4.5 | Role of understorey species
A key difference between the GPP methods is that GPPiso/SF quantified GPP of the trees only whereas GPPPRELES quantified GPP of the whole ecosystem, which included understorey GPP. Understorey GPP was 41 g C m-2 in a 120-year-old Scots pine boreal forest (Kulmala et al., 2011) and 5% of the ecosystem GPP in mixed spruce-pine forest (Palmroth et al., 2019). PRELES estimated understorey GPP at our site to be 7 and 9% of the ecosystem GPP, i.e., 90 g C m-2 y-1 and 123 g C m-2 y-1, on the reference and the fertilised plots in 2013, respectively (Tian et al., under review.). In other words, this preliminary estimate of fertilisation treatment would induce 2% increase of understorey GPP in 2013. A direct comparison of tree GPP between the sap flux/isotopic and PRELES (GPPPRELES-7% and 9%) method would lead to 1369 vs 1194 g C m-2 y-1 in the R plot and 1483 vs 1248 g C m-2 y-1 in the F plot. However this estimate needs more replicates to confirm the understorey contribution to global GPP. As these methods continue to improve, it may become possible to solve for understorey GPP by difference. Note that if a next study shows that the fertilisation significantly increased understorey GPP, then GPPiso/SFwould not detect it, but GPPPRELES would. Future work should explore this possibility.