Technical ReportPDF Available

Freshwater Fishes of the Cape Fold Ecoregion and Climate Change: Volume 1: Synthesis of Research Findings

Authors:
Dallas, HF; Shelton, JM; Paxton, BR
and Weyl, OLF
October 2017
Freshwater fishes of the Cape Fold
Ecoregion and Climate Change
Volume 1: Synthesis of Research Findings
Dallas, HF; Shelton, JM; Paxton, BR & Weyl, OLF. 2017. Freshwater Fishes of the Cape Fold Ecoregion
and Climate Change: Volume 1: Research Synthesis. Prepared on behalf of the Table Mountain Fund
by the Freshwater Research Centre. Pp. 12.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
1 | R e s e a r c h S y n t h e s i s
The Cape Fold Ecoregion (CFE) biodiversity hotspot (Fig
1) of South Africa is home to an assemblage of range-
restricted endemic freshwater fishes, the majority of
which face serious risk of extinction. Principal factors
responsible for this situation include human-linked
destruction of aquatic habitats like the introduction of
non-native freshwater fishes, water abstraction (Fig 2)
and pollution. These threats have resulted in marked
decreases in the distribution ranges and abundances of
many native species over the last century. While
research on the detrimental impacts of non-native fishes
and habitat destruction on native fishes has been
undertaken, the more-recently recognised threat of
climate change has, until now, not been adequately
addressed.
Fig 1. Primary catchments and biomes falling within the Cape
Fold Ecoregion which were included in this study.
Freshwater ecosystems will be among the systems most
strongly impacted by climate change. Several climate
model projections warn of widespread invasions,
extinctions and the redistribution and loss of critical
ecosystem functions within mediterranean regions like
the CFE. Climate change predictions for the CFE include
measurable increases in water temperature and
decreased total runoff over the next 50-100 years. The
predicted decrease in river flows and increase in water
temperatures could potentially increase the risk of
extinction for already-fragmented remnant populations
of native fishes. This situation presents a major challenge
for conservation organisations mandated with devising
strategies to prevent species extinctions. Designing
effective conservation plans to safeguard these species
and associated habitats into the future requires reliable
information on species sensitivities to changes in
temperature and flow, and on how climate change
impacts will influence existing species population trends.
The research presented in this document represents the
first attempt to examine the consequences of climate
change for freshwater fishes in South Africa, if not in
Africa (Fig 3). Despite being widely acknowledged as a
top threat to biodiversity and ecosystem functioning,
climate change impacts are notoriously difficult to
quantify. This is partly because of the challenges
associated with conducting studies at the timescales over
which climate change processes operate, and partly
because of the challenge of uncoupling climate change
impacts from other threats and environmental pressures.
Thus, climate change investigations generally adopt the
approach of examining relationships between species
and temperature, or flow, and then scaling-up those
relationships to estimate changes in species’ distributions
under different climate change scenarios - an approach
that has been adopted in this project.
This research was a collaboration between the
Freshwater Research Centre (FRC), the South African
Institute for Aquatic Biodiversity (SAIAB) and
CapeNature; and was jointly funded by the Table
Mountain Fund (WWF) and the Water Research
Commission (WRC) of South Africa.
Fig 2. Over-abstraction of water is a key factor contributing to
aquatic habitat degradation in rivers in South Africa’s Cape
Fold Ecoregion (CFE)
AIMS AND OBJE CTIVE S
The overall objective of the study was to increase our
understanding of the potential consequences of global
climate change on native and non-native freshwater
fishes in the CFE, and to use this information to inform
conservation efforts, water management and resource
protection in this region. The primary aims of this study
were to:
Determine the vulnerability of native fishes in
the CFE to a changing environment.
IN TRODU CTION
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
2 | R e s e a r c h S y n t h e s i s
Map the current distribution of native and non-
native fish species in the CFE based on existing
and new data.
Evaluate the vulnerability of native fish species
and the threat of non-native fish species in the
CFE under projected climate change.
Characterise flow, habitat and thermal
requirements of target fish species.
Evaluate the likely consequences of climate
change on fish species distributions through
scenario analysis.
Provide recommendations for the conservation
of indigenous fishes in the CFE; thereby
contributing to the identification of high priority
freshwater ecosystems in South Africa that need
to be protected and sustainably managed.
APPROACH
These aims were addressed through a series of
complementary studies, each corresponding to a chapter
in the final report to the Water Research Commission:
Dallas et al. 2017 this document is intended as a
chapter-by-chapter synopsis of that report. The research
employed a range of approaches designed to address
specific gaps in our knowledge about the consequences
of climate change for CFE freshwater fishes. Chapters 3,
4 and 5 comprise a series of field studies that examined
the role of temperature, flow and other environmental
factors in explaining local and regional distribution
patterns of native fish populations in the wild. Chapter
6 estimated upper thermal limits and thermal
preferences of selected native species in an
experimental setting. Chapters 2 and 7 capitalised on
what we know about the factors influencing species
distributions at present and make predictions about the
vulnerability of different species under different climate
change scenarios.
