RH: Lodge and Pretorius • Resource Partitioning in Large Grazing Herbivores Resource Partitioning and Interspecific Competition amongst Large Grazing Herbivores SKYE C. LODGE, Centre for Wildlife Management, University of Pretoria, Private Bag X20 Hatfield, Pretoria, 0028, South Africa YOLANDA PRETORIUS, Centre for Wildlife Management, University of Pretoria, Private Bag X20 Hatfield, Pretoria, 0028, South Africa ABSTRACT Active adaptive management plays a crucial role in the management of Mabula Game Reserve, Limpopo. By studying resource partitioning between the large grazing herbivores in the reserve, management practices can be adapted according to what grasses are selected for most by specific species and how selective those animals are in terms of the species grazed. The selectivity of the animals determines their niche separation in that grass species occur in specific geologies and landscapes. Niche separation will help determine degree of competition between species by looking at where animal species occur and what other species they overlap with. Multiple
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
RH: Lodge and Pretorius • Resource Partitioning in Large Grazing Herbivores
Resource Partitioning and Interspecific Competition amongst Large Grazing
Herbivores
SKYE C. LODGE, Centre for Wildlife Management, University of Pretoria, Private Bag X20
Hatfield, Pretoria, 0028, South Africa
YOLANDA PRETORIUS, Centre for Wildlife Management, University of Pretoria, Private
Bag X20 Hatfield, Pretoria, 0028, South Africa
ABSTRACT Active adaptive management plays a crucial role in the management of Mabula
Game Reserve, Limpopo. By studying resource partitioning between the large grazing
herbivores in the reserve, management practices can be adapted according to what grasses are
selected for most by specific species and how selective those animals are in terms of the
species grazed. The selectivity of the animals determines their niche separation in that grass
species occur in specific geologies and landscapes. Niche separation will help determine
degree of competition between species by looking at where animal species occur and what
other species they overlap with. Multiple factors were taken into consideration when
determining resource partitioning and niche separation such as the size and composition of the
animal group, male to female ratio and animal size. Among these variables, grass species
consumed, height of grass, and soil type were recorded at allocated sites within the reserve to
determine resource partitioning and niche separation amongst large grazing herbivores. Even
though it has been suggested that the size plays an important role in niche separation, the
digestive system of animals plays an equally important role. Data shows that blue wildebeest
(Connochaetes taurinus) and plains zebra (Equus quagga) overlap considerably. While at
certain sites, dominated by Cynodon dactylon (old settlements), large congregations of
2 Lodge and Pretorius
multiple species occur at one time. Sites with Cynodon dactylon are highly selected for in the
late-wet season by most large grazing herbivores due to its high palatability and availability.
The high overlap at old settlements may indicate low quantities of palatable grasses in the
reserve.
KEY WORDS active adaptive management, Mabula Game Reserve, resource partitioning,
niche seperation.
Small fenced reserves and conservation areas are in need of constant monitoring and
management in order to be sustainable. Adaptive management needs to be applied to these
areas in order to accommodate changes in ecological understanding and knowledge, as well as
enabling managers to adapt to unexpected events (van Wilgen and Biggs 2011).This
management approach incorporates research, planning, management and monitoring in
repeated cycles in order to learn what the best ways are to define an achieve objectives
(Pollard and du Toit 2007). Active adaptive management implies that management regimes
need to be flexible and constantly changed as knowledge and understanding of ecological
processes and interactions improve or as environmental conditions or societal values change
(van Wilgen and Biggs 2011).
In order to manage these small properties as best as possible, the history of land-use must
be known. Knowledge of the interaction between land-use, bush encroachment and soil
conditions is important in terms of aiding the understanding of how past land-use in an area
influences sustainable management (Abera and Belachew 2011). Livestock grazing in areas
where settlements used to occur can affect soil nutrient status either through trampling
(directly) or through nutrient recycling via urine and faeces (indirect) (Augustine et al.
