Chromosome counts of Zimbabwean grasses () and an analysis of polyploidy in the grass flora of

G. Davidse, T. Hoshino and B.K. Simon Missouri Botanical Garden, Missouri, U.S.A., Biological Laboratory, Okayama University of Science, Okayama, Japan and Queensland Herbarium, lndooroopilly,

Chromosome numbers and meiotic behaviour are reported Introduction for 64 collections of grasses representing 34 genera and 54 Information about carefully vouchered chromosome numbers species. First chromosome counts are reported for the and behavior is of fundamental importance in broadening our Hylebates and the following 14 species: Agrostis understanding of phylogeny in general (Raven 1975) and continuata, n = 14; mannii, n =7; Eulalia vil/osa, n = 10; Pentaschistis natalensis, n = 14; Oactyloctenium especially polyploidy. Polyploidy is extremely common in the giganteum, n = 5; acraea, n = 10; E. caesia, n = 20; Poaceae, even at the species level (Goldblatt 1980), emphasizing E. nindensis, n =20; E. trichophora, n =20; Hylebates the need for broad-based and intensive sampling. For the rich cordatus, n = 9; Panicum novemnerve, n = 9; P. trichonode Zimbabwean grass flora relatively few cytological data are n = 9; Sacciolepis luciae, n = 18; Setaria orthosticha, n = 9. available. Only one study has specifically been based largely on Counts differing from any previously reported number were obtained for 7 species. Chromosome numbers are now Zimbabwean grasses. In this study Moffett & Hurcombe (1949) known for 22% of the Zimbabwean grasses and 73% of presented 86 counts for 80 species. Additionally, Bourreil et at. these are polyploid. (I972, 1974), Celarier (1957, 1959), Celarier et a/. (1958), S. Afr. J. Bot. 1986, 52: 521-528 Harlan eta/. (1958, 1963, 1964), Li eta/. (1966), Moffett (1944), Phipps & Mahon (1970), Pritchard (1970), Snyder eta/. (1955), Verslag word gedoen oar chromosoomgetalle en meiotiese and Tateoka (I969) have published smaller numbers of counts gedrag van 64 versamelings grasse wat 34 genera en 54 spesies verteenwoordig. Eerste chromosoomtellings word based on Zimbabwean grasses. In our report we present data vir die genus Hylebates verstrek asook vir die volgende 14 for 64 collections representing 55 species and analyze the degree spesies: Agrostis continuata, n = 14; Andropogon mannii, of polyploidy among Zimbabwean grasses. Where we com­ n = 7; Eulalia villosa, n = 10; Pentaschistis natalensis, n = 14; ment on chromosome or base numbers without specific Oactyloctenium giganteum, n =5; Eragrostis acraea, n = 10; reference to original sources, we use information from the E. caesia, n = 20; E. nindensis, n = 20; E. trichophora, n = 20; summary listings of Fedorov (1969), Moore (1973, 1974, 1975) Hylebates cordatus, n =9; Panicum novemnerve, n =9; P. and Goldblatt (1981, 1984, 1985). Taxonomic implications for trichonode n =9 ; Sacciolepis luciae, n =18; Setaria orthosticha, n = 9. Vir 7 spesies is tel lings verkry wat van individual species are discussed where appropriate. aile voriges verskil. Chromosoomgetalle is nou vir 22% Zimbabwegrasse bekend waarvan 73% poliplo'led is. Materials and Methods S.-Afr. Tydskr. Plantk. 1986, 52: 521 -528 Most of the materials used for this study were collected by Davidse, Simon and Pope in February I974, and fixed in the Keywords: Chromosome numbers, cytology, Poaceae, field. However, 12 of the chromosome counts were made on polyploidy, Zimbabwe cultivated at the Missouri Botanical Garden from seeds collected by Crook and Simon from native plants in Zimbab­ we. All counts were made from squashed pollen mother cells. Young inflorescences were fixed in Newcomer's (1953) solution. Anthers were squashed in propio-carmine or aceta­ orcein stain. Photomicrographs and camera Iucida drawings of representative chromosome complements were made for G. Davidse* all preparations. However, illustrations are included in this Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri, report only for new counts or for counts that differ in number 63166 U.S.A, of chromosome behaviour from previous reports. Voucher specimens are deposited at BRI (some), MO and SRGH. T. Hoshino Biological Laboratory, Okayama University of Science, Ridai-cho, Okayama, 700 Japan Results and Discussion A complete list of the species studied, the chromosome B.K. Simon numbers determined, and the voucher specimens is given in Queensland Herbarium, Meiers Road, Indooroopilly, 4068 Table I, where totally new counts differing from any previous Australia (formerly National Herbarium, Harare, Zimbabwe) one for the same taxon are also identified. The classification *To whom correspondence should be addressed used in this report follows the one outlined by Simon (197I) and modified by Bennett (I980). Unless otherwise indicated Accepted 9 1une 1986 in the text, meiosis was regular for all taxa listed in Table I. 522 S.-Afr. Tydskr. Plantk., 1986, 52(6)

