Zoo Biology 16:435–443 (1997)
© 1997 Wiley-Liss, Inc.
Differences in Winter Activity, Courtship,
and Social Behavior of Two Captive
Family Groups of Mexican Wolves
(Canis lupus baileyi)
José F. Bernal* and Jane M. Packard
Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas
The purpose of this study was to determine differences in activity patterns and
social behavior of two groups of endangered Mexican wolves maintained at two
quite different facilities and to determine some of the variables that should be
considered when making specific behavioral comparisons among wolves in this
binational captive breeding program. Quantitative measurements of an Activity
Index and social behaviors were obtained for three individuals in each pack.
Within each age/sex category, activity, aggression, and play were more frequent
in the pack at a zoo facility, compared to the pack at a field station facility.
Frequency of courtship interactions and scent marking were significantly higher
in the field station pack. The packs were similar in the frequency of active sub-
mission, but differed significantly in the pattern of this behavior. Given the large
number of interacting variables and small number of individuals in this study,
we recommend caution in generalizing results to other packs or facilities. Zoo
Biol 16:435–443, 1997. © 1997 Wiley-Liss, Inc.
Key words: activity; social behavior; Species Survival Plan; canid, wolf
Behavior is an important consideration in making management decisions to
implement the goals of a Species Survival Plan [Kleiman, 1994]. For Mexican wolves,
behavioral interactions in a naturalistic enclosure at one field site [Servin, 1991],
general behavior at one zoo, and physiological norms have been reported for Mexi-
can wolves [Reyna Medrano et al., 1993]. However, norms of behavior have not yet
been established for Mexican wolves as has been done for northern subspecies of
wolves [MacDonald, 1980; Packard, 1980; Kreeger et al., 1996].
Received for publication 6 February 1995; revision accepted 2 June 1997.
*Correspondence to José F. Bernal-Stoopen and Dr. Jane M. Packard, Department of Wildlife and
Fisheries Sciences, Texas A&M University, College Station, Texas 77843-2258.
436 Bernal and Packard
A binational program of captive propagation (Mexico and USA) has been quite
successful for Mexican wolves [Siminski, 1993]. At the time this study was con-
ducted (winter 1990), there were only two breeding packs in captivity in Mexico. By
July 1996, the population had increased to 149 Mexican wolves in 30 facilities in
Mexico and the United States [Siminski, P., personal communication], and plans were
in progress for reintroduction of surplus animals to at least one site in the United
States [Parsons and Nicholopolous, 1995].
Although the value of direct observation for assessing behavioral profiles has
been acknowledged, statistical comparisons of behavioral differences of animals at
different facilities present certain difficulties. The purpose of this study was to deter-
mine differences in activity patterns and social behavior of one family group of Mexi-
can wolves maintained on-exhibit at San Juan de Aragón Zoo (SJA Zoo) and another
of similar social structure maintained off-exhibit in a naturalistic enclosure at San
Cayetano (SC Field Station). The zoo exhibit represents minimally enriched design,
whereas the field station enclosure represents a style thought to be suitable for adapting
potential candidates for release from captivity. Our objectives were to assess indi-
vidual variation in wolves at each facility related to age/sex category, time of day,
Each pack consisted of a breeding male, breeding female, and one juvenile
offspring. The history of each pack differed prior to and after the study (Fig. 1).
The exhibit at SJA Zoo (approximately 170 m2) was located in Mexico City
and was open to the public daily between the hours of 0900 and 1700. Its completely
concrete substrate included a dry moat and an upper platform with two indoor pens
(3 m × 2 m). The pack was fed 6 days a week (primarily horse meat supplemented
with chicken parts). Food was provided after the enclosure was cleaned in the morn-
ing, and the wolves were kept during the night in off-exhibit holding pens. General
health care procedures have been described by Ramos Magaña et al. .
In contrast, the SC Field Station was located within pine and oak woods in the
State of Mexico [Ferrusquia-Villafranca, 1993] and was not open to the general pub-
lic. The enclosure (1.3 ha.) provided a diverse, naturalistic environment with soil,
vegetation, a stream, and two holding pens for feeding. Feeding was 6 days per
week, with commercial dog chow supplemented occasionally with chicken parts.
