Mean Chromosome Association Research Articles

Cytology of Paspalum chacoence and P. durifolium and Their Relationship to P. dilatatum

Paspalum chacoence Parodi is a sexual diploid with 20 chromosomes that pair as 10 II at meiosis. Paspalum durifolium Mez. has 60 chromosomes with essentially 30 II during meiosis and reproduces by aposporous apomixis. A yellow-anthered sexual biotype of Paspalum dilatatum, 2n = 4x = 40, was crossed with both species. Meiosis for six P. dilatatum x P. chacoence hybrids, 2n = 3x = 30, was irregular; a mean association of 27.32 I and 1.34 II per cell indicated that these species do not have homologous genomes. Two P. dilatatum x P. durifolium hybrids, 2n = 5x = 50, had irregular meiosis and mean chromosome associations of 20.4 I, 14.61 II, 0.1 III, and 0.01 IV, indicating one homologous genome and a partially homologous genome. On the basis of chromosome pairing in the species and the hybrids, P. durifolium was given the genome formula II J2J2 XX. This is the first indication that a form of the J genome is present in a species of the Quadrifaria group. Implications that P. durifolium may be the hexaploid progenitor of the common apomictic P. dilatatum biotype, 2n = 5x = 50, are discussed.

Cytology of Intra- and Interspecific Hybrids between two Cytotypes of Paspalum notatum and P. cromyorrhizon

Intra- and interspecific hybrids were made between diploid and tetraploid cytotypes of Paspalum notatum Flugge and P. cromyorrhizon Trin, to study their phylogenetic relationships. Two P. notatum (2n = 4x = 40) x P. notatum var. saurae Parodi (2n = 2x = 20) hybrids had 2n = 3x = 30. At metaphase I of meiosis, the mean chromosome associations were 1.93 I, 1.93 II, and 8.07 III, confirming that tetraploid P. notatum is an autotetraploid. Because P. notatum var. saurae has the genome formula NN, P. notatum can be given the formula NNNN. Two hybrids were recovered from the P. cromyorrhizon (2n = 2x = 20) x P. cromyorrhizon (2n = 4x = 40) cross: one had 30, and the other 40, chromosomes. Meiosis was irregular in both hybrids. The mean meiotic chromosome associations for the triploid were 4.81 I, 4.73 II, 4.92 III, 0.1 IV, 0.03 V, and 0.07 VI, and for the tetraploid, 0.84 I, 10.56 II, 0.21 III, 4.15 IV, 0.11 VI, and 0.03 VIII. Of seven P. notatum var. saurae x P. cromyorrhizon (2n = 2x = 20) hybrids, six had 20 chromosomes and a mean chromosome association of 9.04 I, 5.42 II, and 0.04 III, and one had 40 chromosomes and a mean association of 1.2 I, 6.33 II, 0.4 III, 6.13 IV, and 0.07 VI. The frequency of bivalents in the 20-chromosome hybrids and tetravalents in the 40-chromosome hybrid indicates that the genomes of these two species are partially homologous and that both species have forms of the N genome. Two hybrids from the P. notatum x P. cromyorrhizon (2n = 2x = 20) cross had 2n = 3x = 30. The mean pairing behavior was 9.38 I, 9.09 II, 0.78 III, and 0.02 IV, and some cells had as many as 5 III. These findings support the traditional taxonomic treatment that P. notatum and P. cromyorrhizon are separate species.

Chromosome association and fertility in hybrids of Iris laevigata Fisch. x I. ensata Thunb.

Chromosome association at MI of PMCs in hybrids of I. laevigata x I. ensata was examined and compared with that of parental species. The mean chromosome association per cell in the parental species was 0.02I+11.99II for I. ensata (‘Kacho’, 2n=24), 0.11I+15.94II and 0.15I+15.92II for I. laevigata (‘Yukidoro’ and wild type, 2n=32), respectively. On the other hand, the mean chromosome association per cell in the hybrids (2n=28) was 20.22I+3.88II+0.0046IV for ‘Yukidoro’ × ‘Kacho’ and 20.92I+3.54II for wild type × ‘Kacho’. Thus, the hybrids indicated a low homology between I. ensata and I. laevigata, and did not exhibit pollen and seed fertility. The utility of the hybrids was discussed.

