ArticlePDF Available

Response of instream animal communities to a short-term extreme event and to longer-term cumulative impacts in a strategic water resource area, South Africa

Authors:

Abstract and Figures

Disturbance plays an integral part in generating heterogeneity required for ecosystem persistence, but the increased amplitude and duration of disturbances linked to drivers of global change could result in ecosystem shifts or collapse. Biomonitoring over time provides insights into trajectories of ecosystem change. The responses of two instream animal taxa to two contrasting disturbance events, a major flood event and the long-term cumulative effects of land-use changes, were assessed in 1999–2012 by quantifying variation and change in abundance of functional groups based on flow rate sensitivity, water quality and metrics of ecological condition. All metrics recovered to pre-flood conditions within seven months after the flood event. Similarly, cumulative impacts of land use effected significant decreases in some but not all metrics. Indices that did not change, including SASS total score and ASPT, were the result of insufficient consideration of the decrease in the abundance of sensitive taxa specifically, and the abundance of all taxa in general. The decrease in abundance of sensitive taxa could signal imminent collapse in certain metrics. Evidence is also provided for a shift in the structure of fish assemblages linked to the decrease and loss of taxa sensitive to ecosystem degradation caused by the longer-term impacts of land-use change.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=taas20
Download by: [Professor S.H. Foord] Date: 13 February 2016, At: 03:06
African Journal of Aquatic Science
ISSN: 1608-5914 (Print) 1727-9364 (Online) Journal homepage: http://www.tandfonline.com/loi/taas20
Response of instream animal communities to a
short-term extreme event and to longer-term
cumulative impacts in a strategic water resource
area, South Africa
SH Foord & PSO Fouché
To cite this article: SH Foord & PSO Fouché (2016): Response of instream animal communities
to a short-term extreme event and to longer-term cumulative impacts in a strategic water
resource area, South Africa, African Journal of Aquatic Science
To link to this article: http://dx.doi.org/10.2989/16085914.2015.1125336
Published online: 12 Feb 2016.
Submit your article to this journal
View related articles
View Crossmark data
African Journal of Aquatic Science 2016: 1–12
Printed in South Africa — All rights reserved
Copyright © NISC (Pty) Ltd
AFRICAN JOURNAL OF
AQUATIC SCIENCE
ISSN 1608-5914 EISSN 1727-9364
http://dx.doi.org/10.2989/16085914.2015.1125336
African Journal of Aquatic Science is co-published by NISC (Pty) Ltd and Taylor & Francis
This is the nal version of the article that is published
ahead of the print and online issue
An understanding of disturbance and its role in structuring
communities is a major focus in the study of ecosystem
dynamics, particularly with regard to its impact on biodiver-
sity and the significance within the context of ecosystem
resilience and functioning (Folke et al. 2004; Hooper et
al. 2005). This understanding has become even more
relevant within the contemporary context of global change.
The long-term monitoring and assessment of assemblage
responses to disturbance provides a framework for the
evaluation of ecosystem resilience and resistance in
the face of a rapidly changing biosphere (Magurran and
Dornelas 2010). Changes in biodiversity over time are
therefore not only an indication of present ecological status,
but could also be predictive of system resilience and resist-
ance in the face of future change.
Freshwater ecosystems, particularly rivers, are one of the
most threatened ecosystems in the world (Ollis et al. 2006).
These ecosystems cover less than 1% of the Earth’s surface
area and, although often resilient in the face of disturbances
(Meffe and Sheldon 1990), the effects of disturbance such
as land-use change, pollution and alien invasion, as well
as climate change, are integrated into these ecosystems.
Rivers are not only representative of changes in the aquatic
systems but act as sentinels of change in terrestrial and
atmospheric processes (Williamson et al. 2008).
The flow of water represents the driving force sculpting
river channel shape and geomorphology, while water
velocity and the flow regime are two of the drivers that
dictate the kinds of animal life in them (Kleynhans 1999;
Franssen et al. 2006; Kleynhans et al. 2007). Natural flow
regimes are essential for the maintenance or restoration of
aquatic communities (Skelton 2001), while floods not only
reshape channels but also import nutrients and maintain a
dynamic equilibrium to which indigenous organisms have
adapted (Meffe and Sheldon 1990). The impact of large-
scale disturbance events can therefore be evaluated only
within the context of long-term baseline variation (Meffe
1984; Franssen et al. 2006).
Floods as a disturbance factor in arid and semi-arid
regions are particularly unpredictable and often extreme
(Franssen et al. 2006). Organisms exposed to such
unpredictable abiotic perturbations may respond over
evolutionary time by developing morphological and
physiological life-history traits that are set to minimise
these impacts (Meffe 1984; Kleynhans 1999; Smithers
et al. 2001). Absence of floods could impede life cycles
(Meffe and Sheldon 1990) in species such as the lowveld
largescale yellowfish Labeobarbus marequensis, where
increased flow acts as a cue for reproduction (Bell-Cross
and Minshull 1988). Floods are therefore important
Response of instream animal communities to a short-term extreme event
and to longer-term cumulative impacts in a strategic water resource area,
South Africa
SH Foord1* and PSO Fouché2
1 Department of Zoology, Chair in Biodiversity Value and Change, Centre for Invasion Biology, University of Venda,
Thohoyandou, South Africa
2 Department of Zoology, University of Venda, Thohoyandou, South Africa
* Corresponding author, e-mail: stefan.foord@univen.ac.za
Disturbance plays an integral part in generating heterogeneity required for ecosystem persistence, but the
increased amplitude and duration of disturbances linked to drivers of global change could result in ecosystem
shifts or collapse. Biomonitoring over time provides insights into trajectories of ecosystem change. The responses
of two instream animal taxa to two contrasting disturbance events, a major flood event and the long-term
cumulative effects of land-use changes, were assessed in 1999–2012 by quantifying variation and change in
abundance of functional groups based on flow rate sensitivity, water quality and metrics of ecological condition. All
metrics recovered to pre-flood conditions within seven months after the flood event. Similarly, cumulative impacts
of land use effected significant decreases in some but not all metrics. Indices that did not change, including SASS
total score and ASPT, were the result of insufficient consideration of the decrease in the abundance of sensitive
taxa specifically, and the abundance of all taxa in general. The decrease in abundance of sensitive taxa could signal
imminent collapse in certain metrics. Evidence is also provided for a shift in the structure of fish assemblages
linked to the decrease and loss of taxa sensitive to ecosystem degradation caused by the longer-term impacts of
land-use change.
Keywords: ASPT, fish, FRAI, Limpopo province, Luvuvhu River, macroinvertebrates, SASS5
Introduction
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
2
disturbance events influencing the eco-evolutionary adapta-
tions of freshwater organisms, an inherent force dictating
life histories and affecting biotic interactions within rivers
(Rowntree et al. 2000; Robinson 2012).
South African rivers are characterised by a high average
coefficient of variation of annual flow, compared to North
American and European rivers (Rossouw et al. 2005).
The riverine biota are adapted to this variability but, in
the case of large floods and episodic events in particular,
community-structuring forces may switch from biotic to
abiotic (Uys and O’Keeffe 1997). Many lotic organisms are
not only adapted to these harsh conditions but have the
ability to recover rapidly from disturbances (Franssen et al.
2006).
The composition of lotic assemblages is further-
more determined by species-specific tolerance levels of
organisms to water quality variables. The effect of these
variables can act either independently or synergistically
(Dallas and Day 2004). Although water quality is of key
importance to humans and nature, man’s influences on
natural ecosystems often go unrecognised until conditions
gradually change to critical levels (Hickey and Salas 1995).
These gradual changes in drivers would precipitate regime
shifts that are often irreversible.
