The authors have declared that no competing interests exist.
Conceived and designed the experiments: KB LB. Performed the experiments: KB. Analyzed the data: KB LB TH. Contributed reagents/materials/analysis tools: KB. Wrote the paper: KB LB TH.
Allosuckling is a situation when a female nurses a non-filial offspring. It was described in various ungulate species; however for camels this is the first description of this behaviour. The aim of the study was to assess the occurrence of allosuckling in captive camels (
Allonursing or communal nursing, communal suckling, non-offspring nursing in mammals refers to the situation when a lactating female nurses a young which is not her own
The dromedary camel (
The wild Bactrian camel (
The aim of this study was to provide the first description of allosuckling occurrence in camels and to test possible hypotheses explaining this behaviour. The kin selection hypothesis did not seem to be a major factor in this study, as the females were not related to each another. Based on the findings of Zapata
Observations of camels were carried out in zoos mostly from the visitors' area or from the background yards when needed. The observer did not enter animal enclosure and did not affect the behaviour, husbandry, and management of studied animals. The zoo managers were informed and agreed with the research activities.
From 2005 to 2007, we have studied maternal behaviour of Bactrian camels kept in four zoological gardens in the Czech Republic (Praha, Brno, Ostrava, Zlín–Lešná). Nine females (one of them reproduced two times within the observation period) and ten calves (4 males, 6 females), were included in the study. The size of herds ranged between 5 and 11 individuals; including 2 to 3 calves (
Zoo | Year | Adults (M, F) | Nursing F | Calves (M, F) | Total SB | Non-filial SB | Non-filial SB (%) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Total | 373 | 32 |
|
(M - males; F- females; SB – sucking bout).
One of the calves was already weaned by its mother but occasionally sucked from a non-maternal dam.
Zoo | Year | Name | Date of birth | Sex | Mother | Age of calf (months) | Observed hours | No. of sucking bouts | Allo-sucking extent (%) | Allo-sucker | Mean duration ± SE of filial sucking (sec.) | Mean duration ± SE of allosucking (sec.) |
Brno | 2006 | April | 2.4.2006 | F | Isis | 2–5 | 32 | 75 | 0,00 | NO | 22.48±2.38 | |
Brno | 2006 | Gaja | 18.4.2005 | F | Sulika | 13–16 | 32 | 6 | 0,00 | NO | 50.50±22.41 | |
Brno | 2007 | April | 2.4.2006 | F | Isis | 14–17 | 24 | 3 | 100,00 | YES | 37.33±16.83 | |
Brno | 2007 | Polednice | 10.3.2007 | F | Sulika | 2–5 | 24 | 23 | 0,00 | NO | 33.87±6.58 | |
Zlín-Lešná | 2005 | Marek | 17.5.2005 | M | Jade | 0–3 | 22 | 34 | 0,00 | NO | 55.15±10.41 | |
Zlín-Lešná | 2005 | Aštar | 14.3.2005 | M | Klaudie | 3–6 | 22 | 24 | 8,33 | YES | 75.91±15.66 | 10.00±0.00 |
Ostrava | 2006 | 2sameček | 31.3.2006 | M | Vendula | 2–6 | 42,5 | 30 | 3,33 | YES | 62.90±9.03 | 5.00±0.00 |
Ostrava | 2006 | 1samička | 23.3.2006 | F | Čora | 2–6 | 42,5 | 55 | 27,27 | YES | 41.23±5.36 | 36.80±7.39 |
Ostrava | 2007 | 2sameček | 31.3.2006 | M | Vendula | 14–16 | 16 | 8 | 50,0 | YES | 41.25±15.20 | 58.50±24.15 |
Ostrava | 2007 | 1samička | 23.3.2006 | F | Čora | 14–16 | 16 | 11 | 45,45 | YES | 32.50±7.06 | 43.80±14.45 |
Ostrava | 2007 | Kobi | 2.3.2007 | F | Fatima | 2–4 | 16 | 17 | 0,00 | NO | 46.06±10.05 | |
Praha | 2006 | Vanda | 18.1.2006 | F | Lee | 6–8 | 29 | 43 | 4,65 | YES | 37.15±4.96 | 17.50±5.50 |
Praha | 2006 | Víťa | 21.7.2006 | M | Rona | 0–3 | 29 | 44 | 0,00 | NO | 54.16±6.63 |
Zoo | Female name | Birthdate | Arrival to present zoo | Parity till 2005 | Number of calves till 2005 | Parity till 2006 | Number of calves till 2006 | Parity till 2007 | Number of calves till 2007 | Allonursing extent (%) | Allonurser |
Brno | Isis | 2.6.1998 | 24.6.1999 | 2 | 1 | 3 | 2 | 3 | 2 | 0.00 | NO |
Brno | Sulika | 6.3.1992 | 6.3.1992 | 7 | 5 | 7 | 5 | 8 | 6 | 9.38 | YES |
Zlín-Lešná | Jade | 24.2.1997 | 3.9.1998 | 4 | 4 | 4 | 4 | 5 | 4 | 5.56 | YES |
Zlín-Lešná | Klaudie | 3.5.1999 | 3.5.1999 | 2 | 1 | 2 | 1 | 3 | 2 | 0.00 | NO |
Ostrava | Vendula | 03.04.2000 | 30.5.2001 | 1 | 0 | 2 | 1 | 2 | 1 | 35.29 | YES |
Ostrava | Čora | 02.03.2002 | 30.5.2003 | 0 | 0 | 1 | 1 | 1 | 1 | 2.13 | YES |
Ostrava | Fatima | 17.05.1990 | 27.6.2003 | 2 | 0 | 2 | 0 | 3 | 1 | 26.09 | YES |
Praha | Lee | 02.06.1998 | 24.6.1999 | 2 | 1 | 3 | 2 | 4 | 3 | 0.00 | NO |
Praha | Rona | 15.03.1995 | 26.4.1996 | 4 | 1 | 5 | 2 | 5 | 2 | 4.35 | YES |
Camels in all facilities were fed once or twice a day by hay and grasses
We recorded all occurrences of suckling by
For each sucking bout we recorded the identity of the animals, duration of sucking bout, position of sucking calf, which animal terminated the sucking bout (mother, calf, or other). The position of the sucking calf was classified into two classes - antiparallel, when the hind part of the calf was directed toward a cow's head, and lateral, when the calf stands at least in the right angle to the cow's body axis. As the gap between the start of sucking and milk let-down is not documented in camels, we consider all bouts longer than 5 seconds as successful as in other studied species e.g.