Table 1 Chromosome numbers of Zimbabwean Table I Continued grasses Chromosome Chromosome Taxon number Locality and voucher Taxon number Locality and voucher Tribe Danthonieae Tribe Agrostideae Pentaschistis Agrostis natalensis Stapf n 14• lnyanga: Mt. Inyangani, Davidse, continuata Stapf n 14• lnyanga: Mt. lnyangani, Davidse, Simon & Pope 6558. Simon & Pope 6559. Tribe Ehrharteae Tribe Ehrharta Andropogon erecta Lam. n 12 lnyanga: 66 km NE of Rusape, eucomus Nees n = 10 Marandellas: W. of Marandellas, Davidse, Simon & Pope 6512. Davidse 73 - 2628 c. Tribe huillensis Rendle n = 20b Marandellas: W. of Marandellas, Bewsia Davidse 73 - 2612 c. bijlora (Hack.) 2n c. 45 lnyanga: above lnyangombe Falls, mannii Hook. f. n = 7• Inyanga: Mt. lnyangani, Davidse, Goossens Davidse 73-2638 <. Simon & Pope 6561 . Dactyloctenium Cymbopogon giganteum Fisch. n 5• Umtali: 49 km S. of Umtali, caesius (Hook. n = 10 Lomagundi: 64 km NW of & Schweick. Davidse, Simon & Pope 6619. & Arn.) Stapf Salisbury on the Great Dyke, Eleusine Davidse, Simon, Drummond & coracana (L.) n 18 Umtali: 26 km S. of Umtali, Bennett 6476. Gaertn. Davidse, Simon & Pope 6617. Salisbury: Marimba Vlei, Salisbury, Eragrostis Davidse & Simon 6449. acraea de Winter n 10• lnyanga: lnyanga National Park, nardus (L.) n 10 lnyanga: lnyanga National Park, Davidse, Simon & Pope 6586. Rendle Mtarazi Falls, Davidse, Simon & caesia Stapf n = 20• lnyanga: Mt. Inyangani, Davidse, Pope 6590. Simon & Pope 6566. Eulalia capensis (Thunb.) n = 20 Makoni: Rusape, Davidse 10. villosa (Thunb.) n lnyanga: Mt. lnyangani, Davidse, Trin. 73 - 2617 <. Nees Simon & Pope 6560. curvula n = c. 25 Inyanga: lnyanga National Park, Hemarthria (Schrad.) Nees Davidse, Simon & Pope 6536. altissima (Poiret) n 18 Salisbury: University campus, nindensis n = 20• Lomagundi: Rod Camp Mine Road Stapf & C.E. Hubb Davidse 73 - 263 7 c. Ficalho & Hiern. in the Great Dyke, Davidse, Simon, Sorghum Drummond & Bennett 6481 . versicolor n 5 Umtali: Penhalonga, Davidse sclerantha Nees n = 20 Inyanga: Mt. lnyangani, Davidse, Anderss. 73 - 2634 <. Simon & Pope 6576. trichophora n = 20• Melsetter: 78 km S. of Umtali, Tribe Aristideae Coss. & Our. Davidse, Simon & Pope 6622. Aristida Pogonarthria congesta Roem. n = 11 Melsetter: 78 km S. of Umtali, squarrosa n c. 20 Bikita: 150 km SW of Umtali, & Schult. Davidse, Simon & Pope 6626. (Licht. ex Davidse, Simon & Pope 6631. rhiniochloa n = 11 Melsetter: 78 km S. of Umtali, R. & S.) Pilger Hochst. Davidse, Simon & Pope 6624. Tribe Tribe Arundinelleae Brachiaria Danthoniopsis erucijormis (J.E. n = 9 Salisbury: Marimba Vlei, Salisbury, pruinosa C.E. Makoni: 15 km N. of Rusape, Smith) Griseb. Davidse & Simon 6451. Hubb. Davidse, Simon & Pope 6506. humidicola lnyanga: Inyanga National Park, n = 24 Victoria: Kyle National Park, (Rendle) Schweick. Davidse, Simon & Pope 6535. Davidse, Simon & Pope 6659. Digitaria gazensis Rendle n 9 lnyanga: c. 66 km NE of Rusape, jlavida (Stapf) n 12 Lomagundi: Rodcamp Mine Road Davidse, Simon & Pope 6515; C.E. Hubb. in the Great Dyke, Davidse, Simon, lnyanga National Park, Davidse, Drummond & Bennett 6479. Simon & Pope 6537. simplex (Nees) n 30 1nyanga: c. 66 km NE of Rusape, velutina (Forsk.) n 9 lnyanga: Inyangombe Falls, C.E. Hubb. Davidse, Simon & Pope 6518. Beauv. Davidse, Simon & Pope 6529. Hylebates nodiglumis n = 24 Salisbury: Twenty Dales Estate, cordatus n 9• Umtali: main road to K. Schum. Davidse & Simon 6460. Chippindall border, Davidse 73-2633 <. Tribe Bromeae Oplismenus Bromus compositus (L.) n 27 Victoria: above the Glenlivet Hotel, leptocladus Nees n 21 lnyanga: Inyanga National Park, Beauv. Davidse, Simon & Pope 6655. Davidse, Simon & Pope 6540. n 36 Umtali: Murahwa Hill Reserve at Umtali, Davidse, Simon & Pope Tribe CWorideae 6597. Chloris Panicum gayana Kunth n 10 Wankie: Kazuma depression, maximum Jacq. n 16 Salisbury: Salisbury, Davidse & Matetsi, Davidse 73-2618 d' Pope 6690. Davidse 73 - 2626 d. Makoni: 15 km N. of Rusape, S. Afr. J. Bot., 1986, 52(6) 523

