The impact of tree modification by African elephant
(Loxodonta africana) on herpetofaunal species richness in
northern Tanzania
Nabil A. Nasseri, Lance D. McBrayer and Bruce A. Schulte* Department of Biology, Georgia Southern University, Statesboro, GA 30460, U.S.A.
Abstract
In Africa, no other nonhuman animal fulfils the role of
ecosystem engineers to the extent of the elephant. However, little is known about the relationship between elephant modified habitats and species composition of other
animals. Our objective was to sample the herpetofauna
within an Acacia habitat that varied in the degree of elephant impact. If elephant foraging was only modifying but
not degrading or enriching the habitat, then herpetofauna
species abundance and richness were predicted to be
similar in elephant damaged and elephant excluded areas.
We conducted this study at Endarakwai Ranch in northeastern Tanzania for 6 months in 2007 and 2008. We
sampled herpetofaunal species richness and abundance
within high, medium and low elephant damaged areas and
in a plot that excluded elephants. Areas of heavy damage
yielded higher species richness than the exclusion plot.
Species diversity did not differ between the damaged areas
and the exclusion plot. Frogs were more abundant in areas
of high damage; in contrast, toads were found the least in
high damage areas. The results support the notion that
free ranging elephants influence herpetofaunal species
distribution by creating habitat complexity through modifying the woodland area.
Key words: amphibians, habitat modification, reptiles,
savannah
Résumé
En Afrique, à part l’homme, aucun animal ne joue le rôle
d’ingénieur de l’écosystème autant que l’éléphant. Cepen*Correspondence: E-mail: [email protected]
Current address: Department of Biology, Western Kentucky
University, Bowling Green, KY 42101, U.S.A.
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
dant, l’on sait peu de chose de la relation entre les habitats
modifiés par l’éléphant et la composition des autres espèces
animales. Notre objectif était d’échantillonner l’herpétofaune d’un habitat à acacias où le degré d’impact des éléphants était variable. Si l’éléphant, en se nourrissant,
modifiait seulement l’habitat, sans le dégrader ni l’enrichir,
il était prévu que l’abondance et la richesse des espèces de
l’herpétofaune soient similaires à celles des zones endommagées ou désertées par les éléphants. Nous avons
réalisé cette étude à l’Endarakwai Ranch, dans le nord-est
de la Tanzanie, pendant six mois, en 2006 et 2007. Nous
avons échantillonné la richesse et l’abondance, des espèces
de l’herpétofaune dans des zones fortement, moyennement
et peu endommagées par les éléphants et dans une parcelle
d’où ils étaient exclus. Les zones fortement endommagées
contenaient une plus grande richesse en espèces que la
parcelle sans éléphant. La diversité des espèces ne variait
pas entre les zones endommagées et la parcelle d’exclusion.
Les grenouilles étaient plus abondantes dans les zones très
abı̂mées alors que les crapauds étaient là le moins nombreux. Les résultats confortent l’idée que les éléphants en
liberté influencent la distribution des espèces de l’herpétofaune parce qu’ils créent une diversité d’habitats en
modifiant les zones boisées.
Introduction
African elephants (Loxodonta africana Blumenbach) are
ecosystem engineers in that they create and maintain
ecosystems through physically changing the habitat
(Jones, Lawton & Shachak, 1997). Elephants remove
dominant hardy vegetation, which is replaced by quick
growing vegetation thereby transforming dense woodlands
into open grasslands (Laws, 1970; Shannon et al., 2006).
Elephants browse on woody trees, such as acacia (Acacia
1
2
Nabil A. Nasseri et al.
spp.), marula (Sclerocarya birrea Hochst.), mopane (Colophospermum mopane J. Léonard) and baobabs (Adansonia
digitata L.) (Jachmann, 1989; Lewis, 1991; Omondi, Bitok
& Kagiri, 2004). Elephant browsing strategies involve bark
stripping, breaking major branches and uprooting trees
that varies in extent with the time spent in an area, creating a mosaic of altered habitats (Western, 1989).
