Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
Research paper
Relationships between reference site quality and baetid mayfly assemblages in
mountainous streams of the Luvuvhu catchment, South Africa
Pfananani Ramulifho1
, Nick Rivers-Moore2,3
and Stefan Foord4
1
Department of Environmental Sciences, Florida Science Campus, University of South Africa, South Africa
Centre for Water Resources Research, University of KwaZulu-Natal, Pietermaritzburg, South Africa
3
Freshwater Research Centre, Cape Town, South Africa
4
Chair in Biodiversity Value and Change in the Vhembe Biosphere Reserve, Department of Zoology, University of Venda, Thohoyandou,
South Africa
2
With water quality deteriorating rapidly at a global scale, river sections suited to serve as reference sites
are being increasingly lost. It thus becomes critical to develop rapid methods to confirm that previously
monitored sites continue meet the requirements of reliable reference sites. In the absence of pristine sites,
9 near-natural sites, as defined by the Kleynhans (1996) classification, were used as reference sites for the
Luvuvhu River catchment to compare the quality of physico-chemical factors against a biological metric.
Baetid mayfly community structure at a site was chosen as an index of water quality, since this family is
common in all types of freshwaters, highly diverse and adapted to unpolluted running water. Baetid larvae
were sampled monthly from stones-in-current biotopes across 9 sites for over 1 year, between December
2016 and January 2018. A Spearman’s correlation test was used to evaluate the relationship between physicochemical factors and identify redundant variables. Water quality standards were measured against the
national water quality guidelines for aquatic ecosystems. We used a generalized linear model to determine
the effect of physico-chemical variables on baetid species, and canonical correspondence analysis to show
the relationships between baetid species, sites, and physico-chemical variables. A total of 3 039 individuals
belonging to 12 mayfly species were recorded. Our findings indicated that while the physico-chemical
factors were highly variable, they were within favourable ranges to reflect reference site conditions. While
water temperature was the most important driver of baetid community structure in general, as it negatively
affected their abundances, a subset of species (Pseudoponnota sp., Pseudocloeon sp., Acanthiops varius and
Demoulinia crassi) showed clear responses to changes in TDS and stream width. We conclude that specific
baetid species show good potential as biological indicators of reference sites and chronic water temperature
stress, making assessment of reference sites easier.
INTRODUCTION
Despite the recognized importance of rivers in providing critical services to both humans and natural
organisms, their water quality is deteriorating at an alarming rate due to human activities (Tampo
et al., 2020). Worldwide, the quality of water in rivers is increasingly threatened (Dudgeon et al., 2006),
most specifically those in developing countries, due to industrialization, urbanization processes, and
constant changes in land uses (López-López and Sedeño-Díaz, 2015). The quality of river and stream
water is very sensitive to anthropogenic influences (urban, industrial and agricultural activities,
and increasing consumption of water resources), as well as natural processes like soil erosion and
weathering of the earth’s crustal material (Croijmans et al., 2020; Rashid and Romshoo, 2013; Hamid
et al., 2020). In South Africa, extensive efforts of monitoring both the ecological and water quality
conditions of rivers using nationally approved indices (e.g., River Eco-status Monitoring Programme
formerly known as the River Health Programme, also the Rapid Habitat Assessment Methods and
Models, etc.) is the responsibility of the Resource Quality Information Services Directorate of the
national Department of Water and Sanitation (DWS). In the 60-year long records, time-series data
show a growing deterioration of water quality that needs to be addressed more vigorously (Pitman,
2011). This is also reflected in the most recent national ecosystems and biodiversity status report,
which indicated that the condition of natural river ecosystems has declined by 11% between 1999
and 2011 (Skowno et al., 2019). From the 222 stream ecosystems assessed in South Africa, 64%
were found to be threatened and 43% among them were critically endangered. Similarly, in some
developed countries, such as Australia, the United States of America, and some European countries,
the monitoring of streams is a government obligation (López-López and Sedeño-Díaz, 2015;
Couceiro et al., 2012). Sustained action needs to be taken worldwide to prevent further deterioration
of rivers, failure of which might pose a health risk to aquatic life and people.
A key issue in the management and biomonitoring of aquatic systems is the establishment of
reference conditions against which to assess change and ecological trends over time (McDowell
at al., 2013). A practical definition of ‘reference condition’ is the chemical, physical and biological
conditions that can be expected in streams and rivers with minimal or no anthropogenic influence
(Soranno et al., 2011). Reference condition provides a baseline from which to compare changes in
water quality parameters and biological composition. There is a range of methods used to estimate
reference conditions, as mentioned in McDowell at al. (2013). However, in all the methods, the
ISSN (online) 1816-7950
Available on website https://www.watersa.net
251
CORRESPONDENCE
Pfananani Ramulifho
EMAIL
pfananani.ramulifho@gmail.com
DATES
Received: 18 September 2022
Accepted: 19 June 2023
KEYWORDS
bioindicator
ecological integrity
mayflies
reference condition
water temperature
COPYRIGHT
© The Author(s)
Published under a Creative
Commons Attribution 4.0
International Licence
(CC BY 4.0)
biological community of a stressed or disturbed ecosystem is
compared with that of relatively undisturbed reference sites
that have similar environmental conditions, when assessing the
impact of disturbance in multiple sites (Kaboré et al., 2018). If
the test-site community differs from the reference condition site,
the conclusion can be drawn that the site is impacted (Reece and
Reynoldson, 2001). Stream sections that are best suited to serve
as reference condition are increasingly challenging to locate
because of increasingly widespread anthropogenic impacts across
catchments (Soranno et al., 2011).
Many studies make ecological inferences based on the degree of
water quality as reflected by the presence or absence of aquatic
organisms (Aazami et al., 2015; Varnosfaderany et al., 2010;
Venkatesharaju et al., 2010; Beyene et al., 2009; Sharma and Rawat,
2009). There are several good reasons why macroinvertebrates are
useful as indicators of the reference conditions of rivers. These
reasons include their persistence across seasons, their species
diversity, and ubiquitous occurrence in almost all types of the
world’s freshwater ecosystems (Buss and Salles, 2007). Amongst
the macroinvertebrate taxa found in the tropics and the southern
hemisphere, baetidae are more endemic and show more important
adaptation traits to local afrotropical conditions than others
(Barber-James et al., 2008; Gattolliat and Nieto, 2009). Several
studies have demonstrated that baetid community structure
reflects the environmental state of rivers effectively (Kubendran
et al., 2017; Buss and Salles, 2007; Bauernfeind and Moog, 2000).
