Journal archives for November 2023

02 November, 2023

How adder-like is Echiopsis curta?

Comparison of snakes in similar ecosystems in southwestern Australia and the southwestern Cape of South Africa.

The Barrensncf Caledon coast.

Echiopsis curta ( is an elapid, related to Notechis. Itnis the closest counterpart for the viperid, Bitis armata (

Both are southern endemics within their continent, temperate, oligotrophic, heathland, and commonn in the ecosystems studied.

They have the same size and shape: mean length about 40 cm, maximum length 60 cm.

Their head sizes and shapes are similar, and they share vertical slit pupils, pale upper irises, and oblique whitish markings on the upper lips or jowls.

Neither has a caudal lure.

Both are slightly polychromatic, and rely mainly on camouflage.

Both are venomous, sufficient to kill a small mouse within 5 minutes, and front-fanged, and hunt by ambush on the ground without submergence.

Both are fairly catholic in diet (amphibians, lizards, and some small mammals and even birds), with a bias for frogs (Shine).

Both hang on to small prey, but release large prey and follow it to its place of expiry, being incapable of rapid locomotion or constriction.

Both are viviparous, bearing 4-9 (maximum 14-15) offspring in late summer or autumn.

Both have neonates 11-14.5? cm (check for E. curta).

Both emerge in dim light and under cool conditions.

The differences between them are their defences against their own predators.

Echiopsis curta is adder-like in being unusually irascible if provoked, with relatively ling fangs? and potent venom for an Australian elapid of its size, and to this extent is convergent with Bitis in much the same way as the Australkan elapid Acanthophis (Shine).

However, E. curta is far less elaborately camouflaged than B. armata, has a much less indurated skin. The adder has hard, keeled scales, despite having more, not fewer, scale rows than its Australian counterpart.

Echiopsis curta hardly hisses, even when touched (check), whereas B. armata makes anninsistent,mpuffing sound when a potential predatorsddraws near, and before contact is made.

The exact difference in fang length, position, and mobility, and venom-delkverynsystems, are unknown.

This seems to be a classic case if two Good counterparts in habitat, morphology, and foraging ecology, and reproductive behaviour.

They differ simply because of the very different incidences of vertebrate predators in their habitats.

Note that the response of these forms is in terms of greater protectiveness of the adult, not greater numbers of offspring per birth.

A second adder-like elapid occurs at the Barrens, albeit rarely: Acanthophis antarcticus ( and and This more resembles Bitis cornuta, whixh does not reach the Caledon coast.

The theme of lesser defensiveness and greater body size (without lesser litter size) in Australia than southern Africa is epitomised by Notechis scutatus (the Barrens) and Hemachatus haemachatus (Caledon coast).

Both are substantial ()> 1 m long and moderately stocky), viviparous elapids, incapable of rapid pursuit of prey, or rapid escape from their own predators.

Both N scutatus and H. haemachatus have anbhntidy banded pattern, as suited to warnjng an jntruder if a seriously venomojs snake as to camouflage. Both have pale, bright ventrolateral surfaces, conspicuous when the snake moves in alarm.

They often eat large, slow-moving amphibians, while accepting a wide variety of vertebrate prey, and are neither particularlynsecretive bybdaylight nor strictly nocturnal.

Their moist, grassy/sedgey, relatively productive environments support relatively many and diverse predators willing to tackle a cinspi uous, medium-size snake.

These factors may explain why these snakes are among the most dramatically defensive reptiles at the Barrens and the Caledon coast.

Posted on 02 November, 2023 10:26 by milewski milewski | 3 comments | Leave a comment

05 November, 2023

Rock-dwelling agamids on two continents: Ctenophorus vs Agama, part 3: variation in masculine colouration

...continued from

ACKNOWLEDGMENT: I thank Johannes van Rooyen (@johannesvanrooyen ) for educating me about the differences between males and females, in Agama ( The patterns he has pointed out are, in hindsight, now evident to me. However, I was formerly oblivious to them, and field guide-books do not do justice to them, either.

Rock-dwelling agamids in Australia and southern Africa vary extremely in their degree of sexual dimorphism in colouration.

In this respect, they show no evolutionary convergence.

The variation goes along four lines, viz.

  • mature males vary from dull through bright-hued to gaudy, according to the species,
  • females (irrespective of breeding condition) are somewhat bright-hued in Agama planiceps, vs dull in all other spp.,
  • females have bright-hues associated with breeding condition in several spp. of Agama, but no other spp., and
  • mature males of Ctenophorus ornatus, which lack bright hues, differ from females only in the boldness of dark/pale contrast (banding on the tail, plus vertebral stripe).

Further details include the following.

Among mature males, the dullest is Ctenophorus rufescens, whereas the gaudiest are Agama kirkii and A. planiceps. This corresponds, approximately, to a difference between non-gregarious and gregarious, with corresponding polygyny.

Agama anchietae is odd in that it alone

  • among all the spp. studied, has colouration more conspicuous (at least to the human eye) in females than in males, and
  • among the southern African spp., has masculine colouration less conspicuous than that of Australian species (particularly Ctenophorus vadnappa).

This possibly corresponds to A. anchietae being less gregarious than its rock-dwelling congeners in southern Africa, in this way partly resembling the Australian spp. However, the bright hues in females of A. anchietae, in breeding condition, undermine this explanation.

A categorical difference emerging from this study is that females feature conspicuous hues (to the human eye) in no Australian species but all southern African spp. Given that females are the biologically central sex, this degree of intercontinental divergence is surprising indeed.








scroll to third photo in


Fairly typical:

Extensive blue/turquoise:

With prominent pale vertebral stripe:

With yellowish tail and maroon abdomen:



Posted on 05 November, 2023 11:02 by milewski milewski | 11 comments | Leave a comment

10 November, 2023

A photo-guide to the bewilderingly complex colouration of the South African rock-dwelling agamid, Agama atra, part 1

@karooicus @botswanabugs @dinofelis @grant_reed_botswana @chris_whitehouse @joctober @bluehillescape @ryanvanhuyssteen @tom-kirschey @b_akeret @ivanparr @tonyrebelo @jeremygilmore @alexanderr @tyroneping @herping_with_berks @herping_with_sean @peter_erb @m_burger @asimakis_patitsas @max_tibby @grahamarmstrong @rfoster @snakesrcool @calebcam @colin25 @cthawley @agama-girl @johannesvanrooyen @ludwig_muller @karoopixie @alexdreyer @i_c_riddell @ptexis @cr_hundermark @dhfischer @wesselpretorius @richardgill @felix_riegel @vynbos @shauns @riana60

Please see,to%20how%20humans%20become%20tan.

The southern rock agama (Agama atra) presents an unusual opportunity - and a considerable challenge - for an illustrated Post.

This is because it combines

  • a profusion of observations in iNaturalist, and
  • bewildering variation in appearance, particularly colouration.

Chameleons ( are the Iguania ( taken to epitomise the versatility of colouration in lizards. However, an argument can be made that the southern rock agama - belonging to the same suborder - is an equally good example of similar phenomena.

The southern rock agama can be

More broadly, scrolling through the thousands of photos in iNaturalist gives a kaleidoscopic impression, rather than one of a single species.


There is no particular problem identifying Agama atra, because it is the only rock-dwelling agamid over most of South Africa.

However, it is another matter to understand why this species has such an inconsistent appearance.

