FOR several hours a pack of African wild dogs moves quietly through the
dense bush in Zimbabwe鈥檚 Hwange National Park. Suddenly the pack erupts into
action. Two dogs have started to chase a young kudu antelope which was
browsing alone. After a hectic chase, the lead dog leaps and grabs the young
kudu bull by its snout, avoiding the lancing of its short, sharp horns. The
kudu stops in its tracks, paralysed with shock. As it does so, another dog
swerves to the rear and tears a chunk of flesh from the hindquarters of the
animal.
Within 15 minutes, the nine dogs have devoured the 200-kilogram carcass. An
hour later all that remains is a few bones, an uneaten forelimb and a pile of
half-digested grass from the dead animal鈥檚 stomach. While death by
disembowelment may seem like a grizzly end for such a regal antelope, the end
is swift and sure. The African wild dog is an efficient hunter and fast eater.
The swiftness with which wild dogs kill and consume their prey, however,
parallels the swiftness with which the species is disappearing.
Once found from Senegal to South Africa and from the deserts to the top of
snow-capped Mount Kilimanjaro, in the past thirty years the dogs have all but
disappeared from 19 of the 34 countries in which they once lived. Populations
in nine of the remaining countries number fewer than a hundred. Only six
countries support viable populations, and in at least two of these, numbers
have halved in a decade. At the best estimate there are between 5000 and 6000
wild dogs in Africa. Numerically, the wild dog is as threatened as more famous
endangered species such as the black and white rhinos.
Advertisement
Rotten reputation
Among the vast array of Africa鈥檚 mammals, the wild dog has perhaps been
uniquely persecuted by humans. Long seen as vermin by farmers and wildlife
managers, they have been shot in their thousands. But their reputation owes as
much to their 鈥渕essy鈥� method of killing 鈥� disembowelment 鈥� as to the fact that
they occasionally prey on livestock.
The African wild dog is not closely related to other dogs, foxes, wolves or
jackals. It takes its name Lycaon pictus, (painted wolf) from its patterned
coat, an array of yellow, white, brown and black patches. Other common names
include hyena dog, reflecting that it has only four toes on each foot, and
Cape hunting dog. Weighing between 23 and 35 kilograms, the dogs are the size
of German shepherd dogs, but with leaner bodies and longer legs. They eat a
variety of prey, mostly medium-sized antelopes such as impala, Thomson鈥檚
gazelle, wildebeest and kudu.
Because the species is rapidly approaching extinction, conservation is the
top priority. But for this to be successful, scientists must understand the
causes of the decline and gather basic biological information about the
animals. Unfortunately it is a slow process accumulating information, mainly
because the African wild dog is so sparsely distributed. Lions, for example,
are often found at densities of up to one per 5 square kilometres. With wild
dogs, however, it is more common to find one animal per 70 or 100 square
kilometres. At their peak, population density may rise to one dog per 25
square kilometres but in dry areas, such as northern Botswana, it can drop to
one dog per 400 square kilometres.
Extermination plans
Another reason for the paucity of data is that the pack rather than the
individual is often the unit of study. The dogs live in a tightly knit pack,
consisting of an average of 8 to 10 animals. They work as a team in which each
member benefits when a large animal is killed for dinner, a group of pups is
successfully reared or a challenge from a lion or hyena is rebuffed. And each
pack member also encounters the same threats 鈥� disease, competition from other
predators and persecution by humans. But this also limits the available data:
while a study of wild dogs in seven packs may include 100 individuals, in any
given year there may be only seven sets of data. And to make matters worse,
individual packs range over a large area 鈥� typically between 250 to 2000
square kilometres.
Because wild dogs are so thinly dispersed, until recently no one knew for
certain where the existing populations lived or how many were left. In the
past six years, however, research projects have started in eight countries 鈥�
Kenya, Botswana, Namibia, South Africa, Tanzania, Zimbabwe and Senegal. This is filling in some of the gaps in
the genetics, behaviour and ecology of dogs living across Africa. We still
have little idea what has led to their decline, but with such large home
ranges, they are probably particularly sensitive to the loss and fragmentation
of their habitats.
Legal and illegal hunting of the species has also been a problem since the
white settlement of Africa. In the 1960s and 1970s, tens of thousands of the
dogs were shot. These extermination programmes were best documented in the
then Rhodesia: between 1957 and 1974, the government paid farmers and trappers
to kill more than 3000 animals.
