Theor Appl Genet (1986) 71 : 772-783

9 Springer-Verlag 1986

Interspecific hybridization of perennial Medicagospecies using ovule-embryo culture T. J. McCoy and L. Y. Smith USDA/ARS, College of Agriculture, University of Nevada, Reno, NV 89557, USA Received September 12, 1985; Accepted October 14, 1985 Communicated by R. L. Kahler

Summary. New interspecific hybrids between alfalfa (Medicago sativa L.) and several perennial Medicago species were obtained by embryo rescue techniques. The methodology, designated ovule-embryo culture, involved preculturing the fertilized ovule (10 to 20 days post-pollination) for a period of six to 12 days followed by excision and direct culture of the embryo. Placement of the hybrid embryo directly onto culture medium without the interim ovule culture was unsuccessful. Ovule culture to germination without removing the embryo also was unsuccessful. Ovule-embryo culture was essential for recovering interspecific hybrids between diploid alfalfa ( 2 n = 2 x = 16) and the following diploid (2n = 2 x = 16) species: M. hybrida Traut., M. marina L., M. papillosa Boiss., M. rhodopea Velen. and M. rupestris M.B. In addition, trispecies hybrids between M. sativa X M. dzhawakhetica Bordz. F~ hybrids ( 2 n = 3 x = 2 4 ) and either M. cancellata M.B. (2n = 6x= 48) or M. saxatilis M.B. ( 2 n = 6 x = 4 8 ) were obtained from ovuleembryo culture. Media manipulations using M. sativa X M. rupestris F~ and first backcross generation embryos demonstrated the optimum concentration of 12.5 mM NH +, for successful embryo rescue; ammonium salt formulation (whether chloride, nitrate or sulfate) was not critical. From a few thousand crosses, hybrids between M. sativa and either M. rhodopea or M. rupestris were recovered relatively efficiently with 157 and 66 hybrids, respectively. However, only 13 hybrids between M. sativa and M. papillosa were obtained from more than 2,000 crosses, and just two hybrids each have been recovered from the combinations M. sativax M. hybrida and M. sativa x M. marina from 2,000 to 3,000 crosses. The predominant chromosome number between diploid alfalfa and the other diploid perennial species was 2 n - - 2 x = 16. Morphology of the hybrids was generally intermediate. Electrophoretic analysis of the Fa

hybrids and parental clones on uniform or gradient polyacrylamide gels demonstrated that peroxidase phenotypes could be used to confirm hybridity. For all interspecific combinations there was at least one peroxidase isozyme unique to the wild species that was present in the F1 interspecific hybrid.

Key words: Ovule culture - Embryo culture - Embryo rescue - Alfalfa (Medicago sativa L.) - Lucerne - Peroxidase - Isozymes

In~oduetion There has been limited success in interspecific hybridization of alfalfa (Medicago sativa L.) with other Medicago species, although a number of useful traits have been identified in wild Medicago species. Sources of germplasm for alfalfa improvement have been restricted to the M. sativa complex (Barnes et al. 1977). The M. sativa complex is actually one biological species comprised of several members frequently given species rank (e.g., M. falcata, M. media and M. glutinosa). All members of the complex freely intercross (particularly at the same ploidy level), and there is no hybrid sterility in the F1 or later generations (Lesins and Lesins 1979). Germplasm from all members of the complex can be incorporated readily into cultivated germplasm. The genus Medicago consists of both annual and perennial species with a number that contain useful agronomic traits, including disease and insect resistance and potential salt and drought tolerance. Annual species have been found to be resistant to yellow leafblotch caused by Leptotrochila medicaginis Schuepp (Semeniuk and Rumbaugh 1976), Verticillium albo-

773 atrum (Busch and Smith 1981) and race 1 and/or race 2 ofanthracnose caused by Colletotrichum trifolii Bain. (Elgin and Ostazeski 1982). In addition, resistance to the alfalfa weevil (Hypera postica Gyllenhal) has been observed in several annual species (Barnes and Ratcliffe 1969; Shade et al. 1975). Disease resistance also has been identified in perennial species including: resistance to spring blackstem (Phoma medicaginis Malb. and Roum) in M. dzhawakhetica Bordz. and M. suffruticosa Ram. (Renfro and Sprague 1959), Stemphyllium leafspot (Stemphyllium botryosum Wallr.) resistance in M. cancellata M.B. (Borges et al. 1976), Verticillium albo-atrum resistance in M. dzhawakhetica (Busch and Smith 1981), and resistance to race 1 and/or race 2 of anthracnose in six perennial species (Elgin and Ostazeski 1982). In addition, glandular haired genotypes of M. prostrata Jacq. may provide resistance to both alfalfa weevil larvae and potato leafhopper nymphs, Empoascafabae Harris (Sorensen et al. 1983). Sexual crosses have provided only one hybrid by crossing perennial M. sativa with annual M. scutellata (L) Mill (Sangduen et al. 1982). This hybrid was produced following gibberellic acid treatment of peduncles and pedicels after pollination. Interspecific hybrids between alfalfa and the following perennial Medicago species have been recovered by conventional sexual crosses: M. cancellata, M. glomerata Balb., M. papillosa Boiss., M. prostrata, M. rhodopea Velen., M. saxatilis M.B., and with F1 hybrids of M. daghestanica - M. pironae (Lesins and Lesins 1979). Recently, we obtained interspecific hybrids of M. sativa and M. dzhawakhetica by using uneven ploidy levels and a reproductive mutant (McCoy and Smith 1984). Pollination and fertilization studies have demonstrated that the major barrier to recovery of perennial Medicago interspecific hybrids is postfertilization (Oldemeyer 1956; Fridricksson and Bolton 1963 a; Sangduen et al. 1983 b; McCoy 1985); therefore, the recovery of perennial interspecific hybrids should be possible by some form of embryo rescue. For species where sexual crosses failed to produce hybrids, attempts to obtain Medicago interspecific hybrids via embryo rescue have been unsuccessful (Fridriksson and Bolton 1963 b; Elgin et al. 1977). In general, embryo culture or modifications thereof have been highly successful for recovering interspecific hybrids in other genera (Raghavan 1977; Collins et al. 1984). In particular, interspecific hybrids of several forage legumes have been recovered by some form of embryo rescue. These include interspecific hybrids in the genus Trifolium (Phillips etal. 1982; Williams and deLautour 1980) Lotus (Somaroo and Grant 1971; deLautour et al. 1978) and Ornithopus (Williams and deLautour 1981). Crossing experiments with several interspecific combinations conducted in our l a b o r a t o r y confirmed that fertilization occurred b u t embryos a b o r t e d before reaching maturity. Therefore, we a t t e m p t e d to rescue the interspecific h y b r i d embryos by various techniques previously a p p l i e d to legume species. Techniques tested included direct culture o f h y b r i d embryos, previously successful in Arachis (Bajaj et al. 1982) a n d Trifolium (Phillips et al. 1982), a n d in vitro ovule culture, previously successful in Glycine (Newell a n d H y m o w i t z 1982). A modification o f these techniques, designated the o v u l e - e m b r y o culture method, resulted in efficient recovery o f the first interspecific hybrids o f M. sativa and M. rupestris M.B. ( M c C o y 1985). This p a p e r presents (1) the successful application o f the o v u l e - e m b r y o culture technique to the recovery o f

several new Medicago interspecific hybrids including new trispecies combinations, and (2) the m o r p h o l o g y , cytology and peroxidase isozyme patterns o f parents and interspecific hybrids.