An assessment of preferential chromosome pairing at meiosis in Dactylis glomerata

Meiosis was analyzed in 26 full-sib genotypes of tetraploid orchard grass (Dactylis glomerata L., 2n = 4x = 28) that had given apparently nontetrasomic segregation for reaction to the stem rust fungus, Puccinia graminis Pers. f. sp. dactylidis Guyot et Massenot. The mean chromosome associations were 0.24 I, 1.64 rod II, 6.47 ring II, 0.05 III, 0.85 chain IV, and 1.99 ring IV. Optimizations for two classes of mathematical models showed an appreciable excess of bivalents over the number expected upon independent association of long and short chromosome arms. The most probable cause of this bivalent excess was preferential pairing due to slight genomic differentiation in an autotetraploid background. Excess bivalents may have also arisen from close proximity of the two ends of each chromosome upon initiation of pairing. Inequality of chiasma formation between long and short chromosome arms was a minor contributor of excess bivalents. Because of independent segregation of homologous groups, some sibs fit the Kimber–Alonso 2:2 model best, others the 2:1:1.

Chromosome association and pollen fertility in some interspecific hybrids of Lilium

Meiotic observations in PMCs were made in eight kinds of interspecific hybrids in Lilium. Three hybrids, ‘6134-S’ × L. cernuum. L. regale × L. leichtlinii maximowiczii and L. auratum platyphyllum × L. henryi showed respectively the mean chromosome association frequencies of 1.4II, 0.8II and 1.5II per cell at MI which were the lowest values hitherto reported in Lilium hybrids. In L. longiflorum × L. henryi, 33.3° of the cells had 1 or 2 multivalents of three to five chromosomes in addition to uni- and bivalents. 12° bivalents and 12 univalents were invariably observed in the triploid hybrid L. longiflorum × L. cernuum. The hybrids between the species belonging to the different sections of the genus generally showed high pollen sterility, with some exceptions. A more or less remote genomic homology was found between the different sections in Lilium.

Genome Relations among Four Diploid Paspalum Species

Paspalum intermedium Munro ex Morong, P. notatum Flugge, P. pumilum Nees, and P. vaginatum Swartz, all 2n = 2x = 20, were crossed to investigate their cytogenetic relationships. All of the hybrids had 2n = 2x = 20 and irregular meiosis. Three P. intermedium x P. notalum hybrids had a mean chromosome association of 18.43 univalents and 0.78 bivalents per cell, indicating that the two species are not closely related. Eight P. notatum x P. vaginatum hybrids had a mean chromosome association of 18.83 univalents and 0.59 bivalents per cell, indicating that the genomes of these two species are different. Seven P. pumilum x P. vaginatum hybrids and two reciprocal hybrids had a mean of 19.46 univalents and 0.27 bivalents per cell and 19.48 univalents and 0.25 bivalents, respectively, indicating that these two species have different genomes. Paspalum intermedium and P. vaginatum have the genome formulas II and DD. The genomic formula NN is proposed for the diploid form of P. notatum. Since P. pumilum and P. notatum are in the same taxonomic group, P. pumilum may also have the formula NN. None of the hybrids produced fertile seed.

