Cvtologia 43: 643-653, 1978

Cytological Studies in Cyperaceae with Special Reference to its Taxonomy ll

S. P. Rath and S. N. Patnaik CytogeneticsLaboratory, P. G. Departmentof Botany,Utkal University, Bhubaneswar,India. 751004 ReceivedApril 25, 1977

During the recent years a few workers have reported chromosome numbers of various members of the family Cyperaceae from different regions of India. Notable among them are Sanyal (1972), Sanyal and Sharma (1972). Rath and Patnaik (1974a and 1974b), Nijalingappa (1975) and Mehra and Sachdev (1975). A survey of all these reports reveals that not only most of the genera are characterised by series of aneuploid numbers but there is great disparity in the findings of different workers in respect of chromosome numbers for the same species. Moreover certain genera have not been cytologically worked out. This necessitates further study in the family in a more broad based plan. As a continuation of our previous report (Rath and Patnaik 1974a) of chromo some numbers in 24 species of the family, meiotic studies of another 45 species are presented here in the light of which certain taxonomic problems are discussed.

Materials and methods Plants were collected from various parts of the State of Orissa and fixings col lected in the field were generally worked out immediately as even 2-3 days of storing in the usual procedure of fixing in acetic-alcohol (1:3) and transfering to 70% ethyl alcohol induce excessive staining of the cytoplasm. Addition of chloro form to acetic acid and ethyl alcohol mixture though proved slightly better it did not very much improve the staining quality. Suitable cells from freshly prepared slides were immediately photographed as very often the cells become darkened in the process of making the slides permanent. All the negatives of the photomicrographs and the voucher specimens from which the fixings were collected are preserved in the P. G. Department of Botany, Utkal University.

Results and discussion Meiotic and pollen mitotic analysis of 45 species belonging to 12 genera could be done during the present investigation. These findings along with the previous reports are recorded in Table 1. Of these, chromosome numbers of 14 species (*) are reported here for the first time and discrepant chromosome numbers are recorded for 14 species (**). Larger number of species were, however, studied from Cyperus and Fimbristylis and only one species from each of the genera 644 S. P. Rath and S. N. Patnaik Cytologia 43

Bulbostylis, Courtoisia, Lipocarpha and Rhynchospora. Chromosome numbers of 14 species of Cyperus recorded here are found to be in the same aneuploid series as recorded earlier by various workers. Range of chromosome numbers was found to be from n=13 in C. haspan to n=76 in C. rotundus with various intermediate numbers. Cyperus cephalotes, a very distinct species within the genus and kept separate in a single subgenus Anosporum (Clarke in Hooker 1893) was found to be charac terised by n=21 (Fig. 1). C. platystylis, which grows sympatrically with C. cephalotes as the tank floater but morphologically very distinct from it, was found to possess the haploid chromosome number of n=20. Another species C. castaneus though grouped with C. platystylis under the same subsgenus it is a much smaller plant and grows in completely different habitat. This plant was found to possess higher chromosome number of n=49. Cyperus tenuispica, C. pulcherrimus, C. haspan and C. difformis are grouped under the same section of the subgenus Pycnostachys (Clarke in Hooker l.c.). Of these while C. difformis is distinguished out by its capitate spike and very small spikelets, the other three species forms a complex group. Though in the nature of the spikelet they are closely allied, C. tenuispica is differentiated by the large sized plants. Cytologically this species is also distinct from other two in possessing the haploid chromosome number of n=8 (Rath and Patnaik 1974a) which is the lowest chromosome number yet reported for the genus. Cyperus haspan and C. pulcher rimus were both found to be characterised by n=13 (Fig. 4) during the present study. In this connection the finding of n=27 for C. haspan by Mehra and Sachdev (1975) and 2n=16 by Sanyal (1972) are significant and these discrepancies might also be due to the complexity in the morphological characters baffling the correct identifi cation. The other two species of the same subgenus cytologically studied were C. arenarius and C. diffusus. Both of them were, however, morphologically distinct and are plants of specialized habitats, the former a seacoast dweller whereas the latter a hill plant. Two species C, compressus and C. iria are grouped under the same section of the subgenus Choristachys of Clarke (in Hooker l. c.) although morphologically they are very distinct. Both the species were cytologically analysed and exact chromo some numbers could not be determined in them due to extremely small size of the pollen mother cells. But they were found to differ widely in their chromosome numbers. There is also great discrepancies in the findings of chromosome numbers for these two species by previous workers. Cyperus distans and C. nutans are morphologically two very allied species and often it is difficult to differentiate them. As Clarke (Hooker l.c.) remarked it is dif ficult to draw a line between this species (C. nutans) and fine examples of C. distans, because the minor difference in the nature of spikelet fails to be the character of practical value. Different forms collected during the present study though initially assigned to the two species, on careful scrutiny all were identified to C. nutans. Number of plants were also cytologically studied all of which were found to be uniformly characterised by n=28 (Fig. 3) as contrast to the finding of 2n=48 for C. distans (Sanyal 1972). 1978 Cytological Studies in Cyperaceae with Special Reference to its Taxonomy II 645