Materials and methods Plant material Diploid (2n=2x= 16) and tetraploid (2n=4x=32) clones of M. sativa were used in crosses with the other Medicago species. Normal diploids and diploid clones homozygous for jp were used as females. Plants homozygous forjp lack the post-meiotic cytokinesis during microsporogenesis resulting in 4n pollen (McCoy and Smith 1983). The jpjp clones used in this study were self-sterile. In addition the jpjp clones used frequently had abnormal female gametophytes including the presence of supernumerary nuclei in the central cell (unpublished results). Diploid male clones were of two types: normal or homozygous for rp. The rp gene conditions 2n pollen formation by a mechanism genetically equivalent to first division restitution (McCoy 1982). The rprp plants produce both n and 2n gametes, thus allowing the concurrent sampling of two ploidy levels. Normal tetraploid clones were used as male parents, and cytoplasmic male sterile (CMS) clones were used as female parents at the tetraploid level. The following species were used in the interspecific crossing experiments: Medicago cancellata (2n=6x=48), M. dzhawakhetica (2n = 4x = 32), M. hybrida (2n = 2x = 16), M. marina (2n = 2x = 16), M. papillosa (2n = 2x = 16 and 2n = 4x = 32), M. rhodopea (2n = 2x -- 16), M. rupestris (2n = 2x = 16) and M. saxatilis (2n = 6x = 48). Medicago cancellata, M. rhodopea, M. rupestris and M. saxatilis are four species that comprise subsection Rupestres Grossh., section Falcago of the genus Medicago (Lesins and Lesins 1979). The subsection is charcterized by pods having prominent veins either all over the coil face or at the peripheral part. The two hexaploid species M. cancellata and M. saxatilis have a genetic structure described as alloautoploid (Lesins 1970). Medicago cancellata is considered to be composed of four genomes from M. rupestris and two genomes of M. sativa; while M. saxatilis is thought to be derived from four genomes of M. rhodopea and two genomes of M. sativa. The following accessions were used in the crossing experiments: M. canceltata UAG 43 (UAG refers to the University of Alberta Genetics Department and comes from the collection of Dr. K. Lesins), M. rhodopea UAG 493, 494 and 496, M. rupestris UAG 1847 and M. saxatilis UAG 586. One accession of M. dzhawakhetica UAG 98, was used. Earlier we described the morphological characteristics of this species particularly traits that distinguish this species from its close relative, M. papillosa (McCoy and Smith 1984). Medicago hybrida UAG 2028 was used in the crossing experiments. This is the only accession of M. hybrida available in North America. Medicago hybrida is. in Section Suffruticosae of the subgenus Medicago," however, as pointed out by Lesins and Lesins (1979), this species has often been classified in the genus Trigonella. Its pachytene karyotype (Gillies 1972) is distinct from M. sativa, and the species of Section Suffruticosae are unique in that small amounts of fl-zeacarotene are present in the petals (Ignasiak and Lesins 1975). This carotenoid has not been found in any other perennial Medicago species. Additional morphological criteria include yellow flower color and pods that are flat, wide and glabrous.

774 A number of accessions of M. marina were used including U A G 319, 320, 1180, 1208, 1319, 1388, 1472, 1567, 1569 and 1649. Medicago marina is the only species of Section Marinae, subgenus Medicago. Plants of this species are unique to Medicago in that they grow exclusively on seashores in loose sands; therefore, this species has been suggested as a source of salt tolerance (Lesins and Lesins 1979). The whole plant including pods is covered with simple hairs. Three plant introductions of M. papillosa (P.I. 464697, P.I. 464700 and P.I. 464704) were obtained from the USDA Plant Introduction Center at Ames, Iowa. Of the three introductions P.I. 464697 was predominantly tetraploid whereas 464700 and 464704 were predominantly diploid. The M. papillosa introductions had the distinct characteristics representative of M. papillosa described in an earlier publication (McCoy and Smith 1984).

Crossing studies All possible combinations of crosses were made between the various ploidy levels and genotypes of M. sativa and the wild species. Data were tabulated by pods per pollination where a value of 0.01 equals one pod per 100 pollinations. In addition, interspecific F~ hybrids were backcrossed to M. sativa where possible. An attempt also was made to recover trispecies hybrids by crossing the F~ interspecific hybrid to a third species. Suction emasculation was used to prevent self-fertilization when the female parent had some level of self-fertility.

Table 1. Molar concentrations (mM) of four media tested for successful recovery of Medicago interspecific hybrids Component

Ca +2 NO~Na + SO~-2 K+ C1PO~"2 Mg +2 NH~ Mn § Zn +2 Fe +2 BO~-3 IMoO~ 2 Cu +2 Co +2

Media EC2"

B & N"

L2 b

SHc

1.2 2.0 1.6 4.4 1.7 0.9 0.1 2.9 0.031 0.010 0.130 0.024 0.005 0.001 0.0001 0.0001

1.5 18.4 0.1 0.9 9.9 1.5 0.5 0.8 9.0 0.112 0.035 0.130 0.162 0.001 0.0001 -

4.1 33.3 0.6 1.9 23.2 4.1 3.0 1.8 12.5 0.090 0.017 0.090 0.082 0.006 0.002 0.0004 0.0004

1.4 24.8 1.7 24.8 2.7 2.6 1.6 2.6 0.040 0.003 0.050 0.080 0.006 0.0003 0.001 0.0004

Source: Williams and deLautour (1980) b Source: Phillips and Collins (1979) c Source: Schenk and Hildebrandt (1972) a

Ovule-embryo culture Immature pods of all interspecific combinations were removed seven to 28 days after pollination. Pods were surface sterilized by treating for 1.5 min in 95% ethanol followed by 1.5 rain in 25% commercial bleach and two rinses in distilled water. Embryo rescue experiments were conducted using 1) direct culture of immature embryos, 2) culture of fertilized ovules to germination without embryo removal, and 3) culture of ovules containing the immature embryo followed by removal of the embryo and direct culture. All dissections were preformed aseptically under magnification (50 to 100• Several media formulations (Table 1) were evaluated including the L2 medium of Phillips and Collins (1979) minus the growth regulators, SH medium (a culture medium routinely used for alfalfa cell cultures) of Schenk and Hildebrandt (1972), and B & N and EC2 media used by Williams and de Lautour (1980). Sucrose concentration of all media was 25 g/1. In addition to the various media formulations, several modifications of the L2 medium (all minus the growth regulators) were tested to determine the optimum concentration and type of inorganic nitrogen. Concentrations of NH4NO3 tested were 0, 6.25, 12.5, 25 and 50 mM. NH + salts, NI-I4NO3, NH, C1 and (NI-L)2SO4, were tested at a concentration of 12.5 mM NH +. The optimum cultural sequence was determined for each interspecific combination. All in vitro cultures were maintained at 24.5 ~ under continuous illumination (90 ~tEm-~ s-l). For experiments using direct culture of immature embryos and fertilized ovule culture, the ovules were removed at four to five day intervals from seven to 28 days post-pollination (DPP). On a given day, immature embryos were removed from onehalf of the fertilized ovules and the other one-half were used for ovule culture. Additional experiments tested the method of removing and directly culturing the embryo following a period of ovule culture. For these experiments, ovules were placed in culture 14 to 20 DPP and then at three day intervals (from three to 15 days) the embryo was removed and cultured directly on fresh medium.

