Near infrared spectroscopy of forest soils to determine chemical and biological properties related to soil sustainability

• Published in: Forest ecology and management 2002-11, Vol.171 (1), p.121-132
• Main Authors: Ludwig, B, Khanna, P.K, Bauhus, J, Hopmans, P
• Format: Article
• Language: English
• Subjects: Agronomy. Soil science and plant productions; Biological and medical sciences; C mineralisation; Chemical, physicochemical, biochemical and biological properties; Forestry; Fundamental and applied biological sciences. Psychology; General forest ecology; Generalities; Generalities. Production, biomass. Quality of wood and forest products. General forest ecology; Labile C; Microbial biomass; N mineralisation; Physics, chemistry, biochemistry and biology of agricultural and forest soils; Soil P; Soil respiration; Soil science
• ISSN: 0378-1127; 1872-7042
• DOI: 10.1016/S0378-1127(02)00467-X

Sustainability related soil changes due to site disturbance during forest harvesting need to be assessed. Near infrared spectroscopy (NIRS) may be useful for assessing large numbers of samples covering a landscape and dispersed over a long time span. This paper presents an evaluation of the use of NIRS to examine a number of chemical and biological soil properties. One hundred and twenty soil samples were collected from two mountain ash sites in Victoria, Australia, 10 years after forest harvesting. The sites had received a number of site and slash management treatments. Samples were analysed for total contents of C, N, microbial C (Cmic) and N (Nmic), various indices of available soil P and N content in salt solutions. To determine the potentially mineralisable C and N, samples were incubated for 264 days during which CO 2 evolution and N mineralisation were periodically determined. For each sample the reflectance spectrum was collected in the NIR (including the visible) range between 400 and 2500 nm. The sample set was broken up randomly for the calibration ( n=40) and validation ( n=80) and the entire spectrum was used to obtain the most suited mathematical (zeroth, first, second or third derivative) results. The validation showed that NIRS was very useful to predict the contents of C and N, and cumulative respiration and N mineralised after 264 days. The slopes ( a) of linear regression (measured against predicted values) ranged from 0.9 to 1.1 and correlation coefficients ( r) were greater or equal to 0.9. NIRS was quite useful ( r≥0.8, 0.8≤ a≤1.2) to predict Cmic, cumulative respiration and N mineralisation after 53 days, potentially mineralisable C, and Olsen P. For Nmic, the predictions were less satisfactory ( a