Notes on seed traps in terrestrial plant communities

Various seed traps have been used in ecological studies, but the knowledge of the effects of trap types on the number of seeds and species detected is rather sparse. In this paper we review current methods for monitoring seed rain, we re-analyse two published comparisons of seed traps and summarize our own experience with five trap types. We recommend traps with the following attributes: (1) separation of seeds from litter and soil particles, (2) protection against seed predation: and (3) dry preservation of seed material. For all trap types the retrieval rate of seeds is of paramount importance. Advice on the trap area and height above ground depends on the size of plants and the dispersal process studied, since these two parameters have species-specific effects on the quantities of seeds recorded. Large traps reduce the seed records especially for small species which are ant-dispersed or have no effective dispersal mechanism at all. In general, large numbers of small traps are most advisable to assess the internal seed rain of a plant community, whereas fewer large traps are sufficient to detect seed influx by colonizing species. An improved knowledge of the effects of the trap type on the results of dispersal studies will allow more reliable comparisons with existing sets of data. Eventually, standardization of trap types is desirable.