Zs. P. Komáromy szerk.: Studia Botanica Hungarica 14. 1980 (Budapest, 1980)

Rajczy, Miklós: A new measure for comparing the shape of moss leaves

Brid.; Grimmia homodyction Dix.; Hedwigla clllata (Hedw.) P. Beauv.; Pottia heimii (Hedw.) Hamp.; Pottia viridifolia Mitt.; Pottia atarckeana (Hedw.) C. Muell.; Tortula euneifolla (With.) Turn.; Desmatodon convolutus (Brid.) Grout.; Tortula laevipila (Brid.) Schwaegr.; Didymodon trifariuB (Hedw.) Roehl.; Eucladium verticillatum (Brid.) B. S. G.; Encalypta streptocarpa Hedw.; Dialytrichia mucronata (Brid.) Broth.; Cinclidotus fontinaloides (Hedw.) P. Beauv.; Ulota bruchii Homsch.; Splachnum wormskioldii Horn.; Bartramia stricta Brid.; Mnium hornum Hedw.; Orthotheeium rufescens (Brid.) B.

S.G.;

Platyhypnidium riparoides (Hedw. ) Dix. ; Eurhynchium striatura (Hedw.) Schimp.; Grimmia atrata Hornsen.; Plagiomnium ellipticum (Brid.)Kop.; Rhizomnium pseudopunctatum (B. S. G. ) Kop. ; Cinclidium s tygium Sw. The nomenclature is that of the Index Muscorum (Wijk et al. 1959-1969) and that of Koponen (1968) in the Mniaceae. I measured the following five characters: 1. the length of the leaf "i"; 2. the distance from the leaf base to the line where its width is reduced to the half of its maximum "ldm2"> 3

> tne distance from the ieaf base to the line where its width is maximum "1j_." ; 4. the maximum width of the leaf "dm"; 5. the width of the leaf base "db" (Fig. 1, Plate I-II). To get exact groups of the leaves I used average linkage cluster analysis (WPGMA), for the similarity matrix Hummon' s "shared species diversity" function was used (Sneath and Sokal 1973, Hummon 1974). For the analysis characters 1-5 were standardized, and four ratios were calculated as characters 6-9; 6. 100 x dm/1; 7. 10 x LPR; 8. 100 x <

1

dm2-l

dm

)/U 9. 100 x K /1. 'dm' Fig. 1. Measurement of leaf characters RESULTS AND DISCUSSION The dendrogram on Fig. 2. shows the result of cluster analysis. It gives 13 sub- and 4 main categories whose outlines can be easily drawn by means of the two proposed measures. In group I we can find the lanceolate leaves (0.l<dm/l<0

.32 and 0.05<LPR<0.8). Group II contains the one widely triangular leaf (dm/l=0.78, LPR=0.17). In group HI there are the ovate and elliptic leaves (0.2<dm/l<0.8 and 0.5<LPR<6) and in group IV there are the obovate leaves (10<LPR). For details see Table 2. On the plot (F*g. 3) the groups can be recognized more easily because of the logarithmic transformation of LPR. The new measure measure seems to be very useful in numerical studies when a comparison of various leaf shapes is wanted. I wish to express my gratitude to Dr. L.Hajdu for reading and criticizing the manuscript and to Miss M. Hangya for drawing the figures. Plate I-II: The leaves measured (for numbers see Table 1) Numbers 1-20 were redrawn from Dixon et al (1924) and Numbers 31-33 from Orbán (1974).