AUTOCHTHONOUS LEAF ASSEMBLAGES AS RECORDS OF DECIDUOUS FOREST COMMUNITIES - AN ACTUALISTIC STUDY

How well do autochthonous leaf assemblages reflect live plant communit ies? How do leaf assemblages accumulating over different time scales c ompare in paleoecologic information content? Forest-floor leaf assembl ages accumulating over ten-day intervals (referred to here as short-te rm assemblages) and over a five-month season of leaf abscission (refer red to here as a long-term assemblage) were compared with the surround ing community in a modem temperate deciduous forest in northern Ohio. Leaf number in the long-term leaf assemblage is strongly correlated wi th the abundance of taxa (stem number) around the accumulation site an d weakly correlated with both average taxon size (stem circumference) and average taxon distance from the accumulation site. Of the variance in leaf number, 45% is explained solely by stem number and 67% by ste m number and average distance together. Average size explains an insig nificant amount of the variance in leaf number. Like the long-term ass emblage, leaf number in the short-term leaf assemblages is usually str ongly correlated with stem number and usually weakly correlated with a verage taxon size and average taxon distance. However, these patterns are not consistent, and the correlations are highly variable. Similarl y, there is high variability in the degree to which stem number, avera ge taxon size and average taxon distance account for variance in leaf number. Shortterm leaf assemblages are characterized by great fluctuat ions in taxonomic relative abundance, caused by seasonal variation in the timing and rate of leaf abscission among taxa. While autochthonous leaf assemblages accumulating over several months can reflect the sur rounding community with fair accuracy, leaf assemblages accumulating o ver shorter time spans are inconsistent records of the surrounding com munity. The depositional circumstances producing shea-term assemblages (i.e. event burial) may result in well-preserved specimens, but commu nity data from such assemblages should be treated with caution and, if possible, compared with data from contemporaneous long-term assemblag es.