Abstract
M.Sc.
Pesticides and fertilizers, while allowing increases in food production, also have the ability to
find their way into aquatic systems. Irrespective of their route of entry into an aquatic
ecosystem, they may affect aquatic biota by influencing survival, growth and reproduction.
Secondary effects may occur in which populations of organisms are impacted due to a
reduction or elimination of pollution-susceptible species which results in a disturbance of
biological processes and interactions due to water quality impacts associated with
agricultural practices. Biomonitoring techniques are used to assess the integrity of aquatic
ecosystems and provide information on the environmental conditions that have prevailed
within a river. Because aquatic organisms are exposed to their environment and all pollutants
and toxicants thereof, they will cumulatively reflect the conditions which they are exposed to.
This study aims to assess community structure, biotic integrity and feeding traits of aquatic
communities at river sites that have varying adjacent land uses.
The chosen study area falls within the Crocodile (West) Marico Water Management Area
(WMA). Study sites were selected on the Magalies and Crocodile rivers which form
Hartbeespoort Dam at their confluence. Agricultural intensity in North West Province is high
and irrigation farming tends to be located on the large floodplains associated with the middle
Crocodile River. The main economic activity of the WMA occurs in Gauteng, and is
generated by the intense urban and industrial activities of northern Johannesburg. These
practices ensure that water pollution from agriculture and urban land use is a major problem
along segments of the rivers under investigation.
Study sites were chosen based on their adjacent land use and consisted of sites related to
agricultural, urban and natural activities. Environmental driver components that were
assessed included water quality and habitat integrity (IHAS). Biotic response indices were
implemented to assess the integrity of diatom (GDI, SPI, BDI, EPI and %PTV),
macroinvertebrate (SASS5 and MIRAI) and riparian vegetation (VEGRAI) communities.
Principal Component Analysis (PCA) analyses were undertaken on water quality data using
Primer version 6 to determine patterns in water quality between sites. Multivariate
(CLUSTER, NMDS and RDA) and univariate (Margalef’s index, Shannon-Wiener diversity
index and Pielou’s evenness) analyses were performed on macroinvertebrate family data,
macroinvertebrate FFG data, diatom species data and riparian vegetation data using Primer
version 6 and Canoco version 4.5 in order to elucidate differences in community structure per
land use.
Results indicated that particular water quality and habitat impacts were present for each land
use. Comparison of community structure of diatoms taken from sites with varying land uses
showed differences from one another. Relative reference diatom communities comprised of
diatom species that had preferences for clean water, whilst community structures of diatoms
were modified and showed specific change in relation to agricultural and urban water quality
impacts. An increased diversity in air breathing macroinvertebrates was shown at sites with
agricultural practices at high flow, where urban sites were differentiated from agricultural
sites due to the presence of the Hydropsychidae and Hirudinea families. At low flow
macroinvertebrate families making up communities overlapped between land uses. A
difference noted at low flow was that the contribution of the Chironomidae was higher at
urban sites in comparison to agricultural sites, indicating organic water pollution.
The statistical comparison of macroinvertebrate communities, FFGs and riparian vegetation
showed that differences between sites with different land uses were not significant.
Nonetheless, some differences in refined data were noted for the varying land uses.
Considering the macroinvertebrate community make up of the relative reference site, which
was comprised of macroinvertebrate families that were more sensitive and showed
preferences for higher water quality, community structures of macroinvertebrates were
modified and showed change in relation to land use.
Macroinvertebrate FFGs indicated that a change in the input of UPOM at agricultural sites,
and a change in the presence of FPOM at urban sites were responsible for a shift in the FFG
dominance. A difference in riparian integrity was noted between relative reference and test
sites, but could not be easily distinguished between test sites with different land uses.
Riparian integrity was more predictive of macroinvertebrate FFG structure than actual
macroinvertebrate community structure. This indicates that riparian integrity and comparison
with biological traits such as FFGs were useful in showing impacts due to organic matter
inputs.
Overall biotic indices were less useful in distinguishing between urban and agricultural land
uses. It appeared that biotic indices masked the changes in the actual taxonomic
components, erroneously suggesting that sites with different land uses are similar in terms of
ecosystem integrity. It must be noted that integrity indices certainly have an important place
in management of aquatic systems, but it appears to be more useful to utilise taxonomic
make up and biological traits (in this case of FFGs) to show specific impacts, as these are
factors which can be compared across a relatively broad spatial scale.
It can be concluded that sites could be separated according to land use based on community
structure of diatoms and macroinvertebrates, and biological trait analysis of feeding groups. It
was noted that diatom communities were more defined in their response to land use
practices in comparison to macroinvertebrate communities.