- Email: [email protected]
Nest attendance during egg laying in pheasants
ANIMAL BEHAVIOUR, 1999, 58, 159–164 Article No. anbe.1999.1107, available online at http://www.idealibrary.com on
Nest attendance during egg laying in pheasants IRENE PERSSON & GO } RGEN GO } RANSSON
Department of Ecology, Animal Ecology, University of Lund (Received 28 September 1998; initial acceptance 6 November 1998; final acceptance 26 February 1999; MS. number: 6005R)
As precocial bird species hatch synchronously, incubation during the egg-laying stage should be disadvantageous because it makes the embryos develop asynchronously. We established the patterns of nest attendance during egg laying and the start of incubation in ring-necked pheasants, Phasianus colchicus, and tested three hypotheses regarding the advantage of early incubation. To determine nest attendance, we measured egg temperatures in real pheasant nests. Females spent more time on the nest as laying progressed, with an average of 6.4 h at a clutch size of 10. At the start of incubation, nest attendance increased to over 20 h/day. On the day before full incubation, time spent on the nest was positively correlated with the female’s condition and negatively with the number of breeding attempts she had already made that season. The hypothesis that an early start of incubation improves egg viability was rejected, as the predicted relationship between the number of eggs laid after the start of incubation and the number laid before the start of incubation was not significant. We also rejected the possibility that early incubation reduces the risk of nest parasitism, as it was negatively related to the number of females radiotracked around the nest. Our data supported the hypothesis that early incubation reduces the risk of nest predation by shortening the period of exposure, as the number of eggs laid after incubation started was positively related to the number of breeding attempts made by the female, and thus to the perceived predation risk, but was negatively related to the time of season.
(1) Effects of time unincubated on egg viability. Several studies have shown that the hatchability of duck eggs is reduced when incubation starts late (Mather & Laughlin 1977; Arnold et al. 1987; Arnold 1993). Furthermore, preincubation heating of eggs of domestic fowl is generally beneficial when they are stored for more than 7 days (for a review see Mayes & Takeballi 1984). From this hypothesis, we predict that more eggs will be laid after incubation starts in larger clutches. (2) Risk of intra- or interspecific nest parasitism. In waterfowl, females often lay eggs in other females’ nests during the latter’s egg-laying period (Rowher & Freeman 1989 and references therein), thereby leaving the task of incubation to the owner of the nest. An early start of incubation would reduce the risk of nest parasitism. (3) Risk of nest predation. Because of the longer egglaying period, a large clutch is exposed to nest predators for longer than smaller ones are (Perrins 1977). If incubation starts before the clutch is completed, this period of exposure will be reduced. We predict that females who have lost one or more nests to predators would then start incubation earlier. The majority of studies on egg-laying behaviour in precocial species have been done on waterfowl or on gallinaceous birds in captivity. In this paper we study nest
Traditionally it has been thought that precocial bird species do not start incubating until the clutch is completed, as it is important for the eggs to hatch synchronously (Kear 1970; del Hoyo et al. 1992). However, several studies have shown that waterfowl often begin to incubate during egg laying (see for example Caldwell & Cornwell 1975; Cooper 1978; Afton 1979, 1980; Cargill & Cooke 1981; Kennamer et al. 1990; Wilson & Verbeek 1995). Although the eggs in a clutch thereby have different incubation times, they are still able to hatch synchronously. This is possible because the embryos communicate with each other at the end of the incubation period, and adjust their time of hatching (Vince 1964, 1968, 1972; Lauch et al. 1988; Holmberg 1991; I. Persson & G. Andersson, unpublished data). Why do precocial species start to incubate during the laying period even though they do not benefit from developmental asynchrony? Three factors that may make an early start of incubation advantageous have been suggested. Correspondence: I. Persson, Ecology Building, S-223 62 Lund, Sweden (email: [email protected]). G. Go¨ransson is now at the Department of Natural Sciences, University College of Kalmar, P.O. Box 905, S-391 29 Kalmar, Sweden. 0003–3472/99/070159+06 $30.00/0
1999 The Association for the Study of Animal Behaviour
159
1999 The Association for the Study of Animal Behaviour
160
ANIMAL BEHAVIOUR, 58, 1
Figure 1. Egg temperature in an artificial egg placed in a pheasant’s nest at the beginning of the laying period. The arrows indicate the laying of eggs 5–12, and incubation starts after egg number 10.
