Insights and Lessons Learned From the Long-term ......Tech shale 6.6 7.4

Int. J. Environ. Res., 7(3):633-644,Summer 2013ISSN: 1735-6865

Received 8 Oct. 2012; Revised 2 March 2013; Accepted 17 March 2013

*Corresponding author E-mail:[email protected]

633

Insights and Lessons Learned From the Long-term Rehabilitation ofan Iron ore Mine

Unit for Environmental Sciences and Management, North-West University, Potchefstroom, 2520,South Africa

Liebenberg. D.*, Claassens, S. and Van Rensburg, L.

ABSTRACT:A long-term study was conducted between 1985 and 2003 on rehabilitation trials at Sishen IronOre Mine, South Africa, to identify grass species that would survive in the artificial growth medium appliedto the sites, and to determine the most suitable medium for sustainable vegetation growth. Vegetationestablishment was tested at slopes of 18° and 34° and with five different cover materials. After 17 years ofrehabilitation, investigations showed that weathered limestone sloped at 18° produced the highest percentageof plant cover and least erosion. Sixteen grass species were introduced and identified in the survey, withEragrostis sp. and Cenchrus ciliaris as the dominant species. In 2004, new trials were initiated at the samemine to evaluate the effectiveness of different seeding methods and supplementation of the growth mediumwith organic material. After 4 years, hydro seeding was found to be the best method to distribute seeds evenlyand to ensure uniform vegetation growth. Different engineering designs, such as changing the contour lengthand slope, had little influence on the measured parameters. A total of 28 grass species were identified in thesampling plots, with Enneapogon cenchroides and Cenchrus ciliaris as the dominant ones. Dehydrogenaseactivity was used as a proxy for microbial activity, and a positive association was observed between microbialactivity and percentage organic carbon, emphasising the importance of soil organic matter in the soil developmentprocess.

Key words: Enzymatic activity, Microbial community, Mining, Revegetation, Tailings

INTRODUCTIONSouth Africa is a country rich in natural resources

and minerals, and the mining sector substantiallycontribute to the economy, both in terms of minedmaterials and providing employment. However, miningactivities adversely affects the environment andsurrounding communities (Milton, 2001; Mummey etal., 2002). For example, open-cast mines, such as iron-ore mines, lead to the discharge of waste rock materialon potential grazing lands. The rehabilitation of thiscoarse material is extremely difficult due to aridity, windand nutrient-poor soils (Van Rensburg & Rossouw,2008). Sishen mine is located near Kathu, approximately200 km north-west of Kimberley in the Northern CapeProvince. Climate data over a 52 year period shows anaverage annual rainfall of 455 mm for the area, mostly inthe summer months (Weather Bureau, 1984; KumbaResources, 2006). Sishen mine is the largest single-pit

(uninterrupted) iron ore mine in the world and is knownas a source of high-quality iron. The pit was establishedin 1953 and the total mine production since itsestablishment is estimated at 812 Mt, yieldingapproximately 665 Mt of saleable product. The disturbedexpanse-area at the mine covers over 6193 ha.

The ultimate aim of rehabilitating sites disturbedby mining activities is to return a disturbed site tosome resemblance of its former state, or to asustainable usable condition. However, it isrecognised that this rehabilitated condition will mostprobably not match the original condition and allowoptimal land use of the impacted area (Mulligan, 2002).South African legislation imposes a clear obligationon mining companies to prevent environmental damageand defines specific responsibilities associated withmine closure and rehabilitation. According to the

634

Liebenberg. D. et al.

Minerals and Petroleum Resources Development Act(Act 28 of 2002) section 38 (1) (d), any person who is aholder of a reconnaissance permission, prospectingright, mining right, mining permit or retention permitmust, as far as it is reasonably possible, rehabilitatethe environment affected by the prospecting or miningoperations to its natural or predetermined state or to aland use which conforms to the generally acceptedprinciple of sustainable development. This reviewevaluates rehabilitation practices applied at the Sishenmine over 23 years and makes recommendationstowards the best practices for rehabilitation in thestudied context.

Soil formation and the establishment of vegetationcover are essential for achieving rehabilitation ofdisturbed areas (Carroll et al., 2000). Adverse factorsthat can complicate the establishment of vegetationcover include a soil environment with poor physicalcharacteristics, low levels of plant nutrients and organicmatter, pH extremes and the presence of heavy metals.The elevated topography of tailings and minestockpiles accompanied by climatic conditionscharacteristic of arid and semi-arid areas of southernAfrica also deter the establishment of permanent self-sustaining vegetation cover (Milton, 2001). Theinteraction of re-vegetated plants with the physical,chemical and biological components of the soilenvironment, determine whether vegetation will persiston rehabilitated areas (Van Rensburg et al., 2004).Therefore, it is important, when characterising soilquality, to use a selection of all types of soil propertiesrepresenting soil quality as a whole. These shouldinclude properties that are relevant to the chemical,physical and biological aspects of soil that are mostsensitive to management practices and environmentalstress. Such a holistic approach is crucial to promotethe sustainability of environments by providingaccurate information on the status of the soilecosystem as a whole (Ibekwe et al., 2002).

The significance of microbial communities insustainable soil ecosystems has been acknowledgedfor some time. Accordingly, soil microbial propertieshave often been proposed as timely and sensitiveindicators of soil ecological stress or rehabilitationprocesses (Bandick & Dick, 1999; Badiane et al., 2001;Ibekwe et al., 2002; Tate & Rogers, 2002). The analysisof a variety of soil enzymes gives an indication of thediversity of functions that can be fulfilled by themicrobial community (Brohon et al., 2001). Enzymaticactivities such as those of dehydrogenase, ß-glucosidase, urease, and phosphatase show significantcorrelation with total organic carbon, total nitrogen,water-filled pore space, and heterotrophic bacterial andfungal biomass (Aon & Colaneri, 2001).

