the seeds, and impurities. Harvest index was calculated as the ratio of seed yield to total aboveground phytomass at harvest.
Statistical Analyses
Analyses of variance were performed using the GLM procedure of SAS (SAS Institute, 1988). The first analysis included cultivars, locations, years, and their interactions in the model. Across all locations and years, the experimental design was analyzed as a split-plot, with locations as main plots. The location effect was tested against the blocks-within-locations mean square. A pooled error term for testing cultivars, years, and their interactions was formed from interactions of blocks. As the location x year interaction was significant for all traits, each location x year combination was later considered as one environment in subsequent analyses. Variance components were calculated using the VARCOMP procedure of SAS and the restricted maximum likelihood method. The standard error of variance component g was calculated (Becker, 1975) as
 | [1] |
where MSi is the mean square of effect i used to estimate variance component g, fi is the degrees of freedom for MSi, and k is the coefficient for 
2g in the expected mean squares. Broad-sense heritabilities (h2) were estimated as
where 2g is the genetic variance, 2ge is the cultivar x environment interaction variance, and 2 is the residual variance. The standard error of these heritabilities was calculated (Becker, 1975) as
where 2p is the phenotypic variance.
Phenotypic correlations were calculated from the means across the three blocks for each cultivar x environment combination. Genetic correlations were calculated from variance-covariance matrices obtained by the MANOVA statement of the GLM procedure (SAS Institute, 1988) for the cultivar effect and cultivar x environment interaction. Correlations across environments were calculated from the variance-covariance matrices obtained by MANOVA statement of the GLM procedure for the environment effect, the cultivar x environment interaction, the blocks-within-environments effect, and the residual term.
The Wricke's ecovalence parameter was used as a measure for phenotypic stability of any given cultivar (Wricke, 1962). Joint regression analysis of cultivar x environment interaction for seed yield was performed according to Finlay and Wilkinson (1963).
|
RESULTS
|
|---|
In all 3 yr of the study and at all four locations, the growing seasons were characterized by rather favorable conditions with warm temperatures and regular rainfall during the summer. The main exceptions were at Lusignan in Y0 and Y1, with a severe drought period during flowering and pod growth.
The flowering dates of the control cultivar Europe are given in Table 4. Flowering occurred earlier in Etoile in Y1 and Y2 as a consequence of the omission of clipping at this location. At Condom in 1998, clipping was performed early in the season, and therefore the crop flowered early.
Environment and Cultivar Effects
Seed yields averaged 801 kg ha-1 (Table 5), which is higher than the average commercial seed yields in France (450 kg ha-1, J. Hacquet, 2000, personal communication) but similar to seed yields achieved in areas most suited for seed production. The year effect was significant for all traits except mean seed weight and lodging score, while the location effect was significant for all traits except lodging score (Table 5). The year x location interaction was always significant. The effect of year on seed yield was strong, as seed yields were lower in Y0 than in Y1 and Y2. The lowest yields were for two Y0 crops, in Lusignan and Condom, while much higher yields were achieved in three environments, that is, Etoile in Y1 and Y2 and Condom in Y1 (Table 4). Aboveground phytomass varied in a similar manner. The cultivar x year interaction was significant for seed yield, harvest index, and mean seed weight, but very small in comparison with the main effects.
View this table:
[in this window]
[in a new window]
|
Table 5. Grand mean and mean squares for seed yield, aboveground phytomass, harvest index, seed yield components, and lodging score for 12 cultivars grown in four locations in France over 3 yr.
|
|
When considering the different year x location combinations as different environments, the effects of cultivars, environments, and their interaction were significant, except the cultivar x environment interaction for mean seed weight (Table 5). The cultivars ranked similarly in each environment. The highest yields were achieved by the cultivar Radius and the experimental cultivar Lp2507. The lowest yields were from the two grazing-type cultivars Coussouls and Luzelle, and Magali (Table 1).
