AUTHOR=McCalmont Jon , Heinemeyer Andreas , Morison James , Xenakis Georgios , Bell Michael , Wilkinson Matthew , Hastings Astley 

TITLE=Timeseries partitioning of ecosystem respiration components in seasonal, non-tropical forests; comparing literature derived coefficients with evaluation at two contrasting UK forest sites

JOURNAL=Frontiers in Forests and Global Change

VOLUME=Volume 7 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/forests-and-global-change/articles/10.3389/ffgc.2024.1352527

DOI=10.3389/ffgc.2024.1352527

ISSN=2624-893X

ABSTRACT=Understanding flows of organic carbon within ecosystems is key to quantifying impacts of landuse change on the climate. However, while net exchange of CO2 between ecosystem and atmosphere indicates higher-level impactsglobal warming potentials, partitioning into individual flux components is needed to understand sinks and sources, residence times, and sensitivities to land-use impacts. Scaling from research site to region requires modelling evaluated against in-situ measurements, but there is often a mis-match between outputs of process models (e.g. soil heterotrophic respiration (Rh)) and sitemeasured parameters (e.g., total soil surface respiration (Rs) or whole ecosystem respiration (Re) respiration). This study took a literature review approach to determine fractional coefficients for estimating Rh from Re or Rs and considered whether these fractions differed across a year in seasonal forests, where relative contributions of root respiration might be expected to vary between growing and dormant seasons. Compiled timeseries data were grouped by forest type (broadleaf, needleleaf and mixed) and coefficients for the fraction of each component (Rs or Re) that Rh represented were calculated using two approaches: a simple annual mean value over all months, and individual monthly means. These coefficients were then used to estimate Rh separately from higher level fluxes (Re, from eddy covariance and Rs from soil chambers), measured concurrently at two UK forest sites, and compared to Rh estimated from the same datasets using previously published generic coefficients, as well as to concurrently measured Rh and Re. The statistics of the agreement between these separate estimates of Rh within each literature-derived coefficient set were used to demonstrate the performance of the derived factors. Both approaches resulted in much closer convergence of the two separate estimates of Rh (derived from Re or Rs) than the previously published coefficients, particularly for Rh/Re coefficients which had previously been measured under peatland blanket bog rather than forest.