Arbuscular mycorrhiza status of gold and uranium tailings and surrounding soils of South Africa's deep level gold mines. II. Infectivity

Abstract

An AMF infectivity study and spore viability assessment was performed on substrata obtained from gold and uranium mine tailings dumps (‘slimes dams’) in the North West and Free State provinces of South Africa. Three slimes dams in each region were categorized as recently vegetated (RV), old vegetated (OV) and never vegetated (NV), and dams then divided into five zones based on elevation above ground level, steepness and broad chemical differences. Rhizosphere samples were collected from two of three plant species common to all sites; Eragrostis curvula, Atriplex semibaccata and Cynodon dactylon, as well as from bare areas, in order to allow comparisons across all site categories because of the rarity of the grasses on the lower slope of NV dams. Infectivity was determined by the mean infection percentage method from a bioassay of the substrata using Eragrostis curvula cv Ermelo as a host plant. There was no difference in total infectivity between North West and Free State substrata, but within regions, there were differences in infectivity between rehabilitation ages, between zones, and between rhizosphere and bare areas. Toepaddock substrata and veld soil produced the highest total infection levels overall. On both dams and veld, total arbuscular levels differed between rhizosphere and bare substrata, and the percentage of arbuscules (max. 15.4%) and vesicles (max. 22.0%) as a proportion of total infection structures was low. A low correlation between infectivity and total spore numbers was also found. Spore numbers and the numbers of viable spores increased with zone down the slimes dams to the veld, and also differed within zones between rhizosphere and bare substrata with marked interactive effects. Substratum organic matter (SOM) levels differed between regions, and between zones within the North West region increasing with distance down the slopes to the veld, and were strongly correlated with total spore numbers. Substratum pH values and most AMF parameters were positively correlated in the order of RV>OV>NV dams, indicating that natural colonization of acidic NV sites by AMF is at very low rates, and that AMF colonizing RV slopes are not surviving the transition from RV to OV, with the associated increase in acidity, conductivity and decline in plant cover. Substratum conductivity differed between zones in both regions, with minor interaction between region and zone, and was negatively correlated with pH, AMF infectivity and total spore numbers. Our findings demonstrate that the ameliorant effects of liming and irrigation on substratum pH and conductivity are short-lived, but despite the physico-chemical constraints, a significant measurable AMF inoculum potential does exist on all substrata. Amelioration of tailings with organic matter and use of acid and salt-tolerant AMF would be expected to contribute to more persistent AMF communities and vegetation on gold and uranium slimes dams.