Total aboveground phytomass at harvest was highly variable among environments, from 2.4 to 11.3 Mg ha-1, but it did not vary much between cultivars. However, the harvest index varied with both environments and cultivars, with an average value of 12.7% and a range of 7 to 18%.
Seed weight per inflorescence averaged 155 mg and varied from 110 to 224 mg among cultivars. Variation for mean seed weight and number of seeds per pod was more narrow than for seed weight per inflorescence, and both varied similarly as a consequence of environments and cultivars.
Variance components for the different random effects are given in Table 6. For all traits except the number of pods per inflorescence, the environmental variance exceeded the genetic variance. This was especially true for aboveground phytomass at harvest, for which the environmental variance was 90 times larger than the genetic variance.
View this table:
[in this window]
[in a new window]
|
Table 6. Components of variance and broad-sense heritability (h2), and their standard errors in parentheses, for seed yield and its components, measured on 12 cultivars grown at four locations in France over 3 yr.
|
|
Broad-sense heritability for seed yield was moderately high (0.55 ± 0.23), reflecting a large genetic variance and small interaction and residual variances. The large genetic variance was most likely due to the range of cultivars included in the present study, with two pasture-type cultivars which yielded poorly and two high yielding cultivars (Radius in particular). Heritabilities for harvest index, seed weight, and number of pods per inflorescence were high (0.54 to 0.58). Heritability of total aboveground phytomass at harvest was low, reflecting low genetic variance and a large error variance. Mean seed weight and number of seeds per pod also showed low heritabilities as the genetic variation available for these traits appeared to be narrow in the present set of cultivars in comparison to the error term.
Cultivar x Environment Interaction
The genetic variance was three times greater than the interaction variance for seed yield. The ecovalence parameter indicated one high yielding cultivar, Radius, and one low-yielding cultivar, Coussouls, contributed mostly to the interaction. Among environments, the highest yielding (Etoile in Y1 and Y2, and Condom in Y1) and the lowest yielding (Lusignan in Y0) contributed most to the interaction. All three environments without clipping significantly contributed to the interaction. Joint regression analysis showed that, for each cultivar, there was a high correlation between mean seed yield in a given environment and the performance of the cultivar in this environment. Indeed, the R2 of the joint regression varied from 0.89 for Coussouls to 0.99 for Bella. The slopes of the regression varied from 0.55 for Coussouls up to 1.29 for Radius.
Trait Correlations
At the phenotypic level (Table 7), the correlation between seed yield and harvest index was significant but moderate (r = 0.55). The correlation was high at the genetic level, but low at the environmental level. Thus, variation in seed yield among cultivars in a given environment was related to variation in harvest index. This relationship may be partly related to the lodging. Indeed, the genetic correlation of lodging score with seed yield and harvest index were -0.89 and -0.93, respectively. Across environments, the major source of variation in seed yield was the total aboveground phytomass present at harvest (r = 0.94). Differences in aboveground phytomass between environments was possibly due to differences in leaf area. The presence of high yielding environments (Etoile in Y1 and Y2, and Condom in Y1) strengthened this correlation. The environmental correlation between aboveground phytomass and seed yield remains high even when the Y0 data were removed (r = 0.93). The genetic correlation between seed yield and total aboveground phytomass was high (r = 0.79).
View this table:
[in this window]
[in a new window]
|
Table 7. Phenotypic (below diagonal) (n = 144), genotypic (above diagonal), and environmental (underscored) correlations for seed yield and its components, for 12 cultivars grown at four locations in France over 3 yr.
|
|
The correlation between seed yield and seed weight per inflorescence was particularly high at the genetic level (r = 0.91), but only 0.40 at the environmental level. Seed weight per inflorescence explained a major part of the variation among cultivars for any given environment. It was genetically correlated with number of pods per inflorescence (r = 0.91), and to some extent, number of seeds per pod (r = 0.66). This latter correlation was stronger across environments (r = 0.82). The correlation between seed yield and number of seeds per pod was high at the genetic level, but low across environments.