In this document, summaries of the main findings and
conclusions from each chapter are provided, and the key
messages are unpacked under three main themes
including (1) the thermal requirements of CFE native
fishes (Chapters 3, 4 and 6), (2) current threats facing
CFE native fish with a focus on the roles of temperature
and flow as limiting factors (Chapters 3, 4 and 5) and
(3) future threats facing CFE native fish, with a focus on
the roles of temperature and flow (Chapters 2 and 7).
Within each theme (Fig 3), we draw on ‘target species’
(selected data-rich species) as examples to highlight key
aspects of climate change impacts for CFE freshwater
fishes. Finally, we light the way forward for future
research in this field by unpacking key questions that
address the main knowledge gaps identified from the
findings of this research project.
Fig 3. Major themes (centre panels) used to organise key messages emerging from the different studies described in the individual
chapters (summarised in the boxes on the left and right) and feed into managing and mitigating climate change impacts.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
3 | R e s e a r c h S y n t h e s i s
CHA PTER 2: VU LNERAB ILITY O F CFE
FR ESHWA TER FISHES TO CLIM ATE C H ANGE
AND OTH ER HUM AN IM PA CTS
The decline of native freshwater fish populations
throughout the CFE presents a major conservation
challenge that requires identifying management
priorities through assessing species in terms of their
vulnerability to climate change and other human-linked
impacts. In this study we employed the rapid assessment
method of Moyle et al. (2013), to determine the relative
vulnerabilities of different species to climate change and
other human impacts. Eight local freshwater fish experts
conducted vulnerability assessments on 20 native and 17
non-native freshwater fish species present in the CFE.
Results show (1) that native species were generally
classified as being more vulnerable to extinction than
were non-native species, (2) that climate change impacts
are expected to increase the vulnerability of most
native, and some non-native, species, and (3) that
vulnerability hotspots requiring urgent conservation
attention occur in the Olifants-Doring, upper Berg and
upper Breede River catchments in the south-western
corner of the region (Fig 4).
Fig 4. Climate change vulnerability projections for native
freshwater fishes in Cape Fold Ecoregion based on (a) all
species and (b) species for which vulnerability assessments had
high levels of certainty. Species distribution data represent
species records post-2000.
Kilometres
In addition to providing guidance for prioritizing
conservation actions, this study highlights the need for
reliable data on the biology and distribution of many
CFE freshwater fishes (Fig 5), and emphasises that
identification of priority areas for protection should be
based on multiple sources of data.
Fig 5. The charismatic fiery redfin (Pseudobarbus phlegethon)
has been adversely affected by non-native species
introductions and habitat alteration, and now ranks among the
CFE’s must threatened freshwater fishes (photo by Geoff
Spiby).
CHA PTER 3: WE STERN C APE C A SE STUDY:
TH E ROLE OF T EMPER AT U RE I N MEDIATING
NA TIVE A ND NO N-NATI V E FISH
PO PULATIONS
Introduced rainbow trout have invaded many
headwater streams in the CFE and pose arguably the
greatest threat to several species of threatened native
fishes (Fig 6). Trout impacts in these systems appear to
be density-dependent, and we hypothesized that
temperature could be a key factor determining trout
density and corresponding impacts on native fish. In this
study, we took advantage of natural spatial and
temporal thermal heterogeneity in two headwater
streams to investigate the influence of temperature and
other environmental factors on trout density.
We found that temperature was the main factor limiting
trout density during summer surveys (hottest months), with
highest trout densities recorded at relatively cool sites (7
day maximum temp < 24 °C) and trout generally
absent from relatively warm sites (7 day maximum temp
> 27 °C). In comparison, native fish density was best
explained by trout density (inverse relationship) during
summer, but by other environmental variables like
habitat complexity during autumn and spring.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
4 | R e s e a r c h S y n t h e s i s
Fig 6. Introduced rainbow trout pose arguably the greatest
threat to remaining populations of many small-bodied CFE
freshwater fishes, but the strength of their impact is linked to
the thermal dynamics of headwater habitats.
Our results (Fig 8) show that trout distributions expand
during cooler seasons (autumn and spring) when thermal
heterogeneity and maximum temperatures are relatively
low, but contract into thermal refugia (habitat patches
that remain relatively cool) as temperatures and thermal
heterogeneity increase over summer, leaving warmer
habitat patches vacant and able to function as
predation refugia for native fish (Fig 7).
Our findings are somewhat novel given that climate
change is expected to exacerbate non-native species
impacts on native species, and may have the
unexpected outcome of benefiting threatened native fish
populations in CFE headwater streams.
