Revegetation and phytoremediation of tailings from a lead/zinc mine and land disposal of two manganese-rich wastes.

Abstract

The original aims of this project were to investigate the potential for phytoremediation, with emphasis on metal accumulation, of three contrasting industrial processing wastes. These were tailings material (PT) from the decommissioned Pering Pb/Zn Mine (Reivilo, North West Province, South Africa (SA)), smelter slag (SS) from the Samancor Mnsmelter (Meyerton, Gauteng, SA) and electro-winning waste (EW) from MMC (Nelspruit, Mpumalanga, SA). It became evident, however, early in the project, that the use of metal hyperaccumulating plants was not a viable technology for these wastes. The project objectives were thus adapted to investigate alternative remedial technologies. The use of endemic and adapted grass species was investigated to revegetate the PT. In addition, chemically-enhanced phytoremediation was investigated to induce metal hyperaccumulation by grasses grown in the PT (Part 1). Revegetation of the SS and EW were not considered feasible, thus land disposal of these two Mn-rich processing wastes was investigated (Part 2). Part 1 - Revegetation of tailings from Pering Mine The PT was found to be alkaline (pH > 8.0), and consisted mainly of finely crushed dolomite. It was generally nutrient poor with high amounts of readily extractable Zn. It also had a very high P-sorption capacity. Seven grass species (Andropogon eucomus Nees; Cenchrus ciliaris L.; Cymbopogon plurinodis Stapf ex Burtt Davy; Digitaria eriantha Steud; Eragrostis superba Peyr; Eragrostis tef (Zucc.) Trotter and Fingeruthia africana Lehm) were grown in PT treated with different rates of inorganic fertiliser under glasshouse conditions. The fertiliser was applied at rates equivalent to 100 kg N, 150 kg P and 100 kg K ha-1 (full), half the full rate (half) and no fertiliser (0). Seed of C. ciliaris, C. plurinodis, D. eriantha, E. superba and F. africana were collected from Pering Mine. Seed of A. eucomus was collected from the tailings dam of an abandoned chrysotile asbestos mine. These were germinated in seedling trays and replanted into the pots. A commercial variety of E. tef was tested, but due to poor survival this species was subsequently excluded. The foliage and root biomass of the grasses and concentrations of Ca, Cu, Fe, K, Mg, Mn, Pb and Zn in the foliage were determined. The yield of all the grasses increased with an increase in fertiliser rate, with a significant species by fertiliser interaction (p = 0.002). The highest yield was measured for C ciliaris, followed by D. eriantha (4.02 and 3.43 g porI, respectively), at the full fertiliser application rate. Cymbopogon plurinodis was the third highest yielding species, while the yields of E. superba and F. africana were similar. There were positive linear correlations between foliage yield and fertiliser application rate for all grasses. The root biomass of the grasses also increased with an increase in fertiliser application rate. The interaction between grass species and fertiliser level had a non-significant (p = 0.085) effect on the yield of grasses, though there were significant individual effects of species (p < 0.001) and fertiliser (p < 0.001). Digitaria eriantha had the highest root biomass at each fertiliser application rate, followed by C plurinodis and C ciliaris. Similarly to foliage yield, there were positive linear correlations between root biomass and fertiliser application level. Positive, linear correlations were found between foliage yield and root biomass, though the strength of these varied. The weakest correlation was found for D. eriantha (R2 = 0.42) but this was attributed to a moderately high variance in foliage yield and roots becoming potbound. Generally, nutrient concentrations were within adequacy ranges reported in the literature, except for P concentrations. This was attributed to the high P-sorption capacity of the PT. Zinc concentrations were higher than the recommended range for grasses, and also increased with an increase in fertiliser application rate. This was attributed to the high available Zn concentrations in the PT and improved growth of the grasses at higher fertiliser application rates. It was recommended that C ciliaris and D. eriantha be used for revegetation due to high biomass production and that E. superba be used because of rapid growth rate and high self-propagation potential. Both C plurinodis and F. africana can also be used but are slower to establish, while A. eucomus was not a suitable species for revegetation of the PT. Inorganic fertiliser improved the growth of all these species and is recommended for the initial establishment of the grasses. An experiment was conducted to investigate the potential of inducing metal hyperaccumulation in three grass species (C ciliaris, D. eriantha and E. superba) grown in the PT. Grasses were grown in fertilised tailings for six weeks, then either ethylenediaminetetraacetic acid (EDTA) or diethylentriaminepentaacetic acid (DTPA) was added to the pots at rates of 0, 0.25, 0.5, 1 and 2 g kg-I. Grasses were allowed to grow for an additional week before harvesting. The concentrations of Cu, Pb and Zn were determined in the foliage. The interactive effect of species and chelating agent on the uptake of Cu was marginally significant (p = 0.042) and non-significant for Pb and Zn (p = 0.14 and 0.73, respectively). While the addition of the chelating agents resulted in an increase in Pb uptake by the grasses, it did not induce metal hyperaccumulation in the grasses. This was attributed to the ineffectiveness of the chelating agents in the PT in the presence of competing base cations (mainly Ca). The use of this technology was not recommended. Part 2 - Land disposal of Mn-rich processing wastes Chemical characterisation of the SS showed that it was an alkaline (pH > 9.5), Mn-rich silicate (glaucochroite), that generally·had low amounts of soluble and readily extractable metals. Acidic extractants removed high amounts of Mn, Ca and Mg, attributed to the dissolution of the silicate mineral. The EW was highly saline (saturated paste EC = 6 780 mS