Está en la página 1de 10

Available online at www.sciencedirect.

com

Environmental Pollution 155 (2008) 31e40


www.elsevier.com/locate/envpol

Comparison of benthic macroinvertebrate indices for the assessment


of the impact of acid mine drainage on an Irish river below
an abandoned CueS mine
N.F. Gray*, E. Delaney
Centre for the Environment, Trinity College, Dublin 2, Ireland
Received 4 May 2007; received in revised form 21 September 2007; accepted 2 November 2007

Several indices were suitable for AMD impact assessment, although due to AMD and
receiving water variability no single index may be universally reliable.
Abstract
A range of macroinvertebrate indices were compared to assess the most appropriate metric for the assessment of acid mine drainage (AMD)
in a low alkalinity, highly erosional river in south-east Ireland. Differences were found in the ability of indices to discriminate AMD impact with
the Brillouin, BMWP score, Margalef and Shannon Indices the most precise. Taxon richness was also strongly correlated with AMD indicator
parameters (e.g. pH alkalinity, sulphate, Zn and Fe) at impacted sites being an equally reliable metric. The response of the community structure
to AMD in this river does not fulfil the optimum criteria for either diversity or biological indices, which may explain the variation in the success
of different indices seen in this and other studies. The development of indices that model the expected community response to AMD more
accurately or are based on the response of indicator species to AMD pollutants are required.
2007 Elsevier Ltd. All rights reserved.
Keywords: Acid mine drainage; Macroinvertebrates; Rivers; Diversity indices; Biological indices

1. Introduction
The impact of acid mine drainage (AMD) on surface waters
is a worldwide problem that has serious and often widespread
effects (Kelly, 1988; Cherry et al., 2001; Lin et al., 2007). The
nature of AMD makes it an unusual and multifarious pollutant
that is formed when pyritic rocks and ores are exposed to an
oxidizing environment. The oxidation process is complex involving chemical, biological and electrochemical reactions
that vary with environmental conditions (Evangelou and
Zhang, 1995). In the presence of water and oxygen, bacterial
mediated oxidation results in the rapid formation of a highly

* Corresponding author. Tel.: 353 1-8961639.


E-mail address: nfgray@tcd.ie (N.F. Gray).
0269-7491/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.envpol.2007.11.002

acidic and metal rich leachate that can seriously impact surface and ground waters (Singer and Strumm, 1970; Saria
et al., 2006).
The effect on surface waters is dependent on their buffering capacity, but typically results in a reduction in pH, elevated metal concentrations (e.g. Fe, Zn, Cu, Al, Pb, As,
Cd, Mn, Se, etc.), increased sulphate concentration and the
formation of ochre, a stable orange precipitate comprising
iron oxyhydroxides (Gray, 1996). The impacts of AMD on
surface waters have been reviewed by Kelly (1988) and
Gray (1997).
The relationship between AMD and the macroinvertebrate
community has been widely studied (Roline, 1988; Gower
et al., 1995; Malmqvist and Hoffsten, 1999; Battaglia et al.,
2005). The affects of AMD on surface waters can be summarized as acidity, metal toxicity, metal precipitation and

32

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

salinization with the degree of impact ranging from non-detectable to complete destruction of the normal flora and fauna
(Kelly, 1988; Gray, 1997; Cherry et al., 2001; De Nicola and
Stapleton, 2002; David, 2003). However, interpreting the response of riverine communities to AMD has proven difficult.
Riverine studies employ the use of diversity and biological
indices to measure the impact of pollutants using the macroinvertebrate community as indicators of ecosystem health.
This has also been adopted in the assessment of AMD in
rivers. Biological indices are generally used for a particular
type of water pollution, normally organic pollution. In contrast diversity indices are non-specific and measure total environmental stress rather than a specific type of pollutant
(Washington, 1984; Hellawell, 1986). The simplest measure
of diversity is taxon richness although it is susceptible to
sample size. This is overcome by employing diversity indices
that incorporate both taxon richness and abundance. Diversity
indices can be categorized as either dominance indices that
are weighted towards abundance of the commonest species
(e.g. Simpsons Index) or information-statistic indices which
are based on the rationale that diversity in a natural system
is similar to the way information contained in a code or message is measured and will reflect individual taxon abundance
(e.g. Shannon Index and Brillouin Index) (Washington,
1984).
Varying levels of success have been achieved in employing indices in the assessment of AMD impact (Armitage,
1980; Chadwick et al., 1986; Whiting et al., 1994; Nelson
and Roline, 1996), with the majority recommending the
use of the BMWP (Biological Monitoring Working Party
Biotic Index) for routine surveillance at impacted sites. In
a comparative study of the performance of indices in a river
affected by coal mining in Spain and subject to acidic
drainage, Garca-Criado et al. (1999) found the BMWP
(modified for the Iberian Peninsula) and the family richness
of Ephemeroptera, Plecoptera and Trichoptera (EPT) were
most highly correlated with mining impact. Similarly
Malmqvist and Hoffsten (1999) reported reductions in
both taxon richness and EPT with AMD impact in Swedish
streams.
In this study a range of diversity and biological indices are
compared to assess the most appropriate metric for AMD impacted rivers and to examine the appropriateness of the
BMWP as a measure of impact in such circumstances.
A number of separate hypotheses were tested:
H1 That diversity indices reflect the impact of AMD on
macroinvertebrate assemblages more efficiently than biological indices;
H2 That the Simpsons Index (D) is the most suitable diversity index for AMD;
H3 That the BMWP derived score reflects the impact of
AMD on macroinvertebrate assemblages more efficiently
than the average score per taxon (ASPT);
H4 That percentage Diptera, and in particular percentage
Chironomidae, are the most discriminating metrics for
assessing AMD impact in rivers.