Dehydrogenase is present in all microorganisms (Dick,1997) and dehydrogenase activity is regarded as anaccurate measure of the microbial oxidative capacityof soil and therefore of viable microorganisms (Dick,1994; Taylor et al., 2002). Dehydrogenase activity haspreviously been used to estimate the degree ofrecovery of degraded soils (Gil-Sotres et al., 2005). B-Glucosidase activity is useful in the monitoring of soilecosystems for several reasons: it shows low seasonalvariability (Knight & Dick, 2004), plays a central role inthe cycling of organic matter, is the most abundant ofthe three enzymes involved in cellulose degradation,and is rarely substrate limited (Turner et al., 2002).Phosphomonoesterases are involved in organicphosphorus transformations in soil. They have lowsubstrate specificity and are the predominantphosphatases in most soils. They are classified as acidphosphatases and alkaline phosphatases by virtue ofthe pH-optima of activity. Soil microorganisms producealkaline phosphatases while acid phosphatases aremainly produced by plant roots (Criquet et al., 2004).Urease activity is often measured due to its importancein the nitrogen cycle, and it has been widely used toevaluate changes in soil quality due to soil managementprocesses.

MATERIALS & METHODSIn order to address issues on the rehabilitation of

mining waste in the semi-arid and arid areas of SouthAfrica, long-term rehabilitation trials were conductedat the Sishen Iron Ore Mine in 1985. The aim of theoriginal trial was to identify grass species that wouldsurvive in the artificial (non-topsoil) growth mediumapplied to the sites, and to determine the most suitablemedium for sustainable vegetation growth over thelong term (Van Wyk, 1994). Chemical analyses of avariety of potential cover materials available at Sishenin 1986 to serve as a growth medium is shown in Table1. From these, nine different materials and mixtures wereselected according to their nutrient status to evaluatethe establishment of vegetation (Table 2)., A 0.5 m layerof banded ironstone followed by a 1 m layer oflimestone was placed underneath the layer of topsoiland mixed materials, before reaching the iron orediscard. The final evaluation of the trials started in1985 was concluded during 2008.

For the 1987-1990 trials, cover materials wereselected according to results obtained from the levelsurface trials (Table 3).

Ten different treatments were applied to the sites(Table 4). Treatments consisted of a combination ofrehabilitation practices and/or amendments applied toa specific site. First, the ‘engineering’ aspect was takeninto account. A value of 1 indicates a 1 m contour

Int. J. Environ. Res., 7(3):633-644, Summer 2013

635

Material pH Available plant nutrients (ppm) Texture H2O KCl P K Ca Mg Na Coarse Fine Silt Clay Lime 6.8 7.8 <3 33 3211 245 19 34 622 39 5 Red clay 6.4 7.4 <3 152 6310 1557 38 1 19 80 0 Lawa 6.7 8.2 <3 51 1822 124 22 70 28 1 1 Quartzite 6.8 7.9 <3 10 68 9 8 72 27 1 0 Tech shale 6.6 7.4 <3 37 646 346 33 1 23 50 26 Flagstone shale 5.9 6.7 <3 66 141 33 14 64 26 9 1 Black shale 5.7 6.3 <3 80 204 88 14 66 30 4 0 Brown shale 5.9 6.8 <3 173 152 32 10 76 21 2 1 Fine ore 5.9 6.8 <3 54 41 52 54 74 16 9 1 Silt 6.3 7.3 <3 112 1909 264 543 2 7 62 29 Discharge 6.7 7.8 <3 39 419 63 15 72 26 1 1 Diabase 7 8 <3 123 10371 1739 47 9 46 18 27 Red topsoil 5.2 6.7 <3 262 2681 1271 67 26 35 12 27 Dark topsoil 6.2 7.7 <3 343 3666 154 42 18 61 10 11

Table 1. Chemical analyses of the different materials available at Sishen Iron Ore mine

Plot Material Description 1 Fine limestone - 2 Fine iron ore discard - 3 50% fine limestone + 50% discard Reduces the silt content of the limestone,

improves infiltration and increases the nutrient status of the discard

4 70% limestone + 30% discard Reduces the silt content of the limestone, improves infiltration and increases the

nutrient status of the discard 5 70% limestone + 30% discard + 50mm red

topsoil This mixture consists of more fine

material and microorganisms 6 90% limestone + 10% red topsoil Improves texture and contributes fine

material and microorganisms to the limestone

7 90% limestone + 10% dark topsoil Improves texture and contributes fine material and microorganisms to the

limestone 8 70% discard + 30% diabase Improves texture and nutrient status of the

discard 9 50% discard + 30% limestone + 20% diabase Improves texture and nutrient status of the

discard

Table 2. Different materials and mixtures selected in 1986 for level-surface trails to evaluate potential covermaterials

Material 2:3:2 (22) Super phosphate KCl MgO Un-weathered limestone 200 400 600 - Weathered limestone 200 400 300 - Sandy soils + stone 200 400 - - Fine iron ore discard + stone 200 400 200 100 Fine iron ore discard + stone + diabase 200 400 1000 -

Table 3. Different cover materials selected were fertilised according to chemical analysis

636

Long-term rehabilitation of an iron ore mine

length and 2 indicates a 2 m contour length. Thisincluded sites that were physically altered during therehabilitation process by creating contours at varyingdistances from one another. The second treatmentaspect considered was whether hydro seeding or handseeding was applied to a site. Thirdly, treatments eitherincluded the application of a topsoil layer or not. Thelast two columns of Table 5 show whether organicmaterial and/or fertiliser were applied to the sites. Theabsence and presence of these treatment constituentsare indicated by 0 (absence) or 1 (50 t/ha). A value of 2indicates that double the amount was applied and avalue of 3 that triple the amount was applied to thesites.

Plots for the 2004 trials were laid out on twoterraces, three different slope angles were used in theinvestigation (Table 5).

RESULTS & DISCUSSIONThe results of the vegetation properties (% basal

cover) derived from the 1986–1990 slope trials areillustrated in Fig. 1. The irrigated 18 ° slopes had better

Treatment Engineering Hydro seed Topsoil Organic Fertiliser 1 1 0 0 0 0 2 1 0 0 1 0 3 1 0 0 1 1 4 2 0 1 0 0 5 2 1 1 2 2 6 2 0 0 1 1 7 2 1 0 1 1 8 2 1 0 2 3 9 2 1 0 2 1 10 2 0 0 1 1

Upper Slope

Treatment T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T1 T2 T3 T10

Slope angle (°) 24 24 24 24 24 24 24 24 24 24 34 34 34 34 Lower slope

Treatment T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T4 T5 T6 T7 T8 T9

Slope angle (°) 18 18 18 18 18 18 18 18 18 18 34 34 34 34 34 34

grass cover than the non-irrigated 34 ° slopes. In mostof the trials the basal cover percentage increased from1988 to 1990 with Cenchrus ciliaris as the dominantgrass species.

