QUESTIONS AND ANSWERS: Squirrels I

Content Updated: 15th December 2010

QUESTIONS:

Why is the Red squirrel declining in the UK?
What controls the cacheing behaviour of squirrels and how do they find their buried nuts?
Why are some squirrels of the same species different colours?

Q: Why is the Red squirrel declining in the UK? Is it the fault of the Grey?

Short Answer: There is no single, or straightforward, answer to this question. It appears that loss of habitat, disease and competition with the larger Grey squirrels are all factors. Some areas have seen declines in Reds that closely match the spread of the Grey, while the two species have co-existed for years together in other parts of the country. There are several projects underway to both understand and reverse the decline and there has been limited success in removing Greys from some areas of mainland Britain and supporting Red recolonisation. Complete elimination of Greys seems unlikely at this point.

The Details: Archaeological data suggest that Red squirrels (Sciurus vulgaris) have been in Britain for more than 10,000 years; their remains originate from a time just before Britain lost its land bridge with Europe. The Red squirrel has, however, been in serious decline across the UK for about the last 50 years and in northern Italy for the last 20.  What’s more interesting about these declines is that they seem to -- at least in part -- correspond to the introduction and range expansion of an alien species from North America. The Grey squirrel (Sciurus carolinensis) is native to the deciduous woodlands of the eastern USA, where it co-exists with the American Red squirrel (Tamiasciurus douglasi), which despite bearing a striking resemblance to S. vulgaris, is only distantly related to it.

Historically, there have been many introductions of the Grey squirrels into mainland UK forests. There are records of releases (usually from private/pet collections) dating back as far as 1828 (in this case Denbighshire, North Wales), although the first verifiable record is from 1876, when a Mr. T. V. Brocklehurst released a pair of Greys into Henbury Park near Macclesfield in Cheshire, when their attraction as pets waned. It appears that Mr. Brocklehurst started something of a trend and releases continued for the next 50 years. In her 1954 paper to the Journal of Animal Ecology, Monica Shorten lists the documented release sites of Greys and from her data some of the most significant releases -- in terms of helping to establish a wild population -- include Richmond Park in Surrey where 100 were released in 1902, Regent’s Park in London where 91 were released between 1905 and 1907 and Woburn Park in Bedfordshire where ten were released during 1890. There were also several introductions in Scotland (including Edinburgh in 1913). The final documented release of Grey squirrels was in 1929, when two individuals were released into Staffordshire's Needwood Forest.

As early as 1944, it was apparent that Grey squirrels had become well established and that Red squirrels were in serious decline across the country, although opinion has always been divided as to the root cause of this decline because Red squirrel populations have suffered many times throughout their history. For example, Reds were actively hunted in the New Forest (Hampshire) during the 19th Century and, in her book Squirrels, Jessica Holm states that in 1889 nearly 2,300 Reds were shot because they were considered a pest to the timber industry; between 1903 and 1933, the Highland Squirrel Club reputedly killed 82,000 Reds.  Indeed, between 1900 and 1925 there was a dramatic decline in the number of Reds in Britain – so noticeable was the decrease that hunting had been suspended in the New Forest by 1927. Further declines were recorded between 1939 and 1943, which were attributed to the timber demands imposed on the country during the World Wars and a number of exceptionally cold winters.

Today, the decline in the Red squirrel has progressed to such an extent that they now only persist in a few -- isolated -- areas of the UK including the Isle of Wight, Dorset and pockets of Wales. The Red squirrel “stronghold” in the UK is generally cited as Northern England; English Nature estimates that somewhere in the region of 85% of England’s 161,000 Red squirrels live in Cumbria, Durham, Northumberland and North Lancashire. The Red squirrel is still fairly widespread across Scotland and Ireland. On a larger scale, Reds persist in significant numbers throughout much of Europe, from a patchy distribution in Spain to a more catholic distribution through France, Germany and into Poland, Romania and the Ukraine as far east as the southern Urals and Altai mountains in Russia and as far north as northern Finland. According to the Societas Europaea Mammalogica (2004), Reds can also be found in parts of central Mongolia, China and Hokkaido. While large populations persist across much of Europe, however, their future is far from certain since the introduction of Greys to Italy. The first verifiable release in Italy was during 1948, when four squirrels (2 male and 2 female) were released at Stupinigi in Piedmont (a province of Turin in northwest Italy), while subsequent releases include five animals into a park at Villa Groppallo in Genoa Nervi (on the Italian Riviera) during 1966 and three pairs released into an urban part at Trecate (a province of Novara in northwest Italy). A survey in 1996 estimated the Italian Grey squirrel population (the entirety of which appear to be in the northwest provinces) at between 2,500-6,400 individuals.

Despite knowing what is happening to the Red squirrel population, perhaps the most controversial of all current British conservation debates is why this is happening. Currently there are three main schools of thought on the subject: Greys are actively fighting with and driving out Reds; Greys are out-competing Reds (for food, space, breeding sites etc.); and Greys are spreading disease to Reds, which is causing populations to contract. In truth, there is a somewhat smaller school who consider the decline of the Red squirrel may have more to do with some poor choices in respect to our management of their habitat than the introduction of the Grey squirrel, although this is not as widely accepted as the aforementioned theories. Here I will cover these four schools in turn, although the coverage will be brief – if you want to know more, I have given some references that will provide you with a far more comprehensive account.

Greys Actively Attacking and Driving Off Reds
While researching for my article on the Natural History of Tree Squirrels, I read many accounts of Grey-vs.-Red encounters (on websites, it has to be said) that implied some kind of one-upmanship between these animals – that is to say that the writers considered Grey actively ‘turfed’ Red squirrels out of their homes upon arrival. One particular discussion on an Internet board to which I subscribed left me wondering if the Grey squirrels had got together and instigated a coup d’etat against our Nutkin! Upon reflection, it seems reasonable than Greys could evict Reds were they so inclined – after all, a Grey is roughly twice the size of your average Red. There are reports of hostility between these species -- including instances where Greys have apparently killed young Reds -- but, on the whole, encounters seem to emanate a sense of tolerance, if not indifference. Indeed, aggressive encounters are well known between Greys and it seems that they are just as likely to attack and kill a conspecific as they are a heterospecifc. Overall, there are far more observations of these two species feeding amicably together than there are records of aggression between the two. Consequently, if direct aggression is a factor in the decline of Reds, it is certainly not considered to be the primary one.

Greys Out-Competing Reds
This concept has two avenues: greys out-breeding reds; and greys being better adapted to surviving in deciduous woodland than Reds, which consequently leads to them monopolizing resources (i.e. food, dreys, territories, etc.). The suggestion that Greys may out-breed Reds -- in other words, produce more young per season and thus force Reds out by increasing their own population size -- isn’t borne out by the data we have from the field.  Although it is certainly true that Greys have increased their populations at a spectacular rate since their introduction, work by Anglesey Red Squirrel Project biologist Colin Shuttleworth suggests that, in appropriate habitat (i.e. conifer stands), Red squirrels have a fecundity equal to that of Greys in deciduous woodland.

Conventionally, the idea that Grey squirrels may out-compete Reds has maintained a large following. Various authors have presented data showing how the Grey squirrel is better adapted to a life in deciduous stands than the Red. One of the most intriguing aspects of this is the role of acorns. The suggestion is that a pivotal concept in the success of Greys in deciduous woodland has been their ability to digest acorns – something often referred to as the “Phytotoxic Hypothesis”. Phytotoxic comes from the Greek phuton (meaning plant) and so translates to ‘toxic plant hypothesis’. In a fascinating paper to the Proceedings of the Royal Society of London in 1993, Centre for Ecology and Hydrology biologist Robert Kenward and Royal Holloway and Bedford New College zoologist Jessica Holm present data from an experiment looking at the ability of squirrels to survive on English oak (Quercus robur) acorns. During the experiment, the biologists fed six captive squirrels of each species a mixed diet of acorns, peanuts, carrot, apple, hazelnuts and sunflower seeds and found that, while the Greys readily consumed the acorns, the Reds did not. When the diet was altered to include only acorns, Reds were observed to eat them, but suffered significant declines in weight associated with enteritis (intestinal inflammation); all died within 25 days and one died despite being transferred back to a mixed diet as soon as the other two succumbed. Conversely, all the Greys in the experiment were seen to put on weight when fed only acorns.  Overall the scientists considered that this difference in survivability was attributable to the digestibility of the acorns. Based on their faecal analysis, when on an all-acorn diet, Red squirrel faeces contained 40% more water than the Greys' and, when they included the proportion of undigested material, they found that, relative to Greys, Reds had an acorn digestive efficiency of 59%. This difference in digestive efficiency was attributed to Greys having a greater capacity to excrete the polyphenols (especially tannins) that make acorns toxic to most mammals; Greys could reduce pholyphenol concentration by 71%, while Reds could only manage a maximum of 24%. Unfortunately, there are no data to show whether Greys are at a similar disadvantage if fed on diet of only conifer seeds, although a subsequent study by the same authors found that Grey populations could only persist in the conifer forests they looked at if there was a sustained stream of immigrants – this suggests that Greys may be at a competitive disadvantage in conifers.