During the study period, the animals were never captured and entry into the enclo-
sure was avoided as much as possible [Contreras, C., personal communication].
Both packs were observed by the senior author, using continuous focal group
sampling and a tape recorder [Martin and Bateson, 1993]. The pack at the SJA Zoo
was observed (62 hrs) for 2 days a week over 6 weeks during the months of January
to March (1990). The pack at the SC Field Station was observed (161 hrs) between 3
and 5 days a week from December (1989) to April (1990). Hours were distributed
unequally among the following three observation periods defined for analysis: morn-
ing (0700–0959), midday (1000–1259), and evening (1300–1900).
Standard definitions of behavioral categories [Packard, 1980; Zimen, 1981;
Servin, 1991; Derix, 1994] and observational techniques [Martin and Bateson, 1993]
were used. Activity was recorded by scan (sequential instantaneous) sampling with a
5-min sampling interval. The following behaviors were collapsed into the category
Activity Patterns and Social Behavior of Wolves 437
Fig. 1. The histories of the Mexican wolf packs at SJA Zoo and SC Field Station. Symbols represent birth (B), death (D with age in days, weeks
or months), transfers (T), genetic relationships (dashed lines), copulations (dots), presence in pack (solid lines), sex of individuals (male, female or
undetermined), and studbook identities (numbers in circles).
438 Bernal and Packard
of activity: walking, pacing, running, eating, drinking, scent-marking, and interact-
ing. The occurrence of double scent-marking [Mertl-Millhollen et al., 1986] was re-
corded as an index to evaluate the strength of pair-bonding. Double scent-marking
was defined as raised-leg urinations that were oriented on the same spot by both
members of the breeding pair within an observation period. The frequency of court-
ship interactions, scent-marking, active submission, play interactions, and aggressive
interactions were continuously recorded as all-occurrence event sampling. Courtship
behaviors included the following mutually exclusive categories: dance, genital in-
spection, tail avert, mount attempt, and copulatory-tie. Scent-marking included raised-
leg urination, scraping, scent-rolling, and defecation. Social interactions were
categorized as aggression, active submission, and play.
The Activity Index was based on the ratio of instants recorded as active, di-
vided by the total instants an animal was observed. To determine significant differ-
ences in the Activity Index of age/sex categories in each pack, the data were partitioned
by time periods (morning, midday, evening), and the Friedman Test [Lehner, 1979]
was applied, using a computer program [Abacus Concepts, Inc., 1992]. The effect of
time period on activity was analyzed separately for each individual, using the Kruskal-
Wallis Test of Mean Rank [Lehner, 1979]. To examine the effect of location on ac-
tivity, pairwise comparisons were made only between adults in each pack, using the
Mann-Whitney U Test [Lehner, 1979].
To compare the frequency of social behaviors within age/sex categories be-
tween packs, counts for each behavioral category were compared across individuals
using the Chi-square Goodness of Fit Test [Conover, 1980], adjusting expected values to
account for the unequal distribution of observation hrs between packs. The Log-likeli-
hood Ratio Test (G2) and Binomial Test (z) were calculated for more detailed analysis of
behavioral contingencies [Gottman et al., 1990], using a spreadsheet.
The two packs differed in terms of the relative activity patterns of individuals
(Fig. 2; SC Field Station, χ2 = 35.74, P = 0.0001; SJA Zoo, χ2 = 2.27, P = 0.3).
Individuals were not equally likely to be active at each of the three time periods
(Table 1). Activity levels of males differed significantly across packs for morning (U
= 175, P = 0.001) and midday (U = 58.5, P = 0.04), but not for evening periods (U =
74, P = 0.7), due to high activity of the male at the zoo. In a similar pattern, the
activity levels of adult females differed significantly for morning (U = 21, P = 0.001)
and midday (U = 46, P = 0.006), but not for evening periods (U = 65.5, P = 0.5).