Finally, our understanding of climate change impacts on
fishes in these systems could be advanced through (1)
modelling the extent of trout distribution changes under
different climate change scenarios, and (2) investigating
invasions and impacts of warm-tolerant non-native fishes
into headwater sections of CFE streams.
Fig 7. (Left) One of the many pools in the Amandel River that
provide habitat for native and non-native fishes.
Fig 8. Distribution of sites with trout only, trout and native fish together and native fish only along axes of maximum temperature
(Max_7 over the month preceding fish sampling) during autumn, spring, early summer and late summer in the upper Berg River.
Plots on the left represent the densities of trout, and plots on the right show the densities of native fish (redfin, kurper and galaxias
combined) at each site during each season. Bubble sizes are scaled to fish densities.
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Temperature (C, Max_7)
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Trout & native
 


Native only
Trout only
Trout & native
 
Trout only

 
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Early summer
Late summer
Spring
 50

Fish 100 m-2
Trout Native fish

b)


C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
5 | R e s e a r c h S y n t h e s i s
CHA PTER 4: EA STERN C APE C A SE STUDY:
PS EUDOB ARBUS AFER A ND PSE UDOBAR BUS
SW ARTZI AND T HERMA L REGI MES
Non-native black bass species (largemouth bass
Micropterus salmoides, smallmouth bass M. dolomieu and
spotted bass M. punctulatus) and African sharptooth
catfish Clarias gariepinus have invaded streams in both
the upper Gamtoos and Swartkops River systems (Fig 9).
Due to loss of temperature loggers in the Swartkops
River system (Blindekloof tributary), there were
insufficient temperature data with which to assess the
potential effects of temperature on the Eastern Cape
redfin Pseudobarbus afer distributions. However, the
colonisation of sites in the lower Blindekloof River by P.
afer after the disappearance of M. salmoides and M.
dolomieu suggests that temperature is not a limiting
variable (Fig 10). Densities and distribution of
Pseudobarbus swartzi in the Bos River were influenced by
the presence of non-native predatory fishes but could
also be correlated to temperature. Comparisons of
thermal regimes in the rivers with fish distributions
suggest that P. swartzi downstream distributions are
likely to be influenced by stream temperatures. From a
thermal tolerance perspective, the entire range of
occupancy of P. afer and P. swartzi is habitable by non-
native Micropterus spp. and C. gariepinus.
Fig 9. Preparing to undertake a snorkel survey in the
Blindekloof River, a tributary of the Swartkops River in the
Eastern Cape.
Given the strong impacts of Micropterus spp. on redfin
minnows documented elsewhere this puts these
populations at risk if the physical or behavioural factors
limiting the penetration of non-native fishes into
headwater streams are breached. Due to the strong
impacts of non-native fishes on native fish populations
observed elsewhere, it not impossible to attribute the
Fig 10. Pseudobarbus afer density along the Blindekloof River
in relation to the distance from the Swartkops confluence in km.
The maximum value of the Max_7 temperature along a part
of the distribution is shown, as is the reach of the river where
black bass (Micropterus spp.) have previously been recorded.
absence of redfin minnows in the lower stream reaches
entirely to thermal constraints and it is likely that
temperature and non-native fish predation act in
synergy to limit their downstream distributions.
CHAPT ER 5: BREED E RIV E R RED FIN
(P SEUDOB ARBUS BURCHE LLI) POPULATION
SU RVEYS
In this study, we examined the physico-chemical factors
that may be constraining Breede River redfin
Pseudobarbus burchelli populations (Fig 11) at a reach
scale at the limits of the species’ distribution range.
Thirty one of the 58 sites visited during the survey were
dry. This was partly attributable to drought conditions
prevailing at the time of the survey, but also because of
dams and extensive abstraction from tributary reaches
in mountain stream zones (Fig 12). Non-flowing sites had
higher conductivities, lower dissolved oxygen and higher
total dissolved solids values compared to flowing sites.
P. burchelli populations persisted through the summer
months in rivers with low discharges, but were absent at
sites where there was no flow. The species was present
at sites with mean conductivities (115.5 μmS/cm) but
absent from sites with conductivities of > 677 μmS/cm.
They were present at sites with percentage dissolved
oxygen saturation > 80%, pH > 4.9, and at sites that
displayed considerably lower TDS values than the mean
for all sites.
Comparing historical records with extant populations
sampled during the survey suggest that range
restrictions have probably occurred over the last 35-50
years, but that they are likely to have stabilized at their
current levels over the last decade.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
6 | R e s e a r c h S y n t h e s i s
Fig 11. Breede River redfins (Pseudobarbus burchelli) at a site
with healthy flow despite the drought (photo Steve Benjamin).
