INTRODUCTION

Sisyrinchium is a complex polyploid taxon in which the species are not always easily distinguished (1). It is a genus consisting of seventy to eighty species in the Western Hemisphere, best developed in South America (2). Over forty of these species, though, are distributed throughout the United States (1). Determining the number and names of the species in Oklahoma is where the task gets difficult. This is a problem that dates back as far as 1939 when Paul F. Smith wrote The Genus Sisyrinchium in Oklahoma for the Oklahoma Academy of Science. In his paper Smith wrote, “Attention to the Oklahoma species of the genus Sisyrinchium, commonly known as blue eyed grass, was first called when several classmates in taxonomic botany ran down one specimen to three different species, and, strangely enough, found visible evidence to support their contentions among the mounted specimens” (3). It seems little has been done since to remedy this situation. The Flora of North America recognizes six species within Oklahoma: S. angustifolium, S. campestre, S. ensigerum, S. langloisii, S. pruinosum, and S. albidum (1). The USDA PLANTS Database recognizes five species within Oklahoma: S. angustifolium, S. campestre, S. chilense (synonym S. ensigerum), S. langloisii (synonym S. pruinosum), and S. minus (4). An Annotated List of the Ferns, Fern Allies, Gymnosperms, and Flowering Plants of Oklahoma recognizes four species within Oklahoma: S. angustifolium, S. campestre, S. pruinosum, and S. minus (5). Both The Flora of Oklahoma and Keys to the Flora of Oklahoma recognize just two: S. angustifolium and S. campestre (6, 7). These confusions in both the distribution and the nomenclature of the genus were the bases for the two objectives of this project: first to determine the number of species of the genus Sisyrinchium within Oklahoma and second to determine the names of these species.

METHODS AND MATERIALS

The first stage of the research was collecting data. A total of one hundred and fourteen herbarium sheets borrowed from the Bebb Herbarium of the University of Oklahoma were used to make measurements and observations of ten morphological characters. The ten characters that were measured included outer bract length, inner bract length, ratio of outer and inner bract lengths, stem width, stem height, leaf width, presence or absence of broad wings on the aerial stems, inflorescence branching or not branching, nodes geniculate or nongeniculate, and presence or absence of a purple tinge on the base of the spathes. These measurements and observations were then used to create scatter diagrams in an attempt to segregate possible species. Adequately defined clusters were not formed, however, and a new plan of action was enacted. By looking through several taxonomic keys, I was able to determine what I believe to be the three most significant separating characters: presence or absence of broad wings on the aerial stems, inflorescence branching or not branching, and nodes geniculate or nongeniculate. I then studied these three character states in depth to see if there were patterns present. As a part of this evaluation new scatter diagrams were created based on these three character states.

RESULTS

The results of the original scatter diagrams were inconclusive. After analyzing the three characters, presence or absence of broad wings on the aerial stems, inflorescence branching or not branching, and nodes geniculate or nongeniculate, the specimens fell into four distinct complexes, which I referred to as Yellow, Blue, Green and Red in order to remain unbiased (Tables 1-4). The specimens that fell into the Yellow complex exhibited an absence of broad wings on the aerial stems, inflorescence not branching, and nodes nongeniculate (Figure 1). The specimens of the Blue complex showed an absence of broad wings on the aerial stems, inflorescence branching, and nodes geniculate (Figure 2). The specimens of the Green complex displayed an absence of broad wings on the aerial stems, inflorescence branching, and nodes nongeniculate (Figure 3). The specimens of the Red complex demonstrated a presence of broad wings on the aerial stems, inflorescence branching, and nodes nongeniculate (Figure 4). Two new scatter diagrams were then created which were color-coded in order to differentiate between the four complexes (Figure 7-8). This differentiation made it possible to recognize segregated clusters, which completely support the separation into four complexes.

DISCUSSION OF RESULTS

After keying out individual specimens from the four complexes I have come to several conclusions. The Yellow and Red complexes distinctly appear to be the species S. campestre and S. angustifolium, respectively. The presence of a purple tinge on the base of the spathes made it possible to determine that the Blue complex is composed predominately of S. langloisii but the chance still persists that S. pruinosum could be present in the Blue complex. The composition of the Green complex is still undecided at this time. Also after reviewing sheets of S. albidum from other states, it was determined that this species was not within our sample set (Figure 5). To further analyze the complexes, I used a cladistics software program to create a phenogram in order to discern the possible relationships between the four complexes (Figure 6).

ACKNOWLEDGEMENTS

First and foremost, I would like to thank Dr. Bruce Smith for all the time and energy he has devoted to help me with this project. I would also like to thank Ms. Paula Tolbert and my mother for helping me take measurements. I am very grateful to the people at both the Bebb Herbarium and the O.S.U. Herbarium for loaning their specimens. Lastly, I would like to thank Ron Hoggard of the University of Oklahoma for his help in creating the phenogram.

REFERENCES

1. Flora of North America Editorial Committee, eds. Flora of North America, Vol. 26. Oxford University Press, New York (1993) 475 pp.
2. Illustrated Flora of the Northeastern United States and Adjacent Canada, Vol. 1, The Pteridophyta, Gymnospermae and Monocotylednoeae. Hafner Press, New York (1952)
3. Smith, Paul F., The Genus Sisyrinchium in Oklahoma. Proceedings of the Oklahoma Academy of Science. 19, 55-56 (1939)
4. USDA, NRCS. 2002. The PLANTS Database, Version 3.5 (http://plants.usda.gov). National Plant Data Center, Baton Rouge, LA 70874-4490 USA.
5. Taylor, R.J. and C.S. Taylor. 1994. An Annotated List of the Ferns, Fern Allies, Gymnosperms, and Flowering Plants of Oklahoma, 3^rd edition. Published by authors, Durant, Oklahoma. 110 pp.
6. Tyrl, R.J., and S.C. Barber, P. Buck, J.R. Estes, P. Folley L.K. Magrath, C.E.S. Taylor, R.A. Thompson. 1994. Identification of Oklahoma Plants. Oklahoma Flora Incorporated, Noble, Oklahoma. 52 pp.
7. Waterfall, U.T., Keys to the Flora of Oklahoma, 4^th edition. Published by author, Stillwater, OK (1969) 246 pp.





ABSTRACT---- Sisyrinchium is a complex polyploid taxon in which the species are not always easily distinguished. The two objectives of this project were first to determine the number of species of the genus Sisyrinchium within Oklahoma and second to determine the names of these species. The first stage of the research was collecting data. A total of one hundred and fourteen herbarium sheets borrowed from the Bebb Herbarium of the University of Oklahoma were used to make measurements and observations of ten morphological characters. By looking through several taxonomic keys, we were able to determine what we believe to be the three most significant separating characters. These three characters were used to segregate the specimens into four distinct complexes designated by the colors Yellow, Blue, Green, and Red. The complexes Yellow and Red distinctly appear to be the species S. campestre and S. angustifolium, respectively. The complex Blue looks to be composed predominately of S. langloisii and possibly S. pruinosum. The specimens of the complex Green are undetermined at this time.