The 34 ° slopes remained mostly barren in 1990,with 60–80 % of the trial plots non-vegetated. On thelevel-surface trials it was found that fine iron orediscard, topsoil and a mixture of 70 % discard and 30% diabase, produced the best initial results forrevegetation with Eragrostis achinochloidea andCenchrus ciliaris being the dominant grass species.The 90 % limestone and 10 % dark topsoil producedthe poorest results with 63.3 % bare ground in the trialplots. Results obtained in 2003 from the 1985 trial plotsindicated that there was a difference in thedehydrogenase activity for the various treatments after17 years of rehabilitation with 50 % diabase, 50 %discard, and irrigation outperforming the othertreatments in terms of vegetation establishment andmicrobial activity. Multivariate ordination methods canbe used to determine how community compositionvaries with environmental variables, using correlation

Table 5. Site layout with different treatments as identified in November 2004

Table 4. Different treatments applied to the sites at Sishen Iron Ore Mine in 2004

637

Int. J. Environ. Res., 7(3):633-644, Summer 2013

coefficients to analyse the data (Lepš & Šmilauer, 2003).Principal components analysis (PCA) ordinationconsists of two variables that are ordinated together(bi-plots), whereas canonical correspondence analysis(CCA) ordinations consist of three variables that areordinated for comparison (tri-plots). Ordinationdiagrams can consist of points representing speciesand sampling sites, and vectors for quantitativeenvironmental variables. Each species point in thediagram is positioned to represent the average of thesampling site points in which it occurs. Individualvectors represent the quantitative value for eachenvironmental variable. The length of each vectorindicates the gradient of the variables with theordination axes. Dissimilarities are associated with thedistance between the samples; the closer samples areto each other or to a vector, the higher the similaritybetween them. The linear correlation coefficients canbe indicated by the angles between the vectors.Vectors pointing in the same direction have a largepositive correlation, whereas vectors pointing inopposite directions indicate large negativecorrelations. The axes of an ordination diagramdistinguish between different ecological gradientswithin the samples (Ter Braak & Verdonschot, 1995).

Results from the 2007 survey are illustrated in thePCA (Fig. 2) of the relationship between the differentenzymatic activities and the different treatmentsapplied since 2004 (Table 1). The 18 ° (A) sites groupedtogether in the lower part of the ordination togetherwith urease and acid phosphatase activity. The 24 °(B) sites grouped together in the top part of theordination and was associated with ß-glucosidase. The34 ° sites (C) were spread across the top part of theordination and sites C5 and C6 mostly associated withalkaline phosphatase and dehydrogenase activity.

Results from the 2008 survey are illustrated in theCCA (Fig. 3) of the relationship between grass speciesand the different treatments of the sites. Grass speciesgrouped together with the slopes on which itestablished successfully. Sites treated similarly alsogrouped together in the diagram. There was adifference in the dehydrogenase activity for the 10various treatments applied to the different sites.Treatment 5 performed the best, with the addition oftopsoil, organic matter and fertiliser producing thehighest dehydrogenase activity.

Over the period that the rehabilitation trials wereconducted, Cenchrus ciliaris was the dominant grassspecies. In the microbial assays of both 2004 and 2008,treatment 5 (engineered with contours, with theaddition of topsoil, fertiliser, organic material and hydroseeded) had the highest dehydrogenase activity withthe addition of topsoil, organic matter and fertiliserindicating that these sites were the most successful.The number of grass species found in the survey of2008 was high, indicating that other species weremoving in from the surrounding environment. Fourteenspecies were seeded in 2004 and 28 species wereidentified in the survey after 4 years. Topsoil is thepreferable treatment for establishment of vegetationand increased microbial activity (Claassens et al.,2011). However, topsoil is particularly costly anddifficult to obtain and the extensive surface area ofthis mine makes the application of topsoil as a standardrehabilitation practice for mine closure highlyunfeasible. To establish the most effective and efficientway to stabilise and vegetate the slopes of the wastedumps, the mine authorities researched and testeddifferent combinations of soil, slope and vegetation toimprove the aesthetics of the pit and to prepare formine closure in the long term. An overview ofrehabilitation practices and outcomes for 23 years ispresented in Table 6.

Fig. 1. Bar graphs indicating percentage basal cover for sites not irrigated in 1988 and 1990; and percentagebasal cover for sites irrigated in 1988 and 1990

638


Fig. 2. Principal component analysis ordination diagram illustrating the relationship between microbialenzymatic activities and the different treatments in 2008

Fig. 3. Canonical correspondence analysis ordination diagram illustrating the relationship between grassspecies and different treatments of sites. Eigenvalues for the first two axes were 0.136 and 0.071, respectively

639

Int. J. Environ. Res., 7(3):633-644,Summer 2013

Tim

e per

iod

Met

hods

Se

ed m

ixtu

re

Slop

e A

im

Mai

n co

nclu

sions

and

out

com

es

1985

Th

ree

waste

rock

plo

ts us

ing

diffe

rent

cov

er m

ateria

ls: L

ime

mat

erial

, red

Kala

hari

sand

and

un

treat

ed ir

on o

verb

urde

n

Natu

ral

angl

e of

repo

se (3

4 °)

Iden

tify g

rass

sp

ecies

that

wou

ld

surv

ive i

n th

e ar

tifici

al (n

on-

tops

oil)

grow

th

med

ium

.

Kala

hari

sand

was

the

best

mate

rial b

ut er

oded

agai

nst t

he

steep

slop

es. A

ltern

ative

mate

rial n

eede

d to

be te

sted

but a

t lo

wer s

lope

s bec

ause

Kala

hari

sand

ero

sion w

as to

o hi

gh an

d lim

esto

ne w

as n

ot su

stain

able

for v

egeta

tion

esta

blish

men

t

(Van

Wyk

, 200

2).

1986

-199

0 Ch

emica

l ana

lysis

of a

var

iety

of

poten

tial c

over

mat

erial

s ava

ilabl

e on

the m

ine (

Tabl

e 1) t

o se

rve a

s a

grow

th m

ediu

m.

Leve

l su

rface

Ev

aluate

pot

entia

l co

ver m

ateria

ls to

as

sist i

n th

e es

tabl

ishm

ent o

f ve

getat

ion.

