reciprocal allogrooming in dam-reared and hand-reared impala fawns

Ethology 90, 37-51 (1992) 0 1992 Paul Parey Scientific Publishers, Berlin and Hamburg ISSN 0179.1613

Department of Physiological Sciences, University of California, Davis

Reciprocal Allogrooming in Dam-reared and Hand-reared Impala Fawns

MICHAEL S. MOORING & BENJAMIN L. HART

MOORING, M. S. & HART, B. L. 1992: Reciprocal allogrooming in dam-reared and hand-reared impala fawns. Ethology 90, 37-51.

Abstract

Among adult females and males of African antelope impala are unique in their performance of reciprocal allogrooming. The occurrence of this behaviour in neonatal impala fawns was explored in a free-ranging impala herd at the San Diego Wild Animal Park where 5 dam-reared fawns were observed from birth through 10 weeks of age. One-way maternal grooming And reciprocal allogrooming with the dam and non dam partners emerged as distinct behavioural systems. Maternal grooming, directed mostly to the anogenital area, was typical of that seen in other ungulates, and sharply declined over the first two weeks. Reciprocal allogrooming, characterized by alternate exchanges of grooming bouts with a partner in the same manner as in adults, was seen as early as 3-8 d after birth. All fawns were grooming with unrelated adult females by the end of the second week. By week 2 virtually every measure of reciprocal allogrooming by fawns (grooming delivered per hour, reciprocity, and percent of encounters initiated) was as high as for adults. The appearance of this reciprocal allogrooming pattern, especially at such an early age, appears to be unique among ungulates, and possibly mammals in general. Three hand-reared impala fawns, deprived of the opportunity to interact with older herdmates, but having access to impala fawns and heterospecific fawns, were observed from 1-3 mo of age. The hand-reared impala showed no alteration in the occurrence of reciprocal allogrooming behaviour compared with the dam-reared control fawns, indicating that allogrooming experience with older animals was not required for the appearance of reciprocal allogrooming at an early age. Interestingly, hand-reared fawns persisted in grooming heterospecific fawns despite the fact that heterospecifics rarely reciprocated grooming. We postulate that the strong predisposition for impala young to groom others may be related to the threat of tick infestation in the impala’s ecotone habitat.

Corresponding author: Benjamin L. HART, Department of Physiological Sciences, School of Veterinary Medicine, University of California, Davis, C A 95616.

Introduction

Grooming of another conspecific (allogrooming) is common among certain species of primates, rodents, ungulates and even birds. In adults such grooming may take the form of simultaneous mutual grooming, such as seen in equids, or

U.S. Copyright Clearance Center Code Sratement: 01 79-1613/92/9001-0037$02.50/0

38 MICHAEL s. MOORING & BENJAMIN L. HART

alternate reciprocal grooming, such as seen in primates and impala. Allogrooming involving infant animals is almost always unidirectional grooming of the infant by the mother. Simultaneous and reciprocal allogrooming (if it occurs) occurs later, usually in a social context. Both social (bond maintenance, appeasement, alliance formation) and utilitarian (maintenance of pelage, ectoparasite removal) functions have been attributed to allogrooming in adults.

A reciprocal allogrooming system in adult impala, a medium-size antelope existing in woodland areas throughout Eastern and Southern Africa, has been described which appears to be unusual among ungulates in the high degree of reciprocity, and lack of influence of dominance or genetic relatedness in partner preference (HART & HART 1988, 1992). In this behavior one animal turns to another and delivers a bout of 6-12 grooming episodes to the partner’s head or neck. The partner usually reciprocates with a similar bout of episodes to the first animal’s head or neck. The encounter continues in this tit-for-tat-like exchange of bouts until one partner does not return a bout. Allogrooming appears to benefit the partner through removal of ectoparasites, especially ticks, and the animal performing the grooming incurs costs in the way of a modest amount of energy expended, loss of water and electrolytes from saliva used, distraction in vigilance for predators and attrition of the lateral dental elements (see HART & HART 1992). HART & HART (1992) proposed that impala allogrooming may be a candidate for the tit-for-tat model of evolved cooperation described by AXELROD & HAMILTON (1981) and AXELROD & DION (1988).

We report here that the form of reciprocal allogrooming seen in adult impala emerges in captive free-ranging impala fawns, living in a naturalistic environment, during the second week after birth. We also provide evidence that the behavior appears to represent a strong inherited predisposition to deliver grooming to other animals, even heterospecifics, almost immediately after birth. These find- ings reveal a reciprocal allogrooming system that is apparently unique among mammalian species and raises questions about the evolution of this unique system.

The free-ranging herd of 35-38 impala (Aepyceros melumpus) at the San Diego Wild Animal Park (SDWAP) provided an opportunity for detailed observations on the interactions between newborn impala fawns with other fawns and adult herd mates. Preliminary observations had revealed that SDWAP females engage in the same type of reciprocal allogrooming as seen in wild populations in Kenya. Study 1 quantified the pattern of reciprocal allogrooming behavior seen in dam-reared fawns in the free-ranging herd. Study 2 examined the effects of deprivation of the normal herd allogrooming experience on the occurrence of allogrooming in fawns raised by hand from the day of birth.

