Rothwald seed rain data

Given the limited dispersal distances for many temperate and tropical tree species (Ribbens et al. 1994, Nathan and Muller-Landau 2000, Martin and Canham 2010), the effect of spatial variability in seed input at the forest floor is fine-scaled and unfolds at the stand level. However, although seed input into plant communities represents a fundamental template for future plant population structure (albeit not a good proxy for reproductive success of plants), the fine scale spatial variation of seeds at the forest floor, its temporal variation with different sizes of seed crops, and its consequences for tree recruitment are poorly understood. This is partly because of limited data on fine scale spatial variation of seed deposition.

Methods
We established 25 seed traps (0.24 m2) in two study plots, one ha each. For a description of the plots, see e.g. Gratzer et al. (2021) (OIKOS, doi: 10.1111/oik.08826). Seed traps were covered with a wire mesh with a mesh width of 1 cm to avoid seed predation of rodents in seed traps. After preliminary data collection in 2003, we continuously monitored seed rain from 2006 – 2018. Traps were emptied multiple times (once in autumn and once in spring, in three years twice in autumn) throughout the annual cycle, and seed counts were associated with the last seed drop season. Starting in 2008, we identified non-viable (empty) and insect-damaged beech seeds that were predated prior to dispersal in the tree crown (hereafter “predated”) to determine the degree of pollen limitation and predispersal predation (Nilsson & Wästljung 1987).

This study is part of the research project “Sporadic seed production in mast seeding trees”, P30381 of the Austrian Science Fund (FWF). Recordings were started in the framework of the FWF research project ‘Forest dynamics in old growth spruce-fir-beech forests’, P14583.