Cytogenetic Relationships Between Paspalum jurgensii and P. intermedium, P. vaginatum, and P. setaceum var. ciliatifolium1

Paspalum jurgensii Hackel, 2n = 2x = 20, was crossed with P. intermedium Munro ex Morong, 2n = 2x = 20, P. vaginatum Swartz, 2n = 2x = 20, and P. setaceum Michx. var. ciliatifolium (Michx.) Vasey, 2n = 2x = 20. The hybrids were studied cytologlcally to determine their genomic relationships. Three P. jurgensii ✕ P. intermedium hybrids were produced. Each had a chromosome number of 2n = 2x = 20. Meiosis was irregular with a mean chromosome association of 16.48 univalents and 1.70 bivalents per cell. These findings indicate that P. jurgensii and P. intermedium have different genomes which are designated as JJ and II for the respective species. Twelve P. jurgensii ✕ P. vaginatum hybrids were studied cytologically and all had a chromosome number 2n = 2x = 20. Meiosis was very irregular with a mean chromosome pairing behavior of 18.22 univalents and 0.88 bivalents per ceil. Minimal chromosome pairing in these hybrids indicates the two species are not closely related. One hybrid was recovered from the P. jurgensii ✕ P. setaceum var. ciliatifolium cross. It had 20 chromosomes with a mean pairing behavior of 17.73 univalents and 1.14 bivalents per cell at metaphase I. This indicates these species have different genomes. Based on these and previous findings genomic formulas of DD and SS were proposed for P. vaginatura and P. setaceum var, ciliatifolium.

Cytogenetics of Paspalum urvillei ✕ P. intermedium and P. dilatatum ✕ P. paniculatum Hybrids1

Vaseygrass, Paspalum urvillei Steud., 2n = 4x = 40, was crossed with P. intermedium Munro ex Morong, 2n = 2x = 20, and yellow‐anthered dallisgrass, P. dilatatum Poir, 2n = 4x = 40, was crossed with P. paniculatum L., 2n = 2x = 20. Parents and their hybrids were studied cytologically to determine the genome relationships between these species. Twenty‐five P. urvillei ✕ P. intermedium hybrids were produced; however, the plants were small and only six survived. One reciprocal hybrid was recovered. All hybrids studied cytologically had a chromosome number of 2n = 3x = 30. Meiosis was irregular with a mean chromosome association of 11.13 univalents and 9.45 bivalents per cell for the P. urvillei ✕ P. intermedium hybrids and 10.27 univalents and 9.57 blvalents for the reciprocal hybrid. These findings indicated that P. urvillei and P. intermedium have a common genome with some small structural differences. Since P. intermedium is represented by the II genome formula and the second P. urvillei genome is the J genome, P. urvillei has the basic genome formula of IIJJ, the same as that of yellow‐anthered P. dilatatum, a closely related species. One hybrid was recovered from the yellow‐anthered P. dilatatum ✕ P. paniculatum cross. It had 30 chromosomes with a mean pairing behavior of 10.2 univalents and 9.9 bivalents per cell, which indicated that the two species have a homologous genome. Based on chromosome size, it appears that P. panlculatum has the genome formula of JJ. The relationship of P. paniculatum to other Paspalum species with the JJ genome constitution was discussed. Information obtained concerning genome relations of these species support their taxonomic classifications.

Trisomics and aneuploids of ryegrass

Primary trisomics (2 n + 1 = 15), double trisomics (2 n + 1 + 1 = 16) and aneuploids with 24 to 30 chromosomes, as well as a diploid and tetraploids, were found in the progeny of a hypertriploid (2 n = 22) plant of perennial ryegrass, Lolium perenne L. Trisomics and double trisomics differed in their mean chromosome association, chiasma number and spike morphology. A few aneuploids and tetraploids had reciprocal translocations. The diploid, primary trisomics and tetraploids were more fertile than the double trisomics and aneuploids. Most trisomics and aneuploids were probably produced through female transmission. One double trisomic had a high univalent number, a low chiasma number and loose chromosome coiling. Both the extra chromosomes carried secondary constrictions. The gene for desynapsis might be located on one of these chromosomes.