In addition to the data we have concerning specific sources of food, Greys are well known to put on almost twice the weight over winter that Reds achieve and also grow to almost twice the size of Reds. This, coupled with the observation that they have a higher daily energy expenditure (and consequently demand more resources – some authors have suggested that a single Grey can use the same resources as 1.65 Reds), suggests that Greys may exert more competition on Reds than do conspecifics. This is far from universally observed, however. There are examples (e.g. in Norfolk) where Reds disappeared 18 years before the first Grey was sighted; similarly there are records of Greys and Reds co-existing for as long as 16 years. Indeed, there is anecdotal evidence to suggest that the two species have lived together for 60 years or more in some areas.

So, if competition for resources is the reason for the Red displacement, how do we explain these examples? Well, one theory is coexistence through habitat partitioning – in other words, physical and behavioural differences in the foraging ecology of two species, combined with pronounced seasonality in their environment, allows them to share the same area by either exploiting different resources or exploiting the same resources in different ways or at different times. In their 2002 paper to the Journal of Applied Ecology, Oxford University zoologists Jenny Bryce, Paul Johnson and David Macdonald present data from the population of Red and Grey squirrels that have seemingly coexisted for some 30 years in Scotland’s Craigvinean Forest. Bryce and her colleagues found that although Red and Grey ranges overlapped, there was a distinct partitioning of resources within the habitat; Red squirrels chose areas of Norway Spruce (Picea abies), while Greys preferred patches of mixed conifers and broad-leaf trees. While the biologists found no evidence that either species avoided the other, they suggest that this habitat partitioning may have reduced the competition between the species sufficiently to allow coexistence.

While co-existence may be possible under certain conditions or in certain habitat types, work by Luc Wauters, Guido Tosi and John Gurnell in Italy has found that Red squirrels which had a high percentage of their home range overlapping with Greys experienced a lower daily energy intake than those with little or no overlap. Moreover, while the biologists found no evidence that spring body mass was related to the number of Reds with which their territory overlapped, it was negatively correlated with the area shared with Greys (i.e. as the percentage of overlap with Greys increased, the squirrels had put on less body mass body mass by the springtime). Wauters and his colleagues suggest that competition between Reds and Greys for scatter hoarded seeds (and more specifically, cache pilfering by Greys) may serve to reduce the reproductive output of Reds and, consequently, contribute to their decline.

Despite the observation that Reds may be at a competitive disadvantage when searching for food in areas containing Greys, there is little evidence to suggest that the presence of Greys has any significant impact on the type of food consumed by, or on the foraging patterns exhibited by, Red squirrels. Such patterns have been documented (especially during the winter), but the data can be explained by factors other than Grey presence, such as the distribution of the Reds’ preferred tree seeds. Indeed, additional work by the aforementioned authors has shown that, although interspecific competition (i.e. between Reds and Greys) occurred in their Italian study area, it didn’t lead to lower winter survival, spring breeding or a decrease in body condition when compared to sites without Greys.

While competition with Grey squirrels doesn’t appear to have a substantial impact on adult Reds, some data imply that its influence on juvenile and subadult individuals may be more pronounced. Research by John Gurnell, Luc Wauters, Peter Lurz and Guido Tosi published in the Journal of Animal Ecology during 2004 found that, in years when Greys were present in the study area, fewer Red females bred during the summer and there were fewer instances of multiple litters – this lower fecundity was attributed to the females having lower body mass in areas with Greys than in those with only Reds. Furthermore, in their mixed broadleaf sites, Red recruitment rate and juvenile residency decreased with increasing Grey squirrel density. In other words, Red females in woodland with Grey squirrels had lower body weight, produced fewer kittens, of which fewer survived and, of those that did, fewer remained at their natal site. It should be noted, however, that the sample size for the English sites with both species was small and that, although the summer breeding was affected by Greys, they had no discernable impact on spring breeding. Nonetheless, these and previous data from the same authors do suggest that the presence of Greys can lead to a reduction in Red squirrel population recruitment rates (Red recruitment in one mixed site was 13%, while it was 50% at the Red-only site). Dispersing Reds seem less able to settle in areas inhabited by Greys.

Disease-Mediated Decline
Conventionally, it has been considered that competition between Reds and Greys has been the main reason for the decline of S. vulgaris. In a recent paper to Ecology Letters, however, Daniel Tompkins and colleagues report the results of a simulation model, which found that there was an unrealistically slow replacement rate of Red squirrels by Greys when competition alone was considered. That is to say that their model suggests competition alone cannot account for the rate and pattern of Red decline we have witnessed in the UK. Instead, the biologists consider that disease has potentially played a crucial role in this decline.

Perhaps the most off-cited and well publicised disease has been a viral infection that produces symptoms similar to the myxoma virus (Leporipoxvirus), which causes myxomatosis in rabbits. This virus is frequently referred to as “parapoxvirus” (often shorted to “parapox”), although recent taxonomic work suggests that it actually represents a new genus within the Chordopoxviridae family – as such I prefer to follow Moredun Research Institute’s Kathryn Thomas and her colleagues in referring to it simply as a novel squirrel poxvirus (SQPV), until further phylogenetic data are available. (Photo: A young Grey squirrel that succumbed to the pox virus.)

The origin of SQPV is currently unknown although, as many conservationists point out, no records of the disease exist prior to the introduction of the Grey squirrel into the UK. It is, however, interesting to note that the first definite British record -- there are other descriptions of disease with similar symptoms as far back as 1920 -- comes from East Anglia in the 1980s, which is at least 104 years after the first verifiable introduction of Grey squirrels. Whether or not the virus was actually introduced with some Greys, this species certainly seem to act as a reservoir host for it. SQPV seroprevalence -- the number of individuals in a population that test positive for antibodies to an infection -- in Greys is high; one study published in 2000 reports that 136 of the 223 (61%) apparently healthy Grey squirrels tested had antibodies to SQPV (i.e. had been exposed to it at some point during their life), while only four of 140 (3.2%) Reds were seropositive. Perhaps more importantly, all of the seropositive Reds were found dead or dying with symptoms typical of SQPV – these symptoms include ulcerated and bleeding scabs around the eyes, mouth and nose, which later spread to the ventral thorax (chest), inguinal (groin) area and the feet.

In a 2002 paper to the Proceedings of the Royal Society of London, Daniel Tompkins (University of Stirling, UK) and colleagues wrote:

“... grey squirrel seroprevalence to parapox-virus is high in English and Welsh populations, where the red squirrel is almost extinct, but zero in Scottish and Irish populations, where the decline is far less marked and epidemic outbreaks of infection disease have not been documented.”

Interestingly, in the paper, Tompkins and his colleagues report that one Red squirrel in their study recovered from SQPV, despite suffering exudating (weeping) and inflammatory lesions for some six weeks. This is interesting, because it represents the first evidence that the immune system of at least some Red squirrels is capable of fighting the virus if given a suitable environment (i.e. in captivity).  Moreover, because this particular individual was found to have an initial antibody response 38% higher than the other three Reds in the experiment and that this antibody response eventually plateaued at a level seen in wild Grey squirrels, the authors suggest that it may be possible to vaccinate Reds against SQPV.

Currently, there is only one record of suspected SQPV from Sciurus carolinensis, which comes from the examination of a wild, adult squirrel in 1994 that was found in Hampshire (UK). Lesions on its face yielded “parapox-type virus particles”. Overall, the high seroprevalence of SQPV, combined with the lack of observed symptoms in Greys suggest that they are a carrier for the virus and capable of passing it on to Reds – observations on captive Reds along with the fact that the only wild records of SQPV in Reds come from dead or dying species imply that the disease is lethal to them with a matter of days.

Unfortunately, just as we still do not know the virus’ origin(s), assuming that Greys do pass on the disease to Reds, there are no data to confirm a primary route of infection. Several methods of transmission have been speculated upon and these include direct contact, sharing the same feeding stations and that parasites (perhaps spread in shared bedding or nest boxes) may all facilitate SQPV transmission. Indeed, laboratory studies in the USA have demonstrated that the squirrel fibromatosis virus -- often, rather confusingly, referred to as squirrel pox -- can be spread by insects (namely the mosquitoes Aedes aegypti and Anopheles quadrimaculatus), while in a 1995 edition of the National Provident Institution Red Alert UK Newsletter, squirrel biologist Ian Keymer suggests that feeders acting as focal points could increase the risk of disease transmission within and between species. Transmission of disease across feeders is clearly of concern to many squirrel conservation groups and the Cumbria based squirrel preservation charity Red Alert advise that people using squirrel feeders regularly (every two to four weeks) clean and disinfect them. Although rather anecdotal, it is interesting to note that in Reds showing signs of SQPV infection tend to exhibit lesions on the feet, stomach, groin and face that are consistent with areas of the body used during scent-marking – this may add credence to the idea that this disease can be picked up from surfaces to which infected individuals have been in contact.