Nin

e diff

eren

t mat

erial

s and

m

ixtu

res w

ere s

elec

ted a

ccor

ding

to

their

ava

ilabi

lity a

nd n

utrie

nt

statu

s (Ta

ble 2

) and

wer

e app

lied

as co

ver l

ayer

s of 2

0 – 3

0 cm

on

the

surfa

ces o

f the

tria

l plo

ts.

Pl

ots w

ere f

ertil

ised

and

seed

ed in

Ja

nuar

y 19

86

Pl

ots w

ere f

ertil

ised

with

:

Cenc

hrus

cilia

ris,

Enne

apog

on ce

nchr

oide

s, Er

agro

stis e

chin

ochl

oide

a,

Chlo

ris vi

rgat

a, T

ragu

s be

rtero

nian

us, A

ristid

a co

nges

ta, A

ristid

a ad

scen

sioni

s, Sc

hmid

tia

papp

opho

roid

es,

Schm

idtia

kalih

arien

sis,

Stip

agro

stis u

nipl

umis

, St

ipag

rosti

s obt

use,

M

elini

s rep

ens

Gras

s spe

cies f

ound

in al

l of t

he p

lots

incl

uded

: Enn

eapo

gon

cenc

hroi

des,

Chlo

ris v

irgat

a, E

ragr

ostis

ach

inoch

loid

ea,

Cenc

hrus

cilia

ris. S

also

la ka

li wa

s the

dom

inan

t wee

d. T

he

gras

s cov

er in

the

level

surfa

ce tr

ials

wer

e dom

inate

d by

Er

agro

stis e

chin

ochl

oide

a an

d Ce

nchr

us ci

liaris

(Van

Wyk

, 20

02).

Fine

iron

ore

disc

ard,

tops

oil,

and

a mix

ture

of 7

0%

disc

ard a

nd 3

0% d

iaba

se, p

rodu

ced

the b

est i

nitia

l res

ults

for

reve

getat

ion.

The

conc

lusio

n re

ache

d, w

as th

at th

e use

of

tops

oil w

as in

effe

ctiv

e and

disc

ard

coul

d be

used

as a

cov

er

mate

rial o

n le

vel s

urfa

ces a

s it a

chie

ved

the s

ame r

esul

ts an

d wa

s mor

e cos

t-effe

ctive

. The

seed

mix

ture

shou

ld b

e ada

pted

to

incl

ude m

ore s

peci

es fr

om th

e im

med

iate

vicin

ity o

f the

m

ine (

Van

Wyk

, 200

2).

Veg

etati

on es

tablis

hmen

t was

co

mpa

red

with

5 d

iffer

ent c

over

m

ater

ials a

pplie

d 15

0 – 2

00 m

m

thic

k (V

an W

yk, 2

002)

.

Cenc

hrus

cilia

ris,

Cyno

don d

acty

lon,

An

thep

hora

pub

esce

ns,

Arist

ida

cong

esta

, Aris

tida

adsc

ensio

nis,

Erag

rosti

s le

hman

nian

a, E

ragr

ostis

ec

hino

chlo

idea

, En

neap

ogon

cenc

hroi

des,

Chlo

ris vi

rgat

a,

Stip

agro

stis u

nipl

umis

, St

ipag

rosti

s cili

ata,

St

ipag

rosti

s obt

usa,

Tr

agus

ber

tero

nian

us,

Fing

erhu

tia af

rican

a,

Uro

chlo

a br

achy

ura,

Sc

hmid

tia pa

ppop

horo

ides

Use o

f diff

eren

t co

ver m

ateria

ls to

de

term

ine t

he m

ost

effe

ctive

again

st slo

pe an

gles

1988

: 18

° slo

pes h

ad b

etter

gras

s cro

wn

cove

r tha

n th

e 34 °

slo

pe an

d irr

igat

ed sl

opes

had

bet

ter g

rass

estab

lishm

ent t

han

non-

irrig

ated

slop

es. 1

990:

the

basa

l cov

er o

n th

e irri

gated

18

° slo

pes t

reat

ed w

ith un

-wea

ther

ed li

me

rema

ined

cons

tant a

t 2.

6%. I

n m

ost o

f the

oth

er pl

ots t

he b

asal

cove

r inc

reas

ed

from

2 to

4% in

198

8 an

d fro

m 4

to 8.

5% in

199

0. In

all

of th

e pl

ots C

ench

rus c

iliar

is wa

s the

dom

inan

t gra

ss in

199

0. T

he

34 °

slope

s rem

ained

barre

n in

199

0 an

d 60

– 8

0% w

as n

on-

vege

tated

. In

man

y of t

he p

lots

the u

pper

and l

ower

third

of

the s

lope

was

ero

ded.

The

34

° slo

pe co

vere

d wi

th di

abas

e and

sto

ne ex

perie

nced

the

least

eros

ion,

alth

ough

the

vege

tatio

n co

ver w

as sp

arse

. Mos

t of t

he p

lots

wer

e dom

inat

ed b

y th

e we

ed S

also

la ka

li. S

pecie

s dom

inat

ing t

he g

rass

cove

r wer

e En

neap

ogon

cenc

hroi

des,

Cyno

don

dact

ylon,

Cenc

hrus

cil

iaris

and

Anth

epho

ra p

ubes

cens

(Tab

le 2

) (V

an W

yk,

2002

).

Tabl

e 6. O

verv

iew

of r

ehab

ilita

tion

prac

tices

and

out

com

es a

t Sish

en Ir

on O

re m

ine

640


Tw

o se

ts o

f pl

ots

wer

e la

id

out,

the

first

con

sist

ed o

f 10

plot

s, fi

ve 1

8 °

and

five

34

°.

The

sec

ond

set o

f plo

ts w

as

iden

tical

to th

e fi

rst e

xcep

t th

at it

was

irri

gate

d du

ring

the

dry

seas

on (V

an W

yk, 2

002)

.

A r

eplic

ate

of e

ach

trial

irri

gate

d to

de

term

ine

the

vege

tatio

n co

ver i

f th

e tr

ial o

n th

e sl

opes

w

ithou

t irr

igat

ion

was

no

t suc

cess

ful.

