Preorbital Gland Opening in Red Deer (Cervus Elaphus) Calves: Signal of Hunger?1

Preorbital gland opening in red deer (Cervus elaphus) calves: Signal of hunger?1

L. Bartosˇ2,J.Vı´chova´ , and J. Lancingerova´

Ethology Group, Research Institute of Animal Production, CZ-104 01 Prague 10-Uhrˇ´ıneˇves, Czech Republic

ABSTRACT: The opening of the preorbital gland in preorbital gland was associated with feeding and red deer (Cervus elaphus) calves has been hypothesized achieving satiety. In most cases (77%), calves opened to be a signal to the mother that her calf is hungry. their preorbital glands before their meal. Eighty-ﬁve Closing of the gland should indicate that the calf has percent of hungry calves still had preorbital glands open received a sufﬁcient amount of milk, and thus the after the meal, whereas preorbital gland opening was < mother should stop the suckling. We tested the hypoth- decreased (P 0.03) up to 46.6% in satiated calves. esis that the calf signals hunger when its preorbital Despite this, we found a large range in individual re- gland is open and signals satiation when the gland is sponses in preorbital gland opening before and after the meal, and also according to whether the calf reached closed. To test this, the behavior of eight bottle-reared satiety. The heavier calves in this study tended to have calves was monitored for 6 mo. For each meal during their preorbital glands open more frequently than the this time period, opening of the preorbital gland was lighter ones, which may indicate the need for an in- recorded before calves were offered the meal and after creased amount of milk intake. There was no clear trend they ate. Satiation of the calf was estimated (not beg- in the frequency of preorbital gland opening with age. ging for more food after the meal = satiated; begging = In conclusion, preorbital opening is likely to be a signal not satiated). The data set contained 3,116 records of of calf satiety, as suggested earlier. Nevertheless, preor- paired measurements of preorbital gland opening (be- bital glands were not always open when the calves were fore and after the meal). Calves were satiated after hungry and did not vanish during the feeding in all feeding for 90.15% of the meals. The opening of the cases.

Key Words: Cervus elaphus, Hunger, Preorbital Gland Opening, Red Deer Calves, Satiety, Signal

Introduction gested for some cervids (Schaller, 1967; Mu¨ ller- Schwarze, 1975; Macnamara and Eldridge, 1987). In All cervids, with the exception of the moschus, pos- addition to its possible role in marking (Clutton-Brock sess preorbital organs, which consist of a glandular et al., 1982), there seem to be other functions for the region in a pouch adjacent to the nasal corner of the preorbital glands in red deer (Cervus elaphus). It has eye (Schaffer, 1940; Mu¨ ller-Using and Schloeth, 1967). previously been shown that dominant stags open There is no clear evidence of a role across species for this their preorbital glands when they are ﬁghting (Bartos˘, preorbital gland (Gray et al., 1989). The most frequently 1983). Hatlapa (1977) showed that preorbital secretion reported role of preorbital secretions is that they are in red deer has a primary function in establishing used for territorial marking (Gray et al., 1989; Gosling, mother-offspring bonds. Further, it is believed that the 1990). This is typical for many ungulates (Mo et al., opening of the preorbital gland in red deer calves is a 1995; Burger et al., 1999a,b) and has also been sug- signal to the mother that her calf is hungry. Following this, closing of the gland should signal that the calf has received a sufﬁcient amount of milk, causing the mother to stop suckling (Wo¨lfel, 1976, 1983). The interest in 1The authors thank T. DeVries for his kind help with preparing the possible role of the preorbital in signaling mother- the English manuscript. The project was supported by the Grant calf communication has recently been renewed for prac- Agency of the Czech Republic (GA 524/02/0711) and a grant from the tical reasons associated with the worldwide increase in Ministry of Agriculture of the Czech Republic (MZe 0002701402). deer farming (Audenaerde, 2002). Therefore, in this 2Correspondence: POB 10 (phone: +420-267-009-657; fax: +420- 267-710-779; e-mail: [email protected]). study, the hypothesis was tested that the calf signals Received August 31, 2004. hunger when its preorbital gland is open and signals Accepted October 13, 2004. satiation when the gland is closed.

124 Preorbital gland opening in red deer calves 125 feeding. The calves were kept in an enclosure (500 m2) that provided a shelter and ad libitum access to pasture. The observations of preorbital gland opening started on May 30 (the day when the ﬁrst ﬁve calves were placed into the enclosure) and lasted until December 14, when the calves were “weaned” and placed into a herd of calves that were naturally weaned by their mothers. Preorbital opening was recorded during two periods: just before the meal was offered and just after the meal was eaten. The preorbital gland was considered “open” when the back of the gland (a white spot) was visible, and it was considered “closed” when no slit could be distinguished. After the meal, the satiation of the calf was estimated (not begging for more food by vocalizing toward the foster mother, butting or licking her, or attempting to get the bottle after the meal = satiated; begging = not satiated).

