|
Serval kitten courtesy of Jennifer Quincy
Description and Behavior
The serval is well adapted to hunting small prey in long grass: its
legs are slim and relatively long, and shoulder height is about 0.6 m.
Its neck is also elongated, its head is small and delicate, and its
ears are tall. The auditory bullae are correspondingly well-developed,
making up about 22% of skull length (Skinner and Smithers 1990).
Males weigh 9-18 kg (averaging 11-13 kg), and females 9-13 kg,
(averaging 9.7-11 kg: Smithers 1971, Kingdon 1977, Smithers 1978).
Coat color is pale yellow, and is marked with solid black spots along
the sides and bars on the neck and shoulders.
Although 17 subspecies are listed by Allen (1939), their
validity is doubtful (see Appendix I). Smithers (1978) examined
specimens from one locality in southern Africa and found external
characters among them, which had been used to designate six different
subspecies within the sub region. Servals from West Africa most
frequently show a pattern mutation of small speckled spots -- these
so-called servalines were considered a separate species (Felis
brachyura Wagner, 1841) until Pocock (1917a) demonstrated
that the speckled form was a serval morph. Black servals have been
widely recorded (Shortridge 1934, York 1973, Guggisberg 1975).
The heliotype of L. serval was taken near the Cape of Good Hope, but
the serval now appears to have been extirpated from the entire
southern coastal belt of South Africa and most of Cape province (Skead
1980, Stuart 1985) -- although M. Bowland (in litt. 1993)
notes an unconfirmed report from a farmer at George, midway between
Cape Town and Port Elizabeth.
Small mammals, especially rodents, are the serval’s main prey.
Larger rodents are preferred, particularly vlei (swamp) rats (Smithers
and Wilson 1979, Geertsema 1985, Bowland 1990), and also Nile rats
(Geertsema 1976, 1985). Smaller mice are of secondary
importance (Smithers and Wilson 1979, Geertsema 1985, Bowland 1990).
Up to 12 mice were found in one serval stomach from Zimbabwe (Smithers
1978). Birds, reptiles, fish and insects are also taken, although
infrequently when rodents are abundant (Geertsema 1985, Bowland
1990). Geertsema (1985) observed one young male serval, on
a moonlit night, rush into open water to seize one of a group of
feeding flamingos. Geertsema (1985) also found frogs to be a
particularly favorite prey item, with remains occurring in 77% of 56
scats. She saw another young male eat at least 28 frogs in one
three-hour period. Servals do not generally take larger prey as does
the caracal. Single animals have only rarely been observed to kill
duikers and fawns of the smaller antelope species (Rahm 1966, de
Pienaar 1969, York 1973). The detailed studies by Geertsema (1985:
Ngorongoro Conservation Area, Tanzania) and Bowland (1990:
Natal province farmland, South Africa) did not record any instances of
servals taking mammalian prey larger than rodents.
Ecology
The serval locates prey in tall grass or reeds primarily by hearing.
It makes a characteristic high leap as it pounces on a prey animal,
striking it on impact to prevent escape in thick vegetation. A single
pounce may span 1-4 meters and may be over a meter high (Geertsema
1985). Another type of leap is vertical: birds and insects are
seized from the air by "clapping" the front paws together (Smithers
1978) or striking with a downward blow (Leyhausen 1979).
Geertsema’s (1985) four-year study in the Ngorongoro Crater
is the most detailed investigation to date of serval ecology. She
found them to be largely crepuscular, resting in mid-day and
occasionally at night. Females with kittens increase diurnal hunting
activity. Servals on farmland in South Africa’s Natal province were
predominantly nocturnal, possibly a response to human disturbance (Bowland
1990). Through continuous observations (when possible - although
the study animals were habituated, they were not radio-collared),
Geertsema (1985) found that adult males, adult females and
sub-adults spend about 25% of each 24-hour period traveling and
hunting. On average, Ngorongoro servals killed about 16 times within
this period. Independent sub-adults killed more frequently than
adults, but took smaller prey with a lower energetic return. From
nearly 2,000 observations of pounces, Geertsema (1985) found
serval-hunting success to average 49%, with no significant difference
between day and moonlit night. After giving birth to kittens, one
female increased her success to 62% from 48%.
Titan
Biology
Reproductive Season:
(W) Aseasonal, but birth peaks appear to be correlated with wet
seasons, when prey densities are at their highest due to new
vegetative growth (Kingdon 1977, Smithers 1978). Geertsema (1985)
suggests that a peak occurs in the mid- to late dry season in the
Ngorongoro Crater, so that post-rains high prey density coincides with
the raising of older but still dependent kittens.
Estrus (C): 4 days (n=1: Mellen 1989)
Gestation (C): 73 days (n=15; range 70-79) (Stuart and
Wilson 1988)
Litter Size:
(W): 2.5 (n=7; range 1-3) (Smithers 1978);
(C): 1.96 (n=20: Skinner and Smithers 1990); 2.45+0.21
(n=14: Mellen 1989); range 1-5 (Stuart and Wilson 1988)
Age at Independence:
(W) 6-8 months. Newly independent juveniles, tolerated by their
mothers, may circulate within their natal range for periods up to and
over a year (Geertsema 1985)
Age at Sexual Maturity:
(C) 18-24 months (P. Andrews in litt. 1993)
Longevity:
(C) up to 19 years (Green 1991)
Habitat and Distribution
In Sub-Saharan Africa, servals are found in well-watered savannah
long-grass environments (Shortridge 1934, Rosevear 1974, Smithers
1978), and are particularly associated with reed beds and other
riparian vegetation types (Geertsema 1985, Bowland 1990). This
association with water sources means that their distribution is
strongly localized over a wide area and within a variety of habitat
types . They range up into alpine grasslands (Ansell and Dowsett
1988), up to 3,200 m in Ethiopia (Yalden et al. 1980) and
3,800 m in Kenya (York 1973). Servals can penetrate dense
forest along waterways and through grassy patches, but are absent from
the rainforests of Central Africa. A few records from arid parts of
south-western Africa, Ethiopia and Somalia indicate that servals will
occasionally make use of sub-optimal habitats (Shortridge 1934,
Yalden et al. 1980, Stuart and Wilson 1988, A. Simonetta in litt. 1992).