Table 1 Continued Table 1 Continued

Chromosome Chromosome Taxon number Locality and voucher Taxon number Locality and voucher

Davidse, Simon & Pope 6499. Tribe Zoysieae Umtali: Murahwa Hill Reserve at Perot is Umtali, Davidse, Simon & Pope patens n 20 Umtali: 49 km S. of Umtali, 6599. Gandoger Davidse, Simon & Pope 6621. 9a novemnerve n Lomagundi: 61 km NW of "First chromosome count for the species Stapf Salisbury on the Great Dyke, bChromosome count differing from any previous one for this species Davidse, Simon, Drummond & cFrom cultivated plants at MO, the seed collected by A.O. Crook Bennett 6473. ct From cultivated plants at MO, the seed collected by B.K. Simon subalbidum n 18 Victoria: Kyle National Park, Kunth Davidse, Simon & Pope 6667. trichonode n 9" Salisbury: Marimba Vlei, Davidse & Launert & Simon 6450. Tribe Agrostideae Renvoize The chromosome number n = 14 is the first report for Agros­ Rhynchelytrum tis continuata (Figure 1). The large size of the chromosomes nerviglume n 18 lnyanga: c. 66 km NE of Rusape, and the base number x = 7 are characteristic of the genus (Franch.) Chiov. Davidse, Simon & Pope 6517. Agrostis, the tribe Agrostideae, and the subfamily Pooideae. Sacciolepis luciae n 18" lnyanga: lnyanga National Park, Tribe Andropogoneae B.K. Simon Davidse, Simon & Pope 6552. Andropogon mannii, n = 7 (Figure 2) , was previously cyto­ Setaria logically unknown. This species is distributed from Sierra homonyma n = 18 Salisbury: Twenty Dales Estate, Leone to the Sudan and southwards to South (Clayton (Steud.) Chiov. Davidse & Simon 6461. megaphylla n = c. 27b Victoria: forested hills just above & Renvoize 1982). Our plant was collected in the lnyangani (Steud.) Our. & the Glenlivet Hotel, Davidse, Mountains at an elevation of 2 000 m. The number n = 7 Schinz Simon & Pope 6654. is the first for the genus Andropogon. orthosticha K. n 9" Lomagundi: Umvukwo Range, Andropogon is fundamentally a genus of tropical and Schum. ex I km E. of Mtoroshanga Pass, subtropical lowlands. However, one of the four subgenera, Herrm. Davidse, Simon, Drummond & subgenus Andropogon, occupies upland areas in tropical and Bennett 6488. subtropical mountains in Africa and and temperate pumila (Poir.) n 18 Makoni: 15 km N. of Rusape, regions in North America (A. gerardit). Andropogon mannii Roem. & Schult. Davidse, Simon, Drummond & is part of the African montane complex of subgenus Andropo­ Benne// 6507. gon. The other subgenera, Leptopogon, Piestium and Nato­ Umtali: M~romo Tribal Trust solen, have chromosome numbers based on x = 10. Poly­ Land, Davidse 73- 262fl". ploidy is common and a low incidence of aneuploidy has also Stereochlaena been reported. Only in subgenus Andropogon are significant 18b cameronii n Salisbury: Twenty Dales Estate, departures from the base number x = 10 known, namely, (Stapf) Pilger Davidse & Simon 6454. x = 9 in A. lima and A. distachyos (Celarier 1956; Gould Tricholaena 1956; Hedberg & Hedberg 1977), [Saura (1943) also reported monachne n 18 Umtali: 49 km S. of Umtali, 2n = 40, probably mistakenly according to Celarier, for the (Trin.) Stapf & Davidse, Simon & Pope 6620. C.E. Hubb. latter species], x = 8 in A. abyssinicus (Gould 1956), and Urochloa now x = 7 in A. mannii. It seems likely that subgenus Andro­ mosambicensis n 7 Melsetter: 7- 8 km S. of Umtali, pogon is a derived specialized group in Andropogon and that (Hack.) Dandy Davidse, Simon & Pope 6623. this specialization is cytologically manifested in reduced base numbers derived from x = 10. Tribe Pappophoreae Although x = 7 is the characteristic number for the sub­ Enneapogon family Pooideae, this mere numerical coincidence does not 12b cenchroides n Umtali: Mutambara Tribal Trust indicate a phylogenetic relationship between A. mannii and (Roem. & Lands, Davidse 73-2609 c. the Pooideae. Morphologically A. mannii conforms with the Schult.) C.E. Hubb. Andropogoneae, and its chromosomes are medium-sized in Tribe Poeae contrast to the large-sized chromosomes of the Pooideae. Festuca Compare Figure 2 to Figure I. abyssinica n 14 lnyanga: lnyanga National Park, Andropogon huillensis has previously been reported to be A. Rich . Davidse, Simon & Pope 6538. hexaploid, 2n = 60, from by De Wet (1960) and diploid, n = 10, from western Zaire by Dujardin (1978). Tribe Sporoboleae We found a Zimbabwean collection to be tetraploid, n = 20 Sporobolus (Figure 3). Although these data suggest a geographical separa­ panicoides n lOb Umtali: Chitsanga, Davidse tion of polyploid races, much more widespread sampling is A Rich. 73-2611c; Mutambara Tribal Trust Land, Davidse 73-2610 c. needed to establish the true geographical distribution of the Bikita: 150 km SW of Umtali, diploids and polyploids. Davidse, Simon & Pope 6634. Our determination of n = 10 for Eulalia villosa (Figure 4) pi/iferus (Trin.) n 18 Salisbury: Twenty Dales Estate, is the first for this species. Previous counts in the genus have Kunth Davidse & Simon 6455. been almost evenly divided between the base numbers x = 9 524 S.-Afr. Tydskr. Plantk. , 1986, 52(6)

