Forest ecosystem productivity is generally thought to be limited in tropical regions by soil phosphorous (P) availability. However, high rainfall and temperature that cause rapid weathering of soil P also contribute to leaching of other nutrients. We used five, 1-ha forest plots to test whether annual wood tissue production (WP) is related to soil availability of a wide variety of resources and soil properties, including phosphate (PO₄), nitrate (NO₃), ammonium (NH₄), calcium (Ca), potassium (K), magnesium (Mg), aluminum (Al), nitrogen mineralization rate, and net nitrification. Tree growth was monitored for up to four years at five sites distributed across soil types within a wet tropical forest in Costa Rica. We calculated standing mass at two time points based on tree diameter and an allometric scaling equation, but since we did not directly measure leaf production, the mass increment represents only WP. We tested relationships of WP to resources at the whole plot level (1 ha) and also using subplots of 0.164 ha and 0.041 ha to minimize within-subplot resource variation.

Results/Conclusions

Maximum WP (6.34 Mg ha^-1 yr^-1) occurred on younger alluvial soil and minimum WP (3.83 Mg ha^-1 yr^-1) on older residual soil, but production was not associated with soil type. At the 1-ha scale, WP was positively, but insignificantly, related to PO₄ (r² = 0.296, p = 0.343). Neither WP nor standing mass was significantly related to any measured soil resource at any scale, despite three-fold variation in PO₄ availability and two-fold variation in N mineralization. At 1-ha and 0.164-ha, WP was not significantly related to standing mass, while a positive relationship to standing mass (r² = 0.148, p < 0.001) at the 0.041-ha scale was likely due to the influence of large canopy emergent trees in some subplots. We did not measure leaf production, which may constitute a large portion of tropical forest productivity. However, leaf turnover rates are rapid compared to wood, so ecological questions such as forest carbon sequestration potential can be addressed using wood mass accumulation data. We are currently testing how resources vary with topography to more accurately interpolate resources at finer spatial scales, which will enable testing relationships between local resource availability and individual tree growth.