The reproductive versatility of eastern gamagrass.
Cutler and Anderson (1941) provided the earliest, and perhaps the most concise, survey of the genus, including taxonomic, distribution, and species classifications. Tripsacum compromises 15 species in two sections (Cutler and Anderson, 1941; deWet et al., 1981, 1982, 1983a) and is native only to the Western hemisphere. Tripsacum is widely distributed from Connecticut to Kansas and south to Texas in North America, south through Mexico and Central Mexico, and deep into Paraguay and Brazil. The genus consists of diploid (2n = 2x = 36), triploid (2n = 3x = 54), tetraploid (2n = 4x = 72), pentaploid (2n = 5x = 90), and hexaploid (2n = 6x = 108) cytotypes.
Tripsacum is distantly related to maize (Z. mays) and early cytological and genetic studies on Tripsacum reflect the interest in the relationship (Mangelsdorf and Reeves, 1939; Hernandez and Randolph, 1950; Tantravahi, 1968). Historically, cytological and morphological observations indicated Tripsacum and maize were products of collateral evolution with both genera having diverged from a common ancestor (Anderson 1944; Galinat et al., 1964; Kindiger, 1993; Kindiger et al., 1996b). This theory suggested that diploid Tripsacum, with its 18 pairs of chromosomes (2n = 2x = 36), is actually a polyploid species with a basic number of nine pairs and may represent an ancient amphidiploid (n = [9.sup.A] and [9.sup.B]).
Farquharson (1954, 1955) characterized the reproductive processes in Tripsacum by identifying both sexual and facultative apomictic cytotypes. Subsequent embryological analyses confirmed diploids to be exclusively sexual while the polyploid cytotypes reproduced as facultative apomicts (Brown and Emery, 1958; Burson et al., 1990; Sherman et al., 1991; Leblanc et al., 1995). The predominant form of apomictic reproduction in Tripsacum is characterized as being diplosporous pseudogamy of the Antennaria type [Anennaria alpina (L.) Garrtner] (Burson et al, 1990, Leblanc et al., 1995); however, an infrequent Taraxacum Wigg. form of apomixis, with its accompanying first division restitution (FDR) nucleus, occurs at a low frequency (Leblanc et al, 1995; Kindiger and Dewald, 1996).
Embryological investigations indicated that polyploid cytotypes reproduce as facultative apomicts (Leblanc et al., 1995); however, progeny tests of polyploid cytotypes indicated that reproduction was essentially obligate. Backcrosses to polyploid cytotypes have not identified hybrids originating from completely sexual reproductive events (Kindiger and Dewald, 1996).
Until recently, little attention has been directed at the economic and agricultural potential of Tripsacum. Diverse uses ranging from erosion control, land reclamation, and re-establishment of native prairies have provided impetus for renewed interest in the genus.
The primary task of a plant breeder is the development and release of superior genotypes. This is accomplished by understanding a species' reproductive processes and manipulating the plant's genetic architecture. This study focused on identifying the various reproductive characteristics within the various cytotypes and how these novel breeding systems can be used to develop eastern gamagrass as a superior forage grass.
MATERIALS AND METHODS
The Southern Plains Range Research Station maintains an extensive Tripsacum germplasm collection consisting of temperate and tropical germplasm. The materials evaluated in these studies were collected at various sites, most within the continental USA, and are maintained at the Southern Plains Range Research Stn., Woodward, OK. All nursery plantings are established and are maintained in a 20 x 20 grid design without replication. Principal soils at the Woodward location consist primarily of Enterprise fine sandy loam (coarse silts, mixed thermic Typic Ustochrepts). The area is characterized by wide fluctuations in temperature between day and night and summer and winter. Data obtained from 1914 to 1988 at Woodward showed temperatures ranged from a record low of -30 [degrees] C to a record high of 43 [degrees] C. During the same 74-yr interval, the average low and high temperature were -10 [degrees] C and 21 [degrees] C, respectively (Bradford, 1996). Average annual precipitation is 595 mm, but seasonal droughts are common (Fitzpatrick and Boatright, 1938; Nance and Gray, 1978). The data herein represent a culmination of data acquisition and research progress from 1986 to 1996. For simplicity, data obtained from specific crosses are provided to highlight the various reproductive features, forage potential, and breeding strategies that can be utilized in working with Tripsacum (Tables 1 and 2).
Table 1. Woodward identification numbers, ploidy, and origin of plant materials listed in Table 2 and described in this review. Identification number Ploidy Origin(*) WW1000 2x Woodward, OK WW1008 4x Baird, TX WW1082 4x Homestead, FL WW1379 2x Texarkana, TX WW1654 2x Ottawa Co., KS WW1684 2x F1 Hybrid WW1582 2x Ottawa Co., KS WW1722 4x Bowie, TX WW1724 4x Upshur, TX WW1729 4x Navarro, TX WW1766 3x F1 Hybrid (2x x 4x) WW2004 4x Tecalitlan, Jalisco, Mexico WW2031 4x [B.sub.m] derived hybrid (3x x 2x) WW2045 2x F1 Hybrid (2x x 2x) WW2075 2x F1 Hybrid (2x x 2x) WW2128 2x F1 Hybrid WW2167 4x San Antonio, TX WW2190 4x [B.sub.m] derived hybrid (3x x 2x) WW2436 6x [B.sub.m] derived hybrid (WW2167 x Tp112) M-34455 3x (54Tr) + 10(Zea) = 64 USDA Tripsacum Collection (Honduras) M-34445 3x (54Tr) + 10(Zea) = 64 USDA Tripsacum Collection (Columbia) * All accessions listed are T. dactyloides with the exception of WW2004, which is T. maizar, and M-34455 and M-34445, which are T. andersonii.