2003).Areas where settlements used to be, have increased total nitrogen (N) levels in the soil
possible due to manure deposition adding nutrients to the soil (Angassa et al. 2012). On the
3 Lodge and Pretorius
other hand, pH declines have been seen in soils previously used for crop harvesting (Brady
and Weil 2002). This pH decline may be attributed to the depletion of basic ions such as Mg2+
and Ca2+ as well as increased leaching (Brady and Weil 2002). Agassa et al. (2012) shows that
there is a negative correlation between pH and grass biomass. Vegetation richness and
diversity are heavily influenced by the nutrient status of the soil, which indicates how
important soil management should be in the management of reserves and conservation areas
(Agassa et al. 2012).
In a previous study done on Mabula Game Reserve by Smallwood (2009), herbivore
distributions were looked at in terms of the areas that were preferred and what the vegetation
and soil characteristics in those areas were. Smallwood (2009) and Wydeven and Dahlgren
(1985) emphasized the importance of understanding the habitat requirements of wildlife
species as well as the interaction between species and among species (interspecific and
intraspecific competition) for effective management to occur. Wildlife monitoring in small
reserve like Mabula Game Reserve is very important as it can provide the manager with
knowledge that is crucial for assessing and designing management programs as well as
stocking densities, harvesting rates and vegetation management (Pollock et al. 2002). Mabula
Game Reserve has re-established wildlife into the area as it was previously used for livestock
and grain production (Smallwood 2009). It is therefore important to know how the past land
use may have affected both the soil and vegetation characteristics as this will in turn affect the
distribution of grazing herbivores in the reserve and the sustainable stocking rates that the
reserve can manage in terms of what resources are available.
The aim of this study is to determine what different large grazing herbivores select for in
terms of resources and habitat type. The habitat types are defined by their past land use (old
settlement or old field) or whether it is a drainage line. Grass species composition on each site
is studied and related to the habitat type as well as the geology to look at differences in
4 Lodge and Pretorius
resource composition between sites. At a herbivore level, it has been suggested that variations
in body size are the main contributor to resource partitioning (du Toit and Owen-Smith 1989).
Size variation among large grazing herbivores could lead to the utilisation of different quality
forages in varied quantities. It was found that smaller herbivores consumed forage higher in
quality than larger herbivores who consumed higher quantities of bulkier diets of lower
quality forage (du Toit and Owen-Smith 1989, Kleynhans et al. 2010).
Another herbivore aspect is the relationship between body size and gut capacity. Clauss et
al. (2007) stated that gut capacity increases linearly with body size. This further suggests that
the large quantity of forage eaten by larger herbivores enables then to consume forages of
lower quality (de Iongh et al. 2011). A study in the Kruger National Park on ungulate
assemblage showed a negative correlation between nitrogen content of herbivore faeces and
their body size (Codron et al. 2007). These findings support the notion that differences in
body mass aid niche separation among co-existing herbivores (the Jarman-Bell principle, Bell
1971, Jarman 1974).
Digestive strategies of the herbivore species are also important as it plays a role in resource
partitioning and niche separation. The digestive strategy of a herbivore will impact the quality
of food that the herbivore can eat (Kleyhans et al. 2010). Non-ruminants such as zebra (Equus
quagga) and white rhinoceros (Ceratotherium simum) are less efficient compared to
ruminants in terms of nutrient extraction from forage (Duncan et al. 1990). Non-ruminants
can compensate for this as they have higher passage rates (Duncan et al. 1990). The higher
passage rate allow for more efficient processing of low quality forage for non-ruminants
compared to ruminants of comparable size (Duncan et al. 1990).
Megaherbivores such as the white rhinoceros are bulk feeders capable of effectively
utilizing poor quality forage, however, protein levels of their diet is relatively high (Owen-
5 Lodge and Pretorius
Smith 1988). The deviation from large species consuming diets of lower quality can be
explained by the digestive strategy, intake rate limitations and passage rates (Owen-Smith
1988, Clauss et al. 2007). A relationship between digestive efficiency and mean retention time
(MRT) in the gastrointestinal tract occurs (Kleynhans et al. 2010). This supports the statement
that non-ruminants are more efficient. MRT can vary between species and can be used as a
tool in understanding the factors influencing resource partitioning (Kleynhans et al. 2010).
Digestibility and protein content of grasses vary throughout the year (Kleynhans et al.
