Table Of ContentJOURNAL
OF THE
BOMBAY NATURAL HISTORY SOCIETY
December 1998 Vol. 95 No. 3
ANTLER CYCLES AND BREEDING SEASONALITY OF THE CHITAL
{AXISAXIS ERXLEBEN) IN SOUTHERN INDIA1
T. R. Shankar Raman2
{With six text-figures)
Key words: Antler development, fawning, circannual rhythms, age-specific
reproduction, tropical dry evergreen forest, axis deer, Axis axis.
Annual cycles ofantler renewal and casting, and births in a free-ranging, high-density population
ofchital or axis deer were studied for 2.5 years in Guindy National Park in southern India. In all
the three age-classes (yearling, juvenile, and adult males), seasonality was pronounced and birth
seasonality was evident. Occurrence of most births (49%) between December and March, at the
onset ofthe dry season, probablyenabled theenergetically expensive late-lactation period offemales
to coincide with the first flush in food availability after the rains. Only adult males attained peak
hard antler during the months when most conceptions occurred; the monthly percentage in hard
antler, was significantly correlated with fawning 8-9months later. Time-lag correlations and patterns
ofantler development showed that the peak in hard antler ofjuvenile and yearling males occurred
2.5 and 5 months later than in the adult males. Such staggered rutting cycles may reduce inter-
male conflict and increase the chances ofsubordinate age-classes achieving copulation.
Introduction antlers. These bony outgrowths of the cranial
frontal bones occur in 36 of the 40 extant deer
Breeding cycles of deer are known to be species, andare grown and cast at roughly annual
closely linked to annual environmental rhythms intervals in consonance with seasonal sexual
in the temperate regions. In the tropics, where cycles (Goss 1983, Lincoln 1985, 1992b). In all
environmental rhythms are relatively diffuse and species that possess antlers, excluding the
unpredictable, breeding is often considered to be reindeer {Rangifer tarandus), only males carry
markedly less seasonal or aseasonal (Lincoln antlers (Lincoln 1992b). Yearling males in most
1992 a,b). There is evidence, however, thatmany species produce their first set of antlers on
mammal species in the tropics and equatorial reaching puberty, and successive sets of antlers
regions also show distinct seasonality which is in ensuing years increase in size and complexity
influenced by factors such as photoperiod, parallel to the increase in body size. Rutting
rainfall, food availability, and genetics (Bronson males canbe distinguished easilybythe exposed,
1989). bony, hard antlers they carry, in addition to
Deer offer a unique opportunity to study characteristic sexual behaviour (deVos et al.
breeding seasonality due to theirhabitofsporting 1967, Goss 1983, Lincoln 1992b, Mishra and
Wemmer
1987, Schaller 1967). In addition to
'AcceptedMarch, 998
1 antler size and body condition, the reproductive
3CentreforEcological ResearchandConservation,
3076/5,4thCross,Gokulam Park, success of males in polygynous species also
Mysore-570002, India. changes according to the age of the individual
JOURNAL. BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998 377
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
(Clutton-Brock 1982, 1987). Prime-aged adult however, studies have generally documented
males can be expected to rut during the annual some seasonality (Johnsingh 1983, Mishra and
Wemmer
peak in female oestrus. If male-male competi- 1987, Schaller 1967, Sharatchandra
tion for females is high, younger or subordinate and Gadgil 1975). The pulse of seasonality is
males may then be forced to rut at different seen to be weak, at least some males being in
times of the year when adults are in velvet or hard antler throughout the year (Fuchs 1977,
exhausted after their rutting activities (Dunbar Mishra and Wemmer 1987, Schaller 1967). It is
etal. 1990, Hirotani 1994). Monitoring monthly not clear, however, whether the males rutting
changes in the proportion of males of different outside the peak rut are adults or younger males
—
age classes in hard antler, along with fawning whose rut is staggered relative to adults an
peaks, can be used to examine breeding important aspect of age-related reproductive
seasonality and age-related reproductive strate- strategies. This paper presents a comparative
gies ofdeer. account of antler development in different age
The chital or axis deer, Axis axis Erxleben, classes of males, and birth seasonality over 2.5
is an endemic cervid of south Asia occurring years in a free-ranging, high density popula-
between c. 8°N and 30°N in India, Sri Lanka, tion ofchital in southern India. The existence of
Nepal, and Bangladesh (Schaller 1967). This, age-related differences in antler cycles ofmales
coupled with the fact that it is common and is examined in relation to female breeding
widespread within its range, makes it a suitable seasonality and births and compared with those
representative for research on the ecology of from previous studies on chital and other
tropical cervids, which have been poorly studied cervids.