Steven Rushton at the University of Newcastle upon Tyne and his colleagues sum up the SQPV situation quite neatly in there 2000 paper to the Journal of Applied Ecology, in which they write:

“Indeed the grey squirrel-parapoxvirus interaction with the red squirrel could be described as ‘apparent competition’ mediated by an infection agent, as in the case of the pheasant and grey partridge, because the virus gives some advantage to grey squirrels.”

Habitat Loss
The final factor implicated in the decline of the Red squirrel is loss or change of habitat. When the ice sheets of the last Ice Age began to melt about 10,000 years ago, arctic trees like aspen, birch and willow were the first to colonise the landscape. These were followed by pine, hazel, alder, oak, elm, lime, ash, holly, hornbeam and finally maple. Following many thousands of years of interspecific competition for light and space, the so-called “wild-wood” (i.e.  pre-human interference) was complete by about 4,000 years ago, although even before this development was complete mankind began felling. Indeed, it has been estimated that more than 80% of Britain was once covered in forest; the figure today is less than 20%. Historical records show an almost constant process of deforestation until the end of the 19th Century; this deforestation was instigated for various reasons, namely construction in the lowlands and to provide grazing pasture for livestock in the uplands. The First World War reminded us just how important our forests are to us and, at the beginning of the 20th Century, there was a large drive by the newly established Forestry Commission to plant conifer trees in a bid to replace the ancient and slow-growing broadleaf woodlands. Conifers were chosen because they thrived in the acidic soils and cold, wet and exposed environment of the Scottish and Welsh uplands (most of the other prime locations were taken by agriculture). Red squirrel numbers recovered in these plantations and reached their peak numbers in the 20th Century. Contrary to popular misconception, although Greys seem to survive less well in these conifer stands, they can (and do) live in such plantations.

Ultimately, some have suggested that this change in forestry, coupled with changes in some (especially agricultural) land management practices has contributed more to the decline in Reds than competition with Greys or SQPV. Habitat changes (most notably habitat fragmentation) have no doubt played some part in the current condition of our Red squirrel population, although there is little evidence to suggest that it is a major -- let alone unitary -- cause for the decline. Indeed, it could be argued that had afforestation with conifers not commenced at the pace it has, the Red squirrels might not have been able to survive the introduction of Greys at all. Indeed, in his study on the replacement of Reds in eastern England, John Reynolds found no evidence to support the idea that Red squirrel decline was a result of habitat change.

What Can Be Done?
Having looked at some of the potential reasons for the decline in Red squirrels throughout the UK, it is worth considering for a moment some of the measures implemented to try and reverse the trend and re-populate Britain with its native Nutkin. Although monitoring squirrels is not an easy task (most methods are rather imprecise and the number of samples needed to pick up any population change varies according to the size of the area), Red squirrel conservation generally takes three forms: Re-introduction; Habitat Management; and Grey Control.

Reintroduction: During the 17th and 18th Centuries, thousands of Red squirrels were imported into England from Europe for sale in markets; similarly, records of introductions to Scotland date back to 1772 and from Ireland to 1815. Typically, reintroductions were made up of squirrels from elsewhere in the UK (predominantly England), although some individuals were imported from Norway and Sweden. Experiments during which Red squirrels have been released into woodland in order to monitor their progress have proved largely unsuccessful – they are generally either eaten by predators or killed on local roads. One particular study during which Reds were released onto the Goathorn Peninsula of Furzey Island in Poole Harbour (where Greys currently inhabit conifer woodland) found that none survived for more than four months; half were eaten or cached by predators (primarily foxes) and dissection of recovered carcasses revealed that the animals were stressed (suffering hypertrophied adrenals, disease and weight loss).  None of the females showed any sign of reproductive activity and, moreover, none weighed-in above the 300g (10.5 oz.) threshold at which oestrous can be entered. Additionally, the tracking data indicated interference competition; Reds were reluctant to enter traps used by Greys and their ranges tended to overlap less with Greys than with other Reds. It should also be mentioned that reintroductions are typically hampered by International Union for the Conservation of Nature (IUCN) directives, which require that the reasons for the original extinction be fully understood and that said circumstances have been changed for the benefit of the species in question!

Coupled with reintroductions, is promotion of existing populations by supplemental feeding and, while additional feeding stations have been shown to increase over-wintering body condition (and consequently fecundity), they have also been linked to increased mortality on the roads.

Habitat Management: Several woodland regeneration initiatives have been instigated over recent years in a bid to improve the condition of and to re-seed Britain’s woods and forests; these include Community Forest and National Forest projects. This is such a wide topic that those wanting to know more should check out the excellent sites by the Forestry Commission and the Woodland Trust.

Grey Control: Intuitively, controlling numbers of or extirpating Grey squirrels would seem to be the most productive method of promoting re-establishment of Reds. Previous control/eradication programs (for example, five shillings per tail paid by local authorities during the 1950s) have, however, been rather unsuccessful and Grey populations are as healthy as (if not healthier than) ever. More recent attempts at eradication have lead to protests from animal welfare organisations.  Such groups have been a particular thorn in the side of biologists striving to eradicate Greys from northwest Italy. In June of 1997, Italy’s National Wildlife Institute was taken to court by three animal rights groups under charges of illegal hunting, damage to state property and cruelty to animals. The two officers concerned were found guilty of cruelty to animals, although they were subsequently acquitted by the Appeal Court three years later. In Britain, Grey squirrels are shot as vermin by landowners and by conservationists in areas where Greys are encroaching on Red populations. Several surveys have been conducted in order to assess public opinion towards culling programs. The results of one questionnaire sent to organizations and private individuals across the UK who had expressed an interest in squirrel conservation and management were published in the journal Environmental Management in 2002. The responses showed that trapping was the most acceptable method of control, while poisoning was seen as the least acceptable. When the respondents were questioned about their views on immunocontraception (i.e. sterilizing squirrels) as a potential control method, most considered it more humane or acceptable than any of the other methods.  In January 2006 the Department for Environment, Food and Rural Affairs (DEFRA) announced that Grey squirrels in England were to face a widespread cull in a bid to protect the dwindling populations of Reds.  As was expected, public views of the cull are divided. Interested parties can check out DEFRA’s plans, while Animal Aid (opens a PDF in a new window) summarise the anti-cull angle and the Friends of Anglesey Red Squirrels summarize the pro-cull debate.

In Conclusion…
I think that if we take nothing else from these data, we should see that no single hypothesis can fully account for the decline in Red squirrels; even when taken in concert, there are still some situations that fall outside the explanation. Furthermore, without a thorough understanding of all the factors involved in this decline, these factors cannot be resolved and any efforts to resolve the problem could be viewed as superficial, if not futile. At the same time, however, many consider humans to have a duty of care to Red squirrels, arguing that because we are at least partially (if not wholly) responsible for their decline, we are similarly responsible for their reinstatement. As always, it is necessary to assess all the evidence available to us when making these decisions and therefore ensure that knee-jerk reactions are avoided. There seems to be a distinct need for a long-term management plan of the Red squirrel population in the UK to be presented, before any plans for a wide-scale cull of Greys is implemented. After all, it would be rather ironic to “save” the Red from the Grey, only to drive it back to the brink of extinction in some other way. Whatever the correct path, it is to be hoped that any action has not (or will not) come too late to save a creature so indicative (if only historically these days) of the English countryside. (Back to Menu)

Q: What controls the cacheing behaviour of squirrels and how do they find their buried nuts?

Short Answer: Caching appears to be an innate behaviour possessed by squirrels and involves hiding food for later retrieval. How carefully and where about a squirrel chooses to bury a nut is influenced by a variety of landscape features, the type of food object (some are more perishable than others) and the whereabouts of other squirrels - a squirrel will make false caches if it thinks it's being watched.  Relocation appears to be through good spatial memory; the squirrel remembers the location (probably relative to landmarks) and scent may guide it to the cache over the final few centimetres.

The Details: Through the years, many biologists have wondered how squirrels are able to retrieve their buried stores and the studies that have been born from this question have revealed some impressive information about how, where and why squirrels cache surplus food.