and x = 10, the two most common base numbers in the E. curvula, and E. trichophora. In E. capensis (Figure 11) Andropogoneae. the following chromosome associations were observed in about equal frequency in five greenhouse plants: 2rv + 16u; Tribe Arundinelleae 1rv + 18n; 20n; 19n + 2r. The pollen fertility in these plants Cytologically, the small primarily African tribe Arundinelleae as measured by aniline blue staining ranged from 66-75%. is complex with base numbers x = 7, 8, 9, 10, 11, 12 having In E. curvula (Figure 12) we observed multivalents and univa­ been reported (Li eta/. 1966; Phipps & Mahon 1970), and lents in complex associations that made an exact count dif­ several genera have more than one base number. We deter­ ficult. Fragments and bridges were commonly observed at mined Danthoniopsis pruinosa to haven = 12 (Figure 5) in anaphase I. Our count of n = c. 25 agrees with previous one population from the Makoni District and n = 24 (Figure determinations of 2n = 50 for this species by De Wet (1954) 22) in another population from the Victoria District. The and Vorster & Liebenberg (1977). The latter investigators have tetraploid count agrees with the number reported by Li et a!. shown that this species is composed of diploid to hexaploid (1966) for a population from the Transvaal. This contrasts races with reproduction varying from obligate and facultative with the earlier report of 2n = 36 from an unspecific South apomixis to normal sexuality. Meiotic irregularities similar to African locality by De Wet (1960) and suggests the possibility those that we observed were common in the polyploids. It that De Wet's plant may have been a triploid hybrid based seems likely that the plant we studied represents one of the on x = 12. polyploid apomictics. In E. trichophora we commonly observed two quadrivalents (Figure 10) at diakinesis. Tribe Danthonieae Our count of n = c. 20 for con­ Tateoka (1965) has noted that Pentaschistis has two base firms the earlier reports by De Wet & Anderson (1956) and numbers, x = 13 for the Afroalpine species and x = 7 for Moffett & Hurcombe (1949), the latter authors also reporting the South African species. This interpretation has been sup­ the aneuploid number 2n = 42. The uncertainty in our counts ported by the subsequent data of Hedberg & Hedberg (1977) reflect our difficulty in the interpretation of one multivalent. for additional Afroalpine species. Our new count of n = 14 for Pentaschistis natalensis (Figure 6) suggests that the geo­ Tribe Paniceae graphical distinction is not as sharp as originally indicated since Brachiaria humidicola was previously reported from South P. natalensis ranges from to the Transvaal and Natal Africa as octoploid, 2n = 72, by De Wet & Anderson (1956). (Clayton 1970). It seems likely that x = 7 is the primitive The collection we examined was diploid with n = 18 (Figure number and that x = 13 is a secondarily derived number 13). This species is morphologically variable and once addi­ through aneuploid reduction from n = 14. tional studies have been completed, it may be possible to correlate morphological variation with ploidy levels. Tribe Eragrostideae No chromosome number has previously been reported for In the monotypic genus Bewsia, De Wet & Anderson (1956) the ditypic genus Hylebates. Our collection of Hylebates reported 2n = 30 for B. biflora based on root tip counts from cordatus from Umtali had n = 9 (Figure 14). The basic a South African source. We determined a Zimbabwean