Climatic responses of aboveground net primary productivity of a tropical rain forest: A 12-yr study

Background/Question/Methods

How tropical-forest net primary productivity (NPP) responds to climate change will have large implications for the biome’s rich biota and for the global carbon budget.  Current global vegetation process models project declining tropical-forest NPP with higher temperatures and increased drought stress; however, such model findings remain highly uncertain due to the limited field observations.  One approach for determining current climatic sensitivities of tropical-forest productivity is to monitor annual NPP components and assess their relation to interannual climatic variation.   Since 1997 we have monitored many aspects of forest carbon-cycling in the lowland tropical rain forest at La Selva, Costa Rica.  We have continuously assessed the major components of aboveground NPP (ANPP), fine litterfall and aboveground biomass increment, in 18 0.5-ha forest plots distributed over a 600-ha old-growth landscape.  Fine-litterfall is estimated based on biweekly collections of all leaf-, reproductive-, and twig-litter from basket and ground traps in all plots.  Annual aboveground biomass increment for each plot is estimated based on highly quality-controlled (above-buttress) diameter measurements of all woody stems (≥ 10 cm diameter) yearly.  We analyse here the interannual variation in estimated ANPP over the last 12 yr and its relation to interannual climatic variation.

Results/Conclusions

As is generally found for tropical moist forests, at La Selva the largest ANPP component (51%) was leaf litterfall (12-yr mean: 3.4 Mg C/ha/yr).  Annual leaf litterfall increased (non-significantly) with years of higher temperatures and/or greater dry-season intensity (less dry-season rain), but the interyear variation was relatively small (18%).  In contrast, estimated aboveground biomass increment (EABI; 12-yr landscape-scale mean: 2.2 Mg C/ha/yr), the second-largest ANPP component, varied much more among years (31%) and showed two highly significant climatic responses.  A two-factor model (dry-season rainfall - positive effect; nighttime temperatures - negative effect) explained 89% of the interannual variation in EABI over the 12-yr period.  These climatic sensitivities were not, however, shown by the remaining smaller ANPP components, twig litterfall and reproductive litterfall (6% and 10% of estimated ANPP, respectively).  In spite of the marked negative effects of annual temperatures and dry-season intensity on EABI, the lack of such climatic responses by the other measured productivity components resulted in little variation (10%) in total estimated ANPP among the 12 years.  The likelihood of contrasting climatic sensitivities by different components of tropical-forest NPP will need to be factored into the next generation of vegetation process models.