Cyperus michelianus, a species treated under Juncellus and four species of Pycreus studied here, all show same aneuploid variation like the species of Cyperus, though in general these species are characterised by the possession of higher chromo-

Figs. 1-6. 1, Cyperus cephalotes, n=21 in diakinesis. 2, C. michelianus, n=39 in diakinesis.

3, C. nutans, n=28 in diakinesis. 4, C. haspan, n=13 in diakinesis. 5, Mariscus compactus, n=43 in diakinesis. 6, Pycreus globosus, n=40 in diakinesis. All •~2500.

some numbers (Figs. 2 and 6). This trend, however, is more evident in the members of Mariscus (Fig. 5) and Kyllinga where very high and varied chromosome numbers are seen. 646 S. P. Rath and S. N. Patnaik Cytologia 43

Table 1. 1978 CytologicalStudies in Cyperaceaewith SpecialReference to its Taxonomy II 647

Courtoisia is represented in India by a single species and on morphological grounds is considered to be very close to Mariscus. But Courtoisia cyperoides with haploid chromosome number of n=5 was found to stand out distinctly from all other members of Cyperus group. Species of Eleocharis show considerable variation in chromosome nubmers . E. retroflexa ssp. chaetaria with n=5 (Rath and Patnaik 1974a) and E. atropurpurea with n=10 (Fig. 8) and E. geniculata with n=15 (Fig. 9) studied presently show a distinct euploid series. Eleocharis ovata (2n=10), E. afflata (2n=20) and E. congesta (2n=20) studied by Sanyal and Sharma (1972) from India, all conform to the basic number of x=5. Eleocharis palustris studied by both Sanyal and Sharma Table 1 (contd.)

* Chromosome numbers are reported here for the first time . ** Discrepant chromosome numbers are recorded . •õ Authority quoted from Mehra and Sachdev (1975).

(1. c.) and Mehra and Sachdev (1975) were found to be characterised by n=7-8. These species are all smaller plants with much smaller spikes. On the otherhand E. acutangula (=E. frstulosa) and E. dulcis are larger and stouter plants with much longer spikes. While E. acutangula was found to be uniformly characterised by n=27 (Fig. 7) from widely different locality, E. dulcis shows a great amount of variation. Plants of this species studied from Smabalpur was found to possess very high chromosome number of n=c. 108 whereas those studied from Bhu baneswar revealed n=38 (Fig. 10). Identification of specimens were carefully checked and it seems there is a great amount of cytotypic variation within this 648 S. P. Rath and S. N. Patnaik Cytologia 43

