Shaping of shrubs and trees by herbivory in Ithala Game Reserve, Zululand, South Africa

@dhoare @tonyrebelo @jeremygilmore @richardgill @alastairpotts @troos @botaneek @wynand_uys @craigpeter @rob_palmer @ricky_taylor @reubenheydenrych @pieterwinter @adriaan_grobler @andrew_hankey @benjamin_walton @zarek

In August-September 2000, I spent about ten days in Ithala Game Reserve (, with as much time as possible on foot (see

One of my search-images was for visible damage to, and/or shaping of, woody plants by herbivores.

In the case of mistletoes, I have not tried to identify the spp., instead mentioning them as part of their host-spp.


Aloidendron barberae

  • fallen mature specimen (tree), with crown heavily browsed by Diceros bicornis (many branches up to 5 cm diam. chewed off and eaten); note that Dracaena aletriformis, although common at same location, was untouched by D. bicornis or other herbivores


Cussonia spicata

  • dead specimen, formerly 7 m high, with bole broken at a height of 1.5 m by Loxodonta africana; on the felled section, the foliage and some of the bark eaten
  • felled specimen, formerly 6 m high, probably pushed over by L. africana, and now reduced to skeleton
  • specimen with a clear browse-line, the height of which indicates Giraffa giraffa


Osyris lanceolata

  • specimen of shrubby growth-form, with a hedged appearance, with stems of diam. 2.5 cm, at height >1 m, grossly broken, probably by male of Strepsiceros strepsiceros


Ehretia rigida and possibly also Ehretia amoena

  • several specimens shaped by nibbling by G. giraffa
  • a few specimens showing damage by D. bicornis
  • specimen heavily browsed


Ximenia americana and/or caffra

  • shrubby specimen 2.5 m high, heavily browsed
  • specimen of arborescent growth-form, <4 m high, heavily browsed by G. giraffa, to the point of defoliated stems; regenerating fresh shoots that are browsed as soon as they appear


Heteropyxis natalensis

  • shrubby specimen 5 m high, with browse-line corresponding to height of G. giraffa, and signs of heavy browsing of epicormic regrowth after fire, probably by S. strepsiceros


Hippobromus pauciflorus

  • shrubby specimen <2.5 m, heavily browsed


Combretum erythrophyllum

  • specimen showing signs of browsing by G. giraffa


Cola greenwayi

  • shrubby specimen with stems of diam. 1 cm chopped and eaten by D. bicornis


Brachylaena discolor

  • suppressed specimen, heavily browsed


Euphorbia tirucalli

  • mature specimen on edge of riverbank, with bark removed by D. bicornis

Euphorbia ingens

  • several tall specimens, >15 m high, showing obvious damage by Papio ursinus: pale scars from broken-off segments of branches; on the ground and in the crotch of main branches, several discarded sections, still green, 30-60 cm long, from which the green 'wings' (which project about 2.5 cm from the green stems) have been bitten/nibbled, indicating that not just the growing tips but also the older stems (succulent) are eaten


Dovyalis rhamnoides

  • specimen chopped grossly by D. bicornis


Ziziphus mucronata

  • mature specimen with indistinct browse-line, corresponding to height of G. giraffa
  • several specimens in the form of small trees, heavily browsed by G. giraffa, to the point that foliage is regenerating from relatively thick stems


Berchemia zeyheri (

  • specimen closely browsed by G. giraffa and S. strepsiceros, so that crown eliminated and leaves restricted to branches
  • specimen <5 m high, heavily browsed by G. giraffa, to the point of defoliation
  • suppressed specimen <5 m high, heavily browsed
  • sapling specimen with bark stripped by L. africana
  • suppressed specimen 1 m high, hedged by heavy browsing
  • mature specimen at least 12 m high, with a distinct browse-line attributable to G. giraffa


Canthium inerme

  • specimen heavily browsed

Note: in a wider context, the extreme form of spinescence in this species is noteworthy, suggesting adaptation vs stripping by L. africana


Searsia leptodictya

  • specimen fairly heavily browsed, with induced branching configuration right-angled, tending to bar stripping

Searsia pallens

  • in one stand of this species, many broken branches, up to 1 m long, littering the ground, with dead leaves still attached

Searsia pentheri and Searsia rehmanniana (both occur, but are easily confused)

  • specimen on alluvial plain, browsed by D. bicornis (chopped stems 1 cm thick)
  • specimens with chopped stems, of diam. 1 cm, indicating that D. bicornis ate branches up to 1 m long
  • specimen with a lower branch broken by D. bicornis