So, my aim in this Post is to tease apart various factors contributing to the confusing colouration of this lizard.


I have carefully chosen illustrations from the innumerable available on the Web, in an attempt to disentangle the various factors in the colouration.

I restrict the coverage to

  • the dorsal and lateral surfaces of the figure (as opposed to the usually hidden ventral surface of the torso), and
  • hues visible to the human eye (as opposed to ultraviolet).


There seem to be six main factors combining to determine the appearance of the southern rock agama, viz.

  • urgent thermoregulation,
  • growth and development from juvenile ( to adulthood and full (social) maturity,
  • sex,
  • individuality,
  • breeding condition, and
  • mood/emotion (possibly social/sexual excitement but certainly fear, when approached by a potential predator).

These six factors result, respectively, in the following approximate results:

  • conspicuous but temporary darkness vs pallor, presumably according to temperatures,
  • camouflage, particularly in juveniles (,
  • different patterns of colouration in females and males, in complex ways involving both hues and dark/pale contrast,
  • individual/sexual variation in disruptive mottling (camouflage-colouration,, pale vertebral striping, and hues (particularly on the hindquarters and tail in mature males)
  • blue/turquoise heads as a standard in both sexes, plus sex-specific advertisement-patterns elsewhere on the figure, during courtship and territorial defence (masculine) or mating and pregnancy (feminine), and
  • dimming of bright hues (particularly blue/turquoise on the head), within a timespan of several seconds, in apprehension of potential predators.


The whole figure is darkened, to the point of conspicuousness.

The following ( may possibly be a case of darkening while basking, but it is ambivalent because of masculine advertisement (note that the vertebral stripe remains strikingly pale).

Conversely, the whole figure can show pallor to the point of conspicuousness, presumably in reaction to overheating.



Adult females:

Adult males:



Juveniles (sex unknown):

Adult females (most camouflaged when not in breeding condition):

Adult males (camouflaged only when not in breeding condition, and generally less likely than females to be camouflaged):

INDIVIDUAL/TEMPORARY VARIATION IN MOTTLING (with a sexual difference that hypothetically begins in small juveniles)

Juvenile females?:

Juvenile males?:

The following individual adult males ( and are unusual in retaining mottling on the torso, at an age/stage when the head is already capable of full expression of blue.

Does the pattern of mottling on these juvenile individuals ( and indicate female?

extending on to tail and


Females in non-breeding condition:

Males in non-breeding condition:

The following show that, already in juvenile males,


Also please see other recent Posts, including


The following ( shows that feminine colouration (and presumably sexual maturity) arrives in what seems to be large juveniles, long before complete body size is attained.


Also see

Blue: and

Turquoise :

Please see

The following ( and and show that bluish hues start to appear in juveniles.

Perhaps the part of the body most individually variable in colouration in masculinity is the tail (


Dimming of bright hues:



to be continued in

Posted on 10 November, 2023 05:12 by milewski milewski | 55 comments | Leave a comment

15 November, 2023

A photo-guide to the bewilderingly complex colouration of the South African rock-dwelling agamid, Agama atra, part 2: Discussion

...continued from



There is a puzzle w.r.t. pallor, in the photos shown above. This is that the dates and times of day do not suggest high temperatures.

Inferring age and stage of development from photos:

The stage of development, from infancy ( and and and to maturity, is apparent from the proportional size of the head.

In full maturity, females differ from males in having

The sexual difference in the proportional (and absolute) size of the head is shown, within a single photo, in

Infants and small juveniles have camouflage-colouration ( They are conspicuous only when they darken on a pale background (to bask), or show pallor on a dark background (presumably to cool down).

Even adult males remain capable of full camouflage, at least when not in breeding condition (

In large juveniles of both sexes, the head and forelegs start to turn blue.

There is a sexual difference in the mottled pattern on the torso, which - although subtle - is so basic that it seems to appear just after infancy. In females of all ages, the mottling tends

By contrast, in males,

  • the pale vertebral stripe tends to override mottling on the mid-dorsal line,
  • there is no parenthesis-like coalescence of the mottling on the torso,
  • mottling tends to become converted to faint pale spotting (, and
  • all mottling/spotting tends to become obscure with age and breeding condition.

In partial summary, adult females have a distinctive pattern of coalesced mottling on the torso ( and This pattern is not seen in males, which instead tend to feature a pale vertebral stripe (

This difference transcends the conspicuous colouration of breeding condition.

Yellow on the torso ( and is exclusive to females. However, it is not a reliable clue, because it

The following individual adult female ( is unusual in having

  • a distinct pale vertebral stripe, and
  • extensive bluish hue.

A nuance of colouration in the southern rock agama:
In adult females in breeding condition, the yellowish hue on the abdomen ( and is less subject to dimming than is the bluish hue on the head.

This makes sense, because the yellowish has an anatomical position more easily hidden by posture and perspective ( and

Ultimately, the most reliable clues distinguishing female from male, and immature from mature, are subtle. In the following view of an individual adult female (, the best clue to sex is the pattern of reddish-brown mottling on the torso, and the best clue to maturity is the size of the head relative to the rest of the figure.

No subspecies seem to be recognised in the southern rock agama. Populations in Namaqualand and Namibia are distinctive in their large body size and continual, as opposed to seasonal, reproduction ( and

Is it possible that this species is colour-polymorphic w.r.t. masculine colouration?

What I have noticed is as follows:

The first hypothetical colour-morph lacks all hues other than blue/turquoise, and is conspicuous partly by means of dark/pale contrast, with a dark torso and tail, offset by a pale vertebral stripe ( Some individuals in this morph become suffused with blue/turquoise over the whole figure (

The second hypothetical colour-morph features hues other than blue/turquoise ( and and and and

I refer particularly to yellowish on the tail ( and and and and and reddish on the abdomen and anterior surface of the upper hindleg (

This morph lacks dark-pale contrast, and the yellowish on the tail is precocial (

In preparation of this Post, I thrice-examined each of the 5,300 observations (>6,000 photos) of the southern rock agama in iNaturalist, learning more with each of my three consecutive bouts of scrutiny.

After this 10-day effort, much remains obscure/confusing to me. For example, both sexes, plus well-grown juveniles, can feature blue/bluish on the head, while at the same time this hue is inconsistent in all these categories.

One of the few clear findings is that yellowish on the torso is completely diagnostic of femininity ( in this species of lizard. However, even this falls short of being categorical, because yellowish is not apparent in some views of full feminine colouration (

Posted on 15 November, 2023 01:12 by milewski milewski | 12 comments | Leave a comment

17 November, 2023

Why are rock-dwelling agamas absent from Kruger National Park, South Africa?

@tonyrebelo @jeremygilmore @m_burger @johannesvanrooyen @alexanderr @tyroneping @herping_with_berks @herping_with_sean @richardgill @ryanvanhuyssteen @ruan_stander @tom-kirschey @simontonge @ockertvs @rion_c @peter_erb @calebcam @botswanabugs

Rock-dwelling spp. of Agama are widespread and common in Africa, from the Mediterranean coast in the north to Cape Agulhas at the southern tip of the continent.

In the African tropics, several spp. of rock-dwelling Agama may be sympatric. For example, in the Serengeti ecosystem of Tanzania and Kenya, Agama mwanzae (, Agama lionotus ( and Agama dodomae ( all occur.