Instream organisms react to all these physical, chemical
and biological parameters, and are therefore useful
sentinels of the rate and magnitude of these changes,
rendering them a viable alternative in monitoring ecological
integrity, as opposed to traditional reductionistic physico-
chemical methods (Kleynhans 2007). The latter methods
have failed to identify several of the non-point source
impacts associated with modern land-use practices (Karr
2001), giving rise to the development of diagnostic indices
based on instream organisms (Chessman and McEvoy
1998).
The Luvuvhu River, in the north-east of South Africa, lies
within a catchment that has experienced an unprecedented
shift in the type, intensity and duration of disturbance
events over the last two decades. After the first democratic
elections in South Africa in 1994, the region has seen both
agricultural intensification and rapid rural development
(Statistics South Africa 2008). The region also experienced
a 100-year-period flooding event in February 2000, when
exceptionally heavy rains were concentrated in two periods
of three days each (Smithers et al. 2001). The flooding
not only caused damage to infrastructure, but also the
magnitude of the discharge and velocity of the water led to
high levels of impact in the Luvuvhu River and its tributaries
and consequent morphological changes to the river
template and the riparian vegetation (Fouché 2000). The
impact of this on ecosystem processes and assemblages is
largely unknown.
Based on the data collected in the upper reaches of the
A91 tertiary catchment, which not only forms part of a South
African strategic water resource area (Nel et al. 2013), but
is also severely impacted, we describe the response of
instream biota assemblages to a once-off major flooding
event and to the subsequent long-term impacts of extensive
land-use change. We evaluate the changes using biotic
indices and indicator taxon abundance, and we test for
shifts in assemblage structure.
Methods
Study area and the selected sites
The Luvuvhu River Catchment, with a surface area of
c. 3 800 km2 (DWAF 2004), forms part of the larger
Limpopo River system (Figure 1). The Soutpansberg
mountain range forms the main geographic feature
along an east–west axis (Figure 1) of the catchment, with
the Luvuvhu River and tributaries draining their south-
eastern slopes (Berger et al. 2003). The eastern parts
of the Soutpansberg mountains have been identified as
a South African strategic water resource area (Nel et al.
2013). Historically, the physico-chemical parameters of
the silt-laden Luvuvhu River were regarded as ‘good’,
with low electrical conductivity and nutrient concentra-
tions (DWAF 1997). Prior to 1910, human impact on the
river was low, but afforestation with exotics commenced
in the upper catchment around the town of Louis Trichardt
in 1911 and, by 1950, extended to the areas north-west of
the town Thohoyandou (DWAF 1990). The Albasini Dam,
completed in 1952, was the first major impoundment to be
built in the main stem of the Luvuvhu River, after which
time the amount of water abstracted for irrigation increased
dramatically (DWAF 1990). Changes in runoff, recorded
for the ninety-year period from 1900 to 1990, show that
runoff decreased by almost 25%, from c. 389 million m3 to
308 million m3 per annum (DWAF 1990). Historically, the
Luvuvhu River was regarded as perennial, but in recent
times its flow ceased for periods, the first recorded flow
cessation being in 1946 (Moore et al. 1991).
The Thulamela and part of the Makhado local munici-
palities (LM), with estimated populations of 580 800 and
495 200, respectively, constitute the catchment of the upper
and middle reaches of Luvuvhu River (Figure 1). Current
land use in the catchment is characterised by forestry in
the headwaters, independent agricultural development in
the valleys and lower slopes in the west, and large rural
and urban settlements in the east (Anon. 2001; Berger et
al. 2003). Historically, the water requirements for agricul-
ture amounted to 73 million m3 y−1 plus 6 million m3 y−1
for afforestation. In recent years, increased invasive alien
vegetation density is estimated to have reduced the yield
further by 8 million m3 y−1 (DWAF 1997). Although 90% of
the population in the two municipal areas is classified as
rural, urbanisation in the major town of Thohoyandou
and its surrounding areas has increased in recent years
(Statistics South Africa 2008). Future changes in land use
in these regions will not only have local effects, but will have
implications for the Kruger National Park, lower down in the
catchment, as the land use in surrounding areas often is
the most important factor determining the biological quality
of running waters in nature reserves further downstream
(Mancini et al. 2005).
Eight sites in the upper and middle reaches and
tributaries of the Luvuvhu River were selected for this
survey. They were located in the eco-region 5.04 (Figure 1)
as defined by the Water Research Commission (Anon.
2001) and fall within the most heavily impacted area of the
Luvuvhu River catchment, with commercial farming and
forestry in the upstream section and urban/rural sprawl in
the lower reaches. Sites were sampled eight times during
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
African Journal of Aquatic Science 2016: 1–12 3
three distinct periods. In the first period, prior to the one-in-
100-year flood in 2000, sites were surveyed in September
1999. The second period charted recovery at these sites,
which were sampled on four consecutive occasions
between June 2000 and June 2011. In the third period, a
decade later, the sites were surveyed in September 2011,
and in April and September 2012. All the macroinverte-
brate samplings and assessments were conducted by
the first author, while fish sampling and assessment were
conducted by the second author.
Physico-chemical aspects
During each survey the pH, dissolved oxygen content
and concentration, electrical conductivity, total dissolved
substance concentration and temperature were determined
in situ at each site using a handheld Eutech Cyberscan
meter. Water samples were collected and their turbidity and
total suspended solid concentration were determined in the
laboratory.
Applied indices
Instream benthic macroinvertebrates were sampled and the
results processed following a standardised protocol as set
out in the South African Scoring System version 5 (SASS5)
(Dickens and Graham 2002) which targets representa-
tive biotopes at a site. Indices of the protocol are based on
the sensitivity ratings of invertebrate taxa to water quality
conditions. Taxon determinations were mainly to family
level. Calculated metrics included the total SASS5 score
(sum of the sensitivity of all taxa), total number of taxa
caught, and the average score per taxon (ASPT).
Fish were sampled at the sites according to the protocol
described by (Kleynhans 1999) and the collected fish
specimens were identified on-site using the key provided
by Skelton (2001), counted and returned to the environ-
ment. The data were used to calculate fish response
assessment index (FRAI) scores. This index is based on
the environmental tolerances and preferences of fish and
their resultant response to changes in environmental drivers
(Kleynhans 2007). The reference state (RS) for each site
was calculated using historical data (Gaigher 1969) and,
in particular, the frequency of occurrence data provided by
(Kleynhans 2007).
Data analysis
The SASS5, ASPT and FRAI scores, as well as invertebrate
family and fish species richness, were tested for normality
using a Wilk–Shapiro test (Legendre and Legendre 1998),
for independence using the Durbin–Watson test (Durbin and
Watson 1971), and for heteroscedasticity using Levene’s
test for homogeneity of variances (Levene 1960). None
of these assumptions were violated for the metrics, and
therefore analysis of variation (ANOVA) was used to test
the significance of the flood- and long-term effects. Tukey’s
AFRICA
South
Africa
SOUTH
AFRICA
30°30'0" E30°0'0" E
23°0'0" S 23°0'0" S
Bej
10 0 10 km5
Louis Trichardt
Elim
Thohoyandou
Ste
Val Cab
Tsh
Has
Cro
Nan
Granite Lowveld
Gravelotte Rocky Bushveld
Makhado Sweet Bushveld
Makuleke Sandy Bushveld
Soutpansberg
Tzaneen Sour Bushveld
Eco-region 5.04
Town
Study site
Luvuvhu
Limpopo
Figure 1: Map of study region and of the distribution of the eight sample sites in the upper reaches of the Luvuvhu River, Limpopo province.
Refer to Appendix 1 for full site names. Vegetation types (Mucina and Rutherford 2006) and Soutpansberg mountains are listed in the key
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
4
honest significant differences (HSD) test was used to
evaluate pairwise differences between surveys at different
times. This simple linear model was also used to analyse
differences between sites.
