Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

Research paper

ISSN (online) 1816-7950 
Available on website https://www.watersa.net

96

CORRESPONDENCE
Robin Petersen

EMAIL
robin.petersen@sanparks.org

CURRENT ADDRESS
Scientific Services, South African 
National Parks, The Royal Hotel, 
Knysna, South Africa

DATES
Received: 25 April 2022
Accepted: 18 April 2023

KEY WORDS
seasonal
ephemeral
perennial rivers
recharge
discharge
conceptual model
semi-arid

COPYRIGHT
© The Author(s)
Published under a Creative 
Commons Attribution 4.0 
International Licence 
(CC BY 4.0)

The role of groundwater, in general, is often overlooked in freshwater ecosystem management policies and in 
the management of South Africa’s flagship conservation area, the Kruger National Park (KNP). To address this 
gap, a generalised conceptual model of surface water–ground water (sw–gw) interactions in the southern 
and central regions of the KNP was developed. To do this, stable isotope ratios (δ18O and δ2H) of groundwater, 
rainfall and surface water were used to determine the extent to which the base flow of perennial, seasonal 
and ephemeral streams on different geologies (granite vs. basalt) is driven by rainfall or groundwater. These 
results show that the δ18O and δ2H ratios of perennial rivers are similar to that of groundwater, while seasonal 
and ephemeral rivers on basalts have values closer to rainfall. On granite substrates, however, the isotope 
ratios of the seasonal and ephemeral rivers have values closer to groundwater than rainfall. The larger seasonal 
Mbyamiti River had similar isotope ratios to that of groundwater, and the highly ephemeral Nwaswitsontso 
had episodic interaction with groundwater (i.e. isotopic ratios overlap occasionally). These results show that 
decisions necessary for the sustainable management of groundwater resources are better informed when 
the natural interaction, movement, and exchange between groundwater and rivers are understood. This has 
particular relevance for large conservation areas in southern Africa that are expected to experience more 
variable climates in the future with both increases in drought and rainfall intensities.

The use of stable isotopes to identify surface water–groundwater interaction 
in the Kruger National Park, South Africa
RM Petersen1,2 , JM Nel2, T Strydom1, E Riddell3,5, C Coetsee1,6 and E February4

1Scientific Services, South African National Parks, Private Bag X402, Skukuza 1350, South Africa
2Environmental and Water Science Unit, University of the Western Cape, Private Bag X17, Bellville 7530, South Africa
3Conservation Management Services, South African National Parks, Private Bag X402, Skukuza 1350, South Africa
4Department of Biological Sciences, University of Cape Town, HW Pearson Building, University Ave N, Rondebosch, Cape Town 7701, 
South Africa
5Centre for Water Resources Research, University of KwaZulu-Natal, Private Bag X01, Scottsville, Pietermaritzburg 3209, South Africa
6School of Natural Resource Management, Nelson Mandela University, George Campus, George, 6530

INTRODUCTION

Groundwater accounts for almost 90% of Earth’s readily available freshwater yet, because it is out of 
sight, remains underappreciated and poorly managed (Jakeman et al., 2016). While global climate 
change models for southern African savannas predict no change in the amount of total rainfall 
received (Easterling et al., 2000; Frich et al., 2002; Pohl et al., 2017), the models also predict increases in 
rainfall intensity and increases in frequency and severity of droughts (Spinoni et al., 2014). Combined 
with these effects of more variable water supply are development pressures which are resulting 
in increased demands on previously unexploited aquifers (Mussá et al., 2015; Du Plessis, 2019). 
Increasing droughts and anthropogenic demands, paired with poor management of groundwater 
resources, have resulted in multiple problems across the globe. These include overexploitation which 
can result in land subsidence and declines in water table levels, with harmful effects on groundwater-
dependent surface water systems (Vegter and Pitman, 2003; Wada et al., 2010; Wada et al., 2012; 
Wada and Heinrich, 2013). Monitoring and management decisions necessary for the sustainable 
utilisation of groundwater resources are better informed when the natural interaction, movement, 
and exchange between groundwater and rivers are understood (Glazer and Likens, 2012; Gleeson 
and Richter, 2018). In semi-arid ecosystems such as the Kruger National Park (KNP) (Newman  
et al., 2006), groundwater plays an important role in sustaining river baseflows and pools (Zektser and 
Loaiciga, 1993; Parsons, 2004; Hughes et al., 2007; Le Maitre and Colvin, 2008). Given this context, 
large conservation areas such as the KNP are important for understanding natural hydrogeological 
processes largely unaffected by anthropogenic activities. There is a total of 31 584 km of mostly 
second- or higher-order streams flowing across a west−east geological facies from granite to basalt in 
the park. These extensive drainage networks distribute water throughout the park and act as hotspots 
for biodiversity and productivity (Mabunda et al., 2003; Rogers and O’Keeffe, 2003).

