The containment of toxic wastes: I. Long term metal movement in soils over a covered metalliferous waste heap at Parc lead-zinc mine, North Wales

0269-7491(95)00010-0

ELSEVIER

Environmental Pollution, Vol. 90, No. 3, pp. 371-377, 1995 Elsevier Science Ltd Printed in Great Britain 0269-7491/95 $09.50 + 0.00

THE CONTAINMENT OF TOXIC WASTES: I. LONG TERM METAL MOVEMENT IN SOILS OVER A COVERED METALLIFEROUS WASTE HEAP AT PARC LEAD-ZINC MINE, NORTH WALES J i a n m i n Shu* & A. D. B r a d s h a w Department of Environmental and Evolutionary Biology, Liverpool L69 3BX, University of Liverpool, UK

(Received 15 December 1993; accepted 14 February 1995)

Abstract

nomic and effective method of stabilising such wastes, but their toxicity and lack of fertility can restrict plant growth (Bradshaw & Chadwick, 1980). In practice, low fertility problems can be overcome by additions of chemical fertilisers or organic manures. However the toxicity cannot be overcome so easily (Maclean & Dekker, 1976). There are five possible methods that can be applied to overcome the problem of toxicity of heavy metals from mine wastes (Smith & Bradshaw, 1972; Ludeke, 1973; Johnson et al., 1976; Johnson & Bradshaw, 1977; Craze, 1977; Bell et al., 1984; Cairney, 1987a). (1) On the least toxic materials, addition of fertilisers and direct seeding with agricultural or amenity grasses and legumes can be sufficient. (2) Alternatively, to provide a more robust vegetation, addition of fertilisers and direct seeding with native species. (3) On more toxic materials, addition of fertilisers and direct seeding with specific metal tolerant plant cultivars derived from tolerant natural populations. (4) On the most toxic materials, provision of an inert covering material, fertilising and seeding with agricultural or amenity species and legumes. (5) Alternatively, the application of an impermeable barrier layer followed by an inert cover and then fertilising and seeding with suitable species. These methods provide different options which can be chosen in relation to the level of the metals, acidity or similar problems, and the ultimate land use. In particular, if any form of grazing is expected, although a good vegetation cover prevents ingestion of contaminated soil, it is normally important to make sure that the surface vegetation is not in contact with the toxic wastes and is therefore not able to take up metals into its above-ground parts. This can be achieved only if the waste is covered with a layer of inert material. However, soluble toxic elements can move upward through the layer by capillarity (Cairney, 1987b). This can be prevented either by a compacted clay barrier layer under the covering or by providing a covering composed of coarse material so that the pores are too large to allow water rise by capillarity. Because of difficulties

In order to stabilise and contain a toxic metalliferous waste heap at Parc Mine, North Wales, it was covered with 30-40 cm layer of quarry waste in 1977-1978, and sown with a grass/clover seed mixture. This study has examined subsequent metal movement in the cover material and its effect on vegetation. The results, especially when compared with previous observations, give no evidence of upward migration of metals by capillarity in the cover material Sideways movement of leachate, however, appears to be carrying the metals into the cover material on the sloping sides, giving rise to increasing concentrations of heavy metals in the vegetation and dieback in some places. Root growth on the flat top of the heap is greater than on the slope, but the roots have not penetrated the waste and the contents of Pb, Zn and Cd in surface vegetation remain low. Surface covering of toxic waste with coarse materials restricting capillary rise is therefore a valid reclamation technique so long as lateral movement of toxic leachate can be controlled Keywords: Metalliferous waste, land reclamation, heavy metal toxicity, toxic leachates.

INTRODUCTION Plants, animals and man can be seriously affected by heavy metal contamination. The major pathway is from polluted soil; lead, zinc, cadmium and mercury can readily accumulate in soils, particularly in surface soil and become immediately accessible to plant roots (Peterson & Nielson, 1978; Gough et al., 1979). Heavy metal pollution brought about by manufacturing processes can be controlled by reducing emissions. However, it is relatively difficult to control metal pollution arising from mine wastes, because, unless wastes can be removed, they remain as a source from which metals can spread into surrounding areas by wind or water, causing pollution and even the death of adjacent vegetation (Hill, 1977; Shetron & Carroll, 1977). The establishment of a vegetation cover is an ecoPresent address: Chinese Research Academy of Environmental Sciences, Beijing 100012 China. 371

