The degree of regularity of each stage of microsporogenesis in alfalfa (2n=4x=32) and its effect on fertility were studied to understand what is typical in alfalfa cultivars. Alfalfa is essentially a bivalent forming autotetraploid where regular meiotic stages predominate in normal plants, but every plant studied also had some degree of irregularity. This included 17 ‘Vernal’ plants, four ‘Arrow’ plants, three ‘WISFAL’ and one plant each from ‘Saranac’ and ‘Blazer XL’ which were all basically similar in this study. Univalents and the various ways they behave throughout meiosis cause the main irregularities.

A basic mechanism producing univalents that is common to all tetraploid alfalfa plants is the disassociation of some multivalents between pachytene and Metaphase I. This study and others in the literature show a decrease from 2-4 multivalents at pachytene to less than one on the average at Metaphase I. Even when multivalents persisted at Metaphase I in this study, the most common type was a chain of three chromosomes plus the obligatory univalent. Univalents cause irregularities by lying on or off the division plate in Metaphase I, dividing precociously or moving to one division pole, or sometimes doing neither and remaining at the cell equator at Anaphase I. Rare bridges at Anaphase II were interpreted as due to delayed chromatid separation from chromosomes that were either bivalents or univalents in earlier stages. Bivalents occasionally also could lag in Anaphase I and generate subsequent irregularities.

Irregularities of all types resulted in tetrad irregularities. These included micronuclei, and tetrads and pollen of varying sizes. The tetrad stage is a defining stage in terms of the degree of irregularity in a plant. Pollen is somewhat less defining, but still a good indicator of the regularity or irregularity of microsporogenesis in alfalfa. There was no apparent association between the degree of regularity of microsporogenesis and self and cross seed set. Other genetic factors have a larger effect on fertility than the background irregularities observed in this study. The study of irregularities in alfalfa cultivars continues in the abstract by Tsatsenko et al. that reports on 2n pollen frequencies in cultivars.
 
 
 
 
 
 
 
 

___________________

Dep. of Agronomy, Univ. of Wisconsin, 1575 Linden Dr., Madison, WI 53706. Contribution from the Dep. of Agronomy, Univ. of Wisconsin Agric. Exp. Stn., Madison. Research supported by the College of Agricultural and Life Sciences, Univ. of Wisconsin-Madison. Received. *Corresponding author.

___________________

Cultivated alfalfa Medicago sativa behaves as an autotetraploid (2n=4x=32) (Stanford et al., 1972). The four homologous chromosomes pair mainly as bivalents, with fewer than one quadrivalent per cell on the average (Reeves, 1930; Cooper, Brink and Albrecht, 1937; Julen, 1944; Atwood and Grun, 1951; McCoy and Bingham, 1988). Tetrasomic genetic ratios (Stanford, 1951; Bingham, 1973; Quiros, 1982) indicates that the bivalent pairing is random and not preferential.

Most studies of microsporogenesis in cultivated alfalfa usually described either the early stages of meiosis, or focused on the direct outcome as tetrads and pollen (Reeves, 1930; Gillies, 1970; Armstrong, 1971; Gillies and Lesins, 1971; Smith and Murphy, 1986). Patterns of abnormality exhibited by the meiotic chromosomes through the duration of the microsporogenesis have been reported (Atwood and Grun, 1951; Grun, 1951), but have not received major attention until now.

'Vernal' alfalfa was featured in the present study for a number of reasons. Vernal is a benchmark cultivar released in 1953 by the Wisconsin Agriculture Experiment Station. Vernal is still being marketed to farmers. It is recognized for its winterhardiness, bacterial wilt resistance, yield and fine quality (Graber, 1955). Vernal was synthesized from eleven parental clones. Five clones descended from a cross of wild diploid Medicago falcata x cultivated tetraploid M. sativa (Ledingham, 1940), and six clones from the variety 'Cossack' that had been selected for persistence and bacterial wilt resistance. Cossack also contained germplasm from M. falcata. Thus we believe Vernal contains at least 25 per cent M. falcata germplasm. Oldemyer and Brink (1953) studied the fertility of hybrids between M. falcata and M. sativa similar to the ones used as parents of Vernal, and found that fertility was normal.