Study 1 : Development of Reciprocal Allogrooming in Dam-reared Fawns

Ungulate newborn typically receive intensive one-way licking from the mother immediately after birth. For several days or weeks thereafter one-way maternal grooming generally occurs during nursing and is concentrated on the

Reciprocal Allogrooming 39

anogenital region (LENT 1974; LEUTHOLD 1977; MOORING 1989; WALTHER 1984). This maternal grooming remains one-sided and gradually drops out of the mother-offspring interaction repertoire. Allogrooming occurs in a more limited number of species, when the newborn is typically several weeks old, and then the grooming is with the mother. In this study we explored the concept that maternal grooming and infant allogrooming in impala fawns occur as two separate systems.

Methods

Subjects

The impala herd, which numbered 35 individuals in 1988 (yr I ) and 38 in 1989 (yr 2), inhabited the 50-ha "East Africa" enclosure of the SDWAP, along with 13 other species of East African ungulates. The herd size and composition approximated that found in the wild in Zimbabwe (30-150 animals; MURRAY 1982), except that only one mature sire male was present. The area of the enclosure also fell within the natural range of female home-range size reported by MURRAY (1982). The impala were free to graze ad libitum on Kikuyu grass (Pennisetwn culdeniseturn) found throughout the enclosure, and were supplemented with fiber herbivore pellets. The SDWAP appears to be free of ectoparasites inasmuch as blanket-drag tick censusing (ZIMMERMANN & GARRIS 1985) and parasite counts conducted on immobilized animals 0. OESTERHUIS, pers. comm.) routinely revealed no ectoparasites. The observations o n fawns were made from mid-June through August.

Definitions

Impala allogroorn by delivering bouts of 6-12 grooming episodes which consist of licking with the tongue or combing the pelage with the lower incisor-canine complex (MCKENZIE 1990). Episodes are delivered at the rate of about 11s. A complete allogrooming exchange between a pair is termed an encounter. The animal who delivers the first bout of an encounter is the initiator. Self-grooming consists of bouts of either 8-12 oral grooming episodes using the lower incisor-canine complex or 7-9 scratch grooming episodes (HART et al. 1992). The reciprocity index [no. bouts o r episodes delivered/(no. bouts or episodes delivered + received)] is a measure of within-encounter symmetry that can range from the extremes of 0 and 1.0, with perfect reciprocity indicated as 0.5.

Observations

Daylight observations were made with a 15-60 X zoom spotting scope primarily between 6.00 and 12.00 h from a point overlooking the herd at a distance of 5&200 m. We used focal-animal sampling and sampling all occurrences of certain behavior patterns (ALTMANN 1974). Age classes were designated according to age at the start of observations: fawns, &6 mo.; juveniles, 6 1 7 mo.; adults over 17 mo. 83 h of focal observations were conducted on 5 fawns during yr 1 and yr 2. Focal samples on fawns lasted 20 min and were initiated when the focal animal was standing, but continued if the tawn lay down. For comparison of grooming by fawns with older impala focal observations were made on 20 adult females and 5 juvenile impala during yr 1 and 23 adult females and 9 juveniles during yr 2 (total of 243 h). Data on fawns for the first 10 weeks following birth were based on observations of 3 fawns in yr 1 (165, 166, 167) and 2 fawns in yr 2 (185, 188). Of these, 3 fawns were observed from the day of birth, 1 was observed from day 2, and 1 from day 7. All animals were individually marked with ear notches and ear tags affixed at birth and could be individually identified. The data were recorded on a portable computer programmed for real-time event and state recording.

Data Analysis

Data were statistically analyzed using Mann-Whitney, Kruskal-Wallis and Friedman analyses of variance (ANOVA), proportions test, Spearman rank-correlation non parametric procedures, and the one-sample t-test (DANIEL 1990; ZAR 1984). Multi-way A N O V A by ranks were done by performing the F-test on ranked data. Because measures made on the same individuals over time are not

40 MICHAEL S. MOORING & BENJAMIN L. HART

independent, separate correlations were performed on each of the 5 fawns for measures of allogrooming and fawn age. For the fixed-effects analysis (results apply only to animals studied), the pooled Z score, Z,, was calculated as Z, = Z Z / m, where N is the total number of Z scores summed, or 5. For the random-effects model (results can be generalized), the mean of the 5 Spearman correlation coefficients generated for each measure were compared with the null value of zero using the one- sample t-test. Computations were performed on the Stat View program (FELDMAN & GAGNON 1986) and on mainframe SAS version 6.06 (SAS Institute 1989). All tests were two-tailed with the level of significance set at 0.05.