The stable 18O and 2H isotopes in water are ideal tracers for catchment-scale studies on surface 
water−groundwater (sw−gw) interactions and have been applied in many systems around the world 
(Gat, 1971; Thomas and Rose, 2003; Gibson et al., 2005; Kalbus et al., 2006; Praamsma et al., 2009). 
These isotopes have been used to identify sources of groundwater recharge (Praamsma et al., 2009; 
Liu and Yamanaka 2012), discharge locations (Gleeson et al., 2009; Praamsma et al., 2009), and to 
develop a conceptual understanding of sw−gw interaction at a landscape scale (Riddell et al., 2016). 
As groundwater is dependent on the accumulation of rain through time, the stable isotope ratio of 
the groundwater will be a weighted average of rainwater inputs through time. As a result the isotope 
ratio of groundwater tends to be different from that of rain and surface water (Gat, 1971).

https://www.watersa.net
https://creativecommons.org/licenses/by/4.0/
https://creativecommons.org/licenses/by/4.0/
https://creativecommons.org/licenses/by/4.0/
https://creativecommons.org/licenses/by/4.0/
https://orcid.org/0000-0002-2770-9461


97Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

Here, a conceptual model was developed of the interactions 
between groundwater and perennial, seasonal and ephemeral 
streams in the southern and central regions of the KNP. To do this, 
a comparison was made using stable isotope ratios of δ18O and 
δ2H belonging to groundwater, rainfall and surface water. It was 
anticipated that when there is extensive mixing of water among 
rainwater, surface water and groundwater, isotopic ratios will be 
very similar, while limited mixing of these waters will give rise to 
very distinct isotopic ratios (Rozanski et al., 2001). Depending on 
the size and main geologic substrate of the river or stream, the 
extent of interactions will differ. It was predicted on the basis of 
previous work (Rozanski et al., 2001; Xu et al., 2002; Price, 2011) 
that larger perennial rivers will have extensive interactions with 
groundwater reflected in high overlap of isotopic ratios, while 
seasonal streams will have lower interactions with groundwater, 
and ephemeral streams should primarily be rainfall derived and 
have very limited interactions with groundwater.

METHODS

The KNP extends across the Mpumalanga and Limpopo Provinces 
in north-eastern South Africa. It is one of the largest conservation 
areas in Africa (Du Toit et al., 2003), covering almost 2 million 
hectares. It also forms part of the Great Limpopo Transfrontier 
Park, a 3.5 million hectare conservation estate shared with 
neighbouring Zimbabwe and Mozambique. Besides the higher 
elevated areas of the park, i.e., the Lebombo and Malelane 
mountains (~500–800 m), the park is a gently undulating 
landscape between 200 m and 400 m above sea level with a gentle 
gradient to the east (Schutte, 1986; Venter, 1990). Rainfall in the 
KNP is concentrated in a wet season of around 5–8 months at the 
hottest time of the year, alternating with a dry season when there 
is little or no rain (Gertenbach, 1980). Mean monthly maximum 
and minimum temperatures are 26.3°C and 17.5°C, respectively, 
in southern Kruger (i.e. Pretoriuskop) and 29.8°C and 16°C, 
in central Kruger (i.e. at Satara). Mean annual precipitation is  
737 mm at Pretoriuskop and 547 mm at Satara (Zambatis, 2003; 
MacFadyen et al., 2018). Potential evaporation for the Kruger 
region is upwards of 2 000 mm/a (Jovanovic et al., 2015) whilst 
modelled estimates of annual actual evapotranspiration have 
been determined at close to 800 mm/a on granite catchments and  
600 mm/a on basalts (Riddell et al., 2015) Most of this rain occurs 
in the form of thunderstorms and frontal systems, associated with 
the inter-tropical convergence zone (ITCZ), between November 
and March with a peak in January and February (Gertenbach, 
1980). The storms are typically of short duration lasting just 
minutes or a few hours. As a result, the rainfall intensity is often 
high, leading to flash floods in the ephemeral drainage lines 
(Venter et al., 2003). The long-term rainfall of the KNP oscillates 
through periods of above- and below-average rainfall with cycles 
lasting approximately 10 years (Gertenbach, 1980).

The most important lithostratigraphic units are the Basement 
Complex which consists of ancient granitoid rocks of Swazian age 
(> 3 090 Ma), sedimentary and volcanic rock of the Soutpansberg 
Group, and the volcanic rock of the Karoo Supergroup (Venter, 
1990). Granitic rocks occur in the west and the basaltic and 
rhyolitic rocks in the east. A thin north−south strip of sedimentary 
rocks separates the granitic and basaltic rock formations (Bristow 
and Venter, 1986; Schutte, 1986). The influence of geology and 
resulting soils creates a strong correlation with the structure of 
the terrestrial ecosystem (Venter, 1986; Venter et al., 2003). Soil 
profiles generally become shallower as rainfall decreases towards 
the north. This is particularly noted for the coarse-grained soils 
derived from the granitic materials, where soil depths decrease 
from approximately 150 cm in the south near the Pretoriuskop 
area. The Karoo sequence (basalt), which is a predominantly 
flat landscape (low undulation), produces soils that have 

high clay content with olivine-rich clay soils in the northern 
plains and olivine-poor soils in the southern plains. Alluvial 
soils occur along most of the drainage lines in the KNP, the 
extent of which increases as the size of drainage lines increases  
(Venter, 1986; Venter et al., 2003).

The park is drained by two major transboundary river systems, the 
Limpopo system which forms the northern boundary of the park, 
and the Incomati system, of which the Crocodile River forms the 
southern boundary (Venter and Bristow, 1986). The park interior 
is drained by five perennial rivers (i.e., Luvuvhu, Letaba, Olifants, 
Crocodile, and Sabie rivers), which flow west to east across the 
park before flowing into Mozambique and the Indian Ocean. 
Important seasonal and ephemeral rivers that originate in the 
park are the Shingwedzi, Phugwane, and Mphongolo rivers in the 
north, the Tsende, Timbavati, Nwaswitsontso, Ripape, and Sweni 
rivers in the central region, and the Mbyamiti, Nwaswitshaka and 
Mlondozi rivers in the south. These rivers predominantly carry 
water following heavy rainfall during the summer months and, 
depending on the size of the catchment and geology, have flow 
durations that last from a couple of weeks to most of the wet 
season months (Venter and Bristow, 1986).