as compared to temperate zone cervids (Lincoln
1985, 1992a,b, Loudon and Brinklow 1992). Study Area
There have been several studies on the natural
history and ecology of chital in India and Guindy National Park (GNP) is a 2.7 km2
Sri Lanka where it is native (Barrette 1985, 1987, park located in the southwest corner of
Berwick 1974, Eisenberg and Lockhart 1972, Chennai (Madras) city (13°N, 80°E) in south
Johnsingh 1983, Krishnan 1972, Mishra 1982, India. The vegetation, appearing as patches of
Mishra and Wemmer 1987, Miura 1981, Schaller scrub jungle, thickets, and wooded areas,
1967, Sharatchandra and Gadgil 1975, Tak and corresponds to theTropical Dry EvergreenForest
Lamba 1984), and in other countries where it ofChampion and Seth (1968), reclassified as the
was introduced (Abies 1977, Graf and Nichols Albizia amara Boiv. community (Puri et al.
A
1966). Quantitative information on antler cycles 1989). detailed description of the park is
and breeding seasonality is, however, available elsewhere (Raman et al. 1996). The
available in only a few studies (Fuchs 1977, mean annual rainfall is 1215 mm, most of
Grafand Nichols 1966, Johnsingh 1983, Mishra which falls during the northeast monsoon in
Wemmer
and 1987, Schaller 1967, Sharat- October-November, thoughtheparkreceives rain
chandra and Gadgil 1975). Among these, Fuchs during the pre-monsoon (April-May) and the
(1977) also examined antler cycles within southwest monsoon (June-September) as well
different age classes of males for a single (Fig. 1). The park is home to a small native
year. population of about 50 blackbuck (.Antilope
Reports from studies of captive herds of cervicapra L.) and a high-density population
seasonal synchrony in annual antler and birth of about 550 chital (212.3 chital/km2 during
cycles have been conflicting (Bubeniketal 1992, 1991-92, line transect estimate in Raman et al
Loudon and Curlewis 1988). In the wild, 1996).
378 JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998
BREEDING SEASONALITY CF THE CHITAL IN SOUTHERN INDIA
mm Temperature (DegreesC)
Rainfall in
350 40
300
30
250
200
20
150
100
10
50
FMAMJ JASOND
0
0
J
Months
—e— Mean Daily Maximum —e— Mean Daily Minimum
Fig. 1: Rainfall (vertical bars) mean monthly maximum and minimum temperatures for GNP. Data averages
for 1931-1960 from the Climatological Table for Chennai (Minambakkam). Rainfall and temperature
regimes were similar during the study period 1991-93.
(see Raman et al. 1996).
Methods (0600 to 1000 h) and evening (1600 to 1900 h),
the times ofday when animals were most active.