372

Jianmin Shu, A. D. Bradshaw

in finding clay materials and the increased handling required, the former is the more expensive option. A single layer of coarse material will usually be cheaper and easy to find from building wastes, quarry rejects, and even non-toxic wastes from the mine itself. However, in such a system it is possible that capillary rise could still occur or that the plants could root down into the toxic materials. One of the earliest places where this technique was used was at Parc Mine in North Wales. This lead/zinc mine is situated beside a stream in a small wooded valley, Nant Gwydyr (grid ref. SH 787603), with an attractive landscape at an altitude of 100 m, receiving an annual rainfall of approximately 1250 mm. Initially, the covering was found to have worked excellently (EAU, 1983). The present work reports a reinvestigation of its effectiveness 14 years after completion and examines the use of inert coverings more generally. SITE, M A T E R I A L S A N D M E T H O D S

Extraction of lead and zinc ore at Parc Mine, North Wales continued intermittently until 1930. The mine was re-opened in 1952. 8.0 million tons of lead and zinc concentrate were produced from 1952 to 1954 when it was eventually closed. This activity gave a large lagoon-type mine tailings heap about 31m in height and 150 m x 500 m in extent, flat topped with steeply sloping sides and no retaining wall, set on the side of the valley against the edge of a stream (for map see Gao & Bradshaw (1995)). The heap was composed of very fine 'secondary flotation tailings' with high concentrations of lead and zinc (Smith & Bradshaw, 1979). In 1964, it suffered catastrophic erosion in a storm which resulted in a mass of material being carried into the valley of the River Conwy, followed by further continuous erosion. In 1977-1978 it was therefore reshaped and covered with a 25-30 cm layer of a coarse quarry overburden (see Table 1). 12.5 t ha ~ of broiler house litter was spread over the flat top of the heap, and 300 kg ha 1 of I.C.I. No.5 fertiliser was applied to the remainder. The site was sown with a seed mix consisting of the following at a rate of 90 kg ha 1. Festuca rubra var. Merlin - - 60% Festuca rubra var. Highlight - - 10% Agrostis tenuis var. H i g h l a n d - 20% Trifolium repens var. SIO0 - - 10% The grass and clover mixture has been growing well up till now and a flock of sheep with about 15 ewes,

rams and lambs graze the reclaimed area. However, recent inspection showed some areas where signs of surface contamination could be found, although these were restricted to the upper part of the heap where the side slopes more steeply. It therefore seemed possible that some migration of metals might be occurring and causing plant death. If this was a symptom of general changes over the heap as a whole, it would suggest that the site might be gradually deteriorating. The metal contents in the surface covering have therefore been examined. Four profile pits were dug in the cover materials down to the waste itself. Two profiles (T1 and T2) were selected on the fiat top of the heap, and two (SI and $2) on the sloping north side of the heap. Two further profiles ($3 and $4) were also dug in the covering material in the upper part of the heap where there were signs of surface contamination. Samples of the cover material, including all plant roots, were excavated carefully from the sides of the profiles in 5 cm steps, except at the surface where samples were taken at 2.5 cm intervals. Samples of vegetation were also collected from sites T1, T2, S1 and $2. All samples were placed in polyethylene bags and transferred to the laboratory. Each soil sample was divided into two sub-samples, one being used for root measurements, the other for chemical analysis of metals. The water contents and particle sizes were measured from pooled samples from each profile pit. Roots for measurement were hand sorted from samples while they were fresh. After washing, they were air-dried for 2 weeks, and weighed, from which the root weight per gram of covering material was estimated. The total length of the roots per gram of covering material was also measured by the method of Newman (1966). The particle size distribution of the cover material was determined by a series of mesh sieves. Chemical analysis of air-dried samples was made on material passed through a 2-mm mesh sieve. A known weight of material was placed in an acid-washed digestion tube with 10 ml of 4:1 nitric/perchloric acid mixture. Tubes were left overnight for predigestion, then digested at 120°C for 6 h. The digests were made up to 50 ml with double distilled water and filtered through Whatman 42 filter paper. The contents of zinc, lead and cadmium were analysed by atomic absorption spectrophotometry in each of three sub-samples. RESULTS AND DISCUSSION

Table 1. The physical characteristics of the quarry cover material Penetration and distribution of the roots

