POLYMORPHISM IN THE DESMID COSMARIUM ABBREVIATUM VAR . MINUS ( ZYGNEMAPHYCEAE ) AND ITS TAXONOMIC IMPLICATIONS

Polymorphism in a population of the desmid (Zygnemaphyceae) Cosmarium abbreviatum Raciborski var. minus (West & West)Krieger & Gerloff is reported frommaterial collected from 3 stations at the Lily Pond located in the Parque Estadual das Fontes do Ipiranga, Sao Paulo City, southern Brazil. From the examination of approximately 200 plants, 6 different morphological expressions were described and the following remarks made: I) a detailed analysis of sample populations is absolutely necessary for the definition of species and infraspecific categories among desmids; 2) the morphological characteristics presently used for the delimitation of certain species, which are built up of individuals of very small size, seem to have no great taxonomical significance when only isolated individuals are analysed; and 3) taxa, such as Cosmarium angulosum Breb. var. angulosum f. angulosum, C. abbreviatum Rae. var. minus (West & West) Krieg. & Gerl., C. succisum G.S. West var. jaoi Krieg. & Gerl., C. tinctum Ralfs var. tinctum, and C. minimum West & West var. minimum are of doubtful validity and may prove to be I}1erely polymorphic expressions of C. abbreviatum var. minus.


Introduction
When in 1908 Playfair stated that desmids are polymorphic, and that about 90-95070 of the species already described were to be considered morphological expressions of the remaining 5-10%, he was almost literaly burned in public.Playfair's numbers were of course an exaggeration, but his ideas still have considerable validity.Nobody doubts anymore the existence of polymorphism in desmids, and that it surely has to do with taxonomy.
Studies of polymorphism in species of Cosmarium are represented by the works of Starr (1954Starr ( , 1958) ) and Tews (1969), which refer to the variability in the radiation, the first two in C. turpinii Breb.and the latter in C. botrytis Men.var.de pressum West & West, and by Kirk & Cox (1975), which refers to the variability in the form of the semicell outline in C. botrytis Men. in culture; Gerrath (1979a) has studied variability in the morphology of the basal angles of the semicells and pattern of decoration of the cell wall of C. taxichondrum Lund.; and Gerrath (1979b)

Material and methods
The population came from periphyton collection made from 3 different stations in the Lily Pond in the Sao Paulo City Botanical Garden, State of Sao Paulo, southern Brazil, on several occasions during the years 1981 and 1982.The collections were immediately fixed with 4% formalin and examined with a Zeiss Oberkochen regular microscope.Drawings were made using a camera-Iucida attachment.
Cosmarium Abbreviatum Var.Minus (Fig. 1-13) Cosmarium abbreviatum is one in a large group of species built up of very small individuals measuring somewhat withip 8-30 x 8-30J'm.It is a relatively common and widely distributed desmid.As originally described (Raciborski, 1885), it is characterized by its cell, about as long as broad or a little wider than long, 10-25 x 10-29 I'm, deeply constricted at the middle, the sinus narrow and closed.The semicells are transverse angular-oval and 5-sided, the lower lateral margins diverging and straight to a median, rounded angle which is slightly produced, and the upper margins are convergent, straight or retuse, to a relatively broad, truncate apical margin.The cell wall is smooth.In lateral view, the semicells are broadly oval to nearly rounded.In vertical view, the cell is asymmetrically oval, with broadly: rounded poles.' West & West (1900) described a f.minor, which differs from the typical mostly by its said consistently smaller size (9-13 x 8-13J'Iri), and by the narrowly rounded lateral angles which are never produced.Krieger & Gerloff (1969) raised it to the varietal level.'