attendance during laying in a wild population of ringnecked pheasants, Phasianus colchicus, by measuring egg temperatures, a method that avoids disturbing the females. By relating this to clutch size, physical condition, the number of breeding attempts the female has made that season and female density, we try to explain the initiation of full-time incubation by females during the egg-laying period. METHODS We trapped, measured and ringed 62 female pheasants during March and April 1997 in the Revinge area in southern Sweden. We fitted 20-g radiotransmitters (<3% of the body weight) on to the backs of the females; these did not seem to have any adverse effects on the females. The birds were released at the trapping site after a few hours and tracked daily between 1000 and 1400 hours, at the normal time of laying, from the end of April; 43 of the females were radiotracked until the middle of July and all their breeding attempts were followed. Because of nest predation (65% of the nests) or abandonment by the female (25% of the nests), females were often forced to make up to four breeding attempts per season before they produced any chicks. From the tracking data, we discovered ca. 40% of the breeding attempts during the laying period. When a nest was found, we fixed a dummy egg, containing paraffin wax and a thermistor connected to a data logger, into the nest. The wire connecting the egg to the data logger was buried in the soil under the nest, leaving the dummy egg in a natural position in the nest (for details see Persson & Go ¨ ransson 1996). This arrangement did not seem to affect the behaviour of the
females, which did not abandon the nests if the dummy egg was placed in the nest when five or more eggs had been laid. To see whether the extra egg placed in the nest had any effects on the laying sequence, we compared the clutch sizes of the first breeding attempt in nests where the dummy egg was placed in the nest before or after clutch competition; there was no significant difference (XSD=12.82.9 and 12.52.6 eggs, respectively; t test: t20 =0.262, P=0.8). The egg temperature was measured and saved every 16 min, showing the laying and incubation behaviour of the female, as the temperature rises when the female sits on the eggs and vice versa (Fig. 1). The start of incubation was apparent when the female changed from shorter nest visits to almost continuous nest attendance. By counting the eggs in the nests before and after the start of incubation, and comparing this with the temperature graphs, we could establish whether eggs were laid during the incubation period. We captured 28 females on the nest about a week after the start of incubation. We weighed them and measured their tarsus length, then released them. We calculated an index of their physical condition as the residuals of the individual’s weight from the overall regression of weight versus untransformed tarsus length during 1993–1996. Body weight corrected for tarsus length is a good estimate of fat content in other species (e.g. Iverson & Vohs 1982; Johnson et al. 1985). We estimated the risk of nest parasitism as the average number of females per day radiotracked within 100 m of the nest during the egg-laying period. It was calculated both with and without females that had a nest of their own. The presence of females without radiotransmitters
PERSSON & GO } RANSSON: EGG-LAYING BEHAVIOUR
10
8 Time spent on the nest (h)
around the nests studied was probably the same in the whole area, as the radiotagged females were assumed to represent a random sample of the population. In 19 breeding attempts we were able to put a dummy egg with a thermistor into the nest before the start of incubation. Three of these nests were depredated before laying was known to be completed, and so only 16 breeding attempts, all made by different females, could be used in the analysis of egg laying during incubation. In two breeding attempts the thermistor was placed into the nest after the beginning of incubation but before the clutch was completed. This means that the number of eggs laid after incubation started might have been underestimated. In the remaining 14 breeding attempts studied, we recorded at least 1 and a maximum of 7 days of the laying period. All statistical analyses were performed using Systat (Wilkinson 1992), and all probabilities refer to two-tailed tests. Means are given SD. The multiple regressions were made backwards, stepwise with a P value of 0.05 to remove variables and interaction terms. The models were first tested with all possible interactions, and the nonsignificant interactions were then removed one by one.
6
4
2
0
5
10 Egg number
20
15
Figure 2. Time spent on the nest per day during the laying period in relation to egg number in the laying sequence (N=38 days for 12 females; note that two pairs of data points have the same values).
RESULTS During the laying period, females went to the nest once per day, usually around noon. They arrived at the nest punctually; the time difference between the earliest and the latest egg laying of a breeding attempt was on average 1.330.93 h (N=12; two nests studied for only 1 day were excluded). The females stayed on the nest long enough for the egg temperature to rise to 30–35C, that is, a normal incubation temperature (Fig. 1). In 40 days of laying in 14 breeding attempts only one interruption in the laying sequence was observed, and yet that female spent 3.5 h on the nest that day. There was no occurrence of more than one temperature rise per day, thus nest parasitism did not occur while the nests were under observation, unless the owner of the nest happened to skip egg laying the particular day the nest was parasitized. Females spent longer on the nest per day as the laying period progressed (ANCOVA: F1,25 =89.78, P<0.001), from 3.321.19 h when five eggs were laid to 6.720.74 h at a clutch size of 10 (Fig. 2); females differed in nest attendance (F11,25 =9.23, P<0.001). Considering only the last egg laid before the start of incubation, the time spent on the nest was positively related to the physical condition of the female and negatively related to the number of breeding attempts and the interaction between the two (Table 1). A possible relation between the number of eggs laid and the time spent on the nest on the day before the start of incubation was excluded in the multiple regression. In 14 (88%) of the 16 clutches studied by egg temperature measurements, the female laid at least one egg after the beginning of incubation. The most frequent was one egg (69%, 11 nests), but also two (two nests) and three eggs (one nest) were laid after the start of incubation. Only in two nest attempts did the female start to incubate after the clutch was completed. Predators destroyed all
Table 1. Backwards, stepwise multiple regression between the time spent on the nest on the last day of laying before incubation started and the number of breeding attempts, the number of eggs laid before incubation and the physical condition of the female Variable
Physical condition No. of breeding attempts Physical condition × no. of breeding attempts
r
t
P
1.42 −1.29
2.709 −3.782
0.030 0.007
−1.50
−2.449
0.044
Only significant variables are shown: r2 =0.71, N=11.