Plantlets at the unifoliate-leaf stage were transferred to 125 ml Erlenmeyer flasks. After five trifoliate leaves had emerged, plants were transferred to Jiffy-7 peat pellets ~, and grown under high humidity with a 16-h day (light intensity of 105 ~tEm -2 s -1 from cool white fluorescent lamps) until roots emerged from the peat pellet. Following an additional two week hardening period at ambient humidity, plants were transferred to the glasshouse. Interspecific hybrid plants were clonally propagated by vegetative cuttings and, if vigorous, eventually transferred to the field.

Cytology Mitotic and meiotic chromosome analysis of all F1 hybrids and a sample ofbackcross (BC) progeny were conducted according to methods of McCoy and Smith (1984). Pollen stainability and pollen germination were determined by methods of McCoy and Smith (1983).

Enzyme extraction and electrophoresis Electrophoretic assays were conducted on enzyme extracts from plants that were propagated as vegetative cuttings grown in sand. Plants at first bud stage were removed from the sand and the roots were thoroughly washed. The healthiest appearing root tips (1 to 3 cm long) were removed, weighed and then ground in a chilled mortar and pestle with 1 ml trisglycine buffer (pH 8.75 at 4 ~ per 1 g of fresh root tissue. This solution was centrifuged at 18,000• for 8min, the supernatant was then mixed with three crystals of sucrose per ml and used immediately for isoelectric focusing. Pharmacia Gel Electrophoresis Apparatus GE-2/4LS 1 was used for all assays. Assays were conducted with uniform 7% and gradient 4-30% polyacrylamide gels. The buffer used was tris-glycine (pH 8.75) for both gel types. Gels were prefocused

775 for 15 min at 125 V. Samples were then added using 100 ~tl per lane. Uniform gels were run for 2 h at I0~ at 125 V, whereas gradient gels were run for 22 h at 10~ at 125 V. Gels were immediately stained either for peroxidase (PRX) or leucineamino-peptidase (LAP) activity according to the procedures of Quirts (1981). Optimum PRX staining occurred within 30-60 rain, optimum LAP staining was 90 min. 1 Mention of a trademark, proprietary product or vendor does not constitute a guarantee or warranty of the product by the U.S. Department of Agriculture, and does not imply its approval to the exclusion of other products or vendors that may also be suitable

Table 2. Effect of four media on the recovery of M. sativa • M. rupestris F1 hybrids Media

No. of ovules cultured

No. of "viable" embryos recovered

No. of F1 plants obtained

EC2 a B &N" L2 b SHc

118 125 125 118

2 54 73 23

0 7 15 0

a Source:Williams and deLautour (1980) b Source: Phillips and Collins (1979) Source: Schenk and Hildebrandt (1972)

Results Medicago sativa X M. rupestris embryo rescue experiments

Crossing diploid M. sativa, genotype jpjp, with diploid M. rupestris produced good pod set (about one pod per two flowers pollinated); however, no mature seed was obtained. Ovule dissections at two to three day intervals beginning five days post pollination (DPP) demonstrated embryo development similar to intraspecific M. sativa crosses though at a significantly slower rate. Intraspecific M. sativa crosses were at the globular stage of development 7 to 8 DPP, at the heart stage 10-13 DPP and at the torpedo stage 14-19 DPP. In contrast, the interspecific M. sativa - M. rupestris embryos were at the globular stage 9-11 DPP and at the heart stage 12-22 DPP. Interspecific hybrid embryos did not develop beyond the late heart to very early torpedo stage of development. Because globular stage embryos were more easily damaged than heart stage embryos, culture experiments concentrated on embryos 14-20 DPP. No interspecific hybrid embryos were recovered by direct culture of the embryo regardless of media or DPP tested. In all cases, hybrid embryos became necrotic three to four days after culture initiation. Ovule culture until germination of the embryo was also unsuccessful; however, it was noted that embryos continued to develop in the ovule for several days before turning necrotic after 10 days culture in ovulo. Therefore, we attempted to rescue the embryos by removing and directly culturing the embryo after a period of culture in ovulo. Preculturing the fertilized ovule for six to nine days and then removing the embryo resulted in continued development of the embryo. Removal of the embryo prior to four days or after 10 days culture in ovulo was unsuccessful. Initial experiments with this method resulted in the recovery of 15 M. sativa x M. rupestris hybrid plantlets of which five survived transfer to soil and were grown to ma-

turity. This procedure, designated ovule-embryo culthen was used in additional experiments to identify optimal media using the media listed in Table 1. The B & N and L2 media were both acceptable for rescuing M. sativa X M. rupestris hybrid embryos via the ovule-embryo culture method (Table 2). No hybrid plants were obtained when EC2 and SH media were used, though 23 of the 118 ovules cultured on SH media produced viable embryos. The media tested differed substantially in regard to the concentration of inorganic components (Table 1). A major difference in the media that correlated with success rate was the NH + concentration. The two successful media L2 and B & N had a concentration of 12.5 mM NH + and 9.0 mM, respectively. The unsuccessful media SH and EC2 had a NH + concentration of 2.6 mM and 0 mM, respectively. During the time period of testing the four media the original five M. sativa • M. rupestris hybrids flowered, and were found to be male and female sterile. However, excellent pod set (about 80 pods per 100 pollinations) was observed when the F1 was the female parent in backcrosses to M. sativa. Given the probable importance of NH +, the effects of NH + concentration and NI-I~ salt on recovery of BC1 embryos were tested. Fertilized M. sativa - M. rupestris • M. sativa BC1 ovules were used because of the ease of production resulting in greater numbers. Table 3 clearly demonstrates the essential role of NH + in recovering BC~ embryos and plants. In the absence of NH + only one viable BCx embryo, which did not develop into a plant, was obtained from 283 cultured ovules. A low frequency of embryos and plants was obtained at the highest concentration (50 mM), but clearly the optimal NH + concentration was between 6.25 and 25 mM both in recovery of embryos per cultured ovule and plants per cultured embryo. Although NI-I~ concentration is critical for embryo rescue, the NH + salt did not significantly affect the re-