Results

Maternal One-way Grooming

Fawns 165 and 166 were observed immediately following parturition. In both cases the dam began directing grooming to the newborn almost immediately after birth. Grooming consisted of vigorous licking to all parts of the fawn’s body, and continued virtually nonstop until the neonate had been presumably thoroughly cleaned and dried. Fawn 165 received 127 bouts of 4427 licks in 115 min and fawn 166 received 108 bouts of 2598 licks in 73 min following birth. During the first week these fawns received frequent one-way grooming from the dam whenever they suckled. Maternal grooming was associated with nursing in 73 % of the 108 nursing bouts observed. This grooming was directed to the flanks, belly, back, rump, tail, hindleg, and especially the anogenital region of their neonate. Females also delivered one-way grooming (mostly teeth scrapes) to the head and neck of offspring fawns not engaged in suckling. Over both years 62 % of maternal grooming occurred while the fawn was nursing (N= 123), with 95 % of grooming-while-nursing encounters taking place during the first 2 weeks. Considering all maternal grooming, 73 % occurred during week 1 and 91 % occurred within the first two weeks.

Fig. 1: Typical alternation of grooming bouts within an encounter between two impala fawns. Note on the left side the facilitation posture of the groomed individual, who holds its head and neck down, allowing the groomer access to the dorsal neck region. Each side represents one episode of several grooming episodes which are delivered during a bout. Figures drawn from slides by Kim FRALEY


Fig. 2: Typical use of the lower incisor- canine complex in impala mutual grooming during an exchange of grooming bouts between two impala fawns. The lateral dental elements comb through the recipient’s pelage

Reciprocal Allogrooming by Fawns

In the 4 fawns observed during week 1, the first recorded reciprocation of grooming ranged from 3-8 d after birth. In all 4 fawns the first recorded reciprocation was directed to another fawn, and in 3 out of 4 of these instances the encounter was initiated by the younger fawn. The first recorded reciprocal allogrooming with an adult female (dam or non dam) ranged from 5-12 days. In 2 fawns the first recorded reciprocal grooming was with the dam. In the other fawns it was with an unrelated adult female. In one of the fawns who was observed to engage in reciprocal grooming with a nonrelated adult female before the dam, grooming with the dam was not observed until day 62. The structure of the early allogrooming exchanges appeared to be virtually identical in topography to that engaged in by adult impala (HART & HART 1992). Grooming was directed to the head and neck of the partner as bouts of 6-20 episodes with partners alternating in delivering bouts (Fig. I). Teeth scraping with the lateral incisor- canine complex, as described for allogrooming and self grooming in adult impala, was frequently seen (Fig. 2), but not quantifiable because of less than ideal visibility in distinguishing between licking and teeth scraping.

The amount of reciprocal grooming performed with the dam was highly variable among the 5 fawns. In 3 fawns only 2 encounters with the dam were observed, while in the other 2 fawns 17 and 21 encounters, respectively, were observed. Data from the two fawns that engaged in significant amounts of reciprocal grooming with their dam were used to compare reciprocity when allogrooming with the non dam adults (69 and 144 encounters). In this limited sample, the reciprocity with dam and non dam partners was virtually identical


with regard to bouts (0.66 f 0.08 and 0.66 k 0.02; mean f SE) and episodes (0.70 k 0.10 and 0.67 f 0.02). Reciprocal grooming interactions with the dam remained fairly constant over time in respect to grooming with other partners as shown by plotting the proportion of all reciprocal grooming encounters that were either with the dam or non dam partners (Fig. 3).

Young fawns performed a behavior, referred to as lip-licking, in which they attempted to lick the mouth of another impala with small, rapid tongue strokes. This behavior was performed exclusively by fawns between 13 and 42 days of age. 60 % of the 30 observed instances of lip-licking were directed to fawns, and the remaining 40 % directed to non dam adults. Lip-licking occasionally graded into reciprocal grooming by the fawn.

A notable difference between fawns and adults involved the use of the neck presentation posture as an apparent solicitation of grooming from potential partners. Solicitation for grooming was recorded when one animal approached another and lowered its head and neck which had the effect of making the head and neck region accessible to grooming. This behavior appeared to function as an invitation to a potential grooming partner to deliver the first bout inasmuch as it often resulted in grooming being delivered. At other times, the partner rejected the "invitation" by walking away, or the solicitor itself delivered the first bout after an initiating bout from the partner was not forthcoming. The presentation pose was also frequently observed within encounters. In order to examine the frequency of solicitation behavior a subset of the observed encounters was analyzed in which solicitation or lack of solicitation was noted in detail. Con- trasted with adults in which 40.2 % of allogrooming attempts were preceded by solicitation, only 3.4 % of fawn grooming attempts were preceded by solicitation

. ""

2 8 0 - Q)

C c

8 60- C w

2 Q)

4r

With non-dam partners

C w

. ""

8 0 -

60-

4r

With non-dam partners

c P, c 40- ,' '. With Darn # '\

o-., ,' .n-------- ' 8 .