Chromosome behavior in a hypertriploid plant of ryegrass

A hypertriploid (2n = 22) was detected in the progeny of a desynaptic diploid (2n = 14) plant of perennial ryegrass,Lolium perenne L. The hypertriploid did not differ in morphology from its maternal-sib diploids, but showed larger stomata and pollen. The microsporocytes showed a mean chromosome association of 4.3 I + 3.4 II + 3.2 III + 0.3 IV at metaphase I with a mean chiasma number of 14.4 per cell. The 33 types of observed chromosome configurations could be explained by assuming that the plant was trisomie for 6 chromosomes and tetrasomic for one chromosome. Karyotype analysis confirmed the above assumption and revealed tetrasomy of chromosome number VI. However, two chromosome associations, 1I + 4 II + 3 III + IIV and 2 I + 5 II + 2 III + 1 IV suggested the presence of displaced duplications within the genome of ryegrass. The tetrasomic chromosome formed mostly a quadrivalent, which often broke down to form 2 II or 1 I + 1 III and rarely 2 I + 1 II or 4 univalents. Most of the univalents arose from the trisomic chromosomes and divided precociously at anaphase 1, producing diads and tetrads with unequal chromosome numbers. The plant was highly sterile and set no seed on controlled crossing.

Synthetic Hybrids of Agropyron dasystachyum x Elymus glaucus and Sitanion hystrix

Ten unemasculated Agropyron dasystachyum spikes exposed to Elymus glaucus pollen yielded 95 hybrid seeds, of which more than half were viable. The hybrids were morphologically variable and included caespitose and highly rhizomatous plants. The A. dasystachyum and E. glaucus parents were tetraploid (2n = 28) and behaved meiotically as alloploids. Chromosome associations in 10 of their hybrids averaged 0.31 I, 13.82 II, and 0.01 III in 525 metaphase-I cells. Subsequent meiotic stages were basically regular; yet no hybrid produced more than 30% stainable pollen. The hybrids were partially fertile and yielded up to 250 seeds under open pollination. Forty-seven viable hybrid seeds developed in 120 emasculated A. dasystachyum florets pollinated by Sitanion hystrix, and the reciprocal cross yielded four viable seeds from 48 florets. These hybrids and their reciprocals were vegetatively uniform and intermediate between the parent species. An apparent reciprocal difference was noted in awn development. Both parents were tetraploid and regularly formed 14 bivalents at metaphase I. Almost 90% of the 259 metaphase-I cells of six A. dasystachyum x S. hystrix hybrids and four reciprocals also contained 14 bivalents. Mean chromosome associations of 0.21 I, 13.88 II, and 0.01 III were observed in these hybrids. Stainable pollen ranged from less than 1% to 8% in the A. dasystachyum x S. hystrix hybrids, and none produced more than nine seeds. Ready cross-compatibility and almost complete genome homologies verify the previously suspected genetic and phylogenetic relations among A. dasystachyum, E. glaucus, and S. hystrix. These three species, as well as many other closely related Agropyron, Elymus, and Sitanion species, belong in the same genus.

Chromosome association and fertility in tetraploid ryegrass

In colchicine treated diploid ryegrass,Lolium perenne L. (2n=14), in addition to normal diploids, tetraploids and mixoploids, cytologically aberrant plants were observed. The latter included one aneuploid (2n=25), four translocation heterozygotes (diploid and mixoploid), and one mixoploid desynaptic plant which was a chimera of normal and desynaptic tissues as shown by successive clonal propagations. The normal diploids had chromosome association of 7 II, with an average chiasmata frequency of 12 per pollen mother cell and had 85.6 per cent pollen fertility. The tetraploids showed a mean chromosome association of 0.88 I, 4.83 II, 0.29 III and 3.91 IV, with an average chiasmata frequency of 25 per microsporocyte, and had 83.8 per cent fertile pollen. In the tetraploids, pollen fertility was negatively correlated with quadrivalent frequency and positively so with bivalent and trivalent frequencies, there being no significant correlation with univalent frequency. Among the most frequent types of quadrivalents, only the alternate chain quadrivalent frequency was positively correlated (r=+.9297) with the plllen fertility. On this basis, an increased pollen fertility in the succeeding generations of tetraploids could result from a decrease in the frequency of quadrivalent types other than the alternate chains.