m,l) with a near-neutral pH. It had high amounts of soluble Mu, NHt+, S, Mg, Ca and Co. The primary minerals were magnetite, jacobsite (MnFe204) and gypsum. The effect of SS and EW on selected chemical properties of six soils was investigated by means of an incubation experiment, and their effect on the yield and element uptake by ryegrass was investigated in selected soils under glasshouse conditions. Five A-horizons (Bonheim (Ba), Hutton (Hu), lnanda (la), Shortlands (Sd) and Valsrivier (Va» and an Ehorizon (Longlands (Lo» were treated with SS at rates of 30, 60, 120,240 and 480 g kg'l and EW at rates of20, 40,80,160 and 320 g kg'l. Soils were incubated at field capacity at 24 QC and sampled periodically over 252 days. The soil pH, both immediately and over time, increased, while exchangeable acidity decreased after the addition of SS to the soils. The pH at the high rates of SS tended to be very high (about 8). The electrical conductivity (EC) of the soils also increased with an increase in SS application rates and over time. The most marked changes tended to occur in the more acidic soils (e.g. la). In the soils treated with EW, there was generally an increase in the pH of the acid soils (e.g. la) while in the more alkaline soils the pH tended to decrease (e.g. Va), immediately after waste application. There was a general decrease in pH over time, with a concurrent increase in exchangeable acidity, due to nitrification processes. The EC of all the soils increased sharply with an increase in EW application rate, attributed to the very saline nature of the EW. Water-soluble Mn concentrations in the soils treated with SS tended to be below measurable limits, except in the acid la. Iron concentrations decreased with an increase in SS application rate and over time for all soils. The water-soluble concentrations of Mn, Ca, Mg and S increased sharply with an increase in EW application rate in all soils. There was also a general increase in Mn concentrations over time. Iron concentrations tended to be low in the EW-treated soils, while Co concentrations increased as EW application rate increased. Exchangeable (EX, 0.05 M CaCh-extractable) concentrations of Fe, Co, Cu, Zn and Ni were low in the SS-treated soils. The concentrations of EX-Mn tended to increase with an increase in SS application rate in the la soil, but generally decreased in the other soils. There was also a decrease over time, attributed to the high pH leading to immobilisation of Mn. The EX-metal concentrations of the EW-treated soils were generally low, except for Mn. The concentrations of EX-Mn increased sharply as EW application rate increased. The contribution of EX-Mn was calculated to range from 209 to 3 340 mg Mn for EW rates of 20 to 320 g kg-I, respectively. In the Lo soil the expected amount of Mn was extracted at the different EW application rates. In the other soils the EX-Mn concentrations were typically higher than expected. This was attributed primarily to the dissolution ofMn from the EW due to the interaction between soil organic matter and the EW. There was generally an increase in EX-Mn concentrations over time, attributed to the decrease in pH of the soils treated with EW. The above-ground biomass production of ryegrass grown in Lo and Hu soils treated with SS increased at low application rates, but decreased again at the highest rates. The reduction in yield was attributed to an increase in soil pH leading to trace nutrient deficiencies. At the lower SS application rates, nutrient concentrations of the ryegrass tended to be within typical adequate ranges reported in the literature. Of concern was the elevated Mn concentration in the ryegrass foliage, though no toxicity symptoms were seen. This was attributed to the dissolution of the silicate mineral due to soil acidification processes and the possible ameliorating effect of high Ca and Si concentrations on Mn toxicity. The growth of ryegrass was generally poor in the Hu soil treated with EW and it did not survive beyond germination in the Lo soil treated with EW. In the Hu soil plants grew well in the 20 and 40 g kg-I EW treatments, but died at the higher rates. In both cases mortality was thought to be due to the high salinity that resulted in toxicity and osmotic stress in the newly germinated seedlings. The improved growth at the lower rates ofEW, in the Hu soil, was attributed mainly to increased N availability. The concentrations of Mn in the foliage were elevated in the soils treated with EW. A pot experiment was conducted to test the effect of applying either humic acid (HA) or compost (at a rate of 20 g kg-I) with lime (at rates of 0, 5 and 10 Mg ha-I) on the growth and nutrient uptake of ryegrass grown in the Hu soil treated with EW at rates of 0, 10, 20 and 40 g kg-I. A basal P-fertiliser was also applied in this experiment. The highest yields were measured in the treatments receiving either HA or compost at the highest application rate ofEW. The addition oflime did not improve the yield of the HA treatments, but did in the compost treatments. Generally, nutrient concentrations were adequate. The Mn concentrations were markedly lower than expected, and this was attributed to the formation of insoluble Mn-P compounds due to the addition of fertiliser. The effect of either HA or compost on Mn concentrations was not marked, but lime reduced Mn uptake. A leaching column experiment showed that, generally, the Mn was not readily leached through a simulated soil profile, though the addition of compost may enhance mobility. There was also evidence to indicate an increase in salinity and that Co concentrations of the leachate may be a problem. These data suggest that soil organic matter may be a very important factor in determining the release of Mn from the wastes, notably the EW. The land disposal of the SS and EW was not recommended at the rates investigated here, as both showed the potential for Mn accumulation in above-ground foliage, even at low application rates, while high application rates negatively impacted on plant growth. It appears that P-compounds may be beneficial in reducing Mn availability in the EW, but further testing is required.