Lowe et al. (1972) stated that Vernal has made the largest contribution to alfalfa adaptation in the North Central United States. In 1972, Vernal was grown on about 30 per cent of the North Central acreage and was the most widely used alfalfa variety in the USA. By 1977, five commercial varieties had been developed from parents selected exclusively from Vernal, and Vernal was in the pedigree of 20 other varieties (Barnes et al. 1977). By 1998, Vernal had been used as a parent in 140 varieties; it was a grandparent in 216, a great grandparent in 173, and so on for a total of 636 varieties. This means that germplasm from Vernal and its derivatives is represented in more than two-thirds of the 802 alfalfa varieties released from 1962 in which the pedigree could be determined (W.P. Kojis and E.T. Bingham, UW-Madison Unpubl. 1999). These varieties were used in the development of still other varieties in subsequent cycles of alfalfa breeding.

Vernal plants used in the study were grown from certified seed and sampled in the year of establishment. Thus, natural selection was minimal. The complex pedigree of Vernal is discussed in the introduction. The following selected clones and plants were also examined. Clone 'P' of synthetic variety ’Blazer XL’, was selected several years ago as outstanding in vigor, self fertility, and self progeny performance. In addition, clone P is heterozygous (duplex) for the cauliflower head - simple leaf trait. The male sterile clone ’6-4ms’ of ’Saranac’ alfalfa has been used on the project for more than 20 years. It is known that 6-4ms has uniparental transmission of the mitochondrial genome, and biparental transmission of the chloroplast genome (Forsthoefel, et al., 1992). ’WISFAL’ is tetraploid M. falcata developed from several diploid M. falcata plant introductions by backcrossing tetraploids into the diploid via 2n eggs (Bingham, 1991, 1993). ’Arrow’ is a synthetic cultivar the parents of which contained at least 6% Vernal. Arrow clone A-1 was selected from a large population as producing up to 10 seeds per pod. All other Arrow plants were random samples from Arrow.

Flower buds of 17 Vernal alfalfa plants were collected from spaced plants in the field in August 1995. They were fixed in three parts 95% ethanol : one part glacial acetic acid solution for 24 hours and then were stored in 70% ethanol at room temperature. Cytological smear preparations were made according to standard acetocarmine technique. A minimum of 100 pollen mother cells (PMCs) at each stage of meiosis was examined in 11 Vernal plants and a smaller number of PMCs at second meiotic division was examined in six additional Vernal plants, and the other plants in the study. Each type of irregularity was recorded on the basis of PMCs exhibiting it compared to the total number of PMCs examined. Correlation coefficients were calculated for abnormalities observed through the duration of the division.

The above Vernal plants and about 25 additional ones were transferred to the greenhouse where 32 plants were used for fertility studies and the remainder was dropped due to transplant losses, and delayed flowering. Fertility studies were based on about 50 flowers selfed and 50 flowers crossed using bulk pollen from the Vernal population. Similarly, about 30 Arrow plants were transferred to the green house, when self and cross fertility data were collected on 23 plants.

Microsporogenesis in the selected clones and Arrow plants was studied using buds collected in the greenhouse between January and April 1996. Self and cross fertility data were collected on these materials in the same period Vernal alfalfa was studied. Methods were the same as for Vernal, except bulk pollen was from Arrow.

The characteristics of regular stages of microsporogenesis in tetraploid alfalfa from diakinesis to microspores are shown in Figure 1. A complete set of regular stages in tetraploid alfalfa and has not been published previously. Most PMCs in all plants studied had 16 bivalents at diakinesis (Fig. 1a) and Metaphase I (Fig. 1b). Similarly, most PMCs were free of lagging chromosomes and chromosome bridges at Anaphase I (Fig. 1c), the division spindles were aligned properly at Metaphase II (Fig. 1d), and they were free of lagging chromosomes and micronuclei at Anaphase II (Fig. 1e). Simultaneous cytokinesis following Telophase II formed four microspores (Fig. 1f). Although regular stages predominated in all materials, essentially every plant also had some degree of irregularity at every stage (Tables 1 and 2).