Mayflies are characterized by narrow habitat tolerance and
only occur in very clean freshwater, which makes them good
bioindicators for very good water quality (Alhejoj et al., 2023; Buss
and Salles, 2007; Kubendran et al., 2017).
METHODOLOGY
Study area
The study was conducted in the south-eastern streams of the
Soutpansberg Mountains, Limpopo Province, South Africa.
Nine sampling sites were selected along four streams (Dzindi,
Mutshundudi, Lutanandwa and Tshirovha), all of which are
major streams of the Luvuvhu River catchment. These sites are
located in the uppermost 5 km stream segment within their
respective streams, and they hold both an instream and riparian
zone habitat integrity of 60% and 90%, respectively (Kleynhans,
1996). These streams have continuous flow of water throughout
the year during both the dry and rainy seasons. All sites showed
high similarities in their physical characteristics and biotopes
and were in the foothill zone (Rowntree and Wadeson, 1999),
with stream orders of 1 and at elevations of 622–1 022 m amsl.
(Fig. 1). The catchment experiences wet summers from October
to April with peak rainfall in January and February. The mean
annual precipitation is 608 mm, while the mean annual air
temperatures are 17°C in mountainous areas and 24°C near
the Kruger National Park (Singo et al., 2012). The width of the
active channel of the sampled sites ranged from 3.45 m at Thathe
waterfall to 11.42 m at Tshirovha. These sites are near-natural,
with intact vegetation cover and very little to no human impact.
Based on the habitat integrity assessment of Kleynhans (1996),
these sites have limited indigenous vegetation removal, little
exotic vegetation encroachment and water abstraction. Sites were
chosen to represent a pristine gradient of physico-chemical and
environmental conditions and macroinvertebrate community
assemblages.
The interactions between environmental factors and baetid
abundances is crucial since this nexus has potential to enhance
ecosystem services that baetid species provide. Available evidence
shows that they provide many essential services that maintain
and enhance ecosystem function, such as energy flow dynamics
(Boyero et al., 2011; Jacobus et al., 2019). Some baetid species are
good manipulators of organic matter like periphyton and sediment
(Buss and Salles, 2007; Baptista et al., 2006). Baetids process large
amounts of organic matter, allochthonous carbon and nutrients
from riparian vegetation and soil materials (Moulton et al., 2004),
which are used by organisms at higher trophic levels (Wallace and
Webster, 1996; Boyero et al., 2011). According to Wallace and
Webster (1996), most baetid species are generally primary prey
for invertebrate predators and they also contribute in various ways
to energy flow and nutrient cycling. Some filter feeders of mayflies
(including most baetid species) contribute to water purification
and are part of arguably the most important of these predatorprey relationships – as the diet of fish – which is also a driver of
the domestic food and local economy (Jacobus et al., 2019).
Biological and physico-chemical sampling
To our knowledge, no exclusive studies have been undertaken on
the relationship between reference conditions based on physicochemical parameters and baetid community composition. The
similarities of baetid composition between reference condition
sites have not been explored and it is unknown if the physicochemical composition differs across these mountainous rivers.
Furthermore, despite the widespread occurrence of this mayfly
family, it is still unclear if this family is influenced by physicochemical parameters at these sites. We ask the question of whether
the value of sites to still act as reference sites can be quantified using
baetid species. In this survey, we used correlation models to assess
the degree of similarity between physico-chemical characteristics
and baetid species composition at reference sites. Our objective
was to quantify the relative role of physico-chemical factors in
structuring baetid species assemblages, using the Luvuvhu River
catchment as a case study.
At each site, one measurement of four physico-chemical
parameters (water temperature, conductivity, total dissolved
solids (TDS), pH) was taken using a portable pH/EC/TDS/
temperature multi-meter. Other variables (environmental) whose
single reading was measured include elevation, habitat area (stone
size), flow depth, stream width, and flow velocity rate using a Flow
Globe FP101 reader.
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
All biological samples were taken from the ‘stones-in-current’
hydraulic biotope because sub-imagos (nymphs) of many baetid
species inhabit this riffle section of streams and river (Bauernfeind
and Moog, 2000). These sites were all well aerated (Fig. 2) and
provide a home to a variety of macroinvertebrate organisms
(Ramulifho et al., 2020). Each site was sampled on a monthly basis
from December of 2016 to January of 2018. All sites were sampled
within the same single week at daylight during each sampling
month to allow for consistency in weather and flow conditions
across sites. Six stones containing organisms were sampled
at each site using a standard SASS net. All contents from a net
were emptied into a sample bottle and sorted in the laboratory,
and baetid larvae were then identified. Most of the material was
identified to species level, while some early instar larvae were only
identified to the genus or morpho-species level using taxonomic
keys (De Moor et al., 2003). Specimens were preserved in 70%
ethanol and are housed at the reference collection section of the
SARCHi offices, University of Venda.
Data analysis
The baetid composition was analysed using species richness and
abundance metrics. Baetid species abundance data taken from 6
sampling stones at each site was pooled to then represent a single
monthly abundance sample at each site. Biological data were
tested for normality using the normal quantile-quantile (Q-Q)
test which showed that the data were normally distributed.
252
Figure 1. Location of sampling sites in the upstream sections of the south-western side of the Luvuvhu catchment in South Africa
Figure 2. Natural reference sites at Tshirovha potholes (left) and Thathe waterfalls (right) (Photos: Pfananani Ramulifho)
Physico-chemical data was standardized using log10 (x + 1) to
achieve the assumed conditions of normality and homoscedasticity
(Buss and Salles, 2007), while no species in the biological data was
down-weighted. Standardization of physico-chemical data was
necessary to reduce the influence of large differences and double
zeros, to normalize and render data homoscedastic (Clarke and
Gorley, 2006).
To avoid multi-colinearity between physicochemical variables, we
calculated the non-parametric rank-based Spearman correlation
between these variables. The departure from reference condition
of physicochemical variables was made by comparison with the
corresponding standards prescribed for aquatic ecosystems in
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
South Africa (DWAF, 1996). Water temperature should not vary
(standard deviate) from the mean temperature for that specific site
by > 2°C, while the mean TDS should not vary (standard deviate)
by > 15% (DWAF, 1996). Most freshwaters in South Africa are
relatively well buffered and more or less neutral, with pH ranges
between 6 and 8, and pH should not vary (standard deviate)
from the mean values for a specific site by > 0.5 (DWAF, 1996).