Courtship interactions were more frequent at the SC Field Station than at the
SJA Zoo (Table 2). In the SJA Zoo, dance was lower than expected (z = –3.70) and
genital inspection was higher than expected (z = 3.0). The total frequency of mount
attempts was not different than expected by chance in both packs: SJA Zoo (z = –0.22)
and SC Field Station (z = 0.19). The reproductive period was unusually delayed at
the SC Field Station and two peaks of courtship occurred (Fig. 3). The highest peak
was recorded 2 weeks before estrus when the female was not fully receptive to the
male, yet received genital inspection and mounting attempts. During the estrus pe-
riod that resulted in conception (March 29–April 2), distinctive action patterns in-
cluded tail avert and copulation.
In contrast to similarities between the breeding females in each pack, the breed-
Activity Patterns and Social Behavior of Wolves 439
Fig. 2. The relative activity of individuals at the SC Field Station (shaded bars) and SJA Zoo (diago-
nally hatched bars). Horizontal bars indicate means, vertical bars indicate 95% confidence intervals.
TABLE 1. Relative activity patterns differed significantly among different periods of the day for
individuals in Mexican wolf packs at the SJA Zoo and SC Field Station
(mean ± 95% confidence interval) Kruskal-Wallis
Pack Morning Midday Evening test of mean ranka
Individual (n = 8) (n = 13) (n = 5) HaP
Breeding male 96.9 ± 3.6 64.8 ± 21.9 63.6 ± 37.0 8.46 0.0146
Breeding female 92.1 ± 6.8 68.8 ± 19.5 60.9 ± 35.8 11.07 0.0039
Juvenile 98.2 ± 2.7 74.0 ± 18.2 80.0 ± 36.4 7.56 0.0283
SC Field Station
Breeding male 56.9 ± 12.0 22.4 ± 17.7 50.6 ± 9.3 10.87 0.0044
Breeding female 33.5 ± 15.6 8.1 ± 14.7 37.5 ± 13.1 18.52 0.0001
Juvenile 74.8 ± 14.6 35.0 ± 23.7 76.4 ± 9.2 11.08 0.0039
aThe statistic was corrected for ties.
ing males differed significantly in frequency of scent-marking (Table 2). In SJA Zoo,
raised-leg urination was higher than expected (z = 7.09) and scraping was lower than
expected (z = –5.11). Relative to her partner, the female at San Cayetano did more
raised-leg urination (z = 2.29) and scent-rolling (z = 2.77) and less scraping (z = –3.46).
Between January 28, 1990, and March 25, 1990, 10 double scent-marks were ob-
served at the SJA Zoo, compared to 14 double scent-marks observed from December
29, 1989, to April 4, 1990, at the SC Field Station.
The packs differed significantly in total aggression and the pattern of aggres-
440 Bernal and Packard
TABLE 2. Frequency of social interactions and scent-marking in packs at SJA Zoo (SJA) and SC
Field Station (SC)
Comparison ObservedaChi-square Significance
Behavior Pack Individuals (expected) goodness of fit (P value)
Courtship SJA Male/female 122 (81) 29.4 <0.001
SC Male/female 166 (207)
Scent-marking SJA Male/female 118 (102) 3.4 >0.05
SC Male/female 247 (263)
Raised-leg SJA Male 112 (63) 52.9 <0.001
urination SC Male 113 (162)
Aggression SJA Male/female 60 (30) 41.8 <0.001
SC Male/female 47 (77)
Active submission SJA Juvenile 87 (55) 26.7 <0.001
SC Juvenile 108 (141)
Play SJA Juvenile 182 (72) 235.7 <0.001
SC Juvenile 74 (184)
aExpected values were calculated by multiplying the total number of events by the proportion of total
hours that each pack was observed.
Fig. 3. The courtship activity of Mexican wolf pairs at the SJA Zoo and at the SC Field Station
during the 1989–1990 breeding season. The thick horizontal bars indicate periods of estrus as deter-
mined by the occurrence of the tail-avert action.
Activity Patterns and Social Behavior of Wolves 441
sion (Table 2). In the SJA Zoo, male-to-juvenile aggression was higher than ex-
pected (z = 2.39) and female-to-male aggression was lower than expected (z = –4.5).
At the SC Field Station, the male was not aggressive to the female; aggression of the
male to the female at the SJA Zoo occurred, but infrequently.