The rejuvenation of vegetation is stunted when elephants are prevented from moving between forage areas,
leading to permanent, potentially degraded alterations in
the landscape (Birkett & Stevens-Wood, 2005; De Beer
et al., 2006). Yet, elephants encourage succession in areas
that have reached a climax and help control bush
encroachment (Meik et al., 2002). The effects African elephants have on vegetation have been studied extensively
(Osborn & Parker, 2003; Shannon et al., 2006; Boundja &
Midgley, 2009; Chira & Kinyamario, 2009; Guldemond &
Van Aarde, 2007; Ihwagi et al., 2010). However, research
is limited on whether elephants are degrading or
enriching habitats for invertebrate (Haddad et al., 2009) or
vertebrate (Herremans, 1995; Cumming & Brock, 1997;
Pringle, 2008) species.
In some cases, vertebrates are used to assess habitat
quality or otherwise monitor ecosystem conditions, but
precautions are necessary to ensure that the species
selected are viable indicators. Most herpetofaunal species
require strict environments, i.e. refugia and prey source, to
survive and are susceptible to environmental changes.
Recently, the arboreal Kenyan dwarf gecko (Lygodactylus
keniensis Parker) was found to select habitats that became
physically more complex as a result of elephant activity,
i.e. refuges were created by elephants stripping bark and
splintering branches (Pringle, 2008). In a study by Friend
& Cellier (1990), feral pigs and buffalo increased microhabitats for amphibians and to lesser extent reptiles by
expanding ephemeral waterholes through wallowing.
These studies indicate that herpetofauna can be used to
assess the impact megaherbivores have on their surrounding ecosystems.
The objective of the present study was to sample the
reptile and amphibian (herpetofaunal) community within
an Acacia habitat that varied in the degree of elephant
impact. If elephant foraging was degrading the habitat,
then herpetofauna species abundance, diversity and richness were predicted to be higher in areas with lower or no
elephant impact. Conversely, if elephant activities enriched
the habitat, then one or more of these measures of the
herpetofauna would be higher in areas of greater elephant
impact. While amphibians and reptiles as herpetofauna
may be indicators of habitat quality or complexity, the
subgroups and species within these groups differ in habitat
requirements. Therefore, we decided to examine broad
measures such as richness as well as characterize the
abundance of the species within specific sub-groups to
determine the impact of elephant damage on the herpetofauna. Thus, our hypotheses were examined by evaluating the overall herpetofaunal community composition
and then assessing the abundance and species richness of
amphibians, classified as toads (Bufo spp.) and frogs (nonBufo spp.), and reptiles, categorized as non-skinks (nonScincidae) and skinks (Scincidae). These groups were
considered separately because of their different natural
histories and habitat preferences.
Material and methods
Study site
Endarakwai Ranch comprises 4300 ha of mixed savannah
woodland and open savannah located in the in the Kilimanjaro Region of northern Tanzania (S0300.663¢
E3700.113¢) between Amboseli, Mt. Kilimanjaro and
Arusha National Parks (Vyas, 2006; Napora, 2007). The
borders of Endarakwai Ranch are unfenced. On the
southern end of the ranch, there is a permanent 4300 m2
man-made waterhole that is fed by a diversion from the
Ngare Nairobi River, attracting an array of wildlife such as
zebra (Equus quagga burchellii Gray), eland (Taurotragus
oryx Pallas), elephant and wildebeest (Connochaetes taurinus Burchell) (Vyas, 2006). This region experiences a
bimodal seasonal pattern with a short wet and dry season
and a long wet and dry season (Castelda, 2008). Four
volcanic vents surround the study area, and the soil is a
rocky sandy loam.