|
DISCUSSION
|
|---|
These experiments showed wide variation for seed yield potential among cultivars. The cultivars with the lowest yields were Coussouls and Luzelle, which are grazing-type cultivars with a prostrate growth habit. These cultivars were also the most susceptible to lodging. More generally, a strong negative genetic correlation was found between lodging score and seed yield. Lodging is unfavorable to seed setting, as a more compact canopy could limit pollination and possibly induce disease damage to the pods. The lodging-susceptible cultivars showed a low seed weight per inflorescence. For example, the seed weights for Coussouls and Luzelle were 100 and 139 mg per inflorescence, respectively, while the mean value of all cultivars tested was 155 mg. The low seed weights per inflorescence could also be associated with the genetic background of these cultivars, which includes falcata-type progenitors, indicated by the presence of variegated flowers. Crochemore et al. (1998) reported 33 and 28% of plants with variegated flowers in Coussouls and Luzelle, respectively. The lower seed weight per inflorescence on these cultivars may then be associated with lower degrees of tripping, as observed by Goplen and Brandt (1975) on falcata types and with pollinator behavior according to flower color (Steiner et al., 1992a). Thus, the results of the present study do not imply that all grazing tolerant cultivars necessarily produce low seed yields.
For those cultivars with high seed yield potential, the cultivar Radius originated from a breeding program targeting good pod set at low temperatures. The experimental population Lp2507, which yielded similarly to Radius, possesses the long peduncle mutation controlled by a single recessive gene (Bodzon, 1998). This trait could positively contribute to high seed weight per inflorescence.
Variation in seed yield among cultivars was predominantly related to variation in harvest index, as expected since harvest index is calculated from seed yield. Genetic improvement of alfalfa seed yield through a higher harvest index is similar to the objectives in cereal breeding (Donald, 1968) or in grain legume breeding (Huyghe, 1998). The mean harvest index found in the present study (12.7%) was similar to values of 13.6 to 14.4% found by F. Lelièvre (1996, unpublished data) among agronomic treatments. At the genetic level, seed yield was positively correlated with total aboveground phytomass at harvest (e.g., r = 0.79, Table 7). However, it is not possible to draw any conclusion on the relationship between seed production and forage production. Indeed, phytomass at seed harvest also includes seed yield.
Among environments, variation in seed yield was predominantly related to the potential for aboveground phytomass production. Few data are available in the literature to corroborate this relationship. It is interesting that average seed yields achieved in the year of sowing tended to be lower than in the following seasons, as the aboveground phytomass in the year of sowing is usually lower due to lower photosynthetic capacity and poorer crown and root development (Gosse et al., 1988; Khaiti and Lemaire, 1992). This relationship was observed across a range of variation from 2.4 to 11.3 Mg ha-1 of aboveground phytomass at harvest, even though these experiments were optimized for seed production (row spacing and sowing density), pollination, and lodging prevention. If confirmed, the implications of this relationship in terms of agronomic practices would be numerous, especially for defining optimum clipping date and pest control practices. Thus, it seems that both genetic background and agronomic practices may be used to improve seed yield. Indeed, both a higher aboveground phytomass achieved with optimization of agronomic practices and a higher harvest index achieved through choice of cultivar can be combined in seed production.
Seed weight per inflorescence seems to be a reasonable breeding goal for seed yield breeding in alfalfa. This trait showed high broad-sense heritability, high genotypic correlation with seed yield and harvest index, little interaction between cultivars and environments, and it was strongly influenced by the experimental population carrying the long peduncle mutation. No interaction was detected between cultivars and years. Thus, seed weight per inflorescence could be measured in the planting year as an early assessment of yield potential. We calculated the seed weight per inflorescence for each cultivar across the four locations in the planting year. This value showed a significant correlation with mean seed yield in Y1 and Y2 (r = 0.80, P 
0.01), with wide genetic variation for seed weight per inflorescence among the material included in this study. A much wider range of variation was detected among and within populations (Bolaños-Aguilar et al., 2000). Variation in seed weight per inflorescence among environments and cultivars was related both to variation in number of pods per inflorescence and in number of seeds per pod, not to variation in mean seed weight. The seed weight per inflorescence can thus integrate both sources of variation.
