Testis size and onset of spermatogenesis in Cape mountain
zebras (Equus zebra zebra)
B. L. Penzhorn and N. J. van der Merwe
Faculty of Veterinary Science, University of Pretoria, Private Bag X04, Onderstepoort 0110,
Republic of South Africa
Summary. Testis mass of adult Cape mountain zebra stallions (mean 70\m=.\0g) was
appreciably less than that of other zebra species and domestic horses. The histological
appearance of the testes of 11-, 24- and 29-month-old colts was typically prepubertal.
Spermatogenic activity of a 4-year-old stallion obtained at the end of summer was at a
very low level, while a 4\m=.\5-year-oldstallion obtained 6 weeks after the winter solstice
showed a marked increase in spermatogenesis compared with the 4-year-old. Stallions
6\m=.\5\p=n-\19years of age collected in different seasons all showed active spermatogenesis.
Keywords: mountain zebra; testis size; spermatogenesis
Cape mountain zebras (Equus zebra zebra) are rare, with only about 474 individuals in August 1985
(Smithers, 1986). No information is available on their testis size or the onset of spermatogenesis
and their spermatozoa have not been described.
Suitable protected areas in the historic range of the Cape mountain zebra are being restocked
from the population in the Mountain Zebra National Park (MZNP; 32°15'S, 25°41'E). More mares
than stallions have been translocated, resulting in a surplus of stallions in the MZNP population.
Some of these stallions were made available for multidisciplinary research (van Heerden & Dauth,
1985; Horak et ai, 1986; Penzhorn, 1987; Penzhorn & Grimbeek, 1987). The opportunity therefore
arose to measure testis size, describe morphology of spermatozoa, determine age at onset of
spermatogenesis and to freeze epididymal spermatozoa (Barker & Gandier, 1957) for possible
Natural populations of Cape mountain zebras consist of breeding herds and bachelor groups
(Penzhorn, 1984). The breeding herds consist of a single adult stallion, 1-5 mares, and their
offspring. Colts which leave their maternal herds remain in bachelor groups until succeeding in
becoming herd stallions at > 5 years of age. Social hierarchies exist in the bachelor groups, with
younger colts being dominated by the older ones.
Materials and Methods
Eleven Cape mountain zebra colts and stallions of known age ranging from 11 months to 19 years (Penzhorn, 1987),
were collected at random: 2 in October and 3 in each of March, May and July. At the field laboratory the testes and
epididymides were dissected out. The mass of each testis and epididymis of all but 2 animals was determined. The
length, breadth and height of the testes of 6 animals were measured. Each cauda epididymidis that was obviously
distended with spermatozoa was cut into 2-mm sections and the epididymal fluid was collected in ~
1 ml semen
extender consisting of lactose, EDTA, sodium citrate, sodium bicarbonate, glycerol, egg yolk and antibiotics
Smears of epididymal spermatozoa were made on microscope slides for subsequent staining with Spermac (Oettlé,
1986: Stain Enterprises, P.O. Box 12421, Onderstepoort 0110, RSA) and morphological evaluation. The remaining
epididymal spermatozoa were extended 1:3 (v/v) in Tischner's extender, cooled to 4°C over 2 h and frozen on solid
C02 in 005 ml volumes. Sections of each testis (~8 mm thick) and of epididymides not already cut for collection of
spermatozoa were fixed in 10% formalin, dehydrated in graded alcohols, cleared in xylol and embedded in paraffin
wax (Histosec, E. Merck, Darmstadt, West Germany). Sections (5 pm thick) were cut on a Reichert sliding microtome
and stained by the standard Mayer's haematoxylin and eosin technique. Changes in seminiferous tubule diameter
were quantified by measuring two diameters at right angles in cross-sections of 10 seminiferous tubules per testis. The
measurements were made with an ocular micrometer.
The predicted mass of both testes was calculated by means of the power function y 0-035.T0'72, where y is
mass of both testes in grams and is body mass in grams (Kenagy & Trombulak, 1986). This is the function of
the allometric relationship between mass of testes and body mass, based on data from 133 mammalian species.