'. -- -+ p---- Q)

a' 1 0' - - - - - - - -

0

Days From Birth Fzg. 3: Distribution of fawn reciprocal allogrooming encounters between those with dam

and those with non dam partners from 3 to 10 weeks of age

P, ,' '. With Darn # '\

40-

o-., ,' .n-------- ' 8 .

'. -- -+ 2 0 - p----

a' 1 0' - - - - - - - -

0


(p < 0.001). Of the 8 solicitations by fawns that were observed, 3 were directed to adults or juveniles (once to the dam) and 5 to other fawns.

In order to compare allogrooming rate per h and reciprocity by fawns with that by adults, these measures were combined for fawns grooming with all partners because these measures did not significantly differ as a function of partner. The data for fawns are grouped according to week 1, week 2 and weeks 3-10. Since all measures of grooming behavior involving adults and juveniles were virtually identical, the data on these subjects were combined and presented as adults. For the 10 weeks of observation on each fawn there was a significant positive correlation between grooming (bouts and episodes) delivered per h by the fawn to all partners and fawn age (mean rs = 0.63 and = 0.68, respectively for bouts and episodes, p < 0.001). By week 2 the fawns were delivering reciprocal grooming at significantly higher rates per h than were adult animals (Table 1). The number of episodes per bout given by fawns during allogrooming showed no significant correlation with age (p = 0.84) and there was no significant difference in the episodes per bout given by fawns compared with that given by adults (mean, 14.5 and 10.9 episodes per bout respectively).

The mean reciprocity values for the 5 fawns from weeks 2-10 are presented in Table 2 along with reciprocity of adults. The mean values for adults were not significantly different from the expected value of 0.5 for perfect symmetry for

Tubable 1: Comparison of reciprocal allogrooming bouts and episodes delivered per h by fawns and adult females (i k SE)

Subjects Grooming delivered per h

Fawns

Time periods bouts episodes week 1 0.75 (f 0.34) a 12 (f 7) a week 2 6.38 (k 1.34) b 184 (k 116) b weeks 3-10 10.45 (k 2.04) b 154 (k 35) b

Adults 3.82 (k 1.24) a 70 (k 34) a

Different letters in same column represent significant differences (p < 0.05).

Table 2: Comparison of reciprocity index and % initiations of reciprocal allogrooming by impala fawns and adults (X f SE)

Subjects Reciprocity % Initiations bouts episodes

Fawns

Time period

week 2 0.49 (k 0.06) 0.51 (k 0.06) 54 (f 8) weeks 3-10 0.46 (k 0.05) 0.46 (k 0.05) 42 (k 9)

Adults 0.51 (5 0.02) 0.50 (k 0.06) 50 (k 2)


CI

C Q) 0

0.25 L Q) n

J 0.00

Fig. 4: Reciprocity index for bouts (left) and % of encounters initiated (right) by impala fawns allogrooming with dam (D), non dam adults (A), and other fawns (F) at 3-10 weeks of age (i * SE, ns). A reciprocity of 0.5 is exact reciprocity; less than 0.5 means the focal animal receives more than it

delivers

both bouts and episodes, and there was no difference between the reciprocity of adults and that of fawns during either week 2 or weeks 3-10. In fact, by 3 weeks of age when one-way maternal grooming had all but disappeared reciprocal grooming with the dam was also near 0.5 reciprocity (Fig. 4). The percentage of encounters initiated by fawns also reflected symmetrical reciprocity as early as 2 weeks of age (Table2). Throughout weeks 3-10 fawns of focal observations initiated about half of encounters with the dam and other fawns and about a third with non dam adults (ns, Fig. 4). There was no difference in these measures between fawns of week 2 or weeks 3-10 and adults.

Study 2: Allogrooming in Hand-reared Fawns

The precocious development of reciprocal allogrooming in fawns in the free- ranging herd suggested an inherited predisposition to deliver grooming to other animals. Because fawns had the opportunity to interact with older animals who regularly groomed, social learning of allogrooming seemed possible. The removal of 3 impala fawns from the herd at birth, who were then hand-reared together with heterospecific orphaned neonates of various ungulate species, provided an opportunity to observe allogrooming in hand-reared impala fawns deprived of interactions with adults. While not deprived of interactions with impala age- mates (the three fawns could interact among themselves), they were deprived of interactions with older impala.

If social learning is necessary for the normal occurrence of reciprocal allogrooming, then hand-reared fawns deprived of grooming interactions with experienced adult conspecifics should manifest differences in the pattern of reciprocity, amount of allogrooming performed per h, or other measures of allogrooming compared with dam-reared fawns. Because the heterospecific fawns


in the enclosure with the impala fawns were of species that do not normally engage in reciprocal allogrooming, they were not expected to reciprocate grooming to impala. If the initiation of allogrooming relies on the return of grooming for reinforcement, then grooming directed toward heterospecifics should dimin- ish over time (provided that heterospecifics do not generally reciprocate), while that displayed toward conspecifics should continue. O n the other hand, if reinforcement is not involved in a major way in the development of reciprocal allogrooming, then hand-reared fawns should groom the same as dam-reared fawns and grooming initiated with heterospecifics should not decline.