In conclusion, higher yielding cultivars had higher seed weight per inflorescence and higher harvest index. Higher yields may also be achieved through modification of agronomic practices that lead to higher aboveground phytomass.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank Professor Z. Staszewski (Instytu Hodowli i Aklimatyzacji Roslin, Radzikow, Poland) and B. Tharel (Barenbrug-Tourneur-Recherches, France) for providing seeds, and J. Jousse and A. Gilly from the Institut National de la Recherche Agronomique, and S. Bador, L.M. Broucqsault, J.L. Chareyron, C. Fourreau and L. Nardi from FNAMS for their technical assistance. Dr. I. Shield is thanked for his editorial modifications of the manuscript.
|
NOTES
|
|---|
The first author was financially supported by CONACYT (National Science and Technology Council from Mexico), INIFAP (Institute for Forestry, Agriculture, and Livestock Research from Mexico) and SFERE (French Society for Educational Resource Exportation).
Received for publication December 7, 2000.
|
REFERENCES
|
|---|
- Abu-Shakra, S., M. Akhtar, and D.W. Bray. 1969. The influence of irrigation interval and plant density on alfalfa seed production. Agron. J. 61:569–571.[Abstract/Free Full Text]
- Abu-Shakra, S., M.L. Bhatti, and H. Ahmed. 1977. Effect of forage harvest frequency on subsequent seed production and pollen quality. Agron. J. 69:428–431.[Abstract/Free Full Text]
- Askarian, M., J.G. Hampton, and M.J. Hill. 1995. Effect of row spacing and sowing rate on seed production of lucerne (Medicago sativa L.) cv. Grasslands Oranga. N.Z. J. Agric. Res. 38:289–295.
- Becker, W.A. 1975. Manual of quantitative genetics. Washington State Univ., Pullman, WA.
- Boçsa, I., and J. Buglos. 1983. Seed yield and some factors influencing seed setting at the variety level in lucerne. Z. Pflanzenzuecht. 90:172–176.
- Bodzon, Z. 1998. Inheritance of spontaneous long peduncle mutation in alfalfa (Medicago sativa L.). Plant Breed. Seed Sci. 42:3–9.
- Bolaños-Aguilar, E.D., C. Huyghe, B. Julier, and C. Ecalle. 2000. Genetic variation for seed yield and its components in alfalfa (Medicago sativa L.) populations. Agronomie (Paris) 20:333–345.
- Crochemore, M.L., C. Huyghe, C. Ecalle, and B. Julier. 1998. Structuration of alfalfa genetic diversity using agronomic and morphological characteristics. Relationship with RAPD markers. Agronomie (Paris) 18:79–94.
- Donald, C.M. 1968. The breeding of crop ideotype. Euphytica 17: 385–403.[ISI]
- Falcinelli, M. 2000. Temperate forage seed production: conventional and potential breeding strategies. International Herbage Seed Production Research Group Newsletter 31:7–15.
- Finlay, K.W., and G.N. Wilkinson. 1963. The analysis of adaptation in a plant breeding programme. Aust. J. Agric. Res. 14:742–754.[ISI]
- Genter, T., E. Deléens, and A. Fleury. 1997. Influence of photosynthetic restriction due to defoliation at flowering on seed abortion in lucerne (Medicago sativa L.). J. Exp. Bot. 48:1815–1823.