Met

hods

use

d fo

r veg

etat

ion

surv

ey in

198

6-19

90 o

n le

vel a

nd s

lope

tria

ls

Cro

wn

cove

r of i

ndiv

idua

l spe

cies

as

wel

l as

tota

l cro

wn

cove

r wer

e de

term

ined

by

plac

ing

0.5

m² s

urve

y ci

rcle

s ran

dom

ly in

eac

h pl

ot. B

raun

-Bla

nque

t sca

le w

as u

sed

to d

eter

min

e cr

own

cove

r val

ues.

D

ry m

ass

was

det

erm

ined

usi

ng su

rvey

cir

cles

and

cut

ting

the

vege

tatio

n w

ithin

the

circ

les.

The

mat

eria

l rec

over

ed w

as d

ried

at 7

5 °C

and

wei

ghed

.

Sinc

e 19

90 th

e M

ON

ITO

R-p

rogr

am o

f the

Dep

artm

ent P

lant

and

Soi

l Sci

ence

of

Potc

hefs

troo

m U

nive

rsity

was

use

d fo

r ve

geta

tion

surv

eys.

With

this

pro

gram

bas

al c

over

and

sp

ecie

s fre

quen

cy w

as d

eter

min

ed (V

an W

yk, 1

994)

. 20

03

In 2

003,

the

plot

s of

198

7 w

ere

asse

ssed

for

the

last

tim

e by

mea

ns o

f ve

geta

tion

surv

eys

and

the

anal

ysis

of

soil

mic

robi

al a

ctiv

ity b

y m

eans

of e

nzym

atic

act

ivity

. So

il en

zym

atic

act

iviti

es w

ere

assa

yed

acco

rdin

g to

stan

dard

pr

oced

ures

(Ale

f and

N

anni

pier

i, 19

95).

Ass

ays

are

base

d on

the

incu

batio

n of

soil

with

a sp

ecifi

c su

bstr

ate,

fo

llow

ed b

y th

e co

lorim

etric

es

timat

ion

of th

e re

actio

n pr

oduc

t. A

ll en

zym

atic

ac

tiviti

es w

ere

dete

rmin

ed

usin

g ai

r-dr

ied

soil,

exc

ept

dehy

drog

enas

e, w

here

soi

l w

as k

ept a

t fie

ld-w

ater

co

nten

t.

18 °

and

34

°

Aft

er 1

7 ye

ars

of r

ehab

ilita

tion,

wea

ther

ed f

ine

limes

tone

and

fin

e di

scar

d ap

plie

d as

gro

wth

med

ium

pro

duce

d th

e be

st p

lant

co

ver.

Thes

e re

sults

als

o in

dica

ted

that

wea

ther

ed l

imes

tone

se

t at

18

° pr

oduc

ed th

e hi

ghes

t per

cent

age

of p

lant

cov

er a

nd

less

ero

sion

tha

n th

e sl

opes

at

34 °

(V

an R

ensb

urg

et a

l.,

2004

). R

esul

ts o

btai

ned

from

the

sam

plin

g of

the

198

7 tri

al

site

s in

dica

ted

the

pres

ence

of

m

icro

bial

ac

tivity

in

th

e ar

tific

ial

(non

-top

soil)

gr

owth

m

ediu

m.

The

re

was

no

si

gnifi

cant

diff

eren

ce b

etw

een

the

two

diff

eren

t sl

opes

bas

ed

on e

nzym

atic

act

ivity

.

The

resu

lts d

id n

ot a

ddre

ss o

ther

issu

es o

f reh

abili

tatio

n su

ch a

s th

e ap

plic

atio

n of

inor

gani

c fe

rtili

ser t

o ai

d pl

ant g

row

th o

r the

pro

blem

of l

ack

of w

ater

-hol

ding

cap

acity

of t

he

grow

th m

ediu

m. N

ew s

tudi

es w

ere

initi

ated

to e

valu

ate

the

effe

ctiv

enes

s of

dif

fere

nt s

eedi

ng m

etho

ds a

nd th

e ef

fect

of t

he a

dditi

on o

f or

gani

c m

ater

ial t

o th

e gr

owth

med

ium

.

641


2004

Pl

ots w

ere l

aid o

ut o

n two

terra

ces o

f an

east-

faci

ng sl

ope,

with

15 p

lots

on th

e bo

ttom

slop

e an

d 13

plo

ts on

the u

pper

slo

pe. T

hree

diff

eren

t slo

pe an

gles

wer

e us

ed in

the i

nves

tigat

ion

(Tab

le 4)

. Ten

di

ffere

nt tr

eatm

ents

wer

e app

lied

to si

tes

on th

e bot

tom a

nd th

e upp

er sl

opes

(Tab

le 5)

. Tre

atmen

ts co

nsist

ed o

f a co

mbi

natio

n of

reha

bilit

atio

n pra

ctice

s and

/or

amen

dmen

ts ap

plied

to a

spec

ific s

ite.

Sites

were

mon

itore

d usin

g ve

geta

tion

surv

eys.

The

grou

nd a

nd cr

own

cove

r at

all t

he si

tes w

ere e

stima

ted

in th

ree 1

m²

quad

rates

rand

omly

plac

ed ov

er a

50

m

trans

ect (

Van

Rens

burg

et a

l., 2

004)

.

Anth

epho

ra p

ubes

cens

, Ar

istid

a ad

scen

sioni

s, Ar

istid

a co

nges

ta,

Cenc

hrus

cili

aris,

Ch

loris

virg

ata,

Cy

nodo

n da

ctylo

n,

Digi

taria

eria

ntha

, Er

agro

stis

lehma

nnia

na,

Erag

rosti

s tef,

Fi

nger

huth

ia af

rican

a,

Hete

ropo

gon

cont

ortu

s, Sc

hmid

tia

papp

opho

roid

es, a

nd

Trag

us b

erter

onia

nus

18 °,

24 °

and

34 °

Evalu

ation

of

diffe

rent

seed

ing

meth

ods a

nd

appl

icati

on o

f or

gani

c mate

rial

The

vege

tatio

n su

rvey

indi

cated

that

the

treat

men

ts on

the 2

4 ° s

lope

per

form

ed th

e bes

t and

was

less

ero

ded

than

the 3

0 °

slope

s. Ar

eas t

hat w

ere h

ydro

seed

ed ou

t-per

form

ed h

and

sown

are

as an

d we

re c

hara

cter

ised

by a

high

er sp

ecies

fre

quen

cy. A

reas

with

tops

oil a

nd tr

eate

d wi

th a

high

er

appl

icat

ion

rate

of o

rgan

ic m

ater

ial g

ener

ally

per

form

ed

bette

r tha

n tre

atmen

ts wi

th lo

wer

org

anic

treat

ment

. Diff

eren

t en

gine

erin

g des

igns

did

not

hav

e a m

easu

rabl

e inf

luen

ce o

n th

e veg

etatio

n co

mpo

sitio

n an

d co

ver i

n the

trea

tmen

ts.