Mitotic root-tip chromosome counts were used to identify the number of chromosomes in the parental materials and their progeny generated from various crossing schemes. Germination was accomplished by pre-treating the seed following the methods of Ahring and Frank (1968) and Kindiger (1994). Mitotic chromosome counts were made for each Tripsacum accession utilizing the procedure of Sallee (1983). Root tips were taken from actively growing plants obtained from the field, greenhouse, or germination chamber. Microsporocytes and pollen samples were taken directly from the nursery or greenhouse and fixed and stained according to the method of Morrison (1953). Mature pollen was examined through the microscope at 45x or a hand lens (40x magnification) to identify levels of pollen fertility. In some cases, pollen fertility studies were complemented with [KI.sub.2] staining to determine starch content. Pollen grains with drastically deficient or unbalanced chromosome constitutions would be lacking or completely devoid of starch. Aberrant or aborted pollen are typically smaller than the "normal" pollen grains having a balanced chromosome number. The proportion of visible starch in each pollen grain and its relative size strongly correlate with varying chromosome constitutions (1n, 2n, 3n, 4n, etc.) and with various levels of aneuploidy (1n + 1, 2n + 1, 2n + 3, etc.) (Burnham, 1982).
Pollination or Hybridization Studies
Hand crosses were made or attempted in all possible combinations between accessions having different ploidy. The crossing scheme utilized several diploid (2x = 2n = 36), triploid (2n = 3x = 54), tetraploid (2n = 4x = 72), pentaploid (2n = 5x = 90) and hexaploid (2n = 6x = 108) germplasms. All hybrids were generated between 1986 and 1996 either from the field (open pollinations) or the greenhouse (hand pollinations) (Tables 1 and 2). In specific instances, field examinations of the parents and their progeny were made during the summer or in the winter greenhouse. Pollination methods used to generate the various hybrid combinations are described elsewhere (Dewald and Kindiger 1994). Emasculations were performed [TABULAR DATA FOR TABLE 2 OMITTED] to reduce the opportunity for contamination by removing the terminal male section of the infloresence. Pollen contamination was prevented following emasculation by covering the lower female portion of the infloresence with a Lawson #217 pollination bag (Lawson Bags, Northfield, IL).
Female fertility readings were determined by harvesting seed heads from mature individuals. Caryopses were extracted from spikelets by hand. Percent seed set, estimating female fertility, was determined by dividing the number of spikelets containing a caryopsis by the total number of spikelets harvested and then multiplying by 100. Low female fertility readings are not considered a consequence of pollen availability since ample pollen was constantly available throughout the flowering seasons. Controlled greenhouse hand pollinations were sometimes used to augment the summer data obtained from open-pollinated crosses.
PCR-RAPD analyses were performed on DNA samples obtained from lyophilized leaf tissue from 3 grams of fresh leaves harvested from field grown or greenhouse plants. DNA extractions were made following the method of Saghai-Maroof et al. (1984). Individuals were analyzed by RAPD techniques utilizing the protocol developed by Williams et al. (1990). Reactions were conducted on approximately 10 ng plant DNA. Ampli-Taq DNA polymerase and 10x reaction buffer were purchased from Perkin-Elmer (Branchburg, NJ). The RAPD study utilized 220 decamer oligonucleotides available as kits A through K purchased from Operon Technologies (Alameda, CA). Polymerase chain reaction (PCR) amplifications were performed in a M.J. Research (Watertown, MA), PTC-100 thermocycler with a "hot-lid" accessory. The PCR amplification reaction was programmed for 1 cycle of 94 [degrees] C for 1.5 min followed by 34 cycles of 94 [degrees] C for 1 min; 45 [degrees] C for 2 min and 72 [degrees] C for 2 min and then 1 final cycle of 37 [degrees] C for 2 min and 72 [degrees] C for 5 min. Following amplification, RAPD products were held at 4 [degrees] C until they were separated on a 15 g [L.sup.-1] agarose gel in 1 x TBE (Tris-Borate-EDTA) buffer at 120 V for 2 hr and visualized by ethidium bromide staining. Faint bands were disregarded and only bright reproducible bands were considered during the evaluations. Replicate runs were conducted to verify off-type or unexpected bands. RAPD analyses were performed to identify the various reproductive behaviors at each ploidy level, identify the occurrence of apomictic or sexual hybrids, and provide a recognizable DNA fingerprint (genetic profile) of each individual assayed.
RESULTS AND DISCUSSION
Diploids (2n = 2x = 36)
As previously reported, diploids reproduce entirely by classic sexual reproduction (deWet et al., 1983b) (Table 1). As is typical of most sexual reproductive systems, megasporogenesis results in the formation of reduced eggs (ln = 1x = 18), which remain receptive to fertilization by sperm nuclei (Leblanc et al., 1995). Similarily, microsporogenesis in the diploids was also normal, usually resulting in the formation of a tri-nucleate pollen grain possessing two unreduced (1x) sperm nuclei and a vegetative nucleus. Crosses between diploids, in conjunction with crosses using diploids as the maternal or paternal parent, have been successful in generating new sources of diploid, triploid, tetraploid, pentaploid, and hexaploid progeny (Table 1). Though commonly observed in other species, a balanced endosperm chromosome number does not restrict normal seed development in inter-ploidy crosses (Johnson et al., 1980: Lin, 1984).
Although diploid Tripsacum maintains a sexual reproductive process, this does not exclude them from contributing and exchanging germplasm with their apomictic counterparts (i.e., triploids, tetraploids, and hexaploids). Farquharson (1954) was among the first to identify the existence of mixed diploid, triploid, and tetraploid colonies of T. dactyloides, likely resulting from natural hybridizations occurring between sexual diploids and apomictic tetraploids. In addition, controlled greenhouse crosses utilizing diploids as the seed parent and tetraploid cytotypes as the pollen parent were found to readily generate triploid hybrids at a high frequency (Dewald et al., 1992: Dewald and Kindiger, 1994).
Triploids (2n = 3x = 54)
Triploids were easily generated by intercrossing diploids (2n = 2x = 36) with tetraploids (2n = 4x = 72) (Table 1). Triploid hybrids can be completely or partially male sterile and completely or partially female sterile depending on their possession of genes regulating apomictic development (Dewald et al., 1992; Dewald and Kindiger, 1994). All triploids exhibiting maternal fertility are apomictic in their mode of reproduction. All triploids exhibiting complete maternal sterility are expected to be non-apomictic or sexual in their reproductive process. These results were obtained by a combination of crossing studies, chromosome counts, and PCR-RAPD evaluations (Table 1). From numerous diploid x tetraploid crosses, triploid hybrids segregated for varying levels of maternal fertility (Table 3). All seed generated by a triploid was the result of apomictic development or the result of a 2n + n mating. The results of nine 2x x 4x crosses, utilizing different diploid and tetraploid germplasm sources, are provided in Table 3. A high level of uniformity in the segregation of maternal fertility is indicated among the segregations regardless of the maternal or paternal germplasm source. These data indicate a common inheritance-transmission of the gene(s) controlling seed set and a sexual or apomictic mode of reproduction (pooled [[Chi].sup.2] statistic is 4.68; [TABULAR DATA FOR TABLE 3 OMITTED] 0.05 [less than] P [less than] 0.10). Combined with the data generated from a study of apomixis inheritance in apomictic and sexual maize-Tripsacum hybrids (Kindiger et al., 1996c), it is suggested that apomixis in Tripsacum is likely controlled by two dominant, epistatic linked genes.