Observations
The population examined showed variability in the depth (Fig. 1,4), and shape and aperture of the sinus which varied from linear and closed (Fig. 3) to acutangular (Fig. 4); in the shape of the lateral margins of the semicells which varied in the presence and position of a median angular inflation (Fig. 1-2, 4-5); and in the shape of the semicell outline which consequently varied from sub-elliptical (Fig. 4) to sub-rectangular (Fig. 5).As a result, the following 6 main morphological expressions (forms) could be detected: Form 1 -Sernicells transversely angular-oval, lower part of lateral margins straight of almost so, broadly divergent, forming a rounded median angle, upper part usually straight, rarely slightly retuse, convergent, upper angle obtuse, apex broadly truncate, straight or very slightly retuse at the middle (Fig. 2).
Form 6 -Semicells transversely sub-hexagonal, lower and upper parts of lateral margins unequal in length, the lower ones shorter than the upper, straight, median and upper angles obtuse, apex broadly truncate, straight (Fig. 3).
If single individuals are taken for taxonomic identification, and if we follow the usual, classic procedures of taxonomic identification by comparison of illustrations, form 1 would be considered to be a representative of Cosmarium abbreviatum Raciborski var.Clear differentiation of these taxa was complicated by the fact that a complete range of intermediate forms was found.Hence the identification of individual specimens was difficult.Frequently, in the same individual the 2 semicells are morphologically so different that it was impossible to apply just one name to the plant.Indeed, each semicell could be identified as a different species (dichotypical individuals).
Obviously we are dealing with a single population showing considerable variation among individuals.It is impossible to say that this or that individual is more typical than some others.Individual variation only shows which characteristics may safely be used in taxonomic studies, and which are too variable to be used for this purpose.In the present case the shape of the lateral margins of the semicells and consequently that of the semicell outline when only isolated individuals are considered seem to be of minor significance, because no pronounced breaks in these characteristics are noticeable.We must consider them, and chiefly their va~iation, within the population.For this reason, ,such characteristics alone should not be considered for delimitation of species, or even varieties.
Another aspect that should be discussed is the difference of observation by the various workers.All taxa above mentioned are constituted by individuals of very small size (8-28~min lenght and 7-22~min breadth), and the slightest slip of the pencil during the camera-lucid a sketching of the material for identification mayor certainly will -lead to morphological "deviations" from what is mentioned in the literature to be typical for that taxon.This will consequently lead to different taxonomical interpretations of that same material.Fig. 7-13 represent various individual interpretations from differentstlldents at the Phycology Section of the Instituto de Botanica of the very same plant and using the same microscope and drawing equipment.

Conclusions
After careful examination of the almost 200 plants studied, the following remarks can be made: 1.It is hazardous, perhaps impossible, to define species, infraspecific categories, and even genera among desmids without a detailed analysis of sample populations, in order to know the morphological variation of the many characteristics commonly used in taxonomy, as well as the limit of confidence in their use.If this is true, such taxa may be considered synonymous from the nomenclatural viewpoint.4.There obviously are"'-and we must consider them -differences in observation by the various students.In taxa constituted by individuals of very small size, the slightest slip of the pencil during the camera-Iucida sketching of the material for identification may lead to "morphological variations", and consequently to the giving of different names to the very same plant.
, morphological variation of C. quinarium Lund.The present paper represents the results of a study of a natural population of C. abbreviatum Raciborski var.minus (West & West) Krieger & Gerloff.
Cosmarium angulosum Bn!bisson var.angulosum f. angulosum, Cosmarium abbreviatum Raciborski vaT.minus (West & West) Krieger & Gerloff, Cosmarium succisum G.S. West var.)Qoi Krieger & Gerloff, Cosmarium tinetum Ralfs var.tinetum, and Cosmarium minimum West & West var.minimum, all built up of very small sized individuals, is today based fundamentally on the shape of the lateral margins of the semicells and consequently on that of the semicell outline.Such morphological characteristics, instead of being universally accepted, seem to have no great taxonomical significance when only isolated individuals are analysed.We must consider the variation spectrum of these characteristics within the population.3. It is quite probable -and further studies can and must confirm the present one -that the taxa Cosmarium angulosum Bn!bisson var.angulosum f. angulosum, Cosmarium abbreviatum Raciborski var.minus (West & West) Krieger & Gerloff, Cosmarium suecisum G.S. West var.jaoi Krieger & Gerloff, Cosmarium tinetum Ralfs var.tinetum, and Cosmarium minimum West & West var.minimum are merely polymorphic expressions of a single taxonomic category.