Biometric based estimates of net primary production (NPP) in a cool-temperate deciduous forest stand beneath a flux tower

We measured the forest biomass and biometrically derived net primary production (NPP) in a cool-temperate deciduous forest stand beneath a flux tower. NPP was calculated as the sum of the living biomass of new-season tissue in all organs (B) and biomass of new-season tissue lost due to death (L). Annual leaf-litter production was adopted as L. We regarded the total tree growth in the stand as B, and examined three methods for estimating B to discuss the practicality of continuous measurement of NPP to compare with corresponding estimates of eddy-covariance based net ecosystem exchange (NEE). The three methods were diameter at breast height (DBH)-growth allometry by stem analysis of sample trees (SA method), DBH-growth allometry by core sample analysis of sample trees (CS method) and direct measurement of stem growth by tree census (TC method). The total annual tree growth in the forest stand estimated by the SA, CS and TC methods was 2.26, 1.60 and 2.38 Mg ha^-1, respectively, and NPP was 5.64, 4.98 and 5.74 Mg ha^-1. The slope of the regression of DBH against annual tree growth was slightly smaller for the CS method than for the SA method; the CS method underestimated the growth of several sample trees that had no clear main stem and, as a result, greatly underestimated B. To estimate B, the SA and CS methods depend on the use of DBH-growth allometry. Thus, it is difficult to determine species-dependent growth in natural mixed forests by these methods if only a few sample trees are used. In contrast, the TC method can directly and continuously measure the growth of all tree stems. Therefore, the TC method is the most suitable method for measuring NPP through annual measurements of tree diameter and leaf-litter production allowing for direct comparison with eddy-covariance based estimate of NEE.