  • suppressed specimen 1 m high, with small leaves and many spines 1.5 cm long, hedged to a spherical shape
  • several specimens bearing mistletoes; these parasitic plants have been browsed by G. giraffa to the point of hedging, but not killed/eliminated
  • specimen >3 m high, heavily browsed by G. giraffa; basically non-spinescent, so browsing has produced a coral-like effect, rather than hedging

Gymnosporia buxifolia and Gymnosporia senegalensis (see

Gymnosporia nemorosa

  • mistletoe at 1.5 m high, the parasite heavily browsed, probably by S. strepsiceros


Elaeodendron transvaalense

  • specimen 1.5 m high, heavily browsed, with hedged growth-form, bearing mistletoe, also heavily browsed and hedged
  • shrubby specimen (3 stems at ground level) with a browse-line (2.2 m high) corresponding to height of S. strepsiceros


Mystroxylon aethiopicum

  • mistletoe present on tree 5 m high, indicating that the dense populations of herbivores in the reserve have not eliminated these parasites


Diospyros dichrophylla

  • shrubs < 1 m high, not visibly shaped, but said by game-guard, Leonard Gumede, to be a preferred food-plant of D. bicornis

Diospyros lycioides

  • specimen with stem of diam. <1 cm chopped by D. bicornis, at height 1 m
  • scandent, somewhat liane-like specimen in the crown of a specimen of Vachellia nilotica that had been broken down by D. bicornis; in the case of V. nilotica, most of the branches chopped at stem diam. <1 cm, and several branches chopped at stem diam. up to 1.5 cm; however, in the case of D. lycioides, a few branches chopped at stem diam. 1 cm, but most of the plant left intact, showing that D. lycioides was less attractive than V. nilotica to the individual concerned of D. bicornis

Diospyros scabrida

  • several specimens heavily browsed


Euclea natalensis

  • specimen <1.5 m high on alluvial levee, so heavily browsed that <5% of foliage remains
  • specimen heavily browsed by S. strepsiceros, with a branch >1 m long broken by horns
  • several specimens 1 m high, heavily browsed (probably by S. strepsiceros), with only 10% of foliage left
  • specimen 2 m high, heavily browsed, only 5% of foliage present, and this has regenerated since last browsing
  • specimen 2.5 m high, with a 0.7-1 m long branches, of diam. 1-1.5 cm, grossly broken, probably by horns of male S. strepsiceros, with their thinner stems removed, leaving ragged ends. On same specimen, chopped stems, of diam. 0.7 cm, attributable to D. bicornis.
  • many specimens <1.5 m high, suppressed by herbivory, with signs including those of D. bicornis
  • many specimens with branches up to 1 m long, broken and severed by S. strepsiceros
  • specimen, formerly 2 m high, grossly damaged at 1.5 m high, by the horns of male S. strepsiceros; on the broken branches, only leaves and shoots eaten

Euclea racemosa (all plants seem unaffected by herbivory, with no shaping; Giraffa seems to ignore this species despite its heavy browsing of E. natalensis)

Euclea schimperi (apparently ignored by G. giraffa)

  • specimen showing browsing by D. bicornis (chopped stems 1 cm thick)
  • specimen 2.2 m high, heavily browsed by S. strepsiceros


Olea europaea cuspidata

  • suppressed specimen, with coppicing and epicormic growth in which the leaves are diminutive and ovate (heteroblastic), heavily browsed and with dead branches
  • suppressed specimen, 1.2 m high, heavily browsed and hedged
  • a stand consisting of many suppressed specimens, <1.5 m high, with hedged shapes
  • arborescent specimen 5 m high, with unmodified crown, but hedged coppice-growth (small leaves on densely divaricating stems) up to 1.5 m high, at base of bole


Pappea capensis

  • specimen formerly 4 m high, pushed over, broken (snapped bole of diam. >10 cm), and killed by D. bicornis, with many twig-size stems chopped and eaten


Dalbergia armata

  • specimen heavily browsed, hedged
  • ditto


Schotia brachypetala

  • mature specimen at least 12 m high, with browse-line corresponding to height of G. giraffa


Senegalia nigrescens

  • sapling with stem diam. 6 cm, scraped by tragelaphin horns (Leonard Gumede stated that S. strepsiceros, N. angasii, and T. sylvaticus are all known to scrape the bark, presumably to eat cambium)
  • sapling with bark stripped by L. africana
  • many mature specimens (trees) in full bloom but no foliage, with browse-line attributable to G. giraffa, below which no flowers remain


Vachellia karroo
This species occurs in the reserve ( However, I did not find it to be common here, and I did not notice any evidence of herbivory on it.

