Caleonic colouration in the caribou, part 1: Rangifer tarandus terranovae

Naturalists are used to thinking that animals tend to have adaptive colouration that blends into the surroundings.

However, in open environments where it is hard to hide, certain animals opt instead to have conspicuous colouration.

One well-recognised pattern of conspicuous colouration in medium-size to large animals is the pied pattern ( and and

This consists of a dark/pale patchwork, too bold to function disruptively. I.e. pied colouration is too conspicuous for the figure to be camouflaged.

However, there is another conspicuous pattern, less familiar but obvious in certain mammals. I provisionally call this 'caleonic', because it deserves a name, but has not previously been described.

In the caleonic pattern, it is as if the figure is 'highlit' by a flame below.

Caleonic colouration seems to have arisen repeatedly in lineages as diverse as

Animals with caleonic colouration have, in a sense, ‘stretched’ the principle of countershading ( to breaking-point. This extension of pale pelage, upwards, has achieved whole-body conspicuousness instead of crypsis (

Caleonic colouration thus differs from pied colouration in how countershading has been modified. In the latter, pale ventral surfaces have been converted into parts of the large-scale, dark/pale contrast in the pigmentation of the pelage.

It is normal for the ventral parts of the body to be pale, as part of the inconspicuous pattern called cryptic colouration. However, the extension of this pale switches the effect to conspicuousness. This is because the pale, encroaching upwards towards the dorsal side, tends to catch the sunlight at

  • all seasons, and
  • most times of day.

In caleonic colouration, this ‘lateralisation’ of the pale parts of the pelage occurs variously on the

  • cheeks,
  • neck,
  • shoulders,
  • flanks, and/or
  • hindquarters.

Caleonic colouration thus achieves whole-body conspicuousness.

In extreme cases, the 'lateralisation' of pale pelage reaches the dorsal surfaces of the rump, neck, and even back.

In this series of Posts, I hypothesise that a widespread species, the caribou (Rangifer tarandus, has

This arguably makes R. tarandus the only species of mammal in which both patterns (pied and caleonic) occur, depending on the region/subspecies.

The only other species possessing caleonic colouration in only certain populations seems to be the domestic horse (Equus caballus). However, this may have arisen by hybridisation among several wild congeners, during the process of selective breeding (see

The following show pied colouration in R. tarandus: and and and and

The following seem to show caleonic colouration in R. tarandus: and and and and

Pied colouration occurs in the autumn coat of most of the subspecies of R. tarandus, including R. t. groenlandicus and R. t. granti. The darkest features are arranged to provide dark/pale contrast, not crypsis or disruption of the outline of the animal.

These dark parts in this pied colouration are

  • muzzle,
  • forelegs,
  • brisket/chest/lower shoulders, and
  • lower flanks.

In the same figures, the palest features are

  • nose (rhinarium),
  • neck,
  • beard/dewlap,
  • tail, and
  • the narrow area of pale on the rump.

In R. tarandus, caleonic colouration seems to occur in the subspecies found on two systems of islands, far apart geographically (

Firstly, on the island of Newfoundland ( occurs the subspecies R. t. terranovae ( and

Secondly (see part 2 of this series), on the Arctic islands of Canada occurs the subspecies R. t. pearyi ( and and and

I agree with Valerius Geist that R. tarandus on the island of Newfoundland is quite distinct from other forms of ‘woodland caribou’, and that taxonomy took a wrong turn when R. t. terranovae was lumped with R. t. caribou.

The ‘woodland caribou’, in the western part of the boreal zone of North America, is unusually dark for the species (e.g. according to Valerius Geist). By contrast, R. t. terranovae is unusually pale - particularly considering that it lives at a far lower latitude than R. t. pearyi.

The following further illustrate R. t. terranovae:
and and and and second photo in

In the pied pattern, the lower flanks, chest, lower shoulders, and brisket are the darkest parts of the figure. This differs from the caleonic pattern in R. t. terranovae, in which these are the palest parts of the figure. Furthermore, a distinctive feature of R. t. terranovae is an individually variable diagonal border between relative dark and relative pale on the haunch ( and

Detracting from the caleonic pattern is the fact that some individuals of R. t. terranovae have a faint version of the flank-banding seen in other subspecies of R. tarandus.