High-speed roads are another killer. By the 1980s, roads had improved in
many parts of Africa, and as the speed of vehicles increased so did the number
of dogs killed. In some areas, such as the Mikumi National Park in Tanzania
and the areas bordering Hwange National Park in Zimbabwe, vehicles are killing
12 per cent of the adults and 20 per cent of the puppies a year.
Feast or famine
A different cocktail of factors, however, may lead to extinction in each
ecosystem. In the Serengeti National Park and the Masai Mara Game Reserve in
Tanzania and Kenya, an important cause of decline appears to be disease.
Elsewhere, as in South Africa鈥檚 Kruger National Park and in northern Botswana,
competition with other predators is more critical. Lions and hyenas may steal
the dogs鈥� kills or prey on their puppies. Gus Mills and Martyn Gorman of the
University of Aberdeen, working in Kruger, found that wild dogs appear to
avoid hunting in areas with large populations of lions, even if they do
contain abundant prey (快猫短视频, Science, 30 April).
What is becoming apparent is that information on wild dog behaviour and
ecology gleaned from studies in one ecosystem may not always be applicable to
other ecosystems. Wildlife documentaries shot in the open plains of the
Serengeti/Masai Mara ecosystem have provided most people with an image of
large packs of wild dogs relentlessly, and usually successfully, chasing prey
at high speed for many kilometres across the open plain. Early research in the
1970s also focused on wild dogs in this ecosystem. Most of Africa鈥檚 wild
lands, however, are covered in trees and bushes, and there is good evidence
that wild dogs living in bush behave very differently from those living in
treeless plains. The dogs use different hunting strategies, for instance.
A study in the late 1970s in Kruger National Park, and other more recent
studies in Zimbabwe, Botswana and southern Tanzania, show that shorter chases
of only a few hundred metres are the rule there. In areas of thick bush a pack
of dogs will use roads and elephant paths to hunt. A few dogs will trot down
the road while others go off into the bush to flush out the prey. Prey often
escape into the vegetation with chases as often ending in failure as
success.
Another difference is that in the Serengeti most of the animals that the
dogs feed on are migratory, resulting in a 鈥渇east or famine鈥� diet. In woodland
habitats, few of the game species migrate, so that there is a steady source of
food. Consistent availability as well as density of prey may be an important
influence on the number of wild dogs in any particular habitat. Food alone,
however, may not be the only factor limiting the dogs. The number of wild dogs
and their distribution within an ecosystem appears to be governed by a complex
interaction involving not only the distribution of prey and competing
carnivores, such as lions and hyenas, but also that of other packs of wild
dogs.
Dominant bitch
These packs can be as hig as 50 animals, but the dogs usually live in
groups of between six and eight adults with up to 18 puppies. The packs are
really extended families: adult males all related, adult females all related,
but the males and females are not related. In each pack, usually only the
dominant bitch and dominant dog breed. Their puppies are raised communally,
with uncles and aunts, sisters and brothers feeding and protecting the litter
until they reach maturity at one year.
With such strong bonds within a pack, how and why do new packs form? At
some point groups of up to five individuals of a single sex may break off from
the pack. These groups may travel huge distances 鈥� up to 300 kilometres has
been recorded. When two groups of the opposite sex find each other, a new pack
is formed.
Animals which leave the pack in which they were born can be as young as one
year, or as old as four. When they leave varies according to local
circumstances, but it all comes down to a simple question: will an individual
dog leave more copies of its genes in the next generation by staying at home,
or by striking out with its brothers or sisters to form a new pack?
Subdominant animals rarely breed but by staying 鈥渁t home鈥� they improve the
survival rate of the following year鈥檚 litter by protecting the pups from other
predators and by protecting kills from scavengers, which increases the amount
of food available for the puppies. Because more pups survive, and the pups are
close relatives, the subdominant dogs gain a genetic advantage by remaining at
home.
In a small pack of four or five adults, an extra 鈥渉elper鈥� may be
iinstrumental in protecting and feeding pups, but as the pack grows the value
of a 12th or 15th dog is small or non-existent. At some point, it no longer
pays to hang around because the dog has a better chance of passing on its
genes by taking the risk of leaving the pack in which it was born and
attempting to start a new one. Once again, though, other factors will
influence a dogs 鈥渄ecision鈥� to leave, including the density of prey and other
predators and the proximity of other packs of dogs.