Frequency of PMC’s with multivalents was similar in all materials. Microsporocytes with multivalents averaged 28% at diakinesis and 19% at Metaphase I. In the Vernal plants that received intensive study, multivalents averaged 34% at diakinesis and 17% at Metaphase I (Table 1). The decrease at Metaphase I was apparently due to insufficient crossing over to maintain some multivalents, which became bivalents, bivalents and univalents, or a trivalent and a univalent.

Departures from regularity are featured in Figure 2. The irregularities are from the Vernal plants in which a minimum of 100 cells per stage were analyzed, and are considered typical of irregularities in all other plants studies (Table 2). It was sometimes difficult to trace the behavioral patterns of the chromosomes at pachytene due to tight chromosome grouping and overlapping, and no quantitative data were collected. Nonetheless, some abnormalities were photographed including less than perfect pairing (Fig. 2a), an apparent translocation (Fig. 2b), and occasional switching of pairing partners (Fig. 2c). Small loops were observed which may have resulted from a segment deficiency in one of the pairing partners.

Diakinesis is a difficult stage to analyze in alfalfa, but a cell is illustrated with mostly bivalents, a ring quadrivalent, and at least two univalents (Fig. 2d). Metaphase I is easier to analyze and cells with bivalents and multivalents often also contained univalents. When multivalents occurred, a chain of three chromosomes accompanied by a univalent at Metaphase I was common (Fig 2e). Importantly, univalents also were observed in Metaphase I cells where the other chromosomes were bivalents (Fig. 2f) at an overall frequency of 10% (Table 1).

Lagging chromosomes at Anaphase I were common with lagging univalents more common than lagging bivalents (Table 1). Fig. 2g illustrates lagging chromosomes at Anaphase I that likely represent separation of lagging univalents and bivalents. At Metaphase II, univalents that divided precociously at Anaphase I were positioned off the Metaphase plate (Fig. 2h). Distinct chromatin entities positioned off the plate at Metaphase II (Fig 3a) likely were from univalents that remained at the cell equator at Anaphase I. It is interesting that the frequencies of mispositioned chromosomes at Metaphase II (Table 2) reflects the frequency of univalents and lagging chromosomes in Table 1. The correlations between PMCs with mispositioned chromosomes at Metaphase II and lagging chromosomes at Anaphases I and II were r=0.602 and r=0.588, respectively. Although the correlations were not significant at P=0.05 (r=0.671), they were similar. Bridges at Anaphase II (Fig. 3b) were equally rare in all genotypes (0.01-0.02; data not reported in tables) and were interpreted as due to delayed chromatid separation at Anaphase II. Cells at Anaphase II and Telophase II with numerous laggards (Fig. 3c) occurred consistently across genotypes (Table 2) and averaged 24%. These laggards at telephase II probably accounted for the variation in size of nuclei (Fig. 3d), and proceeded to form micronuclei occurring with tetrads of microspores (Fig. 3e). Mature pollen grains also often varied in size (Fig. 3f).

Micronuclei observed at the tetrad stage and microspores and pollen of varying sizes reflected the irregularities seen through the stages of microsporogenesis. Univalents appear to be the fundamental cause of most of the irregularities. The most commonly observed irregularity was lagging chromosomes seen at Anaphase II. In Vernal the correlation between PMCs with lagging chromosomes at Anaphase II and tetrads with micronuclei was highly significant, r=0.953, P=0.01. The number of micronuclei per tetrad varied between one and four. The frequency of tetrads with micronuclei was as low as 0.06 in Vernal-36 and as high as 0.44 in Vernal-11 (Table 2). In Arrow, the range in frequency was 0.12 to 0.23.