SYNTHETIC HYBRIDS OF NEW WORLD AND OLD WORLD AGROPYRONS. IV. TETRAPLOID AGROPYRON SPICATUM F. INERME X TETRAPLOID AGROPYRON DESERTORUM

Emasculated and unemasculated crosses of tetraploid A. spicatum f. inerme X A. desertorum yielded four hybrids. The hybrids were morphologically intermediate between the parent species but resembled A. desertorum more closely than A. spicatum. Both parents behaved cytologically as autoploids. Mean chromosome associations of 0.04 I, 8.60 II, 0.01 III, and 2.67 IV were observed at diakinesis in the 28‐chromosome A. spicatum. The A. desertorum parent contained 30 chromosomes, 2 of which were likely supernumeraries, and averaged 0.03 I, 9.85 II, and 2.57 IV at diakinesis. Three hybrids contained 30 chromosomes, and one had 29. The most common chromosome association in the 30‐chromosome hybrids was 2 I and 14 II; and the average was 3.00 I, 13.40 II, 0.06 III, and 0.01 IV. A. spicatum and A. desertorum chromosomes were usually distinguishable from each other in the hybrid cell on the basis of size. All pairing in the hybrids was attributed to autosyndesis within parental genomes. A. spicatum, A. desertorum, and their hybrids were represented by genome formulas of SSSS, CCCC, and SSCC, respectively. The hybrids produced 5 to 439 seeds under open pollination. Three controlled crosses between the hybrids yielded 2, 5, and 23 seeds, respectively, on 10 maternal spikes in each cross. The prospects of developing a fertile, cytologically stable allotetraploid species from the hybrids appear favorable.

Synthetic Hybrids of New World and Old World Agropyrons. IV. Tetraploid Agropyron spicatum F. Inerme x Tetraploid Agropyron desertorum

Emasculated and unemasculated crosses of tetraploid A. spicatum f. inerme X A. desertorum yielded four hybrids. The hybrids were morphologically intermediate between the parent species but resembled A. desertorum more closely than A. spicatum. Both parents behaved cytologically as autoploids. Mean chromosome associations of 0.04 I, 8.60 II, 0.01 III, and 2.67 IV were observed at diakinesis in the 28-chromosome A. spicatum. The A. desertorum parent contained 30 chromosomes, 2 of which were likely supernumeraries, and averaged 0.03 I, 9.85 II, and 2.57 IV at diakinesis. Three hybrids contained 30 chromosomes, and one had 29. The most common chromosome association in the 30-chromosome hybrids was 2 I and 14 II; and the average was 3.00 I, 13.40 II, 0.06 III, and 0.01 IV. A. spicatum and A. desertorum chromosomes were usually distinguishable from each other in the hybrid cell on the basis of size. All pairing in the hybrids was attributed to autosyndesis within parental genomes. A. spicatum, A. desertorum, and their hybrids were represented by genome formulas of SSSS, CCCC, and SSCC, respectively. The hybrids produced 5 to 439 seeds under open pollination. Three controlled crosses between the hybrids yielded 2, 5, and 23 seeds, respectively, on 10 maternal spikes in each cross. The prospects of developing a fertile, cytologically stable allotetraploid species from the hybrids appear favorable.