Where no specific reference condition criteria were prescribed by
national water quality guidelines (e.g. EC concentrations), peerreviewed reports and articles published from areas of similar
geographic or climatic region to this study with such specification
were used.
253
We used generalized linear mixed model (GLMM) with negative
binomial regression from the ‘MASS’ package and ‘glmer.nb’
function (Nakagawa and Schielzeth, 2013; Jamil and Ter Braak,
2013) to evaluate the relative importance of each physico-chemical
variable on the abundance of baetidae species from all nine sites
combined. GLMM is an extension to the generalized linear
model (GLM) in which the linear predictor contains random
effects in addition to the usual fixed effects (Venables and Ripley,
2002). Since no collinearity existed between physico-chemical
variables, we ran one full model of GLMM with all physicochemical variables. During the analyses, sites were used as a
random factor to account for temporal pseudo-replication, while
all physico-chemical variables were included as fixed variables
(Li et al., 2018). The goodness-of-fit of the models was assessed
using the relations between the residuals (the differences between
observations and predictions by the retained model) and physicochemical variables. We also determined the correspondence of
physico-chemical variables, sites and baetid species during the
sampling period, using the forward addition of correspondence
variables technique in canonical correspondence analysis (CCA)
(Ter Braak and Verdonschot, 1995). The statistical significance of
each variable selected in CCA was judged using a Monte-Carlo
permutation test (Klonowska-Olejnik and Skalski, 2014). All the
statistical analyses were performed in R (R Core Team, 2022).
RESULTS
Sites condition and correlation between
physico-chemical variables
The sampling sites (Table 1) had a mean elevation of 781.09 m,
with the highest at 1 022.94 m and the lowest at 622.38 m.
Conductivity ranged from 1.9 to 42 μs.cm-1, averaging at
21.30 μs.cm-1. Water temperatures ranged from 13.7 to 26.3°C,
with a mean of 18.84°C. TDS ranged from 2.96 to 31.9 mg.L-1 with
6 of the 9 sites fluctuating by over 15% of a mean of 13.57 mg.L-1.
The mean pH value of these sites ranged between 6.50 and 7.56,
with some (55% of sites) having a change (standard deviation) of
> 0.5 over time. Stream flow velocities ranged from 0.2 to 6.4 m.s-1,
with a mean of 1.76 m.s-1. Mean stream flow depth was 13.93 cm
and was highly variable, ranging between 2 and 48.5 cm. Stream
width varied between 2.1 and 20 m, with a mean of 8.03 m. Stones
had a mean size of 12.69 cm and ranged from 4.66 to 46 cm.
No correlation between physico-chemical variables was ≥ 0.7
(Table 2). More than 85% of these variables were positively
correlated. As expected, highest positive correlation was found
between conductivity and TDS (R2 = 0.63), since these two
water quality parameters are related and are used to describe
salinity levels in water. There were highly significant correlations
Table 1. Mean values and standard deviation (±) of physico-chemical parameters, total abundance, number of genus and species richness at the
sampling sites during the period of sampling
Name
Elevation (m)
Stone size (cm)
Stream flow (m³·s¯¹)
Stream depth (cm)
Stream width (m)
Water temperature (°C)
TDS (mg·L¯¹)
pH
Conductivity (μs·cm¯¹)
Abundance (N)
Genera (N)
Species (N)
Upper
Tshirovha Tshirovha Tshirovha Thathe
Tea
Lwamondo Lutanandwa Lutanandwa Average
Lutanandwa potholes
forest
waterfall estate
bridge
745.26
934.02
665.26
622.38
1 022.94 876.59
796.01
651.19
716.14
781.09
12.93
11.46
11.04
13.35
13.67
13.46
15.30
10.87
12.16
12.69
± 3.24
± 3.16
± 3.16
± 3.74
± 1.94
± 3.74
± 3.68
± 3.24
± 3.64
1.12
1.31
1.11
1.95
2.25
2.11
2.03
2.10
1.84
1.76
± 0.61
± 0.81
± 0.83
± 1.02
± 1.03
± 1.03
± 1.00
± 0.85
± 0.99
15.33
14.93
10.80
15.18
10.00
15.88
15.26
13.62
14.33
13.93
± 4.50
± 4.62
± 4.50
± 5.75
± 3.16
± 5.74
± 5.61
± 4.58
± 5.47
5.39
10.49
10.13
11.42
3.05
8.42
8.00
9.45
5.94
8.03
± 1.93
± 2.12
± 2.12
± 2.21
± 2.16
± 2.21
± 2.22
± 2.12
± 2.25
18.04
17.78
18.96
18.11
21.80
18.26
17.81
19.57
19.20
18.84
± 3.36
± 3.11
± 3.00
± 2.91
± 1.61
± 2.90
± 2.79
± 3.12
± 2.84
14.40
14.71
1.93
14.57
7.17
13.74
13.30
15.24
16.33
13.75
± 1.15
± 1.62
± 1.61
± 3.57
± 9.03
± 3.58
± 3.59
± 1.74
± 3.60
7.36
6.76
6.55
7.14
6.50
6.96
7.15
7.56
7.34
7.04
± 0.45
± 0.48
± 0.46
± 0.51
± 0.52
± 0.51
± 0.51
± 0.46
± 0.51
26.97
22.95
22.71
21.90
9.83
19.27
20.53
22.44
25.09
21.30
± 3.03
± 2.70
± 2.71
± 6.17
± 12.86
± 6.13
± 6.16
± 2.96
± 6.17
233.00
211.00
31.00
129.00
4.00
791.00
620.00
494.00
526.00
337.67
± 4.09
± 4.48
± 4.49
± 4.53
± 2.36
± 4.53
± 4.55
± 4.48
± 4.56
4
7
5
7
2
6
6
6
7
5.56
5
8
6
9
2
7
7
7
9
6.67
Table 2. Correlation coefficients (R2) among physico-chemical parameters during the period of sampling. Significance: *p < .05, **p < .01,
and ***p < .001.