Statistical Analyses Figure 1. Proportion of preorbital gland opening of the red deer calves before and after a meal. A greater All data were analyzed using SAS software (Version proportion (P = 0.001) of preorbital gland opening oc- 9.1; SAS Inst., Inc., Cary, NC). The differences in preor- curred before a meal than after a meal. Significance was bital opening before and after a meal were tested using computed by MacNemar’s matched pair data test (n = McNemar’s χ2 test for matched-pair data (McNemar, 3,116). 1947), calculated according to Stokes et al. (2000). To assess the effect of satiety on preorbital opening, we applied an analysis of categorical repeated measure- Materials and Methods ments based on the generalized estimating equation approach (Liang and Zeger, 1986) using the GENMOD procedure. The GENMOD procedure was designed to Animals and Handling model the probability of opening the preorbital gland. This study received approval for animal use and care To account for the repeated measures on the same indi- from the institutional committee. viduals across the period of observation, the analysis Preorbital gland opening was observed in eight arti- was performed with the individual calf in the RE- ficially reared red deer calves (one male, seven females). PEATED statement. The explanatory variables were The calves were collected from a commercial deer farm the classes “satiety of the calf” (yes or no; not begging (Vimperk, Czech Republic) within the first day of their for more food after the meal, “satiety of the calf” = yes; life. They were then weighed, marked by ear tags and begging, ’satiety of the calf’ = no) and the continuous numbered collars, and transported to another facility variable “age of the calf” (in days). Both explanatory where one person reared and fed them over a period of variables and interaction term were tested, but are not 6 mo. The calves were imprinted to a foster mother so reported unless they were statistically significant (P < that they became completely tame. They approached 0.05). Individual differences in preorbital gland opening her, vocalized to her, and when being fed, they butted among the calves were tested by a modified model of the her in the same way they would butt the udder of a logistic regression (PROC GENMOD), with the effect of nursing hind. A bottle containing milk (1,000 mL of 2% the calf as an explanatory variable. The other explana- cow milk, 500 mL of 33% cream, six egg yolks, and tory variables were the same classes as in previous vitamin C [100 mg of ascorbic acid]; Celaskon, SPOFA, model: “satiety of the calf” and “age of the calf.” A record Prague, Czech Republic), prepared after Bartosˇ et al. of preorbital opening was omitted when any of the vari- (1991), was offered to the calves four times each day ables required for the analysis was missing. during the first 4 wk, and decreased with the age of The relationship between preorbital gland opening the calves to one feeding a day at the end of the bottle and the birth weight of the calves was tested using the rearing, which followed the natural process of weaning Cochran-Armitage trend test (Agresti, 1990; Stokes et at the age of 6 mo. Each calf consumed from 80 to al., 2000). 1,500 mL of milk per feeding according to its age and To see whether there was any individual pattern in requirements. The quantity of milk was prepared so preorbital gland opening after the meal, a Spearman that each calf could reach satiety. If it did not, the correlation was applied between the percentage of pre- quantity was either supplemented by milk not con- orbital glands open after the meal when the calf was sumed by another calf and/or was increased for the next satiated and when it was not satiated. 126 Bartosˇ et al.

Figure 2. Proportion of opened or closed preorbital glands before a meal, after a meal when the calf was not satiated and after the meal when the calf was satiated for individual calves. Each bar represents results for one calf and n = 521, 332, 234, 445, 368, 403, 400, and 413 observations per calf (from left to right, respectively).