collec­ chromosome number (x = 9) of this genus is also the most tion to have 2n = c. 45 with irregular meiosis (Figure 23). freque'nt basic number in the tribe Paniceae. We experienced some uncertainty in interpreting chromosome Numerous previous reports have established that Oplismenus associations. Simply considering all of the associations to be compositus is cytologically complex, populations of this species single units, the number of observed units ranged from 22-25. ranging from the diploid to the octoploid level. The two However, we believe that most of the associations can be collections we examined were hexaploid and octoploid. In our interpreted as trivalents, bivalents or univalents. Following this hexaploid collection meiosis was irregular with the following interpretation, we observed the following associations in about association: 1m + 19n + 13, (Figure 15). At anaphase I chro­ equal frequency: 1m+ 21n; 1m+ 20n + 2r; 22n + Ir; 21n mosome fragments and lagging chromosomes were common. + 3,; 20rr + 5r. Pollen fertility in this collection was measured In three collections of Panicum maximum we found a by staining with aniline blue arid was determined to be 89,50Jo. chromosome number of n = 16, the same as that reported Good seed set was obtained in the greenhouse. These lines of by numerous previous investigators. Meiosis was regular evidence suggests that this species is an apomictic polyploid except in Davidse & Pope 6690 where up to four quadrivalents since these characteristics are typical of grasses in which were observed (Figure 17). apomixis has been demonstrated by embryo sac analysis, as in The counts of n = 9 for Panicum novemnerve (Figure 16), Eragrostis curvula (Vorster & Liebenberg 1977), for example. n = 9 for P. trichonode, n = 18 for Sacciolepis luciae (Figure The count of n = 5 (Figure 24) for Dactyloctenium gigan­ 18), and n = 9 for Setaria orthosticha (Figure 25) are the first teum is the frrst report for this species. It is the lowest chromo­ reports for these species, and the base number x = 9 is the some number now known for the genus and suggests that same as previously established for these genera. this is the basic number of the genus. Previously the genus Our count of n = c. 27 for Setaria megaphylla (Figure 19) is was known cytologically from three species and a basic number the first hexaploid record for this species. Previous reports have of x = 10 seemed most probable, although the numerous all been tetraploid. We found n = 18 in Stereochlena cameronii counts for D. aegyptium (L.) Willd. are almost evenly divided (Figure 20) for which Moffett & Hurcombe (1949) previously between those based on x = 9 and x = 10, in addition to reported a hexaploid count 2n = 54 (as Chloridion cameroni1). a number of aneuploid counts. The chromosome numbers of four Eragrostis species were Tribe Pappophorae determined for the first time: E. acraea, n = 10 (Figure 7); The chromosome number of Enneapogon cenchroides was E. ceasia, n = 20 (Figure 8); E. nindensis, n = 20 (Figure 9); previously reported as n = 20 by Reeder & Singh (1968). These E. trichophora, n = 20 (Figure 10). All these numbers are authors also considered the earlier reports of 2n = 36 by De in accordance with the well-established basic number (x = 10) Wet & Anderson (1956) to be in error. Nevertheless, our for the genus. Meiotic irregularities were found in E. capensis, collection had n = 12 (Figure 26). Two or four univalents were S. Afr. J. Bot., 1986, 52(6) 525