776 Table 3. Means (X) and standard errors (SE) of the number of embryos obtained per ovule cultured, and the number of BC1 (M. sativa-M, rupestris• M. sativa) plants obtained per embryo cultured on media containing different concentrations of NH4NO3 Concentration of NH,NO~(mM) a

Viable embryos per ovule cultured (%) b

0 6.25 12.5 25.0 50.0

1.0 32.4 44.8 34.9 18.9

BCa plants per embryo cultured (%)

SE

X

SE

1.4 12.6 11.9 9.9 6.0

0 35.5 61.6 69.8 3.2

10.9 16.3 5.3 4.6

a Media listed are identical L2 medium except for the concentration of NI-hNOs b Mean of 3 experiments consisting of 112, 95 or 76 ovules for each of the 5 concentrations

Table 4. Means (X) and standard errors (SE) of the number of viable embryos per ovule cultured and the number of BCa (M. sativa-M, rupestris X M. sativa) plants per embryo cultured on media containing different NI--ffsalts NI-ff salt

Viable embryos per ovule cultured (%)

BC1 plants per embryo cultured (%)

X~

SE

X~

SE

28.9 22.7 22.8

5.5 3.0 12.8

50.8 37.9 62.8

17.6 11.7 9.1

The only interspecific combination resulting in mature seeds was when diploid M. sativa was crossed with tetraploid M. papillosa. This interspecific combination will be discussed later. For all interspecific combinations, pods produced per pollination was greatest when diploid M. sativa, homozygous for jp, was used as the female parent. Embryo rescue experiments utilizing either direct embryo culture or ovule culture were also unsuccessful for all combinations tested. Direct embryo culture resulted in necrotic embryos within three days after placement on media. No viable embryos were observed if cultured in ovulo for longer than 20 days. The ovule-embryo culture technique was essential for the recovery of four additional Medicago interspecific combinations and two trispecies combinations. Table 6 summarizes the number of hybrid plants obtained for several new interspecific combinations. Pods per pollination varied from 0.01 for M. sativa • M. hybrida to 0.78 for the BC~ of M. sativa - M. rupestris• M. sativa. Medicago sativa X M. rhodopea embryos were rescued relatively easy, whereas hybrids between diploid M. sativa and diploids M. papillosa, M. marina and M. hybrida were extremely difficult to recover (Table 6). Descriptions for each combination presented in Table 6 follow. Medicago sativa • M. rupestris

(NI-I4)2SO4 NI-I4C1 NI-~NO~

a Mean of three experiments consisting of either 36, 65 or 78 ovules cultured for each of the three NH4 salts

covery of viable BCx (M. sativa - M. rupestris X M. sativa) embryos (Table 4). However, NH4C1 was inferior to NI-I,NO3 in the recovery of BCa plants from the subsequently cultured viable embryos; 62.8% of the viable embryos developed into plants on NH, NO~, whereas 37.9% of the embryos resulted in plants on NI-I4C1. Based on these results further experiments testing the potential of this technique for recovering other Medicago interspecific hybrids with known postfertilization barriers were conducted with L2 medium containing 12.5 mM NH4NO~ (original L2). Additional interspecific hybridization experiments

Crossing experiments with M. sativa and M. rupestris have been described (McCoy 1985). Results of crossing various ploidy levels and genotypes of M. sativa with additional Medicago species are summarized in Table 5.

Morphology and cytogenetics have been described for the M. sativa • M. rupestris F1 and BC1 plants (McCoy 1985). Electrophoretic assays of parental clones and F~ hybrids demonstrated that PRX isozyme phenotypes could be used to confirm hybridity. Figure 1 a shows the PRX isozyme phenotypes for the parental M. sativa clone used as the female, the parental M. rupestris clone used as the male, and an F1 hybrid between the two clones. The F1 hybrid had at least 5 isozymes unique to the M. rupestris clone. Medieago sativa • M. rhodopea

Hybrids between M. sativa and M. rhodopea were recovered with the greatest efficiency of all hybrids listed in Table 6. Slightly more than one-half of all flowers pollinated produced pods. Hybrid plants were recovered from 47.3% of the ovules cultured. This efficiency is likely related to the observation that hybrid embryos from this cross developed as far as the torpedo stage, the latest developmental stage observed for any of the hybrid combinations. Experiments testing the optimal period of in ovulo embryo culture used 17 to 20 DPP ovules. All 22 embryos removed after three days culture in ovulo died. Sixteen of 20 embryos removed after six days culture in ovulo and 10 of 26 embryos removed after

777 Table 5. Results of crossing various ploidy levels and genotypes of M. sativa with other Medicago species Cross ~

No. of pollinations b

Pods per pollination c

M. rhodopea (2 • ) M. rhodopea (2 x ) ) x M. sativa (2 • ) ) • M. sativa (rprp) M. rhodopea (2 X ) ) x M. sativa (4 x )

132 2,494 66 51 1,115 60

0.08 0.54 0.05 0.05 0.003 0

M. sativa (2 X ) • M. papillosa (2 X ) M. sativa (I'PjP) X M. papillosa (2 • ) M. papillosa (2 X ) X M. sativa (2 x ) M. papitlosa (2 • ) • M. sativa (rprp) M. sativa (4 • ) X M. papillosa (2 X ) M. papillosa (2 X ) X M. sativa (4 x ) M. sativa (2 X ) X M. papillosa (4 • ) M. sativa (jpjp) • M. papillosa (4 X ) M. papillosa (4 • ) X M. sativa (2 X ) M. papillosa (4 X ) X M. sativa (rprp) M. sativa (4 X ) • M. papillosa (4 X ) M. papillosa (4 • ) X M. sativa (4 X )

52 2,160 92 69 2,340 112 105 1,540 92 160 865 68

0 0.03 0.01 0 0.01 0.01 0.08 0.46 0.04 0.07 0 0

0 0 0 0 0 0 6 785 5 10 0 0

M. M. M. M. M. M.

sativa (2 • ) x sativa (jpjp) • rhodopea (2 x rhodopea (2 x sativa (4 X ) X rhodopea (2 x

Seeds obtained 0 0 0 0 0 0

M. M. M. M. M. M.

sativa (2 X ) X M. marina (2 • ) sativa (jpjp) X M. marina (2 X ) marina (2 • ) X M. sativa (2 • ) marina (2 X ) • M. sativa (rprp) sativa (4 X ) • M. marina (2 X ) marina (2 • ) X M. sativa (4 X )

215 2,016 98 56 1,555 65

0 0.02 0.01 0 0 0

0 0 0 0 0 0

M. M. M. M. M. M.

sativa (2 • ) • M. hybrida (2 X ) sativa (jpjp) X M. hybrida (2 • ) hybrida (2 • ) • M. sativa (2 X ) hybrida (2 • )X M. sativa (rprp) sativa (4 • ) X M. hybrida (2 X ) hybrida (2 • ) x M. sativa (4 X )