Six months (July-December 1990) of Observations were made along established
preliminary observations were made on over 500 transects, forest trails, and roads using a pair of
free-ranging individuals in GNP, concurrent to 7 x 50 binoculars. Accurate classification of
observations of captive chita! in the adjacent individuals was possible because the chital in
Children’s Park zoo. Basedonthese observations, GNP are generally not very shy of humans and
the age-sex and antler-stage classification of canbe approached easily to within 30-50 m, often
m
chital by Schaller (1967) was substantially to less than 20 m. Animals more than 100
A
refined (see below). total of 12,866 individual away, and those not seen clearly, were not
chital were classified into different age-sex and classified. Bellows (rutting call given by adult
antler categories between January 1991 and May males usuallyconsisting of3-5 loudnotes) heard
1993 (sampling with replacement). An average in the field were also recorded.
of444 (range 121-920) chitalwere classifiedeach
month, which comprised, on an average, of232 Age-sex and antler classification
adult females (76-488), 35 yearling females (2-
67), 35 fawns (2-80), 82 adult males (29-170), Females could be classified only as adult
31 juvenile males (6-73), and 32 yearling males or yearling females in the field (Schaller 1967).
(2-74). Around eight field visits were made each Fawns were classified as small fawns
month, with 3-5 trips each in the morning (<2 months) and big fawns (>2 months), based
JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998 379
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
on their height in relation to that of adult antler usually with a small brow
whom
females they almost invariably accom-
tine.
panied. Fawns could not be sexed in the field. (iii) Mature Velvet: Antler reaches nearly
Males were classified into 16 categories (see full length of 20-30 cm and has a
below) representing different stages ofmaturi—ty rounded (not bulbous) tip and a fawn
and antler condition within three age groups colour.
yearlings, juveniles and adults. Having these (iv) Dry Velvet: Antler is mature, pointed,
broad groups facilitated distinguishing the age and covered by a thin but entire
class of individual males irrespective of antler covering of dark, greyish velvet.
condition, by looking at antler pedicel height (v) Peeling: Strips ofvelvet usually hang
(which declines with age), pedicel width, and loose from the antler and portions of
head and body sizes (which increase with age). the bone are visible underneath.
The detailed classification of antler develop- (vi) Hard: Hard bony antler with no trace
ment stages within each age class allows ofvelvet left.
the monitoring of seasonal changes at a
—
finer level and identifying the rutting period (C) Adult Males (age > 3 yr) Six adult male
clearly. classes were considered:
(A) Yearling Males or Spikers, age 10 months
- 2 yr) were classified as follows: Cast Adult, Growing Velvet, Mature
(i) Button Male: Only a stub-like Velvet, Dry Velvet, Peeling, and Hard antler
developing pedicel (c. 2 cm), com- adults were distinguished as in the case of
pletely covered by hair, is visible on Juvenile Males. Antler beam, pedicel, and body
the head. size were larger than for Juvenile Males and
A
(ii) Velvet: blunt, spike-like growing enabled unambiguous classification in the vast
antler covered in fawn-coloured majority ofcases.
‘velvet’ and mounted on a tall c 4-5
( .
cm) pedicel. Analyses
Dry Velvet: A pointed spike-like
(lii)
antler, 10-12 cm long, with a thin grey The percentage of males within each age
covering ofvelvet. class in the different stages of antler condition
(iv) Hard Spiker: Hard bony antler with was computed for each month between March
no trace of velvet. This includes an 1991 and May 1993. Initially (January-March
intermediate stage, namely, yearlings 1991), adult and juvenile males were not
with the velvet peeling offbefore the distinguished, hence the data were not included
hard antler stage. Such individuals in the age-specific analyses.
are rarely seen, presumably because For eachmonth, the number ofsmall fawns
this process occurs injust a few days. (age < 2 months) was expressed per 100 adult
females to examine seasonality in births. The
(B) Juvenile Males (age 2-3 yr) were classified number of bellows heard during 0600-0900 h
as follows: and 1600-1900 h was divided by the number of
(i) Cast Yearling: Only the tall pedicel hours spent in the field in those time blocks to
of the yearling is present, with signs get an index ofbellowing rate (bellows/hour).
of the wound at the burr where the Sinusoidal curves were chosen to fit the
antler was cast. data on the annual antler cycles (Batschelet
(ii) Growing Velvet: Bulbous-tipped 1981).