... ~ ,..~-~ ~ ~-- ..,, ~ •• - c;o •• ...... • " • cSI '-~ ., .....:,.._ •• 4 ol-•o ,. .. y I • 1 ~ 2 ...... 3 . ,.. \ --??* •• .o •\I• •••• ~ :. ··11, •• • .·~·· .. •~ • • •7 ••• 8 J 6 ._ I _ IV • ~· ..._ '- .. .. ,... • . IV ~~· J .,.,, I. I. • • -.• .. I I.. • ov//;0\ , I VI IV~ uv ~ • IV .. '", I I .. IV 10 I - ' ~" 12 .. 11 .~ ....o-· . .., ' ••••• ,... 13 : # . . 16

1------l

Figures 1-21 Camera Iucida drawings of meiotic chromosomes of Zimbabwean grasses. (1) Agrostis continuata, n = 14, diakinesis; (2) Andropogon mannii, n = 7, metaphase II; (3) A. huillensis, n = 20, metaphase I; (4) Eulalia villosa, n = 10, diakinesis; (5) Danthoniopus pruinosa, n = 12, metaphase II; (6) Pentaschistis natalensis, n = 14, metaphase I; (7) Eragrosti:; acraea, n = 10, diakinesis; (8) E. caesia, n = 20, anaphase I; (9) E. nindensis, n = 20, diakinesis; (10) E. trichophora, 21v + 16u, diakinesis; (11) E. capensis, 21v + 16u, diakinesis; (12) E. curvula, lv1 + lv + 3Iv + 21u + 31, diakinesis; (13) Brachiaria humidico/a, n = 18, diakinesis; (14) Hylebates cordatus, n = 9, diakinesis; (15) Oplismenus compositus, lm + 19n + 131, diakinesis; (16) P. novemnerve, n = 9, single complement of metaphase II; (17) Panicum maximum, 4Iv + 8u, diakinesis; (18) Sacciolepis /uciae, n = 18, metaphase I; (19) Setaria megaphyl!a, n = c. 27, diakinesis; (20) Stereoch/aena cameronii, n = 18, diakinesis; (21) Perotis patens, 2m + 17n, metaphase I. Scale line = 10 ~m. 526 S.-Afr. Tydskr. Plantk., 1986, 52(6)

commonly observed in most complements although the in this species, it seems likely that aneuploidy is a factor in the remaining complements showed complete pairing, even at evolution of this species, suggesting that all previous reports metaphase I. Thus, in addition to documenting a diploid race may be correct.

____ 23 • 22

24 25

• • ~.- , 26 • " 27

Figures 22- 27 Photomicrographs of meiotic chromosomes of Zimbabwean grasses. (22) Danthoniopsis pruinosa, n = 24, late diakinesis; (23) Bewsia bif/ora, lm + 20u + 2~, diakinesis; (24) Dactyloctenium giganteum, n = 5, metaphase I; (25) Setaria orthosticha, n = 9, diakinesis; (26) Enneapogon cenchroides, lOu + h metaphase I; (27) Sporobolus panicoides, n = 10, metaphase I. Scale" lines = IO~m .. S. Afr. J. Bot., 1986, 52(6) 527