The results from the tetrad analysis of all the other cultivars supported the meiotic profiles which were established prior to the tetrad stage. The average frequency of tetrads with microspores containing one, two, and even three micronuclei was 0.10 for clone Blaser XL-P, 0.17 for the samples of Arrow, 0.29 for those of WISFAL, and 0.69 for the male sterile clone Saranac 6-4ms (Table 2).

The degrees of regularity of microsporogenesis of the various tetraploid alfalfa materials are graphed in Figure 4. Figure 4 provides profiles of the materials but does not establish cultivar differences because of different sample sizes. Clone Blazer XL-P was selected for its excellent reproductive characteristics and was the most regular. Interestingly, Blazer XL-P had the highest frequency bivalents and lowest frequency of quadrivalents. Saranac 6-4 ms was selected for its stable male sterility, and was as regular as other materials until Anaphase II, at which time irregularities increased probably related to the onset of male sterility.

WISFAL is a tetraploid recently developed from diploid Medicago falcata and had the most irregularities at each stage among the materials with normal fertility. Vernal and Arrow profiles are based on average values involving samples of 17 and 6 plants, respectively, and are very similar.

Rod bivalents are of particular interest because they indicate crossovers took place in only one chromosome arm. About 100 Metaphase I cells were analyzed in five Vernal genotypes (V-4, V-7, V-11, V-20, and V-41). Every plant had some PMCs with rod bivalents at Metaphase I, and the average was 0.25 with one to four rod bivalents and 0.75 with ring bivalent only. For all other plants used in this study, PMCs with ring and rod bivalents were noted at about the same average frequency as in Vernal (data not shown).

Self- and cross-fertility data (Table 3) were collected on plants that were studied cytologically plus additional ones of Vernal, Arrow and WISFAL. Thus, there were cytological and fertility data on 17 of 32 Vernal, 4 of 23 Arrow, and 3 of 6 WISFAL. Among the Vernal, Arrow, and WISFAL plants there was no apparent association between the degree of regularity of meiosis and self- and cross-fertility. One Arrow plant was functionally male-sterile, three others were self-sterile, and mean self-fertility was lower than in Vernal for which all plants produced some self-seed (Table 3). WISFAL was lower than the cultivars in self- and cross-fertility (Table 3) as was also reported in a previous study (Holland and Bingham, 1994).

Alfalfa chromosomes are described as short and subtelocentric with 2.26 average arm ratio (Gilles, 1970; Gillies and Lesins, 1971). According to Stanford, et al. (1972), these features could limit chiasma formation, especially in the shorter chromosome arm. This would explain the presence of 1-4 rod bivalents in about 25% of the PMCs at Met I. The disassociation of some quadrivalents between pachytene and Metaphase I as reported by Armstrong 1971, and in this study, is evidence of an insufficient crossing over to maintain quadrivalents. Importantly, the disassociation of quadrivalents produced bivalents as well as univalents. In fact, this may be the reason alfalfa is a bivalent-forming autotetraploid (Stanford et al. 1972; McCoy and Bingham 1988).

Atwood and Grun (1951), Grun (1951), and Armstrong (1954, 1971) reported multivalent frequencies ranging from 0.21 to 0.89 per cell in plants from cultivars. The same general range was found in this study with an average of 0.28 at diakinesis and 0.19 at Metaphase I. Interestingly, this was also the frequency of multivalents observed in colchicine-doubled cultivated diploids (Obajimi and Bingham, 1972).

Univalents, in 28% of the cells, on the average appeared to be the principal cause of irregularities throughout the stages of microsporogenesis in alfalfa in this and in previous studies by Julen (1944), Atwood and Grun (1951), Grun (1951), Armstrong (1971) and Smith and Murphy (1986). Results from our study on 11 Vernal genotypes indicated there was a high, but not significant correlation between the frequency of PMCs with lagging univalents at Anaphase I, mispositioned chromosomes at Metaphase II, and lagging chromosomes at Anaphase II. Furthermore, there was a highly significant correlation between the frequency of PMCs with lagging chromosomes at Anaphase II and the tetrads containing microspores with micronuclei. One univalent is always produced when there is a multivalent with a chain of three chromosomes, and two univalents are produced when there is insufficient crossovers to hold a bivalent together.