TDS (mg·L¯¹)
Stone size (cm)
Stream flow (m³·s¯¹)
Stream depth (cm)
Stream width (m)
Water temperature (°C)
pH
Conductivity (μs·cm¯¹)
Elevation (m)
TDS
(mg·L¯¹)
1.00
−0.24
−0.11**
−0.27
−0.07
−0.27
0.32
0.63
0.14***
Stone
size (cm)
−0.24
1.00
0.24
0.10**
0.01
0.04
0.18
0.23
0.12***
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
Stream
Stream
Stream
flow (m³·s¯¹) depth (cm) width (m)
−0.11**
−0.27
−0.07
0.24
0.10**
0.01
1.00
0.03
0.12***
0.03
1.00
0.23
0.12***
0.23
1.00
0.21
0.20
0.21
0.06
0.05
0.20
0.19
0.25
0.10**
0.04
0.07
0.04
Water
temperature (°C)
−0.27
0.04
0.21
0.20
0.21
1.00
0.28
0.21
0.14***
pH
Conductivity Elevation
(μs·cm¯¹)
(m)
0.32
0.63
0.14***
0.18
0.23
0.12***
0.06
0.19
0.04
0.05
0.25
0.07
0.20
0.10**
0.04
0.28
0.21
0.14***
1.00
0.25
0.36
0.25
1.00
0.18
0.36
0.18
1.00
254
(p < 0.001) between elevation and three variables (TDS, stone
size and water temperature; R2 < 0.15). Similarly, stream flow and
stream width were significantly correlated with R2 < 0.15. Some
significant correlations (p < 0.01) that were also observed included
those between stream flow and TDS, stream depth and stone size,
and between conductivity and stream width. The lowest negative
correlation had a coefficient of −0.27 and was between TDS and
stream depth, and between TDS and water temperature.
Abundance of baetids and effect of physico-chemical
factors
A total of 3 039 individuals of baetidae belonging to 9 genera
and 12 species were recorded in this study. The highest number
of individuals caught was 28 specimens at Lwamondo (during
low-flow period), while the average catch across the sites was 6
specimens. The highest number of baetid species was recorded
at mid-Lutanandwa and the Tshirovha confluence, with 9 species
(Table 1). The tea estate site had the highest number of individuals
(791), and together with Lutanandwa bridge and Lwamondo sites
had the third highest diversity, with 7 species after Tshirovha
potholes (Table 1). The lowest number of species was recorded at
Thathe waterfall, with only 2 species. The most abundant species
in the streams were Baetis Harissoni and Dabulamanzia media,
and these were also the most widespread species, occurring at
8 of the 9 sites. Centroptiloides bifasciata and Demoulinia crassi
were each limited to 1 site, with low numbers of individuals at
Lwamondo and Upper Lutanandwa (3 and 4, respectively).
The results suggest that, overall, the most significant drivers
of baetid species abundance were water temperature (GLMM:
estimate = −0.26, p < 0.001), followed by conductivity (GLMM:
estimate = 0.12, p < 0.01) and stone size (GLMM: estimate = 0.08,
p < 0.05) (Table 3). An increase in water temperature negatively
affected the abundance of baetid species, as opposed to an increase
of both conductivity and stone size, which had a positive population
effect (Fig. 3). Non-significant drivers of baetid abundance
included stream flow, stream width, stream depth, TDS, pH, and
elevation. Amongst all these drivers, only stream width increased
with a decrease in baetid species abundance (Fig. 3).
Figure 3. Regression plots of baetid species abundance in relation to various physico-chemical gradients during the period of sampling
Table 3. Generalized linear mixed model analyses between Baetidae species and physico-chemical variables at the sampling sites during the
period of survey (significance: *p < 0.05, **p < 0.01, and ***p < 0.001)
Variables
Estimate
Std. error
z-value
p-value
Temperature (°C)
−0.26709
0.04201
−6.357
2.06 x 10 −10 ***
Conductivity (μs·cm¯¹)
0.12966
0.04073
3.184
0.00145**
Stone size (cm)
0.08074
0.03367
2.398
0.01647*
TDS (mg·L¯¹)
0.08446
0.04465
−1.892
0.05853
Stream flow (m³·s¯¹)
0.01838
0.03483
0.528
0.59779
Stream depth (m)
0.01206
0.03614
0.334
0.73861
Stream width (m)
−0.02987
0.06085
0.491
0.62357
Elevation (m)
0.07218
0.13275
0.544
0.58665
pH
0.05846
0.04446
1.315
0.18861
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
255
Site preference of baetid species
The percentage of variance explained by Axes 1−4 of the canonical
correspondence analysis amounted to 76.88%, with the 1st and 2nd
axes explaining 32.4% and 28%, while only a small variation of
9.56% and 6.92% of the total variance was explained by the 3rd and
4th axes, respectively. According to the canonical correspondence
analysis, there are two distinct patterns of baetid species
preferences for sites. The first correspondence component (CA1)
represented a gradient where most physico-chemical variables
(temperature, elevation, stone size, stream flow) load strongly
positively on this component, while TDS and stream flow depth
load negatively. This positive trend of physico-chemical variables
is associated with the Tshirovha site, and Pseudocloeon sp. CA1
is also associated with a decrease in TDS which is closely linked
to the decline in abundance of Pseudoponnota sp. at Tshirovha
forest site. CA2 showed a prominent negative loading of physicochemical variables such as TDS, stream width, stream flow and
stone size, associated with Tshirovha, Thathe waterfall, Tshirovha
potholes, and Tshirovha forest. Further observation showed that
Acanthiops varius is negatively affected, while Demoulinia crassi
abundance increases with an increase in temperature, elevation
and stream depth, and is more closely related to the Upper
Lutanandwa site. The majority of baetid species (8 of 12 species or
66.66%) showed no clear response to changes in physico-chemical
parameters in the study area (Figure 4). These species are clustered
between the nine sampling sites.