2. Materials and methods


2.1. Location
The abandoned CueS mines at Avoca, County Wicklow, in southeast
Ireland has been widely studied since its closure in 1982 (Gray, 1998; Gaynor
and Gray, 2004). The acid mine drainage discharged from the site has seriously
affected the water quality and biological status of the highly erosional and low
alkalinity (<24 mg CaCO3 l1) River Avoca, although there has been a steady
recovery in quality as the acid mine drainage slowly alters in character
(Gaynor and Gray, 2004). This river forms the lower main channel of the
AvonmoreeAvoca catchment (watershed 625 km2) and full details of both
the river and mines are given by Sullivan et al. (1995).
The mines discharge into the Avoca River via a number of adits immediately downstream of the White Bridge (Ordnance Survey Map Reference
T204768). Water and biological samples were taken at monthly intervals during the periods of lowest discharge rate from May to August in 2006, the
only times when the river is reliably accessible for biological monitoring.
The sample locations are shown in Fig. 1. Site 1 is the non-impacted control
site 0.75 km upstream of the White Bridge. Sites 2e5 all show visual signs of
impaction by AMD. Site 2 is located immediately after complete mixing at
2.5 km below the White Bridge and site 3 at 3.6 km. The River Aughrim
is a major tributary that enters the river 7.25 km downstream of the White
Bridge and which has an almost identical discharge rate to that of the
main channel. Site 6 is a control site on the River Aughrim 0.85 km upstream
of the confluence with the River Avoca. Two more impacted sites are monitored below the confluence after complete mixing at site 4 at 8.5 km and
site 5 at 11.0 km downstream of the White Bridge. The river enters the Irish
Sea at Arklow 4 km below site 5. Full details of the chemical and physical
nature of the river have been detailed by Gray (1998) and Gaynor and
Gray (2004).

2.2. Chemical analysis


Water samples were filtered in the field through a Millipore cellulose nitrate membrane with a pore size of 0.45 mm, and stored in high-density plastic
bottles and transported back to the laboratory for analysis in an ice box. Samples were stored at the laboratory in the dark at 4  C. Conductivity, alkalinity
and pH analysis were carried out within 24 h of sample collection using
a WTW LF196 conductivity meter, the Gran titration method (APHA, 1992)
and a Jenway 3030 pH meter, with an Ag/AgCl reference electrode and
temperature compensation, respectively. Sulphate analysis was carried out
by ion-exchange chromatography using a Dionex ICS-1500 analyzer, while
Cu, Fe, Zn, Pb and Cd were analyzed by flame atomic absorption spectrophotometry using a Perkin Elmer 3110 atomic absorption spectrophotometer with
a graphite furnace (APHA, 1992).

2.3. Biological analysis


Macroinvertebrates were collected from riffle sites using a standard hand
net (900 mm mesh) collection technique (ISO, 1994). Collections were made
for a total of 3 min using the replicate sampling technique of Byrne and
Gray (1995). Sub-samples from each site were pooled for each of three replicate collections and after preliminary sorting were stored in 75% alcohol and
subsequently fully sorted in the laboratory with the animals identified to
species level, except for the Chironomidae and Oligochaeta.

2.4. Indices
Five biological and five diversity indices were selected based on their previous use for AMD assessment in freshwaters. The biological indices were
BMWP, ASPT (BMWP average score per taxon), percentage EPT (Ephemeroptera, Plecoptera and Tricoptera), EPT/EPT C (EPT plus the effect of Chironomidae) and the percentage Chironomidae. The diversity indices selected
were the indices of Shannon, Simpson, Menhinick, Margalef and Brillouin.

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

33

Fig. 1. The location of the abandoned CueS mines, and the biological and chemical sampling sites along the Avoca River, in southeast Ireland.

The Biological Monitoring Working Party (BMWP) Biotic Score Index


was devised in 1978 for the surveillance of rivers in the UK (Armitage
et al., 1983), but has been widely adopted within Europe in modified form
(Hawkes, 1997). Scoring is based on 85 families of known tolerance to organic
pollution. The presence of each family provides a single score out of 10 (1 being highly tolerant, 10 highly sensitive) with the cumulative scores being the
BMWP index or score. Due to the variation between BMWP scores due to factors such as sampling variation and seasonality, then it is more reliable to use
the average score per taxa (ASPT) providing a single score out of 10 representing average saprobity (Armitage et al., 1983; Hawkes, 1997).
ASPT

BMWP
N

Shannon Index :

The EPT uses the three most sensitive groups to organic pollution, however, the Chironomidae (C ) are particularly tolerant to organic and other types
of pollution including heavy metals. Therefore it can be included in the biological EPT index to strengthen its ability to measure tolerance to pollution (i.e.
EPT/EPT C).


Abundance of EPT
Abundance of EPT C

H0 

Simpson0 s Index :

 
S
X
ni
ni
loge
Pielou; 1966
n
n
i1

S
X
nini  1
i1

Menhinick0 s Index :
Margalef 0 s Index :

where N is the total number of families used in the calculation of the total
score (i.e. the BMWP). The EPT is the percentage abundance of the three important indicator groups the Ephemeroptera (E), Plecoptera (P) and Trichoptera (T).


Abundance of EPT
100
EPT
Total abundance

EPT=EPT C

The final biological index was the percentage abundance of Chironomidae.


The diversity indices were calculated as:

Brillouin0 s Index :

nn  1

Simpson; 1949

S
I p Menhinick; 1964
n
S1
Margalef; 1951
lnn
P
lnn!  lnni!
Brillouin; 1962
H
n
I0

where S is the number of species in either a sample or population; n the total


number of individuals in a population or community; ni the number of individuals in the ith species.