Caching is the process of storing (or, if you’ll excuse the pun, ‘squirreling away’) food that you don’t immediately need to consume. The purpose of this behaviour (and the associated surplus killing that we see in many carnivores) is to provide the cacher with a larder from which it can acquire food when conditions are harsh. In evolutionary terms, for caching to persist as a behavioural trait, the benefits (i.e.  food when times are hard, such as during the depths of winter) must out-weigh the costs (i.e. time taken to find more food, time taken to bury it, doing all of this instead of looking for a mate or keeping an eye out for predators). Moreover, for caching to be adaptive, the cacher must have a greater probability of getting to its own caches before they’re discovered by someone else; alternatively, the cacher could -- as is known in some birds, rodents and insects -- share caches with relatives (so-called communal caching). The upshot of this is that ecologists can say caching provides both direct and indirect benefits.  Direct benefits are those that the cacher actually sees (e.g. buried food increasing survivorship during the winter), while indirect benefits impact the bloodline (e.g. sharing food with your brothers and sisters helps improve their survivorship and therefore increases the likelihood that at least some of your genes will make it to the next generation).  Caching can also provide more short-term direct benefits, because if the cacher can remember where the goods are stashed, they don’t have to spend as much time searching for food and can use this saved time for other activities, such as finding a mate or raising offspring. Indeed, the fact that even during the spring, when food is typically abundant, squirrels may return to their caches seems to add some weight to this idea.

Now we know why animals cache food, what determines how and where squirrels cache surplus nuts and seeds? It has to be said that the vast majority of studies on sciurid caching have been conducted on Grey squirrels (Sciurus carolinensis), North American Red squirrels (Tamiascurius hudsonicus) and Ground squirrels (Spermophilus spp.) – I am not aware of any studies that have systematically documented the cache retrieval of Eurasian Red squirrels (Sciurus vulgaris).

Many theories have been put forth to try and explain what affects whether a squirrel eats or caches a food item and also what factors determine where abouts the squirrel buries it. One very interesting theory suggests that squirrels decide whether to bury a nut or seed on the basis of its tannin content. Tannins are polyphenolic (i.e. contain carbolic acid) chemicals found in plants; they have a very bitter taste and tend to shrink or constrict biological tissues (i.e. they’re astringents). It is generally considered that tannins have evolved as a defence mechanism in plants because they can bind and precipitate proteins and carbohydrate molecules, causing serious problems for animal tissues. One particular group of tannins (the gallotannins) are metabolised to form gallic and tannic acid in the rumen of herbivores; tannic acid is well known to cause ulceration of the digestive tract and renal failure. Consequently, some consider that burying foods with high tannin content, such as acorns, may allow some of the tannins to leach out, making the seed less toxic. Indeed, studies have shown that leaves buried by North American pikas (Ochotona princeps) lost a significant amount of their tannins, while Meadow voles (Microtus pennsylvanicus) have been observed to cut conifer branches and lay them in the snow for several days before eating them; during this time, the level of tannins in the branch was observed to decrease to that of the vole’s preferred foods. In squirrels, a study by Peter Smallwood and David Peters published in the journal Ecology in 1986 reports that, when presented with food items identical in every respect except for their tannin content, Grey squirrels ate those with high tannin content for significantly shorter periods than those with low tannin content.  Perhaps more importantly, the biologists (both at Ohio State University) found no evidence that tannins had any significant adverse affect on the squirrels’ ability to digest protein, which would make Sciurus carolinensis the only vertebrate currently known to be able to detoxify tannins. Smallwood and Peters suggest that squirrels use the tannins as cues to determine the perishability of the acorn, basing their decision on whether to cache or not on this factor. A more recent paper to the journal Ecological Research by Takuya Shimada at the Kansai Research Center in Japan, however, found no significant changes to the tannin astringency of Sawthorn Oak (Quercus serrata) acorns or horse chestnuts (Aesculus turbinata), even after three months of burial. Shimada suggests that the physical properties of acorns (including their smaller surface-area-to-volume ratio than things such as branches and leaves) may make leaching-out of tannins difficult.

Still, the idea that squirrels may be able to judge the perishability of a nut or seed (sometimes referred to as the “Perishability Hypothesis”) fits nicely with a 1996 paper to Animal Behaviour, which suggests that the perishability of an object may be more important to a squirrel than its handling time. In their fascinating paper, Leila Hadj-Chikh, Michael Steele (Wilkes University, Pennsylvania) and Peter Smallwood looked at how Grey squirrels cached two different types of acorns: white oak (subgenus Leucobalanus) and red oak (subgenus Erythrobalanus). These two acorns have considerably different germination schedules; red acorns undergo a period of winter dormancy before commencing germination in the spring (i.e. they have a low perishability), while white acorns germinate shortly after they mature in the autumn (i.e. they have a high perishability). The biologists found that, regardless of the handling time (i.e. the size of the acorn), squirrels consistently cached red acorns, while consuming the white acorns upon finding them. Moreover, when the squirrels were seen cache white acorns, they excised the embryo before burying it; in acorns, the embryo controls maturation and by removing or killing the embryo, the squirrels can decrease its perishability. In other words, if the squirrel removes the embryo, the acorn will last for longer and they can thus prolong the life of their cache. These results concur with previous experiments, which have also shown that squirrels will eat acorns infested with insect larvae, while caching those that are not – although it is not known whether this relates to the squirrels being aware that insect infestation makes the acorn more perishable, or whether the larvae simply offer a welcome additional source of protein.

Another theory as to what determines whether a squirrel buries or eats a nut is the “Consumption Time Hypothesis”, which as the name implies, proposes that the handling time of an item affects whether it should be eaten or stored. Generally, the larger an item, the longer its handling time – that is to say that big items are more difficult to manipulate, carry and require a bigger hole to be dug than smaller ones and so, overall, take longer to cache. This idea has been largely championed by Lucia Jacobs of the University of California Berkeley who, in a 1992 paper to Animal Behaviour, wrote:

“I hypothesized that a squirrel could increase its foraging efficiency by always choosing the behavioural options, eating or cacheing, that was least time consuming. If one food item took longer to eat than another, and if a squirrel could cache an item in less time than it took to eat it, then a squirrel should preferentially cache those items that took longer to eat.”

Prof. Jacobs tested this theory on five hand-reared adult male Grey squirrels and found that, when deciding whether to cache or eat an item, the squirrels seemed to do whichever took the least amount of time. In direct contrast to the results of Hadj-Chikh et al. (above), Jacobs found that the relative perishability of the food item was less important to the squirrels than the handling time – the squirrels consistently cached the chow blocks she offered rather than the shell-less hazel nuts, despite the fact that the chow block disintegrate quickly when buried in damp ground (i.e. have a high perishability). Jacobs concludes that because the chow blocks took longer to eat than the hazel nuts, the squirrels cached them, even though they had previous experience trying to retrieve buried blocks. Perhaps more interesting still was an article to the journal Natural History, in which Jacobs described how squirrels are apparently predisposed to cache nuts; she wrote:

“The squirrels performed flawlessly from the first day. I was fascinated as I observed one of these miniature squirrels pick up a hazelnut for the first time, search intently for a suitable burying site, and then, with great zest, dig a hole, both paws flying, the nut firmly clenched between tiny teeth, with all the apparent confidence and success of a jaded park squirrel burying its millionth peanut.”

Having touched briefly on some of the possible factors affecting why and how squirrels cache their food, we arrive at the question of where they cache – in other words, what determines how far the squirrels take their nuts? This question takes us back to the idea that squirrels may obtain both direct and indirect benefits from burying their food for later consumption. Indeed, one hypothesis to explain the positioning of caches has been the “Communal Cache Theory”, where squirrels may cache food in an, often centrally-located, area that is well known to other members of their family (for a synopsis of this phenomenon across the animal kingdom, I’d recommend Stephen Vander Wall’s book, Food Hoarding in Animals, published in 1990). In their recent paper to the journal Ethology, Mark Spritzer and Baniel Brazeau investigate this possibility. They considered that, if Grey squirrels cached primarily to gain indirect benefits, then related individuals should live near each other (i.e. they should see kin clustering) so as to facilitate cache sharing . Despite similar studies on Grey and Ground squirrels that support this hypothesis, however, their data showed only very weak evidence of kin clustering, with males more likely to live close to related individuals than females. Instead, they found that squirrels moved nuts toward the centre of their territories, grouped their caches and buried nuts further from their source when competitors were around – these findings strongly support the idea that squirrels cached for direct benefits only, because such behaviours make it easier for squirrels to defend and remember the location of caches. Previous observations that squirrels scatter hoard their food (i.e. bury small amounts in many caches that are spread around, rather than stockpiling it in one hole) and will aggressively defend caches against interlopers add further weight to idea that they cache for direct benefits.

Spritzer and Brazeau’s observation that squirrels took their nuts further away when competitors were nearby implies that there might be more to caching decisions than simply how perishable the item is or how long its handling time is. Indeed, a recent study by Japanese biologists Noriko Tamura, Yuko Hashimoto and Fumio Hayashi yielded similar results to those of Spritzer and Brazeau, in that the further away Japanese squirrels (Sciurus lis) took walnuts (Juglans airanthifolia), the less likely they were to have their cash pilfered by a competitor (primarily the Large Japanese woodmouse, Apodemus speciosus, in this study area).  Indeed, it has recently been established that squirrels make 'fake' caches when being watched in order to try and throw potential pilferers off the scent and that they consider other squirrels a particular threat. In a series of experiments at the University of Exeter campus in Devon, Lisa Leaver and her colleagues found that their Greys took evasive action (i.e. increased the distance between caches and cached with their back to the observer) more often when another squirrel was watching than when a magpie or crow was watching.

