Table 4.

Comparison of various methods used to calculate biomass and NPP at the four sites

	Site
	Rhinelander	ORNL	Duke	KSC
Biomass
Leaves	2002–2008, littertraps. Pre-2002, allometric relationship.	Littertraps.	Littertraps, lagged for pines.	Diameter based allometric functions.
Wood	Diameter based allometric functions. Two functions were used depending on a diameter based cutoff.	Diameter and height based functions relationships, annual measurements of wood carbon density.	Diameter and height based allometric functions, annual measurements of wood density. Sub-sample of full plot.	Diameter based allometric functions.
Coarse-root	Linear function of above-ground tree mass and fine-root mass.	Diameter based allometric functions.	Function of above-ground biomass.	Soil cores, ground-penetrating radar, and allometric functions (when cores and GPR were not taken).
Fine-root	2002–2008, mini-rhizontrons. Pre-2002, allometric relationship.	Mini-rhizotrons.	Soil cores.	Mini-rhizotrons and soil cores.
Production
Leaves	Equal to biomass.	Equal to biomass.	Peak Leaf Area Index divided by species specific SLA.	Biomass increment plus litterfall. Litterfall estimated from littertraps.
Wood	Biomass increment at the tree scale*.	Biomass increment at the tree scale*.	Plot scale biomass increment.	Biomass increment plus litterfall. Litterfall assumed zero.
Coarse-root	"	"	"	Same method as fine roots.
Fine-root	2002–2008, in-growth cores and mini-rhizontrons. Pre-2002, biomass increment plus estimated root litterfall from 2002–2008 mini-rhizotron data.	Mini-rhizotrons.	Biomass multiplied by proportion of annual length production from mini-rhizotrons.	Biomass increment plus litterfall. Litterfall estimated as biomass multiplied by C turnover rate. Turnover rate measured using an isotopic tracer approach.
for details see	Talhelm et al.35	Norby et al.13	McCarthy et al.15	Hungate et al.16,69

*Accounts for mortality such that mortality is not included in this estimate and the minimum NPP for this variable is zero