Kruger National Park ( in South Africa contains various forms of rocky terrain, particularly

Kruger National Park is the best-known part of a larger area, namely the Great Limpopo Transfrontier Park ( This includes an additional category of rocky terrain, namely the Chilojo Cliffs ( and and in Gonarhezou National Park ( in southeastern Zimbabwe (

Elsewhere in southern Africa, rocky terrain is inhabited by various spp. of Agama, namely

These spp. represent three distinct clades within the large and diverse genus (

Some of the above spp. occur sympatrically in Namibia and Northern Cape province of South Africa, so that a given rocky outcrop can be inhabited by two congeners.

Therefore, I find it remarkable that no rock-dwelling agamid occurs in Kruger National Park or elsewhere in Great Limpopo Transfrontier Park.

Extending the puzzle:
An agamid similar to Agama, namely Acanthocercus atricollis (, is abundant in Kruger National Park. This species inhabits trees and fallen logs, rather than rocks.

It is odd that A. atricollis does not extend its habitat on to rocky surfaces in Great Limpopo Transfrontier Park, because

What emerges is a biogeographical and ecological anomaly: what seems to be a gap in Nature, and an 'empty niche'.

Compounding this puzzle are other, ecologically-related lizards.

I refer to genus Cordylus (, which belongs to a different family, viz. Cordylidae.

Elsewhere in South Africa, rocky terrain is co-inhabited by Agama and Cordylus (

This raises the possibility that Cordylus might have usurped the habitat and niche of Agama in Great Limpopo Transfrontier Park.

However this seems not to be the case.

One species of Cordylus ( has been recorded in Kruger National Park. However, as in the case of A. atricollis, it is restricted to trees, not extending to rocks.

A species in a closely-related genus ( is associated with the southern part of the Lebombo range (which extends to Mkuze in Zululand), and is rock-dwelling. However, it seems to be scarce, and has not been recorded in most of Kruger National Park.

Two rock-dwelling spp. of another, mainly tropical, genus in the same family occur in the relevant ecosystems. I refer to

However, their occurrence hardly explains the absence of rock-dwelling agamids. This is because

Can any reader explain why rock-dwelling agamids, despite being so widespread, common, and phylogenetically diverse in Africa, have failed to utilise what seems to be suitable terrain in and around Kruger National Park?

Posted on 17 November, 2023 18:34 by milewski milewski | 10 comments | Leave a comment

23 November, 2023

Bold black-and-white in females of the southern sable antelope (Hippotragus niger niger): anti-predator warning or maternal emulation of masculinity?

For naturalists, a central puzzle of the southern sable antelope (Hippotragus niger niger) is why its fully mature females combine

  • thoroughly conspicuous colouration (on face as well as rest of figure) with
  • lethal-looking horns (sharp-pointed, and straight enough to be deployed forwards, but curved enough to be deployed against a predator on its back).

The sable antelope is one of the few spp. of ruminants - a clade that comprises hundreds of spp. on various continents - that shows this combination.

(Two caveats: this applies only to the nominate subspecies of the sable antelope, and to fully mature females. In this Post I refer particularly to the subspecies Hippotragus niger niger, and to individual females older than eight years ( However, for conciseness I will just call these 'sable antelope' and 'females'.)

At first glance, the pattern in question suggests aposematism (, which by definition is warning colouration vs predators.

However, aposematic colouration seems far-fetched in ungulates, which rely on vigilance and fleeing as their main anti-predatory strategy.

Furthermore, this explanation would fail scrutiny in the following ways:

  • there is limited evidence that the sable antelope is apt to defend itself from predators by means of its horns, thus undermining the explanatory power of aposematism;
  • aposematism typically applies to non-apparent anti-predator defences (e.g. the venom-glands of skunks), whereas the horns of antelopes are fully apparent;
  • the colouration and horns of females emulate those of males - in which the configuration of horns and colouration are more parsimoniously explained intraspecifically, w.r.t. masculine rivalry; and
  • the sable antelope does not show particularly strong maternal protection of infants and small juveniles, which tend to lag behind their mothers at a most vulnerable time of their lives.

A subsidiary puzzle, which overlaps in the conceptual framework with the main puzzle stated above, is as follows.

There are many bovids ( in which females resemble males, in possessing horns, in being maned and/or bearded, and in having similar colouration between the sexes.

However, in most of these, males are not immediately apparent as such, i.e. there is minimal sexual dimorphism in body size and horn size and form.

In the sable antelope, mature males differ enough from females to be easily recognisable. This is because mature males have

The implication is that the appearance and armaments of females may be emulations of masculinity. This seems as plausible as any explanation invoking anti-predator adaptations.

According to Richard Estes (, the emulation of male appearance by female bovids may be explained by the fact that there is sexual non-segregation, throughout the year, in the spp. concerned. Territorial males would tend to ostracise male offspring at the adolescent stage. However, it is in the mothers' interests to keep their adolescent sons in the protection of the group for as long as possible.

Hence - by implication from Estes' hypothesis - females have evolved to match mature males in armaments (horns) and masculine appearance, so that they can protect the adolescents from ostracism for as long as possible.

Estes' explanation seems satisfactory for most lineages of 'plains game' in which there is so little sexual dimorphism that males can be hard to distinguish from females in the field. I refer particularly to wildebeests ( and other alcelaphins, as well as oryxes (

However, this does not fully explain the odd combination of features in the sable antelope. What remains to be explained are that

  • sexual dimorphism remains great enough that, to the human eye, mature males are easily distinguished (by horn length and the black penis-marker) from mature females,
  • masculine colouration is boldly black-and-white,
  • the colouration of mature females matches the masculine boldness,
  • the horns of females are more credibly intimidating than those of alcelaphins or even oryxes, and
  • there is a particular pattern of what looks like 'badger-like warning colouration' on the face.

The anomalies shown by the sable antelope may possibly be explained by its particular habitat and foraging niche, which put a particular premium in maternal defence of adolescent sons.

The sable antelope, unlike most alcelaphins, inhabits relatively nutrient-poor savannas, in which it depends on small patches of relatively nutrient-rich soil and grasses grown to a certain height (about 0.25-1 m). It is well-known that the sable antelope is associated with edges, or 'ecotones' at various scales, rather than homogeneous vegetation.

The patchiness of foraging by the sable antelope is particularly owing to a combination of

  • fires, which occur in the dry season and tend to combust the grass once every few years in the habitat of the sable antelope,
  • interspersion of woodland with drainage lines called 'dambos' (, which tend to have clay-rich soil and to be free of trees, and on which the grasses tend to be fairly palatable, and
  • large mounds of termites, which tend to fairly nutrient-rich and exempt from the fires that sweep the matrix among the mounds, and provide particularly palatable grass in limited quantities.

What this means is that in the sable antelope, unlike most 'plains game', there may be intense competition among the sex/age classes for crucial food, localised in a generally unpalatable type of vegetation. The best grasses would tend to be taken by the territorial male, and adolescent males in particular would tend to marginalised by the masculine aggression of the territorial male.

Estes and Estes (1969, The Shimba Hills sable population. First progress report), referring to Hippotragus niger roosevelti, state on page 13:

"Sub-adult males [2-3 years old] are often found in nursery herds...Adult sable bulls thus show more tolerance than many other territorial antelopes; for instance male wildebeest are ejected from the herd as yearlings...This tolerance is promoted by submissive (= female) behaviour on the part of young males...The tolerance of males within the nursery herds until a relatively advanced age may be adapted to the frequently low population density in this species. Bachelor herds into which young males might go when evicted from the nursery herds often do not exist, which makes young males that much harder to drive from the nursery herd."