The way in which squirrels see each other (in terms of competitor or potential pilferer) is not always straightforward and appears to vary according to the quantity of food available. In a fascinating series of experiments by researchers at the University of Exeter, carried out between May 2006 and July 2007, it was found that their Grey squirrels responded to the presence of other squirrels at their feeding site by adjusting their caching behaviour. The biologists found that, when the squirrels were alone, they took their food off and cached it at distance from the feeding site. When other squirrels were present, however, each squirrel returned to the feeding site more often and took their food shorter distances (presumably so they could return more quickly) to cache it. The biologists concluded that the squirrels saw each other as competitors rather than potential pilferers – so they were more worried about the other animals taking the food that was there, than noticing where the caches were being buried. As the amount of food on the feeding site decreased, however, items were cached at increasing distance from the site presumably to reduce the likelihood of the other squirrels finding and pilfering the cache.

It may seem something of an assumption that squirrels know what each other are doing, but cache pilfering is a well-documented phenomenon among squirrels (indeed, among many species) with pilfering rates in the literature ranging from 1% to 95%. In order to see whether squirrels were able to gain information about what other squirrels were doing around them, and use that information for their own benefit (a phenomenon known as social learning), biologists at Exter University offered Greys on their campus a choice between two identically-looking pots - the subject was made to watch another squirrel empty one pot, before being given the choice of which to open and were rewarded (with food) if they chose the opposite one. In a second set of experiments, squirrels were asked to choose a pot, the full ones of which were marked with a piece of card (no squirrel to watch this time). The researchers found that the squirrels performed better (i.e. chose the correct pot) more often when there was another squirrel involved and concluded that squirrels were more efficient at using cues based on conspecifics than inanimate cue cards. In the wild, it seems likely that the ability to assess feeding opportunities based what other squirrels in the vicinity are doing would be a significant advantage.

The foregoing helps us understand why squirrels are careful about when and where they make their caches, but how does a squirrel decide where to bury a nut and, moreover, how does it find it again afterwards?

Provided that the entire area is suitable for caching, the “Optimal Density Theory” predicts that a squirrel should use the entire 360-deg arc around the resource – this should ensure that caches are placed at a density that prohibits the majority of naive competitors from finding them (think how much more difficult your grocery shopping would be if, rather than putting all the fruit and veg together, the supermarket spread it evenly around the store). There is little evidence, however, to support this concept and many studies have reported that, far from using the entire 360-deg area, caching squirrels use only a small percentage of it. It seems that the most likely reason for caching in a small area (or clustering your caches) is that it improves the likelihood that you’ll be able to find them later on – logically it seems reasonable that the more difficult you make it for others to find your caches, the more difficult you’re likely to find it to recover them several weeks down the line. Consequently, if you put all your caches in a small area around, say your den site, all you need to do is remember the general location. Scatter hoarding your food still means that any competitor is going to have to spend longer exploiting the resource than it may find viable – hence we see a trade-off whereby scatter hoarding my help to offset the costs associated with clustering your caches.

When a squirrel arrives at its preferred cache location, the burying behaviour seems to be rather stereotyped. In her 1989 paper to Natural History, Jacobs described how, after removing the fragrant husk of a hickory nut (presumably to make it less detectable to a competitor), the squirrel used its front paws to dig a hole one-or-two inches deep, into which it forcibly ramed the nut, hitting it with his/her teeth and putting its whole body behind the task. Once the nut was firmly in place, the squirrel set about covering it with dirt before taking great care to replace the leaf litter.

So, how do they find their buried treasure in times of need? Well, much work as been conducted in this area and they all point to much the same answer: even though recovery rates are highly variable (from 26% to 95% of nuts recovered, depending on mast crop), squirrels have good spatial memory that allows them to remember where they have buried their nuts. When it comes to looking for a cache, ethologists (i.e. those who study animal behaviour) recognise three main behaviours: Learned Cache Retrieval (LCR); Reforaging (Rf); and Search by Rule (SbR). Very briefly, if an animal uses episodic memory (i.e. a memory associated with a specific experience) to locate a cache it is said to be using LCR; if it buries the food within the home range simply because it is likely to be easier to find there than anywhere else, then no special mental abilities are required and it is Rf; finally, if an animal has rules (e.g. always cache under a rock) then it can exclude all sites that don’t conform to this rule and is thus considered to be employing SbR cache retrieval. As one might imagine, however, LCR is rather difficult to quantify because it probably exists in all types of cache retrieval.

The question of whether squirrels remember where they have buried their nuts has been banded about in the literature since 1884 and, based on the data we have now, it seems that squirrels have a very high capacity to remember where they’ve buried their nuts. This seems especially evident when one considers that several studies on sciurid caching behaviour have failed to find any evidence to support the supposition that squirrels bury their nuts next to local landmarks, which may then have served to help them find the caches more easily.

In a paper to Animal Behaviour in 1991, Lucia Jacobs and Emily Liman report that, even when squirrels buried nuts in areas where other squirrels had also cached, they retrieved significantly more of their own caches than they did caches of others, even 12 days later. Given that a squirrel was more likely to retrieve its own cache even when the cache of another was closer to it, lead the biologists to conclude that it was working on the basis of spatial memory, rather than odour – although the observation that squirrels always excavated at least one nut from someone else’s cache implies that olfaction certainly can be used to locate buried nuts. Indeed, a study by Denise McQuade (currently at Skidmore College in New York) and colleagues in 1986 found that Grey squirrels had a specific cue hierarchy when looking for caches – in their experiments involving coloured food dishes, they found that the first cue was the location of the dish, then its colour and finally its odour. Furthermore, they even found some evidence to suggest that the squirrels may also remember the type of seed in the cache. Jacobs and Linman go on to report how the squirrels moved from one cluster of two-or-three caches directly to another, with little retracing of their path; this suggests that they can remember a series of locations in relation to each other and use this information to build what the biologists refer to as a “cognitive map”, into which they can encode information on the location of each cache. Finally, the squirrels in this study were seen to dig-up and re-cache some nuts, suggesting that -- in the wild -- cache husbandry might allow them to re-cache hastily-buried nuts and perhaps helps to maintain the all important optimal dispersal that is known to reduce the likelihood of cache pilfering.

To accompany the behavioural observations that squirrels can remember where they bury their nuts, we now have some neurological data. A recent paper to the journal Genes, Brain and Behavior by neurologists at the University of Toronto in Canada documents three times the density of proliferating cells in the dentate gyrus of the Grey squirrel than in the Yellow-pine chipmunk (Tamias amoenus – a small rodent that employs larder caching). The dentate gyrus is an area of the hippocampus (the part of the brain that controls memory and navigation) and represents one of the few areas of the brain in which neurogenesis (i.e. the creation of neurons) is known to occur. The paper also describes the presence of more cells containing the Ki-67 protein (which is associated with the breakdown and reformation of nuclei during neurogenesis) in squirrels than chipmunks. Finally, they found that adult squirrels didn’t undergo the same decline in neuron survivorship with age that chipmunks did. Overall, the neurologists suggest that maintaining this pool of young neurons as they grow older may be necessary for their spatial (i.e. cache retrieval) memory. Moreover, the observation that young squirrels had increased neuron proliferation and young neuron densities than adults, may reflect that juveniles have a greater need for learning (i.e. becoming familiar with their environment and remembering where their food it buried) than adults. Indeed, similar neurological data from Lucia Jacobs seem to support this idea. In a recent paper to the European Journal of Neuroscience, she reports that, in the autumn and spring (when Greys are actively caching and occasionally retrieving their nuts), the squirrels showed a 15% increase in hippocampus size compared to the rest of the year. These data add further weight to the idea that squirrels rely heavily on memory when caching their food.

Squirrels seem to have great potential for recalling the locations of their caches, although it is important to remember that not all caches are excavated and it is highly unlikely that any given squirrel will retrieve all of the nuts it has buried. Consequently, squirrels are considered to play a very important role in the dispersal of trees and in the regeneration of forests. In North America, for example, the Grey squirrel is probably the most important animal for helping red and white oaks to disperse; owing to the differences in the perishability of their acorns, reds tend to disperse further than whites and are likely to be the first oaks to colonise a forest. Here in the UK, it is ironic to think that an introduced species may play a similarly important role in regenerating our native woodland! (Back to Menu)

Q: Why are some squirrels of the same species different colours?