As discussed above, fertile triploids, at least within the T. dactyloides complex, were apomictic in their reproductive behavior. However, 29 partially fertile allotriploids, generated from a single (2n = 2x = 36 T. dactyloides) x (2n = 4x = 72 T. maizar Hernandez and Randolph) cross, exhibited sexual reproductive development. Backcrosses to each parent revealed that the T. dactyloides parent was sexual in its mode of reproduction, while the T. maizar parent produced only apomictic progeny. Root tip counts of progeny generated by the sexual fertile allotriploid hybrids revealed an array of aneuploid constitutions and plant phenotypes. The level of maternal fertility in these allotriploids ranged from I to 5.6% (Dewaid and Kindiger, 1995, unpublished data). A possible explanation for the observed fertility and generation of aneuploid progeny is assumed to be related to the genetic constitution of the allotriploid. The hybrid possesses two doses of T. maizar and one of T. dactyloides genomes. The established abilities of chromosomes from identical or similar genomes to pair and disjoin to the exclusion of a dissimilar genome, offers the most valid explanation for the low level of sexual fertility observed in this individual. This behavior has been readily observed in Agropyron elongatum L.-Triticum aestivum L. (Sears, 1972a,b) and maize-Tripsacum hybrids (Kindiger and Beckett, 1989). Enough dissimilarity between the T. maizar and T. dactyloides genomes could exist to allow the T. maizar chromosomes to form bivalents while the T. dactyloides chromosomes would act as univalents.
Pollen fertility of triploids, regardless of their maternal fertility or sterility ranged from complete to approximately 50% based on the stainability of pollen via [KI.sub.2] staining. Levels of pollen fertility ranged from a low of 3% to a high of 50%. The amount of viable pollen in the male fertile triploids was generally adequate for pollination (Sherman et al., 1991).
For the most part, female sterile or highly female sterile triploids have received only a cursory study. However, female fertile triploids have been well evaluated and provide a robust method for exchanging germplasm between sexual and apomictic cytotypes (Dewald and Kindiger, 1994). Cytological and molecular (RAPD) investigations of progeny generated from fertile apomictic triploids have indicated their ability to undergo sexual polyploidization resulting from 2n + n matings (Kindiger and Dewald, 1994). The resulting progeny generated from such matings are called [B.sub.III] derived hybrids (Bashaw et al., 1992). In these instances, an unreduced egg possessing 54 chromosomes can be fertilized by a reduced sperm nucleus carrying 18 (from a diploid pollen parent) or 36 chromosomes (from a tetraploid parent). Fertilization results in a [B.sub.III] derived hybrid possessing either a tetraploid (2n = 4x = 72) or pentaploid (2n = 5x = 90) constitution. Apomictic reproductive mechanisms are often retained in the [B.sub.III] derived tetraploids but are lost at the pentaploid level. Therefore, for breeding and genetic purposes, a backcross with a diploid pollen source is preferable to that from a triploid or tetraploid. The maternal fertility of a newly generated [B.sub.III], derived tetraploid can be either diminished or enhanced by the presence or absence of the apomictic alleles. Most [B.sub.III] derived tetraploids exhibit high levels of pollen fertility.
Because of the apomictic nature of synthetic and naturally occurring fertile tetraploids, fertile triploids can be used to provide an intermediate step for the creation of new and genetically unique tetraploid germplasm. The introgression of diploid germplasm into apomictic tetraploid germplasm provides for the enrichment of the genetic diversity at the tetraploid level.
Tetraploids (2n = 4x = 72)
Naturally occurring and "synthetic" (generated by controlled crosses in a breeding program) tetraploids reproduce exclusively by apomixis (Table 1). The form of apomixis exhibited in these materials is diplosporous pseudogamy of the Antennaria type (Brown and Emery, 1958: Burson et al., 1990; Sherman et al., 1991; Leblanc et al., 1995). In addition, the rare occurrence of a Taraxacum form of apomixis has been indicated (Leblanc, et al., 1995; Kindiger and Dewald, 1996). With the Antennaria form of diplospory, meiosis is completely omitted resulting in the development of unreduced eggs, genetically identical to the maternal parent (Asker and Jerling, 1992). Subsequently, genetic change is prohibitive. In contrast, the Taraxacum form of apomixis is characterized by the infrequent occurrence of a first division restitution (FDR) nucleus in which the only meiotic division represents a equational division and does not preclude the omission of chromosome pairing and recombination. In the instances where the unreduced egg is a product of a FDR, a low level of genetic change is possible (Peloquin, 1983; Hermsen, 1984).
Progeny tests, utilizing both cytological and RAPD methods to identify the type of progeny generated from tetraploid x diploid and tetraploid x tetraploid crosses, indicate the frequency of FDR events to be between 2 and 4% in T. dactyloides (Table 1, [ILLUSTRATION FOR FIGURE 1 OMITTED]) (Kindiger and Dewald, 1996).
Specifically, only three types of progeny are generated in crosses performed with apomictic tetraploids. These include tetraploid apomicts, [B.sub.III] derived hybrids (5x and 6x), and the occasional apomictic tetraploid offtype arising from an infrequent FDR event (Kindiger et al., 1996a; Kindiger and Dewald, 1996). Hybrids arising from a reduced egg fertilized by a sperm nuclei have not been observed. Embryological studies of several apomictic tetraploids of varying species, revealed the generation of sexual dyads and tetrads at a cumulative frequency of 2.1% (Leblanc, et al., 1995). This observation indicated a facultative component of apomictic reproduction in tetraploids. However, molecular analysis of progeny generated from such a cross between apomictic tetraploids, indicated an absence of viable sexual products or tetraploid hybrids (Kindiger et al., 1996a; Kindiger and Dewald, 1996). These data, suggested that the sexual reproduction of a 4x hybrid derived from a 4x x 4x cross may be rare and may indicate a more obligate nature to apomictic development when considering the generation or procurement of seed.