Estes and Estes, The sable in Rhodesia. Second progress report), referring to Hippotragus niger niger, state on page 12:

"Nursery herds of H. n. niger usually include one adult bull. Although more than one adult male has frequently been recorded by different observers, this is actually an unusual and temporary most cases the supernumerary adult males are simply sub-adults. Males are tolerated in the nursery herds up to around the age of three, by which time H. n. niger bulls are often black except for their hindlegs."

Kingdon's (2015, interpretation has stood the test of time. On page 209, he states:

"Males are driven out into 'bachelor' groups at about 3 years...sable antelopes are secondarily territorial. One indication of their nomadic, open-country origins is the male's reliance on visual self-advertisement. Unlike other woodland and forest antelopes scent-marks are subordinate to the male sable's posturing's and direct herding of the females. Black colouring is both the mark of super-seniority in the colour-coded hierarchy of the female herd and also the central beacon of a defended territory. Battered bushes, dung piles and foot-scrapes in the centre of a 3 or 4 [kilometre-squared] territory may help deter other males but it is the imposition of male physical presence that dominates both hierarchy and territory."

Posted on 23 November, 2023 05:56 by milewski milewski | 17 comments | Leave a comment

24 November, 2023

The morphlings versus the axolotls (how frogs have warped tadpoles into new shapes and sizes), part 2

...continued from


The best examples of morphlings are to be found in true toads (Bufonidae), and particularly the largest species of toads. While some of the African and Asian toads are fairly large as frogs go, the true giants of the bufonid family occur in the Neotropics. Below I show the distribution ranges of two of the largest spp., namely Rhinella marina and R. diptycha. There is some overlap between these two species in the southeastern Amazon basin, but essentially the former is found to the north and the latter to the south. Rhinella marina extends from southernmost Texas all the way to the Amazon. Rhinella diptycha takes over in the caatinga, cerrado, chaco, Atlantic forest, and Pantanal, reaching northern Argentina although it does not extend as far south as the Pampas.

What this means is that, collectively, these two spp. of large toads cover most of South and central America. Even if morphlings were restricted to just these two spp. and no other frogs, their existence would be noteworthy, not so?

Rhinella marina:

Rhinella diptycha:

See 'Cannibalism is Common' in

This was written by Rick Shine or his colleagues, and explains how cannibalistic the large toad Rhinella marina (introduced into Australia and now a major pest) can be. As he points out, it’s not the fully mature individuals that are most cannibalistic, it’s the young adults. These young adults semi-specialise on eating the morphlings of their own species (which Rick Shine of course does not call morphlings, calling them ‘metamorphs’ instead. But the point is that the life history strategy of this large species of true toad, which is presumably typical of Bufonidae generally, involves a level of cannibalism that is systematic rather than being an occasional aberration. Each individual of this species has to survive a veritable gauntlet of cannibalism during the course of its life, in the morphling stage which is like a second infancy after metamorphosis.

(See my other Post about cannibalism in amphibians,


I have been unsure whether the incidence of morphlings is essentially a scaling phenomenon within Bufonidae, or a phenomenon that will remain after corrections for fully mature body size. I’m leaning towards the latter based on a small species of North American toad, namely Anaxyrus debilis.

Anaxyus debilis ( is so small that I would not expect it to have morphlings. However, its metamorphs are larger than I expected, and far larger than those of enormous toads such as Rhinella marina. This makes it clear that A. debilis falls into a different pattern of development, as opposed to just being a 'scaled-down cane toad' as it were, in which the morphlings cease to be remarkably small relative to the adults.

Breeding females of A. debilis have snout-vent length 4.6-5.4 cm, while metamorphs have snout-vent lengths 1.9-2 cm. Please bear in mind that the morphlings of R. marina have snout-vent ca 1 cm. Having metamorphs even smaller than in R. marina is what one would expect from the small species A. debilis if the two toads shared a pattern in common. Instead, the truth is that in A. debilis, compared with R. marina, the adults are smaller and the metamorphs are larger. So the pattern is broken. This suggests that morphlings occur in some, but certainly not all, bufonids. It’s still possible that morphlings occur in all large bufonids, but if so it won’t be just because they grow large.

Adding to the pile of sometimes inconsistent information on the actual sizes of adults and morphlings in true toads (Bufonidae):

I have before me a source that states that in Rhinella marina, the mature toads are male 14 cm and body mass 1 kg, and female up to about 23 cm and up to 1.5 kg. The morphlings are as follows: “In spite of the enormous size of its parents, a newly-formed cane toad is no more than 6-7 mm long.” This refers no doubt to snout-vent length.

Adults of Phrynobatrachus (, with snout-vent length about 2 cm, have body mass ca 0.5 g. I deduce that the body mass of the morphlings of R. marina is usually 0.1 g or less.

Putting the idea behind the invention of the word ‘morphlings’ as simply as possible:

Bufo bufo, the typical true toad, has a mature female body mass ca 100 g. The following value needs checking, but, assuming that the freshly metamorphosed toad has body mass ca 0.1 g (= ten to a paper clip!), these ‘babies’ are only 0.1% of maternal body mass. Yet these metamorphs are widely and unquestioningly called ‘adults’.

The central problem: how can a toad possibly be called ‘adult’ at only 0.1% of mature body mass? Fact is, true toads have BOTH LARVAE AND INFANTS, and ‘morphling’ is the name I suggest for these infants.

A problem with 'toadlet' is that morphlings are not peculiar to toads, and toads are not clearly defined anyway.

The morphlings are, remarkably, proportionately similar to the fully mature stage, i.e. morphlings are as TINY relative to the fully mature stage as infants would be, but are shaped like mature animals, not like infants.

Putting this another way: in true toads (particularly the largest spp.), the life history is divided into two completely different processes. During the larval stage, there is little change in body size but immense change in body form. After metamorphosis, the situation is quite different: there is immense change in body size but minimal change in body form. Of course this is the basic pattern in amphibians in general, but there is extreme polarisation in bufonids.

The trouble with the word ‘infant’ in this context is that it is too vague. Infant can mean any kind of baby. Instead of increasing understanding by boosting precision, it detracts from understanding by boosting vagueness. Putting this another way, imagine a scientific tradition in which frog tadpoles (which are particularly different from adults, even relative to salamander larvae) had no particular name but were just called ‘larvae’ or, worse still, ‘juveniles’. There’s absolutely nothing incorrect about calling frog tadpoles juveniles, because that is indeed what they are. The trouble is that, while correct, this is too vague.

I only discovered at age 63, after a lifetime of particular interest in and study of frogs, that true toads have exceptionally tiny ‘adults’, because this fact, although known for many centuries, has been hidden by the lack of an apt term.

Another problem with calling these juveniles of true toads ‘infants’ is that this would introduce unnecessary confusion between the ‘adult’ infants and the larval infants. If the morphling is really an infant, then why is the tadpole not also an infant? For example, a kangaroo neonate is just as much an infant as a zebra neonate, obscuring the enormous difference in degrees of development between extremely altricial and extremely precocial neonates among mammals. If one said ‘frog infant’ to most people, I suspect that they would imagine tadpoles. For that matter, why don’t we call foetuses in mammals ‘infants’? In Science, the more precise and specific the word used to describe something, the better.