Short Answer: Genetics. The animal's genes code for the colouration of the fur. The differences in colouration are the result of varying amounts of a pigment called melanin that is laid down in the hair as it grows. Grey squirrels are a mixture of several different hair 'patterns', while black and white individuals have varying amounts of this pigment. It is important to recognise that, although different colours, they are not different species. Thus, black squirrels are simply black grey squirrels.

The Details: The short answer is: “genetics”. The coat colouration of squirrels, as for all mammals, is under the control of the animal’s genes – it is the genes that stipulate the colours and patterns in the fur. To understand this a little better it is necessary to take a brief foray into the fascinating world of genetics, embryology and physiology.

Genetics 101
As adults, our bodies are made up of about 10 trillion individual cells, most of which contain 46 strands of deoxyribonucleic acid – a long chain of building blocks, the name of which is frequently (and understandably) shortened to DNA. These 46 strands are called chromosomes and you inherited 23 from your Dad and the other 23 from your Mum – within these chromosomes are the ‘blueprints’ for building another you. Chromosomes are -- figuratively speaking -- divided into individual points, or sections, called genes. These genes contain the details (i.e. the ‘code’) for how to create proteins, which go on to form your tissues and control many biochemical processes within your cells. Not only do genes code for proteins, they also dictate your physical appearance – your genotype (i.e. all of your genes taken together) is largely responsible for determining how you look (geneticists call your appearance your “phenotype”). Now, a given gene may have several different forms (variations on a theme, if you will), each of which may produce different features or lead to different processes; different forms of the same gene are called alleles (differences arise through changes, called “mutations”, to the gene’s code) and some of these alleles have a greater influence on the body than others.

To illustrate the above I will use a very basic, green fingered, example based on the pioneering work of the late, world renowned geneticist, Gregor Mendel. Let’s say you’re growing a sunflower; how tall your plant grows will depend on its genes – for our basic example, we’ll assume that the plant will either be tall or short. If a plant is tall, which we’ll assume is the normal condition, it has the ‘tall allele’ (we’ll abbreviate this to “T”); if it’s short then it has a mutation of the tall gene, the ‘short allele’ (“t”). The tall gene is dominant over the short gene, which means that where the two are present, it should win out – so either combination involving T (i.e. Tt or TT – remember, there are two sets of chromosomes, one from either parent, so one parent gives one T, while the other gives another T) will cause the plant to grow tall. The plant will be short only if both parents give it the short (recessive) allele – i.e. its genetic combination is tt. In cases where one parent contributes a dominant allele and the other provides a recessive one (i.e. Tt), the individual is called heterozygous (from hetero- meaning “different”). Where both parents provide the same allele (i.e. TT or tt) the individual is homozygous (homo- meaning “the same”).

Why have I gone over all of this? Partly to introduce you to the terminology that we’re going to need to explain the different coat colours of squirrels and partly to make a point. The term mutation is frequently associated with somewhat macabre connotations. In genetic terms, a mutation is simply a change to the genetic material; the change could be for the better, for the worse, or it could have no impact one way or the other. So, when we talk about mutations leading to the differences in fur colour, we’re not talking about a comic book style Radioactive Man (or in this case squirrel!), instead we’re referring to a change to the gene which has caused a different colour or pattern to be produced.

Squirrel genetics
Some of the first studies looking at the genetics of coat colour in (Red) squirrels were done by German biologist Herbert Wiltafsky as part of his Ph.D thesis at the University of Köln in the early 1970s. During his captive breeding studies, Wiltafsky found that the colour of the fur on the lower legs and feet is determined by a single gene with a dominant ‘red’ and recessive ‘black’ allele. Wiltafsky also reported that the colour of the tail fur wasn’t the result of a single gene, rather it was probably polygenic (controlled by several genes) – it seems that foot fur colour isn’t related to either the colour of the back or the tail. Assuming this is correct, it goes some way to explaining the rather bizarre-looking, and very rare, piebald squirrels (pictured left).

Work on various aspects of squirrel genetics (typically related to their taxonomy) is underway at several institutions; research specifically into white squirrels is being conducted at several colleges and universities in America. Even today -- some 35 years on from Wiltafsky’s studies -- structured captive breeding studies aimed at yielding empirical data are frustratingly rare and there is still much to be learnt about the genetics of coat colour polymorphism in squirrels. We have, nonetheless, made some important step forward in recent years. We now know that the MC1R gene plays an important role in fur colouration in vertebrates and that there are two crucial components: the MC1R receptor (note lack of italics) and the ASIP protein (sometimes referred to as agouti protein). Fortunately, we don't need to know what these components are, or what they acronyms stand for to get an idea of how colour changes come about.

Colour by numbers
Mammals possess a fairly small selection of colour pigments; perhaps the most well known of these are the carotenes and melanins. In terms of fur colour, it is the melanins that interest us.  Melanin can be divided into two main types: Eumelanin, which is black or brown, and Phaeomelanin, which is paler, ranging from red through to yellow (there is technically a third, Neuromelanin, but this is a dark pigment found only in some brain neurons, so it need not concern us here). The cells that secrete melanin are called melanocytes and the MC1R gene is responsible for regulating the production of melanins by these melanocytes. So, the animal’s genes are responsible not only for controlling the production of melanin, but also for regulating where and how (i.e. clumped or spaced out) the melanocytes are situated within the skin. Differences in melanin production (either in the melanin itself, or the pattern of its production) lead to differences in the colour, and distribution of colouration, of the fur. The white underside of many mammals, for example, occurs when melanocyte activity is suppressed.

First things first: how does hair become pigmented in the first place? Hairs start life as a hair bulb, produced in small sac in the skin called a follicle. How the follicle forms is remarkably complicated (and controlled by many different genes), but the essence is that a hair shaft is formed within the follicle, growing up and out of the pore – the process of follicle generation and hair production is called “anagen”.  As the hair grows, the follicle’s pigment cells (called neural crest melanoblasts, we shall see why later) secrete melanin, which is deposited in the hair shaft and mingles into the hair’s matrix (i.e. into the inner layers). The end colour of the hair depends upon the type and pattern of the melanin within. As the hair grows, the melanocytes are switched 'on' or 'off'' as specified by the corresponding component (i.e. the MC1R and/or ASIP). I won't go into the biochemistry, but basically the type of hormone that activates the MC1R gene dictates whether eumelanin or phaeomelanin is produced by the melanocyte. Pulses of melanin production as the hair grows lead to a banded appearance. The existing data we have on squirrel genetics suggests that it's the MC1R gene that plays the most significant role in squirrel coat colouration, although in other mammals (and perhaps to a lesser extent in squirrels too) the agouti gene is in control. Scientists at the Oak Ridge National Laboratory in Tennessee, for example, have discovered that mice born without the agouti gene are totally black, while those born with a mutated allele (the non-agouti gene) that is ‘always on’ are totally yellow. The reason for this is that when the gene is ‘on’ it causes the secretion of the agouti protein, which stimulates the melanocytes to produce phaeomelanin (leading to a pale yellow band in the hair), when it’s ‘off’ no protein is present and the melanocytes return to producing eumelanin (so a dark band is deposited).

In tree squirrels (those of the genus Sciurus), we see considerable variation in coat colour; both Grey and Red squirrels (S. carolinensis and S. vulgaris, respectively) exhibit white and black forms (called “morphs”) as well as their wild-type (normal) colouration. So, can we be a bit more specific as to what causes these different pelage colours? Well, geneticists consider that all variation from the wild-type arises through mutation (i.e. changes to the animal’s genes) and, although the precise mechanism(s) of melanin deposition in squirrel hair is still unclear, it seems probable that it follows the process observed in other mammals.

Schematic representation of grey squirrel fur types. Wild-type (grey) squirrels get their colouration from a combination of six of the seven hair types, while the brown-black morphs have four and the jet-black morphs (not shown) have only black hair. Black represents the dark eumelanin pigment, while grey represents the lighter phaeomelanin and white indicates little or no pigment present. Diagram based on that published by Helen McRobie, Alison Thomas and Jo Kelly in their 2009 paper to the Journal of Heredity.

In order to best understand the occurrence of black and white morphs, we first need to understand what makes wild-type squirrels grey. Recent research by biologists Helen McRobie, Alison Thomas and Jo Kelly, at the Anglia Ruskin University in Cambridge, has provided an insight into how squirrel colouration works at both a physical and genetic level - their findings were published in a paper to the Journal of Heredity during 2009. The researchers collected hair from the back, sides and belly of 34 squirrels representing all British colour morphs and subjected them to physical (i.e. microscopic) and genetic analysis. It transpires that the each hair could be categorized into one of seven distinct groups (or 'types') according to the pattern of pigmentation (from all black, through various bandings to all white - see above) and each colour morph had differing amounts of each type. Wild-type grey squirrels have six of these hair groups that together provide the grizzled-grey appearance - in other words, the normal grey fur is actually a subtle blend of six different hair types, each with a different pattern of eumelanin and phaeomelanin pigmentation.

