On Extrapolating Nighttime Ecosystem Respiration To Daytime Conditions and Implications for Gross Primary Productivity Estimation

Abstract:

Gross primary productivity (GPP) is a key term in the carbon cycle science. Being difficult or even impossible, at the ecosystem scale to directly quantify, various methods are used to estimate GPP, such as: eddy covariance CO₂ flux partitioning, carbonyl sulfide exchange, sun-induced fluorescence, isotopes of CO₂, and the photochemical reflectance index. The primary source of global GPP estimates is the FLUXNET project within which GPP is estimated in a consistent fashion through eddy covariance flux partitioning at more than 700 sites globally. Since the net ecosystem CO₂ exchange (NEE) reflects net uptake during daytime, when photosynthesis exceeds respiration, and net emission during nighttime due to ecosystem respiration (RECO), the eddy covariance flux partitioning is based on the idea that daytime RECO may be inferred from nighttime NEE direct measurements, and consequently GPP can be obtained by subtracting RECO from NEE. However, the main assumption underlying this approach, which is that a temperature-dependent model of RECO parametrised based on nighttime temperatures may be extrapolated to daytime temperatures, has not been conclusively tested.

This study investigates whether nighttime measurements of RECO provide unbiased estimates of daytime RECO. To this end we used ecosystem respiration chambers in a mountain grassland which, by keeping the vegetation in the dark during the measurement, allowed us to directly quantify RECO during both day and night. These data, pooled by day, night or day and night, were then used to parametrise temperature dependent models of RECO. Results show that day and night RECO do not follow the same relationship with temperature and that RECO inferred by using the nighttime parametrisation overestimates the true respiration. Potential reasons of this observed bias, like the overestimation of daytime mitochondrial respiration and implications for the quantification of GPP are discussed.