It seems basic to the definition of morphlings that they would belong mainly, or only, to large species of frogs. This is because what ‘morphling’ describes is a ‘second babyhood’ after metamorphosis, which makes most sense where the eventual fully mature body size is far greater than that at metamorphosis. Of course, fully mature body size is not the only operative variable; also important is the maximum size of the tadpoles. For example, the huge toad Rhinella marina has both large mature body size and extremely small tadpoles at full development of its larval stage.

Based on this thinking, it seems sure that another example of a frog lineage with morphlings is the Conrauidae of Africa. This family contains the largest of all frogs, Conraua goliath (, which can reach 3.6 kg. A frog that large seems likely to qualify for morphlings just on mature body size alone, but as it happens C. goliath also qualifies in terms of its larva: the tadpoles are of unremarkable size compared with other frogs, reaching only 5 cm long before metamorphosing. Those would be large tadpoles for a bufonid, but they are more or less the same size as those of large ranids.

For comparison, the largest frog in North America apart from Rhinella marina is Lithobates catesbeianus (Ranidae), which reaches a maximum body mass of 0.8 kg and has tadpoles 5-7.5 cm long and up to 18 cm long. The African and North American giant frogs are directly comparable because both are among the more aquatic of frogs worldwide.

The West African giant has tadpoles less than a third the length of the North American giant despite having fully mature mass four-fold greater. Morphlings for sure?

This is rather nice, because what it would mean is that both the largest aquatic frog on Earth (C. goliath) and the largest terrestrial frog (R. marina) on Earth have morphlings.


To North Americans, leopard frogs (Ranidae: Lithobates pipiens and related spp.) are bog-standard frogs, similar to the closest thing to a bog-standard frog in the southwestern Cape of South Africa, namely Amietia fuscigula (Pyxicephalidae). The mature frogs are >10 cm snout-vent length, and the freshly metamorphosed juveniles have a snout-vent length of about 2.5 cm, which is a quarter of the mature dimension. The juveniles grow for a further three years before reaching sexual maturity, and then after that keep growing to some extent to full maturity.

These relative sizes are illustrated below. It may help to know the dimensions of my hand: 10 cm wide at the palm, with the last section (phalanx) of middle finger being about 2.5 cm long.

As you can see, the freshly metamorphosed juvenile frogs are not particularly small relative to the human hand. They are about 2.5-fold longer than the morphlings of toads of comparable mature body size, which means that they presumably weigh an order of magnitude more than morphling toads, likewise freshly metamorphosed from the tadpole stage.

The following photo shows the mature size of Lithobates pipiens or a closely related species. If the palm is 10 cm wide, you can see that this frog exceeds 10 cm in snout-vent length.

The following is another photo of the same type of frog, again showing similar mature dimensions.

I’m not sure that the following juvenile is freshly metamorphosed, but it must be close. As you can see, its snout-vent length is about one inch. It’s certainly a ‘baby’, but not nearly as diminutive as morphling toads (Bufonidae). These relative sizes are quite ordinary for juvenile vs mature vertebrates and there’s no need for a special term for these juveniles. My guess is that, although the fully mature stage is comparable in body mass between leopard frogs and toads, the freshly metamorphosed stage is an order of magnitude different in body mass. That’s why I feel that the word ‘morphling’ is useful for the small juveniles. The morphling toad would weigh about as much as the visible section of upper hindleg of the frog in the photo below. following photo shows a handful of the freshly metamorphosed juveniles. Again, as you can see the body size is about one inch long snout-vent, with a body volume similar to that of the last phalanx of the middle finger of a man. Considering that a small standard paper clip weighs about 1 g, it seems safe to assume that the distal phalanx of the middle finger weighs more than a gram, and that in turn the freshly metamorphosed frog also weighs somewhat more than 1 g. Compare this with the minimum size of morphlings in that paper by Shine and co-authors, in which I remarked that it took 20 morphlings to weigh as much as a paper clip.


Perhaps the most famous of all frogs w.r.t. paternal (fatherly) care, namely Darwin’s frog (Rhinodermatidae: Rhinoderma darwinii, may have been hiding in plain sight as an example of what we call the ‘morphling’ phenomenon.

Nobody with any broader knowledge of frogs can have failed to hear of this species of frog, because the male does something so bizarre, with no parallel in any other animal: it ‘gestates’ the tadpoles in its vocal sac.

By the way, the description ‘vocal sacs’ is rather misleading in the same sense as ‘cheek-pouch’ is misleading for the extensive compartments into which certain hamsters stuff food while foraging. The sacs in question, in R. darwinii, extend from the throat all the way on the ventral surface of the male frog, to the groin and on the flanks almost to the back. Entrance to this modified and extended vocal sac, which creates a space between the skin and the muscles of the body, is gained through a pair of slits inside the mouth. What we’re talking about is a huge and newly-invented cavity, effectively sealed off, into which offspring can be inserted for the purposes of parental care. And from which there is a process of ‘giving birth’ because of the sphincters involved.

The male (snout-vent length 2.2-2.8 cm, slightly less than the female which reaches 3.1 cm) guards the large (diam. 4 mm) eggs until they are nearly hatched (which takes 3-4 weeks), and a noteworthy possibility is that the eggs he chooses are not necessarily the ones he’s fathered because the females lay up to 40 eggs, far more than the male can actually gestate. Just before they hatch, the male takes up to 19 eggs into his mouth and gets them to pass through the paired slits into the ‘vocal sacs’. The eggs hatch about three days later in this body cavity of the male (which may or may not be the father) body and, provided with enough yolk by the mother at the time of hatching, they continue to develop as larvae in this cavity, for 50-70 days. Goicoechea et al. (1986) have shown that the tadpoles nourish themselves partly on secretions inside the male’s sac, which would be an even more remarkable case of male gestation. The male ‘gives birth’ to the offspring when they have partly metamorphosed, with just a stump of the tail remaining and a length (presumably snout to ‘tail’ tip) of about 1cm (which is about the size of a morphling toad).

Please see this video clip of the male ‘giving birth’'s_Frog#p0074thp . You can see from this footage that the ‘neonates’ are small relative to the size of the male, which makes mechanical sense because he has to fit up to 19 into a cavity under his skin. If the adult male is 2.5 cm long and his newborns about 1 cm long, this may not seem like a big difference in length. However, because of overall scaling principles one would not expect the newly metamorphosed frogs to be as small, relative to mature size, as we find in large toads. Considering how tiny the adults are in R. darwinii, I’d suggest that the newborns are small enough to be called ‘morphlings’, which makes sense because a lot of them have to be accommodated in the ‘vocal sac’.

A different way of putting this: morphlings may be consistently about 1 cm long in all species, regardless of the great variation in adult body sizes, because of an allometric exponent. What makes them morphlings is that the babies are small relative to those of other frogs of similar adult sizes.

What constitutes a morphling inevitably depends on the body size of the frog species in question. But the bottom line is that any one inch-long adult male that accommodates more than a dozen newly metamorphosed offspring in a single cavity of his body is almost self-evidently accommodating morphlings, i.e. unusually small metamorphs.