380 JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998
1 ,
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
The curves were ofthe form: To examine temporal differences between
antler cycles of different age classes of males,
yy — A cos -5O- (t - max7) + H time-lag correlations were used as explained
below. This was also used to correlate male peaks
—
where: y percentage ofmales (ofa given age in hard antler with births 8-9 months later (this
— class) in hard antler. corresponds to the gestation period in chital:
t month number (ranging from English 1992, Rao 1984; see Discussion). In
— January = 1 to December = 12). these analyses, a lag ofone unit corresponded to
A amplitude (a measure ofthe extent one month, and a range of 0-12 lag units was
or magnitude of the seasonality employed. Pearson’s product-moment correlation
pulse). coefficients were computed in each case. The lag
—
t the month which is the peak in at which the maximum significant positive
the annual cycle, correlation existed between the antler cycles of
—
H
annual mean of the monthly two age classes ofmales represented the number
percentage of males (of that age of months separating their peaks in hard antler.
class) in hard antler (a constant Similarly, a significant positive correlation
estimated from the data). between the percentage of males of a given
The parameters ofthe curve have biological age class in hard antler in a given month and
relevance (Batschelet 1981: 159). The parameter the number of small fawns observed 8-9
t estimates the month (Jan. = 1 to Dec. = 12) months later indicated whether the peaks in
max x 7
when the peak in antler cycles is reached. hard antler coincided with incidents of mating
Similarly, the estimated value of the peak and conception.
percentage of males in hard antler during the
H
annual cycle is given by A + (when t = t Results
y = A cos (0) + H = A + H). Initial estimates of
A and t were derived iteratively and these were Age-specific antler cycles
used to perform non-linear regression using
SPSS/PC + software (Norusis 1990). Males in hard antlers were seen throughout
the year, with 50% (40%) or more males being
Table
1 in hard antler during 8(10) out ofthe 12 months
BREEDING SEASONALITY OF ALL CHITAL
of the year (Fig. 2a). When considered by age-
MALES CONSIDERED TOGETHER
class, however, more distinctly seasonal patterns
ANDAS SEPARATE AGE CLASSES*
are evident (Fig. 2b-d). The amplitude or extent
MaleAge Amplitude Peak Mean R2 N ofthe seasonality pulse ofall males combined in
Class month monthly % the analysis was lower (A = 18.56), compared to
%
in the seasonality ofadult orjuvenile males coming
hard =
into hard antler (A 28.31 and 34.24,
antler
respectively, Table 1). The extent ofseasonality
(A) (vtmax') (v77)' appeared to be low foryearling males (A = 18.35),
All Ages 13.56 7.10 55.4 84.9 26 but the curve fit was poor in this case R2 = 0.40)
Yearling 18.35 11.35 45.3 39.7 26 (
and the data also suggests a more distinct
Juvenile 34.24 8.67 60.4 79.6 26
Adult 28.31 6.25 56.9 89.0 26 seasonality than forall males consideredtogether
(Fig. 2b vs 2a).
*Parameterestimatesfromnon-linearsinusoidal curve fitting
A
data in Figure 2. For explanation of parameters see text majority (73-91%) of the adult males
(Analyses). came into rut and hard antler between March
JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 95(3), DEC. 1998 381
r
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
Males,
Yearling
(b)
Males,
All
—
(a)
Males
of
antler.