SYNTHETIC HYBRIDS OF NEW WORLD AND OLD WORLD AGROPYRONS: III. AGROPYRON REPENS × TETRAPLOID AGROPYRON SPICATUM

Three hybrids of A. repens, 2n = 42, × A. spicatum, 2n = 28, and two reciprocal hybrids were obtained from emasculated and unemasculated crosses, respectively. The 35‐chromosome hybrids tended to be morphologically intermediate between the parent species but resembled A. repens more closely than A. spicatum. A. repens behaved cytologically as a segmental autoallohexaploid, and A. spicatum acted cytologically as an autotetraploid. Mean chromosome associations of 8.04 I, 12.72 II, 0.41 III, 0.06 IV, and 0.009 V were observed in 116 hybrid cells at metaphase I. Most chromosome pairing in the hybrids was attributed to autosyndesis. A. spicatum, A. repens, and their hybrids were represented by genome formulas of SSSS, R1R1X1X1X2X2, and SSR1X1X2, respectively. Hybrid fertility ranged from 0.02 to 0.69 seeds per spikelet.

Synthetic Hybrids of New World and Old World Agropyrons: III. Agropyron Repens X Tetraploid Agropyron spicatum

Three hybrids of A. repens, 2n = 42, × A. spicatum, 2n = 28, and two reciprocal hybrids were obtained from emasculated and unemasculated crosses, respectively. The 35-chromosome hybrids tended to be morphologically intermediate between the parent species but resembled A. repens more closely than A. spicatum. A. repens behaved cytologically as a segmental autoallohexaploid, and A. spicatum acted cytologically as an autotetraploid. Mean chromosome associations of 8.04 I, 12.72 II, 0.41 III, 0.06 IV, and 0.009 V were observed in 116 hybrid cells at metaphase I. Most chromosome pairing in the hybrids was attributed to autosyndesis. A. spicatum, A. repens, and their hybrids were represented by genome formulas of SSSS, R1R1X1X1X2X2, and SSR1X1X2, respectively. Hybrid fertility ranged from 0.02 to 0.69 seeds per spikelet.

Morphology, Cytology, and Fertility of Diploid and Colchicine‐Induced Tetraploid Crested Wheatgrass1

Treatment of diploid ‘Fairway’ crested wheatgrass, Agropyron cristatum (L.) Gaertn., 2n=14, seeds with colchicine solutions ranging from 0.1 to 0.8% for 12 to 96 hours produced 14 tetraploids. Colchicine solutions of 0.1 to 0.4% applied for 12 to 48 hours appear to be suitable for inducing polyploidy in crested wheatgrass. The induced tetraploids were taller and had fewer but coarser stems than their diploid prototypes. Spike characteristics of the diploids and the tetraploids did not differ significantly. Tetraploid pollen was larger than diploid pollen. Mean chromosome associations of 0.11 I, 7.73 II, 0.06 III, and 3.06 IV where observed in 628 cells at diakinesis and metaphase I. The most common association was 8 II and 3 IV. Irregularities such as laggards, bridges, and unequal chromosome disjunctions occurred in approximately 40% of the anaphase‐I and ‐II cells; however, the induced tetraploids averaged 85.3% stainable pollen. Open‐fertility of the diploids exceeded that of the induced tetraploids, but their mean self‐fertilities did not differ significantly. Induced and natural tetraploids hybridized with relative ease. Many of the hybrids were agronomically desirable and may prove useful in a crested wheatgrass breeding program.

Meiotic Behavior of a Hodo Sorgo ✕ Johnsongrass Hybrid1

Meiosis Of ‘Hodo’ sorgo was slightly more irregular than that reported for other varieties of Sorghum bicolor, notably by having a higher frequency of quadrivalent associations. Meiosis of the fertile interspecific S. bicolor × S. halepense hybrid was almost as regular as that of S. halepense, johnsongrass, and suggests close genomic relationship of the parental species. The 2n chromosome number was 20 for Hodo sorgo, 40 for johnsongrass, and 40 for the interspecific hybrid. The mean chromosome association was: Hodo sorgo, 8.2 II, and .91 IV; jonhsongrass, .19 I, 11.81 II, .09 III, 3.31 IV, .09 V, .28 VI, and .06 VIII; F1 hybrid, .04 I, 7.2 II, .17 III, 4.6 IV, .02 V, .7I VI, .02 VII, and .27 VIII. The frequent occurrence of quadrivalents in the hybrid suggests that part of the parental chromosomes are quite similar if not identical. Although the origin of octoploid (2n=40) johnsongrass is uncertain, experimental evidence substantiates the hypothesis that the species is an autopolyploid derived from S. bicolor.