Although estimates of invertebrate taxon abundance
forms part of the SASS5 protocol, these counts were not
included in the final assessment. In this study, abundance
was incorporated in the analysis by identifying three
sensitivity classes (Dickens and Graham 2002), viz.
(i) ‘tolerant’ taxa, with water quality sensitivity scores
ranging from 1 to 4, (ii) ‘intermediate’ taxa, with scores
ranging between 5 and 8, and (iii) ‘sensitive’ taxa, with
scores ranging between 9 and 15. The SASS protocol
includes an estimation of taxon abundance based on a
logarithmic scale, where 1 is equal to 1 individual observed,
2 to 2–10 individuals, 3 to 11–100, 4 to 101–1 000 and 5 to
numbers in excess of 1 000. These scores were converted
into their log abundance values (1–5), which were then
totalled for all the taxa within each sensitivity class at a site
and for each survey. Similar analysis was done for fish, but
here the actual counts were used, instead of estimates.
Kleynhans et al. (2007) classify fish based on two major
criteria, namely flow and physico-chemical conditions.
The first criterion recognises species that require, and are
tolerant of, high flow conditions and those that are intolerant
of, or sensitive to, high flow. The second criterion identi-
fies three groups, viz. a group that is tolerant, one that is
less tolerant, and one that is sensitive to impaired physico-
chemical param eters. The effects of flooding and long-term
land-use change were analysed using the same linear
model used for indices. All statistical analyses were done in
R (R Development Core Team 2013).
Changes in fish assemblages were mapped through non-
metric multidimensional scaling (MDS) ordination of the
Bray–Curtis similarities between the fish assemblages of the
sites and surveys (Clarke and Warwick 2001). Differences
between the fish assemblages of the eight surveys and
between sites were tested with analysis of similarity
(ANOSIM). Groups identified by this process were then
used as input into SIMPER (similarity percentage), which
allows the identification of species that are responsible for
differences between the a posteriori defined groups. All
assemblage analyses were done in PRIMER 6 (Clarke and
Gorley 2006).
Results
A total of 24 fish species from eight families were collected
during the surveys, with the lowest diversity being observed
at Sterkstroom (Appendix 2). Sixty-six invertebrate taxa
(b)
Sep 1999 Jun 2000 Oct 2000 J an 2001 Jun 2001 S e p 2 011 Apr 2012 Sep 2012
(a) TSS Turbidity (NTU) Conductivity
Downstream
Ste Bej Cab Cro
Tsh Has NanVal
Z-SCORE
2
1
0
−1
2
1.5
1.0
0.5
0.0
−1.0
0.5
SAMPLING DATE
SAMPLE SITE
Figure 2: (a) Spatial trends in turbidity, total suspended solids (TSS) and electrical conductivity, and (b) temporal trends in electrical
conductivity, in the upper reaches of the Luvuvhu River in 1999–2012. Error bars denote SE. Site abbreviations are given in Appendix 1
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
African Journal of Aquatic Science 2016: 1–12 5
were recorded over the period of the study. The flood event
had a negative impact on both the invertebrate and fish
indices, but within one year the scores had recovered to
close to pre-flood conditions (refer text below). During the
third period of sampling, a decade later, the SASS5 scores
had improved to pre-flood conditions, while the FRAI scores
continued to deteriorate.
Physico-chemical variables
Total suspended solids and turbidity values increased
significantly downstream. With the exception of Site 2, a
similar trend was observed in the electrical conductivity
(Figure 2a). Prior to and following the flood, the average
electrical conductivity recorded at the sites remained
relatively stable at c. 100 µS cm−1, but in the second half of
the study it had increased significantly (Figure 2b).
Macroinvertebrate metrics and abundance
Both the SASS5 total scores (F7,57 = 3.9, p = 0.002) and the
total number of taxa (F7,57 = 5, p < 0.001) showed signifi-
cant post-flood decreases, but subsequently recovered
to the levels recorded before the flood (Figure 3a and c).
The average score per taxon (ASPT) recovered within a
year after the flood (Figure 3b), and was then not signifi-
cantly different to scores calculated before the flood (F7,57 =
2.023, p = 0.07). Tolerant taxa increased to levels that were
significantly (F7,57 = 3.4, p = 0.005) higher than before the
flood (Figure 3d), whereas the abundance of intermediate
taxa remained unchanged both after the flood and a decade
later (F7,57 = 2, p = 0.07) (Figure 3e). Taxa sensitive to water
quality decreased significantly after the flood and, 10 years
later, still had not recovered (Figure 3f). Overall, there was
weak support for the significant change in the abundance of
sensitive taxa (F7,57 = 2.1, p = 0.05).
Fish metrics and abundance
Significant changes in fish metrics were observed after the
flood and a decade later (Figure 4). Following the flood, the
fish response assessment index (FRAI) scores recovered
gradually, reaching pre-flood levels by October 2000
(Figure 4a). A decade later the scores had deteriorated to
the levels observed immediately after the flood (F7,57 = 3.6,
p = 0.003) (Figure 4a). A similar pattern was observed in
fish species richness, although it was not significant overall
(F7,57 = 2.1, p = 0.06). As could be expected, species
tolerant of high flows recovered after the flood (F7,57 = 3.11,
p = 0.008), while flow-intolerant taxa never recovered after
the significant reduction caused by the flood (F7,57 = 1.9, p =
0.09) (Figure 5a). Fish tolerant (F7,57 = 1.869, p = 0.09), or
relatively intolerant (F = 0.912, p = 0.5), to deterioration in
river condition showed no real changes in their abundances
(Figure 5b), while intolerant species showed a significant
a
b
ab
ab
ab ab
a
ab
a
b
ab
ab
a
aa
a
a
b
ab
ab
ab
ab
ab
a
b
ab
ab ab
ab
ab
ab
SAMPLING DATE
(a) (b) (c)
(d) (e) (f)
SASS5
ASPT
NUMBER OF TAXA
ABUNDANCE
ABUNDANCE
ABUNDANCE
20
15
10
20
25
15
10
10
12
14
20
25
15
10
2
4
6
8
60
80
100
120
140
160
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
5.0
6.5
Figure 3: Temporal trends in (a) SASS5 total scores, (b) average score per taxon (ASPT) and (c) family richness, and temporal changes
in the abundance of taxa that are (d) tolerant, (e) intermediate and (f) sensitive to water quality in the upper reaches of the Luvuvhu River
in 1999–2012. Arrows indicate the major flood event. Error bars denote SE. Letters above error bars that are in common between surveys
indicate no significant difference
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
6
reduction in their abundances a decade after the flood, in
spite of an initial reduction and subsequent recovery after
the flood (F7,57 = 2.9, p = 0.01) (Figure 5c).
Fish assemblages
Multivariate analysis suggested that fish assemblages at
the headwater sites, Sterkstroom and Beja, were distinct
from the assemblages at the downstream sites (Figure 6a).
There was a significant overall difference in fish assemblages
found by the different surveys (ρ = 0.21, p = 0.01). However,
pairwise comparisons identified three survey groups (Figure
6b). The first group (PreR) included fish assemblages before
the flood, and those in 2001, when the assemblages seemed
to have recovered from the flood. The second group included
fish species found immediately after the flood (PostF),
while the third group represented assemblages observed a
decade later (LongT). Two species, Clarias gariepinus and
Barbus eutaenia, were responsible for 30% of the dissim-
ilarities between the groups (Table 1) and, although all
species decreased in abundance immediately after the flood,
the largest negative impact was on B. eutaenia. Clarias
gariepinus abundance decreased a decade later, but was not
affected by the flood itself (Table 1). Two taxa, Chiloglanis
pretoriae and Labeobarbus marequensis, showed increases
in abundance over the last 10 years (Table 1).