but two clutches; however, in these successful clutches all the eggs hatched 25 days after the start of incubation, including the nest with three eggs laid after incubation started. To test the egg viability hypothesis, we analysed the number of eggs laid after incubation started in relation to the number laid before incubation started, and found a nonsignificant tendency for a negative relation between these variables (F1,14 =3.07, P=0.10). In an attempt to determine which factors might explain the number of eggs laid after incubation started, we used the number of breeding attempts and time of season (measured as number of days after 1 April) as independent variables in a multiple regression. This analysis showed a positive relation between the number of eggs and the number of breeding attempts made by the female (T=4.18, N=16, P=0.001; Fig. 3), as well as a negative relation between the number of eggs and time of season (T= 2.56, N=16, P=0.024). The average number of females per day radiotracked within 100 m of a nest, used as a measure of risk of nest parasitism, ranged from 0 to 2.3 and was not related
161
162
ANIMAL BEHAVIOUR, 58, 1
to the number of eggs laid after incubation started (r= 0.20, N=16, P=0.45). When females with a nest of their own were included in the analysis, we found a nonsignificant tendency for a negative relation between the number of eggs and the number of females in the vicinity (r= 0.47, N=16, P=0.07, average values ranging from 0 to 1.3 females per day). The number of females tracked within 100 m of a nest did not correlate with the number of females caught at the same trap as the nest owner (r=0.012, N=16, P=0.96), thus the density of females did not influence the risk of nest parasitism (the overall density of radiotracked females was 10–12 females/km2). DISCUSSION The egg temperature measurements revealed a fairly regular egg-laying pattern in pheasants. The females visited the nests once per day, at the same time every day, and only one interruption in the laying sequence was registered. Unless females make more interruptions during the early part of the laying period, which we did not study, the average number of eggs laid per day would approach 1.0 instead of the 0.71–0.77 previously recorded in gamefarm and wild pheasants (Buss et al. 1951; Robertson 1991; Wise 1995). This high egg production rate may be a way to shorten the time the nest is exposed to nest predators. As the females stayed on the nest for at least 2 h (up to over 8 h) during the egg-laying period, the egg temperature rose well over the physiological zero (20–27C, for a review see Meijerhof 1992), which is the temperature where the embryo starts to develop. Thus, even without the laying of extra eggs during incubation, development would be asynchronous. This has been shown in mallard ducks, Anas platyrhynchos, which incubate long enough for development to occur in the sixth egg in clutches of 10–12 eggs, leading to a difference in developmental age early in incubation of 1–2 days (Caldwell & Cornwell 1975). A similar pattern was found in northern shovelers, Anas clypeata (Afton 1979). Furthermore, in wood ducks, Aix sponza, larger clutches have greater levels of intraclutch developmental asynchrony than smaller ones (Kennamer et al. 1990). Female pheasants spent longer on the nest as laying progressed (Fig. 2). This is in agreement with the diminishing energy requirements for egg production late in the laying sequence (Drobney 1980), and may be important as camouflage for the eggs when the risk of nest predation is high. The nest attendance was also dependent on the physical condition of the female, those in good condition having a higher nest attendance. In addition to this, both the number of breeding attempts by the female and the interaction between physical condition and number of breeding attempts were negatively related to the time spent on the nest. One explanation for this might be that incubation is started earlier in later breeding attempts. It might be energetically hard to have high nest attendance when many eggs are yet to be laid. King (1973) estimated the daily maximum cost of egg production at 21–30% of the daily energy intake without gain or loss of body
weight in galliforms. Maintaining this high energy requirement until the beginning of incubation might be possible only with a lower nest attendance during laying. As in many other species, pheasants lay smaller clutches in later breeding attempts (Robertson 1991; I. Persson, unpublished data). This may be related to the simultaneous costs of incubation and egg laying, which may make it impossible to lay as large a clutch as in the first breeding attempt. Another explanation for the negative relation between the number of breeding attempts and the time spent on the nest could be that an early start of incubation in combination with high nest attendance may make the developmental asynchrony between the eggs too large for the chicks to hatch synchronously. However, there was no significant relationship between the number of eggs laid after incubation started and the time spent on the nest (r=0.091, N=12, P=0.8). The incubation period is defined as the development time of the embryo during regular attention of the parent (Drent 1975). However, in the majority of studies the incubation period is generalized as the time between laying and hatching of the last egg. This was not applicable in the pheasants we studied, as there was a clear increase in the amount of time spent on the nest, from a maximum of 8.5 h during laying to about 21 h during incubation, giving an obvious start of incubation, usually 1 day before the clutch was completed. A similar pattern was found in red Burmese junglefowl, Gallus gallus spadiceus (Meijer & Siemers 1993), and Canada geese, Branta canadensis (Cooper 1978). This abrupt start of incubation contrasts with that of many waterfowl, which often have a gradual start of incubation and spend over 50% of the day on the nest during laying (e.g. Kennamer et al. 1990; Wilson & Verbeek 1995). Our study shows that it is common for pheasants to lay eggs after the start of incubation, thus creating a developmental asynchrony of 1–3 days in addition to the asynchrony developed during the laying period. As hatching asynchrony is disadvantageous for precocial species, some selective pressure must make the early start of incubation advantageous. The egg viability hypothesis suggests that an early start of incubation enhances the hatchability of the eggs, predicting that the female should start to incubate relatively earlier in larger clutches. However, in our study the number of eggs laid after the start of incubation was not positively correlated with the number laid before the start of incubation, so we can reject the egg viability hypothesis. A second hypothesis suggests that the risk of nest parasitism makes an early start of incubation profitable. This is not applicable in our pheasant population, as we only rarely observed parasitic laying. Furthermore, there was a negative correlation between the number of eggs laid after the start of incubation and the number of females radiotracked around the nest, showing that it was not the density of females that made an early start of incubation profitable. Thus, the nest parasitism hypothesis was also rejected. The nest predation hypothesis predicts that females should start to incubate early to reduce the risk of nest predation. In our study, the number of eggs laid after the
PERSSON & GO } RANSSON: EGG-LAYING BEHAVIOUR
4
Number of eggs
3
2
1
0
1
2 3 Number of nest attempts
4
5
Figure 3. Number of eggs laid during incubation in relation to the number of breeding attempts made. When there was more than one data point on a specific set of values, they were plotted above each other in order to be visible. N=16.
start of incubation was positively related to the number of breeding attempts by the female that season (Fig. 3). This supports the nest predation hypothesis. When a female has lost a couple of nests, her perceived nest predation risk has increased, so she may start incubation earlier within the laying sequence. In this way she may try to avoid another nest being depredated, as both time and her energy reserves are running out. It is also possible that the females are influenced by daylength and try to hurry on the breeding as it is getting late in the season. However, there was an overall negative correlation between the number of eggs laid after the start of incubation and time of season, supporting the nest predation hypothesis. For the first breeding attempt, fewer eggs were laid after the start of incubation later in the season perhaps because the vegetation was higher and denser and the need for camouflage thereby less than earlier in the season. Acknowledgments We thank G. Andersson for assiduous and enthusiastic field work, J.-A r . Nilsson for inspiration and valuable comments during analyses and manuscript writing, and A. Hoodless and an anonymous referee for comments on the manuscript. This study was supported by grants to I.P. from the Foundation of C. F. Lundstro ¨ m and the Bokelund Fund, University of Lund. References Afton, A. D. 1979. Incubation temperatures of the northern shoveler. Canadian Journal of Zoology, 57, 1052–1056. Afton, A. D. 1980. Factors affecting incubation rhythms of northern shovelers. Condor, 82, 132–137. Arnold, T. W. 1993. Factors affecting egg viability and incubation time in prairie dabbling ducks. Canadian Journal of Zoology, 71, 1146–1152.