The relative testis size was then calculated, which is the ratio of observed mass of testes to that predicted by the
The testes of the 29-month-old colt had passed through the inguinal canal but were not yet in the
scrotum. The other animals all had scrotal testes. The mass and dimensions of both testes of 6
stallions are shown in Table 1. The right testis was consistently heavier than the left testis. This was
also found in 2 of the 3 other sets of testes which were weighed. The mean testis mass of 5 adult
stallions was 70-0 g (range 57-5-82-0 g). The mass of paired testes and epididymides of 9 stallions is
shown in Fig. 1. The mean mass of both testes plus their epididymides for adult stallions (>6-5
years old) was 166-4 g (range 1600—180-5 g). In the 4- and 4-5-year-old stallions, the figures were
550 g and 60-5 g respectively.
The predicted mass of both testes calculated by means of the equation of Kenagy & Trombulak
(1986), assuming a body mass of 250 kg (Smithers, 1983), was 269 g. If this is compared to the
calculated mean of 70 g/testis, the relative testis size is 0-52.
Only 2 adult stallions collected in March and one in May had sufficient epididymal spermatozoa
to allow freezing. The May stallion, 6-5 years old and the only herd stallion in the sample, pro¬
duced the largest volume. The spermatozoa resembled those of domestic horses in morphology.
Morphological abnormalities recorded on epididymal sperm smears of 2 of these stallions are
listed in Table 2. If minor acrosomal abnormalities, which may be due to age of the smear or have
resulted from poor technique, are disregarded, the 6-5-year-old stallion had ~20% and the
18-year-old ~50% normal spermatozoa. Major acrosomal abnormalities included lipping,
vacuolation and absence of the acrosome.
The testes of the 11 -, 24- and 29-month-old colts were typically prepubertal in histological
appearance: the epithelium lining the seminiferous tubules consisted mainly of Sertoli cells, with
occasional spermatogonia. Occasional Leydig cells were present in the interstitium between the
Although specimens from the 46-month-old stallion (collected in October) did not fix properly
and no firm conclusions could be drawn as to the spermatogenic activity, no spermatozoa were
visible in the seminiferous tubules or epididymides.
Spermatogenic activity in the testes of a 4-year-old stallion collected in March was at a very low
level. Some seminiferous tubules showed a few pre-leptotene primary spermatocytes, but secondary
spermatocytes, spermatids and spermatozoa were seen in a few seminiferous tubules only, and then
in low numbers. The epididymides contained some fluid and a few spermatozoa. The testes of the
4-5-year-old stallion collected in July showed a marked increase in spermatogenic activity com¬
pared with the 4-year-old. Pre-leptotene spermatocytes were especially numerous, but secondary
spermatocytes, spermatids and spermatozoa were still low in number. The epididymides con¬
tained no spermatozoa. The stallions >6-5 years old all showed active spermatogenesis with all
the spermatogenic cells present in the seminiferous tubules. Leydig cells were not a prominent
The mean diameter of seminiferous tubules (Fig. 1) increased rapidly between 2 and 6 years of
age, before levelling off. The greatest mean diameter was found in the 6-5-year-old herd stallion.
Table 1. Dimensions and mass of the testes of 6 Cape mountain zebras of known age from the Mountain
Zebra National Park
Left testis Right testis
Age Month Length Height Width Mass Length Height Width Mass
(years) collected (mm) (mm) (mm) (g) (mm) (mm) (mm) (g)
2 July 30 18 15 4-5 25 18 15 5-5
4 March 45 32 30 20-5 40 30 25 21-5
4-5 July 40 25 25 19 5 45 30 30 280
14 July 65 45 40 630 65 40 40 730
16 March 70 45 42 73-5 70 45 40 82-0
19 March 60 40 45 610 65 45 30 79-5
Fig. 1. Diameter of seminiferous tubules (mean ± 1 s.d.) and mass of paired testes (hatched
bar) and epididymides (stippled bar) of Cape mountain zebras of known age.
Table 2. Morphology of epididymal spermatozoa* of two Cape
mountain zebra stallions
Age of stallion
6-5 years 18 years
Normal spermatozoa 18 9
Normal spermatozoa with
minor acrosomal abnormalities -20 -50
Acrosome-major + minor >80 >80
-major only 10 32
Head (excluding acrosomal
abnormalities) 8 40
Loose head 24 13
Midpiece 18 9
Tail 3 8
Protoplasmic droplet 0 2
* Values are %; some spermatozoa exhibited more than one abnormality.