Discussion
Jurajda et al. (2006) differentiated between ‘non-erosive’
and ‘erosive’ floods, where the former is laterally expansive
and the latter consists of fast-moving turbulent water that
moves substrates, often with dramatic effects on physical
channel habitat and the riparian zone. Where erosive
floods can reduce fish populations, non-erosive floods
enable fish to move into inundated areas to feed and breed
(Meffe 1984). The substantial deterioration in the ecological
categories of the downstream sites indicates that the 2000
flooding event can be classified as ‘erosive’.
Fish survive erosive floods by (i) remaining close to
submerged structures such as spaces between boulders,
(ii) seeking low-velocity margins or (iii) migrating to
tributaries or backwater pools (Meffe 1984). Jurajda et al.
(2006) found that fish using shelters were less affected
than open-water species. In this study, the open-water
species B. eutaenia, Barbus unitaeniatus and C.
gariepinus decreased, while the numbers of C. pretoriae
and L. marequensis increased. The latter two species are
not only rheophilic, but also prefer habitat where coarse
alluvial materials such as boulders and cobbles dominate
the substrate. Although no direct cause was found, it
is noteworthy that the abundance of C. gariepinus, a
species rated as ‘tolerant to changes in water quality’
(Kleynhans 2007), has declined substantially over the last
10 years.
b
a
b
ab
ab
a
a
a
ab ab
b
a
High flow tolerant
High flow sensitive
Relatively intolerant
Tolerant
Intolerant
(a)
(b)
(c)
ABUNDANCE
20
40
60
80
100
120
140
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
0
20
40
60
80
0
20
40
60
80
0
SAMPLING DATE
Figure 5: Temporal trends in the abundance of fish, classified on
sensitivity to (a) flow, (b) and (c) water quality. Error bars denote
SE. Letters above error bars that are in common between surveys
indicate no significant difference
a
b
ab
ab
ab
ab ab b
a
b
ab
ab
ab
ab ab
b
(a) (b)
FRAI
SPECIES RICHNESS
50
40
10
12
2
4
6
8
60
70
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
Sep 1999
Jun 2000
Oct 200 0
Jan 2001
Jun 2001
Sep 2011
Apr 2012
Sep 2012
Figure 4: Temporal trends in (a) fish response assessment index
(FRAI) scores and (b) fish species richness in the upper reaches of
the Luvuvhu River in 1999–2012. Arrows indicate the major flood
event. Error bars denote SE. Letters above error bars that are in
common between surveys indicate no significant difference
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
African Journal of Aquatic Science 2016: 1–12 7
In the present study instream biota assemblage structure
and metrics (SASS5, ASPT, species richness and FRAI)
recovered within seven months, as was found by Fritz et
al. (2002), where pre-flood assemblages re-established
within eight months after an erosive one-in-50-year flood,
associated with a recovery in headwater populations in less
than three months. However, it is evident that responses
of river biota in both space and time is context-specific.
For example, Franssen et al. (2006) found no evidence to
suggest that recovery of fish assemblage structure was
associated with time since flooding. Jurajda et al. (2006)
observed that, although changes occurred in the fish
assemblage structure and density after flooding in a lowland
medium-sized river, none of these were significantly
different, while Valdez et al. (2001), who hypoth esised
that a test flood of a magnitude exceeding 1 200 m3 for a
minimum period of seven days would not significantly affect
native fish populations, found little effect based on mean
catch per unit effort (CPUE) and catch rates.
Species richness is typically lower in headwater streams
but increases with stream order and consequent habitat
complexity. The latter aspect provides refuge, and has been
Ste
Bej
Val
Ent
Cro
Tsh
Has
Nan
1
1
1
11
1
1
1
2
2
2
2
22
2
3
3
3
3
3
3
3
4
44
4
4
4
4
4
4
5
5
55
5
5
5
6
6
6
6
6
6
6
7
7
7
77
7
7
8
8
8
8
8
8
(a)
(b)
SAMPLES
80
60
40
20
PreR
PostF
Lo ngT
SIMILARITY
Val
Val
Val
Val
Val
Bej
Bej
Bej
Bej
Bej
Nan
Nan
Nan
Nan
Nan
Nan
Has
Has
Has
Has
Has
Has
Has
Has
Ste
Ste
Ste
Ste
Ste
Ste
Ste
Ste
Tsh
Tsh
Tsh
Tsh
Tsh
Nan
Nan
Tsh
Tsh
Tsh
Cro
Cro
Cro
Cro
Cro
Cro
Cro
Cro
Ent
Ent
Ent
Ent
Ent
Ent
Ent
Ent
Figure 6: (a) MDS ordination and (b) cluster analysis of fish assemblage structure of fish collected at sample sites in the Luvuvhu River
in eight surveys during the period September 1999 to September 2012. PreR = assemblages before the flood and after recovery in 2001
following the flood; PostF = assemblages directly after the flood; LongT = long-term responses in assemblages. Site abbreviations given in
Appendix 1
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
8
shown to be an important predictor of fish diversity (Fritz
et al. 2002; Franssen et al. 2006). The effects of flooding
should, therefore, have opposite effects in headwater and
downstream reaches (Franssen et al. 2006). In one of the
first studies on the response to flooding and the relation-
ships between these changes and fish community stability,
Harrell (1978) found that, although flash flooding in desert
streams changed the mean number of specimens as well
as the mean species diversity and evenness, only the
decline in species diversity was statistically significantly
different. This author found no correlation between distance
from headwater and species diversity prior to floods,
although distance from the source and diversity were
positively correlated after the flood. In the present study,
flooding negatively affected diversity and abundance, and
this effect was correlated with distance from the headwater,
with the largest impacts observed at the downstream sites.
Fish species sensitive to and intolerant of deteriora-
tion in river conditions (Kleynhans 2007) showed a signif-
icant reduction, while tolerant and relatively intolerant
species showed no changes (Figure 5b). The fish species
that contributed most to the difference between the period
immediately after the flood (PostF) and that after the initial
recovery (PreR) (Table 1 and Figure 6) are, in descending
order of importance, C. gariepinus, L. marequensis, C.
pretoriae and B. trimaculatus. Of these four species, only
C. pretoriae is regarded by Kleynhans (2007) as intolerant
to changes in physico-chemical parameters, while the other
three species are regarded as tolerant. It should, however,
be noted that five other species, namely B. eutaenia and
Barbus lineomaculatus, both intolerant to water quality
changes, as well as the moderately intolerant species
Micralestes acutidens, Mesobola brevianalis and Barbus
neefi, the abundances of which were reduced by the floods,
were not only less abundant in the last two years of the
survey but also no specimens of B. eutaenia or B. neefi
were collected in the September 2012 survey.
Over the long term, SASS5 scores have recovered to
their pre-flood values (Figure 3a), driven largely by the
increased number of taxa (Figure 3c). There has been a
decrease in ASPT scores over time (Figure 3b) and,
although this trend is not significant, there is substantial
support for a decline in taxa sensitive to water quality and
a significant increase in the abundance of tolerant taxa
(Figure 3d and f).
All metrics and tolerance groups, water quality for
macroinvertebrates, flow and water quality for fish,
recovered after the flood, except for the flow-intolerant
species. The system was therefore remarkably resilient
in the face of such a major flood event. However,
a decade later, fish and invertebrate richness have
changed in contrasting ways. Fish species richness is
on the decrease, and there is also a functional shift, with
flow-intolerant and water quality-sensitive species being
replaced by flow-tolerant taxa and species associated with
impaired water quality. The invertebrate taxon richness
has increased, but sensitive taxa have seen significant
decreases. The latter observation is particularly relevant
to the monitoring of macroinvertebrates. The significant
decrease in the abundance, and therefore the predicted
loss of sensitive taxa, points to a possible collapse in both
SASS5 and ASPT metrics, an observation that would not
be possible if the two metrics were evaluated in isolation,
as is practised in SASS5 assessments. Within the relatively
stable, albeit substantially lower, trend observed for the
SASS5 scores 10 years later, the impact of a decline in
sensitive taxa has been masked by the recent increase in
tolerant and generalist taxa. The resilience of the system
to future disturbances in terms of macroinvertebrate
assemblages and, by extension, of its condition, is therefore
questioned.