Water contents (%) T1 T2 S1 $2 $3 $4

19.2 22.2 27-4 29-5 36.3 34.2

Particle size distribution (%) >5 mm 2-5 mm 0.2-2 mm <0.2 mm

40.9 25.3 22.5 11.3

Figure 1 shows the pattern of root growth and penetration into both the surface and the deeper layers of the cover material. There is a marked decrease in root weights and lengths with increase in depth. In all four sampling sites, more than 90 percent of the roots, in both weight and length, occur in the top 10 cm of the cover material. Such a distribution of roots, especially

The containment of toxic wastes." I 20 0-2 2-5 5-10 10-15 15-20 depth 20-25 below surface 25-30 (cm) 30-35 35-40 40-45 45-5O 50-55

,

l li

below surface

{cm)

80

I

I

I

• root length

•

[]

root weight

mine waste begins

I

40 i

I

60 ,

I

0

80 ,

2O

4O

60

80

I

I

I

I

0-2 2-5 5-10 10-15 '1 15-20 20-25 25-30 30-35 .......................... 35-40 40-45 45-50 50-55

I

J

20

depth

6O

I

TI

II

,

0-2 2-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 5@55

40

373

20

T2

mine waste begins ~

40

......................

60

80

I

Sl ! mine waste begins

0-2 2-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55

S2

mine waste begins

Fig. 1. The distribution of roots in the cover material at four different sampling sites on Parc Mine (root length in cm g-i soil, root weight in mg g ~soil). roots of herbaceous species with shallow root systems, is normal. Comparing rooting in the two top sites (T1 and T2), it seems that the distribution of roots is not related to the thickness of the quarry cover. In both cases the bulk of the roots occur in the upper l0 cm. This is also found in the two slope sites (S1 and $2). The critical finding however is that almost no roots reach down to the interface with the mine wastes and none were found in the first sample taken in the mine waste. In T1, the roots reached the mine waste, but the weight was negligible. Figure 1 gives no indication of any sharp cut-off in rooting just above the mine waste. This suggests therefore that rooting is not reaching the waste, rather than reaching the waste and then being prevented from any further penetration by toxicity. The total amount of roots in the top sites was always greater than that in the slope sites. This result was different from that of the earlier trial using tolerant plants in the untreated waste (Smith & Bradshaw, 1979). Under those conditions total growth was generally greater on the slope than on the top. Unameliorated, the wastes on the top of the heap would have given problems of drought and low fertility. Now with a surface cover these effects have been removed.

Movement of heavy metals Figure 2 indicates that the metal contents in the quarry cover are somewhat greater than in normal, uncontaminated soils, similar to those reported in the earlier survey (Johnson & Bradshaw, 1977, EAU,1983). This contamination was perhaps introduced when the cover

material was spread. An alternative hypothesis is that it is due to capillary rise. The mine tailings contain high concentrations of heavy metals which could be a source for either type of contamination. In the profiles, the metal concentrations in the base of the cover material, adjacent to the mine spoil are elevated. But there is a sharp decrease in metal concentrations between this layer and the pure quarry cover layer immediately above. There is some indication, however, that the concentrations of Zn, Pb and Cd at the 0-20 cm depth are usually less than those in deeper layers at sampling sites T1, T2, S1 and $2, but these differences are not significant. From this it can be inferred that capillary rise, by which the heavy metals in the mine tailings achieve an upward movement, is not occurring. At the same time there is no indication that the roots of the grass penetrate into highly contaminated material. However at sites $3 and $4, although there are also no significant differences (p > 0-05) between concentrations of all metals within the soil profile at depths of 0-25 cm, the mean concentrations of metals in the quarry cover are higher than those in the same layers at sites T1, T2, S1 and $2. The mean concentrations of metals in sites S1 and $2 are slightly but not significantly greater than those in T1 and T2 (p > 0.05). This suggests an effect of slope. Sites T1 and T2 were on the flat top of the spoil heap, sites S1 and $2 on the sides of the heap, and sites $3 and $4 on an even steeper slope where the waste had been placed directly on the side of the mountain. It would be reasonable to infer

374

Jianmin Shu, A. D. Bradshaw 0-2 2-5 5-10 10-15 depth 15-20 below 20-25 surface 25-3O (cm) 30-35 35-40 40-45 45-50 50-55

-7 --I -'1

,

T1

,

[] ~ (ppm)

"~1

[] Pb (ppm) • Zn (ppm)