If so, I suspect that many or most frog lineages with external development (eggs laid out of the water, and development of the larvae within the egg capsules based on yolk) will turn out to have morphlings as well. This would include, for example, Arthroleptis ( in Africa and Eleutherodactylus ( in the Neotropics. This needs confirmation, though.

Rhinoderma darwinii newly born with adult male. This ‘baby’ may not look particularly tiny next to an adult male but please bear in mind that even the adult is only one inch long:

Newborn Rhinoderma darwinii; this individual looks less than 1 cm long to me:

Here are two more examples relevant to morphlings, heterochrony, and the flexibility of development in frogs.

We’ve seen that true toads feature morphlings, which are extremely small ‘adults’. The superficially toad-like Pelobatidae, which occur in Europe and spend much of their lives underground but breed in seasonal pools, turn out to be different, and more like paradoxical frogs in their life history.

Paradoxical frogs (Pseudis) have tadpoles that grow up to 25 cm long (taking four months to grow this large), then metamorphose into adults of only a bit more than 7 cm snout-vent length. Well, in the case of pelobatids such as Pelobates fuscus, the tadpoles can grow to 8-10 cm long or even up to 15-20 cm long in some cases (compared with only about 3 cm long for the tadpoles of even the largest true toads such as Rhinella marina). After metamorphosis the snout-vent length is a mere 2-4 cm, which means shrinkage even if one allows for the fact that the external tail has been lost.

I infer that this aspect of the development of pelobatids differs from their Nearctic counterparts the Scaphiopodidae (, which is perhaps one of the reasons why the spadefoot toads of the Northern Hemisphere, previously all lumped into one family, have now been split into a North American family and a different Eurasian family.

Secondly, in the Hyperoliidae, an African family that includes apparently annual reed-frogs in the genus Hyperolius:
the South African Kassina maculata grows to 6 cm snout-vent length as an adult, but its tadpoles reach up to 13 cm before metamorphosing (the larval phase takes up to 10 months). Here again, there must be shrinkage even allowing for the loss of the external tail.

So European Pelobates and African Kassina seem to be further examples of the phenomenon episomised by South American pseudids.

Again, note the difference:
Sclerophrys pantherina stays small as a tadpole, and fully metamorphoses into a tiny version of the adult, but elsewhere in South Africa (as far south as Zululand) we have the hyperoliid Kassina which does the opposite, growing into a tadpole so large that, even in full maturity, the metamorphosed frog never regains such length in its head and body.

These are extremely different patterns but previously hidden by a lack of suitable terms. I dare say there are naturalists in South Africa who know much about frogs but don’t appreciate this axis of difference, because the literature has not brought it out for what it is.

to be continued in

Posted on 24 November, 2023 00:00 by milewski milewski | 0 comments | Leave a comment

The morphlings versus the axolotls (how frogs have warped tadpoles into new shapes and sizes), part 3

...continued from

Large toads have tadpoles no longer than about 3 cm, which metamorphose into small ‘adults’ with snout-vent length not much more than 1.5 cm (and in some cases, I gather, as little as a third of this).

By contrast, the American bullfrog (Lithobates catesbeianus,, which is about the size of a large toad in maturity and is probably the largest non-toad frog in North America, has tadpoles up to 17.7 cm long. I have not found data on snout-vent length at metamorphosis but if the tadpole is >15 cm then presumably the frog is initially at least 5 cm long, severalfold the corresponding size in toads.

The metamorphs of L. catesbeianus can weigh 5 g, compared to <0.05 g in the case of bufonids with similar mature sizes. That’s a difference of two orders of magnitude in body mass at the stage of metamorphosis.

I think it’s safe to say that in most comparisons of bufonids with ranids of similar mature body mass, we can expect a difference of an order of magnitude in body mass at metamorphosis.

Furthermore, ranids such as Lithobates seem to be a bit like centrolenids, in retaining more of the tail at metamorphosis (when the animal leaves the water) than is true of most families of frogs. Bufonids don’t seem to go in for the retention of a residual tail at all.

So typical toads and typical frogs (ranids, so familiar in Europe and North America, although absent from South Africa where their place is taken by pyxicephalids) are similar in their commonness and fecundity, but differ in their development: bufonids have morphlings whereas ranids do not.

The following illustrations show the body sizes relative to human figures for scale.

The species illustrated is the natterjack toad (Bufonidae: Epidalea calamita, of Europe, but the body sizes and shapes are typical of many bufonids including Sclerophrys pantherina.

The natterjack toad is about average size for a toad, usually about 7 cm snout-vent length in maturity. The tadpoles, which complete their growth within two months, are small and this toad metamorphoses at about 0.7 cm snout-vent length.

Initially the morphlings (which are diurnal, presumably to avoid being cannibalised) are easily mistaken for invertebrates in the poolside herbage. The morphlings are so small relative to the fully mature animal that they take 3-4 years to reach sexual maturity.

The point of all of this is that toads typically metamorphose at remarkably small body size, which means that their life history can best be understood by dividing it into not just the three stages normally described for amphibians, viz eggs, larvae, and adults, but rather into four stages: eggs, tadpoles, morphlings, and adults.

The morphling can be thought of as an adult at larval size, and the development of toads can be interpreted, in a sense, as peramorphic – although I’ve never seen this suggested in any literature or anywhere on the internet.

Bufonidae: Epidalea calamita: mature individual:

Epidalea calamita morphling:

Epidalea calamita morphling:

The following reference invokes peramorphosis in frogs of the family Ceratophryidae ( From the abstract it’s not clear what the basis of peramorphosis is.

The ‘early onset of metamorphic transformations’ mentioned in the abstract is indeed what peramorphosis is all about, but the features concerned in this case seem too subtle or obscure to be specified in the abstract. So I’ll have to delve into the body of the paper itself.;jsessionid=0FB2DFDE4A06FDA17DBD9C02A397C7C9.f03t02?deniedAccessCustomisedMessage=&userIsAuthenticated=false

I assume that the lungs of toads only develop at or after metamorphosis (?known as long ago as 1931 when Noble wrote his book about amphibians).

However, it’s worth noting that in many other families of frogs the lungs start to develop well before metamorphosis, along with the developing legs.

This means that the later stages of the tadpole already possess, and use, lungs in many frogs as well as in the salamanders.

Toads seem to be an exception, which makes sense to me in view of the tiny size of the toad tadpoles.

So in toads the appearance of the lungs coincides with the loss of the external tail, whereas in salamander larvae there is no such coincidence because lungs are already present in the larva, and the tail is not lost; while in most tadpoles there is no such coincidence because the lungs develop before the tail is lost.

The paper below, Cohen & Alford (1993), gives data on the body sizes of morphlings for Rhinella marina.

I infer that a morphling can be defined in this species as having snout-vent length less than 3 cm.

The smallest morphlings seem to have snout-vent length of 9 mm, which is bigger than I thought, at least in the population these authors studied. I can’t understand how a morphling of 0.9 cm could possibly weigh as little as 0.025 g (which if memory serves is the minimum body mass given a paper by Shine). A morphling of length 0.9 cm is certainly at least blowfly size, not housefly size and certainly not fruitfly size. But I would still classify these as morphlings, because even at ca 1 cm they are still tiny relative to fully mature body sizes. So tiny that an additional, even tinier, larval stage seems ‘over the top’.