However, this may perhaps be owing to some degree of anthropogenic mixing with another subspecies.

I hypothesise that the original appearance of the Newfoundland form was truly caleonic, to the exclusion of the flank-banding.

Wilkerson (2010,,%20excerpt.pdf) states that the domestic reindeer (R. t. tarandus) was introduced to the island of Newfoundland early in the twentieth century, and that there was indeed contact between this subspecies (which possesses flank-banding) and the indigenous populations, viz R. t. terranovae. I infer the possibility of hybridisation.

It is hard to see how the two patterns of colouration in R. tarandus, namely pied and caleonic, can be represented as extremes on a continuum. Instead, what seems to have happened is that, in an ancestral form,

  • the whitish at the belly has spread so far up that it has replaced the entire flank-band complex of the pied pattern, while at the same time
  • the legs have gone from basically dark to basically pale, and
  • the neck has acquired a darkish dorsal (nuchal) zone.

It is this combination of pale flanks, pale legs, and a dark nuchal area that makes R. t. terranovae a candidate for caleonic colouration. The combination of extension of pale pelage on to flank and haunch, plus an overall pallour, set this subspecies apart from all other subspecies, besides R. t. pearyi.

However, the following three aspects detract from any simple characterisation of the colouration in R. t. terranovae as caleonic:

  • some individuals do retain a faint version of the flank-banding typical of most subspecies of R. tarandus;
  • the pale of the neck tends to be disjunct from the pale of the shoulders and flanks, separated by a more-or-less vertical tract of pale greyish-fawn pelage; and
  • the pale of the lower haunch tends to give way, ventrally, to a darker tone on the upper hindleg, in some individuals.

What this means is that the pattern in R. t. terranovae - at least in its remaining, hypothetically hybridised form - is not categorically different from that in other subspecies.
In R. tarandus, any description of adaptive colouration must account for the seasonal cycle, in which the pelage moults. The moulting leads from this appearance ( to this (

In winter, the pigmentation fades, owing to wear and weathering. The pelage then falls out in early summer, and re-grows from the darkly-pigmented skin. The colouration initially looks fairly uniform, before the full length of the hairs is achieved and the pigmentation differentiates.

Thus, the fully-differentiated colouration tends to be expressed in autumn, corresponding to the rutting season.

The following two male individuals, probably in September, show some features linking the ‘typical’ pattern above to the pattern typical of other subspecies of R. tarandus: These include the relative darkness of the legs and the faint banding on the flank. These detractions notwithstanding, the colouration remains different from those of any other forms of ‘woodland caribou’ at the same season.

I take the following to be an adolescent male individual in autumn: The side of the body shows a faint version of the banding seen in most subspecies of R. tarandus.

April: and

Early May: and and and
Late May: and
Early June: and and and and and

Mid-June: and
Late June: and
Early July: and 
Mid-July: and and and and and
Late July:
Early August: and
Mid-August: and
Late August:

Early September: and
Mid-September: and and

The following ( seems to conform to caleonic colouration. However, once one understands the cycle of moult and regrowth of the pelage, it is obvious that the effect results mainly from the fact that the faded coat of the previous year has fallen out from the dorsal part of the figure first (as seen more clearly in

to be continued in

Posted on January 30, 2023 09:12 AM by milewski milewski



Rangifer tarandus occurs in Labrador, which is on the mainland near Newfoundland. This form is nominally referred to R. t. caribou, but this subspecific name may not mean much in this far-east location.
Here I show that the caleonic pattern of colouration, found on the islands of Newfoundland, does not occur in mainland Labrador.

The colouration in the George River population of mainland Labrador is similar to R. tarandus in general, and different from e.g. the Avalon Peninsula of the island of Newfoundland as seen in

The following show the appearance of R. tarandus in the George River population of mainland Labrador.

These findings suggest that the colouration of R. t. terranovae, which I take to be restricted to Newfoundland sensu stricto, is distinctive from that on the nearby mainland, in being caleonic.

Posted by milewski about 2 years ago

w.r.t Rangifer tarandus, an author tried to recently change the taxonomy of the Rangifer tarandus into 5-6 different spp. ranked taxa, which is quite the splitting. In the revision (see:, Harding (2022) claims there to be 5-6 species of Rangifer tarandus based on phylogeographic evidence.