2.5. Statistical analysis


Differences between sites were based on the taxonomic composition of the
macroinvertebrate assemblages measured using one-way non-parametric multivariate analysis of variance employing the BrayeCurtis similarity index. Differences were also measured using macroinvertebrate abundance and taxon
richness using one-way ANOVA. The relationships between the biological
data, in terms of abundance and taxon richness, and transformed using the

34

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

various indices under investigation, and the chemical parameters were determined using the correlation of Pearson (De Pauw and Roels, 1988). This approach has successfully been used in similar studies previously (Camargo and
Garca de Jalon, 1995; Zamora-Mu~noz et al., 1995; Garca-Criado et al.,
1999). All the biological parameters showed normal distributions while chemical variables were log-transformed where necessary to make them approximate to normality (Garca-Criado et al., 1999). Average r values for each
site using individual values for pH, sulphate, zinc and iron, that each represents a different component of toxicity and impact, allowed the performance
of the indices to be ranked.

Table 1
The taxon richness and total abundance per site with the monthly mean and
standard deviation (SD)
Site

1
2
3
4
5
6

Taxa

Abundance

Jun

Jul

Aug

Mean

SD

Jun

Jul

Aug

Mean

SD

27
8
11
17
20
26

27
5
9
21
16
11

25
8
11
18
17
23

26.3
7.0
10.3
18.7
17.7
20.0

0.9
1.4
0.9
1.7
1.2
6.5

200
38
32
48
157
321

898
5
41
268
384
116

861
38
85
142
438
447

653
27
53
153
326
293

320
16
23
90
122
134

3. Results
Three zonal communities were identified with significant
differences in the macroinvertebrate assemblages ( p < 0.001).
These were the non-impacted AMD control sites 1 and 6, with
the latter of poorer overall quality than site 1, the heavily impacted zone (sites 2 and 3) and a mildly impacted or recovery
zone (sites 4 and 5) which was clearly influenced by the confluence with the River Aughrim (Fig. 2). Both the taxon richness
and the abundance were also significantly lower in the impacted
sites compared to the control sites showing significant community damage (Table 1). This was also reflected in reductions in
pH and alkalinity indicating acidification from the AMD, and
increases in sulphate and metals commonly associated with
AMD (Gray, 1998).
The impact from the AMD as measured by the indices is
summarized in Table 2. The BMWP score was significantly
correlated against all the key parameters (Table 3), increasing
with pH ( p < 0.001) and alkalinity ( p < 0.05) but decreasing
with sulphate ( p < 0.001), Zn ( p < 0.001) and Fe ( p < 0.001)
concentrations. In contrast the mean BMWP score, the ASPT,
was independent of the abiotic parameters, as were the other
biological indices the EPT and EPT/EPT C. The percentage
Chironomidae was only correlated with increasing Fe concentration ( p < 0.05).
In general diversity indices showed a high degree of correlation with the abiotic parameters. All were significantly

correlated with pH. The Brillouin, Margalef and Shannon Indices were correlated with all five key AMD parameters, in
that order of closeness, and all showing a high negative correlation ( p < 0.001) with Fe (Table 3). The Simpson Index
showed no significant correlation ( p > 0.05) with either alkalinity or zinc and only weak correlations with the other parameters ( p < 0.05). The Menhinick Index was negatively
correlated with pH only ( p < 0.05).
Taxon richness was strongly correlated against pH, sulphate
and zinc ( p < 0.001), and also with alkalinity ( p < 0.05), but
not Fe. Abundance was correlated with pH ( p < 0.001), zinc
( p < 0.01) and sulphate ( p < 0.05), but neither alkalinity
nor Fe ( p > 0.05).
All the indices were tested for associations between each
other (Table 4). While all the diversity indices are expected
to give slightly different results they are all well correlated
with the exception of the Menhinicks Index. Insect orders
such as Ephemeroptera, Plecoptera and Trichoptera were
replaced by metal tolerant Diptera, especially chironomids,
at impacted sites (Table 5). For that reason the EPT was negatively correlated with Chironomidae abundance ( p < 0.001).
Neither the EPT nor the percentage Chironomidae was correlated with either abundance or taxon richness. However, the
BMWP was correlated with both taxon richness ( p < 0.001),
and abundance ( p < 0.01) while ASPT was only correlated
with abundance ( p < 0.01), but not taxon richness ( p > 0.05).
4. Discussion
4.1. Impact of AMD

Fig. 2. BrayeCurtis cluster analysis dendrogram showing similarity between


sampling sites.

The impact of AMD on the biota is due to a number of different factors namely acidity, metal toxicity, metal precipitation and salinization (Gray, 1997). The relative importance
of these factors to the biota varies according to the nature
of the AMD, its dilution, the nature of the receiving water, especially its buffering and assimilative capacities, species tolerance to pollutants, and other ecological and environmental
factors. Battaglia et al. (2005) examined the relative toxicity
of AMD to macroinvertebrates in the water and sediment
(pore water) of impacted streams. They found that the surface
water quality has a greater influence than pore water quality,
with the biological data highly correlated with the concentrations of common metals in the surface water. In this study the
biological data was correlated with a range of chemical

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

35

Table 2
Mean and standard deviation (SD) of selected indices at river sites
Index

EPT (%)
EPT/EPT C
BMWP
ASPT
Simpson
Menhinick
Margalef
Shannon
Brillouin
Chironomids (%)

