Short Communications 597

Table 1. The mate preference of four different experimental groups of zebra finch males which were housed with pairs of zebra finches (ZF) and Bengalese finches (BF) in the sequences shown

Group

Treatment during Mean mate preference (% directed song phrases

0-35 days 35 70 days 80-115 days to the zebra finch female)

A ZF ZF ZF 100 (N=8) B BF BF BF 12 (N=6) C ZF BF ZF 94 (N= 10) D BF ZF BF 77 (N= 10)

siblings. Cross-fostered zebra finches raised alone or with only one sibling court the Bengalese finch female exclusively: only males raised with two or more siblings direct 5% or more of their courtship song towards the conspecific female (Kruijt et al. 1983). This might also explain the difference between Ten Cate et al.'s results and those of Immelmann and the present study, where siblings were transferred with the young male.

Immelmann (personal communication) suggests aggressive interactions between siblings influence mate preference. Other aggressive encounters might also be important. Ten Care et al. (1984) found Bengalese finches were less aggressive than zebra finches but, the greater the number of aggressive and non-aggressive initiatives by Benga- lese finches towards the young, the greater the preference for Bengalese finch.

Social experience plays an important part in the development of sexual preference in male zebra finches. Consequently, subtle differences in experi- mental design can lead to quite different results and caution should be exercised when drawing com- parisons. However, a common factor appears to be the importance of aggression: this might have general applicability to many aspects of learning.

I thank the University of St. Andrews and I.C.I. for their financial support and Peter Slater, Brian Combes and Lucy Eales for comments and advice. 1 also thank Klaus Immelmann and J6rg B6hner for their interesting discussions during my stay in Bielefeld, and last, but not least, Carel ten Cate, for all his stimulating suggestions.

NICKY CLAYTON

Department o f Zoology and Marine Biology, The University o f St. Andrews, Fife K Y I 6 9TS, U.K.

References

Immelmann, K. 1972. The influence of early experience on the development of social behaviour in estrildine finches. Proe. X V Int. Ornithol. Congr. Den Haag. 1970, 316~338.

Immelmann, K. & Suomi, S. J. 1981. Sensitive phases in development. In: Behavioral Development (Ed. by K. Immelmann, G. W. Barlow, L. Petrinovich & M. Main), pp. 395 431. Cambridge: Cambridge Univer- sity Press.

Kruijt, J. P., Ten Cate, C. J. & Meeuwissen, G. B. 1983. The influence of siblings on the development of sexual preferences of male zebra finches. DeN Psychobiol., 16, 233 239~

Ten Cate, C. J. 1986. Sexual preferences in zebra finch (Taeniopygia guttata) males raised by two species (Lonchura striata and Taeniopygia guttata) I. A case for double imprinting. J. eomp. Psyehol., 100, 248-252.

Ten Cate, C. J., Los, L. & Schilperoord, L. 1984. The influence of differences in social experience on the development of species recognition in zebra finch males. Anita. Behav., 32, 852-860.

( Receiced 30 June 1986; rerised 8 September 1986; MS. number: sc-320)

Djungarian Hamster Females Conceive in the Presence of Multiple Sibling Males

Wynne-Edwards & Lisk (1984) found that, although female Djungarian hamsters conceived in the presence of a single male, they copulated but did not conceive when they had full access to multiple males. After mating on a given day, the females mated 'again on day 5, day 9, day 13, day 17 and so on'. Wynne-Edwards & Lisk concluded that 'although a female Djungarian hamster shows behaviourat receptivity when more than one male is present, she does not go through either a pregnancy or a pseudopregnancy following the mating'.

598 Animal Behaeiour, 35, 2

The phenomenon resembles the pericopulatory pregnancy blockage found in deer mice, Peromys- cus maniculatus bairdi, by Dewsbury (1982a) in that in both there is a disruption of pregnancy caused by multiple-male exposure at the time of mating. Both resemble the Bruce effect (Bruce 1959) in that there is a male-induced pregnancy disruption, but differ in that exposure occurs at the time of mating. This difference has imporlant implications for the con- ditions under which one should search for the phenomenon, if it occurs, in nature. The Bruce effect and deer mouse pericopulatory block prob- ably have a common underlying mechanism. For example, neither occurs with females in postpar- tum oestrus (Dewsbury 1985). That the hamster pericopulatory blockage may have a similar mechanism was suggested by the finding that females appear to ovulate but not implant (Wynne- Edwards & Lisk 1984).

During routine breeding of Djungarian hamsters we found an unexpected set of conditions under which the pericopulatory blockage in that species fails to occur. Therefore, systematic data were collected from our breeding colony of Djungarian hamsters, Phodopus sungorus camphelli, which was established in 1981 from stock provided by Michael R. Murphy of the National Institute of Mental Health.