161 2,933 46 71 1,875 50

0 0.011 0 0 0 0

0 0 0 0 0 0

a In all crosses M. sativa (2 x ) clones were normal male and female fertile plants, M. sativa (jpjp) were diploid clones homozygous forjp, M. sativa (rprp) were diploid clones homozygous for rp. At the tetraploid level, M. sativa (4 x ) clones used as females were CMS and M. sativa (4 • ) clones used as males were normal, male- and female-fertile tetraploids b The reason for the greater number of pollinations w h e n j p j p clones or 4 • CMS clones were used in crosses is that tedious emasculations were not required c A value of 0.01 equals one pod per 100 pollinations; based on pods developing 10 DPP

nine days culture in ovulo d e v e l o p e d into h y b r i d plants. After 12 and 15 days culture in ovulo no viable appearing embryos were observed. Most o f the M . s a t i v a • r h o d o p e a hybrids were diploid, although a low frequency o f triploids (4%) a n d tetraploids (15%) also were recovered (Table 7). The triploids and tetraploids likely resulted from 2n gametes from one or b o t h parents, respectively. G e n e r a l m o r p h o l o g y o f the M . sativa • M . r h o d o p e a hybrids was intermediate and most hybrids were vigorous. L e a f size and shape (Fig. 2a) were intermediate. W h e n purple-flowered clones o f M . sativa were used as parents the F1 h a d variegated flowers due

to a blending o f anthocyanins, carotenoids a n d yellow flavonoids. I f yellow-flowered M . sativa clones were crossed the F1 hybrids h a d yellow flowers. Pods h a d the p r o m i n e n t veins characteristic o f the M . r h o d o p e a parent, and they were intermediate in size. Isozyme p h e n o t y p e s o f P R X and L A P confirmed hybridity. F o u r PRX b a n d s unique to M . r h o d o p e a were observed in the FI interspecific h y b r i d (Fig. 1 b). Preliminary cytogenetic analysis was conducted on five diploid M . sativa • M . r h o d o p e a hybrids. In all five hybrids the p r e d o m i n a n t pairing configuration in pollen m o t h e r cells (PMC's) was either eight bivalents or seven bivalents a n d two univalents. F o u r o f the five

778 Table 6. Summary of the number of flowers pollinated, pods per pollination, number of ovules cultured, number of embryos cultured and the number of hybrid plants obtained for various Medicago interspecific combinations Cross

No. of flowers pollinated

Pods per pollination a

No. of ovules cultured b

No. of embryos cultured c

No. of hybrid plants obtained d

M. sativa (2 • )e x M. rupestris (2 • ) FI [M. sativa x M. rupestris] (2 x ) X M. sativa (2 X ) M. sativa (2 X ) x M. rhodopea (2 • ) M. sativa (2 • ) x M. papillosa (2 • ) M. sativa (2 X ) x M. marina (2 x ) M. sativa (2 X ) • M. hybrida (2 x ) F~ [M. sativa (2 X ) x M. dzhawakhetica (4 x )] (3 • ) x M. cancellata (6 x ) F~ [M. sativa (2 • ) x M. dzhawakhetica (4 x )] (3 x ) • M. saxatilis (6 x )

2,645 2,084 2,494 2,160 2,016 2,933 1,560

0.41 0.78 0.54 0.03 0.02 0.01 0.08

451 547 329 83 39 28 75

317 450 256 30 4 12 57

66 (14.6) 323 (56.3) 157 (47.3) 13 (15.7) 2 (5.1) 2 (7.1) 40 (53.3)

2,057

0.09

91

73

52 (57.1)

a A value of 0.01 equals one pod per 100 pollinations; based on pods developing 10 DPP b All ovules cultured on L2 medium. For M. s a t i v a x M , rupestris s and BC1 the NH4NO3 concentration was 25 mM, and for all other combinations the NH4NO3 concentration was 12.5 mM Indicates the number of embryos that appeared to be viable when dissected from the ovule and placed on fresh medium d The number in parenthesis represents the percentage of hybrid plants obtained per ovule cultured e For all crosses the diploid M. sativa clones used as female parents were homozygous forjp

Table 7. Chromosome numbers of hybrid plants obtained via ovule-embryo culture Hybrid

M. sativa x M. rhodopea M. sativa x M. papillosa M. sativa • M. marina M. sativa x M. hybrida [M. s a t i v a x M , dzhawakhetica]x M. eancellata [M. s a t i v a x M , dzhawakhetica]• M. saxatilis

No. of plants tested

No. of plants exhibiting chromosome no. 2 n = 2 x = 16

2n=3x=24

2n=4x=32

47 13

38 13

2 .

7 .

.

.

2 2

2 2

. .

. .

. .

. .

20

-

-

-

15

21

-

-

-

18

2n=6x=48 -

Other" -

1 (34), 2 (40), 1 (45), 1 (46) 1 (40), 1 (41), 1 (46)

a Number in parenthesis indicates chromosome number

h y b r i d s h a d what a p p e a r e d to be a n accessory n u c l e o l u s p r e s e n t in PMC's. Male fertility o f the M . sat i v a X M , r h o d o p e a h y b r i d s e x a m i n e d thus far has r a n g e d from c o m p l e t e sterility up to a b o u t 50% stainable pollen. First backcross (BC1) p r o g e n y were recovered from m a t u r e seed w h e n three of the F1 h y b r i d s were used as males. Viable BCx seed also was p r o d u c e d w h e n the F~ h y b r i d s were used as females a n d crossed with u n r e l a t e d d i p l o i d M. sativa male plants. M e d i c a g o sativa • M . p a p i l l o s a

Hybrids

were recovered from seed w h e n diploid was crossed with tetraploid M. p a p i l l o s a (Table 5). Seeds were recovered in all four crosses M . sativa

i n v o l v i n g diploid M. sativa a n d tetraploid M. p a p i l l o s a regardless o f which species was used as m a l e or female. C h r o m o s o m e n u m b e r was d e t e r m i n e d for 44 h y b r i d s o b t a i n e d from seed a n d all were triploid (2n = 3x = 24). These triploid hybrids were likely c o m p r i s e d of o n e M. sativa g e n o m e a n d two M . p a p i l l o s a genomes. O v u l e - e m b r y o culture was essential for the recovery o f diploid h y b r i d s from crossing diploid M. sativa a n d diploid M. p a p i l l o s a (Tables 5 a n d 6). This h y b r i d c o m b i n a t i o n is difficult to p r o d u c e as d e m o n s t r a t e d b y the m e a n p r o d u c t i o n o f o n l y three pods per 100 p o l l i n a tions a n d the recovery o f o n l y 13 h y b r i d plants (15.7%) from the m o r e t h a n 80 ovules c u l t u r e d (Table 6). O n l y limited data was collected for o p t i m a l period o f in ovulo culture because o f the shortage o f material. Pods that r e m a i n e d o n the p l a n t for l o n g e r t h a n 17 days