2010). These grass characteristics are also negatively correlated with grass height (Kleynhans
et al. 2010). Grasses vary in terms of seasonal patterns of growth and maturation which in
turn affects forage availability and quality (Kleynhans et al. 2010). This could lead to seasonal
differences in resource partitioning and niche separation among herbivores. Herbivores may
move through a reserve to find better resources in the dry season and therefore change their
distribution patterns compared to those seen in the wet season.
Resource partitioning can therefore be influenced by numerous factors such as body size,
grazing composition, digestive strategy and seasonal variation of grass quality and quantity
(Kleynhans et al. 2010). These factors need to be monitored to analyze their effects on
resource partitioning to determine the main driving forces of niche separation.
STUDY AREA
Mabula Game Reserve (MGR) is situated in Limpopo, South Africa, 47 kilometres from
Bela-Bela. The reserve is located in the Waterberg Mountains at 27°54’ S and 24°46’ E.
MGR is divided into two sections. The western section of 8500 hectares was included in the
study. The reserve is surrounded by electrified game proof fencing (Smallwood 2009). The
public are only permitted to use allocated roads within the reserve to go to and from the
6 Lodge and Pretorius
lodges, time-share and whole owner properties. At every other time people must be in a game
vehicle driven by a reserve ranger or personnel that has completed the required test.
Mountainous terrain divided the reserve from north to south into two main geology types,
namely granite and quartzite. The mountainous range covers 12% of the reserve while the
remaining 88% is divided into 67% plains and 21% drainage lines (Bredenkamp and van
Rooyen 1990). The slow weathering of the geology types has resulted in acidic, sandy, loamy
to gravelly soils that have low fertility (Smallwood 2009). The soil fertility is one of the main
reasons why the vegetation in the reserve is low in quality (Smallwood 2009). The soil types
support approximately 122 grass species that are spread throughout the reserve (J. McMillan,
Mabula Game Reserve Ecologist, personal communication).
Mabula Game Reserve occurs in the Savanna Biome, with a unimodal, subtropical savanna
climate (Low and Rebelo 1998). The reserve has an average annual rainfall of 611.3 mm. The
warmest month of the year is January with an average temperature of 23.3 °C (Smallwood
2009). June is the coldest month with a monthly mean maximum temperature of 12.7 °C
(Smallwood 2009).
The land of which is now Mabula Game Reserve was previously used for agricultural
practices and for this reason there are areas spread throughout the reserve that are classified as
old fields. These areas are distinguishable by their clean edges, poor quality grass species and
ridges in the soil from tillage practices. Close to these old fields are areas classified as old
settlements. These old settlements were areas where previously people used to live and keep
their livestock. These areas where then high denuded in terms of vegetation cover due to
trampling by both people and the livestock in the kraals. These sites are distinguished by
clearings where there is very little tree cover and the ground cover is dominated by couch
grass (Cynodon dactylon).
7 Lodge and Pretorius
One of the old fields on the granite geology has been under veld management practices for
the last 10 years (J. McMillan, Mabula Game Reserve Ecologist, personal communication).
Prior to initiation of these practices, the old field was dominated by yellow thatching grass
(Hyperthelia dissoluta). Through the practice of regularly mowing the area once the grass
reaches a certain height, higher quality grasses have begun to grow in between the thatching
grass. The cutting of the grass has also forced the grass to be in a continuous state of growth
thereby making new green growth available all the time. This has encouraged more wildlife,
both in terms of numbers and species, to visit the site especially in the dry season when little
green growth is available.
METHODS
Field studies were conducted in the late-wet and mid-dry season in Mabula Game Reserve.
A total of 27 days was spent doing research in the reserve. The late-wet season observations
were made from 19 March to 22 March and 25 March to 6 April, and the mid-dry season
observations were conducted from 7 July to 16 July. Observations usually started around
07:00 and ended around mid-afternoon with the latest observations being completed at 15:00.
Due to logistical constraints the variations in the time of day during which observations could
be made could not be controlled for leading to a bias in the amount of observations on the part
of the reserve dominated by granite soils.