Arnold, T. W., Rohwer, F. C. & Armstrong, T. 1987. Egg viability, nest predation, and the adaptive significance of clutch size in prairie ducks. American Naturalist, 130, 643–653. Buss, I. O., Meyer, R. K. & Kabat, C. 1951. Wisconsin pheasant reproduction studies based on ovulated follicle technique. Journal of Wildlife Management, 15, 32–46. Caldwell, P. J. & Cornwell, G. W. 1975. Incubation behavior and temperatures of the mallard duck. Auk, 92, 706–731. Cargill, S. M. & Cooke, F. 1981. Correlation of laying and hatching sequences in clutches of the lesser snow goose (Anser caerulescens caerulescens). Canadian Journal of Zoology, 59, 1201–1204. Cooper, J. A. 1978. The history and breeding biology of the Canada geese of Marshy Point, Manitoba. Wildlife Monographs, 61, 1–87. Drent, R. 1975. Incubation. In: Avian Biology, Vol. 5 (Ed. by D. S. Farner & J. R. King), pp. 333–420. New York: Academic Press. Drobney, R. D. 1980. Reproductive bioenergetics of wood ducks. Auk, 97, 480–490. Holmberg, K. 1991. Mallard ducks, mate choice and breeding success. Ph.D. thesis, Stockholm University. del Hoyo, J., Elliott, A. & Sargatal, J. (Eds) 1992. Handbook of the Birds of the World. Barcelona: Lynx Editions. Iverson, G. C. & Vohs, P. A., Jr. 1982. Estimating lipid content of sandhill cranes from anatomical measurements. Journal of Wildlife Management, 46, 478–483. Johnson, D. H., Krapu, G. L., Reinecke, K. J. & Jorde, D. G. 1985. An evaluation of condition indices for birds. Journal of Wildlife Management, 49, 569–575. Kear, J. 1970. The adaptive radiation of parental care in waterfowl. In: Social Behaviour in Birds and Mammals (Ed. by J. H. Crook), pp. 357–392. London: Academic Press. Kennamer, R. A., Harvey IV, W. F. & Hepp, G. R. 1990. Embryonic development and nest attentiveness of wood ducks during egg laying. Condor, 92, 587–592. King, J. R. 1973. Energetics of reproduction in birds. In: Breeding Biology of Birds (Ed. by D. S. Farner), pp. 78–107. Washington, D.C.: National Academy of Sciences. Lauch, M., Nichelmann, M. & Wallschlaeger, D. 1988. Prenatal acoustic communication in muscovy ducklings Cairina moschata. Monatshefte fu¨r Veterina¨rmedizin, 43, 865–867. Mather, C. M. & Laughlin, K. F. 1977. Storage of hatching eggs: the effect on early embryonic development. British Poultry Science, 18, 597–604. Mayes, F. J. & Takeballi, M. A. 1984. Storage of the eggs of the fowl (Gallus domesticus) before incubation: a review. World’s Poultry Science Journal, 40, 131–140. Meijer, T. & Siemers, I. 1993. Incubation development and asynchronous hatching in junglefowl. Behaviour, 127, 309–322. Meijerhof, R. 1992. Pre-incubation holding of hatching eggs. World’s Poultry Science Journal, 48, 57–68. Perrins, C. M. 1977. The role of predation in the evolution of clutch size. In: Evolutionary Ecology (Ed. by B. Stonehouse & C. M. Perrins), pp. 181–191. London: Macmillan. Persson, I. & Go ¨ ransson, G. 1996. Nesting biology studied by egg temperature measurements. In: Proceedings of the International Union of Game Biologists XXII Congress (Ed. by N. Botev), pp. 56–61. Sofia: Pensoft. Robertson, P. A. 1991. Estimating the nesting success and productivity of British pheasants Phasianus colchicus from nest-record scemes. Bird Study, 38, 73–79. Rowher, F. C. & Freeman, S. 1989. The distribution of conspecific nest parasitism in birds. Canadian Journal of Zoology, 67, 239–253. Vince, M. A. 1964. Social facilitation of hatching in the bobwhite quail. Animal Behaviour, 12, 531–534. Vince, M. A. 1968. Retardation as a factor in the synchronization of hatching. Animal Behaviour, 16, 332–335.
163
164
ANIMAL BEHAVIOUR, 58, 1
Vince, M. A. 1972. Communication between quail embryos and the synchronisation of hatching. In: Proceedings of the XVth International Ornithological Congress (Ed. by K. H. Voous), pp. 357–362. Leiden: E. J. Brill. Wilkinson, L. 1992. SYSTAT for the Macintosh: Statistics, Version 5.2 Edition. Evanston, Illinois: SYSTAT.
Wilson, S. F. & Verbeek, N. A. M. 1995. Patterns of wood duck nest temperatures during egg-laying and incubation. Condor, 97, 963–969. Wise, D. R. 1995. Egg production of wild and gamefarm pheasants (Phasianus colchicus) strains under captive breeding systems. Gibier Faune Sauvage, 12, 321–330.