Posted by paradoxornithidae over 1 year ago

Harding wrote "Divergence time estimates of the split between forest (NAL) and barren-ground (BEL) clades range from 135,600 years ago during the penultimate (Illinoian) interstadial (Taylor et al. 2021) to a pre-Illinoian glacial period 300,000 years ago", wouldn't this be a very recent speciation date? For reference, the split between San (Homo sapiens) & Non-San (Homo sapiens) populations of anatomically modern humans is estimated to be 200,000 years old (based on haplogroup divergence times), so according to divergence, that would render human populations as different species, which is obviously not the case.

According to the revision, "The genetic difference estimates between woodland caribou and barren-ground caribou, based on mtDNA, range from FST = 33% to > 50%". The difference between Mbuti and Papuans is FST=37.7%. I may be looking at it incorrectly, & pardon me if that's the case, but this seems like a premature splitting.

Correction: From Wikipedia, "Among this set of 42 world populations, the greatest genetic distance observed is between Mbuti Pygmies and Papua New Guineans, where the Fst distance is 0.4573, while the smallest genetic distance (0.0021) is between the Danish and the English."

Posted by paradoxornithidae over 1 year ago

@paradoxornithidae Many thanks for your comments.

Posted by milewski over 1 year ago

I highly appreciate the informative posts & media you have shared. It may be strange that I ask, but would an fixation index value of 0.4573 (45.73%) be considered normal for a species with no subspecies structure?

Posted by paradoxornithidae over 1 year ago

@paradoxornithidae Many thanks for your appreciative comment. If the value you refer to is genetic, I'm the wrong person to ask. I'm aware that genetics is an obscure field, in which we are invited to accept various assumptions used in various analyses, relating to entities (genetic molecules) which we have no way of observing ourselves. Given that not only Biology, but Science in general, is in the midst of a Replication Crisis (, it is hard to know how reliable or informative genetic analyses really are. I prefer to work on the basis of what I can observe myself: and even in my case, there remain problems of confirmation bias ( and other detractors from true objectivity.

Posted by milewski over 1 year ago

One of those links above ( claims that the extirpation of the Canis lupus (Grey Wolf) stopped the migratory patterns in some herds of Rangifer tarandus (caribu). Interesting, this would imply the they migrate not just seasonally, but to avoid predation by Canis lupus (Grey Wolf) in particular.

Posted by paradoxornithidae over 1 year ago

One of the best photos of fully mature male of Rangifer tarandus terranovae, in autumn:

please scroll to fifth photo in

Posted by milewski over 1 year ago
Posted by milewski over 1 year ago


It is indeed plausible that certain populations of Rangifer tarandus migrated partly to outdistance Canis lupus.

A remarkable pattern, among large mammals, is that no predator follows any migratory population of any prey species over its migratory route.

This is best-exemplified by the Serengeti ecosystem, where migrations are undertaken each year by Connochaetes mearnsi, Equus quagga boehmi, Taurotragus oryx livingstonii, and Eudorcas thomsoni nasalis.

Panthera leo and Acinonyx jubatus do not follow the migrations, although non-breeding individuals do move to some extent with them.

Crocuta crocuta effectively tracks the migrations to some extent. However, this is not done by the predator simply following the prey. Instead, what happens is that the hierarchically inferior females undertake long multi-day excursion from their fixed denning-sites, visiting the migrations mainly to scavenge individuals of C. mearnsi that have died from mishap, and then returning to their offspring with a full udder.

As a result of the failure of any carnivore, particularly the prime breeding females, to follow the migrations, the main consumer of C. mearnsi in the Serengeti ecosystem turns out to be the scavenging bird, Gyps. The vultures are incapable of killing any of the ungulates, but simply undertake daily excursions from their resting/breeding sites, capitalising on corpses resulting from mishap. Even these birds, which enjoy the locomotory economies of soaring flight, do not manage fully to track the migrations as breeding females.