The sampled areas were mixed woodland habitats primarily composed of Acacia tortilis Hayne with some Acacia
mellifera Benth. The disturbed area was unfenced and
located where elephants traversed and fed regularly
(Napora, 2007). The 250 ha exclusion area (locally
known as Rafiki Farm and henceforth called the control)
had been fenced off since 1998 to exclude large herbivores,
such as zebra (E. q. burchellii Gray), eland (T. oryx Pallas),
elephant, African buffalo (Syncerus caffer Sparrman) and
giraffe (Giraffa cameloparadalis Weithofer); the last two
species were uncommon on Endarakwai Ranch. The
majority of the vegetation in the open disturbed area had
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
Impact of African elephants on herpetofauna
some major branches broken off, and more than 50% of
the canopy had been lost because of elephant activities
(Napora, 2007). With 90% of the vegetation in this
condition, a random selection of trap locations was not
possible. Therefore, locations were selected based on specific guidelines detailing levels of habitat damage (Table 1).
Trapping process
Herpetofauna were sampled from August 2007 to
December 2007 and from February 2008 to March 2008.
From December 2007 to January 2008, the personal safety
hazard caused by the large number of elephants present
required the closing of traps. Herpetofaunal composition
and abundances were recorded at a control plot and an
elephant disturbed open woodland savannah near a permanent waterhole at Endarakwai Ranch. To ensure captures were not influenced by tree species, all drift fences
were placed in areas where the only tree species was
A. tortilis. We sampled three different locations within the
control area and three locations within each damage category of high, medium and low, yielding twelve sampling
locations (three trap arrays ⁄ damage site; see Table 1 for
more detail on damage category). Each of the four locations was termed a damage site.
All sampling was conducted using nonlethal drift fences
with pitfalls and funnel traps as well as opportunistic
observation. Drift fences were 10 m long and 0.5 m high
with a total of 120 m of drift fence encompassing 24 pitfall
traps. Two 61 ·20 cm funnel traps were placed on each
side of the drift fence at the midpoint between the two
pitfalls. The funnel traps were constructed of mosquito
wire screening. Cardboard sheets located over the traps
provided shelter. Opportunistic sampling was carried out
when approaching trap locations. All captured individuals
were released within 24 h. Captured individuals were
Table 1 Classification of elephant damage to the vegetation (Napora, 2007) in drift fence areas at Endarakwai Ranch, Tanzania
from August 2007 to March 2008
Damage
category
Operational definition
Low
No damage to main trunk and with
minimal damage to branches and foliage
Medium
Damage to main trunk (not pushed over) and
>50% of branches and foliage damaged
High
Main trunk pushed over and ⁄ or uprooted
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
3
released 15 m due west of the drift fence from which they
were captured as we determined this would be a safe distance as to reduce the chance of a recapture without
overly displacing the captured individual.
Open traps were checked on a daily basis. Drift fences
were damaged occasionally and so pitfalls were closed until
a new drift fence was installed. There were 1976 trapping
days (number of pitfalls open times number of trapping
days). Using field guides, captured herpetofauna were
identified to the lowest taxonomic level possible (Spawls
et al., 2002; Bauer, 2003; Channing & Howell, 2006),
catalogued and photographed. Captured specimens were
uniquely marked with nontoxic paint and toe clippings
(Dodd, 1993; Johnson, 2005).
Statistical analysis
Species richness. Species richness was calculated as the
total number of species captured within a location. We
compared the total species richness of the four levels of
elephant damage sites (control, low, medium and high).
We also made this comparison for amphibians and reptiles
separately.
Mean species richness data that met the assumptions of
equal variance and normality were analysed using a oneway analysis of variance (ANOVA) to test for variation in
species richness between locations. When appropriate, the
Dunnett’s post hoc test was used to compare the different
damaged sites to the control. When the transformed data
did not meet the assumptions of a parametric ANOVA, the
Kruskal–Wallis test was used (Sokal & Rohlf, 1995).
