Tuesday, November 3, 2009
Convention Center, Exhibit Hall BC, Second Floor
Abstract:
Effective measurement of soil C at the field scale requires an understanding of the spatial variability of soil C on a landscape scale. Recent technological advances in soil C measurement contents offer new opportunities in this area. Our objectives were to (1) characterize soil carbon vertically and horizontally, and (2) evaluate NIR as a tool to measure soil carbon. Six fields were studied, each field divided into 5 to 8 EC zones or using NIR map as having similar soil properties within 3 to 4 m distance. Fields were mapped on 20 m transects at 8-10km/hour and probed to 60 cm depth using both Veris NIR (500 to 2200 nm) Spectrophotometer shank and NIR Spectrophotometer Probe. Within each zone, 3 soil profiles were sampled at an equal distance of 3 m for examining total carbon, total nitrogen and bulk density. Samples were analyzed for total C and N with a Thermo-Finnigan Flash EA 1112 and a Leco CN 2000. Spatial variability of these soil properties was evaluated using both Minitab regression procedures and Proc Mixed from SAS 9.1. The results indicated that soil C varied considerably in some fields even with a similar soil type. However in some fields, soil C had low spatial variation. Most of the variation occurred in the surface 30 cm. The field by depth interaction was significant only to 30 cm. Small (3 m triangles) geo-referenced sampling points reduced spatial variation. Surface measurements of soil C by NIR provided a quick assessment of soil C and, soil C predicted by NIRS and measured by dry combustion laboratory measurements was correlated with an R-squared of 0.836. Stratifying fields by soil type and surface soil C mapped by NIRS may help reduce variability for measuring and monitoring soil C at deeper depths.