Nest attendance during egg laying in pheasants IRENE PERSSON & GO } RGEN GO } RANSSON
Department of Ecology, Animal Ecology, University of Lund (Received 28 September 1998; initial acceptance 6 November 1998; final acceptance 26 February 1999; MS. number: 6005R)
As precocial bird species hatch synchronously, incubation during the egg-laying stage should be disadvantageous because it makes the embryos develop asynchronously. We established the patterns of nest attendance during egg laying and the start of incubation in ring-necked pheasants, Phasianus colchicus, and tested three hypotheses regarding the advantage of early incubation. To determine nest attendance, we measured egg temperatures in real pheasant nests. Females spent more time on the nest as laying progressed, with an average of 6.4 h at a clutch size of 10. At the start of incubation, nest attendance increased to over 20 h/day. On the day before full incubation, time spent on the nest was positively correlated with the female’s condition and negatively with the number of breeding attempts she had already made that season. The hypothesis that an early start of incubation improves egg viability was rejected, as the predicted relationship between the number of eggs laid after the start of incubation and the number laid before the start of incubation was not significant. We also rejected the possibility that early incubation reduces the risk of nest parasitism, as it was negatively related to the number of females radiotracked around the nest. Our data supported the hypothesis that early incubation reduces the risk of nest predation by shortening the period of exposure, as the number of eggs laid after incubation started was positively related to the number of breeding attempts made by the female, and thus to the perceived predation risk, but was negatively related to the time of season.
(1) Effects of time unincubated on egg viability. Several studies have shown that the hatchability of duck eggs is reduced when incubation starts late (Mather & Laughlin 1977; Arnold et al. 1987; Arnold 1993). Furthermore, preincubation heating of eggs of domestic fowl is generally beneficial when they are stored for more than 7 days (for a review see Mayes & Takeballi 1984). From this hypothesis, we predict that more eggs will be laid after incubation starts in larger clutches. (2) Risk of intra- or interspecific nest parasitism. In waterfowl, females often lay eggs in other females’ nests during the latter’s egg-laying period (Rowher & Freeman 1989 and references therein), thereby leaving the task of incubation to the owner of the nest. An early start of incubation would reduce the risk of nest parasitism. (3) Risk of nest predation. Because of the longer egglaying period, a large clutch is exposed to nest predators for longer than smaller ones are (Perrins 1977). If incubation starts before the clutch is completed, this period of exposure will be reduced. We predict that females who have lost one or more nests to predators would then start incubation earlier. The majority of studies on egg-laying behaviour in precocial species have been done on waterfowl or on gallinaceous birds in captivity. In this paper we study nest
Traditionally it has been thought that precocial bird species do not start incubating until the clutch is completed, as it is important for the eggs to hatch synchronously (Kear 1970; del Hoyo et al. 1992). However, several studies have shown that waterfowl often begin to incubate during egg laying (see for example Caldwell & Cornwell 1975; Cooper 1978; Afton 1979, 1980; Cargill & Cooke 1981; Kennamer et al. 1990; Wilson & Verbeek 1995). Although the eggs in a clutch thereby have different incubation times, they are still able to hatch synchronously. This is possible because the embryos communicate with each other at the end of the incubation period, and adjust their time of hatching (Vince 1964, 1968, 1972; Lauch et al. 1988; Holmberg 1991; I. Persson & G. Andersson, unpublished data). Why do precocial species start to incubate during the laying period even though they do not benefit from developmental asynchrony? Three factors that may make an early start of incubation advantageous have been suggested. Correspondence: I. Persson, Ecology Building, S-223 62 Lund, Sweden (email: [email protected]). G. Go¨ransson is now at the Department of Natural Sciences, University College of Kalmar, P.O. Box 905, S-391 29 Kalmar, Sweden. 0003–3472/99/070159+06 $30.00/0
1999 The Association for the Study of Animal Behaviour
159
1999 The Association for the Study of Animal Behaviour
160
ANIMAL BEHAVIOUR, 58, 1
Figure 1. Egg temperature in an artificial egg placed in a pheasant’s nest at the beginning of the laying period. The arrows indicate the laying of eggs 5–12, and incubation starts after egg number 10.
