Detecting Moisture Change in Fertilizer Microsites through Soil Color-Moisture Calibrations.

Flatbed scanners have been successfully used to image root growth for some time. Growing interest in quantifying the properties of the soil surrounding the root has led to increased scrutiny as to what soil factors can be measured using image analysis. In this study we calibrate the color of soil with the moisture in order to study moisture distributions surrounding the developing root systems. Soil moisture calibrations were carried out on 13 different soils from around Washington State. The soils were wetted to field capacity and successively dried, weighed for gravimetric moisture, and scanned. Images were collected in RGB and HSV color spaces. In order to apply the calibration curves to the rhiztron studies ArcGIS software was used to discriminate between physical fractions of soil in which color change occurred and which color change did not occur. Calibrations of soil moisture with soil color had r² values ranging from 0.74-0.95. As was found in previous work a generalized soil-moisture calibration between different soils had low correlations (Persson 2005; Sánchez-Marañón et al. 2007). The driving factor of the shift in calibration curves was found to be related most highly with textural and field capacity variations between the soils. The application of soil moisture-color calibrations into high resolution rhizotron images (resolution=5.29 um pixel^-1) proved difficult due to physical heterogeneity of the in-situ soils. In order to detect a noticeable color change due to moisture macro-pores were discriminated from meso-pores and micro-pores. It was found that the aggregates containing meso-pores and micro-pores were responsible for the detectable color change related to moisture change. Calibrations of soil moisture and color were successfully achieved for 13 different soils. The calibrations were applied in high resolution heterogeneous soils being used in rhizotron studies. The establishment of high resolution color moisture calibrations in high resolution heterogeneous rhizotron studies has the capability of greatly increasing our understanding of root soil relationships.