-'I --1 "7 "--'1

mine waste begins

•

I

I

•

1000

0-2 2-5 5-10 10-15 15-20 depth below 20-25 surface 25-30 (cm) 30-35 35-40 40-45 45-5O 50-55

0-2 2-5 5-10 10-15 depth 15-20 below 20-25 surface 25-30 (cm) 30-35 35-40 40-45 45-50 50-55

4,

i

'

II

i

I

•

2000

I,

T2

mine waste begins

i

I

I

'

d

I

I

I

I

1000

2000

3000

Sl

0-2 2-5 5-1o

S2

~

lO-15

15-20 ~ mine waste begins I i

I

I

I

1000

2000

3000

"--'l ---'1 "----I 7

s3

mine waste begins IIIIIII

"7 "--1 -7 --"1 ---i ---i "-'1 ---'!

3000

---I -'-'i -"1 "-1 "-7

0-2 2-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55

I

i

,i

rl

................................

I

I

1000 2000 concentration of metal (ppm)

I

mine waste begins

20-25 25-30 30-35 35-40 40-45 45-50 50-55

0-2 2-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40

i

i

I

I

0

1000

2000

3000

.I~ ~1 ~ ' 1

,

S4 mine waste, b e i n s

= = '

I I I I II

d

I

40-45 t 45-50 50-55 !

o

3000

o;o

20;0

30;0

concentration of metal (ppm)

Fig. 2. Contents of heavy metals in the soil profiles in the cover material and underlying mine waste at six different sampling sites on Parc Mine (metal concentration in mg kg ').

that water from rainfall or ground runoff could permeate through the heap, carrying metals from the mine waste out into the cover by sideways movement. Metal levels in leachate from the heap are appreciable (Gao & Bradshaw, 1995). M e t a l concentrations in vegetation s a m p l e s

The metal concentrations in vegetation samples (Table 2) are elevated compared with vegetation from elsewhere. The Zn and Cd contents in the grasses growing on the slope sites S1 and $2 are significantly elevated. It is understandable that the vegetation growing on the fiat top of the spoil would be likely to obtain less heavy metals from the waste. This matches with the metal levels in the cover profiles. The values for lead however

are not significantly different in T1, SI and $2. The effects of slope seem therefore to be confined to the two soluble elements Zn and Cd. Table 2. Metal concentrations in samples of the surface vegetation (mg kg -t)

Sites

Zn

Pb

Cd

T1 T2 S1 $2

19-3a 19.6a 44-7b 59.8c

9-0a 7.2b 9.3a 10.2a

0-3a 0-6b 1.1c 1. Ic

Within columns for different sampling sites, figures postscripted by different letters differ significantly by T~test with p < 0-05.

The containment o f toxic wastes: I 0-5 5-10 10-15 15-20 20-25 25-30 30-35 depth 35-40 below surface 40-45 46-50 (cm) 50-55 55-60 60-65 65-70 70-75 75-80 80-85 85-90 waste

375

Y Fan Mine i,

[]

Pb

•

Zn

mine waste begins

i

0-5 5-10 10-15 15-20 20-25 25-30 30-35 depth 35-40 below 40-45 surface 45-50 (cm) 50-55 55-60 60-65 65-70 70-75 75-80 80-85 85-90 waste

!

!

5O0O

10000

Minera Mine

mine waste begins .............................

+-

I

!

10000

..........

!

20000

30000

metal concentration (ppm)

Fig. 3. Concentrations of metals in soil profiles on two lead-zinc mine wastes covered with 1 m of colliery spoil (from EAU (1983)) (metal concentration in mg kg l).

PREVIOUS OBSERVATIONS There has been previous work on the durability of the Pare Mine cover, as well as trials elsewhere on the effectiveness of different types of cover. However, this work was carried out a short time after the cover was applied (Johnson et al., 1977; EAU, 1983). Topsoil, unburnt colliery shale and burnt colliery shale were tested for their effectiveness as covers in large field experiments on the wastes of the Minera and Y Fan zinc-lead mines. This showed that after a period of 5 years there were no signs of upward movement of heavy metals from the waste spoils into any of the covers even when their depths varied over a range of 15-90 cm (Fig. 3). Metal concentrations were elevated in the top 50 mm dep in some plots, but this was shown to be due to wind blow of waste from adjacent uncovered areas of mine waste. Table 3. Mean metal concentrations in the surface vegetation sampled in 1980 and 1992 on the top of the heap (mg kg -1 dry weight)