The mass and dimensions of Cape mountain zebra testes were conspicuously less than those of
some other equids. In conspecific Hartmann's mountain zebras (E. z. hartmannae), Joubert (1974)
reported a mean adult testicular mass of 87-6 g. The mean mass of 302 g of both testes plus their
epididymides of adult plains zebra (E. burchelli) stallions from the Kruger National Park (Smuts,
1976) was nearly double that of Cape mountain zebras, although the lower limit of the range ( 175 g)
marginally overlapped with the maximum found in Cape mountain zebras. At 4—4-5 years of age
the mean mass of 229 g of both testes plus epididymides of plains zebras (Smuts, 1976) was nearly
four times greater than that of similarly-aged stallions in our sample. The testis of a young adult
Grevy's zebra (E. grevyi) stallion measured 115 85 75 mm and had a mass of ~475 g, while
that of a solitary territorial stallion and an old stallion had a mass of 600 g and 550 g respectively
(King, 1965). It is not clear whether these figures refer to single or paired testes.
A functional relationship exists in many mammals between relative size of testes and mating
system (Kenagy & Trombulak, 1986). Testes are relatively small in single-male breeding systems,
such as in mountain and plains zebras (Smuts, 1976). Testes are relatively large in multiple-male
breeding systems in which several different males may mate with each female at the same oestrus,
such as in Grevy's zebras (Klingel, 1969). The evolution of large testes can be attributed to high
copulatory frequency and sperm production, but size of testes has undoubtedly evolved in each
species in response to a variety of additional factors beyond the first-order influence of body size
(Kenagy & Trombulak, 1986).
Variations in testis size of domestic horses can be expected due to the variation of factors such
as body size of different breeds, age and season. Swierstra et al. (1974) reported a mean adult
testicular mass of 183 g while Sisson (1975) stated that single testes had a mass of 225-300 g and
measured 100-120 mm in length, 60-70 mm in height and 50 mm in width. Kenagy & Trombulak
(1986), using the data of Cox (1982), reported that a 468 kg stallion would have a combined testes
mass of 416 g, giving a relative testes size of 0-98.
The greater development of the right testis, found in 8/9 of our specimens, contrasted with the
reports for some other equids. In domestic horse and plains zebra stallions, the left testis was more
often the larger (Sisson, 1975; Smuts, 1976; Cox, 1982). According to Hafez (1980), post-natal
growth of the testes of domestic horses begins during the 11th month and the left testis develops
earlier and grows faster than the right one. Cox (1982) found that most testicular growth occurs
during the second winter of life, with further increases up to full maturity.
Virtually all the abnormal sperm heads of the 18-year-old stallion had narrow bases. This could
have been associated with a weak head-neck junction, leading to the high percentage of loose
heads. As protoplasmic droplets may be lost during the staining procedure, the low percentage
found may not reflect the true picture. This male was a displaced herd stallion in a bachelor group.
As a herd stallion he was known to have sired foals. The high percentage of loose heads found in
the semen of the 6-5-year-old stallion may be due to ageing of the spermatozoa in the epididymis,
indicating that he had probably not ejaculated for some time before collection.
In plains zebras, Smuts (1976) demonstrated that mean testicular mass and seminiferous tubule
diameter were significantly greater during summer than during winter, even though the ranges
overlapped extensively. Our sample was too small to allow for similar comparisons.
Although there was some spermatogenic activity in the younger stallions in our sample, active
spermatogenesis began in the 4-5-year-old stallion, which had been collected 6 weeks after the
winter solstice, when daylength was increasing. If daylength is decreasing at the time when a young
stallion reaches the age for the onset of spermatogenesis, the onset may be retarded until the winter
solstice has passed and daylength is increasing again. The 46-month-old stallion was collected in
October (during the southern hemisphere spring), and yet no spermatozoa were seen in any of the
sections. The 4-year-old stallion showing a very low level of spermatogenic activity was collected in
March, at the end of the austral summer.
Although our sample is very small, there are indications that active spermatogenesis in this
population begins at about 4 years of age. In plains zebras, puberty was represented by an interval
ranging from 2 to 4-5 years of age (Smuts, 1976), whereas domestic horse stallions may attain
sexual maturity during their second year (Skinner & Bowen, 1968; Hafez, 1980). The stallions <4-5
years old in our sample were all collected from bachelor groups. Bachelors do not succeed in
becoming herd stallions if < 5 years old. It can therefore be assumed that the young stallions in our
sample were in frequent contact with older, dominant bachelors (Penzhorn, 1984). The role, if any,
that psychological factors may play in retarding puberty in this species has not been elucidated. In a
semi free-ranging population of Cape mountain zebras, a young stallion running with mares in the
absence of other stallions sired a foal when 42 months old (Penzhorn & Lloyd, 1987).
