Crop Science 40:964-967 (2000)
© 2000 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Vestigial Corolla in Flowers of Birdsfoot Trefoil
P.R. Beuselincka and
R.L. McGrawb
a USDA-ARS, Plant Genetics Research Unit, Columbia, MO 65211 USA
b Dep. of Agronomy, University of Missouri, Columbia, MO 65211 USA
beuselinckp{at}missouri.edu
 |
ABSTRACT
|
|---|
A naturally occurring floral mutant is infrequently observed in some populations of birdsfoot trefoil (Lotus corniculatus L.). The petals of mutant flowers are folded inward and do not extend fully, producing a vestigial corolla (vc), while other floral parts remain normal. The style extrudes through a gap between the rudimentary keel and the folded standard. Consequently, the stigma and style extend beyond the corolla and are exposed. The spatial separation between stigma and anthers could facilitate hand-pollination as it eliminates the need for emasculation. The objective of this study was to determine the fertility of L. corniculatus genotypes expressing vc floral characters. Sixteen genotypes of L. corniculatus expressing the vc phenotype were collected from field grown populations. Pollen from vc genotypes was compared with pollen from genotypes of the germplasm MU-81 used as controls. All vc genotypes produced some pollen that appeared normal, although the quantity of such pollen was variable among the vc genotypes. Mean pollen germination among vc genotypes was 22%, compared with 50% for MU-81. Only five of 16 vc genotypes produced pods when crossed to genotypes of MU-81 and both pod and seed set were less than in control crosses. Normal and abnormal ovules were observed in ovaries of vc genotypes and the reduced fertility may have resulted from smaller, incompletely developed ovules. Unidirectional reversion of the vc genotypes, from abnormal to normal flower morphology, was common. Using the vc mutant to eliminate the need for emasculation in hand-pollination programs depends on its fertility. The combination of phenotypic instability and reduced fertility make the vc mutant less desirable for use in breeding programs.
Abbreviations: vc, vestigial corolla • vf, vestigial floret
|
INTRODUCTION
|
|---|
BROADLEAFED BIRDSFOOT TREFOIL is a cross-pollinated, self-incompatible legume (Fairey and Smith, 1999). The inflorescence is a typical umbel consisting of yellow perfect flowers with 10 stamens and a simple pistil (Kirkbride, 1999). A naturally occurring floral mutant has been infrequently observed in some populations of L. corniculatus. The petals of the mutant flowers are folded inward and do not extend fully producing a vestigial corolla (vc) while other floral parts remain normal. The style extrudes through a gap between the rudimentary keel and the folded standard. Consequently, the stigma and style extend beyond the corolla and are exposed. The stamens are embedded in the folded petals. The mutant flowers are smaller than normal flowers of L. corniculatus, ball shaped, and show little color. Mutant flower buds are round, whereas those of normal flowers are boat shaped. The leaves and stems of the mutants are phenotypically normal.
Induced mutants similar to vc have been reported in L. corniculatus. Vestigial floret (vf) was one mutant recovered after exposing seeds of L. corniculatus to chemical mutagenesis (Therrien and Grant, 1982). Flowers of vf mutants had fully developed stigmas, styles, and ovaries, with all other parts being vestigial and slightly protruding beyond the calyx. The style and stigma extruded from the corolla. Anthers were either absent or few, and flowers were completely sterile. The vf mutant was concluded to be tetrasomic-recessive. Webb et al. (1996) observed a floral abnormality that resembled vf, among a line of regenerated L. corniculatus transformed using Agrobacterium tumefaciens. Their transformants were female sterile and male fertile.
Exposed stigmas preclude self-pollination of the stigma because of flower structure abnormalities. The spatial separation between stigma and anthers could facilitate cross-pollination when artificial hybridization is practiced as it eliminates the need for emasculation. Flowers with protruding stigmas would be useful as female parents in such cases, but fertility must be assured.