These preliminary results suggest that although
populations of Breede River redfin persist under drought
conditions in flowing water, they are unlikely to persist in
habitats where flow stops in summer, most likely since
these conditions are associated with water quality
variables which are outside their tolerance ranges. It is
likely that current agricultural impacts will exceed
predicted climate change impacts for some time to
come.
Fig 12. Water abstraction in the Mountain Stream zone in the
Noree River (top) and destruction of aquatic habitats by
bulldozing of stream beds to mitigate flood damage (bottom).
CHA PTER 6: TH ERMAL T OLERA NCES A ND
PR EFERE NCES
The upper thermal limits and thermal preferences of nine
native fish species were determined from five rivers in
the CFE. Upper thermal limits were estimated using the
Critical Thermal Method and thermal preferences using
acute thermal gradient experiments. Hourly water
temperatures recorded in these rivers were used to
generate thermal metrics for characterising thermal
signatures of each river. Both upper thermal limits and
preferences varied significantly between species (Fig
14). Cape galaxias Galaxias zebratus appeared to be
the most thermally sensitive (Fig 13), while Cape kurper
Sandelia capensis was the least thermally sensitive.
Upper thermal limits varied significantly within the genus
Pseudobarbus and the species S. capensis, with
differences primarily linked to the east-west gradient,
where P. afer was less thermally sensitive than its
western counterparts (P. burgi, P. burchelli, P. calidus and
P. phlegethon).
Upper thermal limits and thermal preferences were
correlated, suggesting there is potential to use thermal
tolerance experiments to inform thermal preference,
although further validation of this relationship is
recommended. Water temperature influenced both
upper thermal limits and thermal preference, with both
correlated with several thermal metrics. Significant
differences in upper thermal limits were observed
between summer and winter, and preferred
temperatures shifted seasonally, with species showing a
preference for a range of temperatures rather a single
temperature. Both thermal tolerance and thermal
preference were thus partially dependent on the water
temperature to which the fish had acclimatised.
Fig 13. The Cape galaxias Galaxias zebratus was found to be
the most thermally-sensitive native species, but this may be due
in part to the thermal history of the study site from where the
species was collected (photo Geoff Spilby).
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
7 | R e s e a r c h S y n t h e s i s
CHA PTER 7: PR EDICTIN G THE RESP ONSES
OF ENDEMIC FR ESHWA TER FISH SPEC IES TO
CLIMATE CHANG E IN T H E CAPE FOLD
EC OREGI ON USI NG SP EC IES D ISTRIBUTION
MO DELS
With climate change models predicting an increasingly
dry climate, many rivers may cease flowing for longer
periods of the year with elevated temperatures in the
remaining pools and deteriorating water quality
conditions placing native fish species under increasing
stress (Fig 15). Species Distribution Models (SDMs)
showed that the geographic ranges of all native fish
species will become restricted, but to differing degrees.
The models suggest that most species may be more
sensitive to temperature rather than flow, but that this
may be due to training the models on present-day
distribution data which is influenced by non-native fish
invasions. Both altitudinal (elevation), as well as
latitudinal (north-south) and longitudinal (east-west) shifts
are apparent among the native fish species with a
general trend being a retreat from the more northern
and eastern limits of their range, together with a shift
toward the upper headwater reaches of the eastern
arm of the Cape Fold Belt mountains (Fig 16). The SDMs
predict that populations of non-native smallmouth bass
Micropterus dolomieu may suffer only minor range
restrictions under the given scenarios (Fig 17). On the
other hand, changes predicted for non-native rainbow
trout Oncorhynchus mykiss include range reductions under
flow and rainfall reduction, as well as increased
temperature scenarios, with most pronounced range
contractions into the south-eastern portions of the Cape
Fold mountains.
Fig 15. Many river reaches in the CFE may cease flowing for
long periods during the year.
The SDMs developed in this study provide a potentially
valuable tool for achieving conservation objectives
particularly for identifying climate change refugia for
species at risk. The findings reported on here will
provide useful starting point for informing water
resource management and river rehabilitation priorities,
as well as long term conservation planning.
Fig 16. The non-native smallmouth bass Micropterus dolomieu
was among the species predicted by the species distribution
models (SDMs) to be least affected by climate change impacts
(photo Otto Whitehead).
Fig 14. Thermal tolerance (median, quartiles and range), expressed as CTmax, for all species in the Western Cape ordered from
most thermally-sensitive to least thermally sensitive. All illustrations are ©NRF-SAIAB.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
8 | R e s e a r c h S y n t h e s i s
SU MMARY OF MA IN RE SU LTS A ND
CONCLUSIONS
THERMAL REQUIREMENTS OF NATIVE FISHES
The field (Chapters 3, 4 and 5) and laboratory
(Chapter 6) studies examining the thermal requirements
and limitations of native species both suggest that in
general, CFE freshwater fishes are more sensitive to high
temperatures than are warm-adapted non-native fishes
like centrarchids and clariids, but less sensitive than cold-
adapted non-native salmonids. Native species (P. burgi
and S. capensis, Fig 18) were present at many of the
warmest sampling sites that the cold-adapted non-native
salmonid rainbow trout O. mykiss vacated during warm
summer months, and congregated in cooler thermal
refuge habitats. This distribution pattern may represent
a thermal preference of native species for relatively
warm sites, or avoidance of cooler sites occupied by
predatory, cold-adapted non-native species like O.
mykiss.
