Factor Analysis of Particle Size Specific Mineral Magnetic Measurements on Agricultural Topsoils from the Isle of Man
Booth, Colin A. Fullen, Michael A. Smith, John P. Hallett, Michael D. Walden, John Harris, John Holland, Kim
Booth, Colin A.
Fullen, Michael A.
Smith, John P.
Hallett, Michael D.
Walden, John
Harris, John
Holland, Kim
Editors
Other contributors
Affiliation
Epub Date
Issue Date
2006
Submitted date
2007-02-22
Alternative
Abstract
Agricultural topsoils from the five soil categories of the Isle of Man (British Isles) have been fractionated at discrete particle size intervals and their mineral magnetic properties have been analyzed. The aim was to characterize Manx soils and, with the aid of simultaneous R- and Q-mode factor analysis, evaluate the use of mineral magnetic measurements as an appropriate means of discriminating soil categories. Results show Manx agricultural topsoils contain a range of magnetic concentrations (similar to sedimentary and acid igneous rocks), magnetic mineralogy (greater influence of magnetically soft than magnetically hard minerals) and magnetic domain size (mainly stable single domain and superparamagnetic grains) characteristics, with significant differences between the five identified soil categories for each of the fractionated size class intervals used. Despite the mineral magnetic approach showing considerable potential for classifying Manx soils on the basis of their magnetic properties, large variations exist within individual soil categories. However, compared with previous studies, particle size specific measurements provide a more appropriate means of discrimination than bulk magnetic measurements. Each specific fractionated particle size range accommodates similar abilities of discrimination, yet no single class size is better than any other. Nevertheless, the ability of factor analysis to detect multivariate patterns in mineral magnetic data shows it is a useful data analysis tool for interrogating soil data sets.
Citation
Communications in Soil Science and Plant Analysis, 37(1-2): 249-273
Publisher
Journal
Research Unit
PubMed ID
PubMed Central ID
Embedded videos
Additional Links
Type
Journal article
Language
en
Description
Metadata only
Series/Report no.
ISSN
00103624
1532-2416 (Online)
1532-2416 (Online)