Salso

la ka

li w

as th

e do

min

ant i

nvad

er sp

ecies

in th

e ex

perim

ental

plot

s and

Era

gros

tis sp

. and

Cen

chru

s cili

aris

wer

e the

dom

inan

t gra

ss sp

ecies

(Van

Ren

sbur

g et a

l., 20

04).

The

textu

re o

f the

mat

erial

diff

ered

notic

eabl

y be

twee

n th

e 18

° and

24

° slo

pe tr

eatm

ents,

whi

ch h

ad a

visib

le eff

ect o

n se

edlin

gs. S

eedl

ing

surv

ival

was

ther

efor

e a p

oten

tial

prob

lem

espe

cially

for p

artic

ular

spec

ies s

uch

as

Hete

ropo

gon

cont

ortu

s. 20

06

Veg

etati

on su

rvey

– V

isual

obs

erva

tions

af

ter g

ood d

ownp

ours

of ra

in d

urin

g ea

rly

Janu

ary 2

006

(75 m

m) o

n th

e 200

4 sit

es

(Van

Wyk

, 200

6).

18 °,

24 °a

nd

34 °

Evalu

ate v

eget

ation

es

tabl

ishm

ent o

n th

e di

ffere

nt tr

eatm

ents

and

slope

s afte

r two

gr

owin

g se

ason

s.

From

the

resu

lts o

f the

veg

etatio

n sur

vey

in 2

006,

it wa

s cle

ar

that

the

addi

tion

of to

psoi

l play

ed an

impo

rtant

role

as a

seed

so

urce

, and

it m

ust b

e con

serv

ed d

urin

g th

e min

ing

proc

ess

for r

ehab

ilitat

ion.

Con

clus

ions

reac

hed

durin

g th

is st

udy

indi

cated

Cen

chru

s cili

aris

to b

e the

dom

inan

t gra

ss sp

ecie

s, w

hile

Hete

ropo

gon

cont

ortu

s, D

igita

ria er

iant

ha,

Ant

heph

ora p

ubes

cens

and

Fin

gerh

uthi

a af

rican

a we

re

pere

nnia

l gra

sses

that

coul

d pl

ay a

n im

porta

nt ro

le in

the

reha

bilit

ation

proc

ess.

Mel

inis

repe

ns, E

nnea

pogo

n ce

nchr

oide

s and

Aris

tida

adsc

ensio

nes a

re an

nual

spec

ies th

at

wer

e pe

rform

ing

well

(van

Wyk

, 200

6). C

onto

ur fu

rrows

pl

ayed

an

impo

rtant

role

on b

oth

the 1

8 ° a

nd 2

4 ° s

lope

s be

caus

e it c

reate

d a f

avou

rabl

e env

ironm

ent f

or th

e es

tablis

hmen

t of v

eget

ation

that

serv

ed a

s a se

ed so

urce

for

the d

ownw

ard

slope

. It w

as a

lso c

onclu

ded

that

am

elior

atio

n is

impo

rtant

in th

e esta

blish

men

t of h

ealth

y veg

etati

on. W

ith

an in

crea

se in

per

enni

al gr

asse

s, th

ere w

as a

decl

ine i

n the

fre

quen

cy o

f Sal

sola

kal

i (va

n W

yk, 2

006)

.

642


2007

Ev

alua

te s

oil e

nzym

atic

act

iviti

es:

Deh

ydro

gena

se, β

-glo

cosi

dase

, al

kalin

e an

d ac

id p

hosp

hata

se a

nd

urea

se (A

lef &

Nan

nipi

eri,

1995

). St

atis

tical

ana

lyse

s we

re p

erfo

rmed

us

ing

Can

oco f

or W

indo

ws

4.5

(Bio

met

ris –

Pla

nt R

esea

rch

Inte

rnat

iona

l, W

agen

inge

n, T

he

Net

herl

ands

) (T

er B

raak

et a

l., 1

998)

to

inve

stig

ate

the

relat

ions

hip

betw

een

the

diff

eren

t enz

ymes

and

site

tre

atm

ents

usi

ng a

pri

ncip

al

com

pone

nt a

naly

sis (

PC

A)

mul

tivar

iate

ord

inat

ion

tech

niqu

e.

18 °

, 24

° an

d 34

° Ev

alua

te th

e di

ffer

ent t

reat

men

ts

usin

g so

il en

zym

atic

activ

ities

as

ass

essm

ent

crite

ria o

n th

e di

ffer

ent s

lope

s.

The

resu

lts sh

ow th

at th

e m

ajor

ity o

f the

trea

tmen

ts ap

plie

d to

the

site

s on

the

24 °

slo

pe h

ad a

hig

her

corre

latio

n w

ith β

-glu

cosi

dase

and

dehy

drog

enas

e ac

tiviti

es, i

rres

pect

ive

of th

e tr

eatm

ents

app

lied

(Fig

ure

2). F

urth

erm

ore,

site

s on

the

30° s

lope

trea

ted

with

to

psoi

l, or

gani

c m

atte

r an

d fe

rtilis

er a

lso

show

ed h

ighe

r β-

gluc

osid

ase

and

dehy

drog

enas

e ac

tiviti

es, w

hich

mig

ht

cont

ribu

te to

a fa

ster

rec

over

y of

iron

ore

ove

rbur

den.

4

Trea

tmen

ts ap

plie

d to

the

18 °

slo

pes

indi

cate

d a

stro

nger

as

soci

atio

n w

ith u

reas

e an

d ac

id p

hosp

hata

se a

ctiv

ity.