Individual responses of the fish species also point to the
subtleties of interpretation. This is evident in the signifi-
cant decrease of C. gariepinus, a species thought to be
unaffected by deteriorating water quality but, because of its
sedentary lifestyle and longevity, considered to be a bioindi-
cator of persistent organic pollutants (POPs) (Brink 2010).
The concentrations of total DDT in this river system are now
the highest measured over the past century (Bouwman et
al. 2013).
Table 1: SIMPER results summarising fish species collected at sites in the Luvuvhu River responsible for differences between the groups
identified with ANOSIM. PreR = pre-flood and post-flood recovered, PostF = immediatle post-flood, LongT = long-term response
Species Average abundance Dissimilarity SD Contribution (%)
PreR PostF
Clarias gariepinus 4.18 3.72 1.48 18.59
Barbus eutaenia 2.76 0.2 1.29 11.72
Chiloglanis pretoriae 1.32 1.19 1.42 7.06
Barbus trimaculatus 1.1 1.02 1.02 6.4
PreR LongT
Clarias gariepinus 4.18 0.23 1.11 15.88
Barbus eutaenia 2.76 0.22 1.28 9.89
Labeobarbus marequensis 0.69 2.2 1.09 8.94
Chiloglanis pretoriae 1.32 1.61 1.18 5.31
Barbus unitaeniatus 1.14 0.24 1.1 4.21
PostF LongT
Clarias gariepinus 3.72 0.23 1.47 18.99
Labeobarbus marequensis 0.25 2.2 1.11 10.66
Chiloglanis pretoriae 1.19 1.61 1.22 8.03
Barbus trimaculatus 1.02 0.73 1.12 5.47
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
African Journal of Aquatic Science 2016: 1–12 9
Unfortunately, management responses to any of these
results are often premised on the level of understanding of
a system and on the capacity to take action. We are moving
towards increased understanding, although the ability to
take action is often limited, and we should therefore operate
within the context of watchful tolerance (Scholes 2011).
Acknowledgements — Zikwe Modiba, University of Venda, is
thanked for his assistance in the field and Mike Angliss, Anglo
American Platinum, for the contribution of data and continued
advice. SHF acknowledges financial support from the NRF-DST
South African Research Chair in Biodiversity Value and Change
as well as the DST-NRF Centre of Excellence for Invasion Biology.
References
Anon. 2001. State of the Rivers Report: Letaba and Luvuvhu river
systems. WRC Report No. TT 165/01. Pretoria: Water Research
Commission.
Bell-Cross G, Minshull JL. 1988. The fishes of Zimbabwe.
Bulawayo: National Museums and Monuments of Zimbabwe.
Berger K, Crafford JE, Gaigher I, Gaigher MJ, Hahn N, Macdonald
I. 2003. A first synthesis of the environmental, biological and
cultural assets of the Soutpansberg. Louis Trichardt: Leach
Printers.
Bouwman H, Viljoen IM, Quinn LP, Polder A. 2013. Halogenated
pollutants in terrestrial and aquatic bird eggs: converging
patterns of pollutant profiles, and impacts and risks from high
levels. Environmental Research 126: 240–253.
Brink KA. 2010. Effects of DDT on aquatic organisms in the
Luvuvhu River. PhD thesis, University of Johannesburg, South
Africa.
Chessman BC, McEvoy PK. 1998. Towards diagnostic biotic indices
for river macroinvertebrates. Hydrobiologica 364: 169–182.
Clarke KR, Gorley RN. 2006. PRIMER v6: user manual/tutorial.
Plymouth: PRIMER-E.
Clarke KR, Warwick RM. 2001. Change in marine communities:
an approach to statistical analysis and interpretation. Plymouth:
PRIMER-E.
Dallas HF, Day JA. 2004. The effect of water quality variables on
aquatic ecosystems: a review. WRC Report No. TT 224/04.
Pretoria: Water Research Commission.
Dickens CWS, Graham PM. 2002. The South African Scoring
System (SASS) version 5 rapid bioassessment system for rivers.
African Journal of Aquatic Science 27: 1–10.
Durbin J, Watson GS. 1971. Testing for serial correlation in Least
Squares Regression II. Biometrica 37: 409–428.
DWAF (Department of Water Affairs and Forestry). 1990. Kruger
National Park Rivers Research Programme: Water for Nature.
Hydrology of the Luvuvhu River. Department of Water Affairs and
Forestry, Pretoria.
DWAF (Department of Water Affairs and Forestry). 1997. Report
on the Luvuvhu River IFR workshop, July 1994, Annexure 3:
biophysical aspects: ROIP and IFR reports. Department of Water
Affairs and Forestry, Pretoria.
DWAF (Department of Water Affairs and Forestry). 2004. Luvuvhu/
Letaba Water Management Area: internal strategic perspective.
Department of Water Affairs and Forestry, Pretoria.
Folke C, Carpenter S, Walker B, Scheffer M, Elmqvist T,
Gunderson L, Holling CS. 2004. Regime shifts, resilience, and
biodiversity in ecosystem management. Annual Review of
Ecology and Systematics 35: 557–581.
Fouché PSO. 2000. Riparian vegetation: an assessment of the
Luvuvhu River and tributaries. Technical report prepared for the
Department of Environmental Affairs, Limpopo Province, South
Africa.
Franssen NR, Gido KB, Guy CS, Tripe JA, Shrank SJ, Strakosh
TR et al. 2006. Effects of floods on fish assemblages in an
intermittent prairie stream. Freshwater Biology 51: 2072–2086.
Fritz KM, Tripe JA, Guy CS. 2002. Recovery of three fish species
to flood and seasonal drying in a tallgrass prairie stream.
Transactions of the Kansas Academy of Science. 105: 209–218.
Gaigher IG. 1969. Aspekte met betrekking tot die ekologie,
geografie en taksonomie van varswatervisse in die Limpopo- en
Incomatiriviersisteem. PhD thesis, Rand Afrikaans University,
South Africa.
Harrell HL. 1978. Response of the Devil’s River (Texas) fish
community to flooding. Copeia 1: 60–68.
Hickey JT, Salas JD. 1995. Environmental effects of extreme
floods. Paper presented at US–Italy Research Workshop on the
hydrometeorology, impacts and management of extreme floods.
Perugia, Italy, 13–17 November 1995. Available at http://www.
engr.colostate.edu/ce/facultystaff/salas/us-italy/papers/33hickey.
pdf [accessed 29 January 2016].
Hooper DU, Chapin FS III, Ewel JJ, Hector A, Inchausti P, Lavorel
S et al. 2005. Effects of biodiversity on ecosystem functioning:
a consensus of current knowledge. Ecological Monographs 75:
3–35.
Jurajda P, Reichard M, Smith C. 2006. Immediate impact of
an extensive summer flood on the adult fish assemblage of a
channelised lowland river. Journal of Freshwater Ecology 3:
493–501.
Karr JR. 2001. Biological integrity: a long-neglected aspect of water
resource management. Ecological Applications 1: 66–84.
Kleynhans CJ. 1999. The development of a fish index to assess the
biotic integrity of South African rivers. Water SA 3: 265–278.
Kleynhans CJ. 2007. Module D: Fish Response Assessment Index
in river ecoclassification: manual for ecostatus determination
(Version 2). WRC Report No. TT 330/08. Pretoria: Joint Water
Research Commission and Department of Water Affairs and
Forestry report.