Black/Melanistic Morph: Black morphs of both Red and Grey squirrels are known, although black Red morphs are very rare in the UK and when most people see a black squirrel, it is a black morph of our ubiquitous Grey. Until recently, black Grey squirrels (hence forth, just "black squirrels") were something of an enigma and, even now they are becoming increasingly common in Britain, they still draw great public interest.  Black squirrels are relatively common elsewhere in their range, in North America particularly, and some towns even have them as mascots. The first black squirrel to be recorded from the wild in Britain was seen in the small town of Hitchin, north Hertfordshire during 1912 and these enigmatic animals appear to have spread north and eastward since; they are now relatively common in counties such as Cambridgeshire and Bedfordshire, where they are at least as common as wild-type (grey) squirrels. Quite where these black individuals came from is uncertain, although genetic data collected by the researchers at Anglia Ruskin University suggest that the initial animals were escapees from zoo collections imported from the USA. Before we go on to look at the reason these squirrels are black, it is important to be clear that these are not a different species to the normally-coloured grey squirrels: they are simply a different coloured 'version' of the grey squirrel.

Jet-black squirrels -- we will come on to browns, in a minute -- have only one fur group (pure black) and are the result of hypereumelanogenesis – in other words, there is an excessive production of eumelanin that makes the entire hair black. Based on ORNL’s mouse studies, it seemed feasible that an entirely black appearance could be caused either by a missing/defective agouti gene, or by an allele that’s constantly switched off. The recent work by Helen McRobie and her colleagues, however, suggest that it is a mutation of the MC1R gene that is responsible. During their study 2009 study, the biologists found that jet-black squirrels had a section of the MC1R gene missing. In genetic terms, these black squirrels had a 24 base-pair deletion - the specifics of this aren't important to understand here, but it basically means that the MC1R protein of black squirrels is shorter (by eight amino acid sub-units) than that possessed by the wild-type morph. This mutant MC1R gene 'locks' the melanocytes so they only produce the dark eumelanin pigment.

Not all 'black' squirrels are actually black, even though it may appear to at first glance – some are actually brown-black in colour. The Cambridge biologists found that these brown-black squirrels had four of the hair groups in various combinations on their back, sides and belly. The reason for this particular morph relates to what geneticists refer to as incomplete dominance. Recall the 'tall versus short plant' example, where one allele was dominant over the other and wherever it is present it is expressed - this is fine for cases where dominant and recessive alleles meet, but what happens when you get two different dominant alleles? The result is that there can be a mixing of phenotypes to produce an intermediate result. The genetics can be rather complicated, but the principle is straightforward. Think of mixing paint, where alleles for yellow (Y) and blue (B) were both dominant; where the two occur together (YB) neither is dominant over the other and so they each express their characteristics, giving you green paint. This is a basic example, but the same principle occurs when a wild-type (grey) and jet-black squirrel mate: they produce a brown-black kitten. So, the black gene is incompletely dominant because you need two jet-black parents to be sure of a jet-black kitten. If a jet-black and brown-black squirrel (or two brown-black squirrels) mate they have a 50% chance of producing jet-black kittens, while a mating between a brown-black and wild-type animal could produce only wild-type or brown-black kittens (50% chance of each).

In Britain we tend only to see wild-type, jet-black, black-brown and white morphs but, in North America, some black squirrels have 'frosted' appearance, while others were more subtly 'graded'. In their 1958 paper to the Journal of Mammalogy, Pennsylvania State University biologists William Creed and Ward Sharp divided their black squirrels into three groups.

Group 1 were most common in their study area of the Cameron County forests and exhibited black hairs with narrow ‘buff’ bands on their backs, which gave them a “brownish-black” appearance; their belly fur was predominantly a rusty-brown colour.

Group 2 morphs were jet black except for a scattering of silvery-white hairs on their back and tail; their undersides were also black.

Group 3 squirrels were totally jet black. This was actually the least common colour morph.

The biologists reported that there were squirrels that didn’t fit neatly into any of the above categories; one in particular exhibited both melanism and erythrism (red/ginger colouration), with black hairs on its back having a red tip (giving an overall reddish-black appearance) and a completely red tail! Similarly, during a study of squirrels in the Italian Alps, biologists found ‘red’, ‘brown’ (back and tail dark brown, lower legs and feet red or reddish-brown) and ‘black’ morphs of S. vulgaris. Indeed, it is worth noting that even within wild-type morphs of all squirrels there is colour variation, with some being paler than others. Interestingly the Italian scientists note that, in their squirrels, regardless of the morph the underside was always white suggesting that the situation may be more complex than a single defective or mutated gene – different genes may be responsible for regulating back and stomach fur colour.

White Morph: There are three reasons for a squirrel being white; it can be albino (left), leucistic, or it can be what I shall call a ‘white mutant’. It is important to make the distinction between these conditions, because the underlying genetic causes are entirely different.

The classification of albinism has changed considerably in recent years as the causes have become better understood. One of the most common types of albinism is “OCA1”, in which the sufferer possesses recessive alleles of a gene for the production of an enzyme called tyrosinase; the result is that the tyrosinase enzyme in the melanocyte doesn’t work. Melanin is the end product of a rather complicated biochemical pathway, the starting point of which is the oxidation of the copper-containing amino acid tyrosine – this oxidation requires tyrosinase in order to happen. Without tyrosinase the body cannot make melanin and the skin and hairs lack pigment, appearing white; the eyes also lack pigmentation and the blood vessels that are normally obscured by the melanin are visible, giving the eyes a red appearance.

Leucism, on the other hand, is a form of hypopigmentation – a rare, presumably recessive, gene prevents melanin deposition within the hair. In leucistic animals, the melanocytes are missing from the area altogether (as opposed to the albinos, which have the melanocytes, but often can’t use them). Not only does leucism have a different anatomical profile to albinism, it also has an entirely different cause. In vertebrates, the pigment cells form from a group of cells that start out life lying along the spinal cord – this bunch of cells is called the neural crest. During development (and presumably under genetic control), the cells break away from the neural crest and migrate to various locations across the skin. In leucists, the cells fail to differentiate or migrate from the crest; this affects all pigment production, not just melanins. In some cases, some of the cells migrate, leading to patches with pigment and patches without – this is referred to as partial leucism.

Often the easiest way to separate an albino from a leucist is by the presence of red eyes. We have seen that albinos can lack the ability to create melanin in any of their cells, so their eyes are often unpigmented and appear red because of the haemoglobin in the blood running through the capillaries of the retina and iris. Leucists have normally-coloured eyes; the reason for this is rather complicated, but stems from the origin of the retina during development. Basically, the retinal melanocytes (i.e. the pigment cells of the eye) don’t come from the neural crest; as the embryo grows, a small pouch develops from the neural tube (that goes on to form the ophthalmic cup) and forms the retina. Consequently, because the retina’s pigment cells aren’t from the crest, they’re not affected if the cells fail to migrate or differentiate.

Conceivably, a third possibility is that the animal may be fully capable of producing melanin (so it isn’t an albino) and have melanocytes where it should (so it’s not leucistic), but is white because its genes prevent eumelanin being produced (or promote phaeomelanin production) across most (if not all) of the body. Given that the agouti gene is capable of making mice entirely black or entirely white/yellow, it seems possible a similar mutation in squirrels could yield similar results.

There are several populations of white squirrels found throughout the range of S. carolinensis (white individuals of S. vulgaris are comparatively rare), particularly in the USA. One well known population of white Greys can be found in Brevard County, North Carolina; in this population there are a number of individuals that are totally white, except for a distinctive dark patch on their head and stripe down their back. The dynamics, ecology and genetics of this population are being studied by Brevard College’s White Squirrel Research Institute. Interestingly, biologists at the WSRI report that populations of white individuals tend to pop-up, die out and then re-occur somewhere else; this implies that the white morph may be a spontaneous mutation of the genes that control neural crest splitting or melanocyte migration (so-called “regulator” genes), or that regulate the production of melanin.

More than skin deep?
Some scientists have questioned whether a given colour confers some benefit to a squirrel, or whether the gene that affects the coat colour also influences the animal’s behaviour – genes that control more than one feature are called “pleiotropic”.

During the early 1940s, biologists noticed that melanistic squirrels tended to occur at the northern extent of their geographical range and suggested that having denser, more cryptic (i.e. darker) fur may be an advantage over other colours in wet, dense spruce-fir forests. In genetics, there is a rule called the Hardy-Weinberg Principle, which states that if a particular genetically-controlled trait doesn’t cause disproportionate mortality, it will persist within a population. In other words, if a genetic trait provides the animal with an increased chance of surviving to reproduce, the gene(s) will remain in the population (if it doesn’t, the gene is quite likely to die with the host before it reproduces); in this capacity, a genetic trait is considered to be “adaptive”.