A

cid

phos

phat

ase

activ

ity is

an

impo

rtan

t sou

rce

of p

lant

ro

ots

and

it w

as a

ntic

ipat

ed th

at a

n 18

° slo

pe is

mor

e su

itabl

e fo

r the

esta

blis

hmen

t of v

eget

atio

n. H

owev

er th

e tre

atm

ent a

pplic

atio

ns o

n th

e 24

° sl

ope

can

be

cons

ider

ed to

be

mor

e st

able

ove

r th

e lo

ng te

rm d

ue to

a

stro

nger

cor

rela

tion

of e

nzym

atic

act

iviti

es th

at in

dica

ted

high

er m

icro

bial

activ

ity (V

an R

ensb

urg

& R

ouss

ow,

2008

).

2008

Si

tes

wer

e m

onito

red

usin

g ve

geta

tion

surv

eys a

nd so

il en

zym

atic

act

iviti

es

(Ale

f & N

anni

pier

i, 19

95).

The

gro

und

and

crow

n co

ver o

f all

the

sites

wer

e es

timat

ed in

thre

e 1

m² q

uadr

ates

ra

ndom

ly p

lace

d ov

er a

50

m

trans

ect.7 T

otal

orga

nic

carb

on w

as

dete

rmin

ed a

ccor

ding

to th

e W

alkl

ey-

Bla

ck p

roce

dure

(W

alkl

ey &

Bla

ck,

1934

). St

atis

tical

ana

lyse

s w

ere

perf

orm

ed to

inve

stig

ate

the

rela

tions

hip

betw

een

the

diffe

rent

en

zym

es a

nd si

te tr

eatm

ents

usi

ng a

cano

nica

l cor

resp

onde

nce

anal

ysis

(C

CA

) (C

anoc

o fo

r Win

dows

4.5

) m

ultiv

aria

te o

rdin

atio

n te

chni

que.

18 °

, 24

° an

d 34

° Ev

alua

te th

e di

ffer

ent t

reat

men

ts

usin

g so

il en

zym

atic

activ

ities

an

d ve

geta

tion

cove

r as

asse

ssm

ent

crite

ria o

n th

e di

ffer

ent s

lope

s

Aft

er fo

ur y

ears

of r

ehab

ilita

tion

on th

e ne

w tr

ials

, it w

as

foun

d th

at d

iffer

ent e

ngin

eeri

ng d

esig

ns, i

.e. s

lope

leng

th

and

slop

e, di

d no

t hav

e a

sign

ifica

nt in

flue

nce

on

vege

tatio

n co

mpo

sitio

n an

d sp

ecie

s di

vers

ity. G

row

th

med

ium

was

not

hom

ogen

eous

ly a

pplie

d ov

er th

e si

tes,

an

d th

us le

d to

a v

arie

ty o

f mat

eria

ls fo

und

from

the

botto

m o

f the

slo

pe to

the

top.

Site

s with

no

or li

ttle

addi

tion

of o

rgan

ic m

atte

r and

fert

ilise

r sho

wed

a

corre

spon

ding

ly lo

w e

nzym

atic

act

ivit

y be

caus

e of

the

rock

y m

ater

ial.

Trea

tmen

t 5 h

ad th

e hi

ghes

t de

hydr

ogen

ase a

ctiv

ity, d

ue to

the

addi

tion

of to

psoi

l, or

gani

c m

atte

r and

fer

tilis

er (

Figu

re 3

). C

ench

rus c

ilia

ris,

Enne

apog

on c

ench

roid

es a

nd H

eter

opog

on c

onto

rtus

w

ere

the

mos

t dom

inan

t gra

ss sp

ecie

s in

all

site

s (F

igur

e 3)

.

CONCLUSIONThe main conclusion reached about the

rehabilitation at the Sishen mine was that the additionof topsoil together with organic matter and fertilisercontr ibuted to the fastest recovery of iron oreoverburden and achieved the most success in terms ofvegetation establishment. It was also clear that differentengineering designs did not have a measurableinfluence on the vegetation composition and cover atthe different sites, hence that this factor can bedisregarded in the rehabilitation plan for mine closure.Further studies are desirable to evaluate the progressof the sites in terms of the development of organicmatter and biodiversity. The following insights weregained from the different management practices appliedat the mine:• Vegetation can be established at a 24 ° slope angleon the north- and east-facing slopes. On the westernand southern slopes, successful vegetationestablishment would be possible at a 34 ° slope angleif the practical limitations of reshaping for surface watercontrol can be overcome.• To achieve vegetation success, the application oforganic matter is vital to increase the water-holdingcapacity, as well as the microbial activity of the growthmedium. The importance of the soil carbon fraction inrehabilitated soils is one of the most undervaluedcomponents of rehabilitation. From the trials conductedat Sishen, it is clear that one of the main limitations oftailings materials in terms of vegetation sustainabilityis the lack of organic matter in the growth mediumprofile. Sites treated with compost had significantlyhigher biodiversity and it was found that rootdevelopment, above-ground biomass production, andtherefore the basal and crown cover outperformedareas without compost.• Hydro seeding is the best method for seeding despitethe arid climate at Sishen. This method distributes theseeds evenly and results in a more uniform vegetationcover.• A seed mixture containing the following grass speciesproved to be the most successful – Cenchrus ciliarisbecame the dominating grass species, whereasHeteropogon contortus, Digitaria eriantha,Anthephora pubescens and Fingerhuthia africana areperennial grasses that play an important role insuccessful rehabilitation. Melinis repens, Enneapogoncenchroides and Aristida adscensiones are annualspecies that were found to perform well (Van Rensburg& Roussow, 2008).• Topsoil application is not essential but a fine materialcover layer that can be shaped for surface water controlis required. Topsoil tends to erode notably and shouldonly be used in high priority areas as it is expensive toapply.

• Surface water flow needs to be well managed andbanks should be back-sloped as this has often beenthe main reason for erosion in the past.

Sishen mine is taking a step in the right directionby acknowledging their responsibility toward theenvironment and future generations seriously throughassessing rehabilitation practices and determining whatis best for the mine. Significant contributions made bythis review include the identification of parametersessential for rehabilitation success at Sishen mine andthe establishment of an overview of the history ofrehabilitation practices applied at the mine.

ACKNOWLEDGEMENTSThe authors would like to acknowledge the

management of the mine for their vision andconsciousness. A special word of thanks to D Krugerfor his technical assistance. This research wassupported financially by Kumba resources.

REFERENCESAlef, K. and Nannipieri, P. (Eds.) (1995). Methods in AppliedSoil Microbiology and Biochemistry. New York: AcademicPress.