Vachellia nilotica ssp. kraussiana (see and and

  • partly pushed-over specimen, with stems 1.5 cm thick chopped by D. bicornis, indicating the wholesale eating of spinescent branches up to 60 cm long
  • in one area, many specimens, of former height at least 4 m, pushed over or broken down by D. bicornis; some specimens were still alive, by virtue of a remaining connection in the xylem
  • specimen formerly 3.5 m high, pushed over by D. bicornis, with bole (basal diam. 15 cm) snapped
  • several flat-topped specimens, formerly <5 m high, pushed over by D. bicornis, with chopped branches of diam. up to 2 cm
  • several specimens, formerly <3 m high, pushed over by D. bicornis, with most of the branches chopped off and eaten at stem diam. up to 5 cm

Vachellia robusta
This species seems to be unattractive to herbivores. I saw a mature specimen (tree, similar to but smaller than already in full foliage, apparently untouched by G. giraffa.

Vachellia tortilis

  • specimen formerly 3 m high, pushed over by D. bicornis; stem broken but crown still alive


A noteworthy difference in destructive foraging between Loxodonta africana and Diceros bicornis is that the former eats Cussonia spicata, but the latter does not (Leonard Gumede, pers. comm.).

Papio ursinus foraged destructively on Euphorbia ingens, to the point of facilitating Diceros bicornis by dropping fresh, edible litter consisting of partly-eaten sections of green stems. The discarded sections remain alive for some time, by virtue of the broken cross-sections sealing over.

In 2000, all the individuals I saw of Sclerocarya caffra were free of visible damage. My explanation is that L. africana was still underpopulated in the reserve at the time.

Searsia rehmanniana, which is even more abundant than Euclea natalensis in the reserve, seems to be ignored by S. strepsiceros (and G. giraffa). This is consistent with the fact that S. rehmanniana is as much associated with the Fynbos Biome as with subtropical savanna.

However, Searsia pallens ( contradicts the above pattern. It occurs mainly in the Fynbos Biome. Yet, it was apparently damaged by males of S. strepsiceros in Ithala Game Reserve, and I repeatedly observed G. giraffa foraging on it in August-September 2000.

The growth-form of V. nilotica in Ithala Game Reserve, in its maturity, is a sturdy but low tree, up to 4.5 m high with a flat crown up to 4.5 m wide, and a bole of diameter similar to the human waist.

Vachellia nilotica seemed to be associated mainly with parts of Ithala Game Reserve that had been ploughed for crops before 1973. However, this species was scarce in the lower-lying parts of the reserve.

In this area, the typical, flat-topped crown was lower than the maximum height of G. giraffa. However, I saw no evidence that the flatness of the crown was owing to shaping by herbivores. Instead, it seemed intrinsic to the growth-form of the subspecies indigenous to this area.

Vachellia nilotica seemed to be a staple of D. bicornis. However, I also observed G. giraffa foraging on it repeatedly.

Game-guard Leonard Gumede stated that, in Ithala Game Reserve, V. nilotica attracted foraging by both L. africana and D. bicornis. However, L. africana stripped the bark, whereas D. bicornis pushed over the small trees, then chopping off and eating branches.

As the plant grows, the flat shape of the crown of V. nilotica was attained when the plant was only 2 m high, with a basal diameter of only about 12 cm.

A local feature is that some individuals of Diospyros lycioides adopt a liane-like growth-form, climbing particularly into the crowns of V. nilotica. During my visit, I observed a specimen that had climbed to a height of >6 m, by means of this growth-form.

(This habit of D. lycioides, of climbing into the flat crown of Vachellia, has not been captured in any of the 807 observations currently posted in iNaturalist.)

Dichrostachys cinerea was, in Ithala Game Reserve, eaten by D. bicornis, but not preferred (Leonard Gumede, pers. comm.). At the time of my visit, D. cinerea (which differs phenologically from V. nilotica) had few green leaves.

Berchemia zeyheri is attractive to both L. africana and G. giraffa, but the former strips the bark, whereas the latter eats the foliage.

The fact that Elaeodendron transvaalense had a browse-line at only 2.2 m high indicates that this species is more attractive to S. strepsiceros than to G. giraffa. This seems to contradict the situation in the Maasai Mara National Reserve in Kenya, where Giraffa tippelskirchi produces a browse-line on Elaeodendron buchananii ( and and and

Both Euclea, consisting of relatively nutrient-poor 'evergreens', and Vachellia, consisting of relatively nutrient-rich 'deciduous' plants, vary in palatability according to the species.