779

Fig. l a-f. Peroxidase isozyme phenotypes for Medicago sativa, M. hybrida, M. marina, M.papillosa, M. rhodopea, and M. rupestris, and the F~ interspecific hybrids. Diploid M. sativa clones were used as the female parent in all crosses. The origin is at the top, and the anodal front is at the bottom. Arrows show bands unique to the wild Medicago species that were present in the F~ interspecific hybrid, a Peroxidase isozyme phenotypes of the M. sativa parent (S), the F~ hybrid and the M. rupestris parent (R) on gradient (4-30%) gels; b peroxidase isozyme phenotypes of the M. rhodopea parent (R), the F1 hybrid and the M. sativa parent (S) on gradient (4-30%) gels; c peroxidase isozyme phenotypes of the M. sativa parent (S), the M.papillosa parent (P) and four F1 hybrids on uniform 7% gels; d peroxidase isozyme phenotypes of the M. sativa parent (S), the M. papillosa parent (P) and two F1 hybrids on uniform 7% gels; e peroxidase isozyme phenotypes of the M. marina parent (M), a reference M. sativa plant (St), the M. sativa parent (Sp) and the Fx hybrid on gradient (4-30%) gels; f peroxidase isozyme phenotypes of the M. hybrida parent (H) the M. sativa parent (S) and the FI hybrid on gradient (4-30%) gels

turned brown and the ovules contained inviable embryos. Ovules 14 to 16 DPP were used, and similar to previous combinations all 11 embryos removed after three days culture in ovulo became necrotic in less than four days. Of the 13 hybrid plants obtained six were derived from embryos removed after six days culture in ovulo, three were from embryos cultured nine days in ovulo and four were from embryos cultured 12 days in ovulo. All 13 hybrids recovered from ovule-embryo culture were diploid (Table 7). Whereas, triploid hybrids from M. sativa ( 2 x ) • papillosa (4x) were vigorous and flowered profusely, the diploid hybrids were extremely weak. Six of the hybrids died and only two surviving hybrids flowered. Only a few flowers were groduced in an erratic manner preventing meiotic analysis of these diploid hybrids. Although the diploid hybrids were extremely weak, leaf morphology (Fig. 2 b) and growth habit were intermediate. Medicago papillosa parents had a prostrate growth habit, whereas M. sativa was upright; the F1 hybrids were prostrate to slightly ascending. Peroxidase isozymes on uniform 7% polyacrylamide gels were used to definitively identify the hybrids. Figure 1 c shows that one M. sativa parent was heterozygous producing two bands near the anodal front, and another M. saliva parent was single banded homozygous (Fig. 1 d). The M. papillosa parent was homozygous for a faster migrating band than the M. sativa parents. Figure 1 c also shows the peroxidase phenotypes of four F1 hybrids between M. sativa and M. papillosa. All four have the unique band of M. papillosa and either the faster migrating M. sativa band (2 plants) or the slower migrating band (2 plants). When the cross was made with the homozygous M. sativa parent, the F~ hybrids had the band from M. sativa and the unique band of M. papillosa (Fig. 1 d). Medicago sativa x M. marina

Hybrids between M. sativa and M. marina were difficult to produce. Only two pods per 100 pollinations were set, and plants were recovered from only 5% (2 of 39) of the cultured ovules (Tables 5 and 6). Similar to M. sativa x M. papillosa crosses the pods turned brown after 17 DPP, and no viable embryos were observed. Both plants recovered were derived from embryos cultured in ovulo for nine days. The ovules were cultured 15 DPP. The two M. s a t i v a x M , marina hybrids recovered were diploid (Table 7), and they were extremely weak and slow to develop. Shoots on both plants became chlorotic and died back after about five leaves reached the fully expanded stage. Additional shoots developed from the crown; however, neither plant flowered, pre-

780 ovules cultured 15 DPP. One of the hybrid plants was derived from an embryo cultured for nine days in ovulo, and the other hybrid was from an embryo cultured 12 days in ovulo. Both M. s a t i v a x M , hybrida hybrids were diploid (Table 7), but no meiotic analysis was conducted due to lack of floral initiation. Although M. sativa X M. hybrida hybrids were more vigorous than M. sativa - M. marina they were still very weak. Morphology of the hybrids was intermediate for leaf size and shape (Fig. 2 d), and growth habit was prostate similar to M. hybrida. Four PRX isozymes that were present in the F1 hybrid were unique to M. hybrida (Fig. 1 f). Trispecies hybrids

Fig. 2a-d. Leaves of parent plants and F1 interspecific hybrids, a M. sativa (SAT), the F1 hybrid and M. rhodopea (RHO). b M. sativa (SAT), the Fa hybrid and M.papillosa (PAP); c M. sativa (SAT), the F1 hybrid and M. marina (MAR); d M. sativa (SAT), the F~ hybrid and M. hybrida (HYB)

cluding meiotic analysis. Morphology of the hybrids was more similar to M. marina than M. sativa. Leaflet shape was nearly intermediate (Fig. 2 c); however vegetative parts were covered with simple hairs similar to M. marina.

Figure l e shows four PRX isozymes unique to M. marina that were observed in an F1 hybrid. In Fig. 1 e, Sp refers to the M. sativa parental clone used to produce the Fa M. s a t i v a x M , marina hybrid. The Sr band is a reference M. sativa clone that is a progeny of Sp crossed with another M. sativa plant. Medicago sativa X M. hybrida

Similar to M. marina, hybrids between M. sativa and M. hybrida were difficult to produce. Approximately one pod was produced per 100 pollinations (Table 5), and only two hybrid plants were recovered from 28 cultured ovules (Table 6). Both hybrid plants were from

As a part of experiments to produce novel ploidy combinations in Medicago, attempts were directed toward the formation of hexaploids by crossing sterile, triploid M. sativa x M. dzhawakhetiea with the naturally occurring hexaploid species M. cancellata and M. saxatilis. However, no mature seed was recovered from either cross. Although both crosses resulted in less than 10 pods per 100 pollinations, the efficiency of plants obtained per ovule cultured was greater than any of the F1 combinations (Table 6). In both cases, greater than 50% of the cultured ovules produced a plant. Ovules cultured 14 to 20 DPP were used to determine the optimal period of in ovuto culture. For the trispecies combinations with M. cancellata and M. saxatilis all embryos removed after three days culture in ovulo turned necrotic. For the trispecies combination with M. cancellata, 22 hybrid plants were obtained from 30 embryos cultured in ovulo six days, 12 hybrids were derived from 15 embryos cultured in ovulo nine days and 6 hybrids were recovered from 10 embryos cultured in ovulo 12 days, The trispecies combination with M. saxatilis behaved in a similar manner with 18 hybrids from 25 embryos cultured in ovulo six days, 20 hybrids from 25 embryos cultured in ovulo for nine days, 11 hybrids from 15 embryos cultured in ovulo for 12 days and three hybrids from 15 embryos cultured in ovulo for 15 days. The majority of the trispecies hybrids obtained from both crosses were at or near the hexaploid level (Table 7). These hybrids likely resulted from the union of an unreduced egg (2n = 3x = 24) from the M. sativa x M. dzhawakhetica parent and a normal gamete from the hexaploid species. Vigor of the trispecies hybrids was generally good and even exceeded the parents in some cases (visual observation). Morphology of the trispecies hybrids was intermediate, and generally plants appeared to be a blend of characteristics from the three species involved. Male fertility of the trispecies hybrids ranged from

781 complete sterility to greater than 90% stainable pollen in one trispecies hybrid. Trispecies hybrids had workable levels of female fertility in crosses with hexaploid M. sativa and backcrosses to M. saxatilis and M. cancellata.