attendance during laying in a wild population of ringnecked pheasants, Phasianus colchicus, by measuring egg temperatures, a method that avoids disturbing the females. By relating this to clutch size, physical condition, the number of breeding attempts the female has made that season and female density, we try to explain the initiation of full-time incubation by females during the egg-laying period. METHODS We trapped, measured and ringed 62 female pheasants during March and April 1997 in the Revinge area in southern Sweden. We fitted 20-g radiotransmitters (<3% of the body weight) on to the backs of the females; these did not seem to have any adverse effects on the females. The birds were released at the trapping site after a few hours and tracked daily between 1000 and 1400 hours, at the normal time of laying, from the end of April; 43 of the females were radiotracked until the middle of July and all their breeding attempts were followed. Because of nest predation (65% of the nests) or abandonment by the female (25% of the nests), females were often forced to make up to four breeding attempts per season before they produced any chicks. From the tracking data, we discovered ca. 40% of the breeding attempts during the laying period. When a nest was found, we fixed a dummy egg, containing paraffin wax and a thermistor connected to a data logger, into the nest. The wire connecting the egg to the data logger was buried in the soil under the nest, leaving the dummy egg in a natural position in the nest (for details see Persson & Go ¨ ransson 1996). This arrangement did not seem to affect the behaviour of the
females, which did not abandon the nests if the dummy egg was placed in the nest when five or more eggs had been laid. To see whether the extra egg placed in the nest had any effects on the laying sequence, we compared the clutch sizes of the first breeding attempt in nests where the dummy egg was placed in the nest before or after clutch competition; there was no significant difference (XSD=12.82.9 and 12.52.6 eggs, respectively; t test: t20 =0.262, P=0.8). The egg temperature was measured and saved every 16 min, showing the laying and incubation behaviour of the female, as the temperature rises when the female sits on the eggs and vice versa (Fig. 1). The start of incubation was apparent when the female changed from shorter nest visits to almost continuous nest attendance. By counting the eggs in the nests before and after the start of incubation, and comparing this with the temperature graphs, we could establish whether eggs were laid during the incubation period. We captured 28 females on the nest about a week after the start of incubation. We weighed them and measured their tarsus length, then released them. We calculated an index of their physical condition as the residuals of the individual’s weight from the overall regression of weight versus untransformed tarsus length during 1993–1996. Body weight corrected for tarsus length is a good estimate of fat content in other species (e.g. Iverson & Vohs 1982; Johnson et al. 1985). We estimated the risk of nest parasitism as the average number of females per day radiotracked within 100 m of the nest during the egg-laying period. It was calculated both with and without females that had a nest of their own. The presence of females without radiotransmitters
PERSSON & GO } RANSSON: EGG-LAYING BEHAVIOUR
10
8 Time spent on the nest (h)
around the nests studied was probably the same in the whole area, as the radiotagged females were assumed to represent a random sample of the population. In 19 breeding attempts we were able to put a dummy egg with a thermistor into the nest before the start of incubation. Three of these nests were depredated before laying was known to be completed, and so only 16 breeding attempts, all made by different females, could be used in the analysis of egg laying during incubation. In two breeding attempts the thermistor was placed into the nest after the beginning of incubation but before the clutch was completed. This means that the number of eggs laid after incubation started might have been underestimated. In the remaining 14 breeding attempts studied, we recorded at least 1 and a maximum of 7 days of the laying period. All statistical analyses were performed using Systat (Wilkinson 1992), and all probabilities refer to two-tailed tests. Means are given SD. The multiple regressions were made backwards, stepwise with a P value of 0.05 to remove variables and interaction terms. The models were first tested with all possible interactions, and the nonsignificant interactions were then removed one by one.
6
4
2
0
5
10 Egg number
20
15
Figure 2. Time spent on the nest per day during the laying period in relation to egg number in the laying sequence (N=38 days for 12 females; note that two pairs of data points have the same values).
RESULTS During the laying period, females went to the nest once per day, usually around noon. They arrived at the nest punctually; the time difference between the earliest and the latest egg laying of a breeding attempt was on average 1.330.93 h (N=12; two nests studied for only 1 day were excluded). The females stayed on the nest long enough for the egg temperature to rise to 30–35C, that is, a normal incubation temperature (Fig. 1). In 40 days of laying in 14 breeding attempts only one interruption in the laying sequence was observed, and yet that female spent 3.5 h on the nest that day. There was no occurrence of more than one temperature rise per day, thus nest parasitism did not occur while the nests were under observation, unless the owner of the nest happened to skip egg laying the particular day the nest was parasitized. Females spent longer on the nest per day as the laying period progressed (ANCOVA: F1,25 =89.78, P<0.001), from 3.321.19 h when five eggs were laid to 6.720.74 h at a clutch size of 10 (Fig. 2); females differed in nest attendance (F11,25 =9.23, P<0.001). Considering only the last egg laid before the start of incubation, the time spent on the nest was positively related to the physical condition of the female and negatively related to the number of breeding attempts and the interaction between the two (Table 1). A possible relation between the number of eggs laid and the time spent on the nest on the day before the start of incubation was excluded in the multiple regression. In 14 (88%) of the 16 clutches studied by egg temperature measurements, the female laid at least one egg after the beginning of incubation. The most frequent was one egg (69%, 11 nests), but also two (two nests) and three eggs (one nest) were laid after the start of incubation. Only in two nest attempts did the female start to incubate after the clutch was completed. Predators destroyed all
Table 1. Backwards, stepwise multiple regression between the time spent on the nest on the last day of laying before incubation started and the number of breeding attempts, the number of eggs laid before incubation and the physical condition of the female Variable
Physical condition No. of breeding attempts Physical condition × no. of breeding attempts
r
t
P
1.42 −1.29
2.709 −3.782
0.030 0.007
−1.50
−2.449
0.044
Only significant variables are shown: r2 =0.71, N=11.