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Mean

SD

68.6
84.7
157
7.4
0.88
1.31
4.5
2.43
2.32
9.57

24.7
18.9
12.8
0.26
0.07
0.43
0.37
0.21
0.16
10.9

71.2
73.7
50
7.6
0.76
1.75
2.33
1.67
1.29
25.17

24.6
24.5
7.0
0.42
0.07
0.35
0.13
0.17
0.28
23.8

58.8
46.0
64
6.6
0.87
1.68
2.73
1.80
1.51
44.6

20.3
28.2
10.0
0.86
0.18
0.28
0.29
0.43
0.34
24.8

53.1
55.7
116
6.9
0.85
2.57
4.10
2.47
2.19
30.2

15.7
24.4
18.7
0.16
0.08
0.48
0.25
0.33
0.31
17.7

58.2
64.7
98
6.7
0.87
1.20
3.33
2.40
2.25
19.2

14.4
10.6
10.3
0.28
0.04
0.29
0.37
0.14
0.10
3.38

45.3
63.7
111
7.0
0.85
1.33
3.70
2.37
2.24
25.2

20.9
25.9
39.4
0.29
0.06
0.14
0.75
0.25
0.25
19.1

parameters that characterize the addition of AMD, namely


pH, alkalinity and sulphate which are unaffected by sorption
processes and the two commonest metals found in the leachate Fe and Zn. Sulphate is considered the best indicator of
AMD as it is unaffected by neutralization (Gray, 1996) and
has been used as the reference indicator in previous AMD impact studies (Garca-Criado et al., 1999). The pH and alkalinity act as a reference for possible acidification, while Zn
references metal toxicity. Finally the Fe concentration was
used to assess the effect of ochre formation, which was found
to decrease with distance from the input of AMD in earlier
studies (Gray, 1998). The critical factors responsible for disrupting the macroinvertebrate community in AMD impacted
rivers appears to be pH (Allen et al., 1996), metal concentrations (Soucek et al., 2000; Kim and Chon, 2001) or a combination of the two (Cherry et al., 2001). In practice there is
often a clear negative relationship between pH and metal concentration in AMD impacted rivers, confirmed in the present
study (Zn p < 0.001), with the effects on the biota of each
difficult to separate.
A reduction in both the diversity and abundance of macro-invertebrates is commonly associated with AMD impacted rivers,
which is the common response of macroinvertebrate assemblages to metal contamination with species often eliminated

in the most impacted sections (Gray, 2005) (Fig. 3). This community response is clearly seen in the Avoca River (Table 1).
Ephemeroptera, Plecoptera and Trichoptera (EPT) are normally
excluded, although they have varying sensitivities to both
metals and pH, while the abundance of Diptera, in particular
Chironomidae, usually increases due to tolerance to both these
abiotic factors (Gray, 1996; Malmqvist and Hoffsten, 1999;
Courtney and Clements, 2002; De Nicola and Stapleton,
2002). Four hypotheses were tested in this study.
4.2. H1: that diversity indices reflect the impact of AMD
on macroinvertebrate assemblages more efficiently than
biological indices
Biological indices are based on the response of indicator organisms to organic pollution and have been developed from
the concept of saprobity, with each organism having a discernible tolerance to the presence of degradable organic material
(Washington, 1984; Hellawell, 1986). In contrast, diversity indices depend on the concept that environmental stress, including pollutants, alters the community structure by reducing
taxon diversity and increasing the size of the populations
within that community. So an unstressed site has a macroinvertebrate assemblage comprised of high taxon diversity with

Table 3
Pearson correlations (two-tailed) r of indices against key indicator abiotic AMD parameters
Index

pH

Alkalinity (mg CaCO3 l1)

Sulphate (mg l1)

Zn (mg l1)

Fe (mg l1)

Average r

Average r excluding
alkalinity

Biotic indices
BMWP
ASPT
EPT (%)
EPT/EPT C
Chironomids (%)

0.758***
0.161
0.077
0.171
0.300

0.476*
0.145
0.102
0.052
0.146

0.709***
0.206
0.035
0.174
0.299

0.775***
0.010
0.020
0.224
0.365

0.489*
0.246
0.155
0.366
0.532*

0.641**
0.154
0.078
0.197
0.328

0.683**
0.156
0.072
0.234
0.374

0.684**
0.472*
0.237
0.704***
0.764***
0.499*
0.748***

0.536*
0.431
0.435
0.528*
0.636**
0.639**
0.830***

0.710***
0.530*
0.326
0.760***
0.795***
0.373
0.441

0.624**
0.447
0.340
0.642**
0.717***
0.523*
0.683**

0.652**
0.482
0.367
0.672**
0.750***
0.575
0.711***

Diversity indices and abundance


Shannon
0.679**
Simpson
0.495*
Menhinick
0.470*
Margalef
0.697***
Brillouin
0.805***
Total abundance
0.787***
Taxon richness
0.825***

0.513*
0.306
0.234
0.520*
0.585**
0.317
0.568*

The level of significance is ***p < 0.001, **p < 0.01 and *p < 0.05.

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

Table 5
Relative abundance of the key groups in the River Avoca showing mean percentage EPT and Chironomidae per site

0
0.193
0.290
The level of significance is ***p < 0.001, **p < 0.01 and *p < 0.05.