2 Results calf” (χ (7) = 29.84; P = 0.001), the “satiety of the calf” 2 (χ (1) = 68.18; P < 0.001) and the third term interaction During the 196 d of observations, 3,967 meals were of the “identity of the calf” × “satiety of the calf” × “the 2 offered to the calves. After the removal of incomplete age of the calf” (χ (15) = 307.46; P < 0.001). Figure 3 records (i.e., records for which the feeding person did (top) shows the individual differences in the predicted not recognize whether the preorbital glands were open value of preorbital gland opening before the meal plot- or closed, whether the calf was satiated after the meal, ted against the age of the calf. The predicted probability etc.), the data set had 3,116 records of paired measure- of preorbital gland opening was further plotted against ments of preorbital opening (before and after the meal) the age when the calf was not satiated (Figure 3, mid- for which all required variables were observed. The dle) and when it was satiated after its meal (Figure calves were satiated after feeding for 90.15% of the 3, bottom). meals. A significant trend was found in the frequency of The McNemar test revealed significant differences preorbital opening according to the birth weight of the between the percentages of preorbital glands opened calves (Cochran-Armitage trend test: Z = −24.60; P < 2 before vs. after the meal (χ (1) = 519.38; P < 0.001). 0.001). The probability of the preorbital glands being The gland was mostly open before the meal (77.02%); open increased (from 19.93 to 83.45%) with increasing however, after the meal, the percentage of preorbital birth weight of the calves (Figure 4). There was a high opening decreased to 50.45% (Figure 1). Spearman correlation between the percentage of preor- To assess the effect of satiety on preorbital gland bital glands open after the meal when the calf was opening of the calves after the meal, the data set was satiated and when it was not satiated (rs = 0.83; n = 8; analysed using a logistic regression model. The resul- P = 0.01). tant model only contained the effect of the “satiety of 2 the calf” (χ (1) = 5.14; P < 0.03). The odds ratio of 6.66 Discussion estimated for the “satiety of the calf” indicates that the odds of the satiated calves of having their preorbital The opening of the preorbital gland in bottle-fed red glands open was 6.66 times the odds of the unsatiated deer calves was found to be associated with feeding and calves (confidence limits 3.40 to 13.05). the attainment of satiety. In most cases, calves opened We found large individual variability among the their preorbital gland before their meal. Those calves calves in the frequency of preorbital opening (Figure that were hungry after their meal had their preorbital 2). The final logistic regression model contained the gland still open after the meal, whereas preorbital statistically significant effects of the “identity of the gland opening was found to be decreased in satiated Preorbital gland opening in red deer calves 127