Yearling percentage
hard
in
(b)
100,
monthly Males,
J0|;U0 pjBLj Uj 06^U0OJ0d j0|;ue pjeij uj 060;u0oj0
c|
as
Adult
1
curves
(d)
and
sinusoidal
7
Males,
o as
-
- CO Juvenile
- <
— modelled
H JZ (c)
cycles
antler
of
;
Ma!es
Seasonality
All
oooooooooo
(a) 2:
Qr—
10 J0|}us pjeij uj 06e^U0OJ©d Fig.
J0|;u9 pj0q uj 06e;u0OJ0d
382 JOURNAL, BOMBAYNATURAL HISTORYSOCIETY, 95(3), DEC. 1998
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
and July, with the peak in hard antler being in and juvenile males. Thus, the percentage of
July (tmax = 6.3, Fig. 2d). Most of their rutting yearling males in hard antler peaked around
activity appearedconcentratedbetweenApril and December (January in 1991 and December in
July, the period when the number ofbellows per 1992, tmm= 1 1.3, Fig. 2b). Thus adult andjuvenile
hour was highest. Thus, available data for April- male peaks were separated by about 2.4 months,
December 1991 showed that the percentage of while juvenile and yearling male peaks were
adult males in hard antler during a given month separated by 2.7 months. This was confirmed by
was positively correlated with the number of time-lag correlationanalyses. Comparisons were
bellows per hour that month (r = 0.85, df= 7, made between the antler cycles (monthly
2-tailed P = 0.004). percentage in hard antler) ofjuveniles and adults
Juvenile males, in contrast to adult males, and between juveniles and yearlings. In each
came into hard antlermuch later, reaching apeak case, they were each out of phase by about 2-3
around late September (t = 8.7, Fig. 2c) when months as shown by the higher and significant
max
most adult males began to cast antlers (August- positive correlations at the corresponding lag
October). Juveniles cast antlers mostly during (Fig. 3). Thus, yearling and adult males were
November-January as was evident from the fall out ofphase by about 5.1 months.
in the hard antler curves for the two age classes These distinctly age-specific seasonal
(see Fig. 2, and Antler development). Yearling patterns were evident in 1991 and 1992, and even
males (spikers) were out ofphase withboth adult in the data for the first five months of 1993. Only
Lag (Months)
Fig. 3: Time-lag correlations between different age classes ofmale chital in the monthly percentage in hard
antler. Pearson product-moment correlation values falling outside the range encompassed by the horizontal
P <
dotted lines are significant (2-tailed 0.05).
JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998 383
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
(a) Button Ve vet K\1 Pry Velvet [ZD Hard
I I I
0D) Hi Oast I 1 Growing Velvet 1\D Velvet I I Velvet dry Peeling I I Hard
OO
1
80 \
\
\
60
\
Percentage
40
J
Ml
20
II..
wlFMAMJ JASONDJ FM1 AMJ JASONDJ FMAM
(c) Hi Cast I I Growing Velvet Z3 Velvet I Velvet drryy LID Peelin'g9 LID Hard
OO
1
80
60
I
\
40 -
v
-
20
-J
O FJ ML AMJ JASONDJ FMAMJ JASONDJ FMAMZL
J
Month
Fig. 4: Antler development in (a) Yearling Males, (b) Juvenile Males, and (c) Adult Males during 1991-93
as monthly percentage of individuals in different stages of antler development.
384 JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC 1998
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
ifmales ofall age classes are pooled, is an overall Fawning seasonality
flatter curve obtained, indicating milder
seasonality. It is clear, however, that even the Births of fawns also clearly indicated
pulse ofage-specific seasonality is not absolute, seasonality (Fig. 5). The proportion of small
as about 20% of the adult males were in hard fawns per 100 females peaked in February in
antler even during the lowest month (December, 1992 and 1993. In 1991, a peak in births is also
but see Discussion). apparent in April, the reason for which is
unknown. During 1991-92, most small fawns
Antler development (49.2%) were noticed between December and
according to age class March. Fawns two weeks to one month old could
be seen moving with females. Based on the sizes
Consideration ofdifferent stages ofantler of fawns in the field, my observations indicated
development reinforces the above patterns. that most births in GNP occurred between mid-
Yearling males in the ‘Button Male’ stage were December and mid-February. Very few births
observed as early as November-December, when occurredbetweenJuly andNovember. Only adult
they were 10-12 months old (assuming that they males showed a significant correlation between
were born during the birth peak in January- rutting (monthly percentage in hard antler) and
A
February). few button males were seen till fawning after a lag of8-9 months, corresponding
about May-June the following year (Fig. 4a). to the gestation period (Fig. 6).