A nonparametric equivalent of the Dunnett’s test to compare damaged areas to the control was applied (Zar, 1984)
by calculating the q statistic as follows: q = (RA ) RB) ⁄ SE
Where RA is the sum of ranks of a group and RB is the sum
of ranks of the control area. The sum of ranks for each
damage area was calculated and compared to the control
area.
Species abundances. Species abundances were calculated as
total number of individuals captured within a trap location.
We analysed the difference in mean abundance, defined as
the total number of individuals captured across each
replicate of a damage category divided by three, for the
focal herpetofaunal groups (skinks, non-skinks, toads and
frogs). The data that met the assumptions were analysed
using a one-way ANOVA. If there was a significant
difference because of damage site, then the damage sites
4
Nabil A. Nasseri et al.
were compared to the control with Dunnett’s test (Sokal &
Rohlf, 1995) to determine whether herpetofauna used
elephant modified habitats differentially. Data not meeting
assumptions were analysed using the Kruskal–Wallis test,
and differences between sites were tested with the nonparametric analogue of Dunnett’s test.
All statistical analyses were tested to a 95% confidence
limit (a = 0.05) using JMP 7.0.1 (SAS Institute Inc., Cary,
NC, USA). All descriptive statistics are displayed as mean
(±SE) or in percentages. Only data on reptiles within the
order Sauria and amphibians within the order Anura were
analysed. Snakes captures were rare because the traps
were not designed to catch snakes and thus snakes were
excluded from the analyses.
Results
Species richness and abundance
Species richness of the herpetofauna showed a trend
towards greater richness in areas with more elephant
damage to the woody vegetation (F3,8 = 3.18, P = 0.08;
8.7 ± 0.33; Fig. 1). Specifically, the high damage area was
found to have significantly greater species richness than
the control region (Dunnett’s test, P = 0.05). Eighteen
herpetofaunal species, nine saurian and nine anuran, were
sampled in areas of high elephant damage (Table 2).
Medium damage areas were comprised of twelve species
(eight saurian, four anuran), while areas of low damage
had eleven species (seven saurian, four anuran). The
control site had the lowest species richness with only eight
species (five saurian, three anuran). The mean abundance
of herpetofauna captured per damage area ranged from
9.7 ± 2.7 to 14.3 ± 2.2, but the overall abundance did
not significantly differ among the four damage areas
(ANOVA: F3,8 = 0.65, P = 0.61).
Anuran composition across damage sites
Six species comprising four families of non-bufonid frogs
were captured across all study sites (Table 2). Kassina
senegalensis Duméril and Bibron and Xenopus victorianus
Ahl were the only two species of frogs that were captured
outside of high damage areas. Kassina senegalensis was the
only species to be captured in every damage category with
higher prevalence in high (N = 6) and medium (N = 3)
damage areas. Only two species of toads were sampled in
this study (Table 2). Both species were found within all the
damage categories. However, Bufo gutturalis Power
(N = 41) was almost four times more abundant than
Bufo xeros Tandy et al. (N = 11). Seventy per cent of
B. gutturalis individuals were captured in areas of medium
damage and the control; similarly, 60% of the B. xeros
specimens were sampled in the medium damage and
control areas.
Seven species of frogs were captured between the
different trapping sites with the majority being sampled
in high damage areas (Table 2). Frog species richness
significantly differed between the damage areas (F3,8 =
6.46, P = 0.02) with species richness being highest in high
damage areas when compared to the control (P = 0.01). As
only two species of toads (B. gutturalis and B. xeros) were
captured, no difference was detected in species richness
between damage sites (H3 = 2.75, P = 0.43; Fig. 2).
Damage site had a significant effect on toad abundances
but not on frog abundance, although there was a trend
(F3,8 = 6.65, P = 0.01; F3,8 = 3.07, P = 0.09, respectively). Toad abundance was significantly lower in areas of
high damage when compared to the control (P = 0.04);
conversely, frogs were significantly more abundant in the
high damaged sites than the control site (P = 0.05; Fig. 3).