The significance of the morphling stage in toads is that:

this stage combines larval body size with adult form;

toads have essentially two consecutive ‘baby’ stages in their life history, viz tadpole and morphling;

morphlings are subject to cannibalism by juveniles regardless of whether they are also cannibalised by adults;

morphlings are part of a life history strategy of extreme fecundity (enormous clutches of eggs, up to 30,000 by a single mother), in which parental care is replaced by parental ‘hyperinvestment’.

The morphlings of Anaxyrus terrestris (, which I take to have snout-vent length of ca 8 mm, can be as small as 0.055g according to the figures in

For comparison, a house fly (length 6mm) has body mass of 0.012g. This means that even the smallest morphlings of Anaxyrus terrestris are about five-fold heavier than the average house fly.

Fully mature Anaxyrus terrestris reaches snout-vent 8 cm or up to 9.2 cm, which is large for a frog.

So the morphlings are small but certainly much larger than fruit fly size.

The hylid Litoria caerulea ( is a large frog. Its tadpole reaches about 5 cm long including the tail. The freshly metamorphosed (tailless) frog has a snout-vent length of about 1.6 cm, which means that it is initially about the same size as a man’s thumbnail. That means about the size of those big dung beetles one sees in elephant faeces. It can live >20 years, and over that time it grows to a snout-vent length of 10 cm.

Compare this with bufonids. Although the mature cane toad (Rhinella marina) is bigger than Litoria caerulea, its tadpole is smaller, reaching only 3.1 cm long including the tail. The freshly metamorphosed (tailless) stage is about 0.7 cm long snout-vent, about the size of a blowfly. This then goes on to grow even more than is the case in Litoria caerulea.

So there is a difference between this hylid and this bufonid, in relative size of tadpole and freshly metamorphosed frog. The difference in body mass is about an order of magnitude, in the case of both the fully-grown tadpole and the freshly metamorphosed ‘adult’ (tailless for the first time).

There is certainly a difference in life history strategy here. It remains debatable whether this difference justifies the introduction of a new term, viz ‘morphling’ (which needs objective criteria), for the extremely small freshly metamorphosed stage of the bufonid, which is more fecund than the hylid.

The value of making this distinction would be more apparent in a comparison with Pseudis (, which has a much larger tadpole again and does not seem to grow much after metamorphosis.

Incidentally, Hyperolius ( has a tadpole that is larger, relative to the mature size of the frog, than is the case for Litoria caerulea. This is partly because the fully mature Hyperolius is so small. No matter how we define ‘morphling’, Hyperolius would certainly not qualify as possessing such a stage in its life history. The whole concept of a ‘morphling’ probably only matters in large frogs, i.e. frog species in which the fully mature stage has a snout-vent length of say >5cm.

Posted on 24 November, 2023 18:49 by milewski milewski | 0 comments | Leave a comment

25 November, 2023

Diet of the sable antelope, part 1: Hippotragus niger roosevelti, with special mention of osteophagy

@paradoxornithidae @dejong @matthewinabinett @jakob @jwidness @simontonge @tandala @ldacosta

In the late 'sixties, the late Richard Estes ( spent 10 weeks in the Shimba Hills in coastal Kenya. The dates were 17 Oct.-8 Dec. 1968 and 1-14 March 1969.

His purpose was to study a rare and threatened subspecies of large bovid, namely the eastern sable antelope (Hippotragus niger roosevelti,

Richard Estes was, at the time, an authority on the western white-bearded wildebeest (Connochaetes mearnsi), and he went on to become the global authority on the sable antelope.

The eastern sable antelope was originally restricted to a limited strip of coastal East Africa, straddling the border between Kenya and Tanzania.

Its habitat was the northernmost, attenuated form of miombo woodland (, in which only one species of Brachystegia persists under the equatorial regime of bimodal seasonal rainfall.

Estes undertook this study nearly 60 years ago. At that time, the eastern sable antelope still remained in small numbers on the coastal plain in Kenya, in attenuated woodland of Brachystegia spiciformis ( and Afzelia quanzensis (, and at the edges of fire-free forest containing Manilkara and Diospyros. These populations at low altitudes, which Estes did not field-study, have since been exterminated.

The species has survived in Kenya only in the Shimba Hills. Here, the vegetation ( is a picturesque forest/savanna mosaic, only vaguely related to the northernmost miombo woodland.

The Shimba Hills were protected, before Estes' study, mainly for the purposes of watershed and forestry (with part of the area cleared for plantation of Pinus). The National Reserve was designated in 1968 ( and

The main purpose of this Post is to record, electronically, previously unpublished information that risks being lost to Science with the passage of time.

In this, part 1, the topic is the diet of the eastern sable antelope in the Shimba Hills.

My reference (xeroxed) is:

*Richard D Estes and Runhild K Estes (1969) The Shimba Hills sable population. First progress report. National Geographic Society, hippotragine antelope study.

The above report includes data from:

+Table 15 in Glover P E (1969) Report on an ecological survey of the proposed Shimba Hills National Reserve. East African Wildlife Society. 148 pages (as referred to in the above report). This study was made in March through May 1968.

In the following dietary list (as well more generally as in all three parts of this series of Posts), I have updated the species-names, several of which have been revised/synonymised since 1969.



*Cyperus hemisphaericus


*+Andropogon schirensis
eaten intensively in October and March

*+Ctenium concinnum

+Cymbopogon caesius

*+Digitaria milanjiana
eaten intensively in October and November

*+Diheteropogon amplectens

*Eragrostis perbella

+Eragrostis racemosa

*Hylebates chlorochloe

*+Hyparrhenia filipendula
eaten intensively in October and November

*+Hyperthelia dissoluta

+Megathyrsus infestus

*+Megathyrsus maximus
eaten intensively in November and March

*+Panicum trichocladum

*Paspalum orbiculare

*Setaria parviflora

*Setaria sphacelata
eaten intensively in March

*+Setaria trinervia (possibly synonymous with sphacelata)

+Sporobolus pyramidalis

*Sporobolus sp.

*+Urochloa brizantha
eaten intensively in November and March

*unidentified grass, possibly Pennisetum sp.
eaten intensively in November



*Justicia sp.


*Crotalaria emarginata

*Galactia argentifolia



*Rourea coccinea ssp. boiviniana


+Albizia adianthifolia

+Albizia gummifera


*+Securidaca longipedunculata


+Ximenia caffra

The following excerpts from the text in Estes and Estes (1969) are relevant to diet.


Pages 9-10:

"The sable is primarily a grazer on grasses of medium height, preferring the greenest and tenderest available plants. The Longo Magandi herd often specialised on one or two common grasses for a few days while the plants were at the preferred stage of growth, changing to others when the first ones became a little taller and coarser. Plants were usually not grazed shorter than about four inches. Two areas that were burnt in November and December were unutilised by sable until the leaf table exceeded six inches, although tall unburnt grassland was the only alternative. However, the unburnt grassland offered an understorey of green grass. The usual grazing method is to gather in a clump of blades with dexterous lip movements, then to bite off and chew a length of up to a foot...Generally speaking, the sable grazed most intensively around the edges of the copses, in hollows and on termite mounds in open grassland - i.e. in the places where the lushest, tenderest grasses grow...[several] of the grasses, which grow most abundantly in these situations but are not among the commonest grasses, were heavily grazed by sable so long as they were green and young: Megathyrsus maximus, Digitaria milanjiana, and Urochloa brizantha. An unusual habit of Shimba Hills sable is bone-chewing. It was observed on 30 different occasions, and appeared to be a practice indulged in by all herd members. It was usual for an animal to spend up to half an hour chewing a bone, meanwhile frothing at the mouth, and even to transport small pieces from one place to another when the herd moved. The sites of an elephant and of a buffalo skeleton were among the places most frequently visited by the Longo Magandi herd. At those places animals were seen apparently seeking pieces of bone to chew. Long bones were evidently preferred, but probably because few remained at these sites, sable were also seen to chew pieces of pelvis and even vertebrae. Glover also comments upon bone-chewing behaviour. The pronounced deficiency of calcium and phosphorus in Shimba Hills soils is a probable explanation. Yet the sable and other wildlife had failed to discover salt and bone meal put out on a cleared piece of ground the previous November by the following March. Although it was sited near a route not infrequently used by the sable, apparently it remained undetected. The obvious solution would be to establish a lick at one of their 'bone yards'."