EPT
EPTC
BMWP
ASPT
SIMP
MENH
MARG
SHAN
BRIL
CHIR
ABUN
TAXA

0
0.922***
0.186
0.708***
0.396
0.340
0.133
0.234
0.069
0.809***
0.004
0.027

0
0.327
0.552*
0.656**
0.286
0.287
0.524*
0.363
0.963***
0.104
0.186

0
0.184
0.496*
0.206
0.923***
0.731***
0.805***
0.400
0.662**
0.969***

0
0.359
0.151
0.035
0.114
0.002
0.465*
0.669**
0.014

0
0.057
0.426
0.875***
0.772***
0.771***
0.252
0.406

0
0.072
0.030
0.233
0.224
0.667**
0.290

0
0.709***
0.727***
0.379
0.510*
0.926***

0
0.956***
0.645**
0.300
0.674**

0
0.500*
0.470*
0.785***

0
0.718***

TAXA
ABUN
CHIR
BRIL
SHAN
MARG
MENH
SIMP
ASPT
BMWP
EPTC
EPT

Table 4
Pearson correlations (two-tailed) r of indices percentage EPT (EPT), EPT/EPT C (EPTC), BMWP, ASPT, Simpsons (SIMP), Menhenicks (MENH), Margalefs (MARG), Shannons, (SHAN), Brillouins
(BRIL) indices, percentage Chironomidae (CHIR), the total abundance (ABUN), and taxon richness (TAXA)

36

Mean taxa
richness (%)

Site 1

Site 2

Site 3

Site 4

Site 5

Site 6

Ephemeroptera
Plecoptera
Trichoptera
EPT
Chironomidae

43.8
9.3
8.6
61.7
12.7

12.3
17.3
28.4
58.0
37.0

5.1
14.0
7.0
26.1
55.0

16.6
8.9
8.7
34.2
31.0

16.6
15.7
4.2
36.5
27.1

25.8
11.4
5.2
42.4
30.1

most species having a low abundance while at stressed sites


the diversity is low with most species having a high abundance
(Hellawell, 1986).
Table 3 summarizes the correlation of both the biological
and diversity indices selected to the key AMD parameters measured along the AMD pollution gradients that occur in the
Avoca River. Apart from the total BMWP score, which showed
a high correlation with all the abiotic parameters (average r
value 0.641, p < 0.01), none of the biological indices normally
recommended for use with AMD impacted rivers were correlated with the abiotic parameters (Table 3). This is contrary
to the findings of Malmqvist and Hoffsten (1999) who reported
significant correlations between the EPT and AMD impact.
Only the percentage Chironomidae showed a weak but significant correlation with Fe ( p < 0.05). In the River Avoca the
substrate at the most impacted sites, which are normally coated
with ochre, become colonized during the summer months with
a luxuriant growth of periphyton comprised mainly of filamentous iron bacteria and the metal-tolerant algae Hormidium
rivulare, although the growths are heavily encrusted with
iron hydroxide precipitate. This becomes a temporary seasonal
niche for chironomids in the impacted areas of the river. The
dominance of EPT taxon in the non-impacted river and the
dominance of Diptera, in particular Chironomidae at impacted
sites, did produce correlations between the groups when used
for indicator purposes (Table 4). This is also seen with the
ASPT but not the BMWP score which is more influenced by
taxon richness (Armitage et al., 1983) to which none of the
other biological indices were correlated.
In contrast the diversity indices display a high degree of
correlation with all the abiotic parameters (Table 3). This is
in direct contradiction to Chadwick et al. (1986) who found
that diversity indices did not reflect the level of mine drainage
at their sites. While Chadwick et al. (1986) and Garca-Criado
et al. (1999) both found the Shannon Index to be correlated to
pollution levels caused by AMD they concluded that the index
was of limited discriminatory power and displayed inconsistency between seasons. In this study the Shannon Index was
significantly correlated with all the key AMD parameters, although not as strongly as either the Brillouin or Margalef Indices (Table 3). The hypothesis, that diversity indices are
more precise indicators of AMD impact in rivers than biological indices is rejected due to the reliability seen in this study
and that reported by Garca-Criado et al. (1999) of the BMWP
score.

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

37

Fig. 3. Generalized model of the effect of a discharge of acid mine drainage on river macroinvertebrate community structure.

4.3. H2: that the Simpsons Index (D) is the most suitable
diversity index for AMD
Several studies have shown the Simpsons Index (D),
which is based on the theory of runs to be superior to those
based on information theory (e.g. Shannon and Brillouin)
and more realistic in the interpretation of freshwater ecosystems (Washington, 1984; Lande, 1996). However, both
Routledge (1979) and Keylock (2005) have highlighted the
potential usefulness of the Shannon Index (H0 ) in such
studies.
In this study the Simpson Index was least able to interpret
the impact of AMD (Table 3) although still significantly correlated with pH ( p < 0.05), sulphate ( p < 0.05) and Fe
( p < 0.05). While it was strongly correlated with all the other
diversity indices ( p < 0.001), except Margalefs Index
( p > 0.05), it was unrelated to either total abundance or taxon
richness ( p > 0.05). In contrast the Brillouin, Margalef and
Shannon Indices were all significantly correlated with the abiotic parameters (Table 3). The Brillouin Index (H ) has been
successfully used with AMD impacted rivers in a number of
studies (Bradfield, 1986), including the assessment of the reclamation techniques in a stream contaminated by AMD from
an abandoned mine in southwest Virginia (Matter and Ney,
1981). In the current study it performed best out of all the
diversity indices tested followed by the Margalef (I0 ) and
Shannon Indices. The Simpson Index was least sensitive to
changes in AMD impact in the river. Therefore the hypothesis
is rejected.
The main difference between the Shannon and Brillouin
Indices is that the former does not change with abundance
as long as the proportional abundance remains constant,
while the latter does change. Dominance indices such as
Simpsons Index are biased towards the dominant taxon as
can be seen by its independence from both taxon diversity
and abundance. Where the total abundance of individuals is
important then the Brillouin Index is normally superior, and
in this study it more significantly correlated to both abundance ( p < 0.05) and taxon diversity ( p < 0.001) than the
Shannon Index (Table 4). Like all information indices, both
the Shannon and Brillouin Indices pay more attention to
rare species.