Tribe Sporoboleae CELARIER, R.P. 1959. Cytotaxonomy of the Andropogoneae. Three of our collections of Sporobolus panicoides uniformly III. Subtribe Sorgheae genus Sorghum. Cytologia 23: 395 - 418. CELARIER, R.P., MEHRA, K.L. & WULF, M.L. 1958. had n = 10 with regular pairing and no meiotic irregularities Cytogeography of the Dichanthium annulatum complex. (Figure 27). Moffett & Hurcombe (1949) had previously Brillonia 10: 59 - 72. reported a count of 2n = 24 for this species from Zimbabwe. CLAYTON, W.D. 1970. Gramineae (Part 1). Flora of tropical Although this may indicate aneuploidy in this species, it seems . Crown Agents for Oversea Governments and more likely that the collection was misidentified. Unfortunately Administrations, London. CLAYTON, W.O. & RENVOIZE, S.A. 1982. Gramineae (Part 3). we have not been able to check a voucher. The genus is Flora of tropical East Africa. Balkema, Rotterdam. cytologically complex with basic chromosome numbers of DAVIDSE, G. & POHL, R.W. 1974. Chromosome numbers, x = 6, 9, and 10 having been documented. meiotic behavior, and notes on tropical American grasses (Gramineae). Can. J. Bot. 52: 317 - 328. Tribe Zoysieae DE WET, J.M.J. 1954. Chromosome numbers of a few South African grasses. Cytologia 19: 97- 103. Our count of n = 20 (Figure 21) for Perotis patens agrees with DE WET, J.M.J. 1960. Chromosome numbers and some a number of other reports although De Wet & Anderson (1956) morphological attributes of various South African grasses. Am. reported 2n = 36. We observed two trivalents in the four cells J. Bot. 47: 44-50. with adequately spread chromosomes. DE WET, J.M.J. & ANDERSON, L.J. 1956. Chromosome numbers in Transvaal grasses. Cytologia 21: I - 10. Conclusions DUJARDIN, M. 1978. Chromosome numbers of some tropical A.frican grasses from western Zaire. Can. J. Bot. 56: Primarily following Bennett (1980), the known Zimbabwean 2138 - 2152. grass flora consists of 522 species and 149 genera. Including the FEDOROV, A.A. 1969. Chromosome numbers of flowering present report, 117 (220Jo) species and 55 (37%) genera are plants. V .L. Komarov Botanical Institute, Leningrad. now known chromosomally from at least one count of Zim­ GOLDBLATT, P. 1980. Polyploidy in angiosperms: babwean origin. . In: Polyploidy: biological relevance, Analyzing the occurrence of polyploidy in the chromo­ ed. Lewis, W.H. pp. 219 - 239, Plenum Pub!. Co., New York. somally known Zimbabwean grasses, and utilizing all known GOLDBLATT, P. 1981. Index to plant chromosome numbers counts of a species for this purpose, 45% of the species show 1975 - 1978. Monogr. Syst. Bot. 5: 1- 553. infraspecific polyploidy. Examining this further, 27% of the GOLDBLATT, P. 1984. Index to plant chromosome numbers species are known only as diploids, 31% as both diploids and 1979-1981. Monogr. Syst. Bot. 8: 1-427. polyploids, and 42% only as polyploids. Thus 73% of Zim­ GOLDBLATT, P. 1985. Index to plant chromosome numbers babwean grasses are polyploid in some part of their complete 1982-1983. Monogr. Syst. Bot. 13: 1- 224. GOULD, F.W. 1956. Chromosome counts and cytotaxonomic distributional range. This compares to estimates of 64% for notes on grasses of the tribe Andropogoneae. Am. J. Bot. 43: tropical American grasses (Davidse & Pohl 1974), 70% for 395-404. Texas grasses (Gould 1968), and ranging between 64%-70% GOULD, F.W. 1968. Chromosome numbers of Texas grasses . for the entire family (Stebbins 1956; Goldblatt 1980). Our still Can. J. Bot. 46:1315 - 1325. relatively low level of cytological knowledge of Zimbabwean HARLAN, J.R., CELARIER, R.P., RICHARDSON, W.L., grasses and the high incidence and high degree of variability BROOKS, M.H. & MEHRA, K. 1958. Studies on old world bluestems. II. Tech. Bull. Okla. agric. Exp. Stn. T-72. of polyploidy in the Poaceae emphasizes the need for much HARLAN, J.R., RICHARDSON, W.L. & DE WET, J.M.J. additional cytological data from Zimbabwe, as well as other 1963. Improving old world bluestems for the South. Progress areas of the world. Report 1962. Okla. Agric. Exp. Sin. Mimeographed. HARLAN, J.R., RICHARDSON, W.L. & DE WET, J.M.J. Acknowledgements 1964. Improving old world bluestems for the South. Progress Field work by Davidse was made possible by NSF grant GB Report 1963. Okla. Agric. Exp. Stn. Mimeographed. 