Remotely-sensed indicators of N-related biomass allocation in Schoenoplectus acutus

Background/Question/Methods

Many coastal marshes depend on plant belowground net primary productivity (BNPP) to accrete soil and alleviate flooding as sea level rises. Therefore, understanding BNPP may help control wetland loss. Wetland BNPP also contributes to long-term storage of soil organic carbon or peat and BNPP estimation may assist carbon accounting. Eutrophication reduces BNPP for wetland plants that conform to the balanced-growth hypothesis (e.g. decreased BNPP and increased aboveground net primary productivity (ANPP) when nutrients are not limiting). We extended the balanced-growth hypothesis to a common freshwater macrophyte, Schoenoplectus acutus. In an N-addition experiment, we tested whether remotely-sensed spectral leaf N indices could predict % leaf total N, BNPP:ANPP ratios and BNPP of S. acutus. We installed 4 plots each in an N-addition and control treatment within a wetland impoundment. Following 5 months of summer growth, we harvested plants and measured leaf N and plant net primary productivity (NPP).  Prior to harvest, we measured plant spectral reflectance over 350 – 2500 nm with a portable spectroradiometer and used these data to simulate 164 hyperspectral Hyperion satellite bands.

Results/Conclusions

N addition did not alter NPP, but reduced BNPP 36% and increased ANPP 71%. We correlated % leaf N with published indices and all possible normalized difference (ND), simple band ratio (SR) and first order derivative ND (FDN) and SR (FDS) spectral indices based on simulated Hyperion bands. FDS_1235,549  and FDS_743,1316  strongly correlated with % leaf N. The best index, FDS_1235,549, was significantly related to BNPP. Ours in the first demonstration of relationships among spectral N indices and BNPP, provided N increases with environmental N and plants conform to the balanced-growth hypothesis. Because many wetland emergent plants fit these criteria, remote-sensing is promising for estimating relative differences in wetland BNPP. Spectral N indices should aid side-wide productivity estimation and assist regional monitoring.