Abstract
Non-forest ecosystems (predominant in semi-arid and arid regions) contribute significantly to the increasing trend and interannual variation of land carbon uptake over the last three decades, yet the mechanisms are poorly understood. By analysing the flux measurements from 23 ecosystems in Australia, we found the the correlation between gross primary production (GPP) and ecosystem respiration (Re) was significant for non-forest ecosystems, but was not for forests. In non-forest ecosystems, both GPP and Re increased with rainfall, and, consequently net ecosystem production (NEP) increased with rainfall. In forest ecosystems, GPP and Re were insensitive to rainfall. Furthermore sensitivity of GPP to rainfall was dominated by the rainfall-driven variation of LAI rather GPP per unit LAI in non-forest ecosystems, which was not correctly reproduced by current land models, indicating that the mechanisms underlying the response of LAI to rainfall should be targeted for future model development.
Highlights
Net ecosystem production (NEP) is a small difference between two large fluxes, namely gross primary production (GPP) and ecosystem respiration (Re)
Most Australian non-forest ecosystems are shrublands and savannas that together significantly contributed to the IAV of global land carbon uptake over the last three decades[1]
This study has further demonstrated that it is the rapid response of canopy LAI, and to much less extent the response of GPP/ LAI that is the dominant the large contribution of Australian non-forest ecosystems to global land sink IAV
Summary
Net ecosystem production (NEP) is a small difference between two large fluxes, namely gross primary production (GPP) and ecosystem respiration (Re). Annual rainfall and mean annual temperature drive much of the inter-annual variability in GPP and Re6–8, for non-forest ecosystems. Previous studies found that the net ecosystem carbon balance of Australian www.nature.com/scientificreports/. Non-forest ecosystems was principally driven by year-to-year fluctuations in rainfall via changes in both ecosystem GPP and Re10–12. To identify the key mechanisms controlling the IAV of NEP for Australian terrestrial ecosystems, we analyzed carbon fluxes from 23 Australian eddy flux sites, which together covered all major Australian ecosystem types[13]. This study will identify the key mechanisms causing interannual variability in terrestrial ecosystem C balances of Australian forest and non-forest ecosystems, and assess whether those mechansims are correctly represented by some of the most advanced global land models. To achieve the second aim, we compared the observed variances of log-transformed GPP, LAI, GPP/LAI and the covairnace between log-transformed LAI and GPP/LAI with the simulations from four process-based ecosystem models from the TRENDY (Trends in net land-atmosphere carbon exchange) compendium[14]
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call