Lycaon pictus, the closest counterpart in the Serengeti ecosystem for C. lupus, is particularly unsuccessful in following the migrations. This is because a) as in all the other meat-eaters involved, the breeding sites are fixed rather than mobile, b) the predator is too small-bodied to kill the two largest-bodied migratory spp. (E. q. boehmi and T. o. livingstonei) to any important extent, c) the canid is both subordinate to P. leo and C. crocuta and rather rare and sporadic in the ecosystem, and d) L. pictus, unlike C, lupus, P. leo, or C. crocuta, rarely scavenges.

Few biologists would be surprised to hear that an ungulate might migrate partly to evade predators. It comes as more of a surprise to realise that no predator - and least of all the breeding females - seems energetically capable of simply following migratory ungulates.

Posted by milewski over 1 year ago
Posted by milewski over 1 year ago

@maxallen @alex_wall @michalsloviak @matthewinabinett

I wondered about the current genetic purity of Rangifer tarandus terranovae, because Long (2003) records 22 'introductions' of a total of 384 individuals, 1961-1982.

I refer to and and and and

The way Long (2003) presents the information, the animals were introduced to Newfoundland from elsewhere (which would have compromised the indigenous stock, resulting in a subspecific hybrid today).

However, I suspect that the reality is that all of these we reintroductions, based on the residual population, e.g. on the Avalon Peninsula of Newfoundland (

For example, in the 1950's, when Mike Nolan ( and and and and started to protect the Newfoundland caribou all by himself, the population on the Avalon Peninsula had declined to a mere 84 individuals.

As the population built up over several decades (until Nolan retired in 1980), repeated reintroductions were made to various parts of Newfoundland. Many of these were from the population on the Avalon Peninsula.

Nolan and the team he came to lead are likely to have documented their reintroductions meticulously, compared to any introductions made from sources beyond Newfoundland by sundry other agencies.

Therefore, I infer that Long had no information of any real 'introductions', and merely presented the history of reintroductions, which was ultimately extremely successful.

I now doubt that other subspecies were resorted to, given that the indigenous subspecies did actually survive in Newfoundland even in the 1950's.

If I am right in my reading of the literature, then R. t. terranovae still remains in 'pure' form, and its 'genetic pollution' - implied by Long (2003) - is a false alarm.

Also see

Posted by milewski over 1 year ago

All adult males of Rangifer tarandus, in hard antler, have dark faces.

However, this is particularly striking in R. t. terranovae, notwithstanding that this darkness excludes, at least in some individuals, the forehead, muzzle-tip, and masseters.

In the case of the forehead, the pale patch is extensive, down to the level of the eyes. In the case of the pale muzzle-tip (, this pattern differs from other subspecies other than R. t. pearyi.

The faces of adult females without antlers, like those of males in hard antler, are peculiar-looking, with pale, broad ring posterior to the rhinarium.

Posted by milewski over 1 year ago

The legs of Rangifer tarandus terranovae are pallid, unlike those of other subspecies besides R. t. pearyi.

This pallor on the legs is consistent with the colouration of the animal as a whole. However, at least in males in hard antler, the legs are not whitish. Instead, they are a nondescript and rather untidy pale greyish.

Posted by milewski over 1 year ago

In Rangifer tarandus terranovae, the brisket is not dark, as it is in some other subspecies of the species.

In most subspecies of Rangifer tarandus, there is a pale patch posterior to the elbow (e.g. and

This feature is usually absent in R. t. terranovae, which instead features a pale patch on the haunch.

In R. t. terranovae, a striking and fairly consistent feature is a somewhat diagonal divide between the 'ground-colour' on the dorsal part of the haunch and the whitish on the ventral part of the haunch. Whether this feature conforms to the caleonic pattern or not, it is absent from the migratory (barren-ground) and woodland/mountain types of Rangifer tarandus.

Posted by milewski over 1 year ago

The colouration of the neck in R. t. tarandus differs from that in other subspecies.

Firstly, the dorsal part of the neck is not as pallid as expected for the species; it is as if the 'ground-colour' extends in an anterior direction on the dorsal surface of the neck, to the posterior surface of the head.