The reserve was divided into granite (eastern side of the mountainous range) and quartzite
(western side of the mountainous range) geologies. Each geology type was allocated five
sites: two old fields, two old settlements and one drainage line. Each day the sites were
visited. Once at a site, observations were made as to whether wildlife were grazing on the site
or not. If wildlife were present, the species was identified using binoculars (Steiner SkyHawk
Pro 10x24). Only large herbivore species that are grazers or mixed feeders seen grazing on a
8 Lodge and Pretorius
site were recorded. Along with the herbivore species name, the group size, sex ratio, relative
ages of group (adult or juvenile), GPS co-ordinate (Garmin eTrex Vista) of the vehicle upon
herbivore sighting, direction of sighted herbivores (north, south, west, east etc.), approximate
distance from vehicle, date, time, geology type and site name were recorded. After all of the
information is recorded the herbivores are observed in terms of where they were grazing. The
point of grazing was then approached on foot in order to identify the grass specie(s) that were
grazed. Grass blades and stems with edges that were still green (had not turned brown or
started turning brown) were assumed as being eaten by the observed herbivore species.
Animal tracks were also used to verify assumptions.
At the point of grazing a sample of more than 30 grams of the same grass species was
taken using a sickle. The sample was then placed in a paper bag and labelled accordingly. The
height of the identified eaten grass was also recorded using a disc pasture metre. Once the
sample and height reading had been recorded, the disc pasture metre was used to record the
biomass of the surrounding grasses. A radius of 10 metres around the point of grazing was
established and ten random points were then measured and the most dominant grass species
under the disc was identified and recorded.
This exact procedure mentioned above was used for the late-wet season observations. In
the mid-dry season it was difficult to identify grass species as well as the grasses that were
recently grazed. It was therefore decided to omit the steps of identifying the grasses and
taking samples due to the possibility of incorrectly identifying grazed grass and grass species.
Further sampling was done in the form of line transects. Line transects of 50 metres were
done on each site using a 50 metre tape measure. The species at each one meter mark was
identified and recorded. If the grass was grazed it was recorded accordingly. At one metre
marks where no grass species were found a dot was made during recording meaning that
9 Lodge and Pretorius
nothing was found and the grass species identified was observed 10 centimetres past the one
metre mark. If nothing was found at the position 10 centimetres past the metre mark another
dot was made until a grass tuft occurred at a point. Hits were also recorded. This meant that
the grass species occurred directly under mark.
Chemical analysis was done on samples taken from the most frequently eaten grasses.
Some grass species were repeated in order to determine whether differences in nutrient quality
exist between geology types. Each grass species sampled had a respective late-wet season and
mid-dry season sample so that differences in nutrient quality between these seasons could be
analysed. Chemical analysis was done at Nutrilab, University of Pretoria. Chemical analyses
done were dry matter (DM), neutral detergent fibre (NDF), acid detergent fibre (ADF), and
mineral analyses. The mineral analyses included phosphorus (P), calcium (Ca), copper (Cu),
iron (Fe), manganese (Mn), and zinc (Zn). Each grass sample was analysed in duplicate (as
per Nutrilab procedure) to ensure that errors were minimised.
Data analyses
GPS co-ordinates were converted to a format readable by ArcGIS 10.1. ArcMap was used
to create maps of Mabula Game Reserve using GIS layers provided by the reserve. GPS co-
ordinate of observations in both the late-wet and mid-dry seasons were entered into ArcMap
and layers were created for different herbivores to determine changes in site selection
patterns. Tables and figure were then used to represent data in the figures created in ArcGIS
for clarification of information. Tables of grass species consumed per animal species and their
relative frequencies of consumption were also constructed.
RESULTS
The sites monitored in this study have been labelled accordingly for ease of reference:
10 Lodge and Pretorius
Granite: Old field 1 F1G
Old field 2 F2G
Old settlement 1 S1G
Old settlement 2 S2G
Drainage line DG
Quartzite: Old field 1 F1Q
Old field 2 F2Q
Old settlement 1 S1Q
Old settlement 2 S2Q
Drainage line DQ
S1GS2G
F1G
F2GDG
S1Q
S2Q
F1Q
F2Q
DQ
11 Lodge and Pretorius
Figure 1: Position of sites and their respective abbreviated names in Mabula Game
Reserve. A total of ten sites were observed with five on granite geology and five on quartzite
geology.