Page 6:

"The grassland consists mainly of tufted perennials, separated by bare ground; basal cover averages probably less than 5 percent. While there are many species, tall stalks of Hyparrhenia filipendula and Hyperthelia dissoluta appear dominant following the long rains unless burnt off during the dry season. Andropogon spp., especially A. dummeri and A. schirensis, are co-dominant and may actually be more plentiful though less conspicuous. In Makin's (p. 14) view, Andropogon is an 'unpalatable and poorly nutritious grass which characterises burnt-over and infertile soils.' This genus is nevertheless heavily utilised by sable in most of the areas we have investigated. A far more unpalatable grass, which grows in widely separated clumps and dominates on gravelly slopes, is Trachypogon spicatus []; neither it nor Cymbopogon caesius was ever seen to have been grazed."

For the crucial role of mineral nutrients in the diet and drinking water of the sable antelope, please also see

to be continued in

Posted on 25 November, 2023 16:50 by milewski milewski | 8 comments | Leave a comment

28 November, 2023

Flora and vegetation of Shimba Hills National Reserve in coastal Kenya, including a description (1969) by Richard Estes

@dejong @richardgill @zarek @dianastuder @kai_schablewski @peakaytea @marcoschmidtffm @wasinitourguide @troos @craigpeter @bartwursten

Also see



The following is the main reference to the flora of Shimba Hills National Reserve:[5:ACOTPO]2.0.CO;2.short[5:ACOTPO]2.0.CO;2/ANNOTATED-CHECKLIST-OF-THE-PLANTS-OF-THE-SHIMBA-HILLS-KWALE/10.2982/0012-8317(2005)94[5:ACOTPO]2.0.CO;2.short

The following is a thesis on the topic of the effects of Loxodonta africana on the vegetation:

The following publication is also relevant:


The following is an electronic record of the writings - now at risk of permanent loss in paper form - of a noteworthy researcher ( more than half a century ago.

I refer to pages 4-6 in Estes, Richard D and Estes, Runhild K (1969) The Shimba Hills sable population: First progress report. National Geographic Society, hippotragine antelope study.

'Anon. 1968', 'Glover', and 'Makin' refer to the following:

  • Anon. (1968) A reconnaissance inventory survey of the indigenous forest areas of Kenya. Part 2: Shimba Hills sampling unit. Spartan Air Services Limited. Ottawa, Canada. 59 pages.
  • Glover P E (1969) Report on an ecological survey of the proposed Shimba Hills National Reserve. East African Wildlife Society. 148 pages.
  • Makin J (1968) The soils in the country around Shimba Hills Settlement, Kikoneni and Jombo mountain. Soil Survey Unit, Department of Agriculture, Kenya.

In the following verbatim transcript, I have updated and corrected the scientific names.


[start of verbatim transcript]

"The flora of the Shimba Hills is rich and varied. Glover collected over 1000 species in four months. An inventory of trees found in forest reserves (Anon., 1968) lists 109 species of trees.

The Forest Department 1:50,000 map..., prepared from aerial photographs taken in 1964, distinguishes five different vegetation types including...pine plantations. Glover classifies the vegetation into five main types (not counting the plantations) and over 10 subtypes. The types shown on the map are listed below, with brief descriptions of Glover's main subtypes and the commonest species.


Glover distinguishes three subtypes:


A fairly dense woodland 30-40 feet high with an understorey. This type is common in unburnt, drier areas of the coastal hinterland south and west of the Shimba Hills, but is confined mainly to drainage lines and the northwestern tip of the reserve. A Manilkara ( is often the dominant tree, with Elaeodendron schweinfurthianum (, Searsia natalensis (, Antidesma membranaceum (, Apodytes dimidiata (, Zanthoxylum holtzianum (, Diospyros loureiriana (, etc.


More open, fire-resistant secondary bush has invaded extensive areas of formerly open grassland. The commonest trees and shrubs are Tetracera boiviniana (, Securidaca longipedunculata (, Rourea coccinea ssp. boiviniana (, Ochna purpurea (, Ozoroa mucronata (, Stereospermum kunthianum (, Ormocarpum kirkii ( and Dichrostachys cinerea (

'Sagebrush': Another type of invading woody growth, composed of Vernonia zanzibarensis ( and Lantana camara ( is popularly referred to as 'sagebrush'. It forms a dense low shrubby growth which infests a considerable part of the grassland, especially the Forest Department's plantations. While the Lantana is an exotic, Glover points out that Vernonia infestation represents a normal stage in the succession back to forest. As far as is known, neither is used by herbivores, although birds eat the fruits - and spread the seeds - of the Lantana. Both species are vulnerable to fire.


The origin of the Shimba Hills grasslands is indeterminate. There is some evidence, in the form of truncated soils, that they represent old cultivation sites maintained by fire, rather than edaphic grasslands created by seasonal waterlogging (Makin; Glover). The invasion of bush and sagebrush in the absence of burning bears out the view that Shimba grasslands are a fire subclimax.

The grassland consists mainly of tufted perennials, separated by bare ground; basal cover averages probably less than 5 percent. While there are many species, tall stalks of Hyparrhenia filipendula ( and Hyperthelia dissoluta ( appear dominant following the long rains unless burnt off during the dry season. Andropogon spp., especially A. dummeri [now synonymjsed with the following sp.] and A. schirensis (, are co-dominant and may actually be more plentiful though less conspicuous. In Makin's (p. 14) view, Andropogon is an 'unpalatable and poorly nutritious grass which characterises burnt-over and infertile soils.' This genus is nevertheless heavily utilised by sable [Hippotragus niger roosevelti] in most of the areas we have investigated. A far more unpalatable grass, which grows in widely separated clumps and dominates on gravelly slopes, is Trachypogon spicatus ( ); neither it nor Cymbopogon caesius ( ('pepper grass') was ever seen to have been grazed.

The lushest pastures are to be found on the edges of the copses, upon old termite mounds, and growing in depressions. Here Megathyrsus maximus (, ...Digitaria milanjiana (, including D. mombasana), Urochloa brizantha (, and Setaria trinervia ( grow most abundantly, along with Hyparrhenia filipendula. Other common constituents of the open grassland include Eragrostis spp. and Ctenium concinnum (

In places, especially on ridges, the open grassland is dotted with single trees or clusters of doum palms, in particular the symmetrically branched Hyphaene coriacea ("

[end of verbatim transcript]


Posted on 28 November, 2023 01:03 by milewski milewski | 7 comments | Leave a comment