1980 1992

Zn

Pb

Cd

50.0 35.9

5.4 8.9

<0-3 0.8

At Pare Mine, the same investigation indicated no signs of upward movement of metals (Fig. 4) after 3 years (1978-1980). However there was a general elevation of lead and zinc in the cover material, mentioned previously. It was suggested then that this contamination could have occurred during the operation of spreading the material, or that since the original cover material came from a quarry in a zone of mineralisation the metal content of the original material might have been elevated. There were patches of elevated zinc and lead concentrations at certain places within the cover material which supported the idea of mixing during spreading rather than upward migration of metals. Although it cannot be assumed that sampling would have been in exactly the same place, a comparison of the 1980 and 1992 values (Fig. 4) provides a critical test of migration. This shows that there has been little change, again providing evidence that little movement of metals has occurred. The metal concentrations in the surface vegetation in 1980 were relatively low: 13-48 mg kg -~ for zinc, <3.3-13 mg k g ' for lead, and 0.3-0.8 mg kg -I for cadmium. A separate bulk analysis gave Pb 54, Zn 50, Cd 0.3 mg kg 4. These values are in the same range as the present investigation, although present values for S1 and $2 suggest some increase in Pb and Cd (Table 3).

376

Jianmin Shu, A. D. Bradshaw

0-2 2-5 5-10 10-15 15-20 depth below 20-25 surface 25-30 30-35 (cm) 35-40 40-45 45-50 50-55

0-2 2-5 5-10 10-15 depth 15-20 below 20-25 surface 25-30 (cm) 30-35 35-40 40-45 45-50 50-55

erence except in the layer directly in contact with the waste heap. M o v e m e n t by capillarity does not seem to be taking place.

Zn

I

Zn (1980) Zn (1992)

[]

l L

mine waste begins III

.......

I ......

IIIII

"= . . . . . .

~=

!

......

I

I

I

!

1000

200O

3000

40O0

I I

5000

Pb • Pb (198o) [] Pb (1992) I

mine waste begins I

Ij .

.

.

.

.

.

I

!

I

I

I

/

200

400

600

800

1000

o.2 0

2-5 5-10 10-15 depth 15-20 below 20-25 surface 25-30 (cm) 30-35

~ ~ 35-40 40-45 45-50

Cd

•

Cd (1980)

[]

Cd (1992)

mine waste begins .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

50-55 0

I

I

I

I

10

20

30

40

A physical analysis (Table 1), however, suggests that the particle sizes of the quarry material were not too large to prevent the formation of pores where some water could ascend by capillarity. Visual inspection confirmed this. Restriction of upward movement would also have been assisted by the high rainfall experienced at the site. The elevated metal concentration in the quarry cover on the slope sites S1 and $2 match earlier results, suggesting mixing. However, in $3 and $4 levels were noticeably elevated and some areas of vegetation were killed. This suggests that lateral movement of leachate is a factor which can cause increases in metal contents on slopes. This is supported by the water contents in the quarry cover which were greater on slopes than on the fiat top (Table 1). These results are very similar to those reported by Borgeghrd and Rydin (1989) for copper tailings covered with morainic material. However their results suggest a rather greater upward m o v e m e n t of heavy metals. This could have been due to a combination o f the more continental climate of the site and a higher content of fine particles and organic matter in the cover material. Coarse inert cover materials appear to provide a very effective barrier to upward movement of toxic elements in moist temperate climates. They can therefore be used as a long term treatment of toxic metalliferous mine wastes. They will sustain satisfactory plant growth if nutrient deficiencies are relieved. But they do not prevent the downward migration o f water into the waste. The consequences of this are discussed in the following paper (Gao & Bradshaw, 1995).

50

metal concentration (ppm)

Fig. 4. A comparison of the concentrations of metals in soil profiles at the same site on Parc Mine in 1980 (from EAU (1983)) and 1992 (metal concentration in mg kg l).