Crystalline basement aquifers dominate the park and are 
classified into three hydrogeological domains (Fischer et al., 
2009; Du Toit, 2017): (i) composite aquifers comprised of a 
variable thickness of regolith overlying bedrock with the upper 
part frequently fractured; (ii) deep fractured aquifers composed 
mainly of crystalline material (igneous and metamorphic rocks) 
characterized by a complex arrangement of interconnected 
fracture systems; and (iii) alluvial aquifers, where alluvial material 
overlies or replaces the weathered overburden which can be found 
along with river systems such as the Sabie and Sand rivers. The 
depths of groundwater vary from 1.2 m to 40 m and the aquifers 
are low-yielding at 1.5–3 L/s and rarely exceeding 5 L/s (Vegter, 
2003; Du Toit, 2017). The average groundwater recharge rate 
for the entire KNP is estimated at 12 mm/a, 2.3% of the average 
rainfall (Vegter, 2003; Du Toit, 2017). The groundwater flows from 
a regional perspective from west to east through the park, closely 
following the surface water drainage regions (Du Toit, 2017).

Study sites

To describe the nature of sw−gw interaction, i.e., groundwater 
recharge by surface water, or discharge of groundwater into 
surface water, an existing classification system was used whereby 
streams are classified according to streamflow characteristics 
(Xu et al., 2002; Lerner, 2003; Vegter and Pitman, 2003): i.e., 
ephemeral, seasonal and perennial. This resulted in the following 
classifications: Nwaswitsontso (ephemeral) and Sweni (seasonal) 
rivers in the Satara region, and the Mbyamiti (seasonal) and 
Mlondozi (ephemeral) rivers in the south. The Sabie and Sand 
rivers are two major perennial rivers flowing from west to east 
through the study area. Five sites were selected, representative of 
the main geological units in the park. Three boreholes, Rietpan, 
Mafutsu and Sweni-hide, are located in the Karoo Sequence 
consisting of the Letaba formation (basalts) and the intrusive 
Jozini formation (rhyolite), while two boreholes, Jock and 
Msanimond, are located in the Basement complex that consists 
of the Nelspruit granite suite and Orpen gneiss (granites). These 
boreholes are located between 200 m and 500 m from important 
seasonal and ephemeral rivers and between 30 and 60 km from 
the perennial Sabie and Sand rivers (see Fig. 1).

Sampling and analytical procedures

Samples of groundwater, surface water and precipitation were 
collected at the end of each month for two dry seasons (May–Oct 
2010 and 2011) and one wet season (Nov–Apr 2011). Cumulative 



98Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

monthly rainfall samples were obtained from permanent rainfall 
gauges installed close (10–50 m) to the boreholes. The rain 
gauges are standard polypropylene 100 mm gauges with 30 mm 
of paraffin oil to prevent evaporation, and thereby minimise the 
risk of enriching of the remaining water (February et al., 2007b). 
A total of 45 rainfall samples collected throughout the sampling 
period were used to develop a local meteoric water line (LMWL) 
with the equation for the best fit line using a regression analysis 
with an equation δ2H = 8.66 δ18O + 2.23 (R2 = 0.58). Surface 
water samples were collected from the seasonal and ephemeral 
Mbyamiti, Sweni, Mlondozi, and Nwaswitsontso rivers, and 
Sabie and Sand perennial rivers. Water samples collected from 
the associated rivers and streams were sampled directly (grab 
samples), about 10–15 cm below the water surface, either in pools 
or in-stream when rivers were flowing. Groundwater samples 
were collected from the selected boreholes by purging the aquifer 
using a submersible pump powered by a small generator. The 
water samples were collected directly from the outlet pipe after the 
electrical conductivity (EC) stabilized, or three borehole volumes 
were abstracted to ensure that water representative of the aquifer 
was sampled. To avoid evaporation, all samples were stored in 

insulated bottles and followed the sampling protocol described by 
Weaver et al. (2007).

Water samples (n = 176) were analysed for 18O/16O ratios using 
the CO2 equilibrium method of Socki et al. (1992), while 2H/H 
ratios were obtained through the closed tube zinc reduction 
method of Coleman et al. (1982). Isotopic ratios of both hydrogen 
and oxygen were determined using a ThermoFinnigan Delta 
XP Mass Spectrometer (Thermo Fisher Scientific, Waltham, 
Massachusetts, USA). Internal standards were run to calibrate the 
results relative to Vienna Standard Mean Ocean Water (V-SMOW) 
as well as to correct for drift in the reference gas. The analytical 
uncertainty is approximately 2‰ for δ2H and 0.2‰ for δ18O  
(February et al., 2007b).