The experimental evaluations undertaken in both the
Western and Eastern Cape suggest that thermal
tolerance and preference varies significantly between
species, with G. zebratus most thermally sensitive, while
S. capensis was least thermally sensitive (Fig 19).
Differences in environmental thermal history within
systems is an important source of variation in thermal
tolerance and preference. At a broader scale,
differences in thermal sensitivity of species between the
Western and Eastern Cape may be attributable to
differences in the thermal regimes of the systems where
they evolved.
Fig 18. Native species like Pseudobarbus burgi and Sandelia
capensis were present at many of the warmest sites that the
cold-adapted non-native salmonid rainbow trout O. mykiss
Fig 17. Predicted probabilities of occurrence for the Breede River redfin Pseudobarbus burchelli for four climate change scenarios.
Warmer colours (orange and red) show areas predicted by the model to be more suitable for the species.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
9 | R e s e a r c h S y n t h e s i s
vacated during warm summer months (photo Steve Benjamin).
However, CTmax values represent the extreme upper
temperature that a fish can tolerate for a relatively
short time period. The upper temperature that a fish can
tolerate in the wild is usually estimated over a more
biologically-meaningful time scale, for example by using
the chronic stress threshold value such as the Maximum
Weekly Allowable Temperature (MWAT). However, the
conservation status of these fishes prevents the
establishment of these MWAT experimentally. Instead
one is reliant on the in-situ water temperature data,
whereby 7-day moving average of daily mean or
maximum water temperatures (Mean_7 and Max_7)
can be used as an estimate of a chronic stress threshold.
Thus, relating CTmax values to maximum stream
temperatures (be it thermal maxima, or 7-day means
and maxima) should be considered a conservative
approach to assessing the thermal suitability of stream
sites for a species.
Fig 19. Experimental evaluations suggest that thermal
tolerance and preference varies significantly between species,
and Sandelia capensis was found to be the least thermally
sensitive native species.
Several of the field sites supporting the highest densities
of native fish recorded in our surveys were also the
warmest sites, with Mean_7 (7-day moving average of
the daily mean temperature) values typically exceeding
22 C in the Amandel River, and 23 C in the Berg River.
This observation suggests that native fish may be
actively selecting the warmest sites, which may represent
a thermal preference or the avoidance of cooler sites
more suitable for non-native O. mykiss in these systems.
Future survey work should focus on investigating thermal
preferences of native fish species in the wild, in the
context of laboratory-derived thermal limits and
preference values. The role of thermal history as a
source of variation in thermal tolerance and preference
is identified as a key question to be addressed in future
research.
CURRENT THREATS FACING NATIVE FISH
POPULATIONS
The vulnerability analysis (Chapter 2) and field survey
chapters (Chapters 3, 4 and 5) offer insights into the
present-day role of temperature and other factors in the
imperilled status of CFE freshwater fishes. Results
indicate that habitat degradation and the impacts of
non-native fishes currently pose the greatest threats to
the native species of interest in this study. From a habitat
perspective, over abstraction of water is identified as a
key threat to P. burchelli (Chapter 5), and pollution a
key threat to P. afer (Chapter 3), and associated habitat
factors may also have played a strong role in
fragmenting native fish distributions in the past.
A combination of over abstraction and severe drought
created situations where P. burchelli was pushed beyond
the limits of its environmental tolerance (Chapter 5). This
survey suggested that the species’ environmental
tolerance was exceeded at sites where flow ceased,
conductivities were high and dissolved oxygen
concentrations were low. Similarly, there was also some
evidence that low dissolved oxygen levels in summer
may be a limiting factor for P. burchelli in the Amandel
River (Chapter 3). Ultimately, over-abstraction is likely
to interact with other climate change impacts to reduce
flow and dissolved oxygen levels, and raise water
temperatures beyond a species’ environmental tolerance
limits (Fig 20).
Fig 20. Over abstraction of water, pollution and associated
effects on stream habitat may have played a strong role in
fragmenting CFE native fish distributions in the past.
On the other hand, temperature does not appear to be
an obvious determinant of patterns in species
abundance observed in the wild at present.