For example, Euclea racemosa and Vachellia robusta seem to be ignored by herbivores, whereas Euclea natalensis and Vachellia nilotica are grossly damaged.

Euclea racemosa tends to occur as a small tree, up to 5 m high, rather than a multi-stemmed shrub. It tends to be untouched by herbivores, despite the obvious browsing on E. natalensis nearby.

The browsing looked heavier on Euclea natalensis than on Diospyros lycioides. Many specimens of the former were maintained as shrubs 2 m high, whereas the latter was free to climb up into the flat crown of e.g. V. nilotica, but tended not to be eaten even when V. nilotica was felled by D. bicornis.

During this visit to Ithala Game Reserve, I found that G. giraffa, S. strepsiceros, and D. bicornis all foraged on E. natalensis in ways that broke stems.

Euclea natalensis seems to be a local staple for both G. giraffa and S. strepsiceros, both of which are fully-populated in the reserve. It is not hedged by herbivory, but many individuals seem suppressed. Males of S. strepsiceros forage wastefully on this species, by breaking the branches.

Euclea natalensis is a staple of D. bicornis in Ithala Game Reserve, seemingly preferred over Dichrostachys cinerea. It is also heavily browsed by G. giraffa (including juveniles). On severaI occasions, I watched G. giraffa eating whole sprays of foliage, and even the dead leaves still attached to the stems after fire.

The following is a particularly detailed excerpt from my field notes from this visit to Ithala Game Reserve in late August-early September 2000:

"I watch an adult female individual of G. g. giraffa foraging wholesale on an upright, 5 m-high, rather spindly, tree of Euclea (probably natalensis), at the full height of the giraffe. The plant has fresh shoots (whitish, not reddish), but it is not these that the animal takes, particularly. Instead, it strips wholesale whole sprays of leaves (including new shoots in a minor percentage of mouthfuls), pausing to chew these very substantial mouthfuls before proceeding. It extends its tongue to the maximum (perhaps up to 20 cm from the upper lip) to hook around the bare part of a stem, forcefully drawing the branch to be grasped between the tongue and the upper lip, then it strips the whole bunch of leaves (perhaps about 20 mature leaves at a time) off and chews them, plus the smaller (still green) stems and small, growing leaves. On one occasion, the stem is snapped in the process, so that the animal takes the whole branch system in its mouth, chewing the leaves and then dropping the brown length of bared stem, about 22 cm long. With each occasion, the force of grabbing the branch tugs the whole crown jerkily, and with each stripping the branches snap back - an unusually forceful mode of foraging for giraffes. I watch as the animal concentrates on and mutilates this plant, as opposed to merely taking a few bites. It takes about six such sprays of foliage, after which I disturb it by approaching on foot. On closer examination, I find that the animal has significantly (?30%) defoliated the plant. There are many similar individuals of Euclea available. So, the victimisation seen on this occasion suggests that when the animal finds a 'sweet' individual, it indulges in 'bulk foraging' on it, at this time of year. Right next to the victimised plant is a tree of Berchemia zeyheri, so it is not that there are no alternatives available. Based on a week of opportunistic observation here, I would guess that, at this time of year, Euclea natalensis contributes about 10% of food mass consumed by the local population of G. g. giraffa, although requiring only <3% of foraging time. I.e. G. g. giraffa does not eat E. natalensis preferentially, and indeed it passes by most individuals of this species that it encounters; however, when it does eat it, it does so wholesale, by stripping leaves and whole shoot systems up to 30 cm long. The fact that I saw repeated extension of the tongue at this individual plant suggests that the animal had already stripped off all the easily available branches by the time I arrived."

Notes on diet of hook-lipped rhino (Diceros bicornis):

In Ithala Game Reserve, the genera Searsia, Gymnosporia, Diospyros, and Euclea are all abundant, and each contains several spp. here. In much of the reserve, these four genera collectively dominate the vegetation.

This is not ideal habitat for Diceros bicornis. However, it had (as at 2000) proven to be sufficient for a significant contribution to the conservation of this species.

During my visit, I learned of the following being eaten: Sanseviera hyacinthoides, Aloidendron barberae, Euphorbia tirucalli, Dalbergia armata, Dichrostachys cinerea, Vachellia nilotica, and Euclea natalensis. Diceros bicornis did not seem to eat V. nilotica and D. cinerea out of proportion to their abundance. However, their commonness in the reserve means that these mimosas were staples here.