RESULTS

Rainfall and groundwater

There were seasonal variations in 18O in rainfall as more depleted 
values were observed during the wet season (δ18O = −1.36 ‰) 
compared to that of rainfall occurring during the dry season 

Table 1. Borehole locations and their relative distance from the rivers sampled 

Borehole Latitude Longitude River name Distance from borehole (km)

Rietpan −24.8969 31.91421 Mlondozi 1.398

Mafutsu −25.0643 32.00947 Mlondozi 3.538

Sweni-hide −24.4736 31.97269 Sweni 0.050

Jock −25.2775 31.92509 Mbyamiti 9.365

Msanimond −24.6138 31.70972 Nwatwitsontso 0.244

Figure 1. Map of the study area in the southern section of the Kruger National Park showing location of the boreholes (circles) and rivers 
(triangles) that were sampled



99Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

(δ18O = −0.53‰). As most of the rain at the study site falls during 
the hottest time of the year, the resulting LMWL is composed 
of evaporatively enriched water and plots out below the global 
meteoric water line (Gat, 1971). Groundwater samples had the 
most negative isotope ratios of all the samples collected, with 
mean values of δ18O = −2.96‰ and δ2H = −31.67‰, std dev = 
1.17 and 7.37, respectively. The low standard deviations among 
samples suggest that groundwater over the entire area exhibits 
similar values.

Surface water–groundwater (sw–gw) interaction in 
perennial rivers

To evaluate sw–gw interactions and the temporal variability of 
these interactions, the isotopic values of the perennial rivers, 
Sabie and Sand, were compared with that of groundwater sampled 
from the paired boreholes and to that of rainwater collected at 
these boreholes. The mean isotopic ratios of the water from the 
Sabie and Sand rivers were δ18O = −2.55‰ ± 0.15‰ (mean  
± standard error), n = 15; δ2H = −27.19‰ ± 1.07‰ for the Sabie, 
and δ18O = −2.68‰ ± 0.09‰, n = 8; δ2H = −28.91‰ ± 1.09‰ for 
the Sand. These values are very similar to that of the groundwater 
(δ18O = −2.96‰ ± 0.15‰, n = 42 and δ2H = −31.6‰ ± 0.92‰) 
and noticeably different from that of rainwater (δ18O = −0.60‰ 
± 0.30‰, n = 38, δ2H = −4.54‰ ± 3.01‰; (Fig. 2).

Surface water–groundwater interaction along seasonal 
and ephemeral streams on basalts

The mean isotopic ratios of groundwater sampled at the boreholes 
of Sweni-hide, Mafutsu and Rietpan (δ18O = −2.86‰ ± 0.02‰, n 
= 42, δ2H = −32.5‰ ± 9.05‰) were more depleted than that of 
the water from the Mlondozi and Sweni rivers (δ18O = −1.58‰  
± 3.5‰, n = 25, δ2H = -6.06‰ ± 9.7‰) as well as rainfall (δ18O = 
−0.41‰ ± 0.32‰, n = 26; δ2H = −2.75‰ ± 18.5‰,). River δ18O 
became increasingly enriched as the season progressed, with 
values above 5‰ at the end of the dry season in 2011 (Fig. 3).

Surface water−groundwater interaction along seasonal 
and ephemeral streams on granites

Mean values for surface water in the Mbyamiti River during the 
wet season when the river experienced high flows were δ18O = 
−1.26‰ ± 0.92‰, n = 8 and δ2H = −28.61‰ ± 6.81‰, and the 
groundwater (BH-Jock) values were δ18O = −2.11‰ ± 0.58‰, 
n = 7 and δ2H = −29.81‰ ± 3.90‰ (Fig. 4). During the dry 
season, when the river experienced low flows, mean values for 
surface water were δ18O = −2.50‰ ± 0.05‰, n = 8 and δ2H = 
−19.38‰ ± 5.81‰, while the groundwater values were δ18O = 
−4.15‰ ± 0.02‰, n = 4 and δ2H = −21.49‰ ± 1.44‰. Along 
the ephemeral Nwaswitsontso River, a high-rainfall event during 

Figure 2.  Mean δ18O and δ2H values (± 1 SE) of surface water values from the perennial Sabie (diamonds) and Sand (squares) rivers compared to 
groundwater (triangles) and rainfall (circles). Open symbols are wet season and closed symbols dry season.

Figure 3. Mean δ18O and δ2H values (± 1 SE) of seasonal and ephemeral Sweni (diamonds) and Mlondozi (squares) rivers on the basalt substrate, 
relative to groundwater (triangles) and rainfall (circles) from the Mafutsu, Sweni-hide and Rietpan boreholes and rain gauges. Open symbols are 
wet season and closed symbols dry season.



100Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

the wet season (January 2011) generated streamflow that lasted 
for about 1 month, then progressively receded to shallow pools, 
eventually drying up completely at the end of March 2011. Surface 
water isotope ratios during the flow period were δ18O = −1.55‰  
± 1.61‰, n = 5 and δ2H = −21.64‰ ± 0.86‰ and the groundwater 
(BH-Msanimond) values were δ18O = −2.48‰ ± 2.44‰, n = 4 and 
δ2H= −40.13‰ ± 0.66‰ (Fig. 4). This river did not flow during 
the 2009–2011 drought.

DISCUSSION

The extensive similarity in δ2H and δ18O values between perennial 
rivers and groundwater indicate high interchange of water 
between these systems. Perennial rivers, such as the Sabie and 
Sand within the KNP, are always connected to the groundwater 
system. During the dry season when these rivers experience low 
flows and throughout the wet season during high-flow periods, 
groundwater is consistently discharged into these rivers. As a 
result, these two rivers act as catchment drains, intersecting 
the regional groundwater piezometric surface and maintaining 
baseflow conditions, as has been shown for perennial rivers in 
semi-arid ecosystems elsewhere (Xu et al., 2002; Lerner, 2003; 
Vegter and Pitman, 2003). The hydrological interaction and 
connectivity are critical in similar semi-arid systems where 
groundwater-dependent ecosystems are reliant on groundwater 
discharge during dry seasons (Colvin et al., 2003; Parsons, 2004).