Discussion

As early as 1963 Fridriksson and Bolton (1963b) attempted to recover Medicago interspecific hybrids, but were unsuccessful. To our knowledge, this represents the first report of embryo rescue of Medicago interspecific hybrids via in vitro methods. The critical technique applied in this report involved the period of preculturing the ovule before excising and directly culturing the embryo. Perhaps new techniques such as the culture of fertilized pods demonstrated for intraspecific hybrids of annual and perennial Medicago species (Wang et al. 1984) will be utilized in the future. In addition, somatic hybridization via protoplast fusion is likely to be realized in the near future for sexually incompatible species similar to the recently produced fusion of the sexually compatible, taxonomic species M. sativa and M. falcata (Teoule 1983). Studies on media effects demonstrated the essential role of NH + for the recovery of plants following ovuleembryo culture similar to the essential role of reduced nitrogen for somatic embryogenesis from alfalfa cell cultures. Walker and Sato (1981) definitively showed that NH + was necessary for somatic embryogenesis. Subsequent research demonstrated an additional threefold increase in somatic embryos with the addition of proline (Stuart and Strickland 1984a), and the optimum medium included a combination of NH + and proline (Stuart and Strickland 1984 b). Our results were similar to the results of Walker and Sato (1981), in that there was little or no difference when NO~-, SO~ 2, or C1- salts of NH + were used. The essential role of NH + regardless of salt formulation also has been demonstrated for Gossypium hirsutum L. (cotton) embryos (Stewart and Hsu 1977, 1978). The succesful recovery of interspecific hybrids of several forage legumes by the transplanted nurse endosperm technique (Williams and deLautour 1980) also may have been due to NH +. Williams and deLautour (1980) found that their B & N and EC5 media were suitable for recovering embryos from globular stage to germination. Of the six media that they tested, only B & N and EC5 have NH + present. Because efforts were directed toward media for optimal recovery of M. sativa • M. rupestris F1 and BC1 hybrids, other hybrid combinations may require additional media modifications. For example, the NH + concentration of 12.5 mM may be too high for some interspecific hybrids. Media modifications involving additional sources of reduced nitrogen may identify c o m -

ponents that greatly improve the efficiency of recovering Medicago interspecific hybrids. The ovule-embryo culture method was found to be essential for the recovery of several new Medicago interspecific hybrids as well as trispecies hybrids. In addition, the ovule-embryo culture method has significantly improved the recovery of backcross generations, including backcrossing the triploid M. sativa • M. dzhawakhetica hybrids to both diploid and tetraploid M. sativa, and backcrossing triploid M. sativa • M. papillosa hybrids to both diploid and tetraploid M. sativa. Of the interpecific hybrid combinations listed in Table 6, only M. sativa • M. rhodopea and M. sativa • M. papillosa had been reported previously. The only previous report of M. sativa X M. rhodopea hybrids described the unsuccesful recovery of hybrids when diploid M. sativa and diploid M. rhodopea were crossed (Lesins 1972). Only two hybrids were recovered; one was triploid from crossing diploid M. sativa with a chromosomally doubled tetraploid M. rhodopea clone, and the other hybrid (derived from crossing tetraploid M. sativa and the somatically doubled tetraploid M. rhodopea) had 31 chromosomes (Lesins 1972). Lesins (1972) proposed that the hybrids with unbalanced chromosome numbers were recovered due to the removal of some incompatability factor. Our results demonstrate that diploid hybrids are easily recovered from the cross diploid M. sativa x diploid M. rhodopea. The artificial media must be providing nutrients capable of nurturing the hybrid embryos first in the presence, then in the absence of endosperm. There is only one previously reported cross of diploid M. M. papillosa (as M. dzhawakhetiea) that resulted in a single diploid hybrid (Clement 1963). Otherwise the only hybrids recovered from seed were derived from crossing diploid M. sativa and tetraploid M. papillosa. Resultant hybrids were triploid containing one M. sativa genome and two M. papillosa genomes (Lesins 1961; Lesins and Lesins 1979). We also have found that triploid hybrids are recovered from seed when diploid M. sativa is crossed with tetraploid M. papillosa particularly when the diploid M. sativa parent is homozygous forjp. Although we report the first hybrids between M. sativa and M. marina, Fridriksson and Bolton (1963a) previously had documented that the hybridization barrier was post-fertilization. Fertilization and early stages of embryo development were observed when diploid M. marina was crossed with either diploid or tetraploid M. sativa. The diploid hybrids we recovered from crossing diploid M. sativa, homozygous for jp, as the female with diploids of M. papillosa, M. marina and M. hybrida were all weak and never flowered. Additional crossing experiments with diverse parental clones may yield more vigorous interspecific hybrids. In addition, embryo rescue may lead to the recovery of hybrids from crossing tetraploid M. sativa and chochicine-doubled clones of the diploid Medicago species. sativa•

782 Isozyme phenotypes were found to be useful in hybrid confirmation. Peroxidase phenotypes o f all species used had at least one isozyme unique to the wild species that was present in the F1 interspecific hybrid. Isozyme analysis m a y prove to be useful for increasing the efficiency o f introgressing useful traits from wild Medicago species, similar to the methods discussed by Tanksley (1983) for tomato in particular and all crop species in general.