During the following months, yearling males
were mostly in velvet and dry velvet till about Discussion
August-October, when many of them started
turning from dry velvet into hard spikers over The antler cycle of chital in Guindy
the winter. Most yearlings cast their antlers the National Park appears, at first glance, to indicate
following January-February, and the approxi- a very diffuse seasonality as at least 50% of the
mately 2-yearoldmales now entered thejuvenile males are in hard antler for 8 out of 12 months
age class. of the year. This has prompted observers in the
The juvenile males were in growing and past to conclude that chital breed aseasonally,
mature velvet till April-May; they cleaned their even though each male may have its own yearly
antlers of velvet, achieving their peak in hard seasonal cycle (Inverarity 1895, Krishnan 1972,
antler between June and September (Fig. 4b). Lincoln 1992a,b, Loudon and Curlewis 1988).
Interestingly, these were the months when The present study shows, however, thatthe antler
most adult males had just cast their antlers cycle of individual age classes of males reveals
(Fig. 4c). A majority of juveniles cast their distinct seasonal patterns. Mostofthe adultmales
hard antlers after December and entered the (73-91%) are in hard antler between March and
adult male age category as 3-year olds. These July, with a peak in May-June. In contrast,
individuals then came into hard antler earlier juvenile and yearling males peak about 2.5 and
than in the previous year, more or less in 5 months later.
phase with the other adult males in the Such staggered antler cycles of different
population. Between September andDecember, age classes of deer have been reported even in
most adult males had growing antlers. By temperate cervids exhibiting greater seasonality
the following April, 60-75% of the adults ofbreeding such as the Sika deer Cervus nippon
had passed through the dry velvet and (Miura 1984a) and reindeer (Leader-Williams
peeling phases to attain hard antlers and begin 1988). In red deer Cervus elaphus, dominant
to rut. males tend to cast antlers (and come into hard
JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC. 1998 385
BREEDING SEASONALITY OF THE CHITAL IN SOUTHERN INDIA
Month
Fig. 5: Seasonality offawning in chital in Guindy National Park from 1991-93.
antler) earlier and produce larger antlers than I discuss the factors influencing the breeding
subordinate ones (Bartos 1990). Yearling male seasonality patterns in male and female chital,
chital have also been reported to come into hard and compare these with other cervids.
antler several months after the peak in the adult
rut (Fuchs 1977, Schaller 1967, Schaller and De Age-related antler cycles
1971). Age-related differences in antler cycles and reproduction in males
and rutting have been noted in other Indian deer
species: the swamp d—eer or barasingha (Cervus It is well established that the hard antler
duvauceli duvauceli young stags cast antlers stage in tropical and temperate cervids is
later in the year, Singh 1984, and my personal associatedwithhightestosterone levels, enlarged
observation), the hardground barasingha (C. d. testes, and heightened sexual activity in males
—
branderi young stags rut later than adults, (Goss 1983, Lincoln 1985, 1992b, Axis axis
— :
Martin 1975), and the sambar (C. unicolor Loudon and Curlewis 1988). The rut is an
young males cast antlers later than adults, energetically expensive period for males of
Richardson 1972). In spite of the occasional differentspecies due to the demands ofterritorial
description of these patterns, the processes defence, mate searching, bellowing, sparring and
underlying themremainpoorly understood. Here fighting activities (Clutton-Brock et al. 1979,
386 JOURNAL, BOMBAYNATURALHISTORYSOCIETY, 95(3), DEC 1998