The data were analysed using Statistical Analysis Systems (SAS) version 9.2. Frequency counts for prediction (i) were analysed by computing chi-square test (PROC FREQ). The output contained cell or cells counts less than 5, hence Pearson exact chi-square was used. For other data we used Generalised Linear Mixed Model (GLMM) for analysing numeric variables (PROC MIXED) or categorical variables (PROC GLIMMIX for binary distribution). To account for repeated measures, all mixed model analyses but one were performed using individual camel ‘calf’ nested within the ‘herd’ as a random effect. In unbalanced designs with more than one effect, the arithmetic mean for a group may not accurately reflect response for that group, because it does not take other effects into account. Therefore, we used least-squares-means (LSMEANs) instead. LSMEANs are, in effect, within-group means appropriately adjusted for other effects in the model. LSMEANs were computed for each class and differences between classes were tested by t-test. For multiple comparisons we used the post hoc Tukey-Kramer adjustment.
We combined predictions (ii) and (iv) into one GLMM for binary distribution modelling the probability for a calf to suck from non-maternal dam. Fixed effects were ‘age of the calf’ (a continuous predictor that ranged from 1 to 17 months), ‘number of calves’ taking part in the sucking bout (a categorical factor with levels 1 to 3), ‘nursing females’ (a categorical factor with levels 2 and 3 females per herd),‘sex of the calf’ (a categorical factor with male and female levels), and ‘sucking order’ (a categorical factor with levels the 1st, the 2nd, and the 3rd calf coming to suck). None of the non-filial calves sucked in antiparallel position, therefore the effect ‘Position’ (antiparallel or lateral) could not be applied. For prediction (iii), we applied a GLMM for binary distribution modelling the probability for a dam to terminate the sucking bout. Fixed effects were ‘relatedness’ (filial sucking and non-filial sucking), ‘position’, ‘age of the calf’, ‘sex of the calf’, ‘nursing females’, and ‘birth order’ (the birth order of the calf within the season and herd). Primarily we were interested in testing the effect of the ‘relatedness’ alone and/or in an interaction with ‘position’. Given that none of the non-filial calves sucked in the antiparallel position, the effect of ‘position’ had to be omitted. We also examined various combinations of the other fixed effects (i.e., ‘relation’, ‘age of the calf’, ‘sex of the calf’, ‘nursing females’, and ‘birth order’) on the termination of sucking by the dam. For prediction (v) we applied GLMM with duration of the sucking bout as a dependent variable. Fixed effects were ‘age of the calf’, ‘age of the dam’ (4 to 17 years), ‘number of calves’, ‘relatedness’ in interaction with ‘position’ and in interaction with ‘sex of the calf’.
Over the three years of study (2005–2007; 164 hours within 26 days of observation in total) we have recorded 373 sucking bouts (
Filial calves sucked from their mothers mostly standing in the antiparallel position (62.17% of cases), while non-filial calves suckled exclusively in the lateral position (n = 32, difference Pearson exact chi-square test p = 3.04 * 10−13,
Four non-filial calves were involved in a sucking bout without the presence of filial calves five times (15.6% of cases), standing in a lateral position (
Termination of sucking by the dam was not affected by any of the tested factors either when they entered the model alone or in any combination with other factors. Non-filial calves never sucked in anti-parallel position, so we could not test the effect of position to termination. Of the non-filial calves which sucked without a presence of filial calf, sucking was terminated by the calf three times, once by the dam and in one case we did not see who terminated the bout.