The packs did not differ significantly in total frequency of active submission
(Table 2), but differed significantly in the pattern of active submission (G2 = 8.18; df
= 1; P < 0.01). In the SJA Zoo, active submission from juvenile to male was lower
than expected (z = –2.12), but was higher than expected (z = 2.12) between juvenile
and female. At the SC Field Station, the juvenile was equally submissive to the adult
male and female. Play behavior was significantly more frequent at the SJA Zoo than
at the SC Field Station (Table 2). At the SC Field Station, play between the male and
juvenile was higher than expected (z = 8.02).
The direct-observation techniques employed in this study were useful for com-
paring activity and frequency of specific behaviors across packs for individuals in
similar age/sex categories. In making statistical comparisons, we found that it was
necessary to account for differences in activity relative to time of day and differ-
ences in the total time each pack was observed.
Initially, we were surprised that activity was greater at the zoo than the field
station, because we expected the natural vegetation to function as behavioral enrich-
ment. However, considering the higher pattern of morning activity in the zoo, we
hypothesize high activity could have been associated with the daily pattern of clean-
ing, occasional separation at night, and/or stimulation by vocalization of coyotes
(Canis latrans) in an adjacent pen. In addition, observation conditions also varied
between both enclosures; visibility at the Field Station Pack was obscured by veg-
etation and wolves may have been active when they were out of view.
We interpret the differences in courtship behavior in the two packs as more
related to individual differences (and possible pathology) than to the physical environ-
ment at the two facilities. The pattern of courtship and estrus shown by the pair at the zoo
was more like that described for other subspecies [Packard, 1980]. Although the higher
frequency of scent-marking, rolling, and scraping at the Field Station might have been
related to the soft substrate and vegetation, it was not indicative of the pattern usually
associated with a strong pair-bond. Double scent-marking is usually considered indica-
tive of a strong bond in wolves. Although it occurred in both packs, it was low relative to
the prolonged courtship and delayed estrus at the field station. The relatively low fre-
quency of raised-leg urination and high frequency of scraping by the male at the field
station also deviated from the norm. Since the policy was “hands off” at the field station,
no physiological data were available to assess possible pathology.
The relationships between the juvenile and breeding male differed greatly be-
tween packs. The breeding male at the zoo was more likely to direct aggression and
less likely to receive active submission (derived from food begging) and play from
the juvenile. In the wild, both the male and female actively participate in the rearing
of cubs [Packard et al., 1992]. We interpret this difference in juvenile/adult relation-
ships as related to the practice of isolating the female at the SJA Zoo prior to parturition
and during the first 3 months postpartum. At the SC Field Station, both the breeding
male and female were present and participated actively in the rearing of the pup.
442 Bernal and Packard
Direct behavioral observations are useful in making comparisons between indi-
vidual wolves living in different facilities. Comparisons at the two sites indicate the
zoo facility is more appropriate for intensive husbandry and production of pups.
However, the practice of separating the male and female near parturition should be
considered carefully, since the social bond between juvenile and adult male is impor-
tant to the goals outlined by the Species Survival Plan. In these two packs, the quiet,
undisturbed, naturalistic environment of the larger enclosure did not stimulate the
activity that may be required for physical conditioning of wolves selected for rein-
troduction. Similar techniques of comparing activity and social behavior might be
used in choosing between individuals as candidates for reintroduction and assessing
the suitability of individuals and pairs for reintroduction.
We thank the Secretaría de Desarrollo Social (SEDESOL) and the San Juan de
Aragón Zoo for their support during the study. Specifically we thank the following
persons for their enthusiasm during the study: Luis Gerardo López Lemus, Carlos
Contreras, Gerardo Tapia, Ariel Alarcon, José Maria Reyes, Eleazar Loa, Mauro
Reyna, Patricia Reyes, Gerardo López Islas, Bernardo Manriquez, Guillermo Islas,
and Fernando Camacho. We also thank Gerardo López Islas, Fernando Cervantes,
Gerardo Ceballos, Mauro Ivan Reyna, Carmen Vazquez, Wendy Brown, Peter
Siminski, and Ed Spevak for reviewing earlier versions of the manuscript. This pa-
per is dedicated to the memory of Josefa, a very special Mexican wolf.
Abacus Concepts, Inc. STATVIEW. Berkeley, CA,
Abacus Concepts, Inc., 1992.