Crosses between diploids (2n = 2x = 36) and tetraploids (2n = 4x = 72) typically produce triploid hybrids (2n = 3x = 54). However, diploid hybrids are infrequently generated. Chromosome counts in the developing pollen grain of tetraploids indicated a wide array of chromosome numbers. Pollen with varying numbers of chromosomes (other than 1n = 1x = 36) are likely the result of 3:1 disjunction events from quadrivalent pairing during meiosis. In situations of low pollen competition, pollen with unbalanced chromosome numbers can compete against pollen with a balanced chromosome constitution and generate hyrids with aneuploid or unexpected chromosome numbers (Pfahler, 1975; Ottaviano and Mulcahy, 1986). Diploids obtained from a 2x x 4x cross are likely generated from these events.
Apomictic tetraploids also possess the sexual polyploidization mechanism observed in their apomictic triploid counterparts. [B.sub.III] hybrids resulting from 2n + n matings typically occur 5 to 8% of the time (Table 1) (Kindiger and Dewald, 1994, 1996). However, some accessions have been observed to generate [B.sub.III] hybrids as often as 50% of the time (Kindiger and Dewald, 1995, unpublished data).
Pentaploids (2n = 5x = 90)
Pentaploids are generated from 2n + n matings following a 4x x 2x crossing scheme. To date, pentaploids have been completely female sterile and somewhat male sterile (0-12%). Pentaploids are usually weak, diminutive plants that are unsuitable for breeding or genetic experiments. Their agronomically inferior nature and sterility are likely due to an inbalance of genomes and alleles. It is typical of individuals possessing aneuploid constitutions or unbalanced genomes to display diminutive, inferior phenotypes. Individuals exhibiting such phenotypes are usually accompanied by a reduction in plant vigor and increased sterility (Birchler and Newton, 1981). However, in rare instances, pentaploids can exhibit a extremely vigorous phenotype and be somewhat male fertile (Kindiger and Dewald, 1995, unpublished data). In mixed colonies of Tripsacum possessing both diploid and tetraploid cytotypes, pentaploids should naturally occur, but may generally lack the competitive edge for long-term survival. Because of their inferior phenotype, studies utilizing these individuals have been limited. No breeding value has yet been assessed for these materials and their development/existence is presently regarded as a genetic novelty.
Hexaploids (2n = 6x = 108)
Hexaploids are generated during a 4x x 4x cross where the unreduced egg is fertilized by a 2n sperm nuclei. As with pentaploids, hexaploids are also the result of 2n + n matings and are considered [B.sub.III], derived hybrids. Like their pentaploid counterparts, hexaploids can exhibit diminutive phenotypes [ILLUSTRATION FOR FIGURE 2 OMITTED]. However, robust and vigorous hexaploids were occasionally generated. Unlike pentaploids, hexaploids provide a valuable bridge for moving germplasm, previously confined within apomictic genomes, to sexual diploids through 2x x 6x crossing schemes (Table 2).
Hexaploids are often but not exclusively female sterile. The low frequency of hexaploids exhibiting female fertility are apomictic in their mode of reproduction (Table 2). In addition, regardless of their maternal sterility-fertility, hexaploids typically exhibit 50 to 75% pollen fertility. Pollen is plentiful and outcrosses are likely abundant in nature. During controlled pollinations between sexual diploids x hexaploid cytotypes, diploid hybrids are the usual result. Triploid, tetraploid and hybrids with aneuploid constitutions have also been obtained but at a much lower frequency (Table 2). Although the tetraploid parents utilized to generate the hexaploid hybrids were apomictic, all diploids generated from a 2x x 6x crossing scheme were sexual. This result could indicate that the gene(s) controlling apomictic expression were not transferred to the diploid level. This is highly unlikely. It is more likely that the the apomictic gene(s) do not express at the diploid level and a gene(s) dosage requirement, which is not present in diploids, is necessary for apomictic reproduction. This possibility is in agreement with some prevailing ideas relating to the inheritance and expression of apomixis (Harlan et al., 1964; Mogie, 1992). The major discrepancy between numbers of diploid, triploid, tetraploid and, aneuploid hybrids generated from a 2x x 6x cross is likely the result of pollen competition. Pollen possessing a more balanced chromosome constitution can maintain a selective advantage over those with a different constitution, resulting in a distortion of the anticipated segregation ratios (Pfahler, 1975; Ottaviano and Mulcahy, 1986). Similar to tetraploids, chromosome counts in the developing pollen grains of hexaploid individuals at the first mitotic metaphase indicate a wide range in chromosome number. Chromosome counts ranged from a low of 16 to the high 90s. This is because of the various chromosome pairing and disjunction configurations that can be obtained in a hexaploid situation. These include several possible combinations of univalents, bivalents, trivalents, quadrivalents, pentavalents, and hexavalents followed by adjacent I and II as well as alternate I and II segregation events. Any of these events occurring separately or collectively would produce balanced, unbalanced, aborted, and even aneuploid gametes that can be identified to some extent by observing chromosome numbers in the developing pollen grain at the first mitotic division following meiosis (Morrison, 1953; Kindiger and Beckett, 1985). Likely, during heavy pollinations, pollen carrying a haploid constitution fertilize the egg more often than pollen with other constitutions. Diploid hybrids generated from 2x x 6x crosses are highly variable and possess a wide diversity of alleles contributed from the apomictic cytotypes. The ability to generate hexaploid materials and to cross diploid x hexaploid germplasm, provides an excellent method for the transfer of tetraploid germplasm to the diploid level.
Individuals possessing aneuploid constitutions have limited value for breeding purposes but can provide useful materials for genetic studies. The first deliberate attempt to generate individuals with aneuploid constitutions involved the use of 2x x 3x crosses (Sherman et al., 1991). Though extensive karyotyping was not performed on the hybrids, numerous individuals with varying chromosome numbers were identified. Chromosome counts ranged from normal individuals with 2n = 36 and 2n = 3x = 54 chromosomes to aneuploids with chromosome numbers ranging from 37 (36 + 1) to 51 (36 + 15).