Methods

Observations were conducted on 3 hand-reared fawns (177, 179, and 183) from 25 July to 2 Sep. 1988 at the Infant Isolation Unit, in a 297-m’ fenced enclosure. All animals were marked. The 3 impala fawns were housed together with 4-7 heterospecific fawns that were periodically added and removed from the enclosure. Heterospecifics were of the following species: addra gazelle (Gazella duma rzrficolhs), barasingha (Cervus duvauceli devauceli), Burmese thamin (Cervus eldi thamin), fringe- eared oryx (Oryx gazella callotzs), gemsbok (Oryx gazella gazella), Nubian ibex (Capra ibex nubtana), roe deer (Capreolus capreolus capreolus), Russian saiga (Saiga tatarica tatarica), sand gazelle (Gazella subguttorosa marica), and Angolan springbok (Antzdorcas marsupialis angolensis).

Observations totalling 53 h were made by pairs of observers during 2-h sessions, 5 d per week between 7.00 and 10.00 h. O n e observer watched the animals while the other recorded the information spoken by the first observer onto a standardized data sheet. Information on allogrooming and self-grooming were recorded in random order from the 3 impala fawns during 20-min focal animal samples as in Study 1. Fawn ages during the observation period ranged between 34 and 93 days. Similar data were extracted from the 20-min focal samples of 5 dam-reared fawns of Study 1 for the same time period and fawn ages as with the hand-reared fawns (total of 26 h). The mean age at the start of study was 44 days and 50 days for dam-reared and hand-reared fawns, respectively.

Because grooming activity was recorded differently for hand-reared than dam-reared fawns (verbal counting versus punching a counter key on the computer), and counting proved to slightly underestimate the episodes per bout, the counting method used with dam-reared fawns was adjusted by applying a conversion factor of 0.87 calculated from the inter-observer reliability data, in which both counting methods were used to record the same grooming events.

Additional observations were counted on a yearling female (169) that had been hand-reared one year earlier and released into the East African enclosure. She did not join the impala herd but instead associated with Thomson’s and Grant’s gazelles (Gazella thomsonii; G. granrz). Grooming data were collected on this subject during 1-2 h focal samples, for a total of 13 h using the same methods as described for the dam-reared fawns.

Results

For some measures of allogrooming, the data were combined for the total observational period on the basis that Study 1 found no relationship within this age period and rate of grooming, amount of grooming per encounter, reciprocity, or percent of encounters initiated. The behavioural topography of allogrooming by hand-reared impala fawns was indistinguishable from that typical of dam- reared fawns examined in Study I. In grooming encounters with conspecifics, bouts of mostly teeth-scraping episodes alternated back and forth between partners as typically seen with dam-reared fawns. Grooming was initiated by approaching the prospective partner and delivering the first bout, as previously described for the herd fawns.


A quantitative comparison between dam-reared and hand-reared fawns is confounded by the fact that hand-reared fawns could not groom with impala adults and dam-reared fawns had no access to heterospecifics. To obtain a comparison of overall allogrooming activity the numbers of bouts and episodes delivered per h to all partners were compared (Table 3). No differences were found. For other analyses only data from hand-reared and dam-reared fawns when grooming only with conspecific fawns were utilized. Such a comparison revealed that although bouts delivered as well as bouts and episodes received per encounter did not differ, dam-reared impala delivered significantly more episodes per encounter (Table 4). The percentage of allogrooming initiations to conspecifics by hand-reared fawns that were reciprocated (43.3 %) was not significantly different from the reciprocation to initiations by dam-reared fawns (47.3 Yo). As expected, there was no difference between rearing groups in the proportion of encounters initiated nor in the frequency with which fawns solicited grooming with the presentation posture (2.6 % and 3.4 Yo for hand- reared and dam-reared fawns respectively).

Tuble 3: Rates of reciprocal allogrooming delivered per h by dam-reared and hand-reared impala fawns grooming with all partners (2 5 SE)

Fawn type Behavioral measure bouts/h episodelh

Dam-reared 178 181 182 I85 188 X

Hand-reared 177 179 183 X

-

-

4.64 2.69 9.38 3.74 9.19 5.93

7.87 5.85 6.67 6.80

92 31 81 23

158 77

51 47 55 51

Includes dam-reared grooming with adult female impala and other fawns and hand-reared grooming with heterospecific and conspecific fawns.