Kleynhans CJ, Louw MD, Moolman J. 2007. Reference frequency
of occurrence of fish species in South Africa. WRC Report No.
TT 331/08. Pretoria: Joint Water Research Commission and
Department of Water Affairs and Forestry report.
Legendre P, Legendre L. 1998. Numerical ecology. Amsterdam:
Elsevier Science.
Levene H. 1960. Robust tests for equality of variances. In: Olkin I,
Alto P (eds), Contributions to probability and statistics. Stanford:
Stanford University Press. pp 278–292.
Magurran AE, Dornelas M. 2010. Biological diversity in a changing
world. Philosophical Transactions of the Royal Society B 365:
3593–3597.
Mancini L, Formichetti P, Anselmo A, Tancioni L, Marchini S,
Sorace A. 2005. Biological quality of running waters in protected
areas: the influence of size and land use. Biodiversity and
Conservation 14: 351–364.
Meffe GK. 1984. Effects of abiotic disturbance on coexistense of
predator-prey fish species. Ecology 68: 1525–1534.
Meffe GK, Sheldon AL. 1990. Post-defaunation recovery of fish
assemblages in southeastern blackwater stream. Ecology 71:
657–667.
Moore CA, van Veelen M, Ashton PJ, Walmsley RD. 1991.
Preliminary water quality guidelines for the Kruger National Park
rivers. SANParks, South Africa.
Mucina L, Rutherford MC (eds). 2006. The vegetation of South
Africa, Lesotho and Swaziland. Strelitzia 19. Pretoria: South
African National Biodiversity Institute.
Nel J, Colvin C, Le Maitre D, Smith J, Haines I. 2013. South
Africa’s strategic water source areas. Report No: CSIR/NRE/
ECOS/ER/2013/0031/A. Council for Scientific and Industrial
Research, Pretoria.
Ollis DJ, Dallas HF, Esler KJ, Boucher C. 2006. Bioassessment
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
10
of the ecological integrity of river ecosystems using aquatic
macroinvertebrates: an overview with a focus on South Africa.
African Journal of Aquatic Science 31: 205–227.
R Development Core Team. 2014. R: a language and environment
for statistical computing. Foundation for Statistical Computing,
Vienna, Austria. Available at: http://www.R-project.org.
Robinson CT. 2012. Long-term changes in community assembly,
resistance, and resilience following experimental floods.
Ecological Applications 22: 1949–1961.
Rossouw LM, Avenant MF, Seaman MT, King JM, Barker CH,
du Preez PJ et al. 2005. Environmental water requirements in
non-perennial systems. WRC Report No. 1414/1/05. Pretoria:
Water Research Commission.
Rowntree KM, Wadeson RA, O’Keeffe J. 2000. The development
of a geomorphological classification system for the longitudinal
zonation of South African rivers. South African Geographical
Journal 82: 163–172.
Scholes RJ. 2011. A framework for deriving and triggering
thresholds for management intervention in uncertain, varying and
time-lagged systems. Koedoe 53: Art. #987. 8 pp.
Skelton P. 2001. A complete guide to the freshwater fishes of
southern Africa. Pretoria: Struik.
Smithers JC, Schulze RE, Pike A, Jewitt GP W. 2001. A
hydrological perspective of the February 2000 floods: a case
study in the Sabie River catchment. Water SA 27: 325–332.
Statistics South Africa. 2008. Community survey 2007: basic
results – municipalities. Statistical release P0301.1. Pretoria:
Department of Statistics.
Uys MC, O’Keeffe JH. 1997. Simple words and fuzzy zones:
early directions for temporary river research in South Africa.
Environmental Management 21: 517–531.
Valdez RA, Hoffnagle TL, McIvor CC, McKinney T, Liebfried WC.
2001. Effects of a test flood on fishes of the Colorado River in
Grand Canyon, Arizona. Ecology 71: 657–667.
Williamson CE, Dodds W, Kratz TK, Palmer MA. 2008. Lakes
and streams as sentinels of environmental change in terrestrial
and atmospheric processes. Frontiers in Ecology and the
Environment 6: 247–254.
Manuscript received 3 September 2015, revised 17 November 2015, accepted 24 November 2015
Associate Editor: C Thirion
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
African Journal of Aquatic Science 2016: 1–12 11
Site name Code River Altitude
(m asl) Latitude Longitude Geomorphological zone
(after Rowntree et al. 2000)
Sterkstroom Ste Sterkstroom 870 23.0511° S 30.0880° E B – Mountain stream
Beja Bej Luvuvhu 765 23.0990° S 30.0705° E D – Upper foothills
Valdezia Val Luvuvhu 709 23.0830° S 30.1720° E D – Upper foothills
Cabbage farm Cab Lutanandwa 637 23.0750° S 30.3233° E E – Lower foothills
Tshino Tsh Luvuvhu 584 23.1152° S 30.3899° E E – Lower foothills
Crocodile ventures Cro Dzindi 574 23.0102° S 30.4983° E D – Upper foothills
Hasani Has Luvuvhu 546 23.0843° S 30.4731° E E – Lower foothills
Nandoni Nan Luvuvhu 490 23.0950° S 30.6031° E E – Lower foothills
Appendix 1: Location and classification of sites in the Luvuvhu River and tributaries surveyed in September 1999 and September 2012
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
Foord and Fouché
12
Family Species Sterkstroom Beja Valdezia Cabbage
Farm Tshino Crocodile
Ventures Hasani Nandoni
RS PS RS PS RS PS RS PS RS PS RS PS RS PS RS PS
Characidae Micralestes acutidens X X X X X X X X X X X X
Cyprinidae Labeobarbus marequensis
Barbus annectens
Barbus eutaenia
Barbus lineomaculatus
Barbus neefi
Barbus paludinosus
Barbus toppini
Barbus trimaculatus
Barbus unitaeniatus
Barbus viviparus
Labeo cylindricus
Labeo molybdinus
Mesobola brevianalis
Opsaridium peringueyi
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Clariidae Clarias gariepinus X X X X X X X X X X X X X
Amphiliidae Amphilius uranoscopus X X X X X X X X X X X X X X X X
Mochokidae Chiloglanis swierstrae
Chiloglanis paratus
Chiloglanis pretoriae
X
X
X X
X X
X
X X
X
X X
X
X
X
X
X
X
X
X
X
Cichlidae Oreochromis mossambicus
Coptodon rendalli
Tilapia sparrmanii
Pseudocrenilabrus philander
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
Anguillidae Anguilla mossambica
Anguilla bengalensis labiata X
X
X
X
X
X
X
X
X
X X
X
X X
X
X
X
X
Mormyridae Marcusenius macrolepidotus X X X X X X X X
Gobiidae Glossogobius giuris X
Schilbeidae Schilbe intermedius X X X X
Appendix 2: Historic and observed fish biodiversity at sites in the Luvuvhu River and tributaries in 1999–2012. RS = reference conditions, including 1999 survey results; PS = results of
post-1999 surveys
Downloaded by [Professor S.H. Foord] at 03:06 13 February 2016
... area of approximately 5 940 km 2 , of which 24% is protected in the Kruger National Park (KNP) downstream of the upper catchment Heath and Classen, 1999). The main river channel is the Luvuvhu River, and together with its tributaries (Mutshindudi, Lutanandwa, Dzindi and Mutale Rivers), begin their downstream journey from the highlands of the Soutpansberg Mountain range flowing in an eastern direction through a diverse landscape with various aquatic habitat types before joining the transboundary Limpopo River (Foord and Fouché, 2016;DWAF, 2012;Angliss et al., 2001;Kleynhans, 1996;Heath and Classen, 1999). The Luvuvhu River is the main source of water for the greatest diversity and richness of species within the Makuleke wetland of international importance (Ramsar site) (Lalley, 2014). ...