4.4. H3: that the BMWP derived score reflects the impact
of AMD on macroinvertebrate assemblages more
efficiently than the average score per taxon (ASPT)
As in previous studies (Garca-Criado et al., 1999), the
BMWP score was found to be an excellent index of AMD
(Table 3). Yet the ASPT was uncorrelated with all measured
parameters ( p > 0.05) and, unlike the BMWP score, was not
correlated with any of the diversity indices. ASPT was not significantly different between sites ( p > 0.05) and showed that
the saprobic ratings for all the sites over the entire sampling
period as measured by the ASPT were within a single unit
(6.6e7.6) offering little discrimination even though significant
differences in macroinvertebrate assemblages were obvious
(Fig. 2). The variation between sites were small: June, mean
7.0 (coefficient of variation 6.0%); July, 7.4 (6.9%); and
August 6.8 (8.8%). Within site variation was site 1, mean
7.4 (coefficient of variation 3.5%); site 2, 7.6 (5.5%); site 3,
6.6 (13%); site 4, 6.9 (2.8%); site 5, 6.7 (4.2%) and site 6,
7.0, (4.1%) (Table 2). ASPT was, however, significantly correlated to EPT ( p < 0.001), EPT/EPT C ( p < 0.05), percentage Chironomidae ( p < 0.05) and abundance ( p < 0.01). The
BMWP score was not correlated to any other biological index
( p > 0.05) (Table 4).
Armitage et al. (1983) recommend that the ASPT should be
used in preference to the BMWP score in river quality assessment, which is more susceptible to sampling variation, site selection and seasonality. They observed that additional taxa do
not necessarily increase the value of ASPT as they do the
BMWP score and may even lead to decreases. However, this
study accepts the hypothesis and confirms the usefulness of
the BMWP score in AMD impact assessment. The number
of individual scores making up total BMWP score at each
site is very strongly correlated to taxon richness ( p < 0.001)
(Table 4). Unlike diversity indices, where taxon richness is
used in their calculation, biological indices should not be dependent on taxon richness, which is confirmed by the independence of the other biological indices tested in this study.
However, the more taxa the greater the BMWP score and as
saprobity, which controls the individual score out of ten, is
not a factor in terms of presence of absence of species in
AMD impacted rivers as shown by similar ASPT site values,

38

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

then taxon richness becomes a major factor in the success of


the BMWP score as a metric for AMD.
4.5. H4: that percentage Diptera, and in particular
percentage Chironomidae, is the most discriminating
metric for assessing AMD impact in rivers
The Diptera are known to be tolerant to a wide range of pollutants, including metals and acidity, and as such are frequently recorded at AMD impacted sites (Winner et al.,
1980). It follows that if this is correct then these groups can
be considered as indicators of the metal contamination and
acidification associated with AMD.
As discussed previously, apart from the BMWP the only
other biological index to be significantly correlated with
AMD key parameters was the percentage Chironomidae being
correlated only with Fe ( p < 0.05). The percentage of the total
abundance comprised by chironomids did increase in the impacted sites (Table 5), reflecting the changing niches in the
river and the tolerance of the group to both acidification and
metal toxicity. However, the percentage Chironomidae was
not correlated with either abundance or taxon richness
( p > 0.05), nor metal concentrations ( p > 0.05) as recorded
by Winner et al. (1980) (Table 4). In this study neither Diptera
nor Chironomidae proved to be a precise metric and so the hypothesis is rejected.

so variable, resulting in different community responses to


each single component of the impact (i.e. inert (toxic) solids,
metal toxicity, acidity, and salinization) (Gray, 2005), it is unlikely that such an index will be either simple to develop or
subsequently to interpret. Garca-Criado et al. (1999) found
that members of the orders Ephemeroptera, Plecoptera and
Trichoptera were largely responsible for variations in the macroinvertebrate community, which is also seen in this study and
suggest that in order to develop a biological index for AMD
that sensitive and tolerant families within these orders should
be identified. A macroinvertebrate index based on families has
been developed by Davy-Bowker et al. (2005) for detecting
the impact of acidity in streams. The Acid Waters Indicator
Community Index (AWIC) distinguishes acid sites by the indicative loss of acid sensitive families, which are far more
common than acid tolerant families, rather than a characteristic
acid water tolerant assemblage of families. This may hold the
key to the development of an AMD biological index. Macroinvertebrate communities at AMD impacted sites are not characterized by specific obligate AMD families or taxa (e.g. acid
tolerant or metal tolerant) as these are generally equally common in non-impacted zones, rather it is the absence of large
numbers of sensitive families at impacted sites that may provide the basis of a new index. However, the good correlation
of the BMWP score with AMD impact suggests that it could
be improved by revising family scores, currently based on saprobity, on their tolerance to acidity and metal contamination.

4.6. Selection of AMD indices


Indices behave differently in terms of their ability to assess
AMD impact using macroinvertebrates as bioindicators, with
the Brillouin, BMWP score, Margalef and Shannon Indices
the most precise in this study of an erosional acidic river.
The results from this study show a significant reduction in
both abundance and taxon richness in response to AMD.
This is consistent with other similar studies who attributed
this reduction to the loss of Ephemeroptera that are particularly sensitive to pH (Courtney and Clements, 1998; Battaglia
et al., 2005), although Winterbourn and Collier (1987) who
studied streams over a pH range of 3.5e8.1 in New Zealand
found no correlation between pH and taxon richness. Therefore abundance and taxon richness are important criteria as
is the presence of individual indicator taxon or groups.
The negative correlations of Zn with both taxon richness
( p < 0.001) and abundance ( p < 0.05) (Table 3) recorded in
this study have also been noted in other studies (Thorp and
Lake, 1973; Armitage, 1980). Malmqvist and Hoffsten
(1999) found a similar significant negative relationship between taxon richness and both Zn and Cu, although in the
present study the concentrations of Cu in river water were
too close to detection limits for possible associations to be
reliable.
Some success has been achieved with developing specific
biological indices to assess acidification (Raddum et al.,
1988) and heavy metal pollution (Clements et al., 1992). So
in theory the development of a similar biological index for
AMD should be feasible. However, as the impact of AMD is