Temperature generally did not emerge as a key
predictor of native species distributions in the field
studies. This finding is consistent with the result that
experimentally-determined CTmax values for target
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
10 | R e s e a r c h S y n t h e s i s
species (Chapter 6) were not exceeded in the wild
(Chapters 3 and 4) and the observation in the Amandel
and Berg Rivers that native species were abundant at
the warmest sites during the summer surveys (Chapter 3).
There was some evidence, however, that temperature
may influence native fish distributions seasonally at the
different field sites (Chapter 3), but in general it was
difficult to separate the influence of temperature from
other correlated habitat factors. The presence of non-
native species remained by far the most important
determinant of native species distributions in all systems
where target species distributions were studied,
especially during summer. However, temperature does
appear to be a key factor determining non-native
species distributions. For example, in the Blindekloof
River in the Eastern Cape (Chapter 4), non-native
centrarchids which impact strongly on native P. afer
appear to only establish at downstream sites that are
relatively warm. The SDMs for centrarchids (Chapter 7)
do not forecast notable range contractions like those
forecast for most native species (including P. afer), which
may increasingly give centrarchids the upper hand as
climate change impacts manifest. Unlike salmonids,
centrarchid impacts on CFE native fish are not density-
dependent (impacts can be severe even where
centrarchid densities are relatively low), elevating the
conservation concern around potential range expansions.
Conversely, the distribution and impact of O. mykiss in
the Western Cape appears to be limited to relatively
cool sites (Chapter 3), which has clear consequences for
the distribution of native species like P. burchelli, P. burgi
and S. capensis. Thus, temperature may have a strong
indirect influence on native species distributions through
limiting the distribution of predatory non-native species
and thereby facilitating seasonal predation refugia in
certain habitats (Fig 21). Indeed, the Species Distribution
Models for O. mykiss (Chapter 7) forecast notable
range contractions which may relieve pressure on native
species in certain headwater streams.
FUTURE THREATS FACING CFE NATIVE FISH
There was convergence of the results from the
vulnerability analysis (Chapter 2) and Species
Distribution Models (SDMs) (Chapter 7) by way of the
prediction that native fish species are expected to be
adversely effected by climate change predictions for
the CFE. The SDMs predict that most species will be more
sensitive to changes in temperature than changes in flow
although from Chapter 3 we know that they are
sensitive to no-flow conditions. The SDMs forecast
notable range contractions for all target species under
the scenario of a 2 °C increase in temperature, and
scenarios with a 2 °C temperature increase together
with a 20% flow reduction. Expected changes in both
temperature and flow under climate change scenarios
also came through strongly as future limiting factors in
the vulnerability analysis. However, there were also
some discrepancies between the SDMs and expert-
based predictions for some of the species. For example,
while the SDMs warn of substantial range contractions
for P. burchelli and S. capensis, the expert-based
predictions suggest the opposite that species ranges
may increase. This is because the distribution of O.
mykiss, the greatest threat to these species at present, is
expected to contract upstream as headwaters warm,
increasing the availability of predation refugia. In this
respect, it should be noted that the SDMs did not include
non-native species as a predictive factor.
In general, the SDMs forecast range contractions from
east to west, and into increasingly higher-altitude
upstream habitats for many species (e.g. for P. burchelli,
S. capensis and G. zebratus). These range shifts highlight
the conservation and management importance of
undisturbed, high-altitude perennial habitats in the
winter rainfall portion of the region. Correspondingly,
the vulnerability analysis highlights headwater
catchment areas in the south-west corner of the CFE as
vulnerability hotspots that deserve special conservation
attention (Fig 22).
Fig 21. Temperature may have a strong indirect influence on
native species distributions through limiting the distribution of
predatory non-native species.
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
11 | R e s e a r c h S y n t h e s i s
Finally, both approaches indicate that the warm-
adapted non-native species such as centrarchids and the
clariids are unlikely to change much under climate
change scenarios. On the other hand, both methods
forecast range contractions for cold-adapted non-native
salmonids in the region, which may have positive
benefits for native species on which salmonids prey in
some situations (see Chapter 4 for details).
Fig 22. The SDMs and vulnerability analysis both identify
headwater catchment areas in the south-west corner of the CFE
as vulnerability hotspots that deserve special conservation
attention.
TH E WAY FORW ARD
While this collaborative research project has improved
our understanding of the potential consequences of
global climate change on native and non-native
freshwater fishes in the Cape Fold Ecoregion, it became
evident that there are still gaps in our knowledge and
further steps are needed to fully integrate the research
outputs into management and conservation planning.
Recommendations for future research are detailed
below.
DISTRIBUTION AND IMPACT OF NON-NATIVE SPECIES
IN THE CFE
Results indicate that the impacts of non-native fishes,
together with habitat degradation, currently pose the
greatest threats to the native species in the CFE.
Understanding these impacts is crucial if further invasions
and habitat deterioration are to be halted. Specifically:
What is the role of natural and artificial
barriers in the future of CFE native fishes?