Distribution patterns of wet and dry seasons
At each site, animals were observed and the GPS coordinates recorded. This helps
determine changes in distribution. The overall distribution of animals shows differences in
12 Lodge and Pretorius
site selectivity between the late-wet and mid-dry seasons (Figure 2 and Figure 3). A general
movement from granite to quartzite geology was seen in the mid-dry season as well as
significant increases in visits to both F1G and DQ. In the late-wet season, large grazing
herbivores were observed more frequently on S1Q, when compared to observations made in
the mid-dry season, where the frequency of visits to other old settlements remained similar
(S1G, S2G and S2Q). Large grazing herbivores were also more widely spread across the
reserve in the late-wet season. Herbivores did not visit DG frequently in both the late-wet and
mid-dry seasons; however, in comparison, the frequency of observations made at DQ
increased significantly in the mid-dry season. There was a slight increase in observations at
F2G; however, overall very few herbivores were observed grazing at this site. The most
utilised site by all herbivore species was F1G (Figure 2 and Figure 3). On most days there
were more than one species grazing on this site at the same time with an average of 3-4
species being observed at the same time.
13 Lodge and Pretorius
Figure 2: Overall distribution of large grazing herbivores in the late-wet and mid-dry
seasons. The GPS points were plotted onto the Mabula Game Reserve map with the geology
layer to show differences in distribution over the two dominant geology types, granite on the
eastern side of the mountainous ridge and quartzite on the western side of the mountainous
ridge.
14 Lodge and Pretorius
F1G F2G S1G S2G DG F1Q F2Q S1Q S2Q DQ0
5
10
15
20
25
30
35
40
45
50
Mid-dry seasonLate-wet season
Site
Freq
uenc
y
Figure 3: Frequency (number of observations) of visits by large grazing herbivores to sites
compared over the late-wet and mid-dry seasons.
Species specific distribution changes - Some of the herbivore species mentioned in Table 1
were observed on more occasions than others due to their specific habitat preferences. Buffalo
along with roan antelope and blesbok were not observed at all in the late-wet season, on the
allocated sites, with slightly more sightings in the mid-dry season. On the other hand,
waterbuck and nyala were monitored at all in the mid-dry season. Zebra and blue wildebeest
observations remained similar, regarding each species, thereby indicating the possibility of the
two species moving together and therefore being observed at similar frequencies.
15 Lodge and Pretorius
Table 1: Large herbivore species (common and scientific names) that were monitored on
Mabula Game Reserve and the number of times they were observed grazing in the late-wet
and mid-dry seasons.
Number of observations
Common name Scientific name Late-wet Mid-dry
Blesbok Damaliscus pygargus phillipsi 0 5
Blue wildebeest Connochaetes taurinus 20 18
Buffalo Syncerus caffer 0 4
Eland Tragelaphus oryx 4 5
Gemsbok Oryx gazella 1 7
Impala Aepyceros melampus 11 47
Nyala Tragelaphus angasii 4 0
Red hartebeest Alcelphus buselaphus 7 13
Roan antelope Hippotragus equinus 0 1
Tsessebe Damaliscus lunatus 2 3
White rhinoceros Ceratotherium simum 4 9
Zebra Equus quagga 26 26
Waterbuck Kobus ellipsiprymnus 1 0
In Table 2 and Figure 4, the observations of both blue wildebeest and zebra are shown for
the late-wet and mid-dry seasons to illustrate the similarity in habitat preference. Zebra and
wildebeest were the most observed species, apart from impala, over the duration of the study.
They were found on a various sites in the reserve. Site DQ was only visited once by zebra in
the late-wet season while blue wildebeest were not found there at all; however, increased
16 Lodge and Pretorius
observation of both species were made on site DQ during the mid-dry season. Another site
that was not visited by blue wildebeest in the late-wet and mid-dry seasons was DG while
zebra only grazed at DG once during the mid-dry season. Both the points on the map (Figure
4 and values in Table 2) show that both blue wildebeest and zebra tend to occur on the same
sites.
Table 2: Frequency of blue wildebeest and zebra observations per site in the late-wet and