Crosses with tetraploids and hexaploids can also provide similar results. In a series of 2x x 6x crosses, the predominant hybrids generated were diploids (2n = 2x = 36). However, a low occurrence of primary trisomics were observed (Table 2) as well as a single tetraploid.
From crosses using a female fertile and sexual T. dactyloides-T. maizar allotriploid hybrid, progeny with aneuploid constitutions have also be generated (Kindiger and Dewaid, 1995, unpublished data). Properly utilized, genetic and cytogenetic evaluations with aneuploid stocks can provide a useful tool for mapping studies and research devoted to identifying dosage dependant traits (Birchler, 1991).
Methods for Horizontal Gene Flow
The generation of triploid and hexaploid genotypes offer innumerable possibilities for improving and widening the genetic diversity available in the genus. Triploids provide the means to introgress sexual germplasm into a tetraploid cyto-type while hexaploids provide for the transfer of tetraploid germplasm to sexual, diploid cytotypes [ILLUSTRATION FOR FIGURE 3 OMITTED]. The generation of such cytotypes is firmly established from greenhouse pollinations and is apparently operating in nature (Farquharson, 1954, 1955). In addition, a preliminary survey of Tripsacum germplasm maintained at the USDA facility in Miami, FL, has identified one naturally occurring hexaploid (T. peruvanium deWet and Timothy) (M-34502) originally obtained from San Martin, Peru. These two crossing schemes have likely produced a significant portion of the taxonomic diversity within the genus (Cutler and Anderson, 1941; deWet et al., 1981, 1982, 1983a).
Tripsacum andersonii Gray
Among all the Tripsacum species, T. andersonii is unique. It is a vigorous, perennial species native to Central and South America. Cytological and molecular analysis of T. andersonii has suggested that it is a natural Tripsacum-Zea hybrid carrying 54 Tripsacum chromosomes (2n = 3x = 54) and a haploid set (n = 10) of Zea chromosomes (Levings et al., 1976; deWet et al., 1983b; Talbert et al., 1990). Based on the reproductive behaviors exhibited in apomictic triploid and tetraploid Tripsacum, T. andersonii is a likely [B.sub.III] derived hybrid generated from a 2n + n mating event with a Zea (n = 10) complex. Tripsacum andersonii is completely male sterile but can set an occasional seed (deWet et al, 1978). Maternal sterility is estimated to be 99% or greater. Propagation is usually by rootstock and stolon cuttings (deWet et al., 1983b). Recent research has suggested the genomic composition of T. andersonii to be comprised of Tripsacum species from the section Fasiculata with its Zea component being from Zea luxurians Durieu and Ascherson (Talbert et al., 1990; Larson and Doebley, 1994).
In 1994, four seed obtained from uncontrolled crosses made with T. andersonii clones (M-34455 and M-34445) maintained at the USDA-ARS National Clonal Germplasm Repository, Miami, FL, were analyzed for their chromosome number and constitution. Among the three germinating seed, each had a chromosome number of 82. Karyotypes of the three individuals indicated that they possessed a tetraploid Tripsacum genomic constitution (4x = 72) and a haploid set of Zea (1x = 1n = 10). The result of this event was probably caused by a 2n + n mating with an unknown diploid Tripsacum pollinator. This behavior has been previously observed in other apomictic Tripsacum accessions (Kindiger and Dewald, 1994). In 1995, these plants were backcrossed with diploid T. dactyloides as the pollen parent. Seed obtained from this backcross generated individuals with the following chromosome numbers: two possessed about 82 chromosomes and were likely generated by apomixis; one carried about 100 chromosomes and was the likely result of a 2n + n mating event; and one carried 72 chromosomes (2n = 4x = 72) with an apparent total loss of its Zea group. These data indicated that the intergeneric hybrid T. andersonii, exhibited similar reproductive behaviors found in other polyploid and apomictic Tripsacum species. In addition, the possibility for initiating gene flow from T. andersonii to other Tripsacum species has been achieved in controlled greenhouse conditions. Cytological and molecular studies are continuing with these materials.
Crosses of T. dactyloides with all other Tripsacum species (Cutler and Anderson, 1941) are readily accomplished, regardless of ploidy or the taxonomic section a species has been assigned (Brink and deWet, 1983). All interspecific hybridizations generated at the Woodward location were performed under controlled greenhouse conditions. No interspecific crossing barriers have been observed and this includes T. andersonii. A comparison of the reproductive behaviors in T. dactyloides, T. floridanum Porter ex Vasey, T. maizar, and T. zopilotense Hernandez and Randolph and various interspecific hybrids (data not presented) have indicated identical reproductive characteristics. This includes T. andersonii, which also generated fertile [B.sub.III]. derived hybrids when backcrossed to Tripsacum. The generation of fertile interspecific hybrids with identical reproductive attributes suggested that some of the taxonomic relationships within the genus Tripsacum are likely artificial. Support of this assumption is provided by a recent survey of chloroplast genome variation among 24 accessions (15 species) of Tripsacum. That study revealed limited genetic differentiation among the accessions surveyed (Larson and Doebley, 1994) and failed to support the separation of section Tripsacum from section Fasciculatum. Within limits, interspecific crosses occur naturally, providing an opportunity to expand the genetic diversity across species. Natural crossing barriers such as topography, precipitation, taxonomic distribution, date of flowering, and elevation likely place limitations on the success of such hybridizations.
The Gynomonoecious Sex Form
The classic Tripsacum inflorescence exhibits male spikelets in the upper (tassel) section and female spikelets below on the same raceme ([ILLUSTRATION FOR FIGURE 4 OMITTED], top). For seed production purposes, this "normal" infloresence is inadequate for producing the quantity of seed required by the commercial seed industry. Dewald and Dayton (1985) reported the discovery of a mutant form of T. dactyloides that exhibited conversion of previously male flowers to seed setting structures. This characteristic provided for 10 to 25 more female florets and a subsequent increase in seed yield (Jackson et al., 1992). The mutant was given the assignment GSF-1, indicating its gynomonoecious phenotype (Dewald and Dayton, 1985). Subsequent genetic studies indicated that a single recessive allele conferred complete or nearly complete feminization to the male portions of the inflorescence (Dewald et al., 1987). The allele controlling the trait has been assigned the genetic nomenclature of gsf1 (Blakey et al., 1994; Dewald and Kindiger, 1994) and individuals possessing a homozygous gsf1 genotype exhibit paired pistillate spikelets and bisexual spikelets in the upper, previously male portion of the inflorescence ([ILLUSTRATION FOR FIGURE 4 OMITTED], bottom). By means of the methods addressed above, the trait has been successfully transferred from a sexual diploid and introduced into a fertile apomictic triploid (Dewald and Kindiger, 1994, 1996). Subsequent backcrosses will eventually result in the generation of an apomictic, tetraploid form via a 2n + n mating event (Kindiger and Dewald, 1994). The development of an apomictic, gynomonoecious tetraploid is expected to greatly improve the commercialization of this species as a pasture-hay crop and potential perennial grain species.