The objective of this study was to determine the fertility of L. corniculatus expressing vc floral characters.
|
Materials and methods
|
|---|
Sixteen genotypes of L. corniculatus expressing the vc phenotype were collected from field-grown populations of the germplasm MU-81 and `Norcen'. Five ramets of each genotype were vegetatively propagated using stem cuttings. Three genotypes of MU-81 that produced normal flowers were similarly propagated for use as controls. All plants were transplanted into a commercial peat-based growing medium (Promix, Premier Horticulture, Dorval, QC, Canada) and maintained in a greenhouse.1
Natural lighting was supplemented with high pressure sodium and metal halide lighting of 447 µmol photon m-2 s-1 to provide a 24-h day length. Greenhouse temperatures were maintained between 18 and 25°C. Plants were trimmed and fertilized as needed.
Morphology and viability of fresh pollen from vc genotypes were compared with pollen from control genotypes. Pollen grains were collected from anthers of 10 flowers of each genotype, placed on a glass slide, then microscopically examined for gross differences in size and shape. A single drop of acetocarmine was added and nonviable pollen was distinguished by its failure to stain. Counts of stained and unstained pollen were made among 100 grains from each of four random fields of view. Pollen viability was measured as the percentage of stainable pollen.
Fresh pollen grains were distributed on an agar-based germination medium (Spiss and Hittle, 1979). Pollen germination for each genotype was microscopically assessed from counts of 50 pollen in each of four random fields of view at 10x. A one-way classification analysis of variance was performed on arcsin-transformed percentage of stained pollen and germination data. Means of vc and control genotypes were compared using a protected LSD (P 
0.05).
Reciprocal matings between vc and control genotypes were conducted to determine fecundity. Flowers of control genotypes were emasculated 2 d before anthesis, then hand-pollinated 2 d later with fresh pollen from each of the 16 vc genotypes or a mix of pollen from the other control genotypes. Pollen from vc genotypes was collected from flowers that were removed from donor plants and dissected to liberate the anthers and pollen. All flowers within an umbel were pollinated by one pollen source and tagged. Flowers of each of the 16 vc genotypes were hand-pollinated using mixtures of pollen from the other vc genotypes or the three control genotypes. Emasculation was not required on the vc genotypes because the stigmas were physically separated from the anthers. Pollinated plants were grown in greenhouse conditions as described above. Pods were harvested at physiological maturity and hand-threshed. Numbers of pods and seed set per pod were calculated for each cross. A one-way classification analysis of variance was performed on seed data. Means of vc and control genotypes were compared using a protected LSD (P 0.05).
Seeds produced in the crosses described above were mechanically scarified and then surface-sterilized in a 20% (v/v) commercial bleach solution and rinsed three times with sterile distilled water. The seeds were germinated according to Maxon (1994), except that the seeds were plated on water agar (1g agar 100 mL-1 water), then germinated under continuous lighting of 70 µmol photons m-2 s-1. Germination was expressed as a percentage of the number of germinated seed at 12 d. After 12 d, seedlings were transferred from petri dishes into commercial soil mix and grown in the greenhouse. Plants received natural lighting until they were induced to flower by extending the photoperiod beyond 16 h using supplemented lighting as described above. Type of flower development for vc x vc, control x vc, vc x control, and control x control progeny was recorded.
Flowers of vc and control genotypes were collected and mature pistils dissected from floral buds. Pistils were fixed in FPA50 (formalin, propionic acid, 50% ethanol 50:50:90 by volume) for 24 h at room temperature. Fixed pistils were transferred to 70% ethanol and stored at 4°C until use. Fixed pistils were placed in clearing fluid (Herr, 1971) for 24 h at room temperature. Then they were placed on a Raj slide (Smith, 1973) and ovules were dissected from the ovary. Ovaries were examined for any abnormalities. Dissected ovules were examined for comparative development of megagametophytes derived from vc or control genotypes with a Leitz Diaplan (Leitz/Wild, Rockleigh, NJ) microscope with Normarski interference. Microscope images of ovules were digitized and measured using imaging software.
|
Results and discussion
|
|---|
Anthers of normal L. corniculatus flowers, from the control genotypes, produced a considerable amount of globose-prolate pollen. Anthers from vc plants did not dehisce easily and produced less pollen than normal flowers. Globose-prolate pollen comprised 75% of the pollen from both vc and control genotypes. Pollen with irregular shape was considered abnormal. All vc genotypes produced some pollen that appeared normal, although the quantity was variable.
