Predicting Soil Respiration from Plant Productivity (NDVI) in a Sub-Arctic Tundra Ecosystem

• Published in: Remote Sensing
• Main Authors: Azevedo, Olivia, Parker, Thomas C, Siewert, Matthias B, Subke, Jens-Arne
• Other Authors: NERC Natural Environment Research Council, Biological and Environmental Sciences, Umea University, orcid:0000-0002-6220-8865, orcid:0000-0002-3648-5316, orcid:0000-0001-9244-639X
• Format: Article in Journal/Newspaper
• Language: English
• Published: MDPI 2021
• Subjects: Abisko; CO2 flux; LAI; modelling; plant functional type; SOC; vegetation index; Arctic; Tundra
• Online Access: http://hdl.handle.net/1893/32839; https://doi.org/10.3390/rs13132571; http://dspace.stir.ac.uk/bitstream/1893/32839/1/remotesensing-13-02571.pdf

Soils represent the largest store of carbon in the biosphere with soils at high latitudes containing twice as much carbon (C) than the atmosphere. High latitude tundra vegetation communities show increases in the relative abundance and cover of deciduous shrubs which may influence net ecosystem exchange of CO2 from this C-rich ecosystem. Monitoring soil respiration (Rs) as a crucial component of the ecosystem carbon balance at regional scales is difficult given the remoteness of these ecosystems and the intensiveness of measurements that is required. Here we use direct measurements of Rs from contrasting tundra plant communities combined with direct measurements of aboveground plant productivity via Normalised Difference Vegetation Index (NDVI) to predict soil respiration across four key vegetation communities in a tundra ecosystem. Soil respiration exhibited a nonlinear relationship with NDVI (y = 0.202e 3.508 x , p < 0.001). Our results further suggest that NDVI and soil temperature can help predict Rs if vegetation type is taken into consideration. We observed, however, that NDVI is not a relevant explanatory variable in the estimation of SOC in a single-study analysis.