Open Collections

Applying different partitioning and modelling approaches to investigate the effects of nitrogen fertilization and light inhibition on net ecosystem exchange using long-term eddy-covariance data

University of British Columbia

There is a lack of knowledge of the processes controlling the net ecosystem production (NEP) of forest stands after disturbance (e.g., logging and nitrogen (N) fertilization) in the Pacific Northwest. To answer these questions, long-term eddy-covariance (EC) measurements of carbon-dioxide (CO₂) above a Douglas-fir stand on Vancouver Island, Canada were used in this study. Additionally, answering these questions also requires that NEP can be accurately partitioned into gross primary production (GPP) and ecosystem respiration (Re) to understand the biophysical controls of their C dynamics. Therefore, an intensive stable C isotope campaign was conducted and various partitioning techniques were deployed in this study. The Physiological Principles Predicting Growth (3-PG) model was modified to compare its predictions with EC-measured data to better understand the effects of N fertilization on CO₂ and water vapour fluxes. Application of N fertilizer to this stand led to a short-term (first two years) increase in GPP followed by little change over the long term. Re increased over the short-term (first year), while it was appreciably suppressed over the long term. N fertilization resulted in an average increase in NEP by 170 g C m-² year-¹ on average and led to an average increase in annual water use of 15%. The light-inhibition effect of daily Re for the stand during the peak growing season was estimated to be 37% by comparing the nighttime relationship and the stable C isotope methods. The daytime intercept partitioning method only partly accounted for light inhibition effects. By applying four gap-filling models using different partitioning techniques (nighttime relationship, daytime intercept, ANN–nighttime, and FVS–ANN models) on the 18-years of measured data, it was estimated that the mean annual NEP totals ranged from 195 to 238 g C m-² year-¹, hence the choice of the gap-filling model would have a great impact on long-term C budgets. There were even larger differences in Re and GPP estimates. Our findings have implications for the interpretation of EC measurements, a widely-used data source for understanding terrestrial C cycling. I further argue that the consideration of light inhibition of daytime Re in terrestrial ecosystem models is critical.

Text

2020-08-19

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/75605

Geography

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/