The control genotypes and all but one of the vc genotypes produced pollen that stained red in acetocarmine solution. Pollen from vc genotypes had a lower percentage staining than the control genotypes (Table 1)
. Mean percentage staining of control pollen was significantly higher
at 97% compared with 76% for the vc genotypes. Variation for percentage staining of pollen among the sixteen vc genotypes ranged from 0 to 94%.
View this table:
[in this window]
[in a new window]
|
Table 1 Overall stainability and in vitro germination of pollen from the vestigial corolla (vc) mutant of Lotus corniculatus in comparison with MU-81
|
|
Pollen germination occurred as early as 1 h after placement on agar medium. Mean in vitro germination of pollen from vc genotypes was significantly lower
at 
22% compared with 51% for pollen of control genotypes (Table 1). Pollen germination among vc genotypes ranged from 0 to 48%. No pollen germinated from the vc genotype that did not produce stainable pollen.
Pollinations of emasculated flowers of control genotypes with pollen from the individual vc genotypes successfully produced pods and seeds for all crosses, except for the vc genotype that failed to produce pollen that either stained or germinated (Table 2) . Mean pod set from these crosses averaged 23.7% compared with 25.9% when pollen from control genotypes was used. Mean seed per pod for control x vc crosses were not significantly different (P > 0.05) from that observed for the control x control crosses. These findings further indicate that pollen from vc genotypes was functional. The mean number of seed per pod for the control was below the normal average (10–15 seeds pod-1) as reported by Beuselinck and Grant (1995), but this result might be an expression of incompatibility since the pollen mixture used was derived from only three genotypes. In contrast, few pods and significantly fewer
seeds resulted from pollinations of vc genotypes with pollen from the controls. Pods were obtained on only five of the sixteen vc genotypes. Mean pod set was 1.9% for the vc x control crosses with an average of 4.2 seeds per pod. Pollinations of vc x vc produced only one pod in 121 crosses (0.8%) and that pod had three seeds.
View this table:
[in this window]
[in a new window]
|
Table 2 Comparison of pod and seed set on the vestigial corolla (vc) mutant genotypes and MU-81 pollinated with pollen of vc mutant genotypes and MU-81
|
|
Pollination results are consistent with data from other crumpled petal mutants. Reduced female fertility was noted in floral mutants of red clover (Trifolium pratense L.) (Taylor and Snead, 1986), pea (Pisum sativum L.) (Sharma and Aravindan, 1971), and cowpea [Vigna unguiculata (L.) Walp.] (Rachie et al., 1975). Reduced fertility appears coincident in flowers with an exposed stigma. Protrusion of the stigma was positively associated with female sterility in white cabbage (Brassica oleracea sp. capitata L.) (Nieuwhof, 1961), in canola (Brassica napus L.) (Polowick and Shawney, 1988), and in tomato (Lycopersicon esculentum Mill.) (Currence, 1944; Rick and Robinson, 1951). Despite this trend, exsertion of the style does not always imply sterility, as demonstrated by crumpled petal mutants of lentil (Lens culinaris Medikus) (Sharma and Sharma, 1980).
All plants grown from seed produced between vc x control genotypes and reciprocal crosses produced normal flowers only (Table 3)
. Of the three seed produced by the vc x vc crosses, only one plant grew to maturity and it produced flowers characteristic of the vc genotypes.
View this table:
[in this window]
[in a new window]
|
Table 3 Number of seed germinated, seedlings produced, and type of flower produced by crosses between normal MU-81 and vestigial corolla (vc) mutant genotypes
|
|
Microscopic examination of cleared ovules of control genotypes contained features of mature megagametophytes of L. corniculatus and agree with the observations of Rim et al. (1990). In contrast, three types of ovules were observed in ovaries of vc genotypes: (i) normal megagametophyte except that the nucellus was exposed and frequently extruded from the integuments, (ii) normal megagametophyte except inner integuments and nucellus were shorter than normal, and (iii) no internal development of a megagametophyte. The first two ovule types were the most common. Additionally, ovules from vc ovaries were significantly shorter than ovules of the control genotypes (Table 4)
, but were of comparable width.