Figure 3. Predicted probability (computed by logistic regression) of preorbital gland opening plotted according to age of the calf before the meal (top panel), after the meal when the calf was not satiated (middle panel), and after the meal when the calf was satiated (bottom panel). Individual calves were characterized by the same style of the curve in the top, middle, and bottom panels. 128 Bartosˇ et al. mother and/or an expression of the emotional stage of the calf dealing with its individual strategy to get milk. The preorbital gland signal of calf satiety was most likely also affected by some other factor or factors. One good candidate for this could be excitement. It has been reported that the preorbital glands of fighting males are open (Bartos˘, 1983) and are also commonly open in rutting stags. Despite this, it is not yet clear whether the open preorbital gland serves as a visual or an olfac- tory signal. In conclusion, preorbital gland opening was not always present when the calves were hungry and did not always close during feeding. This suggests that for practical purposes, preorbital gland opening should be accepted with caution as a reliable indicator of calf satiety as recommended by Wo¨lfel (1976, 1983). Figure 4. Proportion of preorbital glands open (black bars) and closed (hatched bars) in calves according to Literature Cited their birth weight and significance of the Cochran-Armi- Agresti, A. 1990. Categorical Data Analysis. John Wiley & Sons, Inc., tage trend test. Each bar represents results for one calf New York, NY. and n = 445, 234, 521, 332, 403, 368, 400, and 413 observa- Audenaerde, P. M. F. 2002. European view on the deer farming indus- tions per calf (from left to right, respectively). try of the world. Pages 73–76 in Proc. 2002 NADeFA Annu. Conf. and the World Deer Farming Cong. III. J. B. Wood, ed. NADeFA, Austin, TX. Bartosˇ, L. 1983. Some observations on the relationships between preorbital gland opening and social interactions in red deer. calves. This was highly evident during the critical pe- Aggress. Behav. 9:59–67. riod of the age of the calves up to d 30, when the calves Bartosˇ, L., J. Sˇ iler, S. Losos, H. Herrmann, Z. Podeˇbradsky´,M. were fully dependent on milk (Bartos˘ et al., 2001). Chlupsa, V. Barnet, M. Havlıćˇek, P. Rezaćˇ , H. Faltysova´,P. ˇ Therefore, we can conclude that preorbital opening is Ryba´rˇ, J. Lancingerova´,Z.Steˇtkova´, L. Ambrozek, V. Glasna´k, J. Mojto, M. Landa, I. Hanouskova´, and M. Novotna´. 1991. Zku- likely to be a signal of calf satiety as recommended by sˇenosti z prvnıćh farem jelenu˚ a danˇ ku˚ vCˇ echaćh a na Moraveˇ. Wo¨lfel (1976, 1983). Nevertheless, the preorbital gland U´ stav veterina´rnı´ osveˇty, Pardubice, Czech Republic. was not always open when the calves were hungry or Bartosˇ, L., D. Vanˇ kova´,J.Hyańek, and J. Sˇ iler. 2001. Impact of willing to get milk, and it did not vanish in all cases allosuckling on growth of farmed red deer calves (Cervus elaphus). Anim. Sci. 72:493–500. during the feeding. Burger, B. V., J. Greyling, and H. S. C. Spies 1999a. Mammalian A large range was found in individual response in exocrine secretions. XIV: Constituents of preorbital secretion of preorbital gland opening both before and after the meal, steenbok, Raphicerus campestris. J. Chem. Ecol. 25:2099–2108. and also according to whether the calf reached satiety Burger, B. V., A. E. Nell, H. S. C. Spies, M. Leroux, and R. C. Bigalke by eating. A high correlation between the percentage 1999b. Mammalian exocrine secretions. XIII: Constituents of preorbital secretions of Bontebok, Damaliscus dorcas dorcas, of preorbital glands open after the meal when the calf and Blesbok, D-d. phillipsi. J. Chem . Ecol. 25:2085–2097. was satiated and when it was not satiated suggests a Clutton-Brock, T. H., F. E. Guinness, and S. D. Albon 1982. Red high general individual predisposition for opening its deer: Behavior and ecology of two sexes. The Univ. of Chicago/ preorbital glands. A good example is the calf shown by a Edinburgh Univ. Press, Edinburgh, U.K. solid line on Figure 3, who tended to have its preorbital Gosling, L. M. 1990. Scent marking by resource holder: Alternative mechanisms for advertising the costs of competition. Pages 315– glands opened regardless of whether it was before or 328 in Chemical Signals in Vertebrates 5. D. W. Macdonald, D. after the meal and whether it achieved satiety. This Mu¨ ller-Schwarze, and S. E. Natynczuk, ed. Oxford Univ. Press, may reflect some sort of agitation for the dam’s atten- Oxford, U.K. tion, perhaps due to frustration from the social stress Gray, D. R., P. F. Flood, and J. E. Rowell 1989. The structure and function of muskox preorbital glands. Can. J. Zool. 67:1134– of living in a group of calves. Unfortunately, we do not 1142. have data to assess this. The heavier calves in this Hatlapa, H. H. 1977. Zur biologischen Bedeutung des Praörbitalor- study tended to have their preorbital glands open more gans beim Rotwild, Pra¨gung, Individualgeruch, Orientierung . frequently than the lighter calves, which may suggest Berl. Muench. Tieraertzl. Wochenschr. 90:100–104. the need for a larger amount of milk. There was no Liang, K. Y., and S. L. Zeger 1986. Longitudinal data analysis using generalized linear models. Biometrika 73:13–22. clear trend in the frequency of preorbital gland opening Macnamara, M., and W. Eldridge 1987. Behavior and reproduction with age. Even though the period of the full dependency in captive pudu (Pudu puda) and red brocket (Mazama ameri- on milk finished at the age of approximately 30 d cana), a descriptive and comparative analysis. Pages 371–386 (Bartos˘ et al., 2001), the calves in the present study in Biology and Management of the Cervidae. C. W. Wemmer, ed. Smithsonian Inst., Washington, DC. kept opening their preorbital glands for a long time McNemar, Q. 1947. Note on the sampling error of the difference afterwards. This may indicate some role of preorbital between correlated proportions or percentages. Psychometrika glands opening in social contact with the surrogate 12:153–157. Preorbital gland opening in red deer calves 129

Mo, W.-P., B. V. Burger, M. Le Roux, and H. S. C. Spies 1995. Mamma- Schaller, G. B. 1967. The Deer and the Tiger: A Study of Wildlife in lian exocrine secretions IX. Constituents of the preorbital secre- India. Univ. of Chicago Press, IL. tion of oribi, Ourebia ourebi. J. Chem. Ecol. 21:1191–1215. Stokes, M. E., C. S. Davis, and G. G. Koch 2000. Categorical data Mu¨ ller-Schwarze, D. 1975. A note on the use of the antorbital gland analysis using the SAS system. 2nd ed. SAS Inst., Inc., Cary, NC. in marking by Eld’s deer. Appl. Anim. Ethol. 1:301–303. Wo¨lfel, H. 1976. Vorlaüfiger Bericht u¨ ber einige neue Beobachtungen Mu¨ ller-Using, D., and R. Schloeth 1967. Das Verhalten der Hirsche. zur muterlosen Aufzucht des Rothirsches (Cervus elaphus). Z. Pages 1–60 in Handbook of Zoology. Ku¨ kenthal Handb. Zool. Koeln. Zoo 19:16–19. 10:1–60. Wo¨lfel, H. 1983. Zur Jugendentwicklung, Mutter-Kind-Bindung und Schaffer, J. 1940. Die Hautdru¨ senorgane der Saügetiere. Urban und Feindvermeidung beim Rothirsch (Cervus elaphus) II. Z. Jagd- Schwartzenberg, Berlin, Germany. wiss. 29:197–213.