Births, deaths, dispersal and pack formation form the core of what makes
wild dog populations go up or down. To understand the population dynamics of
wild dogs, however, a scientist must be able to identify individuals and
follow them through their lives. Fortunately, wild dogs are easily
distinguished from one another by their distinctive coat patterns. By making a
set of identification cards, each with a photo of the left and right side of a
dog, researchers can recognise individual animals in a population.
Tourist snaps
Another valuable source of supplementary information comes from photographs
taken by visitors to an area. For example, tourists鈥� photos of wild dogs have
allowed the researchers in Kruger and Hwange National Parks to collect
extensive data on many animals outside their central areas.
This photographic identification also gives a clue to the number of wild
dogs that once roamed Africa. If you count the dogs found in different types
of habitat, then calculate the relationship between habitat type and density,
you can extrapolate how many dogs lived in various habitats in Africa. While
we will never know the exact figure, this method suggests that about 200 000
individuals reflects the right order of magnitude.
An intensive photographic survey would provide an accurate census of the
present population, but such surveys are labour-intensive, time-consuming and
rely on a high density of tourists with cameras, a resource not found in many
parts of Africa. A less accurate, but more practical method is to conduct a
mail survey among people working in wildlife areas around Africa. Results of
such a survey, conducted in 1988 by John Fanshawe, a biologist from BirdLife
International in Cambridge, and Lori Frame, a freelance American ecologist,
suggest that there are now fewer than 5000 wild dogs in Africa, with fewer
than 3000 of these in protected areas.
In response to this decline, the World Conservation Union (IUCN) and
Tanzania鈥檚 Department of Wildlife sponsored a workshop to assess the viability
of the existing population of wild dogs. Held in Arusha, Tanzania, in March
1992, the workshop brought together experts and wildlife managers from the
seven eastern and southern African countries with the largest wild dog
populations.
Because wild dogs die from numerous causes, the only way to investigate how
patterns of births and deaths affect population growth and decline is through
computer simulation. This is because the dogs have high mortality and birth
rates so that numbers can move up and down relatively quickly. Computer
simulations show that even populations as large as 100 individuals are likely
to become extinct in a 100 years. But if several populations of up to 100
animals are linked, either by continuous habitat or through wildlife
corridors, the probability of extinction drops dramatically. Political
initiatives to keep corridors open between populations are critical if wild
dogs are to survive.
The best way to reduce the chance that wild dogs will become extinct is to
establish and protect as many populations as possible. One way to increase the
number of populations is to reintroduce animals into areas where they are
known to have lived in the past. The history of re-establishing wild dog
populations, however, is a chequered one. For example, a reintroduction
programme in South Africa鈥檚 Hluhluwe/Umfolozi Game Reserve, sponsored by the
Natal Parks Board, met with limited success. Twenty-three dogs were released
in the reserve in 1981, most of which had been bred in zoos. Since then the
population has fluctuated between 3 and 30 individuals. In the late 1980s, two
attempts to reintroduce wild dogs to Etosha National Park in Namibia failed,
the first because of vaccine-induced distemper, and the second because the
dogs did not understand the need to avoid other predators 鈥� every one was
killed by lions.
The misfits
A better strategy might be to reintroduce dogs into some of the large
national parks 鈥� those covering 5000 square kilometres or more. Only in such
vast areas of wilderness can a reintroduction programme have any hope of long-
term survival. Potential locations in eastern Africa include Virunga and
Garamba National Parks in Zaire, Akagera National Park in Rwanda and Murchison
Falls National Park in Uganda. But what will we use as the source population
for reintroduction attempts? Across Africa, researchers have been collecting
blood and tissue samples for the first genetic study of Lycaon pictus. Derek
Girman and Robert Wayne, geneticists at the University of California, Los
Angeles, found that the wild dogs living in southern and eastern Africa are
genetically and morphologically different from each other, and should
therefore be considered as two distinct subspecies. This suggests that
southern wild dogs should not be introduced into eastern or western African
countries. Yet recent data compiled by Bruce Brewer and Stephanie Rhodes of
Brookfield Zoo in Chicago show that all 300 of the world鈥檚 captive wild dogs
are of southern African origin. The 鈥渇it鈥� could not be worse 鈥� we most need to
reintroduce dogs into eastern or western Africa, but all the captive dogs come
from the continent鈥檚 southern region.