Secondly, the pale of the neck tends not to be continuous with the pale of the shoulders, despite the general theme of pallor in R. t. terranovae. Instead, there tends to be a narrow vertical tract of 'ground-colour' separating the neck from the shoulders and scapula. The pallid shoulders, ventral half of flanks, and haunches are not mere extensions, in a posterior direction, of the pallor on the neck. If anything, the neck is overall less pallid in R. t. terranovae than jn R. t. granti/groenlandicus in autumn, and the pale of the neck remains separate from the pallor on the rest of the figure.

Overall, the neck conforms rather poorly to the caleonic pattern, in R. t. terranovae.

Posted by milewski over 1 year ago

There are two different patterns of conspicuous colouration within Rangifer tarandus terranovae, namely one with dark/pale contrast in autumn-early winter, and a pallid one in late winter-spring. The former includes the rutting season, and the latter includes the birthing season.

At the height of summer, terranovae shows dark/pale contrast in a 'vertically-stratified' way. The dark occurs on the upper surfaces: head, neck (other than a short pale 'beard'), scapula, withers, back, upper flanks, upper haunches, rump, and probably short stripe on upper surface of tail. The pale occurs on the muzzle-tip, beard, brisket, shoulders, forelegs, lower flanks, chest, belly, lower haunches, hindlegs, and a patch on the rump.

This means conspicuous colouration throughout the annual wear cycle of the pelage. In autumn it has a caleonic pattern. However, by the end of winter, wear and weathering of the hairs produces a general pallor (except for part of the face, which remains dark in some individuals).

The dark/pale, 'stratified' appearance of the height of summer is produced by retention of the wooly, whitish pelage below the level of the lower flank (including the 'beard'). Above this level, the emergent new coat of pelage is relatively dark, except on the neck (where only the nuchal area is darkish). On most of the neck, even the short new hairs are relatively pale from the start.

Once all the old hairs have been shed, the new, growing pelage features fairly dark upper parts to the torso, bordering abruptly with the pale of the rump; pale grey on the lower flanks, shoulders, haunches, and legs, plus pale grey on most of the neck; and dark on most of the head. The lower level (ventral parts) of the figure already seems sheeny at this stage, despite the shortness of the hairs.

In the fresh coat of late summer, the tail is largely white, with minimal dark in its upper side.

The age-categories in the population diifer chronologically in the annual cycle of renewal of pelage. The most recent cohort of juveniles, at the height of summer, shows most individuals still whitish, some individuals having pale greyish torso and pale legs. This is because the juveniles change colour after the adults.

Posted by milewski over 1 year ago


During the rutting season, males, female, and juveniles form mixed groups.

At this time, the faces are grey rather than blackish, and the rest of the figures are pallid, including the pale greyish tract from from withers to rump, above the level of the mid-flank and upper haunch. What I have noticed to be consistent is that, even where the upper body is pale greyish, the dorsal edge (the top of the back, from withers to rump) is pale. I.e. there is no dark accentuation of the dorsal silhouette, and there may actually be pale accentuation instead.


A close-up view of what I take to be an adult female individual in hard antler in winter shows a whitish figure with little more than a) a faint remnant of the crisp delineation between the white adjacent to the anus and the darker pelage at the junction of rump, haunch, and buttock, b) a faint, fuzzy stripe along the mid-flank, and c) a shorter but equally faint, fuzzy stripe along the middle of the side of the face, as if the darkness of the face has been nearly wiped out by encroachment of pale from the muzzle on one side, the masseters on the other side, and the orbits and forehead above.

The brisket, too, is pale, despite the relative darkness of the ventral edge of the torso (perhaps owing to soiling rather than pigmentation).

Posted by milewski over 1 year ago

Newfoundland, the exclusive habitat of Rangifer tarandus terranovae, is an island about the same size as Cuba, Iceland, or Ireland, and larger than Sri Lanka or Tasmania (the last two being similar to each other in area).

Posted by milewski over 1 year ago

The following, buried in the Web by vague labelling, is what I take to be Rangifer tarandus terranovae:

Posted by milewski over 1 year ago

An additional worthwhile photo of Rangifer tarandus terranovae:

Posted by milewski over 1 year ago
Posted by milewski over 1 year ago
Posted by milewski over 1 year ago

The following, of Rangifer tatandus tarandus, shows the clearest and simplest expression of the dark flank-band in this species:

Posted by milewski over 1 year ago

Add a Comment

Sign In or Sign Up to add comments