SYNTHETIC AGROPYRON‐ELYMUS HYBRIDS. I. ELYMUS CANADENSIS × AGROPYRON SUBSECUNDUM

Twenty‐four seeds were formed in 34 hand‐emasculated E. canadensis florets exposed to A. subsecundum pollen. Fifteen of the 24 seeds germinated, and 12 plants were raised to maturity. Each of the 12 plants proved to be a hybrid. The hybrids were morphologically intermediate between the parents, although they favored A. subsecundum in general appearance. The parent plants were meiotically regular and behaved cytologically as strict allotetraploids, 2n = 28. Fourteen bivalents were formed in 12.6% of the hybrid cells interpreted at metaphase I. All other hybrid cells contained various combinations of univalents, bivalents, and multivalents. Mean chromosome associations of 0.22 univalents, 11.70 bivalents, 0.11 trivalente, 0.23 quadrivalents, 0.01 pentavalents, and 0.55 hexavalents were observed in 135 hybrid cells. All hybrid pollen examined was shriveled and non‐staining. Five hybrids produced eight seeds, and seven hybrids were completely sterile. The ready cross‐compatibility of E. canadensis and A. subsecundum and the relatively good chromosome pairing in their hybrids suggest a much closer relationship between the parent species than is implied by the prevailing taxonomic treatment. Structural rearrangements appear to be responsible for the relatively few differences between E. canadensis and A. subsecundum genomes and for the sterility of the hybrids. Cytological data from this and other investigations indicate a close relationship between many self‐fertilizing Agropyron, Elymus, and Sitanion species. It is postulated that this composite of self‐fertilizing Agropyron, Elymus, and Sitanion species originated by hybridization between A. spicatum and an unidentified Hordeum species.

GENOME ANALYSIS OF AGROPYRON REPENS × AGROPYRON CRISTATUM SYNTHETIC HYBRIDS

Hexaploid A. repens, 2n = 42, and diploid A. cristatum, 2n = 14, were hybridized and gave rise to two 28‐chromosome reciprocal hybrids. Approximately 1% of hand‐emasculated florets of both parent species produced viable hybrid seed following controlled pollination. Early embryo abortion prevented greater hybrid seed set on A. repens, whereas failure of fertilization appeared to be the major cause of poor hybrid seed set on A. cristatum. Reciprocal differences in hybrid vegetative and spike morphology were striking. The A. repens × A. cristatum hybrid was vigorous, highly rhizomatous, and bore abundant spikes whose morphology was intermediate between that of the parent species. A. cristatum × A. repens hybrids were weak, non‐rhizomatous with frequently‐malformed spikes. Mean chromosome associations of 0.10 I, 20.10 II, and 0.43 IV were observed in 134 metaphase‐I cells of A. repens. Subsequent meiotic stages were regular except for occasional laggards and bridges at anaphase I and II. Metaphase‐I chromosome associations averaged 0.07 I and 6.97 II in 124 A. cristatum cells. Chromosome pairing in the hybrids was highly variable and averaged 11.45 I, 7.58 II, 0.44 III, and 0.02 IV per cell in 187 cells interpreted. From 5 to 14 laggards appeared in every hybrid cell at anaphase I. Bridges were observed in approximately 25% of the anaphase‐I cells. Similar irregularities were observed at anaphase II. Pollen viability was estimated as 3%, and the hybrids failed to set viable seed. On the basis of chromosome pairing in the species itself and in the hybrids, A. repens was designated as a segmental autoallohexaploid with a genome formula of the type A1A1A2A2BB. Although A. repens and A. cristatum chromosomes paired occasionally, the genomes of the 2 species were essentially non‐homologous. Some of the interpretational difficulties of genome analysis were discussed.