A comparison of the degree of regularity and the meiotic patterns exhibited throughout microsporogenesis in Vernal, Arrow, WISFAL, and the individual clones Blazer XL-P and Saranac male sterile 6-4ms indicates that they are basically similar. However, clone Blazer XL-P which was selected as a very fertile clone was consistently the most regular of all entries in the study. The derived tetraploid nature of WISFAL probably accounted for the high frequency of lagging univalents, and, consequently, the mispositioned chromosomes at Metaphase II. Toward the end of microsporogenesis, the PMCs of Saranac male sterile clone 6-4ms demonstrated the highest frequency of abnormalities, likely reflecting the expression of cytoplasmic male-sterility.

Regularity of microsporogenesis in alfalfa appears to be a matter of degree, with even the most regular plant containing occasional univalents and microspores with micronuclei. As discussed by Smith and Murphy (1986), a threshhold level of irregularity apparently must be exceeded before fertility is significantly affected. In the Arrow and Vernal cultivar populations, plants with a similar background of meiotic irregularities ranged from nearly self-sterile to quite self-fertile. This indicated that genetic factors affecting other fertility parameters had a greater impact on fertility than the range of meiotic regularity observed in this study.

The high levels of fertility in the selected clones, none of which was perfectly regular, also suggests that genetic factors not associated with meiotic regularity are important. This study and the one by Smith and Murphy (1986) suggest that when selection is practiced for improved fertility, meiotic regularity is improved to a much smaller degree.

Armstrong, J.M. 1953. Cytological studies in alfalfa polyploids. Can. J. of Botany. 32:531-542.

Table 2. Frequencies of meiotic regularities and irregularities during second division of meiosis in PMCs from 17 Vernal plants, four Arrow plants, three WISFAL plants, one Blazer XL clone, and male sterile clone Saranac 6-4ms. Measure based on at least 50 cells per stage per plant. Standard deviations range from 0.10-0.21; cv(%) range from 20-80%.

Table 3. Number of plants sampled (n), means, and ranges for self-fertility, cross-fertility, and the ratio of self-fertility to cross-fertility of cultivars and clones.
 
 

Fig. 1. Regular meiosis in PMCs from cultivar Vernal: a. Late prophase showing 16 bivalents (Vernal-18); b. Metaphase I (Vernal-41); c. Late anaphase I (Vernal-46); d. Metaphase II (Vernal-43); e. Anaphase II (Vernal-27); f. Association of four microspores (Vernal-43); (100/1.32).

Fig. 2. First meiotic division in PMCs from cultivar Vernal featuring various patterns of chromosome pairing: a. Pachytene chromosomes with a less than perfect pairing (Vernal-20); b. Pachytene chromosomes involved in an apparent translocation (Vernal-20); c. Pachytene chromosomes engaged in a double nonhomologous arm exchange (Vernal-7); d. Late prophase chromosomes, bivalent and quadrivalent associations (Vernal-11); e. Metaphase I with multivalent and univalent (Vernal-7); f. Early metaphase I with two univalents (Vernal-43); g. Late anaphase I with multiple laggards undergoing delayed division (Vernal-20); h. Telophase I with univalents remaining at the cell equator (Vernal-4); (100/1.32).

Fig. 3. Irregularities in PMCs from cultivar Vernal exhibited during second meiotic division: a. Metaphase II with chromosomes positioned distantly off the metaphase plate (Vernal-43); b. Anaphase II with a chromosomal bridge (Vernal-20); c. Late anaphase II with numerous laggards (Vernal-11); d. Late telophase II with laggards associated in a separate chromatin body (Vernal-43); e. Microspores of various sizes and chromatin contents (Vernal-7); f. Pollen grains of heterogeneous size and stainability (Vernal-18); (100/1.32).

Fig. 4. Frequency of PMCs with regular stages of meiosis from metaphase I to tetrads. Cultivars Vernal and Arrow, WISFAL, clone Blaser XL-P, and male sterile clone Saranac 6-4ms.

Previous Page