Intergenic Hybridization and Conservation of Genomes
Early studies devoted to elucidating the evolution of maize were perhaps the primary motivation that fostered modern cytogenetic and molecular studies of the Tripsacum genus (Mangelsdorf and Reeves, 1931, 1939; Cutler and Anderson, 1941; Anderson, 1944). Fertile maize x Tripsacum hybrids can be generated with relatively little effort through the use of appropriate maize and Tripsacum parents (Weatherwax and Randolph, 1955; Kindiger and Beckett, 1992). Tripsacum x maize hybrids are more difficult to generate and are relatively rare. Fertile Tripsacum x maize hybrids are extremely rare and only one such individual has been identified in our program (Dewald and Kindiger, 1996, unpublished data). The generation of such hybrids have identified common genomic relationships between the maize and Tripsacum genomes (Maguire, 1961, 1962; deWet et al., 1972: Galinat, 1973; Rao et al., 1974; Kindiger and Beckett, 1990) and recent molecular studies have clearly identified a conservation of linkage groups among the Gramineae (Whitkus et al., 1992; Ahn and Tanksley, 1993; Bennetzen and Freeling, 1993; Blakey, 1993). Comparative mapping strategies capitalizing on the potential conservation of linkage groups between maize and Tripsacum can be useful tools to identify regions within the Tripsacum genome possessing alleles or genes conferring favorable agronomic traits (Hake and Walbot, 1980; Blakey, 1994).
Utilizing a conserved group of maize and Tripsacum specific restriction fragment length polymorphisms (RFLP) markers, Blakey (1994) identified a possible association of the ts2 allele of maize (located to the short arm of chromsome 1) to the gsf1 allele of Tripsacum (assigned to linkage group I). Based on this possible conservation of linkage groups, a test of allelism was performed by generating hybrids between maize stocks carrying ts2 and a Tripsacum stock carrying gsf1. The intergenic cross was successful in generating maize-Tripsacum hybrids that expressed either a tassel seed or gynomonoecious phenotype. The successful test of allelic non-complementation suggested that the two genes expressed identical levels of genetic function and regulation (Kindiger et al., 1994). These results have also been confirmed utilizing a tetraploid, gynomonoecious T. dactyloides accession to generate a ts2/gsf1/gsf1 hybrid.
In another instance, by utilizing data available from the maize and Tripsacum genetic maps, maize and Tripsacum RFLP markers were used to identify the position of gene(s) regulating apomictic development in apomictic Tripsacum and apomictic maize-Tripsacum hybrids and regions of genome synteny (Kindiger et al., 1996b).
Research topics such as these can provide information that will be useful in cross-species mapping studies as well as increasing our understanding of phylogenetic relationships between the two taxa.
This study highlights the various reproductive mechanisms in Tripsacum. In addition, two major methods for facilitating horizontal gene flow between sexual and apomictic cytotypes of T. dactyloides were detailed. Utilizing the described crossing schemes, a breeder can effectively sample and manipulate the genetic diversity found within the genus. The ability to introgress germplasm across ploidies, regardless of reproductive method (i.e., sexual vs apomictic) provides an additional dimension to breeding superior cultivars of Tripsacum, which had previously not been realized. The data presented in this paper provide some clarification on the diversity found within the genus and also information on how the sexual and apomictic reproductive systems of Tripsacum can be manipulated for line improvement. In addition, the ability to utilize information generated from decades of maize genetic research offers tantilizing possibilities for reducing the time frame necessary to manipulate and identify the location of useful genes in Tripsacum.
Abbreviations: PCR-RAPD, polymerase chain reaction-random amplified polymorphic DNA; FDR, first division restitution; RFLP, restriction fragment length polymorphism.
Ahring, R.M., and H. Frank. 1968. Establishment of eastern gamagrass from seed and vegetative propagation. J. Range Manag. 21:27-30.
Ahn, S., and S.D. Tanksley. 1993. Comparative linkage maps of the rice and maize genomes. Proc. Natl. Acad. Sci. (USA) 90: 7980-7984.
Anderson, E. 1944. Cytological observations on Tripsacum dactyloides. Ann. Mo. Bot. Gard. 31:317-323.
Asker, S.E., and L. Jerling. 1992. Apomixis in plants. CRC Press, Boca Raton, FL.
Bashaw, E.C., M.A. Hussey, and K.W. Hignight. 1992. Hybridization (N + N and 2N + N) of facultative apomictic species in the Pennisetum agamic complex. Intl. J. Plant Sci. 153:466-470.
Bennetzen, J., and M. Freeling. 1993. Grasses as a single genetic system: genome composition, collinearity and compatibility. Trends Genetics 9:259-292.
Birchler, J.A., and K.J. Newton. 1981. Modulation of protein levels in chromosomal dosage series of maize: the biochemical basis of aneuploid syndromes. Genetics 99:247-266.
Birchler, J.A. 1991. Chromosome manipulation in maize. p. 531-559. In P.K. Gupta and T. Tsuchiya (ed.) Chromosome engineering in plants: Genetics, breeding and evolution, Part A. Elsevier, New York.
Blakey, C.A. 1993. A molecular map in Tripsacum dactyloides, eastern gamagrass. Ph.D. diss., Univ. Missouri-Columbia (Publ. No. 9412465).
Blakey, C.A., C.L. Dewald, and E.H. Coe. 1994. Co-segregation of the gynomonoecious sex form1 gene (gsf1) of Tripsacum dactyloides (Poaceae) with molecular markers. Genome 37:809-812.