... Climate in the catchment ranges from humid and wet areas in the mountains (west of the catchment) with a Mean Annual Precipitation (MAP) of 1 800 mm, to dry and hotter areas in the eastern lowlands of the catchment where MAP is 400 mm per annum (DWAF, 2012). Changes in runoff have been recorded for a ninety-year period from 1900 to 1990, with periodic cessation in flow in recent times for some streams that were historically perennial (Foord and Fouché, 2016). ...
... Increasing variability of flow over time will lead to increased extremes in daily flow and sub-daily flow metrics. Foord and Fouché (2016) of daily and sub-daily data show that the use of daily flows on their own did not provide support for the assertion that temporal partitioning still exists within flow patterns to allow for the persistence of both fish flow guilds (Dallas and Rivers-Moore, 2018;Bogan and Lytle, 2007). The implication of this is that the use of sub-daily flows in ecological assessments becomes important for assessing aquatic ecological resilience. ...
Article
As the impacts of the anthropocene intensifies, there is an increasing need to understand how these changes affect both daily and sub‐daily stream flow variability, timing and flow quantities, as these are some of the most influential drivers of spatial and temporal dynamics of stream biota. In this paper, long‐term changes in flow patterns of a strategic water source area (Luvuvhu Catchment) in an arid region of southern Africa were quantified, focusing on the relation between daily and sub‐daily flow and its potential impact on fish biota of the catchment. Long‐term temporal trends in stream flow were modelled using Generalized Least Squares (GLS), while sub‐daily and daily mean flow of the same stations were compared using a suite of metrics. Periods of similar stream flow patterns were identified using K‐means cluster analysis. A spreadsheet rule‐based model was developed linking fish communities to streamflow patterns providing a predictive framework for fish assemblage responses to stream flow classes. Long term reduction in flow in the Luvuvhu Catchment has a strong seasonal component, with significant decreases during the wet season, not linked to long‐term rainfall patterns. The flow regime of the Luvuvhu river system has become more variable over time. Several sub‐daily flow metrics were positively related to daily flow metrics. Oscillating flow conditions and the loss of intermediate flow states may permanently exclude certain fish flow guilds. However, temporal partitioning is only evident when sub‐daily metrics are considered, highlighting their importance for assessing ecological resilience.
... Other studies in the LRC recognize the need to limit both water abstractions and changes to the river flow regimes by setting environmental standards that guard and maintain river ecosystem characteristics and functions [22,23]. The Luvuvhu River catchment is located in the Luvuvhu-Letaba Water Management Area, with a catchment size of approximately 5941 km 2 ( Figure 1), with 14 quaternary catchments, with the main river channels being the Luvuvhu, Mutale, Mutshindudi and Dzindi Rivers [23,24]. The Luvuvhu River catchment is important to forestry, macadamia, avocado and banana plantations in the headwaters, agricultural development in the valleys and lower slopes (western side) and rural and urban settlements (eastern side) [25], while the rivers and the riparian vegetation of the Luvuvhu catchment are home to many bird and animal species [26]. ...
... The mountainous areas of the LRC receive the highest mean annual precipitation of 1800 mm as compared to 400 mm in the lowland areas. However, the mean annual precipitation for the entire catchment varies between 610 to 800 mm [24,27], with a mean annual runoff of 519 million m 3 [27]. The variability of streamflow has increased within the Luvuvhu River Catchment over an 86-year period, indicating that factors such as anthropogenic activities and reservoir development could be impacting on the river streamflow [22,29]. ...
... The mountainous areas of the LRC receive the highest mean annual precipitation of 1800 mm as compared to 400 mm in the lowland areas. However, the mean annual precipitation for the entire catchment varies between 610 to 800 mm [24,27], with a mean annual runoff of 519 million m 3 [27]. The variability of streamflow has increased within the Luvuvhu River Catchment over an 86-year period, indicating that factors such as anthropogenic activities and reservoir development could be impacting on the river streamflow [22,30]. ...
Article
Full-text available
Rivers are now facing increasing pressure and demand to provide water directly for drinking, farming and supporting industries as a result of rapidly growing global human population. Globally, the most common practice for catchment managers is to limit water abstraction and changes to stream flow by setting environmental flow standards that guard and maintain the natural ecosystem characteristics. Since the development of the environmental flow concept and methods in South Africa, very few studies have assessed the institutional constraints towards environmental flow implementation. This study determined stream flow trends over time by fitting simple linear regression model to mean daily stream flow data at three selected stations in the Luvuvhu River Catchment (LRC). We also conducted a literature search to review, firstly the response of aquatic organisms (fish and macroinvertebrate) to changes in habitat conditions and secondly on local challenges affecting the sustainable implementation of environmental flow regime and related water resources management strategies. All the three stream flow stations show decreasing stream flow volume of 1 and 2 orders of magnitude faster in some stations with the possibility that flow will cease in the near future. Qualitative analyses from both local and international literature search found that the main challenges facing the implementation of sustainable flow strategies and management are absence of catchment management agency, lack of understanding of environmental flow benefits, limited financial budget, lack of capacity and conflict of interest. Rivers with changing stream flows tend to lose sensitive species. The development of scientifically credible catchment-wide environmental flow and abstraction thresholds for rivers within the LRC would make a major contribution in minimizing the declining stream flow volumes. Monitoring and reporting should be prioritized to give regular accounts of the state of our rivers.
... The literature on the relationships between hydro-environmental variables and assemblage structure of aquatic organisms is limited for Afrotropical streams, particularly for the northern regions of South Africa, where it is largely restricted to rapid biological assessments (Foord & Fouché, 2016). Here, we explore the relationship between stream flow, water temperature, and other important covariates that drive mayfly assemblage structure in the Luvuvhu catchment. ...
... Rivers of this catchment have shown a substantial decrease (>53%) in stream flow volume over the last 80 years (Odiyo, Makungo, & Nkuna, 2015). Kleynhans (1996) classified streams in the Luvuvhu River as fairly natural, but recent agricultural intensification and the expansion of human settlements have had substantial impacts on instream biota (Foord & Fouché, 2016), and the flow regime has consequently been altered considerably (Ramulifho, Ndou, Thifhulufhelwi, & Dalu, 2019). ...
Article
Full-text available
Freshwater organisms are threatened by changes in stream flow and water temperature regimes due to global climate change and anthropogenic activities. Threats include the disappearance of narrow‐tolerance species and loss of favorable thermal conditions for cold‐adapted organisms. Mayflies are an abundant and diverse indicator of river health that performs important functional roles. The relative importance of key hydro‐environmental factors such as water temperature and flow volumes in structuring these communities has rarely been explored in the tropical regions of Africa. Here, we investigate the response of mayfly species diversity to these factors in the Luvuvhu catchment, a strategic water source area in the arid northeastern region of South Africa. Mayfly larvae were sampled monthly in stones‐in‐current biotopes across 23 sites over a one‐year period. The relationship between these environmental drivers and mayfly diversity was modeled using linear mixed effects models (LMMs) and a model‐based multivariate approach. Threshold Indicator Taxa Analysis (TITAN) was used to model the response of mayfly species to important gradients and identify thresholds of change. Site‐specific characteristic were the most important predictor of mayfly diversity, and there was considerable variation over time, with mayfly diversity peaking during winter. Along this, gradient temperature was the best predictor of assemblage structure, with five out of six reliable indicator species being cold‐adapted, and a community threshold response at 19°C. Results support laboratory‐based thresholds of temperature for mayfly species survival and development, extending empirical evidence to include field‐based observations. Increased global (climate change) and local (riparian vegetation removal, impoundments) changes are predicted to have negative impacts on mayfly diversity and ultimately on ecosystem function.