The GLMM model revealed that the probability for a calf to suck from non-maternal dam was affected by ‘age’ of the calf (F(1,358) = 3.96, p = 0.047,
The mean (± SE) sucking duration was 42.93±2.22 s (range 5–270), the mean duration of filial sucking bout 43.50±2.37 s (range 5–270) and the non-filial sucking bout 36.78±5.47 s (range 5–121).
The GLMM model showed that duration of sucking bouts was dependent on the ‘number of calves’ taking part in the sucking bout (F(1,191) = 17.19, p<0.0001), so the sucking bouts involving more than one calf were longer than those involving just one calf, either filial or non-filial (
Comparison between allonursing and non-allonursing dams is shown in
In this study we brought the first description of allosuckling occurrence in camels. The results have shown that allosuckling occurred in 5 out of 10 calves from 4 camel herds containing more than one calf in different zoos and different seasons. The allosucking calves were in all cases the older ones in the herd, while the youngest calf from the herd never allosucked. The only herd in the study without allosucking occurrence was the Brno Zoo in 2006, where two female calves from different mothers were kept together. Although the data were not included in the study, one author (Karolína Brandlová) observed the allosuckling occurrence there out of the range of the recording time (Gaja allosucked from Isis). These data further imply that allosuckling is common in the captive camels, comparable to captive guanaco
Up to three calves (always one filial and one or two non-filials, there were no more calves in the herd than three) were involved together in a sucking bout. The herd with the highest incidence of allonursing (25%) was the only herd with 3 calves providing the largest number of allonursing possibilities. The earliest allosucking was reported in 50-days-old calf. The youngest calves in the herd were never seen allosucking, despite of the fact that they had the possibility to do it just after joining the herd where other nursing female was present. In other ungulates, except in zebra
Regarding the behaviour of calves, our results widely correspond with the milk-theft hypothesis. We confirmed that (i) allosucking calves sucked only in the lateral (other than antiparallel) position. That may have helped the calf to remain undetected by the nursing female or decreased the probability of being threatened by her. Higher incidence of allosuckling in lateral position was confirmed also by Zapata
As predicted (ii), in all cases when more calves were sucking together, the non-filial calves joined filial calf during sucking non-maternal dam and the probability of allosuckling was higher when there were more calves involved in a sucking bout as reported by Ekvall
We failed to find any support for the prediction (iii). We did not record any case of non-filial calf sucking in anti-parallel position, so we could not assess any influence of sucking position on the termination by females. This could mean that calves which tried to allosuck close to the females head were not successful. Females might have refused to nurse them and calves then learned how to approach the non-maternal dam safely and successfully as reported by Zapata
In agreement with our prediction (iv), the incidence of allosucking increased with age of the allosucking calf as skills of the calf to outwit the non-maternal dam increased or as increased the motivation of a calf due to the weaning process of weaning. At least one of the allosucking calves was already weaned. It corresponds with the findings of Ekvall
The suckling bouts generally lasted longer in filial calves in antiparallel position than in non-filial ones in lateral position as we expected (v). Although the sucking duration itself should not be used as a predictor of milk intake, it can reveal the level of maternal investment
On five occasions a non-filial calf was allosucking with no other calf present (ii). This could be simply a mistake from the dam, considering the fact that mentioned allosucking bouts were considerably shorter than those including also the filial calf. On the other hand, however, we cannot reject entirely the possibility that in some cases the dams tolerated certain individuals in need as an altruistic act as was reported for red deer
Termination of sucking bouts by the females did not differ during sucking events involving filial and non-filial calves. We could not test termination of a sucking bout involving a non-filial calf in an antiparallel position (iii) in comparison with a filial calf in the same position, because none of the non-filial calves has ever been seen in the antiparallel position. Taking into account that non-filial calves were sucking more often in the presence of the filial calf and that the non-filial calf was located more distant to the head of the dam, one could presume that the female would terminate equally sucking of filial and non-filial calves when trying to terminate the non-filial sucking. This may explain generally low level of terminations of non-filial sucking bouts and the tolerance of females.
The increasing incidence of allosucking in older calves (iv) may also imply higher tolerance of nursing females to calves that are more familiar to them as they had lived longer in the same herd than the newborn calves.
The fact that some dams allonursed while others did not, and the fact that at least some of the calves sucked regularly and very successfully suggests a possible strategy of compensation of nutritional requirements by the young as seen in red deer and cattle
Age differences among calves in herds were larger in the zoos in this study (the first calf born in January while the last in July, see
Our results correspond with those of Zapata
The results of the study support the hypothesis of ‘milk theft’, being mostly performed by calves behaving as opportunistic parasites. Nevertheless, tolerance of the camel females to non-filial calves may also suggest that at least in part allosuckling in camels might be adaptive trait, despite the fact it is mostly performed by calves which have the occasion to get surplus milk from a non-maternal female as opportunistic parasites.
The first and most important acknowledgement is for all zoological gardens, Zoo Praha, Zoo Ostrava, Zoo Zlín-Lešná, and Zoo Brno, with their keepers and curators which provide us the best working conditions for camel observations. We are deeply grateful to Jaroslav Šimek, Jan Pluháček, Luis Ebensperger, and two anonymous reviewers for their useful comments and suggestions.