In the end it is likely that individual populations of wild dogs will
become extinct; relocating populations from other habitats or linking
populations through wildlife corridors will be the only things which will
ensure long-term survival of the wild dogs in these areas. Fortunately for
wild dogs, scientists have begun making plans for the conservation of the
species before the situation is critical, but it is still likely that in the
next twenty years wild dogs will exist only in extremely large areas of
relatively undisturbed habitat. As human populations grow, the areas of
suitable habitat decline, and national parks and protected areas will offer
the only refuge. Ironically, the future survival of the species may depend on
the stewardship of the wild dog鈥檚 worst enemy, humans.
Perils of handling wild dogs
STUDYING highly mobile endangered species such as the African wild dog can
present a dilemma as any intervention and handlng of wild animals carries
risks. But data critical to conservation cannot be collected without some
intervention 鈥� disease screening, genetic studies and fitting radio
transmitters to study ecology and mortality of animals all involve giving
anaesthetics and/or collection of tissue samples. While no instances have been
recorded of wild dogs dying while under anaesthesia, for many species, such as
rhinos, just putting an animal to sleep could kill it.
Anaesthesia is now less risky and new techniques such as recovery of
genetic material from faeces will minimise some sorts of intervention. Even
so, how do researchers balance these risks against the urgent need to collect
information critical to conserving the species?
Samples taken from captive and wild Lycaon have shown that disease appears
to be relatively common. Anthrax was confirmed in both the Luangwa Valley in
Zambia, and in wild dogs living in the Selous Game Reserve in Tanzania.
Diseases have been found that are common in domestic dogs such as rabies,
distemper, parvovirus, babesiasis and canine ehrlichiosis. This suggests that
increasing numbers of humans, and the accompanying increase in domestic dogs,
pose a serious threat to wild dogs.
In 1990, rabies was confirmed in the Serengeti ecosystem from two samples
taken by Peter Kat and his co-workers in the Masai Mara in Kenya, and in a
single wild dog carcass sampled by Sarah Gascoyne and her associates in
Tanzania鈥檚 Serengeti National Park. Believing that rabies was a threat to the
reemaining fragmentary population, dogs in each population were vaccinated in
an attempt to provide protection.
Despite the vaccination, by 1992 nearly every pack of wild dogs in the open
plains, had disappeared or died, in most cases from unknown causes. This
鈥渓ocal extinction鈥� marked the end of a decline in the Serengeti population
from 77 adults and yearlings in 1970, to 26 adults and yearlings in 1977, and
somewhere between 12 and 31 individuals from 1976 to 1991.
But was the death of the dogs an inevitable result of disease in the
Serengeti ecosystem, or did the intervention of scientists hasten or cause
this population crash?
Wild dogs have been studied in the Serengeti on and off since 1967, but
starting in 1985, members of each pack were fitted with radio collars. Roger
Burrows from the University of Exeter, who studied wild dog behaviour in the
Serengeti during the late 1980s and early 1990s, has suggested that handling
the wild dogs to fit radio collars and subsequent vaccination may have
stressed the dogs, activating latent infections which precipitated the early
death of dogs in these packs.
In a recent paper, Burrows and two other scientists present data consistent
with this hypothesis 鈥� the patterns of mortality in the Serengeti dogs are
correlated with changes in research practice. The data are also consistent
with the explanation that a disease, either rabies or distemper, could have
wiped out the study population even if there had been no handling or
vaccination.
To investigate the more general effects of handling, one of us (Joshua
Ginsberg) has worked with researchers involved in wild dog projects in Masai
Mara, Kruger, Hwange, Selous and northern Botswana and compared data on the
survival of 135 animals which had been handled with the survival of 305
animals which had not been handled. We found no difference in survival rates
between the two groups of dogs. Some of these animals have worn collars for up
to four years and in Kruger National Park, the study area in which handling
was most intensive, the population has doubled since handling began.
If handling is implicated in early mortality of wild dogs in the Serengeti,
it must have involved a complex interaction of factors such as nutrition,
stress and the prevalence and virulence of disease not found in other
ecosystems.
At present, there is no reason to assume a causal link between intervention
and population decline, but Burrow鈥檚 hypothesis has focused attention on the
need for further study of vaccination of endangered species and the effects of
handling-induced stress. Studies of stress hormones in wild dogs are now under
way in Tanzania.