Given the potential for increases in non-native
species distributions, how will existing and new
distribution barriers come into play?
What are the impacts of warm-adapted non-
native fishes other than centrarchids on CFE
native fishes?
THERMAL LIMITS, TOLERANCES AND PREFERENCES
This project has generated the first data on thermal
tolerance and preferences for freshwater fish in South
Africa. From this research, it became evident that certain
aspects necessitate further examination, including:
What are biologically meaningful estimates of
upper thermal limits for CFE freshwater fish
species in the wild?
What is the role of environmental thermal
history on expressed tolerances and
preferences?
How robust are the relationships between
thermal preference and tolerance for CFE
native species?
How does habitat heterogeneity influence
thermal refugia at the reach scale?
RIVER FLOW, HABITAT AND WATER QUALITY
What are the preferences and tolerances for
climate change-linked environmental variables
other than temperature for native CFE fish?
FORECASTING CHANGES IN SPECIES VULNERABILITIES
AND DISTRIBUTIONS
How will vulnerability analyses and SDMs
change as pending species complexes are
described?
Can non-native fish distributions and associated
impacts be incorporated into SDMs as
predictors, and if so how does this influence
predictions of native fish distributions under
climate change scenarios?
INCORPORATING KNOWLEDGE GENERATED ON
CLIMATE CHANGE AND FRESHWATER FISHES OF THE
CFE INTO CONSERVATION AND MANAGEMENT PLANS
How do the SDMs developed in this study align
with FEPAs to identify likely climate change
refugia for each species?
IDENTIFY AND ESTABLISH LONG-TERM SITES FOR
MONITORING WATER TEMPERATURE AND FLOW
Long-term data are inherently valuable for
evaluating the effects of climate change on
aquatic ecosystems. What criteria should be
used for strategically selecting sites to serve as
sentinel sites for evaluating the potential
impacts of climate change? How can
monitoring of these sites be integrated into
existing institutional business plans in the future
C l i m a t e C h a n g e a n d F r e s h w a t e r F i s h
12 | R e s e a r c h S y n t h e s i s
unless otherwise specified. In addition to some of the
authors, the following researchers undertook species
assessments for the vulnerability study and provided
useful comments Chapter 2: Dean Impson
(CapeNature), Riaan van der Walt (CapeNature),
Martine Jordaan (CapeNature), Albert Chakona (South
African Institute for Aquatic Biodiversity), Sean Marr
(South African Institute for Aquatic Biodiversity) and
Bruce Ellender (South African Institute for Aquatic
Biodiversity).
Two MSc students undertook the thermal tolerance and
preference studies. Ms Jody-Lee Reizenberg - MSc at
University of Cape Town (2015 2016), MSc awarded
in June 2017. Ms Lesley Bloy - MSc at Rhodes
University (2016 2017).
Several interns from the Freshwater Research Centre
assisted with field surveys, data consolidation and
analysis, including Jenna Bowker, Nonkanyiso Zungu,
Chloe Wallace, Martin Emanuel, Tumisho Ngobela, Toni
Olsen, Brad Robertson and Lily Bovim (DST/NRF Intern -
2017). We are most grateful to the landowner De Wet
Conradie for provided access to sites on the Amandel
River via his farm, Kanetvlei.
The Project Team (L to R): Jeremy Shelton, Riaan van der Walt,
Helen Dallas, Nick Rivers-Moore, Dean Impson, Tumisho
Ngobela, Olaf Weyl, Bruce Paxton, Martine Jordaan, Jody-
Lee Reizenberg.
and what alternative monitoring actions can be
incorporated?
RE FEREN CES
Dallas HF, Shelton JM, Paxton BR, Weyl OLF,
Reizenberg J, Bloy L & Rivers-Moore N. In Press.
Assessing the effect of climate change on native and non-
native freshwater fishes of the Cape Fold Ecoregion, South
Africa. Water Research Commission Report for Project
K5/2337. Water Research Commission, Pretoria, South
Africa.
Shelton J, Chakona A, Ellender B, Esler K, Impson D,
Jordaan M, Marr S, Ngobela T, Paxton B, Van der Walt
J, Weyl O & Dallas H. In Press. Vulnerability of Cape
Floristic Region freshwater fishes to climate change and
other human impacts. Aquatic Conservation: Marine and
Freshwater Ecosystems.
ACK NOWLEGEMENTS
This project was funded by the Table Mountain Fund
(WWF) TM 2490) and the Water Research
Commission (K5/2337) Several individuals guided the
project through our Reference Group including Prof
Paul Fouché (University of Venda), Mrs Debbie Muir
(Department of Environmental Affairs), Mr Dean Impson
(CapeNature), Dr Sean Marr (University of
Limpopo/South African Institute of Aquatic Biodiversity),
Dr Andrew Gordon (Department of Water and
Sanitation), Dr Neels Kleynhans (Department of Water
and Sanitation), Mr Jan Venter (East Cape Parks/Nelson
Mandela University) and Prof Karen Esler (Stellenbosch
University).