Figs. 7-13. 7, Eleocharis acutangula, n=27 in diakinesis. 8, E. atropurpurea, n=10 in pollen mitotic metaphase. 9, E. geniculata, n=15 in pollen mitotic metaphase. 10, E. dulcis, n=38 in metaphase I. 11, Rhynchospora wightiana, n=10 in diakinesis. 12, Scirpus erectus, n=37 in metaphase I. 13, S. squarrosus, n=20 in diakinesis. All •~2500. 1978 CytologicalStudies in Cyperaceaewith Special Reference to its Taxonomy II 649 species. This needs a thorough sampling and cytological survey of specimens through the entire range of its distribution. Seven species of Fimbristylis were cytologically studied during the present investigation. These considered along with the previous findings in the genus have important bearing on the taxonomy of the group. Fimbristylis umbellaris was found to possess n=3 (and 2n=6) which is the lowest chromosome number yet recorded not only for Fimbristylis but for the family Cyperaceae as a whole. The three chromosomes in the haploid set were found to be quite distinct in the size as well as in the position of the centromere. Though varied chromosome numbers have been recorded only in a few species of Fimbristylis, most of the species are characterised by chromosome numbers which conform to x=5 as contrast to most of the genera of Cyperaceae except, however, Bulbostylis (Bulbostylis barbata pos sessing n=5, Fig. 14) which is an allied genus to Fimbristylis. With the dis covery of n=3 in a species of Fimbristylis it may be possible to trace the evolution of higher chromosome numbers in the family. This finding is also significant in correlating Cyperaceae with Juncaceae (n=3 in Luzula purpurea) the latter being considered as a progenitor of the families Cyperaceae and Gramineae. While Fimbristylis cinnamometorum, F. ferruginea and F. subbispicata were found to be characterised by n=5 (Figs. 15 and 16), Fimbristylis pluristriata (n= 15) was found to be at hexaploid level and F. tenera (n=20) was at octoploid level. Number of species like F. bisumbellata. F, argentea etc. have been reported earlier to be tetraploids with n=10. Thus a distinct euploid series is seen in this genus unlike other genera in the family. Fimbristylis complanata studied presently is also found to possess the haploid chromosome number of n=10 (Fig. 17). However, there is a great deal of confusion between F. complanata and F. quinquangularis and plants studied by us earlier and described under these two species (Rath and Patnaik 1974a) have all been assigned to F. quinquangularis. Cytologically of course all of them were found to possess n=5. But in view of the present finding our earlier report of n=5 for F. complanata is revised. The contradictory reports of chromosome numbers in F. dichotoma complex by various workers referred above are, however, highly significant and need certain clarification. The confusion in the identification due to extreme overlapping of characters and sympatric growth of the species of this group has been further magnified by the nomenclatural changes. F. dichotomy (Linn.) Vahl (=F. diphylla sensu Clarke in Hooker 1.c.) was found to be a diploid with n=5 whereas F, bisumbellata (Forsk.) Bubani (=F. dichotoma sensu Clarke in Hooker l. c.) is a tetraploid with n=10 (Rath and Patnaik 1974a). F. podocarpa Nees and Meyen, also another member of this complex and often indistinguishable from the above two species in external morphology is somewhat distinct by the possession of con spicuous gynophore of the nut. Plants differentiated on this basis were found to be characterised by n=10 (Fig. 18), though there is earlier report of n=15 for F. podocarpa (Sanyal and Sharma 1972). Fimbristylis pluristriata stat. nov. (=F. diphylla var. pluristriata Clarke) is another member of this complex which could be differentiated from others by its profusedly hairy leaves with silvery appearance and multistriate nature of the nuts. Large number of plants of this form were 650 S. P. Rath and S. N. Patnaik Cytologia 43

Figs. 14-19. 14, Bulbostylis barbata, n=5 in diakinesis. 15, Fimbristylis cinnamometorum, n=5 in diakinesis. 16, F. ferruginea, n=5 in diakinesis. 17, F. complanata, n=10 in diakinesis. (•ª, one separating bivalent. •ª•ª, two overlapping bivalents). 18, F. podocarpa, n=10 in diakinesis.

19, F. pluristriata, n=15 in diakinesis. All •~2500. 1978 Cytological Studies in Cyperaceae with Special Reference to its Taxonomy II 651 cytologically studied which revealed the haploid chromosome number of n=15 (Fig. 19). From morphological and cytological considerations a distinct specific status has been suggested for this taxon (Patnaik and Rath under pulbication). Of the six Indian species of Fuirena, five species have been cytologically worked out (Rath and Patnaik 1974b and Nijalingappa 1977) and the genus was found to be characterised by three distinct basic chromosome numbers of x=18, 19, 26. Fuirena uncinata, F. ciliaris and F. trilobites are a group of very closely allied species. While F. uncinata was found to possess the haploid chromosome number of n=18, F. ciliaris and F. trilobites were both characterised by n=19. Thus aneuploidy seems to have played a role in the evolution of these species. However, F. wallichiana (n=38) is the only tetraploid species (Nijalingappa 1972) known in the genus. F. umbellata stands out as a distinct species by its ecological preference, robust habit and comparatively larger spikelets. Cytologically also this species is distinct by the possession of very different basic chromosome number of x=26 and comparatively smaller sized chromosomes than those of the other species of the genus. Lipocarpha is represented in India by two species viz. L. chinensis and L. sphace lata which are very closely allied. Clarke (Hooker 1893) remarks that L. sphacelata is largely confused with L. chinensis (=L. argentea) but can be usually distinguished by more purple heads and shorter squamellae and style. However, chromo somally the two are quite distinct. L. chinensis has been earlier reported to be char acterised by n=13 whereas L. sphacelata presently worked out revealed the haploid chromosome number of n=19. Thus the trend of aneuploidy is evident in this small genus. One species of the genus Rhynchospora viz. R. wightiana was cytologically studied which revealed 10 bivalents at diakinesis (Fig. 11). This conforms to the basic num ber of x=5 reported earlier in R. tenuis (2n=10) studied from Netherlands (Gadella and Kliphuis 1964). Number of other species studied from outside India have been reported to be characterised by another basic number of x=13. Five species of Scirpus have been cytologically studied from this locality, two of which have been reported earlier (Rath and Patnaik 1974a). Large number of plants of S. articulatus revealed uniformly n=16 and our earlier report of n=15 was possibly of an aberrant plant and therefore should be discounted in the light of the present study. The chromosome number of this species has also been re ported by Sanyal and Sharma (1972) to be 2n=32. Number of species of this genus now studied by us as well as by other workers show the same trend of high aneuploid variation ranging from n=14 to n=55 (Figs. 12 and 13). However, now the genus Scirpus has been splitted in to several genera by some taxonomists and in the light of the cytological findings the taxonomic revisions should be evaluated. From a gross cytological standpoint Fimbristylis (also Bulbostylis) with its low chromosome numbers (lowest number recorded for the family) coupled with large sized chromosomes and euploid mechanism of evolution stands out distinct and considered primitive within the family. The genus also lacks in diffuse centromere as contrast to many genera in the family. Of course, whether the absence of diffuse 652 S. P. Rath and S. N. Patnaik Cytologia 43 centromere should be considered as an advanced character (Sanyal and Sharma 1972) is an open question. The genera like Carex, Eleocharis, and Scirpus etc. characterised by diffuse centromere exhibit high degree of aneuploidy and the development of this character might be a later specialization. However, as there is sporadic aneuploid variation in Fimbristylis, euploid mechanism has been found to be responsible for the evolution of a few species in more or less all the genera of the family.