Both D. bicornis and G. giraffa foraged on V. nilotica. However, only the former foraged on it destructively.

Diceros bicornis is capable of pushing over even mature trees of Euphorbia tirucalli (Leonard Gumede, pers. comm.). In 2000, the only specimens of E. tirucalli left in the reserve were trees about 4 m high, in protected situations, such as the edges of earth banks on the edges of drainage lines. Where the plants were protected by their topographical position, D. bicornis resorted to using its anterior horn to prise off the bark, which was presumably eaten.

Also see

Publicado el noviembre 1, 2022 10:52 TARDE por milewski milewski



According to, a browse-line is well-known in Berchemia discolor in southeastern Zimbabwe.

Please scroll to the 4th photo in

For a browse-line made by Giraffa giraffa giraffa on B. discolor, please scroll in

Publicado por milewski hace más de 1 año

The following observation provides further evidence that G. giraffa in Ithala Game Reserve persecutes/victimises certain individual plants of Euclea natalensis and Searsia pallens:

"Adult female individual of Giraffa giraffa, with small juvenile. The adult forages on two shrubs of Euclea natalensis (2.5 m high) by wholesale stripping. After stripping a large mouthful, it spends more than five, and up to 10, chews on it, before swallowing. These individual plants have previously been repeatedly stripped, and are reduced to a mere fraction of their former foliage. The juvenile nibbles for several minutes on a 2-m high shrub of Searsia pallens, also heavily browsed, with new shoots but only a fraction of its full foliage. I go over to examine the freshly-stripped stems. On one stem of E. natalensis, the woody skeleton (diam. 0.4 cm) is exposed for more than 15 cm."

Publicado por milewski hace más de 1 año

Gymnosporia buxifolia, showing maximum spinescence elsewhere:

Publicado por milewski hace más de 1 año
Publicado por milewski hace más de 1 año

Fascinating, thanks for posting all these Ithala records.

Publicado por jeremygilmore hace más de 1 año

@jeremygilmore You're most welcome.

Publicado por milewski hace más de 1 año

Diospyros lycioides is odd, in its genus and family, in featuring both liane-like laxness and a capacity for semi-spinescence.

Publicado por milewski hace más de 1 año

Diceros bicornis ate euphorbia mainly in the dry season, and relied on it for water in drought, despite the provision of an artificial waterhole. This megherbivore generally kills the individuals of Euphorbia ingens at which it forages. Diceros bicornis pushes over the plant, then eats some of the fallen stems; the plant subsequently dies, and rapidly decays.

This paper contains records of Papio eating euphorbia. However, Loxodonta africana (in contrast to D. bicornis) does not forage on Euphorbia ingens.

Diceros bicornis also eats (only in the dry season) the basal stems of the latex-rich liane, Fockea multiflora ( and and and and

A noteworthy finding, mentioned in this paper, is that this waterhole was shunned by Hippotragus niger (and to a lesser extent Aepyceros melampus and Kobus ellipsiprymnus), to the point of death by drought.

Publicado por milewski hace más de 1 año

Hitchins P M (1968) Records of plants eaten by mammals in the Hluhluwe Game Reserve, Zululand. The Lammergeyer 8: 31-39.

DICEROS BICORNIS recorded eating the following (among others):

Ammocharis coranica leaves (
Aspilia mossambicensis leaves
Solanum campylacanthum stems, leaves, and fruits (
Searsia chirindensis, Searsia dentata and Searsia pentheri leaves and branches
Mystroxylon aethiopicum schlechteri leaves and branches (
Gymnosporia spp. leaves and branches
Putterlickia sp. leaves and branches (
Tarchonanthus (possibly trilobus and/or parvicapitulatus) stems and leaves ( and
Diospyros lycioides branches and leaves (
Euclea crispa branches and leaves
Euclea divinorum branches and leaves
Euclea schimperi branches and leaves
Asparagus sp. stems and leaves
Canthium inerme stems and leaves
Grewia occidentalis stems and leaves (
Rhoicissus tomentosa and Rhoicissus tridentata stems and leaves ( and
Ozoroa spp. branches and leaves
Apodytes dimidiata branches and leaves

Nyala angasii and Strepsiceros strepsiceros recorded eating (among others) Sideroxylon inerme

Publicado por milewski hace más de 1 año

Agregar un comentario

Acceder o Crear una cuenta para agregar comentarios.