CSIRO Jeff Baldock1, Dr Peter Macreadie2, Dr Jeff Kelleway3, Dr Oscar Serrano4, Dr Paul Lavery4, Ms Christina Asanopoulos5, Dr Joey Crosswell6, Dr Catherine Lovelock7, Dr Matt Hayes7, Dr Andy Steven6
1CSIRO Agriculture and Food, Glen Osmond, Australia, 2Centre for Integrative Ecology, School of Life and Environmental Sciences, Deakin University, Burwood, Australia, 3School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong, Australia, 4School of Science & Centre for Marine Ecosystems Research, Edith Cowan University, Joondalup, Australia, 5School of Agriculture, Food and Wine, University of Adelaide, Glen Osmond, Australia, 6CSIRO Oceans and Atmosphere, Brisbane, Australia, 7University of Queensland, St Lucia, Australia
Coastal blue carbon environments (mangroves, tidal marshes and seagrass meadows) can contain significant stocks of soil organic carbon and can accumulate additional organic carbon through the capture and retention of organic materials derived from autochthonous and allochthonous sources. Assessing soil organic carbon stocks present in these environments requires quantification of soil organic carbon concentration. Developments in the combined use of infrared spectroscopy and partial least squares regression analyses (IR/PLSR) have demonstrated an ability to provide cost effective measurement of organic carbon concentration in agricultural soils and additionally provide values for the concentrations of inorganic carbon and total nitrogen from one analysis. The objective of this study was to assess the capability of IR/PLSR analyses to provide accurate values for total (TC), organic (OC) and inorganic (IC) carbon and total nitrogen (TN) concentrations in soil samples derived from blue carbon environments. A total 1201 samples were used. All TC, OC, IC and TN analytical data were acquired using automated dry combustion analysers (LECO TruMac, C-144 or CNS-2000) with the application of acid pretreatment to soils containing carbonate. Independent sets of 300 and 901 samples were used respectively to develop and then validate IR/PLSR predictive algorithms. Robust IR/PLSR models were obtained for TC, OC, IC and TN concentrations given the values derived for the coefficient of determination (R2=0.91-0.96), ratio of performance to deviation (RPD = 4.8-5.6) and ratio of performance to interquartile range (RPIQ = 2.5-3.8). After calibration, the concentrations of TC, OC, IC and TN of the blue carbon soils could be robustly predicted using a single IR scan. The IR/PLSR approach therefore provides a cost-effective alternative approach to quantifying the concentration of carbon and nitrogen in blue carbon soils.
Biography: Jeff Baldock is a research scientist working with CSIRO studying the cycling of organic carbon in a range of natural environments.