but two clutches; however, in these successful clutches all the eggs hatched 25 days after the start of incubation, including the nest with three eggs laid after incubation started. To test the egg viability hypothesis, we analysed the number of eggs laid after incubation started in relation to the number laid before incubation started, and found a nonsignificant tendency for a negative relation between these variables (F1,14 =3.07, P=0.10). In an attempt to determine which factors might explain the number of eggs laid after incubation started, we used the number of breeding attempts and time of season (measured as number of days after 1 April) as independent variables in a multiple regression. This analysis showed a positive relation between the number of eggs and the number of breeding attempts made by the female (T=4.18, N=16, P=0.001; Fig. 3), as well as a negative relation between the number of eggs and time of season (T= 2.56, N=16, P=0.024). The average number of females per day radiotracked within 100 m of a nest, used as a measure of risk of nest parasitism, ranged from 0 to 2.3 and was not related
161
162
ANIMAL BEHAVIOUR, 58, 1
to the number of eggs laid after incubation started (r= 0.20, N=16, P=0.45). When females with a nest of their own were included in the analysis, we found a nonsignificant tendency for a negative relation between the number of eggs and the number of females in the vicinity (r= 0.47, N=16, P=0.07, average values ranging from 0 to 1.3 females per day). The number of females tracked within 100 m of a nest did not correlate with the number of females caught at the same trap as the nest owner (r=0.012, N=16, P=0.96), thus the density of females did not influence the risk of nest parasitism (the overall density of radiotracked females was 10–12 females/km2). DISCUSSION The egg temperature measurements revealed a fairly regular egg-laying pattern in pheasants. The females visited the nests once per day, at the same time every day, and only one interruption in the laying sequence was registered. Unless females make more interruptions during the early part of the laying period, which we did not study, the average number of eggs laid per day would approach 1.0 instead of the 0.71–0.77 previously recorded in gamefarm and wild pheasants (Buss et al. 1951; Robertson 1991; Wise 1995). This high egg production rate may be a way to shorten the time the nest is exposed to nest predators. As the females stayed on the nest for at least 2 h (up to over 8 h) during the egg-laying period, the egg temperature rose well over the physiological zero (20–27C, for a review see Meijerhof 1992), which is the temperature where the embryo starts to develop. Thus, even without the laying of extra eggs during incubation, development would be asynchronous. This has been shown in mallard ducks, Anas platyrhynchos, which incubate long enough for development to occur in the sixth egg in clutches of 10–12 eggs, leading to a difference in developmental age early in incubation of 1–2 days (Caldwell & Cornwell 1975). A similar pattern was found in northern shovelers, Anas clypeata (Afton 1979). Furthermore, in wood ducks, Aix sponza, larger clutches have greater levels of intraclutch developmental asynchrony than smaller ones (Kennamer et al. 1990). Female pheasants spent longer on the nest as laying progressed (Fig. 2). This is in agreement with the diminishing energy requirements for egg production late in the laying sequence (Drobney 1980), and may be important as camouflage for the eggs when the risk of nest predation is high. The nest attendance was also dependent on the physical condition of the female, those in good condition having a higher nest attendance. In addition to this, both the number of breeding attempts by the female and the interaction between physical condition and number of breeding attempts were negatively related to the time spent on the nest. One explanation for this might be that incubation is started earlier in later breeding attempts. It might be energetically hard to have high nest attendance when many eggs are yet to be laid. King (1973) estimated the daily maximum cost of egg production at 21–30% of the daily energy intake without gain or loss of body
weight in galliforms. Maintaining this high energy requirement until the beginning of incubation might be possible only with a lower nest attendance during laying. As in many other species, pheasants lay smaller clutches in later breeding attempts (Robertson 1991; I. Persson, unpublished data). This may be related to the simultaneous costs of incubation and egg laying, which may make it impossible to lay as large a clutch as in the first breeding attempt. Another explanation for the negative relation between the number of breeding attempts and the time spent on the nest could be that an early start of incubation in combination with high nest attendance may make the developmental asynchrony between the eggs too large for the chicks to hatch synchronously. However, there was no significant relationship between the number of eggs laid after incubation started and the time spent on the nest (r=0.091, N=12, P=0.8). The incubation period is defined as the development time of the embryo during regular attention of the parent (Drent 1975). However, in the majority of studies the incubation period is generalized as the time between laying and hatching of the last egg. This was not applicable in the pheasants we studied, as there was a clear increase in the amount of time spent on the nest, from a maximum of 8.5 h during laying to about 21 h during incubation, giving an obvious start of incubation, usually 1 day before the clutch was completed. A similar pattern was found in red Burmese junglefowl, Gallus gallus spadiceus (Meijer & Siemers 1993), and Canada geese, Branta canadensis (Cooper 1978). This abrupt start of incubation contrasts with that of many waterfowl, which often have a gradual start of incubation and spend over 50% of the day on the nest during laying (e.g. Kennamer et al. 1990; Wilson & Verbeek 1995). Our study shows that it is common for pheasants to lay eggs after the start of incubation, thus creating a developmental asynchrony of 1–3 days in addition to the asynchrony developed during the laying period. As hatching asynchrony is disadvantageous for precocial species, some selective pressure must make the early start of incubation advantageous. The egg viability hypothesis suggests that an early start of incubation enhances the hatchability of the eggs, predicting that the female should start to incubate relatively earlier in larger clutches. However, in our study the number of eggs laid after the start of incubation was not positively correlated with the number laid before the start of incubation, so we can reject the egg viability hypothesis. A second hypothesis suggests that the risk of nest parasitism makes an early start of incubation profitable. This is not applicable in our pheasant population, as we only rarely observed parasitic laying. Furthermore, there was a negative correlation between the number of eggs laid after the start of incubation and the number of females radiotracked around the nest, showing that it was not the density of females that made an early start of incubation profitable. Thus, the nest parasitism hypothesis was also rejected. The nest predation hypothesis predicts that females should start to incubate early to reduce the risk of nest predation. In our study, the number of eggs laid after the
PERSSON & GO } RANSSON: EGG-LAYING BEHAVIOUR
4
Number of eggs
3
2
1
0
1
2 3 Number of nest attempts
4
5
Figure 3. Number of eggs laid during incubation in relation to the number of breeding attempts made. When there was more than one data point on a specific set of values, they were plotted above each other in order to be visible. N=16.
start of incubation was positively related to the number of breeding attempts by the female that season (Fig. 3). This supports the nest predation hypothesis. When a female has lost a couple of nests, her perceived nest predation risk has increased, so she may start incubation earlier within the laying sequence. In this way she may try to avoid another nest being depredated, as both time and her energy reserves are running out. It is also possible that the females are influenced by daylength and try to hurry on the breeding as it is getting late in the season. However, there was an overall negative correlation between the number of eggs laid after the start of incubation and time of season, supporting the nest predation hypothesis. For the first breeding attempt, fewer eggs were laid after the start of incubation later in the season perhaps because the vegetation was higher and denser and the need for camouflage thereby less than earlier in the season. Acknowledgments We thank G. Andersson for assiduous and enthusiastic field work, J.-A r . Nilsson for inspiration and valuable comments during analyses and manuscript writing, and A. Hoodless and an anonymous referee for comments on the manuscript. This study was supported by grants to I.P. from the Foundation of C. F. Lundstro ¨ m and the Bokelund Fund, University of Lund. References Afton, A. D. 1979. Incubation temperatures of the northern shoveler. Canadian Journal of Zoology, 57, 1052–1056. Afton, A. D. 1980. Factors affecting incubation rhythms of northern shovelers. Condor, 82, 132–137. Arnold, T. W. 1993. Factors affecting egg viability and incubation time in prairie dabbling ducks. Canadian Journal of Zoology, 71, 1146–1152.