Summary

Chromosome number of 45 species belonging to 12 different genera of Cyperaceae were determined through meiotic and pollen mitotic analyses. Of these 14 cases were proved to be new reports. While larger number of species were studied from Cyperus and Fimbristylis only one species from each of the genera Bulbostylis, Courtoisia, Lipocarpha and Rhynchospora had been investigated. In the light of these cytological findings the taxonomic problems in different groups of species in the family have been discussed.

Acknowledgement

The authors express their gratitude to Dr. B. Padhi, Professor of Botany, Utkal University, Vani Vihar, Bhubaneswar for facilities and encouragements throughout the progress of this work. Thanks are due to the University Grants Commision for the financial assistance.

Literature cited

Baquar, S. R. 1969. IOPB Chromosome number reports xxiv. Taxon 18: 683. Gadella, T. W. J. and Kliphuis, E. 1964. Chromosome numbers of flowering plants collected in Suriam. Acta. Bot. Neerl. 13: 432-433. Heiser, C. B. and Whitaker, T. W. 1948. Chromosome number, polyploidy and growth habit in California weeds. Amer. J. Bot. 35: 179. Hicks, G. C. 1928. Chromosome studies in Cyperaceae with special reference to Scirpus. Bot. Gaz. 86: 295-317. Hooker, J. D. 1893. The flora of British India. Vol. 6:Reeve & Co. Ltd. Lend. Mehra, P. N. and Sachdev, S. K. 1975. Cytology of some W. Himalayan Cyperaceae. Cytologia 40: 497-515. Nijalingappa, B. H. M. 1972. IOPB Chromosome reports xxxviii. Taxon 21: 683. - . 1975. Cytological studies in Fimbristylis (Cyperaceae). Cytologia 40: 177-183. - . 1977. Cytological studies in Fuirena Rottb. (Cyperaceae). Curr. Sc. 46 (4): 121-122. Patnaik, S. N. and Guru, K. M. 1968. A note on the cytotaxonomy of two species of Eleocharis R. Br. Samb. Univ. Journ. 1: 20-24. Rath, S. P. and Patnaik, S. N. 1974a. Cytological studies in Cyperaceae with special references to it's taxonomy I. Cytologia 39: 341-352. -. and -. 1974b. A note on the cytotaxonomy of East Indian species of the genus Fuirena Rottb. Bot. Mag. Tokyo 87: 333-336. - and -. 1977. A Note on the cytology of Fimbristylis umbellaris (Lamk.) Vahl. Bet. Mag. Tokyo 90: 79-81. Sanyal, B. 1972. Cytological studies on Indian Cyperaceae II. Tribe Cypereae. Cytologia 37: 33-42. 1978 Cytological Studies in Cyperaceae with Special Reference to its Taxonomy II 653

- and Sharma , A. 1972. Cytological studies in Indian Cyperaceae I. Tribe Scirpeae. Ibid. 37: 13-32. Tanak, N. 1937. Chromosome studies in Cyperaceae I. Cytologia, Fujii Jub. 2: 814-821. - . 1939. Chromosome studies in Cyperaceae VII. Chromosome number and pollen develop ment of Fimbristylis. Bot. Mag. Tokyo 53: 480-488. - . 1941. Chromosome studies in Cyperaceae XII. Pollen development in five genera, with special reference to Rhynchospora. Bot. Mag. Tokyo 55: 55-65.