DISCUSSION
Physico-chemical variables and baetid community
composition
Strong positive correlations between physico-chemical variables
which were highly significant (e.g., elevation to TDS, stone size
and water temperature) were observed in this study at numerous
sites (Table 2). This was expected from sampling sites which
are seemingly influenced by both closely related sources (the
Soutpansberg mountains) and land uses, as shown in Abowei
(2010). This low variation within physico-chemical variables
was characteristic of all sampling sites. The absence of industrial
activities at these sites is evident by pH levels for sampled sites which
were all within the recommended South African aquatic system
pH range of 6–8 (DWAF, 1996), and also as observed by Monyai
et al. (2016). Acidic effluents from industrial activities are known
to cause low pH levels in rivers (e.g., mine drainage, paper, tanning
and leather industries). The visual evidence from stream water
showed no black or brown (tea-coloured) water or any filamentous
algae (Fig. 2), which is usually caused by changing pH levels. The
water temperature standard for sustaining aquatic life is 20–30°C
(Weldemariam, 2013). This study was dominated by sites with
relatively lower temperature range within the accepted thresholds
(Table 1). This could be due to forest cover at the sites, which reduces
light incidence keeping the stream water temperature at low values
(Siegloch et al., 2014; Klonowska-Olejnik and Skalski, 2014), and
the effect of altitude The concentration of TDS and EC at the nine
sites was well within the WHO standard for inland surface water of
1 000 mg.L-1 and 300 μs.cm-1 (WHO, 2011). This was expected at
all these sites due to the absence of practices such as enrichment by
soaps and detergents from people washing or bathing in streams,
which would result in high levels of TDS and EC, placing stress
on aquatic species (Monyai et al., 2016). This might also mean the
absence of land-use practices such as overgrazing, non-contour
ploughing, removal of riparian vegetation and forestry operations
adjacent to these sites. These practices accelerate erosion or result
in increased loads of suspended solids in rivers (Monyai et al.,
2016; Adu and Oyeniyi, 2019). Most of the environmental variables
recorded in this study were within levels prescribed by DWAF and
WHO and should be able to support aquatic life (DWAF, 1996;
WHO, 2011; Weldemariam, 2013).
It is evident from this study that not all species have the same
response to environmental parameters (Table 3 and Fig. 4).
Water temperature explained the most significant amount of
variation in relative abundance, as has been reported in other
studies (Ramulifho et al., 2020; Buss and Salles, 2007; Jacobus
et al., 2019; Adu and Oyeniyi, 2019; Bauernfeind and Moog, 2000).
Water temperature in these sites is driven by riparian vegetation.
Riparian vegetation is vital for maintaining and ensuring suitable
water temperature and the amount of light available as it also
forms a buffer area for the stream. A similar study of undisturbed
sites by Klonowska-Olejnik and Skalski (2014) found that the
intactness of riparian vegetation is one of the most important
factors structuring communities. Another similar study of upper
catchment sites by Svitok (2006) concluded that mayfly abundance
was most strongly related to elevation, which also relates to the
climatic variable of air temperature. The findings of this signal
important concerns regarding potential species movement and
survival in the face of climate change predictions.
Figure 4. CCA plot with Axes 1 and 2 showing variation between sites, physico-chemical variables and Baetid species. The contribution of each
variable is proportional to the length of the arrow. Temp = water temperature; Elev = elevation; TDS = total dissolved solids; Width = stream flow
width; Depth = flow depth; Flow = stream flow.
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
256
Eight species of baetid did not show any conclusive response to
physico-chemical variables and sites. Only a few species were
associated with specific sampling sites (Fig. 4). Pseudoponnota
sp., Pseudocloeon sp., and A. varius showed a considerable degree
of preference for TDS, temperature and stream width linked
to Tshirovha potholes, Tshirovha forest, and Thathe waterfall,
which are sites found in close proximity each other and on
the same stream. Ubiquitous species such as Baetis harissoni
and Dabulamanzia media showed no preference for measured
conditions in streams. These species are generalist in their nature
in the Luvuvhu catchment as they tolerate a range of conditions
(Ramulifho et al., 2020). Studies globally have largely used
different species of baetids as valid biological indicators of water
quality because they are highly sensitive to substrate changes
(Kubendran et al., 2017; Buss and Salles, 2007; Bauernfeind
and Moog, 2000). In this study, it is evident that baetids showed
varied tolerance levels to pollution, but generally are considered
intolerant organisms and require water of good quality to survive,
as also shown by Alhejoj et al. (2014).
Benefits of baetids in confirming reference site quality
Results from this study further enhance the use of baetid species
as a low-cost indicator for aquatic reference sites that allow
quick, widespread, long-term, routine monitoring and direct
comparison of sites, time periods and studies (Butana et al.,
2010). The use of natural variation of baetid species in reference
conditions also helps to avoid the setting of quantitative limits
or targets of physico-chemical factors (of reference sites) that
are either too restrictive or impossible to meet in the face of
changing land use and rapid industrialization (McDowell et al.,
2013). Thus, even in areas where there is a deficiency of physicochemical and environmental data, by using baetid species there is
still a possibility that reference conditions may well be established.
This biological approach to reference site selection enables the
measurement of a natural continuum of the substantial benefits
of baetid species ecosystem processes, such as nutrient cycling,
algal distribution, retention and distribution of organic matter,
and predator-prey interactions (Jacobus et al., 2019; Sartori and
Brittain, 2015; Wallace and Webster, 1996).
The usefulness of this research lies in its contribution towards
closing an existing gap on a biological index of baetids species in
reference sites in South Africa. This research establishes preliminary
baseline biological characteristics of the potential reference sites in
mountain rivers, as opposed to widely used selection criteria like
chemical and physical (i.e., abiotic) factors as surrogates (Agboola
et al., 2020). The biological indices are widely recommended
and a valuable tool in monitoring macroinvertebrate response,
reference conditions and anthropogenic disturbances in rivers
in many regions including Europe (Lewin et al., 2013) and west
Africa (Kaboré et al., 2018). If this is adopted for local streams,
accurate estimation of reference conditions based on biological
indices will provide information on anthropogenic impacts and
stress for sites in upstream catchments and potential areas for
restoration of reference conditions (McDowell et al., 2013).
CONCLUSIONS
Our findings indicated that the physico-chemical factors at the
selected sites are highly variable but are still in a favourable range
for reference site conditions. Direct effects of measured physicochemical factors on the entire baetid community were evident
largely for Pseudoponnota sp., Pseudocloeon sp., Demoulinia crassi
and Acanthiops varius. Since the presence or absence of certain
mayflies was strongly influenced by water temperature, TDS, and
stream width (as observed from the models), this study confirmed
that these species are a powerful tool as descriptors of reference
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
sites. These results are of relevance for protection of these species
and reference sites in catchments in South Africa.
ACKNOWLEDGMENTS
This study was supported financially and logistically by the
National Research Foundation and the Department of Science and
Technology through the South African Research Chairs Initiative
(SARChI) Chair on Biodiversity Value and Change in the Vhembe
Biosphere Reserve, hosted and supported by the University of
Venda. We thank the Department of Zoology, University of Venda,
for providing laboratory and field work facilities.