The geology of the catchment associated with the Sweni and 
Mlondozi rivers is primarily composed of basalts and rhyolite, 
which have high clay content (Venter, 1986; Venter et al., 2003). 
The permeability of the surface and subsurface materials can 
greatly affect recharge and discharge processes; recharge is less 
likely to occur in areas that have finer-grained, less permeable 
sediments (Healy, 2010). Therefore, in areas of finer-grained 
sediments there would be decreased infiltration and enhanced 
surface runoff. These processes are observed on the Sweni and 
Mlondozi rivers as rainfall−runoff during the wet season generates 
streamflow; however, as rainfall decreases, streamflow subsides, 
resulting in standing pools of water along certain reaches of the 
river. Evaporation results in the water collected from these pools 
in the dry season having extremely enriched isotope ratios. These 
streams can be classified as detached or remote (Xu et al., 2002; 
Lerner, 2003; Vegter and Pitman, 2003), whereby the stream bed 

materials are impervious and the piezometric head is always 
below the stream bed material.

The stream bed material of the seasonal Mbyamiti and the ephemeral 
Nwaswitsontso rivers on the granite substrate is underlain by 
alluvial deposits, weathered porous granite material (coarse river 
sand) and rock. This combination of material promotes recharge, as 
the resulting soils have relatively high permeability and are capable 
of transmitting water rapidly (Healy, 2010). Typically, in seasonal 
rivers, the piezometric surface of the groundwater fluctuates 
alternately above and below the stream stage (Xu et al., 2002; Lerner, 
2003; Vegter and Pitman, 2003). As a result, when the Mbyamiti 
River experiences peak flows during the wet season, groundwater 
is recharged from the river bed. Conversely, during the dry season, 
groundwater is discharged into the river, maintaining low flows 
and pools. Therefore, the Mbyamiti River is alternately an influent 
and effluent (intermittent) stream. The Nwaswitsontso River only 
flows after high-rainfall events. In typical ephemeral rivers, the 
groundwater piezometric surface is at all times below the stream 
bed level (Xu et al., 2002; Lerner, 2003; Vegter and Pitman, 2003). 
As a result, during the wet season, after significant rainfall events, 
groundwater experiences indirect recharge through the river bed 
which acts as a preferential pathway (recharge sink).

This study has described the hydrogeological interactions between 
a range of river flow conditions and their geological setting, which 
should allow inferences to the region more broadly. Whilst it would 
be speculative to ascribe climate change impacts on these processes 
in the absence of a detailed modelling study, our observations 
do support recent opinion on the resilience of groundwater 
systems under climate change in dry subtropical regions of Africa 
(Cuthbert et al., 2019), where the positive feedback of the spatial-
temporal intensification of precipitation anomalies could lead to 
increased recharge, particularly through ephemeral drainages, 
thus being made available for discharge to perennial surface water 
systems, despite possible drying trends overall.

Due to the complexities associated with quantifying and concep-
tualising sw−gw interactions (Sophocleous, 2002), particularly in 
semi-arid drainage regions where groundwater discharge into a 
river is not evenly distributed along its length but rather along 
discrete locations which include along fractures, faults and dykes, 
it is not uncommon to have different reaches gaining or losing 
water within a particular stream or river (Newman et al., 2006).  

Figure 4. Mean δ18O and δ2H values (± 1 SE) for water from the granite substrate, ephemeral Nwatwitsontso (diamonds) and Mbyamiti (squares) 
rivers relative to groundwater (triangles) and rainfall (circles) from the Jock and Msanimond boreholes and rain gauges. Open symbols are wet 
season and closed symbols dry season.



101Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

We, therefore, caution against the oversimplification of sw−gw 
interaction on a river or catchment scale. Nonetheless, the find-
ings presented in this paper contributes to the understanding of 
sw−gw interaction to aid in further development and refining of 
sustainable groundwater resource management in the KNP.

CONCLUSION

In semi-arid African savanna, surface water and groundwater re-
sources are intrinsically linked. Understanding these hydrologi-
cal processes to manage the natural interaction, movement, and 
exchange between groundwater and rivers will be critical for the 
sustainable management of water resources within the KNP, as 
well as extrapolating this to the broader Lowveld catchments tak-
ing into account future climate change and accruing these benefits 
downstream of the protected area.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the Mellon Foundation for 
the funding provided as part of the Junior Scientist Programme, 
SANParks, Ms Shamiela Davids for her support and expertise 
at the University of Cape Town isotope lab, the field assistance 
provided by Mr Renson Thethe and Mr Velly Ndlovu from 
SANParks, Ms Chenay Simms for producing the map and Dr 
Stefanie Freitag-Ronaldson, Dr Rina Grant and Dr Izak Smit for 
their guidance and mentorship during the project.

ORCID

Robin Petersen 
        https://orcid.org/0000-0002-2770-9461

REFERENCES

BRISTOW J and VENTER F (1986) Notes on the Permian to recent 
geology of the Kruger National Park. Koedoe 29 85–104. https://doi.
org/10.4102/koedoe.v29i1.522

COLEMAN ML, SHEPHERD TJ, DURHAM JJ, ROUSE JE and 
MOORE GR (1982) Reduction of water with zinc for hydrogen 
isotope analysis. Anal. Chem. 54  993–995. https://doi.org/10.1021/
ac00243a035

COLVIN C, LE MAITRE D and HUGHES S (2003) Assessing terrestrial 
groundwater dependent ecosystems in South Africa. WRC Report 
No. 1090-2/2/03. Water Research Commission, Pretoria.