40630. We are grateful to Mrs A.O. Crook for providing us HEDBERG, I. & HEDBERG, 0 . 1977. Chromosome numbers of with seeds of Zimbabwean grasses. G. Pope assisted us with afroalpine and afromontane angiosperms. Bot. Notiser 130: 1-24. the field work and we appreciate his help. We thank the LI, Y.H., LUBKE, R.A. & PHIPPS, J.B. 1966. Studies in the director and staff of SRGH for making their facilities available Arundinelleae (Gramineae). IV. Chromosome numbers of 23 for the collecting and shipping of the specimens. We appreci­ species. Can. J. Bot. 44: 387-393. ate the determination of two voucher specimens by S. Renvoize. MOFFETT, A.A. 1944. Note on the cytology of Rhodes grass. Hoshino's participation in this research in St. Louis was made Rhodesia agric. J. 44: II - 13. possible through the Foundation of Private School Personnel MOFFETT, A.A. & HURCOMBE, R. 1949. Chromosome numbers of South African grasses. Heredity 3: 369-373. (Japan) and Dr Peter H. Raven. MOORE, R.J. 1973. Index to plant chromosome numbers 1967- 1971. Regnum veg. 90: I - 539. References MOORE, R.J. 1974. Index to plant chromosome numbers 1972. BENNETT, K.E. 1980. Keys to Zimbabwean grass species. Kirkia Regnum veg. 90: I - 108. 11: 169-286. MOORE, R.J. 1975. Index to plant chromosome numbers BOURREIL, P., GESLOT, A., GORLIER, M. & DE WINTER, 1973- 1974. Regnum veg. 96: I - 257. B. 1972. Contribution to the caryological study of the African NEWCOMER, E.H. 1953. A new cytological and histological grass Aristida rhiniochloa Hochst. based on specimens from the fixing fluid. Science 118: 161. Southern Hemisphere. Bothalia 10: 555-563. PHIPPS, J.B. & MAHON, J.B. 1970. Studies in the BOURREIL, P., RICHARD, M.L. & GHIGLIONE, C. 1974. In: Arundinelleae (Gramineae). IX. Chromosome numbers of 21 IOPB chromosome number reports. Taxon 23: 809. species. Can. J. Bot. 48: 1419- 1423. CELARIER, R.P. 1956. Cytology of Andropogon distachyus L. PRITCHARD, H.J. 1970. Meiosis and embryo sac development in Bull. Torrey bot. Club. 83: 183- 191. Urochloa mosambicensis and three Paspalum species. Austr. J. CELARIER, R.P. 1957. Cytotaxonomy of the Andropogoneae. II. agric. Res. 21: 649- 652. Subtribes Ischaeminae, Rottboelliinae, and the Maydeae. RAVEN, P.H. 1975. The bases of angiosperm phylogeny: Cytologia 22: 160-183. cytology. Ann. Mo. bot. Gdn. 62: 724- 764. 528 S.-Afr. Tydskr. Plantk., 1986, 52(6)

REEDER, J.R. & SINGH, D.N. 1968. Chromosome numbers in STEBBINS, G.L. 1956. Cytogenetics and evolution of the grass the tribe Pappophoreae (Gramineae). Madrono 19: 183 - 186. family. Am. J. Bot. 43: 890 - 905. SAURA, F. 1943. Cariologia de gramineas. Univ. Nac. Buenos TATEOKA, T. 1965. Chromosome numbers of some grasses from Aires Fac. Agron. Vet. 10: 344 - 353 . . Bot. Mag., Tokyo 78: 306 - 311. SIMON, B.K. 1971. Rhodesian and Zambian grass lists. Kirkia 8: TATEOKA, T. 1969. Cytological evidence for the affinities of the 3 - 83 . genus Streblochaete. Bull. natn. Sci. Mus., Tokyo 12: SNYDER, L.A., HERNANDEZ, A.R. & WARMKE, H.E. 1955. 161 - 163. The mechanism of apomixis in ciliare. Bot. Gaz. VORSTER, T.B. & LIESENBERG, H. 1977. Cytogenetic studies 116: 209 - 221. in the Eragrostis curvula complex. Bothalia 12: 215 - 221.