The study was financed by a research grant from the University of Pretoria. We thank the
National Parks Board, especially Dr V. de Vos, for inviting us to participate in the project; Dr E. E.
Oettlé for staining the semen smears; Dr D. H. Volkmann for describing the morphological abnor¬
malities of the spermatozoa; Professor R. O. Gilbert for commenting on the manuscript; and Mrs
I. Cornelius for drawing the figure.
Barker, C.A.V. & Gandier, J.C.C. (1957) Pregnancy in a Penzhorn, B.L. & Grimbeek, R.J. (1987) Incisor wear
mare resulting from frozen epididymal spermatozoa. in free-ranging Cape mountain zebras Equus zebra
Can. J. comp. Med. 21, 47-51. zebra. S. Afr. J. Wildl. Res. 17, 99-102.
Cox, J.E. (1982) Factors affecting testis weight in normal Penzhorn, B.L. & Lloyd, P.H. (1987) Comparisons of
and cryptorchid horses. J. Reprod. Fert., Suppl, 32, reproductive parameters of two Cape mountain ze¬
129-134. bra (Equus zebra zebra) populations. J. Reprod. Fert.,
Hafez, E.S.E. (1980) Horses. In Reproduction in Farm Suppl. 35, 661-663.
Animals, pp. 387-408. Ed. E. S. E. Hafez. Lea & Sisson, S. (1975) Equine urogenital system. In Sisson and
Febiger, Philadelphia. Grossman's The Anatomy of the Domestic Animals,
Horak, I. G., Knight, M. M. & de Vos, V. (1986) Parasites vol. I, pp. 526-550. Ed. R. Getty. W. B. Saunders,
of domestic and wild animals in South Africa. XX. Philadelphia.
Arthropod parasites of the Cape mountain zebra Skinner, J. D. & Bowen, J. (1968) Puberty in the Welsh
stallion. J. Reprod. Fert. 16, 133-135.
(Equus zebra zebra). Onderstepoort J. vet. Res. 53,
127-132. Smithers, R.H.N. (1983) Mammals of the Southern
Joubert, E. ( 1974) Notes on the reproduction in Hartmann African subregion. University of Pretoria, Pretoria.
zebra Equus zebra hartmannae in South West Africa. Smithers, R.H.N. (1986) South African Red Data
Book—Terrestrial Mammals. South African National
Madoqua Ser. I. 8, 31-35.
Kenagy, G.J. & Trombulak, S.C. (1986) Size and func¬ Scientific Programmes Report No. 125, 1-126.
tion of mammalian testes in relation to body size. J. Council for Scientific and Industrial Research,
Mammal. 67, 1-22. Pretoria.
King, J.M. (1965) A field guide to the reproduction of the Smuts, G.L. (1976) Reproduction in the zebra stallion
Grant's zebra and Grevy's zebra. E. Afr. Wildl. J. 3, (Equus burchelli antiquorum) from the Kruger
99-117. National Park. Zoologica africana 11, 207-220.
Swierstra, E.E., Gebauer, M.R. & Pickett, B.W. (1974)
Klingel, H. (1969) Zur Soziologie des Grevy-Zebras. Reproductive physiology of the stallion. I. Spermato¬
Zool. Anz., Suppl. 33, 311-316.
Oettlé, E.E. (1986) Using a new acrosome stain to evalu¬ genesis and testis composition. /. Reprod. Feri. 40,
ate sperm morphology. Vet. Med. 81, 263-266.
Tischner, M. (1979) Evaluation of deep-frozen semen in
Penzhorn, B.L. (1984) A long-term study of social organ¬ stallions. J. Reprod. Fert., Suppl. 27, 53-59.
isation and behaviour of Cape mountain zebras van Heerden, J. & Dauth, J. (1985) Serum potassium
Equus zebra zebra. Z. Tierpsychol. 64, 97-146. and sodium concentrations in dead and hypoxic
Penzhorn, B.L. (1987) Descriptions of incisors of known- dogs, shot mountain zebra Equus zebra zebra and
age Cape mountain zebras, Equus zebra zebra, from chemically immobilized laboratory rats and ground
the Mountain Zebra National Park. Onderstepoort J.
vet. Res. 54, 135-141.
squirrels. S. Afr. J. Wildl. Res. 15, 32-36.
Received 21 September 1987