CUTHBERT M, GLEESON T, MOOSDORF N, BEFUS KM, 
SCHNEIDER A, HARTMANN J and LEHNER B (2019) Global 
patterns and dynamics of climate-groundwater interactions. Nat. 
Clim. Change 9 137–141. https://doi.org/10.1038/s41558-018-0386-4

DU PLESSIS A (2019) Evaluation of southern and South Africa’s 
freshwater resources. In: Du Plessis A (ed.) Water as an Inescapable 
Risk. Current Global Water Availability, Quality and Risks with a 
Specific Focus on South Africa. 147–172. Springer Link. https://doi.
org/10.1007/978-3-030-03186-2_7

DU TOIT J, ROGERS K and BIGGS H (2003) The Kruger Experience: 
Ecology and Management of Savanna Heterogeneity. Island Press, 
Washington DC.

DU TOIT WH and VERSTER H (2017) Establishing a groundwater 
monitoring network in the Kruger National Park: Monitoring the 
fluctuation of groundwater levels under un-impacted conditions 
to determine short, medium and long term trends and to estimate 
TPC levels for abstraction boreholes. DWS Report No. GH 4264. 
Department of Water and Sanitation, Pretoria.

EASTERLING DR, MEEHL GA, PARMESAN C, CHANGNON SA, 
KARL TR and MEARNS LO (2000) Climate extremes: observations, 
modeling, and impacts. Science 289 2068–2074. https://doi.org/ 
10.1126/science.289.5487.2068

FISCHER S, WITTHÜSER K, BIRKE M, LEYLAND R and SCHNEIDER 
M (2009) Regional description of the groundwater chemistry of 
the Kruger National Park (KNP) using multivariate statistics. In: 
Groundwater Conference of the Geological Society of South Africa, 
16–18 November 2009, Somerset West, South Africa.

FRICH P, ALEXANDER LV, DELLA-MARTA P, GLEASON B, 
HAYLOCK M, TANK AK and PETERSON T (2002) Observed 
coherent changes in climatic extremes during the second half of the 
twentieth century. Clim. Res. 19  193–212. https://doi.org/10.3354/
cr019193

GAT JR (1971) Comments on the stable isotope method in regional 
groundwater investigations. Water Resour. Res. 7 980–993. https://
doi.org/10.1029/WR007i004p00980

GERTENBACH WD (1980) Rainfall patterns in the Kruger National 
Park. Koedoe 23 35–43. https://doi.org/10.4102/koedoe.v23i1.634

GIBSON J, EDWARDS T, BIRKS S, ST AMOUR N, BUHAY W, 
MCEACHERN P, WOLFE B and PETERS D (2005) Progress in 
isotope tracer hydrology in Canada. Hydrol. Process. 19  303–327. 
https://doi.org/10.1002/hyp.5766

GLAZER AN and LIKENS GE (2012) The water table: The shifting 
foundation of life on land. Ambio 41  657–669. https://doi.org/ 
10.1007/s13280-012-0328-8

GLEESON T, NOVAKOWSKI K, COOK PG and KYSER TK (2009) 
Constraining groundwater discharge in a large watershed: Inte- 
grated isotopic, hydraulic, and thermal data from the Canadian 
shield. Water Resour. Res. 45. https://doi.org/10.1029/2008WR007622

GLEESON T and RICHTER B (2018) How much groundwater can we 
pump and protect environmental flows through time? Presumptive 
standards for conjunctive management of aquifers and rivers. River 
Res. Appl. 34 83–92. https://doi.org/10.1002/rra.3185

HEALY RW (2010) Estimating Groundwater Recharge. Cambridge 
University Press, Cambridge. https://doi.org/10.1017/CBO9780511 
780745

HUGHES DA, PARSONS R and CONRAD JE (2007) Quantification 
of the groundwater contribution to baseflow. WRC Report No. 
1498/1/07. Water Research Commission, Pretoria.

JAKEMAN AJ, BARRETEAU O, HUNT RJ, RINAUDO J-D, ROSS A, 
ARSHAD M and HAMILTON S (2016) Integrated groundwater 
management: an overview of concepts and challenges. In: Jakeman 
AJ, Barreteau O, Hunt RJ, Rinaudo J-D and Ross A (eds) Integrated 
Groundwater Management. 3–20. Springer Link. https://doi.org/ 
10.1007/978-3-319-23576-9_1

JOVANOVIC N, MU Q, BUGAN RDH and ZHAO M (2015)  
Dynamics of MODIS evapotranspiration in South Africa. Water SA 
41 (1) 79–90. https://doi.org/10.4314/wsa.v41i1.11

KALBUS E, REINSTORF F and SCHIRMER M (2006) Measuring 
methods for groundwater? surface water interactions: a review. 
Hydrol. Earth Syst. Sci. Discuss. 10 873–887. https://doi.org/10.5194/
hess-10-873-2006

LE MAITRE DC and COLVIN CA (2008) Assessment of the contri-
bution of groundwater discharges to rivers using monthly flow 
statistics and flow seasonality. Water SA 34  549–564. https://doi.
org/10.4314/wsa.v34i5.180652

LERNER D (2003) Surface water-groundwater interactions in the 
context of groundwater resources. In: Xu Y and Beekman HE (eds) 
Groundwater Recharge Estimation in Southern Africa. IHP Series 
No. 64. UNESCO, Paris.

