Meteorological and functional response partitioning to explain interannual variability of CO2 exchange at an Irish Atlantic blanket bog

Abstract

This study aims to develop models to describe CO2 fluxes in terms of environmental and meteorological variables and their variation over an Atlantic blanket bog in Glencar, southwest Ireland. Ten full years (September 2002-August 2012) of data were included in the assessment of CO2 flux and micro-meteorological data. Models were based on non-gapfilled growing season data, using complete calendar years for annual models, and the entire time-series for weekly models, whilst taking interaction between variables into account for increased model accuracy. This was to determine which environmental variables were most influential in directly controlling CO2 exchange on a long- and short-term basis. A homogeneity of slopes (HOS) model was used to determine if there was any ecosystem response to indirect effects (functional response) of environmental or meteorological interannual variation. This model uses multiple regression analysis to determine if the ecosystem response can be better predicted as a linear function of the variables using a single slope model for all years, or a separate slopes model for each year. The separate slopes model gave a different (and improved) outcome for both daytime and night-time CO2 fluxes, and so functional responses were deemed to have occurred. The contribution to variation of day and night-time net ecosystem exchange (NEEday and NEEnight respectively) was then separated into four components: indirect functional responses, direct interannual meteorological variability, direct week to week meteorological variability, and random error, which identified 13.8%, 36.6%, 28.2% and 21.4% respectively of the variation in NEEday, as well as 23.4%, 24.4%, 22.2% and 30% respectively of the variation in NEEnight. Water table level (WTL) had the greatest influence upon functional variation of NEE at the Glencar blanket peatland, and comparisons of modelled NEEday with leaf area index (LAI) measurements verified the estimate of functional contribution using the separate slopes model. The significance of interannual variation (IAV) and functional responses on NEE at Glencar suggests that it is a resilient ecosystem which might be able to adapt to environmental or climatic changes, although given current climate change predictions, it is likely to have a reduced carbon dioxide sink status in the future. © 2014.