CONCLUSIONS

As a support for growing vegetation, quarry waste has a weakness in being deficient in nutrients. However this is not difficult to overcome. At Parc Mine, broiler house litter and fertiliser were applied at the outset, followed by some subsequent fertiliser treatment. The clover in the original seed mix will also have been important in the maintenance of adequate nitrogen. It has persisted, forming about 20% of the sward. Quarry waste appears to have behaved excellently as a barrier to upward movement of metals. The evidence provided by chemical analysis suggests that no upward movement of the heavy metals has occurred over the 15 years since the waste was treated. The concentrations of metals in different layers showed no significant diff-

ACKNOWLEDGEMENTS This work was carried out under the auspices o f an Academic Links with China Scheme o f the British Council. The analysis o f the data was considerably assisted by Dr J. Y Guo. We are grateful to Dr M. S. Johnson and Dr R. M. Bell for their comments on a preliminary draft of the paper.

REFERENCES

Bradshaw, A. D. & Chadwick M. J. (1980). The Restoration of Land. Blackwell Scientific Publications, Oxford. Bell, R. M., Parry, G. D. R. & Gildon A. (1984). Isolating metal contaminated waste by covering-potential migration of metals into surface soil and vegetation. Conservation & Recycling, 7, 99-105. Borgeghrd, S.-O, & Rydin, H. (1989). Biomass, root penetration and heavy metal uptake in birch in a soil cover over copper tailings. J. AppL Ecol., 26, 585 95. Cairney, T. (ed.) (1987a). Reclaiming Contaminated Land, Blackie, Glasgow. Cairney, T. (1987b). Soil cover reclamation. In Reclaiming Contaminated Land, ed. T. Cairney, pp. 144-169. Blackie, Glasgow.

The containment o f toxic wastes." I Craze, B. (1977). Restoration of Captains Flat mining area. J. Soil Conserv. Service, 33, 98-105. EAU (Environmental Advisory Unit). (1983). Movement of Metals in Reclaimed Metal Contaminated Ground; Site Survey Report. Department of Botany, University of Liverpool, Liverpool. Gao, Y. & Bradshaw, A. D. (1995). The containment of toxic wastes 2. Metal movement in leachate and drainage at Parc Mine, North Wales. Environ. Pollut., 90, 379-82. Gough, L. P., Shacklette, H. T. & Case, A. A. (1979). Element Concentrations Toxic to Plants, Animals and Man. Geological Survey Bulletin 1466. United States Government Printing Office, Washington, DC. Hill, J. R. C. (1977). Factors that limit plant growth on copper, gold, and nickel-mining wastes in Rhodesia. Trans. Inst. Min. Metall., 86A, 98-109. Johnson, M. S. & Bradshaw, A. D. (1977). Prevention of heavy metal pollution from mine waste by vegetative stabilisation. Trans. Inst. Min. Metall., 86A, 47-55. Johnson, M. S., Bradshaw, A. D. & Handley, J. F. (1976). Revegetation of metalliferous fluorspar mine tailings. Trans. Inst. Min. Metall., 85A, 32-7. Johnson, M. S., McNeilly, T. & Putwain, P. D., (1977). Revegetation of metalliferous mine spoil contaminated by

377

lead and zinc. Environ. Pollut., 12, 261-77. Ludeke, K. L. (1973). Vegetation stabilisation of tailings disposal berms of Pima Mining Company. In Tailing Disposal Today, ed. C. L. Aplin & G. D. Argall, pp. 606-14. Miller Freeman, San Francisco, CA. Maclean, A. J. & Dekker, A. J. (1976). Lime requirement and availability of nutrients and toxic metals to plants grown in acid mine tailings. Can. J. Soil ScL, 56, 27-36. Newman, E. I. (1966). A method of estimating the total length of root in a sample. J. Appl. Ecol., 3, 136-46. Peterson, H. B. & Nielson, R. F. (1978). Heavy metals in relation to plant growth on mine and mill wastes. In Environmental Management of Mineral Wastes, ed. G. T. Goodman & M. J. Chadwick, pp. 297-309. Sijthoff & Noordhoff, The Netherlands. Shetron, S. G. & Carroll, D. A. (1977). Performance of trees and shrubs on metallic mine mill wastes. J. Soil & Water Conserv., 3, 222-5. Smith, R. A. H. & Bradshaw, A. D. (1972). Stabilisation of toxic mine wastes by the use tolerant plant populations. Trans. Inst. Min. Metall., 81A, 230-7. Smith, R. A. H. & Bradshaw, A. D. (1979). The use of metal tolerant plant populations for the reclamation of metalliferous wastes. J. Appl. Ecol., 16, 595-612.