5. Conclusions
1. Indices behave differently in terms of their ability to identify AMD impact.
2. Diversity indices such as the Brillouin and Margalef are
equally as effective as the BMWP score in the River
Avoca, although differences in the performance of indices
are recorded by other studies on different rivers. This suggests that no single index may be universally reliable due
to wide variability of the impacts possible.
3. The ideal criteria for the application of either diversity or
biological indices are not recorded in the impacted River
Avoca in terms of community response to pollution. This
is also seen in other AMD impacted rivers.
4. Therefore, specific indices need to be developed that more
precisely measure the expected community structure arising from AMD impact than existing metrics, or alternatively indices based on AMD indicator species.
5. Taxon richness is a reliable measure of AMD impact
although subject to error from potential drift.

References
Allen, S.K., Allen, J.M., Lucas, S., 1996. Dissolved metal concentrations in
surface waters from west-central Indiana contaminated with acidic mine
drainage. Bulletin of Environmental Contamination and Toxicology 56,
240e243.

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40


APHA, 1992. Standard Methods for the Examination of Water and Wastewater, 18th ed. Joint publication of the American Public Health Association,
American Water Works Association, and Water Pollution Control Federation, Washington, DC.
Armitage, P.D., 1980. The effects of the mine drainage and organic enrichment
on benthos in the River Nent system, Northern Pennines. Hydrobiologia
74, 119e128.
Armitage, P.D., Moss, D., Wright, J.F., Furse, M.T., 1983. The performance of
a new biological water quality score system based on macroinvertebrates
over a wide range of unpolluted running water sites. Water Research 17,
333e347.
Battaglia, M., Hose, G.C., Turak, E., Warden, B., 2005. Depauperate macroinvertebrates in a mine affected stream: clean water may be key to recovery.
Environmental Pollution 138, 132e141.
Bradfield, A.D., 1986. Benthic Invertebrate Population Characteristics
Affected by Water Quality in Coal-Bearing Regions of Tennessee.
Water-Resources Investigations Report: 85-4227. US Geological Survey,
Nashville, TN.
Brillouin, L., 1962. Science and Information Theory. Academic Press, New
York.
Byrne, C., Gray, N.F., 1995. Effect of habitat type on the biological monitoring
of acid mine drainage in rivers. Fresenius Environmental Bulletin 4,
589e596.
Camargo, J.A., Garca de Jalon, D., 1995. Assessing the influence of altitude
and temperature on biological monitoring of freshwater quality: a preliminary investigation. Environmental Monitoring and Assessment 35, 227e
238.
Chadwick, J.W., Canton, S.P., Dent, R.L., 1986. Recovery of benthic invertebrate communities in Silver Bow Creek, Montana, following improved
metal mine wastewater treatment. Water, Air and Soil Pollution 28,
427e438.
Cherry, D.S., Currie, R.J., Soucek, D.J., Latimer, H.A., Trent, G.C., 2001. An
integrative assessment of a watershed impacted by abandoned mine land
discharges. Environmental Pollution 111, 377e388.
Clements, W.H., Cherry, D.S., Van Hassel, J.H., 1992. Assessment of the impact of heavy metals on benthic communities at the Clinch River (Virginia): Evaluation of an index of community sensitivity. Canadian
Journal of Fisheries and Aquatic Science 49, 1686e1694.
Courtney, L.A., Clements, W.H., 1998. Effects of acidic pH on benthic macroinvertebrate communities in stream microcosms. Hydrobiologia 379, 135e
145.
Courtney, L.A., Clements, W.H., 2002. Assessing the influence of water and
substratum quality on benthic macroinvertebrate communities in a metalpolluted stream: an experimental approach. Freshwater Biology 47,
1766e1778.
David, C.P., 2003. Establishing the impact of acid mine drainage through
metal bioaccumulation and taxa richness of benthic insects in a tropical
Asian stream (The Philippines). Environmental Toxicology and Chemistry
22, 2952e2959.
Davy-Bowker, J., Murphy, J.F., Rutt, G.P., Steel, J.E.C., Furse, M.T., 2005.
The development and testing of a macroinvertebrate biotic index for detecting the impact of acidity on streams. Archiv fuer Hydrobiologie 163,
383e403.
De Nicola, D.M., Stapleton, M.G., 2002. Impact of acid mine drainage on benthic communities in streams: the relative roles of substratum vs. aqueous
effects. Environmental Pollution 119, 303e315.
De Pauw, N., Roels, D., 1988. Relationships between biological and chemical
indicators of surface water quality. Verhandlungen Internationale Vereinigung Limnologie 23, 1553e1558.
Evangelou, V.P., Zhang, Y.L., 1995. A review: pyritic oxidation mechanisms
and acid mine drainage prevention. Critical Reviews in Environmental Science and Technology 25 (2), 141e199.
Garca-Criado, F., Tome, A., Vega, F.J., Antoln, C., 1999. Performance of
some diversity and biotic indices in rivers affected by coal mining in northwestern Spain. Hydrobiologia 394, 209e217.
Gaynor, A., Gray, N.F., 2004. Trends in sediment metal concentrations in the
River Avoca, southeast Ireland. Environmental Geochemistry and Health
26, 411e419.