Article
Full-text available
Ecosystems are characterised by complexity: high connectivity, the presence of positive and negative feedback loops, non-linear, abrupt and sometimes irreversible changes, delays between cause and effects, and uncertainties in observations, understanding and prediction. ‘Adaptive management’ is the preferred approach for the rational management of such systems. Where the management objective is to allow natural feedbacks and adaptive processes to operate as much as possible – as it is in many areas set aside for biodiversity conservation – a key issue is defining the thresholds that will trigger management intervention. This paper outlines and illustrates a logical process for doing so, taking into account the characteristics of complex, continuously changing ecosystems and the reality of information that is partial and understanding that is always provisional. After identifying a key ecological process that is believed to have an element of irreversibility beyond a certain point, the process has several steps, (1) define an indicator of the system state, (2) set a limit of acceptable change and add a safety margin, (3) project the indicator forward using a model, including uncertainty, (4) note the time when the indicator might transgress the safety-buffered limit and (5) subtract ecosystem and management response times. If the resultant time is at hand, an action is indicated – if not, the action is to continue to monitor the situation and refine the observations and models.Conservation implications: Ecosystems are characterized by abrupt and sometimes irreversible changes. The challenge that face conservationists and managers are to identify which of these changes are likely to be irreversible and at what levels this will occur. This paper describes a logical process that enable mangers to determine which ecological processes have levels of irreversibility and monitor their status at all times. Once these processes are nearing the levels that are undesirable management actions can be invoked to prevent this from happening.
Article
Full-text available
A biological integrity index that uses attributes of fish assemblages was developed and applied to a part of the Crocodile River (Incomati System, Mpumalanga Province). Fundamentally, the fish assemblage integrity index (FAII) is based on the fish species expected to be present in biological (fish habitat) segments which are sections of river with relatively homogeneous fish habitat. Within this framework fish are categorised according to an intolerance index which takes into account trophic preferences and specialisation, habitat preferences and specialisation, requirement for flowing water during different life-stages and association with habitats with unmodified water quality. This intolerance index, the expected frequency of occurrence and expected health of fish species in a particular fish habitat segment was used to formulate an index for the situation expected under minimally impaired conditions which was used as the comparative basis for the observed (sampled) situation. The observed situation was expressed as a fraction of the expected situation to arrive at a relative FAII index value which was grouped into one of six descriptive fish assemblage integrity index classes. It was found that the index reflected several aspects of the modifications that have occurred in the Crocodile River, i.e. water quality modifications, flow modifications and introduction of alien fish. However, flow releases from Kwena Dam hampered fish sampling while the floods that occurred in the summer prior to the survey, extensively modified physical habitat conditions that existed for some years. It was concluded that the index provides a broad, synoptic estimation of the biological integrity of the river. It is highly likely that the index in its current form provides an underestimation of the biological integrity due to the species expected to be present, being based on all species listed for a segment, and not on the fish species expected to be present in the habitats actually sampled. It is suggested that refinement of the index should include development of improved methodology to determine not only the fish species expected per segment, but also species expected per habitat type. In its current form, the index has the potential to provide qualitative, descriptive criteria for the desired ecological condition or integrity of rivers for management purposes in terms of the new South African Water Law. The development of numerical criteria will, however, require development in terms of the statistical properties (i.e. statistical power) of the index.
Article
Full-text available
We investigated the presence, levels, relationships, and risks of HCHs, DDTs, chlordanes, mirex, PCBs, and brominated flame retardants (BFRs) in terrestrial and aquatic bird eggs from an area in South Africa where DDT is used for malaria control. We found one of the highest ΣDDT levels reported this century; 13 000 ng/g wm (wet mass) in Grey Heron eggs which exceeds critical levels for reproductive success (3000 ng/g wm) calculated for Brown Pelicans, with a no-effect level estimated at 500 ng/g wm. Even higher ΣDDT levels at 16 000 ng/g wm were found in House Sparrow eggs (possibly the highest ever recorded for sparrows), with a maximum of 24 400 ng/g wm. Significant eggshell thinning in Cattle Egrets (33% between thickest and thinnest) was associated with increased levels of p,p′-DDT and p,p′-DDE. There were indications of unknown use of DDT and lindane. Relative to DDT, PCBs and BFRs levels were quite low. Ordinated data showed that different terrestrial pollutant profiles converged to a homogenised aquatic profile. Converging profiles, high levels of DDT in heron and sparrow eggs, and thinning eggs shells, indicate risk and impacts at release, in the aquatic environment, and in between. If characteristic life-strategies of birds in warm areas (e.g. longer-lived and fewer eggs per clutch) increases the risk compared with similar birds living in colder regions when both experience the same environmental pollutant levels, then malaria control using DDT probably has more significant impacts on biota than previously realised. Therefore, risk assessment and modelling without hard data may miss crucial impacts and risks, as the chemical use patterns and ecologies in Africa and elsewhere may differ from the conditions and assumptions of existing risk assessment and modelling parameters. Considera- tion of other findings associated with DDT from the same area (intersex in fish and urogental birth defects in baby boys), together with the findings of this study (high levels of DDT in bird eggs, eggshell thinning in the Cattle Egrets, and the apparent absence of breeding piscivore birds in the sprayed area) are strongly suggestive of negative impacts from DDT spraying for Malaria control. Our data presents strong arguments for an expedited process of replacing DDT with sustainable methods.
Article
The construction of biotic indices that use macroinvertebrates to assess pollution and other anthropogenic disturbances of rivers and streams often requires that each taxon be assigned a number indicating its level of sensitivity. A problem in constructing such indices is that individual taxa may vary quite widely in sensivity, depending on the nature of the particular disturbance. One possible means of overcoming this problem is to construct a suite of indices, each assembled using sensitivity numbers targeted to a particular impact. In order to test this idea, we sampled macroinvertebrates from rivers in south-eastern Australia subjected to three different types of anthropogenic disturbance: operation of large darns, discharge of effluent from municipal sewage treatment plants, and contamination by metals originating from historical mining. Using macroinvertebrate data from sampling sites with varying levels of exposure to disturbance, we developed sensitivity numbers for macroinvertebrate families for individual rivers and combinations of rivers with the same disturbance type. Sensitivity numbers calculated for individual families differed significantly according to disturbance type in several cases. Gastropod molluscs (family Thiaridae) were tolerant of dam effects but sensitive to sewage and metals, whereas coenagrionid damselfly nymphs, elmid beetles and ostracods were most tolerant of sewage. Corydalid alderfly larvae, leptophlebiid mayfly nymphs, lestid damselfly nymphs, libellulid dragonfly nymphs and scirtid beetle larvae were most tolerant of metals. Indices constructed using sensitivity numbers for a particular disturbance were generally most responsive to that disturbance, but there was considerable generality in responses as well as substantial variability between different rivers with the same disturbance type. In particular, macroinvertebrate communities at sites downstream of dams responded quite variably, probably because of substantial differences in release regimes. We conclude that the approach has merit but requires considerable further development and testing, as well as consideration of the levels of specificity and diagnostic strength that are appropriate or achievable.
Article
Flooding on the Devil's River, Texas, caused changes in both physical and biological structure of the river. Habitat types became less distinct and generally more riffle-like. Fish species diversity, evenness and number of species decreased. Higher diversities were expressed nearer the river's terminus, indicating flood-resistance there. Prior to the flood 18 species associations existed; 13 existed after flooding, only three of which were present prior to flooding. Six species dominated both pre- and post-flood community biomass and maintained the same rank-order dominance in both flood regimes. These (Dionda episcopa, Notropis proserpinus, N. amabilis, Astyanax mexicanus, Notropis venustus and Dionda diaboli) occupied slightly different habitats or utilized different resources of the same habitat in the post-flood regime. All are adapted to a flood-prone environment and may constitute the greatest degree of faunal stability in the ecosystem.