Bradford, J.A. 1996. Climate history of the Southern Plains Range Research Station, Woodward, OK 1914-1988. USDA-ARS, S.P.R.R.S., Woodward, OK.
Brink, D.E., and J.M.J. deWet. 1983. Supraspecific groups in Tripsacum (Gramineae). Sys. Bot. 8:243-249.
Brown, W.V., and W.H.P. Emery. 1958. Apomixis in the Gramineae: Panicoideae. Am. J. Bot. 45:253-263.
Burnham, C.R. 1982. Details of the smear technique for studying chromosomes in maize. p. 107-118. In W. Sheridan (ed.) Maize for biological research. Plant Molecular Biology Assoc., Charlottesville, VA.
Burson, B.L., P.W. Voigt, R.A. Sherman, and C.L. Dewald. 1990. Apomixis and sexuality in eastern gamagrass. Crop Sci. 30:86-89.
Cutler, H.C., and E.G. Anderson. 1941. A preliminary survey of the genus Tripsacum. Ann. Mo. Bot. Gard., 28:249-269.
Dewald, C.L., and R.S. Dayton. 1985. Registration of gynomonoecious germplasm (GSF-1 and GSF-II) of eastern gamagrass. Crop Sci. 25:715.
Dewald, C.L., B.L. Burson, J.M.J. deWet, and J.R. Harlan. 1987. Morphology, inheritance and evolutionary significance of sex reversal in Tripsacum dactyloides (Poaceae). Am. J. Bot. 74:1055-1059.
Dewald, C.L., C.M. Taliaferro, and P.C. Dunfield. 1992. Registration of four fertile triploid germplasm lines of eastern gamagrass. Crop Sci. 32:504.
Dewald, C.L., and B. Kindiger. 1994. Genetic transfer of gynomonoecy from diploid to triploid eastern gamagrass. Crop Sci. 34: 1259-1262.
Dewald, C.L., and B. Kindiger. 1996. Registration of fertile gynomonoecious germplasm. Crop Sci. 36:219.
deWet, J.M.J., L.M. Engle, C.A. Grant, and S.T. Tanaka. 1972. Cytology of maize-Tripsacum introgression. Am. J. Bot. 59:1026-1029.
deWet, J.M.J., J.R. Harlan, and A.V. Randrianasolo. 1978. Morphology of teosintoid and Tripsacoid maize (Zea mays L.). Am. J. Bot. 65:741-747.
deWet, J.M.J., D.H. Timothy, K.W. Hilu, and G.B. Fletcher. 1981. Systematics of South American Tripsacum (Gramineae). Am. J. Bot. 68:269-276.
deWet, J.M.J., J.R. Harlan, and D.E. Brink. 1982. Systematics of Tripsacum dactyloides (Gramineae). Am. J. Bot. 69:1251-1257.
deWet, J.M.J., D.E. Brink, and C.E. Cohen. 1983a. Systematics of Tripsacum section Fasciculata (Gramineae). Am. J. Bot. 70: 1139-1146.
deWet. J.M.J., G.B. Fletcher, K.W. Hilu, and J.R. Harlan. 1983b. Origin of Tripsacum andersonii (Gramineae). Am. J. Bot. 70:706-711.
Farquharson, L.I. 1954. Natural selection of tetraploids in a mixed colony of Tripsacum dactyloides. Proc. Ind. Acad. Sci. 63:80-82.
Farquharson, L.I. 1955. Apomixis and polyembryony in Tripsacum dactyloides. Am. J. Bot. 42:737-743.
Fitzpatrick, E.G., and W.C. Boatright. 1938. Soil Survey of Woodward County, Oklahoma. United States Department of Agriculture-Bureau of Chemistry and Soils, Series 1932, No. 28.
Galinat, W.C., R.S. Chaganti, and F.D. Hager. 1964. Tripsacum as a possible amphidiploid of wild maize and manisuris. Bot. Mus. Leafl. Harvard Univ. 20:289-316.
Galinat, W.C. 1973. Intergenomic mapping of maize, teosinte and Tripsacum. Evolution 27:644-655.
Hake, S., and V. Walbot, V. 1980. The genome of Zea mays, its organization and homology to related grasses. Chromosoma 79:251-270.
Harlan, J.R., M.H. Brooks, D.S. Borgaonkar, and J.M.J. deWet. 1964. Nature and inheritance of apomixis in Bothriochloa and Dicanthium. Bot. Gaz. (Chicago) 125:41-46.
Hermsen, J.G. Th. 1984. Mechanisms and genetic implications of 2n gamete formation. Iowa State J. Res. 58:421-434.
Hernandez, X.E., and L.F. Randolph. 1950. Descripcion de los Tripsacum zopilolense, spp. nov. Folleto Tecnico No. 4 de la Oficina de Estudios Especiales, SAG, Mexico, D.F.
Jackson, L.L., C.L. Dewald, and C.C. Bohlen. 1992. A macromutation in Tripsacum dacotyloides (Poaceae): Consequences for seed size, germination and seedling establishment. Am. J. Bot 79:1031-1038.
Johnson, A.A., T.P.M. Den Nijs, S.J. Peloquin, and R.E. Hanneman, Jr. 1980. The significance of genetic balance to endosperm development in interspecific crosses. Theor. Appl. Genet. 57:5-9.
Kindiger, B., and J.B. Beckett. 1985. A hematoxylin staining procedure for maize pollen grain chromosomes. Stain Tech. 60:265-269.
Kindiger, B., and J.B. Beckett. 1989. Cytological evidence supporting a procedure for directing and enhancing pairing between maize and Tripsacum. Genome 33:495-500.
Kindiger, B., and J.B. Beckett. 1990. Cytological evidence supporting a procedure for directing and enhancing pairing between maize and Tripsacum. Genome 33:495-500.
Kindiger, B., and J.B. Beckett. 1992. Popcorn germplasm as a parental source for maize x Tripsacum dactyloides hybridization. Maydica 37:245-249.
Kindiger, B. 1993. Aberrant microspore development in hybrids of maize x Tripsacum dactyloides. Genome 36:987-997.
Kindiger, B. 1994. A method to enhance germination of eastern gamagrass. Maydica 39:53-56.
Kindiger, B., and C.L. Dewald. 1994. Genome accumulation in eastern gamagrass, Tripsacum dactyloides (L.) L. (Poaceae). Genetica 92:197-201.