LIU Y and YAMANAKA T (2012) Tracing groundwater recharge  
sources in a mountain-plain transitional area using stable 
isotopes and hydrochemistry. J. Hydrol. 464  116–126. https://doi.
org/10.1016/j.jhydrol.2012.06.053

MABUNDA D, PIENAAR DJ and VERHOEF J (2003) The Kruger 
National Park: a century of management and research. In: Du Toit 
J, Rogers K and Biggs H (2003) The Kruger Experience: Ecology and 
Management of Savanna Heterogeneity. Island Press, Washington 
DC. 3–21.

MACFADYEN S, ZAMBATIS N, VAN TEEFFELEN AJ and HUI 
C (2018) Long‐term rainfall regression surfaces for the Kruger 
National Park, South Africa: a spatio‐temporal review of patterns 
from 1981 to 2015. Int. J. Climatol. 38 2506–2519. https://doi.
org/10.1002/joc.5394

MUSSÁ F, ZHOU Y, MASKEY S, MASIH I and UHLENBROOK S 
(2015) Groundwater as an emergency source for drought mitigation 
in the Crocodile River catchment, South Africa. Hydrol. Earth Syst. 
Sci. 19 1093–1106. https://doi.org/10.5194/hess-19-1093-2015

NEWMAN BD, VIVONI ER and GROFFMAN AR (2006) Surface 
water-groundwater interactions in semiarid drainages of the 
American southwest. Hydrol. Process. 20 3371–3394. https://doi.
org/10.1002/hyp.6336

https://orcid.org/0000-0002-2770-9461
https://doi.org/10.4102/koedoe.v29i1.522
https://doi.org/10.4102/koedoe.v29i1.522
https://doi.org/10.1021/ac00243a035
https://doi.org/10.1021/ac00243a035
https://doi.org/10.1038/s41558-018-0386-4
https://doi.org/10.1007/978-3-030-03186-2_7
https://doi.org/10.1007/978-3-030-03186-2_7
https://doi.org/10.1126/science.289.5487.2068
https://doi.org/10.1126/science.289.5487.2068
https://doi.org/10.3354/cr019193
https://doi.org/10.3354/cr019193
https://doi.org/10.1029/WR007i004p00980
https://doi.org/10.1029/WR007i004p00980
https://doi.org/10.4102/koedoe.v23i1.634
https://doi.org/10.1002/hyp.5766
https://doi.org/10.1007/s13280-012-0328-8
https://doi.org/10.1007/s13280-012-0328-8
https://doi.org/10.1029/2008WR007622
https://doi.org/10.1002/rra.3185
https://doi.org/10.1017/CBO9780511780745
https://doi.org/10.1017/CBO9780511780745
https://doi.org/10.1007/978-3-319-23576-9_1
https://doi.org/10.1007/978-3-319-23576-9_1
https://doi.org/10.4314/wsa.v41i1.11
https://doi.org/10.5194/hess-10-873-2006
https://doi.org/10.5194/hess-10-873-2006
https://doi.org/10.4314/wsa.v34i5.180652
https://doi.org/10.4314/wsa.v34i5.180652
https://doi.org/10.1016/j.jhydrol.2012.06.053
https://doi.org/10.1016/j.jhydrol.2012.06.053
https://doi.org/10.1002/joc.5394
https://doi.org/10.1002/joc.5394
https://doi.org/10.5194/hess-19-1093-2015
https://doi.org/10.1002/hyp.6336
https://doi.org/10.1002/hyp.6336


102Water SA 49(2) 96–102 / Apr 2023
https://doi.org/10.17159/wsa/2023.v49.i2.3992

PARSONS R (2004) Surface water:groundwater interaction in a 
Southern African context. WRC Report No. TT 218/03. Water 
Research Commission, Pretoria.

POHL B, MACRON C and MONERIE P-A (2017) Fewer rainy days and 
more extreme rainfall by the end of the century in Southern Africa. 
Sci. Rep. 7 1–7. https://doi.org/10.1038/srep46466

PRAAMSMA T, NOVAKOWSKI K, KYSER K and HALL K (2009) Using 
stable isotopes and hydraulic head data to investigate groundwater 
recharge and discharge in a fractured rock aquifer. J. Hydrol. 366 
35–45. https://doi.org/10.1016/j.jhydrol.2008.12.011

PRICE K (2011) Effects of watershed topography, soils, land use, and 
climate on baseflow hydrology in humid regions: A review. Progr. 
Phys. Geogr. 35 465–492. https://doi.org/10.1177/0309133311402714

RIDDELL ES, NEL J, FUNDISI D, JUMBI F, VAN NIEKERK A and 
LORENTZ SA (2015). Ephemeral hydrological processes in savannas. 
WRC Report No. TT 619/14. Water Research Commission, Pretoria.

RIDDELL ES, KILIAN W, VERSFELD W and KOSOANA M (2016) 
Groundwater stable isotope profile of the Etosha National Park, 
Namibia. Koedoe 58 1–7. https://doi.org/10.4102/koedoe.v58i1.1329

ROGERS KH and O’KEEFFE J (2003) River heterogeneity: ecosystem 
structure, function and management. In: Du Toit J, Rogers K and 
Biggs H (2003) The Kruger Experience: Ecology and Management of 
Savanna Heterogeneity. Island Press, Washington DC. 189–218.