So, the simple observation that many populations of black squirrels seem to do well in the wild, suggests that the colour does confer some benefit. Indeed, there are more differences between melanistic and wild morphs than meets the eye. A study of S. vulgaris from Finland, by zoologists Paavo Voipio and Raimo Hissa at the University of Turku, found that black morphs had longer and denser undercoat fur than wild-types. Similarly, other research has shown that, at temperatures of -10-deg C (14 -deg F) of less, completely black individuals of S. carolinensis seem to experience (proportionally) almost 20% less heat loss, are more than 10% more tolerant to the cold and have a lower basal metabolic rate than grey morphs.

In recent years it has been suggested that black squirrels are larger and more aggressive than their wild-type conspecifics. There are, as far as I am aware, no data to support the claim that black morphs are physically bigger than grey morphs (although the aforementioned difference in fur density may account for black looking larger than greys) and the theory that black morph mutation is linked to higher testosterone levels has yet to confirmed. Studies on the behaviour of black and wild-type morphs have yet to find any differences. In a 1990 paper to the American Midland Naturalist, State University of New York biologists Eric Gustafson and Larry VanDruff report the findings of their study on the behaviour of black and wild-type Grey squirrel (S.  carolinensis) between February 1982 and March 1983 in Syracuse, New York. The researchers found that black and grey morphs were equally as wary of approaching humans and dogs and, at feeding stations, neither morph was dominant. Additionally, the scientists observed that both morphs sunned themselves in the same ways and for the same lengths of time. Overall, Gustafson and VanDruff conclude that behavioural differences can’t explain the distribution of the colours and neither morph is likely to have an advantage when it comes to mating.

Reasons to be … colourful
Taking into account that there don’t seem to be any differences in the behaviour of the colour morphs, the idea that melanism provides a selective advantage in cold, moist climates seems the most plausible explanation for the perpetuation of the black morphs (human selection aside). One common observation of colour morphs is that animals in humid environments tend to be more highly pigmented than those of the same species in less humid environments; this idea was first documented by Constantin Wilhelm Lambert Gloger in 1833 during his studies of bird plumage and is now referred to as Gloger’s Rule. Melanistic morphs of squirrels aren’t, however, found in particularly humid environments and it seems unlikely that Gloger’s Rule explains their distribution. Instead, it seems thermoregulation (the ability to maintain a body temperature above that of your environment) and camouflage may offer more plausible explanations.

During their study of Red squirrels in the Italian Alps, biologists at the University of Insubria and the University of Turin found that black morphs were indeed to be found the dense, moist conifer forests.  In their 2004 paper to Mammalia, the researchers wrote:

“We suggest that the combination of a denser and more cryptic fur in black morphs gives them a selective advantage over other coat colour morphs in wet, dense spruce-fir forests of the Italian Alps”.

Being melanistic may also make them more difficult for predators to spot and the biologists note that:

“… red morphs seem better camouflaged in mixed broadleafs. In contrast, black morphs are more cryptic [harder to spot] in dense conifer forests, particularly those dominated by species with dark-grey bark such as Norway spruce and fir, where they are more common than in less denser forests with more larch and/or pines."

Ultimately, in genetic terms, whether black morphs are better suited than wild-types to dense forests (or the northern extremes of their ranges) because they are better able to cope with cold, damp conditions, or because they’re less vulnerable to predators (or both!) is largely immaterial. The fact that having black fur is an advantage, in at least some circumstances (regardless of why), should be enough for the gene to remain in the population. So, why do we see black squirrels in urban areas, where it’s not particularly cold, or humid and they probably stand out more than, say, grey morphs? Well, it’s entirely possible that thermoregulation and crypsis don’t tell the whole story – there could be some other factor(s) involved that we have yet to identify. Nonetheless, if the black morph allele is a random, and/or homogenous, mutation (as is proposed for the leucism allele) it could theoretically pop-up anywhere. Moreover, a melanistic population would only fail to become established if there was some selective disadvantage to being black rather than grey. In other words, as long as black morphs do equally as well in urban environments as grey morphs (which appears to be the case), there’s no reason why they shouldn’t survive and reproduce, hence passing on the trait.

The idea of improved thermoregulation and camouflage over wild-type morphs in certain environments may explain why the melanistic populations persist, but what about white ones? White morphs tend not to be particularly common in wild populations; the reason for this is largely that they are very much easier to spot (they lack the camouflage conferred by the wild-type morph) and this increases their chances of being killed by a predator. In the case of albinos, however, being more obvious to things trying to eat you is just one of several problems.

Albinos tend to have poorer eyesight than their pigmented conspecifics. Just outside of the retina, there is a layer of pigmented cells called the Retinal Pigment Epithelium (or RPE for short) that serve to nourish (supply with blood) and protect its cells. In albinos, the lack of tyrosinase leads to the poor development of the RPE and light entering the eye -- which would ordinarily be absorbed by the melanin -- scatters, flooding the eye with light and dazzling the retina. Furthermore, rods (the visual cells on the retina sensitive to changes in light levels – used for scoptic or ‘twilight/greyscale vision’) require a chemical called dihydroxyphenylalanine (abbreviated to DOPA) in order to develop properly. Unfortunately for albinos, tyrosinase is needed to form DOPA and, as such, albinos tend to suffer both a reduction in the number of rod cells on the retina and a higher proportion of rods that are abnormally low in the visual pigment rhodopsin. Albinos may also suffer abnormal (i.e. simpler) connections between the retina and the brain.

It’s not difficult to see how being easily dazzled or having generally poor eyesight could be a disadvantage for an animal that spends much of its time jumping around in trees. Despite the problems, however, many observers have pointed out that white squirrels tend to retain much of their mastery of the treetops; if the individuals are leucistic or ‘white morphs’, this is to be expected (neither are known to suffer the visual defects found in albinos). Shouldn’t we, however, expect albino squirrels to be in serious danger of falling? The answer is “maybe not”, because albino squirrels may not be as prone to retinal defects as other mammals. In an interesting 1998 paper to the journal Vision Research, Glen Jeffrey (at University College London) and Jona Estive (at Barcelona University) found that the eyes of the albino S. carolinensis they looked at suffered only about a 5% reduction in the number of central nerve cells on the retina – this is in comparison to an average reduction of about 25% in most albino mammals. If Jeffrey and Estive’s results are representative (they only examined two albino squirrels), they may explain why albino squirrels seem to do pretty well in the wild.

Squirrels also have a water-soluble yellow pigment in the lenses of their eyes, which has two peaks of absorption in the ultraviolet: one at 265nm (UVC) and another at 370nm (UVB). In other words, this pigment screens out UVB and UVC rays, acting like a pair of sunglasses.  Moreover, if you’ve ever used yellow-tinted sunglasses, you may have noticed that they tend to increase colour contrast by removing or reducing chromatic aberration -- the “blue haze” well known by photographers -- caused by different wavelengths (i.e. colours) of light being focused at different points in the eye. This pigment was first documented, quite accidently in 1930, by Gordon Walls while he was dissecting a freshly-killed python. In a short communication to the journal Science some ten years later, Walls reported that the pigment was present in the lenses of the albino and normally pigmented Grey squirrels he studied. This yellow pigment seems to be present in albinos (which suggests it’s not a type of melanin), although it only appears to filter out UV light and as such it doesn’t offer much (if any) protection from the amount of visible light flooding and scattering in the eye. Consequently, it seems that an albino squirrel is still just as susceptible to being dazzled and suffering the retinal damage associated with uncontrolled light as other albino mammals.

So, if albinos are more susceptible to vision defects and, along with non-albino white morphs run a higher risk of being spotted by predators, does being white confer any advantages? Well, to the best of my knowledge, there are no data to suggest there is any difference in fur length or anatomy in white and wild-type morphs. Conceivably, given that tree squirrels don’t hibernate, in areas where snowfall is common during winter, a white morph might have a selective advantage over a grey or red morph (i.e. stand out less against a snowy backdrop). In many cases, however, the reason populations of white squirrels (albino or otherwise) persist has a more obvious cause: humans. Towns in the USA where white squirrels are commonplace (e.g. Brevard County in North Carolina, Marionville in Missouri and Olney in Illinois to name a few) tend to be fiercely protective of their pallid rodents – indeed the squirrels are veritable tourist attractions. This human protection means that, for a squirrel, being white is beneficial (grey morphs are even trapped and sent packing!) and this permits the persistence of the ‘white allele’ within the population.

In conclusion, we have seen that the colour morphs of squirrels are under genetic control; they’re a result of changes (mutations) to the gene(s) responsible for producing and/or distributing melanin in the body. In some instances these colours come with physiological and/or biochemical downsides (as in the case of albinism); at times they may make the animal more conspicuous to predators. Nonetheless, in certain environments some colours seem to be advantageous, so the traits remain. Where morphs exist outside of these environments, they do so presumably because either being a given colour doesn’t put them at a competitive disadvantage, or (as is the case with some populations of white morphs) because humans select for the colour. (Back to Menu)

Return to TOP