Tuble 4: Grooming bouts and episodes delivered and received per encounter for hand-reared and dam-reared fawns when grooming with conspecific fawn partners (i + SE)

Fawn type Behavioral measure bouts episodes

delivered received delivered received

Hand-reared 2.08 (5 0.18) a 1.92 (? 0.17) a 15.6 (+ 2.24) a 14.4 (+ 1.53) a Dam-reared 2.45 (+ 0.40) a 1.68 (+ 0.14) a 24.6 (!c 3.68) b 16.6 (k 2.76) a

Different letters in same column represent significant difference (p < 0.05).

Rcciprocal Allogrooming 47

Fig. >; Hand-reared impala fawn delivering grooming to a heterospecific fawn, who passively receives the grooming without attempting to reciprocate

The hand-reared fawns approached and initiated grooming with heterospecific fawns the same as with conspecifics (Fig. 5 ) . However, heterospecifics reciprocated only 3.4 Yo of all grooming initiations by hand-reared fawns, compared with 43.3 "/o reciprocated by hand-reared conspecifics. When an initiation was returned it consisted of a mean of 1.4 bouts (9.2 episodes). Bouts and episodes delivered to heterospecifics were significantly less than those to conspecifics (bouts 1.38 and 2.08; episodes 11.7 and 15.6 respectively). Episodes per bout given to conspecifics did not differ from those to heterospecifics (7.4 and 8.4 episodes per bout respectively). To examine whether grooming initiations to heterospecifics decreased over time, the study period was divided into two time phases, the first 20 d (phase I ) and the last 19 d (phase 2 ) . Initiations to heterospecifics by each fawn were compared between the two time phases. No difference was evident in the mean number of initiations per h or in the mean percentage of all allogrooming encounters (both conspecifics and heterospecifics) that were directed toward heterospecifics (Table 5). Furthermore, if initiations to non reciprocating heterospecifics diminished over time, one would expect to see fewer such initiations by older fawns. O n the contrary, older fawns (ix., fawns with lower identification numbers) tended to initiate grooming to heterospecifics at higher rates than younger fawns (Table 5 ) . Interestingly, on the 10 occasions when heterospecific fawns initiated grooming to impala, none of these initiations were reciprocated.


Table fi: Allogrooming initiations directed to heterospecific fawns by hand-reared impala fawns during Phase 1 (the first 20 d of the study) and Phase 2 (the last 19 d) as indicated by initiations per h

and % of all allogrooming encounters (conspecifics and heterospecifics)

Fawn No. Initiations per h % initiations Phase 1 Phase 2 Phase 1 Phase 2

177 3.60 3.21 47.8 35.7 179 0.66 1.83 28.6 38.9 183 0.68 0.52 18.8 13.3 X (k SE) 1.64 (k 0.98) 1.85 ( 5 0.78) 31.7 (k 8.5) 29.3 (k 8.1) -

Fawns with lower identification numbers are older.

During the 13 h of observations on female impala 169 (hand-reared the previous year), who stayed only with Thomson’s and Grant’s gazelle, she initiated grooming to Thomson’s gazelle 14 times and to Grant’s gazelle 2 times (X 1.38 bouts and 14 episodes per h). These grooming initiations were never reciprocated by the heterospecific. The rate at which 169 delivered grooming, while less than that of herd females (X 3.92 bouts and 75 episodes per h), fell within the 95 % confidence intervals for grooming given per h by females in the herd. The main difference is that impala herd females exchanged several bouts per allogrooming encounter while 169 delivered one bout during an encounter, which was the pattern seen with fawns grooming with heterospecifics.

Discussion

Longitudinal observations of impala fawns at the SDWAP from birth to 10 weeks of age revealed two distinct allogrooming systems, maternal one-way grooming and reciprocal allogrooming. The changes in maternal one-way grooming were similar to those of other ungulate species. Intensive generalized licking of the fawn’s entire body immediately after birth completely scrubbed and dried the newborn and stimulated it to suckle (see MOORING & RUBIN 1991 for details of early nursing). Throughout the first week the dam directed one-way grooming to the fawn. One-way maternal grooming frequently occurred while the fawn suckled, the mother directing licks to the posterior region of the fawn, especially the anogenital region. Maternal grooming quickly declined and was rarely seen after 14 days. Anogenital grooming of young by the dam is typical of ungulate species in which neonates hide out for the first weeks of life (LENT 1974; WALTHER 1984). Although impala fawns hide out, they d o so for only a few h (JARMAN

1976) or a few d (SCHENKEL 1966) after birth. The basic pattern of reciprocal allogrooming, as seen in adults, occurred by

week 2 when fawns were observed grooming unrelated fawns and adults. The behavioral topography, including use of the incisor-canine complex for teeth scraping of the head and neck of the partner, was the same as with adults. From week 2 fawn allogrooming was no more likely to be directed to the dam than to any other adult female in the herd. Grooming rate, reciprocity, and percentage

Reciprocal Allogroorning 49

initiations by fawns during week 2 and beyond were as high as values for adults. Such a pattern is unique among ungulates, in which reciprocal allogrooming, if it does occur, develops slowly over many months and/or attenuates with age (WALTHER 1984).