Saurian composition across damage sites
Fig 1 Mean (±SE) herpetofauna species richness based on damage
site at Endarakwai Ranch, Tanzania from August 2007 to March
2008 (* indicates significant difference at P = 0.05)
Seven species of skinks (Scincidae) and six species of nonskinks (non-Scincidae) that comprised four families were
sampled within the different damage sites (Table 2). The
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
Impact of African elephants on herpetofauna
5
Table 2 Herpetofaunal species captured at Endarakwai Ranch, Tanzania from August 2007 to March 2008
Order
Reptilia
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Sauria
Amphibia
Anura
Anura
Anura
Anura
Anura
Anura
Anura
Anura
Anura
Family
Species
High
Agamidae
Gerrhosauridae
Gekkonidae
Gekkonidae
Gekkonidae
Lacertidae
Scincidae
Scincidae
Scincidae
Scincidae
Scincidae
Scincidae
Scincidae
Total
Agama agama Linnaeus, 1758
Gerrhosaurus flavigularis Wiegmann, 1828
Hemidactylus squamulatus Tornier, 1896
Lygodactylus laterimaculatus Pasteur, 1964
Lygodactylus picturatus Peters, 1868
Latastia longicaudata Reuss, 1864
Leptosiaphos kilimensis Stejneger, 1891
Lygosoma afrum Peters, 1864
Lygosoma sundevalli Smith, 1849
Trachylepis brevicollis Wiegmann, 1837
Trachylepis striata Peters, 1844
Trachylepis varia Peters, 1867
Panaspis wahlbergii Smith, 1849
13
Arthroleptidae
(Hyperoliidae)
Bufonidae
Bufonidae
Ranidae
Ranidae
Ranidae
Hyperoliidae
Pipidae
Unknown
Total
Leptopelis bocagii Günther, 1865
1
Medium
Low
Control
1
1
2
1
1
1
4
2
1
1
1
6
16
1
2
1
2
3
15
3
1
2
1
1
1
4
1
12
1
2
9
15
1
Bufo gutturalis Power, 1927
Bufo xeros Tandy et al., 1976
Cacosternum sp.
Ptychadena mascareniensis Duméril & Bibron, 1841
Tomopterna tandyi Channing & Bogart, 1996
Kassina senegalensis Duméril & Bibron, 1841
Xenopus victorianus Ahl, 1924
Unknown
9
5
2
1
2
2
6
7
1
27
Total
2
1
6
1
1
7
1
5
5
1
2
7
19
58
1
17
3
1
3
24
7
2
12
4
1
7
1
17
17
41
11
1
2
3
11
14
1
85
saurian fauna was distributed similarly throughout the
damage sites. Lygosoma afrum Peters and Panaspis
wahlbergii Smith were the only two species captured within
all three damage sites and the control area. Species richness for skinks and non-skinks did not show significant
difference across the damage sites (F3,8 = 0.31, P = 0.82;
F3,8 = 0.41, P = 0.75, respectively); in the same way, as
mean abundances did not differ significantly across damage sites (F3,8 = 0.35, P = 0.79; F3,8 = 0.23, P = 0.87,
respectively).
Discussion
Fig 2 Mean (±SE) toad and frog species richness by damage sites
at Endarakwai Ranch, Tanzania from August 2007 to March
2008 (* indicates significant difference at P = 0.05)
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
We observed that high damage to A. tortilis trees caused by
elephants favoured significantly greater species richness of
herpetofauna than the control area from which elephants
and other large herbivores were excluded. While the foraging and movement of other herbivores can modify
6
Nabil A. Nasseri et al.
2002; McCauley et al., 2006). The majority of the species
captured within the study sites are strict insectivores.