Aon, M. A. and Colaneri, A. C. (2001). Temporal and spatialevolution of enzymatic activities and physico-chemicalproperties in an agricultural soil. Appl. Soil Ecol., 18, 255-270.

Badiane, N. N. Y., Chone, J. L., Pate, E., Masse, D. andRouland, C. (2001). Use of soil enzyme activities to monitorsoil quality in natural and improved fallows in semi-aridtropical regions. Appl. Soil Ecol., 18, 229-238.

Bandick, A. K. and Dick, R. P. (1999). Field managementeffects on soil enzyme activities. Soil Biol. Biochem., 31,1471-1479.

Brohon, B., Delolme, C. and Goudon, R. (2001).Complimentarily of bioassays and microbial activitymeasurements for the evaluation of hydrocarbon-contaminated soils quality. Soil Biol. Biochem., 33, 383-891.

Carroll, C., Merton, L. and Burger, P. (2000). Impact ofvegetative cover and slope runoff, erosion, and water qualityfor field plots on a range of soil and spoil materials on centralQueensland coal mines. Aus. J. Soil Res., 38, 313-327.

Claassens, S., Jansen van Rensburg, P.J., Maboeta, M.S.and Van Rensburg, L. (2011). An application of space-for-time substitution in two post-mining chronosequences underrehabilitation. S. Afr. J. Plant & Soil, 28 (3), 151-162.

Criquet, S., Ferre, E., Farnet, A. M. and Le Petit, J. (2004).Annual dynamics of phosphatase activities in an evergreenoak liner: influence of biotic and abiotic factors. Soil Biol.Biochem., 36, 1111-1118.

Dick, R. P. (1994). Soil enzyme activities as indicators ofsoil quality. In Doran JW, Coleman DC, Bezdicek DF. andStewart BA, (Eds.), Defining soil quality for a sustainableenvironment, SSSA Special Publication 35, Madison: SSSA.


643

Dick, R. P. (1997). Soil enzyme activities as integrativeindicators. In Pankhurst, C.E, Doube BM, Gupta VVSR,(Eds.), Biological indicators of soil health. New York: CABInternational.

Gil-Sotres, F., Trasar-Cepeda, C., Leiros, M. C. and Seoane, S.(2005). Different approaches to evaluating soil quality usingbiochemical properties. Soil Biol. Biochem., 37, 877-887.

Ibekwe, A. M., Kennedy, A. C., Frohne, P. S., Papiernik, S.K., Yang, C. H. and Crowley, D. E. (2002). Microbialdiversity along a transect of argonomic zones. FEMSMicrobiol. Ecol., 39, 183-191.

Knight, T. R. and Dick, R. P. (2004). Differentiating microbialand stabilized P-glucosidase activity relative to soil quality.Soil Biol. Biochem., 36, 2089-2096.

Kumba Resources, (2006). An independent competentpersons report on the mining assets of Exxaro ResourcesLimited: Environmental Management. Retrieved August 6,2010 from http://www.exxaro.com/pdf/icpr/g/em/coal.htm

Lepš, J. and Šmilauer, P. (2003). Multivariate analysis ofecological data using CANOCO. Cambridge:University Press.

Milton, S. J. (2001). Rethinking ecological rehabilitation inarid and winter rainfall regions of southern Africa. S. Afr. J.Sci., 97, 1-2.

Mulligan, D. R. (1997). Environmental Management in theAustralian Minerals and Energy Industries: Principles andPractices. Sydney: University of New South Wales Press.

Mummey, D. L., Stahl, P. D. and Buyer, J. S. (2002). Microbialbiomarkers as an indicator of ecosystem recovery followingsurface mine reclamation. Appl Soil Ecol, 21, 251-259.

Tate, R. L. and Rogers, B. F. (2002). Soil ecosystemproperties, microbial diversity, and ecosystem assessments.(In Violante A., Huang P.M., Bollag J.M., Gianfreda L,(Eds.), Developments in soil science. New York: Elsevier.

Taylor, J. P., Wilson, B., Mills, M. S. and Bums, R. C.(2002). Comparison of microbial numbers and enzymaticactivities in surface soils and subsoils using varioustechniques. Soil Biol Biochem, 34, 387-401.

Ter Braak, C. F. and Smilauer, P. (1998). Canoco referencemanual for user’s guide to Canoco community ordination,ver. 4. New York: Microcomputer Power.

Ter Braak, C. F. and Verdonschot, P. F. M. (1995). Canonicalcorrespondence analysis and related multivariable methodsin aquatic ecology. Aquat. Sci., 57, 153-298.

Turner, B. L., Hopkins, D. W., Haygarth, P. M. and Ostle,N. (2002). Â-Glucosidase activity in pasture soils. Appl.Soil Ecol., 120, 157-162.

Van Rensburg, L., Claassens, S. and Jansen van Rensburg, P.J. (2004). Progress Report: Quantifying the extent ofrehabilitation of iron ore tailings at the Sishen Iron Mine.Potchefstroom: North-West University, PotchefstroomCampus, School of Environmental Sciences andDevelopment.

Van Rensburg, L. and Rossouw, J. (2008). Progress Report:Quantifying the extent of rehabilitation of iron ore tailingsat the Sishen Iron Ore Mine: Microbiological activityassociated with different treatment options. Potchefstroom:North-West University, Potchefstroom Campus, School ofEnvironmental Sciences and Development.

Van Wyk, S. (1994). ‘n Strategie vir die rehabilitasie vanversteurde mynbougebiede in Suidelike Afrika. Thesis,Potchefstroom University for Christian Higher Education,Potchefstroom, South Africa.

Van Wyk, S. J. (2002). An analytical investigation of thebiophysical factors that inhibit successful ecologicalrestoration of gold tailings dams. Dissertation,Potchefstroom University for Christian Higher Education,Potchefstroom, South Africa.

Van Wyk, S. (2006). Report on visual observation for Kumba(Sishen) on the vegetation of the experimental site.Potchefstroom: UFAAM Ecological consultant.

Walkley, A. and Black, I. A. (1934). An examination of theDegiareff method for determining SOM and a proposedmodification of the chromic acid titration method. Soil Sci.,37, 29-38.

Weather Bureau, (1984). Climate of South Africa, Climatestatistics up to 1984 WB 40. Republic of South Africa.Department of Environmental Affairs.


644

Documents

Insights and Lessons Learned From the Long-term ......Tech shale 6.6 7.4