Microscopic examinations of ovules from normal flowers of control genotypes revealed that mature megagametophytes of L. corniculatus contained all the complements of a normal, polygonum type of embryo sac, except for the lack of antipodals, as was reported earlier by Rim et al. (1990) and Wojciechowska (1963). In contrast, most of the ovules from vc genotypes had no micropyle and the outer integument failed to surround the nucellus. In 75% of the aberrant ovules, a portion of the nucellus was exposed and no micropyle was present. Absence of the micropyle might have contributed to sterility since it is through the micropyle that the microgametophyte enters the megagametophyte. Failure of the integuments to surround and protect the gametophyte was cited as the cause of sterility in alfalfa (Medicago sativa L.) (Bingham and Hawkins-Pfeiffer, 1984). The mechanism of sterility was found to be the incomplete integument development that left the nucellus and female gametophyte exposed (unprotected) and as a result, both nucellus and gametophyte ruptured and were destroyed.
Analysis of ovary development in the vc mutants revealed another probable cause of sterility. One of the abnormalities observed was rupture of the adaxial suture of the ovary resulting in exposure of some ovules. The rupture was apparently caused by the anther(s), which rests on the suture of the ovary. Open ovaries were reported in a yellow lupine (Lupinus luteus L.) mutant (Edwardson and Corbett, 1959) and in the ms-3 mutant of pea (Myers, 1984). In both cases, female fertility was lower than normal. It is more likely that exposed ovules do not survive if they are successfully fertilized.
Reversion of the vc genotypes, from producing abnormal to normal flower morphology, was common. The reversion was gradual, with abnormal flowers produced by inflorescences of a stem, but then at a later date, the inflorescenses would be a mix of abnormal and normal flowers, and finally the complete stem and/or plant might produce flowers of normal appearance. Reversion was unidirectional; that is, no vc genotypes having to the production of normal flowers ever produced abnormal vc flowers.
Even though some functional ovules and pollen are produced, the vc floral mutants do not set pod under field conditions, possibly because they are unattractive to insect pollinators. The unattractiveness is a result of crumpling of the petals that usually act as a landing platform for pollinators. The presence of vc mutants in natural populations of L. corniculatus could be due to pollen transfer between vc and normal genotypes, since the vc mutants were semifertile. However, the combination of abnormal flowers unattractive to pollinators and nondehiscent anthers suggests that pollen transfer from vc mutants to normal plants does not occur. It is more likely that vc mutants in a population result from spontaneous mutation. The ability of the vc mutants to revert to normal types suggests that this floral mutant is controlled by a mutation susceptible site or transposable element.
Like other crops, L. corniculatus can benefit from hybridization techniques employed by plant breeders. The emasculation of flowers required for controlled pollination is tedious and laborious. The vc mutant with its exposed stigma could facilitate hand pollination without emasculation. In addition, because of its easily identifiable phenotype, the vc mutant could be used as a genetic marker. The combination of phenotypic instability and reduced fertility when used as the maternal parent make the vc mutant less desirable for use in breeding programs. Consequently, the future utility of the vc mutant is probably limited to its characterization as a floral anomaly.
|
ACKNOWLEDGMENTS
|
|---|
The authors wish to recognize the contributions that J. Samek, I. Nyrikabbi, and L. Gebrehiwot made to the culture and maintenance of the research materials and data collection.
|
NOTES
|
|---|
Contribution of the Missouri Agric. Exp. Stn. Journal Series no. 12915.
1 Mention of a trademark, vendor, or proprietary product does not constitute a guarantee or warranty of the product by the USDA or the Univ. of Missouri and does not imply its approval to the exclusion of other products or vendors that may also be suitable. 
Received for publication July 2, 1999.
|
REFERENCES
|
|---|
- Beuselinck P.R., Grant W.F. Birdsfoot trefoil. In: Barnes R.F., et al. , ed. Forages. 5th ed. Vol. 1. An introduction to grassland agriculture. Ames: Iowa State Univ. Press, 1995:237-248.