DATA AVAILABILITY STATEMENT
Data, models or codes that support the findings of this study are
available from the corresponding author upon request.
AUTHOR CONTRIBUTIONS
Pfananani Ramulifho conceived the study, wrote the initial draft of
the manuscript and performed data analyses. Nick Rivers-Moore
and Stefan Foord edited and commented on the manuscript. All
authors contributed to discussions that shaped the manuscript.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ORCIDS
Pfananani Ramulifho
https://orcid.org/0000-0002-1589-7899
Nick Rivers-Moore
https://orcid.org/0000-0002-6546-4215
Stefan Foord
https://orcid.org/0000-0002-9195-2562
REFERENCES
AAZAMI J, ESMAILI-SARI A, ABDOLI A, SOHRABI H and VAN
DEN BRINK PJ (2015) Monitoring and assessment of water health
quality in the Tajan River, Iran, using physicochemical, fish and
macroinvertebrates indices. J. Environ. Health Sci. Eng. 13 29.
https://doi.org/10.1186/s40201-015-0186-y
ABOWEI JFN (2010) Salinity, dissolved oxygen, pH and surface water
temperature conditions in Nkoro River, Niger Delta, Nigeria. Adv. J.
Food Sci. Technol. 2 (1) 36–40.
ADU BW and OYENIYI EA (2019) Water quality parameters and
aquatic insect diversity in Aahoo stream, southwestern Nigeria. J.
Basic Appl. Zool. 80 15 https://doi.org/10.1186/s41936-019-0085-3
AGBOOLA OA, DOWNS CT and O’BRIEN G (2020) A multivariate
approach to the selection and validation of reference conditions in
KwaZulu-Natal Rivers, South Africa. Front. Environ. Sci. 8 584923.
https://doi.org/10.3389/fenvs.2020.584923
ALHEJOJ I, ELIAS S and KLAUS B (2014) Mayflies (Order
Ephemeroptera): an effective indicator of water bodies conditions
in Jordan. Int. J. Sci. Res. Environ. Sci. 2 (10) 361–370. https://doi.
org/10.12983/ijsres-2014-p0361-0370
ALHEJOJ I, HIASAT TH, SALAMEH E, HAMAD AA and
AL KUISI M (2023) Use of the aquatic mayfly (Insecta: Ephemeroptera) as environmental bio-indicator in Jordan. Int. J. Design Nat.
Ecodyn. 8 (1) 133–139. https://doi.org/10.18280/ijdne.180115
BAPTISTA DF, BUSS DF, DIAS LG, NESSIMIAN JL, DA SILVA ER,
NETO AD, DE CARVALHO SN, DE OLIVEIRA MA, ANDRADE
LR (2006) Functional feeding groups of Brazilian Ephemeroptera
nymphs: Ultrastructure of mouthparts. Int. J. Limnol. 42 (2) 87–96.
https://doi.org/10.1051/limn/2006013
BAUERNFEIND E and MOOG O (2000) Mayflies (Insecta: Ephemeroptera) and the assessment of ecological integrity: a methodological
approach, (pp. 71–83). In: Jungwirth M, Muhar S and Schmutz S
(eds) Assessing the Ecological Integrity of Running Waters. Springer,
The Netherlands. https://doi.org/10.1007/978-94-011-4164-2_6
257
BEYENE A, ADDIS T, KIFLE D, LEGESSE W, KLOOS H and
TRIEST L (2009) Comparative study of diatoms and macroinvertebrates as indicators of severe water pollution: case study of the
Kebena and Akaki rivers in Addis Ababa, Ethiopia. Ecological
Indicators 9 381–392. https://doi.org/10.1016/j.ecolind.2008.05.001
BOYERO L, PEARSON RG, DUDGEON D, GRAÇA MA, GESSNER
MO, ALBARIÑO RJ and CALLISTO M (2011) Global distribution
of a key trophic guild contrasts with common latitudinal diversity
patterns. Ecology 92 (9) 1839–1848. https://doi.org/10.1890/102244.1
BULUTA S, FINAU I, BRODIE G and HODGE S (2010) A preliminary
study into the potential of mayflies (Ephemeroptera: Baetidae and
Caenidae) as bio-indicators of stream health in Fiji. S. Pacific J.
Nat. Appl. Sci. 28 (1) 82–84. https://doi.org/10.1071/SP10011
BUSS DF and SALLES FF (2007) Using Baetidae species as biological
indicators of environmental degradation in a Brazilian river basin.
Environ. Monit. Assess. 130 365–372. https://doi.org/10.1007/s10661006-9403-6
CHRISTIDIS F, CONNOLLY NM, NOLEN JA, PEARSON RG,
CLAIR R, CAIRNS A and DAVIS L (2017) Stream macroinvertebrate assemblage uniformity and drivers in a tropical bioregion.
Freshwater Biol. 62 (3) 544–558. https://doi.org/10.1111/fwb.12884
CLARKE KR and GORLEY RN (2006) Primer v.6: User Manual/
Tutorial. PRIMER-E.
COUCEIRO SR, HAMADA N, FORSBERG BR, PIMENTEL TP
and LUZ SL (2012) A macroinvertebrate multimetric index to
evaluate the biological condition of streams in the Central Amazon
region of Brazil. Ecol. Indic. 18 118–125. https://doi.org/10.1016/j.
ecolind.2011.11.001
CROIJMANS L, DE JONG JF and PRINS HH (2020) Oxygen is a
better predictor of macroinvertebrate richness than temperature –
a systematic review. Environ. Res. Lett. 21;16(2):023002. https://doi.
org/10.1088/1748-9326/ab9b42
DE MOOR IJ, DAY JA and DE MOOR FC (2003) Guides to the
freshwater invertebrates of Southern Africa. Vol. 7: Insecta I
(Ephemeroptera, Odonata and Plecoptera). WRC Report No. TT
207/03. Water Research Commission, Pretoria.
DUDGEON D, ARTHINGTON AH, GESSNER MO, KAWABATA ZI,
KNOWLER DJ, LÉVÊQUE C, NAIMAN RJ, PRIEUR-RICHARD
AH, SOTO D, STIASSNY ML, SULLIVAN CA (2006) Freshwater
biodiversity: importance, threats, status and conservation
challenges. Biol. Rev. 81 163–182. https://doi.org/10.1017/S14647
93105006950
DWAF (Department of Water Affairs and Forestry, South Africa)
(1996) South African Water Quality Guidelines. Volume 7: Aquatic
Ecosystems. Department of Water Affairs and Forestry, Pretoria.