Conover, W.J. PRACTICAL NONPARAMETRIC
STATISTICS. New York, John Wiley and Sons,
Derix, R.R.W.M. THE SOCIAL ORGANIZATION
OF WOLVES AND AFRICAN WILD DOGS:
AN EMPIRICAL AND MODEL-THEORETI-
CAL APPROACH. Utrecht, Universiteit Utrecht,
Ferrusquia-Villafranca, I. Geology of Mexico: A
synopsis. Pp. 3–108 in BIOLOGICAL DIVER-
SITY OF MEXICO: ORIGINS AND DISTRI-
BUTION. T.P. Ramamoorthy; R. Bye; A. Lot; J.
Fa; eds. Oxford, Oxford University Press, 1993.
Gottman, J.M.; Roy, A.K. SEQUENTIAL ANALY-
SIS: A GUIDE FOR BEHAVIORAL RESEARCH-
ERS. Cambridge, Cambridge University Press,
Kleiman, D.G. Foreword: Animal behavior stud-
ies and zoo propagation programs. ZOO BIOL-
OGY 13:411–412, 1994.
Kreeger, T.J.; Pereira, D.L.; Callahan, M.; Beckel,
M. Activity patterns of gray wolves housed in
small vs. large enclosures. ZOO BIOLOGY
Lehner, P. HANDBOOK OF ETHOLOGICAL
METHODS. Colorado State University, Garland
STPM Press, 1979.
MacDonald, K.B. Activity patterns in a captive wolf
pack. CARNIVORE 3:62–64.
Martin, P.; Bateson, P. MEASURING BEHAVIOUR.
Cambridge, Cambridge University Press, 1993.
Mertl-Millhollen, A.S.; Goodman, P.A.; Kling-
hammer, E. Wolf scent marking with raised leg
urination. ZOO BIOLOGY 5:7–20, 1986.
Packard, J.M. DEFERRED REPRODUCTION IN
WOLVES (Canis lupus). Ph.D. Dissertation, Uni-
versity of Minnesota, 1980.
Packard, J.M.; Mech, L.D.; Ream, R.R. Weaning in
an Arctic wolf pack: Behavioral mechanisms. CA-
NADIAN JOURNAL OF ZOOLOGY 70:1269–
Parsons, D.R.; Nicholopolous, J.E. An update on
the status of the Mexican wolf recovery pro-
gram. Pp. 141–146 in ECOLOGY AND CON-
SERVATION OF WOLVES IN A CHANGING
WORLD. L.N. Carbyn; S.H. Fritts; D.R. Seip;
eds. Edmonton, University of Alberta Press,
Ramos Magaña, X.; Mata Castro, F.; López Islas,
G.; Islas Dondé, G. Lobo Méxicano (Canis lu-
pus baileyi) aspecto zootecnico en el Zoológico
San Juan de Aragón. Pp. 51–56 in MEMORIAS
Activity Patterns and Social Behavior of Wolves 443
PRIMER SIMPOSIUM NACIONAL SOBRE EL
LOBO GRIS MEXICANO. M.I. Reyna Medrano;
G. López Islas; B. Manrique Nobara; eds. Mexico
City, Zoológico de San Juan de Aragón, 1993.
Reyna Medrano, M.I.; López Islas, G.; Manrique
Nobara, B., eds. MEMORIAS PRIMER SIMPO-
SIUM NACIONAL SOBRE EL LOBO GRIS
MEXICANO. Mexico City, Zoológico de San
Juan de Aragón, 1993.
Siminski, P. The Mexican wolf recovery effort
in the United States of America. Pp. 212–219
in MEMORIAS. 4° CONGRESO LATINO
AMERICANO DE PARQUES ZOOLOGICOS,
ACUARIOS Y AFINES. Puebla, Zoológico
Sede African Safari Puebla, 1993.
Servín, J. Algunos aspectos de la conducta social
de lobo Mexicano (Canis lupus baileyi) en
cautiverio. ACTA ZOOLOGICA MEXICANA
Zimen, E. THE WOLF: HIS PLACE IN THE
NATURAL WORLD. London, Souvenir Press Ltd.,