Arthropods and other invertebrates may target elephant
damaged trees to use for feeding and nesting (Larsson
et al., 1983; Harmon et al., 1986). Therefore, the abundance and diversity of prey may be important factors that
attracted these species of herpetofauna to elephant modified versus control areas.
Fig 3 Mean (±SE) toad and frog abundance by damage sites at
Endarakwai Ranch, Tanzania from August 2007 to March 2008
(Asterisks indicate a significance between damage areas for
abundance of toads (*) or frogs (**) at P = 0.05)
habitat, in our study area, African elephants are the most
likely herbivore to impact heavily the woody vegetation.
Therefore, the difference in herpetofauna richness is likely
to be attributable to the exclusion of elephants from the
control area.
Habitat selection is a nonrandom process based on
appropriate habitat characteristics for particular organisms (Goldsbrough, Shine & Hochuli, 2006). Viable refuge
availability is a primary driving force in habitat selection
by herpetofaunal species (Toft, 1985; Meik et al., 2002;
Pringle, 2008). The breaking of branches and uprooting of
trees by elephants result in increased coarse woody debris
that herpetofauna use as refuges, hunting areas and
breeding grounds (Greenberg, 2001). In addition, the
craters and mounds created by uprooted trees form habitats for numerous organisms, some of which may be food
sources for reptiles and amphibians (Guo, 1996; Olff &
Ritchie, 1998) and presumably these craters could also
serve as temporary breeding pools for amphibians during
the rainy seasons.
Species richness was nearly twice as great in areas of
high damage when compared to the control site
(8.7 ± 0.33 and 4.67 ± 0.33, respectively). The high
number of frog species found in high damage areas compared to control areas contributed to this difference. The
increased complexity of the modified area generates new
habitat for a diverse array of frog species. Craters and
coarse woody debris formed by uprooted and broken trees
augmented the number of refuges against predators and
desiccation and elevated perches for calling and foraging
resources. Although high damage areas supported more
species, prey abundance and type are also factors in habitat selection by herpetofauna (Toft, 1985; Meik et al.,
Long-term implications
This study provides evidence that the exclusion of elephants could lead to lower species richness of herpetofauna
because of a reduction in habitat complexity. The habitat
mosaic created by free ranging elephants helps to meet the
habitat requirements of both generalists and the more
sensitive species (e.g. frogs). However, more research is
needed to determine the full impact free ranging elephants
have on other species in their ecosystems. For example, our
study design was not effective at sampling arboreal species
and snakes. Acacia tortilis has very smooth bark that offers
very little protection to arboreal lizards (Meik et al., 2002),
but through bark stripping and breaking minor branches
elephants can create habitats that would otherwise not
exist (see Pringle, 2008). Similarly, herbivory by elephants
is a driving force in ant colonies inhabiting Acacia spp.
trees (Palmer et al., 2008), so food web dynamics need to
be considered. For example, elephant feeding in sand forest
in South Africa leads to an increase in species richness of
spiders through the creation of new microhabitats (Haddad
et al., 2009). Examination of hypotheses related to trophic
dynamics and habitat selection by invertebrate and vertebrate species within savannah woodlands will facilitate a
more efficient and practical landscape management for free
ranging elephants.
Acknowledgements
Research permits were issued by the Tanzanian Wildlife
Research Institute and the Commission for Science and
Technology (Nos. 2006-302-CC-2006-175 and 2008108-NA-2007-120). Peter Jones permitted us to work at
Endarakwai Ranch and the project was facilitated with the
help of all those that live and work there. The study was
approved by the IACUC at Georgia Southern University
(I06039). Assistance was provided by our respective
institutions, including a Georgia Southern University
Academic Excellence award, and a grant from the National
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
Impact of African elephants on herpetofauna
Science Foundation, Award Nos. 02-17062, -17068 and 16862 to B. A. Schulte, T. E. Goodwin and the late L.E.L.
Rasmussen, respectively.