References

Bajaj YPS, Kumar P, Singh MM, Labana KS (1982) Interspecific hybridization in the genus Arachis through embryo culture. Euphytica 31 : 365-370 Barnes DK, Bingham ET, Murphy RP, Hunt O J, Beard DF, Skrdla WH, Teuber LR (1977) Alfalfa germplasm in the United States: genetic vulnerability, use, improvement and maintenance. USDA Agric Res Serv Bull 1571 Barnes DK, Ratcliffe RH (1969) Evaluation of annual species of Medicago as sources of alfalfa weevil resistance. Crop Sci 9:640-642 Borges OL, Stanford EH, Webster RK (1976) Sources and inheritance of resistance to Stemphyllium leafspot of alfalfa. Crop Sci 16:458-461 Busch LV, Smith E (1981) Susceptibility of Ontario-grown alfalfa cultivars and certain Medicago species to Verticillium albo-atrum. Can J Plant Pathol 3:169-172 Clement WM Jr (1963) Chromosome relationships in a diploid hybrid between Medicago sativa L. and M. dzhawakhetica Bordz. Can J Genet Cytol 5:427-432 Collins GB, Taylor NL, DeVerna JW (1984) In vitro approaches to interspecific hybridization and chromosome manipulation in crop plants. In: Gustafson JP (ed) Gene manipulation in plant improvement. Plenum Press, New York London, pp 323 383 deLautour G, Jones WT, Ross MD (1978) Production of interspecific hybrids in Lotus aided by endosperm transplants. N Z J Bot 16:61-68 Elgin JH Jr, McMurtrey JE III, Schaeffer GW (1977) Attempted interspecific hybridization of Medicago scutellata and M. sativa. Agron Abstr 69:54 Elgin JH Jr, Ostazeski SA (1982) Evaluation of selected alfalfa cultivars and related Medicago species for resistance to Race 1 and Race 2 anthracnose. Crop Sci 22:39-42 Fridriksson S, Bolton JL (1963a) Development of the embryo of Medieago sativa L. after normal fertilization and after pollination by other species of Medicago. Can J Bot 41:23-33 Fridriksson S, Bolton JL (1963b) Preliminary report on the culture of alfalfa embryos. Can J Bot 41:439-440 Gillies CB (1972) Pachytene chromosomes of perennial Medicago species. 3. Unique karyotypes of M. hybrida Trautv. and M. suffruticosa Ramond. Hereditas 72:303-310 Ignasiak T, Lesins K (1975) Carotenoids in petals of perennial Medicago species. Biochem Syst Ecol 2:177-180 Lesins K (1961) Interspecific crosses involving alfalfa. 1. Medicago dzhawakhetica (Bordz) Vass. x M. sativa L. and its peculiarities. Can J Genet Cytol 3:135 152 Lesins K (1970) Interspecific crosses involving alfalfa. 5. Medicago saxatilisx M. sativa with reference to M. cancellata and M. rhodopea. Can J Genet Cytol 12:80-86

Lesins K (1972) Interspecific crosses involving alfalfa. 7. Medicago sativa X M. rhodopea. Can J Genet Cytol 14:221-226 Lesins K, Lesins I (1979) Genus Medicago (Leguminosae): a taxogenetic study. Dr W Junk, The Hague, The Netherlands McCoy TJ (1982) The inheritance of 2n pollen formation in diploid alfalfa Medicago sativa. Can J Genet Cytol 24:315 to 323 McCoy TJ (1985) Interspecific hybridization of Medicago sativa L. and M. rupestris M.B. using ovule-embryo culture. Can J Genet Cyto127:238-245 McCoy TJ, Smith LY (1983) Genetics, cytology and crossing behavior of an alfalfa (Medicago sativa) mutant resulting in failure of the post meiotic cytokinesis. Can J Genet Cytol 25:390-397 McCoy TJ, Smith LY (1984) Uneven ploidy levels and a reproductive mutant required for interspecific hybridization of Medicago sativa L. and Medicago dzhawakhetica Bordz. Can J Genet Cyto126:511-518 Newell CA, Hymowitz T (1982) Successful wide hybridization between the soybean and a wild perennial relative, G. tomentella Hayata. Crop Sci 22:1062-1065 Oldemeyer RD (1956) Interspecific hybridization in Medicago. Agron J 48:584 585 Phillips GC, Collins GB (1979) In vitro tissue culture of selected legumes and plant regeneration from callus cultures of red clover. Crop Sei 19:59-64 Phillips GC, Collins GB, Taylor NL (1982) Interspecific hybridization of red dover (Trifolium pratense L.) with T. sarosiense Hazsl. using in vitro embryo rescue. Theor Appl Genet 62:17-24 Quiros CF (1981) Starch gel electrophoresis technique used with alfalfa and other Medicago species. Can J Plant Sci 61 : 745-749 Raghavan V (1977) Applied aspects of embryo culture. In: Reinert J, Bajaj YPS (eds) Plant cell, tissue and organ culture. Springer, Berlin, pp 375-397 Renfro G, Sprague EW (1959) Reaction of Medicago species to eight alfalfa pathogens. Agron J 51 : 481-483 Sangduen N, Kreitner GL, Sorensen EL (1983a) Light and electron microscopy of embryo development in perennial and annual Medicago species. Can J Bot 61:837-849 Sangduen N, Kreitner GL, Sorensen EL (1983b) Light and electron microscopy of embryo development in an annual x perennial Medicago species cross. Can J Bot 61:1241 to 1257 Sangduen N, Sorensen EL, Liang GH (1982) A perennial• annual Medicago cross. Can J Genet Cytol 24:361-365 Schenk RU, Hildebrandt AC (1972) Medium and techniques for induction and growth of monocotyledonous and dicotyledonous cell cultures. Can J Bot 50:199-204 Semeniuk G, Rumbaugh MD (1976) Reaction of some perennial and annual Medicago species and cultivars to the yellow leafblotch disease caused by Leptotrochila medicaginis. Plant Dis Rep 60:596-599 Shade RE, Thompson TE, Campbell WR (1975) An alfalfa weevil larval resistance mechanism detected in Medicago. J Econ Entomo163:399-404 Somaroo BH, Grant WF (1971). Interspecific hybridization between diploid species of Lotus (Leguminosae). Genetica 42:353-367 Sorensen EL, Horber EK, Stuteville DL (1983) Development of glandular-haired alfalfas with multiple pest resistance. In: Stuteville DL (ed) Proc 18th Central Alfalfa Improv Conf. Manhattan Kan, p 28

783 Stewart JM, Hsu CL (1977) In-ovulo embryo culture and seedling development of cotton (Gossypium hirsutum L). Planta 137:113-117 Stewart JM, Hsu CL (1978) Hybridization of diploid and tetraploid cottons through in-ovulo embryo culture. J Hered 69:404-408 Stuart DA, Strickland SG (1984a) Somatic embryogenesis from cell cultures of Medicago sativa L. 1. The role of amino acid additions to the regeneration medium. Plant Sci Lett 34:165-174 Stuart DA, Strickland SG (1984b) Somatic embryogenesis from cell cultures of Medicago sativa L. 2. The interaction of amino acids with ammonium. Plant Sci Lett 34:175-181 Tanksley SD (1983) Introgression of genes from wild species. In: Tanksley SD, Orton TJ (eds) Isozymes in plant genetics and breeding, part A. Elsevier, Amsterdam. pp 331-337

Teoule E (1983) Somatic hybridization between Medicago sativa L. and Medicago falcata L. CR Acad Sci, Ser III 297:13-16 Walker KA, Sato SJ (1981) Morphogenesis in callus tissue of Medicago sativa: the role of ammonium ion in somatic embryogenesis. Plant Cell Tissue Organ Cuk 1:109-121 Wang JW, Sorensen EL, Liang GH (1984) In vitro culture of pods from annual and perennial Medicago species. Plant Cell Rep 3:146-148 Williams EG, deLautour G (1980) The use of embryo culture with transplanted nurse endosperm for the production of interspecific hybrids in pasture legumes. Bot Gaz 141:252 to 257 Williams EG, deLautour G (1981) Production of tetraploid hybrids between Ornithopus pinnatus and O. sativus using embryo culture. N Z J Bot 19:23-30