Measurement of pore size distribution in a lamellar Bt horizon using epifluorescence microscopy and image analysis

Abstract

A staining technique suitable for epifluorescence microscopy of polished block surfaces was developed using proprietary dyes (Sanford Co.). The dyes greatly enhanced the contrast between particles and pores, thereby allowing, with only minimal processing, the production of images suitable for automated pore and particle size measurements. An algorithm, the “multi-directional minimum chord” (MDMC) method, was developed for measurement of pore size as it relates to the hydraulic properties of soils. Pore chord lengths were measured in two or more directions through a set of points on a grid. The smallest dimension was retained as being most representative of the effective capillary diameter, or hydraulic diameter, of the pore. Chord length distributions measured by the MDMC method were compared to those obtained by the unidirectional chord intercept method. These techniques were applied to a study of lamellar Bt horizon formation in sandy glacial outwash in the Anoka Sand Plain, Minnesota, U.S.A. The lamellae are nearly horizontal, occur from 0.6 to > 2 m depth, and vary in thickness from 0.5 to 6 cm. It was hypothesized that differences in pore size distribution initially present in the sands caused changes in soil hydraulic properties that may have been responsible for the initial formation of lamellae. Undisturbed samples containing lamellae and the underlying interlamellar horizons were collected using Kubiena boxes, oven-dried at low temperature, cast in resin, sawed in half, polished, and stained. Digitized images of lamellae and the uppermost part of the interlamellar horizons were taken using epifluorescence microscopy and then edited to remove clays to simulate conditions present before the formation of the lamellae. Pore and particle size distributions measured on the images by the MDMC method showed that significant differences existed between the lamellae and the uppermost part of the underlying interlamellar horizons. In all cases a coarser over finer pore size discontinuity existed at the bottom of the lamellar horizon, in support of the proposed mechanism for the initiation of lamellae formation.