- Bingham E.T., Hawkins-Pfeiffer J. Female sterility in alfalfa due to a recessive trait retarding integument development. J. Hered. 1984;75:231-233.[Abstract/Free Full Text]
- Currence T.M. A combination of semi-sterility with two simply inherited characters that can be used to reduce the cost of hybrid tomato seed. Proc. Am. Soc. Hortic. Sci. 1944;44:403-406.
- Edwardson J.R., Corbett M.K. A macro mutation in yellow lupine (Lupinus luteus L.). J. Hered. 1959;50:167-170.[Free Full Text]
- Fairey D.T., Smith R.R. Seed production in birdsfoot trefoil, Lotus species. In: Beuselinck P.R., ed. Trefoil: The science and technology of Lotus. Madison, WI: CSSA, 1999:145-166 CSSA Spec. Publ. 28..
- Herr J.M., Jr. A new clearing-squash technique for the study of ovule development in angiosperms. Am. J. Bot. 1971;58:785-790.
- Kirkbride J.H., Jr. Lotus systematics and distribution. In: Beuselinck P.R., ed. Trefoil: The science and technology of Lotus. Madison, WI: CSSA, 1999:1-20 CSSA Spec. Publ. 28..
- Maxon S. Rules for testing seeds. Assoc. Official Seed Anal. 1994;16:58.
- Myers, J.R. 1984. Male sterility, structure sterility, and 2n pollen in the pea (Pisum sativum L.). Ph.D. diss. Univ. of Wisconsin-Madison. (Diss. Abstr. 45-1084B).
- Nieuwhof M. Male sterility in some cole crops. Euphytica 1961;10:351-356.
- Polowick P.L., Shawney V.K. High temperature induced male and female sterility in canola (Brassica napus L.). Ann. Bot. (London) 1988;62:83-86.[Abstract/Free Full Text]
- Rachie K.O., Rawal K., Franckowiak J.D., Akinpelu M.A. Two outcrossing mechanisms in cowpea, Vigna unguiculata (L.) Walp. Euphytica 1975;24:159-163.
- Rick C.M., Robinson J. Inherited Defects of floral structure affecting fruitfulness in Lycopersicon esculentum. Am. J. Bot. 1951;38:639-652.
- Rim Y.W., Beuselinck P.R., McGraw R.L., Somers D.A. Megagametophyte development in Lotus corniculatus, L. conimbrecensis, and their protoplast fusion hybrid. Am. J. Bot. 1990;77:1084-1094.
- Sharma, B., and R.V. Aravindan. 1971. "Crumpled petal" a new mutation in Pisum sativum causing mechanical sterility. Pisum Newsl. 3:50–51.
- Sharma S.K., Sharma B. Note on gamma-ray-induced crumpled petal mutation in lentil. Indian J. Agric. Sci. 1980;51:119-120.
- Smith B.B. The use of a new clearing technique for the study of early ovule development, megasporogenesis, and megagametogenesis in five species of Cornus L. Am. J. Bot. 1973;60:322-338.
- Spiss L., Hittle C.N. Pollen germination and storage in birdsfoot trefoil (Lotus corniculatus). Acta Agrar. Silvestria, Ser. Agrar. 1979;18:267-277.
- Taylor N.L., Snead L.D. Inheritance of rudimentary corolla in red clover. J. Hered. 1986;77:371-372.[Abstract/Free Full Text]
- Therrien M.C., Grant W.F. Induction of two mutants in birdsfoot trefoil (Lotus corniculatus) by x-rays and chemical mutagens. Can. J. Plant Sci. 1982;62:957-963.
- Webb K.J., Gibbs M.J., Mizen S., Skøt L., Gatehouse J.A. Genetic transdormation of Lotus corniculatus with Agrobacterium tumefaciens and the analysis of the inheritance of transgenes in the T1 generation. Transgenic Res. 1996;5:303-312.
- Wojciechowska B. Embryological studies in the genus Lotus. Part 1. Fertilization and seed development following open- and self-pollination of Lotus corniculatus. L. Genet. Polonica 1963;4:53-63.
TOP
ABSTRACT
INTRODUCTION