References
Bauer, A.M. (2003) On the identity of Lacerta punctata Linnaeus
1758, the type species of the genus Euprepis Wagler 1830,
and the generic assignment of Afro-Malagasy skinks. Afr. J.
Herpetol. 52, 1–7.
Birkett, A. & Stevens-Wood, B. (2005) Effect of low rainfall and
browsing by large herbivores on an enclosed savannah habitat
in Kenya. Afr. J. Ecol. 43, 123–130.
Boundja, R.P. & Midgley, J.J. (2009) Patterns of elephant impact
on woody plants in the Hluhluwe-Imfolozi park, Kwazulu-Natal,
South Africa. Afr. J. Ecol. 48, 206–214.
Castelda, S.M. (2008) Waterhole Dynamics and Chemical Signals of
African Elephants (Loxodonta africana). Thesis (M.Sc.), Georgia
Southern University, Statesboro.
Channing, A. & Howell, K.M. (2006) Amphibians of East Africa.
Cornell University Press, Ithaca, NY.
Chira, R.M. & Kinyamario, J.I. (2009) Growth response of
woody species to elephant foraging in Mwea National Reserve,
Kenya. Afr. J. Ecol. 47, 598–605.
Cumming, D.H.M. & Brock, M.B. (1997) Elephants, woodlands and
biodiversity in southern Africa. S. Afr. J. Sci. 93, 231–236.
De Beer, Y., Kilian, W., Versfeld, W. & Van Aarde, R.J. (2006)
Elephants and low rainfall alter woody vegetation in Etosha
National Park, Namibia. J. Arid Environ. 64, 412–421.
Dodd, C.K. Jr (1993) The effects of toe clipping on sprint performance of the lizard Cnemidophorus sexlineatus. J. Herpetol. 27,
209–213.
Friend, G.R. & Cellier, K.M. (1990) Wetland herpetofauna of
Kakadu National Park, Australia: seasonal richness trends,
habitat preferences and the effects of feral ungulates. J. Trop.
Ecol. 6, 131–152.
Goldsbrough, C.L., Shine, R. & Hochuli, D.F. (2006) Factors
affecting retreat-site selection by coppertail skinks (Ctenotus
taeniolatus) from sandstone outcrops in eastern Australia. Austral Ecol. 31, 326–336.
Greenberg, C.H. (2001) Response of reptile and amphibian communities to canopy gaps created by wind disturbance in the
southern Appalachians. For. Ecol. Manage. 148, 135–144.
Guldemond, R. & Van Aarde, R. (2007) The impact of elephants on
plants and their community variables in South Africa’s Maputaland. Afr. J. Ecol. 45, 327–335.
Guo, Q. (1996) Effects of bannertail kangaroo rat mounds on
small-scale plant community structure. Oecologia 106, 247–
256.
Haddad, C.R., Honiball, A.S., Dippenarr-Schoeman, A.S., Slotow,
R. & Van Rensburg, B.J. (2009) Spiders as potential indicators
of elephant-induced habitat changes in endemic sand forest,
Maputaland South Africa. Afr. J. Ecol. 48, 446–460.
2010 Blackwell Publishing Ltd, Afr. J. Ecol.
7
Harmon, M.E., Franklin, J.F., Swanson, F.J., Sollins, P., Gregory,
S.V., Lattin, J.D., Anderson, N.H., Cline, S.P., Aumen, N.G.,
Sedell, J.R., Lienkaemper, G.W., Cromack, K. Jr & Cummins, K.W.
(1986) Ecology of coarse woody debris in temperate ecosystems.
Adv. Ecol. Res. 15, 133–302.
Herremans, M. (1995) Effects of woodland modification by African
elephant Loxodonta africana on bird diversity in northern Botswana. Ecography 18, 440–454.
Ihwagi, F.W., Vollrath, F., Chira, R.M., Douglas-Hamilton, I. &
Kironchi, G. (2010) The impact of elephants, Loxodonta africana,
on woody vegetation through selective debarking in Samburu
and Buffalo Springs National Reserves, Kenya. Afr. J. Ecol. 48,
87–95.
