B. Papp szerk.: Studia Botanica Hungarica 38. 2007 (Budapest, 2007)
Csontos, Péter, Tamás, Júlia; Balogh, Lajos: Thousend-seed weight records of species from the flora of Hungary, II. Dicotyledonopsida
naturalised. Families Asteraceae (45 species) and Rosaceae (32 species) have the greatest representation in this list. Of Rosaceae, several data came from the rose collection maintained by Géza Facsar at the Soroksár Botanical Garden (FACSAR 1993). Following seed weight measurements the samples were deposited in the seed collections of both the Department of Plant Taxonomy and Ecology of the Eötvös Loránd University, Budapest, and the Botanical Department of the Hungarian Natural History Museum (BP). Thirty-seven among the studied species are protected members of the Hungarian flora (e.g. Crambe tataria, Doronicum hungaricum, Dracocephalum austriacum, Telekia speciosa, Vincetoxicum pannonicum, etc.). Seed weight studies on protected species may improve seed quality assessment and propagation success of these valuable species, as it was published for Vinco, herbacea (BÉNYEl-HlMMER et al 1999). A further benefit of seed weight data in conservation activities can be achieved when they are used for determining sowing density of matrix species in habitat restoration projects. Statistical analysis of seed weight data among species groups resulted significant differences in two of the three cases of possible pair-wise comparisons (Table 1). Table 1. Differences of mean seed weights of three distinct species groups. Seed weights were considered as the logarithm of raw data, and analysed by the KruskalWallis test (non-p arametric ANOVA) (INSTAT 1997). Species groups compared Mean rank difference Level of significance annuals and biennials vs. non-woody -8.411 P > 0.05 ns perennials annuals and biennials vs. woody perennials -112.66 P < 0.001 non-woody perennials vs. woody perennials -104.25 P < 0.001 ns = not significant The group of woody perennials showed higher average seed weight than the other two species groups. This corresponds to the finding of THOMPSON and RABINOVITZ (1989) and underline a loose, but existing positive correlation between plant size and seed weight. The statistical similarity of seed weights between short lived and long lived herbaceous plants may indicate that longer life-span does not necessarily cause larger body size in non-woody perennials. An additional
