Foliar Stoichiometry is Marginally Sensitive to Soil Phosphorus Across a Lowland Tropical Rainforest

The distribution of nutrients, both vertically and horizontally in a forest, has long been theorized to influence primary productivity. Working at La Selva Biological Station, Costa Rica, we gathered the most comprehensive foliar samples to date for a lowland tropical rainforest to measure horizontal and vertical trends in foliar nutrients. The mean traits of foliage from forest floor to top-of-canopy were determined at 45 plots placed across the landscape in a stratified random design. Area-basis foliar N and P for these vertically integrated columns varied by a factor of 3, while foliar N:P and mass-basis foliar N and P varied by a factor of 2. The variance in plot-level foliar N:P and P was best explained by total soil P, while variance in foliar N was best explained by soil pH (regression trees: r(2) >= 0.20, p <= 0.01). Other soil, topographic, and forest structure factors offered no additional explanatory power for variation of foliar nutrients from plot to plot. To explore vertical trends, we aggregated the data across the landscape into 2 mvertical segments. We found that foliar N:P was unrelated to height in the canopy, and that areabasis foliar N and P increased with height in the canopy (linear regression: r(2) = 0.82 and r(2) = 0.65 respectively, p < 0.0001 for both). We compared these vertical trends to those of the eight other elements quantified in the leaves, and the only other element enriched with height was potassium (K). Vertical nutrient enrichment was driven by increases in leaf mass per area (LMA), not mass-basis concentrations. Altogether, these findings suggest that, even in diverse tropical rainforests, foliar chemistry may reflect environmental constraints.