Mr Dean Impson, Dr Martine Jordaan and Mr Riaan van
der Walt provided invaluable advice throughout the
project. Dean Impson also assisted with selecting study
sites, refining sampling protocols, fish surveys and
assisting with the identification of experts for the
vulnerability study.
Fish distribution data were provided to the project team
by Dr Roger Bills of the South African Institute for
Aquatic Biodiversity (SAIAB) and Dr Martine Jordaan of
CapeNature (State of Biodiversity SOB) database.
The illustrations used in this report have been made
available by the South African Institute for Aquatic
Biodiversity. The original paintings are by Elizabeth Tarr
and David Voorvelt and were produced for A Complete
Guide to the Freshwater Fishes of Southern Africa
(2001) by Paul Skelton. All illustrations are ©NRF-
SAIAB. All photographs were taken by the project team
... Formal and informal information sources were used to compile a database of occurrence records of all Black Bass species in South Africa (see Supplementary Table I). Formal distribution records housed at the South African Institute for Aquatic Biodiversity (SAIAB, unpublished data), Ezemvelo KwaZulu-Natal Wildlife (EKZN Wildlife, unpublished data), Mpumalanga Tourism and Parks Agency (MTPA, unpublished data) and the Cape Fold Ecoregion (CFE) (Dallas et al. 2017; CapeNature unpublished data) were used. These were supplemented with data from reviews by De Moor and Bruton (1988) and . ...
Article
Full-text available
Black Bass, a collective name for members of the centrarchid genus Micropterus, are native to North America, but have been introduced globally to enhance recreational angling. This study assessed the distribution of Micropterus salmoides, M. dolomieu and M. punctulatus in South Africa using both formal (survey-based) and informal (tournament data and social media) information sources. Analysis of the distribution data showed habitat bias between the data sources. Survey data from formal information sources were dominated by locality records in riverine environments while those derived from informal information sources focused more on lacustrine habitats. Presence data were used to develop niche models to identify suitable areas for their establishment. The predicted distribution range of M. salmoides revealed a broad suitability over most of South Africa, however, the Cape Fold Ecoregion and all coastal regions were most suitable for the establishment for both M. dolomieu and M. punctulatus. Flow accumulation and precipitation of coldest quarter were the most important environmental variables associated with the presence of all Black Bass species in South Africa. In addition, anthropogenic disturbance such as agricultural activities were associated with the presence of both Smallmouth Bass and Spotted Bass. An extensive area-based invasion debt was observed for all Micropterus spp. The potential for further spread of Black Bass in South Africa is of ecological concern because of their impact on native biota.
Article
Native freshwater fish populations throughout South Africa's Cape Fold Ecoregion (CFE) are in decline as a result of human impacts on aquatic habitats, including the introduction of non‐native freshwater fishes. Climate change may be further accelerating declines of many species, although this has not yet been studied in the CFE. This situation presents a major conservation challenge that requires assigning management priorities through assessing species in terms of their vulnerability to climate change. One factor hindering reliable vulnerability assessments and the concurrent development of effective conservation strategies is limited knowledge of the biology and population status of many species. This paper reports on a study employing a rapid assessment method used in the USA, designed to capitalize on available expert knowledge to supplement existing empirical data, to determine the relative vulnerabilities of different species to climate change and other human impacts. Eight local freshwater fish experts conducted vulnerability assessments on 20 native and 17 non‐native freshwater fish species present in the CFE. Results show (1) that native species were generally classified as being more vulnerable to extinction than were non‐native species, (2) that the climate change impacts are expected to increase the vulnerability of most native, and some non‐native, species, (3) that vulnerability hotspots requiring urgent conservation attention occur in the Olifants‐Doring, upper Berg and upper Breede River catchments in the south west of the region, (4) that in addition to providing guidance for prioritizing management interventions, this study highlights the need for reliable data on the biology and distribution of many CFE freshwater fishes, and (5) that identification of priority areas for protection should be based on multiple sources of data.
Assessing the effect of climate change on native and nonnative freshwater fishes of the Cape Fold Ecoregion, South Africa
  • H F Dallas
  • J M Shelton
  • B R Paxton
  • Olf Weyl
  • J Reizenberg
  • L Bloy
  • N Rivers-Moore
Dallas HF, Shelton JM, Paxton BR, Weyl OLF, Reizenberg J, Bloy L & Rivers-Moore N. In Press. Assessing the effect of climate change on native and nonnative freshwater fishes of the Cape Fold Ecoregion, South Africa. Water Research Commission Report for Project K5/2337. Water Research Commission, Pretoria, South Africa.