Total organic carbon in soils and its relation with manganese concentrations in soils and vegetation close to an abandoned manganese mine

This study aimed at quantifying the total organic carbon (TOC) present in soils within the proximity of the Kgwakgwe Mn oxide ore abandoned mine, Botswana, and establish its relationship with Mn concentrations in soils and vegetation based on multivariate and Geographical Information Systems (GIS) analytical techniques. Four hundred soil samples and 200 vegetation set samples were obtained from a 4 km² area close to the abandoned mine. The TOC in soil samples were determined using a carbon/hydrogen/moisture determinator, and Mn concentrations in soils and vegetation by atomic absorption spectrophotometry. Results were processed using the statistical package for social science (SPSS), GIS, and Remote Sensing (RS) techniques with the Integrated Land and Water Information System (ILWIS), Geosoft Oasis Montaj and ArcGIS software packages. The values for TOC in the soil samples from the study area ranged from 0 wt % to 7.91 wt %, with a mean of 1.90 wt %, and at the control area, from 4.07 wt % to 4.86 wt %. The range of concentrations of Mn in soils was from 36 mg/g to 24908 mg/g and for Mn concentrations in the vegetation samples from 26 mg/g to 3611 mg/g with a mean of 598 mg/g. Results of correlation coefficients depicted very weak negative association except Mn in soils/Mn in leaves which was weak but positive. The statistical data yielded four clusters as follows: cluster one consisted mainly of Mn in leaves, cluster two was constituted of Mn in soils, and cluster four had TOC. Cluster three was dominated by the three parameters but with negative t statistic. The spatial presentation of data presented revealed little or no vegetation in the south eastern area and those close to the mine workings, and some significant vegetation in the north western part of the study area. The low TOC in the soils is associated to low vegetation cover which is considered to have been influenced by the soil clay fraction mineralogy and high concentrations of Mn.

Keywords
Bonferroni adjusted clustering, cluster analysis, image processing, spatial presentation