In North Africa, relict populations may still be found in humid
scrub and mixed woodlands of Morocco’s Atlas Mountains (Lambert
1966) and northern Tunisia and Algeria (Gouttenoire 1954, De
Smet 1989). The last confirmed record from Algeria is of an animal
killed by a French hunter in 1936 in Arzew (north-west coast), said to
be the last in the area. There have been scattered reports of serval
occurrence throughout northern Algeria during the 1980s, but
zoologists have not been able to confirm them (De Smet 1989, K. de
Smet in litt. 1993). Surviving animals are likely to have been
isolated from sub-Saharan populations for at least 6-7,000 years (Swift
1975).
Population Status
Global: Category 4
Regional (sub-Saharan Africa): Category 3
Regional (North Africa): Category 2(A)
IUCN: Not Listed
Smithers (1978) reviewed the serval’s distribution and
concluded that its range has remained largely intact, shrinking only
in the extreme north and south due to habitat loss in the wake of
increasing urbanization and changes in land use (C. Stuart in litt.
1993). Possibly servals were never very numerous in North Africa,
and water sources in the region are likely to be focal points of human
use and settlement. However, servals are highly tolerant of
agricultural development, which fosters increased rodent densities, as
long as there is sufficient water and shelter available (Bowland
1990). Kingdon (1977) notes that the serval has adapted
well to the cultivation-fallow mosaic that is widespread over the
moister regions of Africa. Degradation of forests to savannah in West
Africa probably favors the species.
Geertsema (1985) found minimum home ranges in Ngorongoro to
be 11.6 km2 for one adult male and 9.5 km2 for
one adult female over four years. The male’s home range overlapped
those of at least two adult females, while the ranges of three adult
females showed minimal overlap. Bowland (1990) found larger
home ranges for servals on South African ranchland: 16-20 km2
for two adult females and 31.5 km2 for one male, monitored
for 4-5 months during the spring and summer.
Protection Status
Protection Status: CITES Appendix II
National Legislation:
Not protected over most of its range
Hunting prohibited:
Algeria, Botswana, Congo, Kenya, Liberia, Mozambique, Nigeria, Rwanda,
South Africa (Cape province only)
Hunting regulated:
Angola, Burkina Faso, Central African Republic, Ghana, Malawi,
Senegal, Sierra Leone, Somalia, Tanzania, Togo, Zaire, Zambia
No legal protection:
Benin, Cameroon, Ethiopia, Gabon, Gambia, Guinea Bissau, Ivory Coast,
Lesotho, Malawi, Mauritania, Morocco, Namibia, Niger, South Africa,
Sudan, Swaziland, Tunisia, Uganda, Zimbabwe
No information:
Burundi, Chad, Djibouti, Guinea
(IUCN Environmental Law Centre 1986, Smithers 1986, Hecketsweiler
1988)
Principal Threats
Wetland conservation is the key to serval conservation. Wetlands
harbor comparatively high rodent densities compared to other habitat
types, and form the core areas of serval home ranges (Geertsema
1985, Bowland 1990). Of secondary importance is degradation of
grasslands through annual burning followed by over-grazing by domestic
hoofstock, leading to reduced abundance of small mammals (F. Hurst
in litt. 1991, Rowe-Rowe 1992).
Trade in serval pelts has been reported from many countries (Yalden
et al. 1980, Sayer and Green 1984, Myers 1986, Cunningham and Zondi
1991; L. Gadsby, F. Hurst in litt. 1991, E. Abe in litt. 1993);
they are frequently marketed as "cheetah" or "leopard". While the
scale of the harvest and its effect upon populations is difficult to
judge, the pelt trade appears to be primarily domestic (especially for
ceremonial or medicinal purposes) or tourist-oriented, rather than
international commercial exports (WCMC unpubl. data; see Table
1 in Part II Chapter 4). The serval’s localized distribution around
water sources may increase its vulnerability to hunting; it will also
climb a tree when chased by hounds (Stuart 1985).
Servals occasionally kill domestic poultry and only rarely young
livestock (sheep and goats): studies of their diet in farming areas in
Zimbabwe (Smithers 1978) and South Africa (Lawson 1987)
found no evidence that predation was a problem. Bowland (1990)
pointed out that problem animals, which raid chicken coops, could be
easily live-trapped for translocation. Although 17% of Namibian
farmers who indicated that servals were present on their land reported
livestock predation, none took any control measures (legally
permissible), indicating that the problem is not serious. For
comparison, 36% of the farmers reporting stock predation by African
wildcats took control measures (Joubert et al. 1982). The
serval’s preference for rodent prey should actually benefit farmers:
Geertsema (1985) calculated that an adult serval would eat some
4,000 rodents a year.
Serval kitten
|