39

Gower, A.M., Myers, G., Kent, M., Foulkes, M.E., 1995. The use of macroinvertebrate assemblages in the assessment of metal-contaminated streams.
In: Harper, D.M., Ferguson, A.J.D. (Eds.), The Ecological Basis for River
Management. Wiley, Chichester, pp. 181e192.
Gray, N.F., 1996. A substrate classification index for the visual assessment of
the impact of acid mine drainage in lotic systems. Water Research 30,
1551e1554.
Gray, N.F., 1997. Environmental impact and remediation of acid mine drainage: a management problem. Environmental Geology 30, 62e71.
Gray, N.F., 1998. Acid mine drainage composition and the implications for its
impact on lotic systems. Water Research 32, 2122e2134.
Gray, N.F., 2005. Water Technology, second ed. Elsevier, Oxford.
Hawkes, H.A., 1997. Origin and development of the Biological Monitoring
Working Party Scoring System. Water Research 32, 964e968.
Hellawell, J.M., 1986. Biological Indicators of Freshwater Pollution and Environmental Management. Elsevier Applied Science, London.
ISO, 1994. Water Quality. Methods for Biological Sampling. Guidance on
Hand-net Sampling of Benthic Macro-invertebrates. EN ISO 27828:1994.
International Organization for Standardization, Geneva.
Kelly, M., 1988. Mining and the Freshwater Environment. Elsevier Applied
Science, London.
Keylock, C.J., 2005. Simpson diversity and the ShannoneWeiner index as special cases of generalized entropy. Oikos 109, 203e207.
Kim, J., Chon, H., 2001. Pollution of a water course impacted by acid mine
drainage in the Imgok creek of the Gangreung coal filed, Korea. Applied
Geochemistry 16, 1387e1396.
Lande, R., 1996. Statistical partitioning of species diversity and similarity
among multiple communities. Oikos 76, 5e13.
Lin, C., Wu, Y., Lu, W., Chen, W., Liu, Y., 2007. Water chemistry and
ecotoxicity of an acid mine drainage-affected stream in subtropical China
during a major flood event. Journal of Hazardous Materials 142, 199e207.
Malmqvist, B., Hoffsten, P., 1999. Influence of drainage from old mine deposits on benthic communities in central Swedish streams. Water Research
33, 2415e2423.
Margalef, R., 1951. Diversidad de especies en las comunidades naturals. Publicaciones del Instituto Biologica Applicada Barcelona 6, 59e72.
Matter, W.J., Ney, J.J., 1981. The impact of surface mine reclamation on headwater streams in Southwest Virginia. Hydrobiologia 78, 63e71.
Menhinick, E.F., 1964. A comparison of some species e individuals diversity
indices applied to samples of field insects. Ecology 45, 859e861.
Nelson, S., Roline, R.A., 1996. Recovery of stream macroinvertebrate community from mine drainage disturbance. Hydrobiologia 339, 73e84.
Pielou, E.C., 1966. Shannons formulae as a measure of specific diversity: its
use and misuse. American Naturalist 100, 463e465.
Raddum, G.C., Fjellheim, A., Hesthagen, T., 1988. Monitoring of acidification
by the use of aquatic organisms. Verhandlungen Internationale Vereinigung Limnologie 23, 2291e2297.
Roline, R.A., 1988. The effects of heavy metals pollution of the upper Arkansas River on the distribution of aquatic macroinvertebrates. Hydrobiologia
160, 3e8.
Routledge, R.D., 1979. Diversity indices: which ones are admissible? Journal
of Theoretical Biology 76, 503e515.
Saria, L., Shimaoka, T., Miyawaki, K., 2006. Leaching of heavy metals in acid
mine drainage. Waste Management Research 24, 134e140.
Simpson, E.H., 1949. Measurement of diversity. Nature (London) 163, 688.
Singer, P.C., Strumm, W., 1970. Acidic mine drainage: the rate-limiting step.
Science 167, 1121e1123.
Soucek, D.J., Cherry, D.S., Currie, R.J., Latimer, H.A., Trent, G.C., 2000. Laboratory to field investigation in an integrated assessment of an acid mine
drainage-impacted watershed. Environmental Toxicology and Chemistry
19, 1036e1043.
Sullivan, M., Gray, N.F., ONeill, C., 1995. Synoptic overview of the Avocae
Avonmore catchment and the Avoca Mines. Technical Report: 26, Water
Technology Research, Trinity College. University of Dublin, Dublin,
ISBN 1-872220-27-4.
Thorp, V.J., Lake, P.S., 1973. Pollution of a Tasmanian river by mine effluents.
II. Distribution of macroinvertebrates. Internationale Revue der Gesamten
Hydrobiologie 58, 885e892.

40

N.F. Gray, E. Delaney / Environmental Pollution 155 (2008) 31e40

Washington, H.G., 1984. Diversity, biotic and similarity indices: a review with
special relevance to aquatic ecosystems. Water Research 18, 653e694.
Whiting, E.R., Mathieu, S., Parker, D.W., 1994. Effects of drainage from a molybdenum mine and mill on stream macroinvertebrate communities. Freshwater Ecology 9, 299e311.
Winner, R.W., Boesel, M.W., Farrell, M.P., 1980. Insect community structure
as an index of heavy metal pollution in lotic ecosystems. Canadian Journal
of Fisheries and Aquatic Sciences 37, 647e655.

Winterbourn, M.J., Collier, K.J., 1987. Distribution of benthic invertebrates in


acid, brown water streams in the South island of New Zealand. Hydrobiologia 153, 277e286.
Zamora-Mu~noz, C., Sainz-Cantero, C.E., Sanchez-Ortega, A., AlbaTercedor, J., 1995. Are biological indices BMWP0 and ASPT0 and their significance regarding water quality seasonally dependent? Factors explaining
their variations. Water Research 29, 285e290.