ROZANSKI K, FROEHLICH K and MOOK WG (2001) Environmental 
Isotopes in the Hydrological Cycle. Principles and Applications. 
Surfacewater. IHP-V: Technical Documents in Hydrology, Vol. 3 (ed. 
WG Mook), No. 39, UNESCO, Paris.

SCHUTTE I (1986) The general geology of the Kruger National Park. 
Koedoe 29 (1) a517. https://doi.org/10.4102/koedoe.v29i1.517

SOCKI RA, KARLSSON HR and GIBSON EK (1992) Extraction 
technique for the determination of oxygen-18 in water using 
preevacuated glass vials. Anal. Chem. 64  829–831. https://doi.org/ 
10.1021/ac00031a026

SOPHOCLEOUS M (2002) Interactions between groundwater and 
surface water: the state of the science. Hydrogeol. J. 10 52–67. https://
doi.org/10.1007/s10040-001-0170-8

SPINONI J, NAUMANN G, CARRAO H, BARBOSA P and VOGT J 
(2014) World drought frequency, duration, and severity for 1951-
2010. Int. J. Climatol. 34 2792–2804. https://doi.org/10.1002/joc.3875

THOMAS J and ROSE T (2003) Environmental isotopes in hydrogeology. 
Environ. Geol. 43  532–532. https://doi.org/10.1007/s00254-002-06 
77-x

VEGTER J (2003) Hydrogeology of Groundwater: Region 19: Lowveld. 
WRC report No. TT 208. Water Research Commission, Pretoria.

VEGTER J and PITMAN W (2003) Recharge and stream flow. In: Xu 
Y and Beekman HE (eds) Groundwater Recharge Estimation in 
Southern Africa. IHP Series No. 64. UNESCO, Paris. 109–123.

VENTER F (1986) Soil patterns associated with the major geological 
units of the Kruger National Park. Koedoe 29  125–138. https://doi.
org/10.4102/koedoe.v29i1.525

VENTER F (1990) A classification of land for management planning in 
the Kruger National Park. PhD thesis, University of South Africa, 
Pretoria.

VENTER F and BRISTOW J (1986) An account of the geomorphology 
and drainage of the Kruger National Park. Koedoe 29  117–124. 
https://doi.org/10.4102/koedoe.v29i1.524

VENTER FJ, SCHOLES RJ and ECKHARDT HC (2003) The abiotic 
template and its associated vegetation pattern. In: Du Toit J, 
Rogers K and Biggs H (2003) The Kruger Experience: Ecology and 
Management of Savanna Heterogeneity. Island Press, Washington 
DC. 83–129.

WADA Y and HEINRICH L (2013) Assessment of transboundary 
aquifers of the world-vulnerability arising from human water 
use. Environ. Res. Lett. 8 024003. https://doi.org/10.1088/1748-
9326/8/2/024003

WADA Y, VAN BEEK LP, SPERNA WEILAND FC, CHAO BF, WU YH 
and BIERKENS MF (2012) Past and future contribution of global 
groundwater depletion to sea‐level rise. Geophys. Res. Lett. 39. 
https://doi.org/10.1029/2012GL051230

WADA Y, VAN BEEK LP, VAN KEMPEN CM, RECKMAN JW, 
VASAK S and BIERKENS MF (2010) Global depletion of 
groundwater resources. Geophys. Res. Lett. 37. https://doi.org/10. 
1029/2010GL044571

WEAVER JM, CAVE L and TALMA AS (2007) Groundwater sampling. 
WRC Report No. TT 303. Water Research Commission, Pretoria.

XU Y, TITUS R, HOLNESS S, ZHANG J and VAN TONDER G (2002)  
A hydrogeomorphological approach to quantification of ground-
water discharge to streams in South Africa. Water SA 28  375–380. 
https://doi.org/10.4314/wsa.v28i4.4910

ZAMBATIS N (2003) Determinants of grass production and 
composition in the Kruger National Park. MSc thesis, University of 
Natal, Pietermaritzburg.

ZEKTSER I and LOAICIGA HA (1993) Groundwater fluxes in the global 
hydrologic cycle: past, present and future. J. Hydrol. 144 405–427. 
https://doi.org/10.1016/0022-1694(93)90182-9

https://doi.org/10.1038/srep46466
https://doi.org/10.1016/j.jhydrol.2008.12.011
https://doi.org/10.1177/0309133311402714
https://doi.org/10.4102/koedoe.v58i1.1329
https://doi.org/10.4102/koedoe.v29i1.517
https://doi.org/10.1021/ac00031a026
https://doi.org/10.1021/ac00031a026
https://doi.org/10.1007/s10040-001-0170-8
https://doi.org/10.1007/s10040-001-0170-8
https://doi.org/10.1002/joc.3875
https://doi.org/10.1007/s00254-002-0677-x
https://doi.org/10.1007/s00254-002-0677-x
https://doi.org/10.4102/koedoe.v29i1.525
https://doi.org/10.4102/koedoe.v29i1.525
https://doi.org/10.4102/koedoe.v29i1.524
https://doi.org/10.1088/1748-9326/8/2/024003
https://doi.org/10.1088/1748-9326/8/2/024003
https://doi.org/10.1029/2012GL051230
https://doi.org/10.1029/2010GL044571
https://doi.org/10.1029/2010GL044571
https://doi.org/10.4314/wsa.v28i4.4910
https://doi.org/10.1016/0022-1694(93)90182-9