The only difference seen in the allogrooming between fawns and adults was in solicitation and lip-licking behavior. Although nearly half of all adult grooming initiations involved neck presentation of the erstwhile partner, fawns rarely used such solicitation to invite grooming. It may be that the solicitation behavior seen in the adults was acquired as a result of repeated grooming interactions with the same herd-mates, since in observations of wild impala populations in Kenya, solicitation behavior was virtually never seen in adults that allogroomed (HART & HART 1992). Thus, the general absence of solicitation by fawns at SDWAP may be more representative of the loosely structured herds of impala in Kenya, and the occurrence of frequent solicitation may be an artifact of captivity. Lip-licking which was exhibited only by young fawns and later disappeared, may be a transitionary stage in the appearance of reciprocal allogrooming.

Although the number of subjects is limited, the results of the study on hand- reared impala fawns indicate that interactions with older experienced groomers is not a prerequisite for early occurrence of impala allogrooming. Analyses of grooming rates of hand-reared fawns revealed no difference from rates of dam- reared fawns in terms of the number of bouts or episodes delivered per h. This is not to say that learning plays no role in the developmental appearance of reciprocal allogrooming. Although social interactions with adults could be important in the fine-tuning of reciprocal allogrooming as fawns mature into adults, this was not revealed by the present study. Self-grooming could possibly have stimulated the expression of allogrooming and/or interactions with other impala fawns may have provided sufficient stimulation to allow the expression of allogrooming. Testing the latter possibility would require raising impala fawns in isolation from conspecifics.

The hand-reared fawns were in a pen with a variety of fawns of other species, such that impala conspecifics represented only 22-33 O/O of potential allogrooming partners available at any given time. However, 67 YO of impala grooming initiations were to conspecifics, showing that impala fawns preferentially initiated grooming to other impala. Nonetheless, 33 % of initiations were to heterospecific fawns that reciprocated only 3.4 % of the time. When heterospecific reciprocation occurred there was a mean of 1.4 bouts. The percentage of all allogrooming encounters that were directed to heterospecifics, and the rate per h of initiations to heterospecifics, did not decline significantly between the first half and the last half of the study. Evidently, feedback from grooming initiations is not required for grooming initiations to be continued. This view is supported by the yearling female impala who regularly initiated grooming with Thomson’s and Grant’s gazelle herd-mates despite the fact that the gazelles were never observed returning grooming. After 6 mo with little or no reciprocation, she still groomed heterospecifics. The observations on this female also argue against the idea that impala fawns continued to initiate grooming to heterospecifics based on regular reinforcement of allogrooming from conspecifics. However, the finding that

Ethology, Vol. 90 (1) 4


hand-reared impala delivered usually just one bout per encounter to heterospecifics, as opposed to a mean of 2 bouts per encounter with conspecifics, shows that the impala fawns did not follow with a second bout when the heterospecific did not reciprocate (episodes per bout did not differ).

Grooming has been shown in cattle to be effective in removing ticks (BENNETT 1969; SNOWBALL 1956). In antelope, interspecific comparisons reveal that species with the highest grooming rate have the fewest ticks per m2 body surface (HART et al. 1992). Because ticks take large blood meals which adversely effect growth rates and can deplete reserves that can be important for intraspecific competition in escaping from predators, weaning young or competing for territory (reviewed by HART 1990), the threat of tick infestation should be a strong selection pressure. Impala and other browsing African antelope possess a lateral dental combing apparatus (incisors and canines) aligned so as to effectively remove ticks and other ectoparasites (MCKENZIE 1990). Comparative observations of self grooming in antelope, including impala, suggest that self grooming is likely to be driven by a central nervous system timing mechanism much as has been described for rodents following removal of peripheral input (FENTRESS 1973, 1988). The strong predisposition of neonatal impala fawns to groom a partner suggests a type of central programming.

Several lines of evidence indicate that impala are subject to greater threat of tick infestation than many sympatric antelope that inhabit grassland savanna. Environmental censuses of woodland and riverine vegetation, such as inhabited by impala, have generally found higher tick densities than in open grassland areas (LONDT & WHITEHEAD 1972; NORVAL 1977). Sympatric gazelle species inhabiting open grassland adjacent to impala habitat do not reciprocally allogroom (HART & HART 1988) and impala self-oral groom more than comparably sized antelope (HART et al. 1992). The implication is that the threat of tick infestation in the ancestral environment has been a strong selective pressure favoring the evolution of reciprocal allogrooming of the head and neck which cannot be reached by the animal with its mouth for self grooming.