Jachmann, H. (1989) Food selection by elephants in the miombo
biome, in relation to leaf chemistry. Biochem. Syst. Ecol. 17, 15–
24.
Johnson, M.A. (2005) A new method of temporarily marking lizards. Herpetol. Rev. 36, 277–279.
Jones, C.G., Lawton, J.H. & Shachak, M. (1997) Positive and
negative effects of organisms as physical ecosystem engineers.
Ecology 78, 1946–1957.
Larsson, S., Oren, R., Waring, R.H. & Barrett, J.W. (1983) Attacks of mountain pine beetle as related to tree vigor of ponderosa pine. For. Sci. 29, 395–402.
Laws, R.M. (1970) Elephants as agents of habitat and landscape
change in East Africa. Oikos 21, 1–15.
Lewis, D. (1991) Observations of tree growth, woodland structure
and elephant damage on Colophospermum mopane in Luangwa
Valley, Zambia. Afr. J. Ecol. 29, 207–221.
Mc Cauley, D.J., Kessing, F., Young, T.P., Allan, B.F. & Pringle,
R.M. (2006) Indirect effects of large herbivores on snakes in an
African savanna. Ecology 87, 2657–2663.
Meik, J.M., Jeo, R.M., Mendelson, J.R. & Jenks, K.E. (2002) Effects of
bush encroachment on an assemblage of diurnal lizard species
in central Namibia. Biol. Conserv. 106, 29–36.
Napora, E.S. (2007) Chemical signaling and resource use by African
elephants (Loxodonta africana). Thesis (M.Sc.), Georgia Southern
University, Statesboro.
Olff, H. & Ritchie, M.E. (1998) Effects of herbivores on grassland
plant diversity. Trends Ecol. Evol. 13, 261–265.
Omondi, P., Bitok, E. & Kagiri, J. (2004) Managing human-elephant conflicts: the Kenyan experience. Pachyderm 36, 80–86.
Osborn, F.V. & Parker, G.E. (2003) Towards an integrated approach for reducing the conflict between elephants and people: a
review of current research. Oryx 37, 80–84.
Palmer, T.M., Stanton, M.L., Young, T.P., Goheen, J.R., Pringle,
R.M. & Karban, R. (2008) Breakdown of an ant-plant mutualism follows the loss of large herbivores from an African savanna. Science 11, 192–195.
Pringle, R.M. (2008) Elephants as agents of habitat creation for
small vertebrates at the patch scale. Ecology 89, 26–33.
Shannon, G., Page, B.R., Duffy, K.J. & Slotow, R. (2006) The role
of foraging behaviour in the sexual segregation of the African
elephant. Oecologia 150, 344–354.
8
Nabil A. Nasseri et al.
Sokal, R.R. & Rohlf, F.J. (1995) Biometry: The Principles and
Practice of Statistics in Biological Research. W H Freeman, New
York.
Spawls, S.K., Howell, K., Drewes, R. & Ashe, J. (2002) A Field Guide
to the Reptiles of East Africa. Natural World, San Diego.
Toft, C.A. (1985) Resource partitioning in amphibians and reptiles. Copeia 1, 1–21.
Vyas, D. (2006) Sexually Dimorphic Development Patterns of
Chemosensory Behaviors in African Elephants (Loxodonta africana).
Thesis (M.Sc.), Georgia Southern University, Statesboro.
Western, D. (1989) The ecological role of elephants in Africa.
Pachyderm 12, 42–45.
Zar, J.H. (1984) Biostatistical Analysis, 2nd Edn. Prentice-Hall,
Englewood Cliffs, NJ.
(Manuscript accepted 26 August 2010)
doi: 10.1111/j.1365-2028.2010.01238.x
2010 Blackwell Publishing Ltd, Afr. J. Ecol.