EDOKPAYI JN, ODIYO JO, POPOOLA OE and MSAGATI TAM
(2016) Assessment of trace metals contamination of surface water
and sediment: a case study of Mvudi River, South Africa.
Sustainability 8 135. https://doi.org/10.3390/su8020135
GATTOLLIAT JL and NIETO C (2009) The family Baetidae (Insecta:
Ephemeroptera): synthesis and future challenges. Aquat. Insects 31
(sup1) 41–62. https://doi.org/10.1080/01650420902812214
GUSTAFSON MP (2008) Effects of thermal regime on mayfly
assemblages in mountain streams. Hydrobiologia 605 (1) 235–246.
https://doi.org/10.1007/s10750-008-9357-5
HAIDEKKER A and HERING D (2008) Relationship between benthic
insects (Ephemeroptera, Plecoptera, Coleoptera, Trichoptera) and
temperature in small and medium-sized streams in Germany: A
multivariate study. Aquat. Ecol. 42 (3) 463–481. https://doi.org/
10.1007/s10452-007-9097-z
HAMID A, BHAT SU and JEHANGIR A (2020) Local determinants
influencing stream water quality. Appl. Water Sci. 10 (1) 1–16.
https://doi.org/10.1007/s13201-019-1043-4
HEATH RGM and CLASSEN M (1990) An overview of the pesticide
and metal levels present in populations of the larger indigenous fish
species of selected South African rivers. WRC Report No. 428/1/99.
Water Research Commission, Pretoria.
JACOBUS LM, MACADAM CR and SARTORI M (2019) Mayflies
(Ephemeroptera) and their contributions to ecosystem services.
Insects 10 170. https://doi.org/10.3390/insects10060170
JAMIL T and TER BRAAK CJF (2013) Generalized linear mixed
models can detect unimodal species-environment relationships.
PeerJ 1 e95. https://doi.org/10.7717/peerj.95
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
KABORÉ I, MOOG O, OUÉDA A, SENDZIMIR J, OUÉDRAOGO
R, GUENDA W and MELCHER AH (2018) Developing reference
criteria for the ecological status of West African rivers. Environ.
Monit. Assess. 190 2. https://doi.org/10.1007/s10661-017-6360-1
KARR JR (1999) Defining and measuring river health. Freshwater
Biol. 41 221–234. https://doi.org/10.1046/j.1365-2427.1999.00427.x
KLEYNHANS CJ (1996) A qualitative procedure for the assessment
of the habitat integrity status of the Luvuvhu River (Limpopo
system, South Africa). J. Aquat. Ecosyst. Health 5 41–54. https://doi.
org/10.1007/BF00691728
KLONOWSKA-OLEJNIK M and SKALSKI T (2014) The effect of
environmental factors on the mayfly communities of headwater
streams in the Pieniny mountains (west carpathians). Biologia 69 (4)
498–507. https://doi.org/10.2478/s11756-014-0334-3
KUBENDRAN T, SELVAKUMAR C, SIDHU AK, NAIR A and
KRISHNAN SM (2017) Baetidae (Ephemeroptera : Insecta) as
biological indicators of environmental degradation in Tamiraparani
and Vaigai river basins of Southern Western Ghats, India. Int. J.
Curr. Microbiol. Appl. Sci. 6 (6) 558–572. https://doi.org/10.20546/
ijcmas.2017.606.066
KUSANGAYA S, WARBURTON ML, ARCHER VAN GARDEREN E
and JEWITT GPW (2013) Impacts of climate change on water
resources in southern Africa: A review. Phys. Chem. Earth 67 47–54.
https://doi.org/10.1016/j.pce.2013.09.014
LEWIN I, CZERNIAWSKA-KUSZA I, SZOSZKIEWICZ K,
EWA-LAWNICZAK A and JUSIK S (2013) Biological indices
applied to benthic macroinvertebrates at reference conditions of
mountain streams in two ecoregions (Poland, the Slovak Republic).
Hydrobiologia 709 183–200. https://doi.org/10.1007/s10750-013-14
48-2
LI L, ZHENG B and LIU L (2010) Biomonitoring and bioindicators
used for river ecosystems: definitions, approaches and trends.
Proced. Environ. Sci. 2 1510–1524. https://doi.org/10.1016/j.proenv.
2010.10.164
LI S, YANG W, WANG L, CHEN K, XU S and WANG B (2018)
Influences of environmental factors on macroinvertebrate assemblages: differences between mountain and lowland ecoregions,
Wei River, China. Environ. Monit. Assess. 190 152. https://doi.
org/10.1007/s10661-018-6516-7
LÓPEZ-LÓPEZ E and SEDEÑO-DÍAZ JE (2015) Biological indicators
of water quality: the role of fish and macroinvertebrates as
indicators of water quality. In: Armon R and Hänninen O (eds)
Environmental Indicators. Springer, Dordrecht. https://doi.org/10.
1007/978-94-017-9499-2_37
MANGADZE T, DALU T and FRONEMAN PW (2019) Biological
monitoring in southern Africa: a review of the current status,
challenges and future prospects. Sci. Total Environ. 648 1492–1499.
https://doi.org/10.1016/j.scitotenv.2018.08.252
MCDOWELL RW, SNELDER TH, COX N, BOOKER DJ and
WILCOCK RJ (2013) Establishment of reference or baseline
conditions of chemical indicators in New Zealand streams and
rivers relative to present conditions. Mar. Freshwater Res. 64
387–400. https://doi.org/10.1071/MF12153
MONYAI M, MAKHADO RA and NOVHE NO (2016) Water
quality of the Luvuvhu River and its tributaries within the
Thulamela Local Municipality, Limpopo Province, South Africa.
Afr. J. Sci. Technol. Innovat. Dev. 8 (5-6) 439–445. https://doi.org/10.