Arnold, T. W., Rohwer, F. C. & Armstrong, T. 1987. Egg viability, nest predation, and the adaptive significance of clutch size in prairie ducks. American Naturalist, 130, 643–653. Buss, I. O., Meyer, R. K. & Kabat, C. 1951. Wisconsin pheasant reproduction studies based on ovulated follicle technique. Journal of Wildlife Management, 15, 32–46. Caldwell, P. J. & Cornwell, G. W. 1975. Incubation behavior and temperatures of the mallard duck. Auk, 92, 706–731. Cargill, S. M. & Cooke, F. 1981. Correlation of laying and hatching sequences in clutches of the lesser snow goose (Anser caerulescens caerulescens). Canadian Journal of Zoology, 59, 1201–1204. Cooper, J. A. 1978. The history and breeding biology of the Canada geese of Marshy Point, Manitoba. Wildlife Monographs, 61, 1–87. Drent, R. 1975. Incubation. In: Avian Biology, Vol. 5 (Ed. by D. S. Farner & J. R. King), pp. 333–420. New York: Academic Press. Drobney, R. D. 1980. Reproductive bioenergetics of wood ducks. Auk, 97, 480–490. Holmberg, K. 1991. Mallard ducks, mate choice and breeding success. Ph.D. thesis, Stockholm University. del Hoyo, J., Elliott, A. & Sargatal, J. (Eds) 1992. Handbook of the Birds of the World. Barcelona: Lynx Editions. Iverson, G. C. & Vohs, P. A., Jr. 1982. Estimating lipid content of sandhill cranes from anatomical measurements. Journal of Wildlife Management, 46, 478–483. Johnson, D. H., Krapu, G. L., Reinecke, K. J. & Jorde, D. G. 1985. An evaluation of condition indices for birds. Journal of Wildlife Management, 49, 569–575. Kear, J. 1970. The adaptive radiation of parental care in waterfowl. In: Social Behaviour in Birds and Mammals (Ed. by J. H. Crook), pp. 357–392. London: Academic Press. Kennamer, R. A., Harvey IV, W. F. & Hepp, G. R. 1990. Embryonic development and nest attentiveness of wood ducks during egg laying. Condor, 92, 587–592. King, J. R. 1973. Energetics of reproduction in birds. In: Breeding Biology of Birds (Ed. by D. S. Farner), pp. 78–107. Washington, D.C.: National Academy of Sciences. Lauch, M., Nichelmann, M. & Wallschlaeger, D. 1988. Prenatal acoustic communication in muscovy ducklings Cairina moschata. Monatshefte fu¨r Veterina¨rmedizin, 43, 865–867. Mather, C. M. & Laughlin, K. F. 1977. Storage of hatching eggs: the effect on early embryonic development. British Poultry Science, 18, 597–604. Mayes, F. J. & Takeballi, M. A. 1984. Storage of the eggs of the fowl (Gallus domesticus) before incubation: a review. World’s Poultry Science Journal, 40, 131–140. Meijer, T. & Siemers, I. 1993. Incubation development and asynchronous hatching in junglefowl. Behaviour, 127, 309–322. Meijerhof, R. 1992. Pre-incubation holding of hatching eggs. World’s Poultry Science Journal, 48, 57–68. Perrins, C. M. 1977. The role of predation in the evolution of clutch size. In: Evolutionary Ecology (Ed. by B. Stonehouse & C. M. Perrins), pp. 181–191. London: Macmillan. Persson, I. & Go ¨ ransson, G. 1996. Nesting biology studied by egg temperature measurements. In: Proceedings of the International Union of Game Biologists XXII Congress (Ed. by N. Botev), pp. 56–61. Sofia: Pensoft. Robertson, P. A. 1991. Estimating the nesting success and productivity of British pheasants Phasianus colchicus from nest-record scemes. Bird Study, 38, 73–79. Rowher, F. C. & Freeman, S. 1989. The distribution of conspecific nest parasitism in birds. Canadian Journal of Zoology, 67, 239–253. Vince, M. A. 1964. Social facilitation of hatching in the bobwhite quail. Animal Behaviour, 12, 531–534. Vince, M. A. 1968. Retardation as a factor in the synchronization of hatching. Animal Behaviour, 16, 332–335.
163
164
ANIMAL BEHAVIOUR, 58, 1
Vince, M. A. 1972. Communication between quail embryos and the synchronisation of hatching. In: Proceedings of the XVth International Ornithological Congress (Ed. by K. H. Voous), pp. 357–362. Leiden: E. J. Brill. Wilkinson, L. 1992. SYSTAT for the Macintosh: Statistics, Version 5.2 Edition. Evanston, Illinois: SYSTAT.
Wilson, S. F. & Verbeek, N. A. M. 1995. Patterns of wood duck nest temperatures during egg-laying and incubation. Condor, 97, 963–969. Wise, D. R. 1995. Egg production of wild and gamefarm pheasants (Phasianus colchicus) strains under captive breeding systems. Gibier Faune Sauvage, 12, 321–330.