Consequences for seed distributions of intra-crop variation in wing-loading of wind-dispersed species

We used a computer simulation to quantify how intra-crop variation in wing-loading in a wind-dispersed species affects the seed distribution around a parent plant. We used a data set of seed distributions generated from a previous field study using artificial fruits varying in seed mass or fruit area dispersed from a tower into a tropical forest. For this study, the spatial distribution from each hypothetical parent's fruit crop of 1000 was calculated by randomly drawing locations of dispersed fruits from the previous data base. Three parents with contrasting fruit crops were used to test two hypotheses: 1) Increasing within-parent variance in wing-loading (= weight/area), while maintaining the mean, will lead to an increase in the area and uniformity of the seed distribution, without changing the mean dispersal distance. 2) Decreasing within-parent mean wing-loading, (which also decreases variance), will lead to an increase in mean dispersal distance, area, and uniformity of the seed distribution. The hypotheses were tested under four wind speeds.Increasing variance in wing-loading resulted in increasing the area and uniformity of density of the seed distribution without changing mean dispersal distance. Decreasing mean and variance in wing-loading resulted in increasing the area and uniformity of density of the seed distribution, as well as increasing the mean dispersal distance. Similar results occurred whether the differences in wing-loading arose by altering seed mass or fruit area. The effect of wind speed was consistently greater than the effect of parent. Generally, the same pattern of parent effects on seed distributions occurred, regardless of wind condition.The effects on seed distributions differed for alterations in mean versus variance, specifically in whether mean dispersal distance was increased. How selection may act on intra-crop mean and variance in wing-loading will depend on additional factors, e.g. the relative importance of distance, area, and density on seedling recruitment and the relative costs for crop size and seedling establishment of making fruit crops of a given mean and variance.