- Goldman, A., and A. Dovrat. 1980. Irrigation regime and honey bee activity as related to seed yield in alfalfa. Agron. J. 72:961–965.[Abstract/Free Full Text]
- Goplen, B.P., and S.A. Brandt. 1975. Alfalfa flower color associated with differential seed set by leaf-cutter bees. Agron. J. 67:804–806.[Abstract/Free Full Text]
- Gosse, G., G. Lemaire, M. Chartier, and F. Balfourier. 1988. Structure of population of Medicago sativa L. and dynamics of stem competition for light during regrowth. J. Appl. Biol. 25:609–617.
- Hacquet, J. 1990. Genetic variability and climatic factors affecting lucerne seed production. J. Appl. Seed Prod. 8:59–67.
- Hutmacher, R.B., J.J. Steiner, S.S. Vail, and J.E. Ayars. 1991. Crop water stress index for seed alfalfa: influences of within-season changes in plant morphology. Agric. Water Manage. 19:135–149.
- Huyghe, C. 1998. Genetics and genetic modifications of plant architecture in grain legumes: a review. Agronomie (Paris) 18:383–411.
- Khaiti, M., and G. Lemaire. 1992. Dynamics of shoot and root growth of lucerne after seeding and after cutting. Eur. J. Agron. 1:241–247.
- Knapp, E.E., and L.R. Teuber. 1994. Selection progress for ease of floret tripping in alfalfa. Crop Sci. 34:323–326.[Abstract/Free Full Text]
- Kowithayakorn, L., and M.J. Hill. 1982. A study of herbage and seed production in lucerne (Medicago sativa L.) under different plant spacing and cutting treatments in the seeding year. Seed Sci. Technol. 10:3–12.
- Lorenzetti, F. 1981. Relationship between dry matter and seed yield in leguminous forage plants. p. 57–74. In G. van Bogaert (ed.) Breeding high yielding forage cultivars combined with high seed yield. Rep. of the meeting of the Eucarpia Fodder Crops Section, Merelbeke, Belgium. (publisher unknown).
- Rosellini, D., F. Lorenzetti, and E.T. Bingham. 1998. Quantitative ovule sterility in Medicago sativa. Theor. Appl. Genet. 97:1289–1295.[ISI]
- SAS Institute. 1988. SAS user's guide: Statistics. 6th ed. SAS Inst., Cary, NC.
- Steiner, J.J., P.R. Beuselink, R.N. Peaden, W.P. Kojis, and E.T. Bingham. 1992a. Pollinator effect on crossing and genetic shift in a three-flower-color alfalfa population. Crop Sci. 32:73–77.[Abstract/Free Full Text]
- Steiner, J.J., R.B. Hutmacher, S.D. Gamble, J.E. Ayars, and S.S. Vail. 1992b. Alfalfa seed water management: Crop reproductive development and seed yield. Crop Sci. 32:476–481.[Abstract/Free Full Text]
- Taylor, A.J., and V.L. Marble. 1986. Lucerne irrigation and soil water use during bloom and seed set on a red-brown earth in south-eastern Australia. Aust. J. Exp. Agric. 26:577–581.
- Viands, D.R., P. Sun, and D.K. Barnes. 1988. Pollination control: mechanical and sterility. p. 931–960. In A.A. Hanson et al. (ed.) Alfalfa and Alfalfa Improvement. Agron. Monogr. 29. ASA, CSSA, and SSSA, Madison, WI.
- Wricke, G. 1962. Uber eine Methode zur Erfassung der Ökologischen Streubreite in Feldversuchen. Z. Pflanzenzuecht. 47:92–96.
This article has been cited by other articles:
|
|
|
|
 
T. Zhang, X. Wang, J. Han, Y. Wang, P. Mao, and M. Majerus
Effects of Between-Row and Within-Row Spacing on Alfalfa Seed Yields
Crop Sci.,
March 19, 2008;
48(2):
794 - 803.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. Dordas
Foliar Boron Application Improves Seed Set, Seed Yield, and Seed Quality of Alfalfa
Agron. J.,
June 5, 2006;
98(4):
907 - 913.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