1080/20421338.2016.1219504
MOULTON TP, SOUZA ML, SILVEIRA RML and KRSULOVIĆ
FAM (2004) Effects of ephemeropterans and shrimps on
periphyton and sediments in a coastal stream (Atlantic forest, Rio
de Janeiro, Brazil). J. N. Am. Benthol. Soc. 23 868–881. https://doi.
org/10.1899/0887-3593(2004)023<0868:EOEASO>2.0.CO;2
NAKAGAWA S and SCHIELZETH H (2013) A general and simple
method for obtaining R 2 from generalized linear mixed effects
models. Meth. Ecol. Evol. 42 133–142. https://doi.org/10.1111/j.2041210x.2012.00261.x
OKONKWO JO and MOTHIBA M (2005) Physico-chemical characteristics and pollution levels of heavy metals in the rivers in
Thohoyandou, South Africa. J. Hydrol. 308 122–127. https://doi.
org/10.1016/j.jhydrol.2004.10.025
PITMAN WV (2011) Overview of water resource assessment in
South Africa: current state and future challenges. Water SA 37 (5)
659–664. https://doi.org/10.4314/wsa.v37i5.3
258
RAMULIFHO PA (2020) Modelling flow and water temperature in
the Luvuvhu catchment and their impact on macroinvertebrate
assemblages. PhD thesis, University of Venda.
RAMULIFHO PA, FOORD SH and RIVERS-MOORE NA (2020)
The role of hydro-environmental factors in Mayfly (Ephemperoptera, Insecta) community structure: identifying threshold
responses. Ecol. Evolut. 10 (14) 6919–6928. https://doi.org/10.1002/
ece3.6333
RASHID I and ROMSHOO SA (2013) Impact of anthropogenic
activities on water quality of Lidder River in Kashmir Himalayas.
Environ. Monit. Assess. 185 4705–4719. https://doi.org/10.1007/s10
661-012-2898-0
REECE PF, REYNOLDSON TB, RICHARDSON JS and
ROSENBERG DM (2001) Implications of seasonal variation
for biomonitoring with predictive models in the Fraser River
catchment, British Columbia. Can. J. Fish. Aquat. Sci. 58 1411–1418.
https://doi.org/10.1139/f01-076
ROWNTREE KM and WADESON RA (1999) A hierarchical
geomorphological model for the classification of selected South
African rivers. WRC Report No. 497/1/99. Water Research
Commission, Pretoria.
SARTORI M and BRITTAIN JE (2015) Order Ephemeroptera. In:
Thorp J and Rodgers DC (eds) Thorp and Covich’s Freshwater
Invertebrates: Ecology and General Biology (4th edn.). Academic
Press, New York. 873–891. https://doi.org/10.1016/B978-0-12-385
026-3.00034-6
SHARMA RC and RAWAT JS (2009) Monitoring of aquatic
macroinvertebrates as bioindicators for assessing the health of
wetlands: a case study in the Central Himalayas, India. Ecol. Indic.
9 118–128. https://doi.org/10.1016/j.ecolind.2008.02.004
SIEGLOCH AE, SURIANO M, SPIES M. and FONSECA-GESSNER
A (2014) Effect of land use on mayfly assemblages’ structure
in Neotropical headwater streams. An. Acad. Bras. Ciê. 86 (4)
1735–1747. https://doi.org/10.1590/0001-3765201420130516
SINGO LR, KUNDU PM, ODIYO JO, MATHIVHA FI and NKUNA TR
(2012) Flood frequency analysis of annual maximum stream flows
for Luvuvhu River Catchment, Limpopo Province, South Africa.
In: 16th SANCIAHS Hydrology Symposium, 1–3 October 2012,
University of Pretoria.
SKOWNO AL, POOLE CJ, RAIMONDO DC, SINK KJ, VAN
DEVENTER H, VAN NIEKERK L, HARRIS LR, SMITH-ADAO
LB, TOLLEY KA, ZENGEYA TA, FODEN WB, MIDGLEY GF
and DRIVER A (2019) National Biodiversity Assessment 2018:
The status of South Africa’s ecosystems and biodiversity. Synthesis
Report. South African National Biodiversity Institute, an entity of
the Department of Environment, Forestry and Fisheries, Pretoria.
214 pp.
Water SA 49(3) 251–259 / Jul 2023
https://doi.org/10.17159/wsa/2023.v49.i3.4023
SVITOK M (2006) Structure and spatial variability of mayfly
(Ephemeroptera) communities in the upper Hron river basin.
Biologia 61 (5) 547–554. https://doi.org/10.2478/s11756-006-0089-6
TAMPO L, LAZAR IM, KABORÉ I, OUEDA A, AKPATAKU
KV, DJANEYE-BOUNDJOU G, BAWA LM, LAZAR G and
GUENDA WA (2020) A multimetric index for assessment of
aquatic ecosystem health based on macroinvertebrates for the
Zio river basin in Togo. Limnologica 83 125783. https://doi.org/
10.1016/j.limno.2020.125783
TER BRAAK CJF and VERDONSCHOT PFM (1995) Canonical
correspondence analysis and related multivariate methods in
aquatic ecology. Aquat. Sci. 37 130–137.
THIRUPATHAIAH M, SAMATHA CH and SAMMAIAH C (2012)
Analysis of water quality using physico-chemical parameters in
lower manair reservoir of Karimnagar district, Andhra Pradesh
Environ. Sci. 3 (1) 172–180. https://doi.org/10.6088/ijes.201203013
1017
VARNOSFADERANY NM, EBRAHIMI E, MIRGHAFFARY N
and SAFYANIAN A (2010) Biological assessment of the Zayandeh
Rud River, Iran, using benthic macroinvertebrates. Limnologica 40
226–232. https://doi.org/10.1016/j.limno.2009.10.002
VENABLES WN, and RIPLEY BD (2002) Random and mixed effects.
In: Modern Applied Statistics with S. Statistics and Computing.
Springer, New York, NY. https://doi.org/10.1007/978-0-387-217062_10
VENKATESHARAJU K, RAVIKUMAR P, SOMASHEKAR RK
and PRAKASH KL (2010) Physico-chemical and bacteriological
investigation on the river Cauvery of Kollegal Stretch in Karnataka.
J. Sci. Eng. Technol. 6 (1) 50–59. https://doi.org/10.3126/kuset.
v6i1.3310
WALLACE JB and WEBSTER JR (1996) The role of macroinvertebrates
in stream ecosystem function. Annu. Rev. Entomol. 41 115–139.
https://doi.org/10.1146/annurev.en.41.010196.000555
WELDEMARIAM MM (2013) Physico-chemical analysis of GudBahri
river water of Wukro, Eastern Tigrai, Ethiopia. Int. J. Sci. Res. Publ.
3 (11) 1–4.
WHO (World Health Organisation) (2011) Guidelines for DrinkingWater Quality (4th edn). World Health Organisation Press, Geneva.
259