Kindiger, B., and C.L. Dewald. 1996. A system for genetic change in apomictic eastern gamagrass. Crop Sci. 36:250-255.
Kindiger, B., C.A. Blakey, and C. Dewald. 1994. Sex reversal in maize x Tripsacum hybrids: Allelic non-complementation of ts2 and gsf1. Maydica 40:187-190.
Kindiger, B., V. Sokolov, and C. Dewald. 1996a. A comparison of apomictic reproduction in eastern gamagrass (Tripsacum dactyloides (L.) L.) and maize-Tripsacum hybrids. Genetica 97: 103-110.
Kindiger, B., D. Bai, and V. Sokolov. 1996b. Assignment of gene(s) conferring apomixis in Tripsacum to a chromosome arm: Cytological and molecular evidence. Genome 39:1133-1141.
Larson, S.R., and J.R. Doebley. 1994. Restriction site variation in the chloroplast genome of Tripsacum (Poaceae): Phylogeny and rates of sequence evolution. Syst. Bot. 19:21-34.
Leblanc, O., M.D. Peel, J.G. Carman, and Y. Savidan. 1995. Megasporogenesis and megagametogenesis in several Tripsacum species (Poaceae). Am. J. Bot. 82:57-63.
Levings, C.S., III, D.H. Timothy and W.W.L. Hu. 1976. Cytological characteristics and nuclear buoyant densities of corn, teosinte, tripsacum and corn-tripsacum hybrids. Crop Sci. 16:63-66.
Lin, B.-Y. 1984. Ploidy barrier to endosperm development in maize. Genetics 107:103-115.
Maguire, M. 1961. Divergence in Tripsacum and Zea chromosomes. Evolution 15:394-400.
Maguire, M. 1962. Chromatid interchange in allodiploid maize-Tripsacum hybrids. Can. J. Genet. Cytol. 5:414-420.
Mangelsdorf, P.C., and R.G. Reeves. 1931. Hybridization of maize, Tripsacum and Euchlaena. J. Hered. 22:329-343.
Mangelsdorf, P.C., and R.G. Reeves. 1939. The origin of Indian corn and its relatives. Texas Agric. Exp. Stn. Bull. No. 574.
Martin, W.F. 1963. Pasture and range plants, 7th Printing. Phillips Petroleum Company, Bartelsville, OK.
Mogie, M. 1992. The genetic control of apomixis. p. 139-196. In M. Mogie (ed.) The evolution of asexual reproduction in plants. Chapman & Hall, London.
Morrison, J.W. 1953. A new technique for pollen grain study in the Gramineae. Can. J. Agric. Sci. 33:399-401.
Nance, E., and F. Gray. 1978. Modern detailed survey of the Southern Great Plains Field Station, Woodward Research Report, USDA-SCS Bulletin No. P-774.
Ottaviano, E., and D.L. Mulcahy. 1986. p. 101-120. Gametophytic selection as a factor of crop plant evolution. In Barigozzi, C. (Ed.) The Origin and Domestication of Cultivated Plants. Elsevier, Amsterdam.
Peloquin, S.J. 1983. Genetic engineering with meiotic mutants. p. 311316. In D.L. Mulcahy and E. Ottaviano (ed.) Pollen biology and implications for plant breeding. Elsevier, New York.
Pfahler, P.L. 1975. Factors affecting male transmission in maize (Zea mays L). p.115-124. In D.L. Mulcahy (ed.) Gamete competition in plants and animals. North Holland, Amsterdam.
Polk, D.B., and W.L. Adcock. 1964. Eastern gamagrass. Cattleman 50:82-84.
Rao, B.G.S., and W.C. Galinat. 1974. The evolution of the American Maydeae. I. The characteristics of two Tripsacum chromosomes (Tr7 and Tr13) that are partial homeologs to maize chromosome 4. J. Hered. 65:335-340.
Saghai-Maroof, M.A., K.M. Soliman, R.A. Jorgensen, and R.W. Allard. 1984. Ribosomal DNA spacer-length polymorphisms in barley: Mendelian inheritance, chromosomal location and population dynamics. Proc. Natl. Acad. Sci. (USA) 81:8014-8018.
Sallee, P.J., and G. Kimber. 1983. A technique for the preparation of somatic chromosomes of rice. Cereal Research Comm. 11:53-55.
Sears. E.R. 1972a. Agropyron-wheat transfers through induced homoeologous pairing. Can. J. Genet. Cytol. 14:736.
Sears, E.R. 1972b. Chromosome engineering in wheat. Stadler Genetics Symp. 4:23-38.
Sherman, R.A., P.W. Voigt, B.L. Burson, and C.L. Dewald. 1991. Apomixis in diploid x triploid Tripsacum dactyloides hybrids. Genome 34:528-532.
Stubbendieck, J., S.L. Hatch, and C.H. Butterfield. 1993. North American range plants. 4th ed. Univ. Nebraska Press, Lincoln, NE.
Talbert, L.E., J.F. Doebley, S. Larson, and V.L. Chandler. 1990. Tripsaturn andersonii is a natural hybrid involving Zea and Tripsacum: Molecular evidence. Am. J. Bot. 77:722-726.
Tantravahi, R.V. 1968. Cytology and crossability relationships of Tripsacum. Ph.D. Thesis. Bussey Inst., Harvard University Publication, Cambridge, MA.
Tucker, L. 1835. The Genesee Farmer and Gardner's Journal. Vol. V.
Weatherwax, P., and L.F. Randolph. 1955. Cytogenetic aspects of the origin and evolutionary history of corn. p. 16-61. In G.F. Sprague (ed.) Corn and corn improvement. Academic Press, New York.
Whitkus, R., J. Doebley, and M. Lee. 1992. Comparative genome mapping of sorghum and maize. Genetics 132:1119-1130.
Williams, J.G.K., A.R. Kubelik, K.J. Livbak, J.A. Rafalski, and S.V. Tingey. 1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 18:6531-6535.
|Printer friendly Cite/link Email Feedback|
|Author:||Kingiger, B.; Dewald, C.L.|
|Date:||Jul 1, 1997|
|Previous Article:||Apomixis and sexuality in some Paspalum species.|
|Next Article:||Forage quality variation in the U.S. alfalfa core collection.|