Acknowledgements We are grateful to the Zoological Society of San Diego for granting us access to the impala of the

San Diego Wild Animal Park. Thanks go to S. BAKER, L. KILMAR, D. LINDBURG, and J. OSTERHUE for permission to conduct various aspects of the study. E. RUBIN and C. MICHELSON helped facilitate the observations at the Infant Isolation Unit, while R. MASSENA and R. RIECHES and their staff cooperated with the herd observations. We are particularly grateful for the professional work of the volunteer observers: S. COW, J. DEMARCO, B. DISHER, M. EGGERT, K. FRALEY, and J. RIZZA. MSM was funded in 1988 by a Distinguished Scholar Award from the University of California, Davis, and in both 1988 and 1989 by awards from Sigma Xi, the Scientific Research Society, and Graduate Research Awards from the University of California, Davis.

Literature Cited ALTMANN, J. 1974: Observational study of behavior: sampling methods. Behaviour 49, 227-267. AXELROD, R. & DION, D. 1988: The further evolution of cooperation. Science 242, 1385-1390. _ - & HAMILTON, W. D. 1981: The evolution of cooperation. Science 211, 1390-1396.


BENNEIT, G. F. 1969: Boophilus microplus (Acarina: Ixodidae): Experimental infestations on cattle

DANIEL, W. W. 1990: Applied Nonparametric Statistics. 2nd ed. PWS-Kent, Boston. FELDMAN, D. S. & GAGNON, J . 1986: StatView 512+: The Professional, Graphic, Statistics Utility.

FENTRESS, J . C. 1973: Development of grooming in mice with amputated forelimbs. Science 179,

_ - 1988: Expressive contexts, fine structure, and central mediation of rodent grooming. Ann.

HART, B. L. 1990: Behavioral adaptations to pathogens and parasites. Neurosci. Biobehav. Rev. 14,

_ - & HART, L. A. 1992: Reciprocal grooming in impala (Aepyceros melumpus). Anim. Behav., in press.

_- -- , MOORING, M. S. & OLUBAYO, R. 1992: Biological basis of grooming in antelope. Anim. Behav., in press.

HART, L. A. & HART, B. L. 1988: Auto and social grooming behavior in impala. Ann. N.Y. Acad. Sci. 525, 3 9 9 4 0 2 .

JARMAN, M. V. 1979: Impala Social Behaviour: Territory, Hierarchy, Mating and the Use of Space. Adv. in Ethology 21, Suppl. to Ethology.

LENT, P. C. 1974: Mother-infant relationships in ungulates. In: The Behaviour of Ungulates and its Relation to Management. (GEIST, V. & WALTHER, F., eds.) IUCN Publ. New Ser. Morges 24, 14-55.

LEUTHOLD, W. 1977: African Ungulates: A Comparative Review of their Ethology and Behavioral Ecology. Springer-Verlag, Berlin.

LONDT, J. G. H. & WHITEHEAD, G. B. 1972: Ecological studies of larval ticks in South Africa (Acarina: Ixodidae). Parasitology 65, 4 6 9 4 9 0 .

MCKENZIE, A. A. 1990: The ruminant dental grooming apparatus. 2001. J . Linn. SOC. 99, 117-128. MOORING, M. S. 1989: Onrogeny of allogrooming in male deer (Odocozleus hemionus). J. Mammal.

-. - & RUBIN, E. S. 1991: Nursing behavior development of Kenyan impala at the San Diego Wild

MURRAY, M. G. 1982: Home range, dispersal and the clan system of impala. Afr. J. Ecol. 20,

NORVAL, R. A. I. 1977: Ecology of the tick Amblyomma hebraewm Koch in the eastern Cape

SAS INSTITUTE 1989: SAS/STAT User's Guide, Version 6, Vol. 2 .4th ed. SAS Institute Inc., Cary. SCHENKEL, R. 1966: O n sociality and behavior in impala. 2. Saugetierkd. 31, 177-205. SNOWBALL, G. J. 1956: The effect of self-licking by cattle on infestations of cattle tick. Australian J.

WALTHER, F. R. 1984: Communication and Expression in Hoofed Mammals. Indiana Univ. Press,

ZAR, J. H. 1984: Biostatistical Analysis. 2nd ed. Prentice-Hall, Englewood Cliffs. ZIMMERMAN, R. H . & GARRIS, G. I. 1985: Sampling efficiency of three dragging techniques for the

collection of non parasitic Boophilus microplus (Acari: Ixodidae) larvae in Puerto Rico. j. Econ. Entomol. 781, 6 2 7 4 3 1 .

restrained from grooming. Exp. Parasitol. 26, 323-328.

Brain Power, Calabasas.

704-705.

N.Y. Acad. Sci. 525, 18-26.

273-294.

70, 434-437.

Animal Park. Zoo. Biol. 9, 329-339.

253-269.

province of South Africa. I. Distribution and seasonal acitivity. J. Parasitol. 63, 734-739.

Agric. Res. 7, 227-232.

Bloomington.

Received: November 2, 1990 Accepted: September 23, 1991 (G. Barlow)

reciprocal allogrooming in dam-reared and hand-reared impala fawns

Documents