Animals were housed in clear polycarbonate cages, 48 x 27 x 13 cm for four or more animals and the breeding pairs, or 29 x 19 x 13 cm for three or fewer animals. The animals were given fresh water and Purina Laboratory Rodent Chow 5001 ad libitum and were under a reversed 16 h light:8 h dark photoperiod. Sibling groups were weaned at 21 days of age. After noting that inbred litters often appeared in the sibling group cages, we selectively manipulated the sex ratio of sibling groups at weaning to create groups with either a single male

(one male with one, two, or three females) or multiple males (two or three males with one or two females) present. Sibling groups were checked daily for inbred litters, Upon the birth of an inbred litter, the date and number of pups in the litter were recorded and the pups were removed and killed. Data were recorded for each sibling group to an age of 150 days.

The results of the data collection can be seen in Table |. Sibling groups in all six sex ratio conditions produced inbred litters. Overall, 86% (18) of 21 sibling groups were found to produce at least one inbred litter. Sibling groups with either one male or multiple (two or three) males produced approxi- mately the same percentage of inbred litters (82% versus 90%). The latencies to first litter for the sibling groups were not significantly different for single male and multiple male litters (P>0.05 , Mann-Whi tney U-test). Neither the mean number of litters produced before the 150 day criterion nor the mean number of pups per litter differed signifi- cantly for the single- and multiple-male sibling groups.

These results are not inconsistent with those of Wynne-Edwards & Lisk (1984). Rather, they help delineate the range of conditions under which the phenomenon they described does and does not occur. What is especially surprising is the improve- ment in breeding with incestuous pairings. It has generally been found that such conditions disrupt reproductive activity (e.g. Dewsbury 1982b). Argu- ments that inbreeding is beneficial have been developed by some authors (e.g. Shields 1982) and these may apply.

Both Dewsbury (1982a) and Wynne-Edwards & Lisk (1984) speculated that the pericopulatory block might have evolved to defer reproduction during times of possible social instability. As the presence of multiple-male siblings appears less

Table 1. Sex ratio, number of sibling groups, percentage of sibling groups with inbred litters, latency to first litter, number of litters, and mean number of pups per litter for Djungarian hamster sibling groups

"., Sibling Latency from groups with weaning

Number No. sibling inbred to first Mean no. Mean no. of males groups litters litter (days) litters pups/litter

1 I1 82 I08.8 (12.9) 1.6 (0.6) 3"5 (0"8) 2o r3 10 90 94.0 (6.0) 2.0 (0.8) 4.1 (1.0)

Means are presented with 95'!/,, confidence intervals. There were 16 females in the one-male condition (seven in 1F litters, six in 2F, and three in a 3F litter) and 15 females in the multi-male condition (five 2M2F litters, three 2M 1F litters, and two 3M IF litters).

Short Communications 599

likely than that of multiple strange males to signal impending instability, these results can be taken as consistent with that hypothesis. However, the hypothesis fails in other respects; for example in explaining the lack of effect in postpartum oestrus.

Although the adaptive significance is not yet clear, the finding of a limiting set of conditions for the pericopulatory block provides one additional brick in building an understanding of functional aspects of pregnancy blockage. Successful breeding within litters is a surprising result; however, Djun- garian hamsters do conceive in the presence of multiple sibling males.

This work was supported by grant BNS-8520318 from the National Science Foundation.

BRUCE FERGUSON DONALD A, DEWSBURY

Department of Psychology, University o[ Florida, Gainesville, FL 32611, U.S.A.

References

Bruce, H. M. 1959. An exteroceptive block to pregnancy in the mouse. Nature, Lond., 184, 105.

Dewsbury, D. A. 1982a. Pregnancy blockage following multiple-male copulation or exposure at the time of mating in deer mice, Peromyseus manicuhttus. Beha~:. Ecol. Sociobiol., I1, 37~42.

Dewsbury, D. A. 1982b. Avoidance of incestuous breed- ing between siblings in two species of Peromyscus mice. Biol. Behar., 7, 157-168.

Dewsbury, D. A. 1985. Studies of pericopulatory block- age and the gestation period in deer mice (Peromyscus maniculatus). Horm. Behat,., 19, 164~ 173.

Shields, W. M. 1982. Philopao3., Inbreeding. and the Evolution of Sex. Albany, N.Y.: State University of New York Press.

Wynne-Edwards, K. E. & Lisk, R. D. 1984. Djungarian hamsters fail to conceive in the presence of multiple males. Anim. Behae., 32, 626-628.

(Received l August 1986; retised 5 August 1986." MS. number. AS-385)

Geomagnetic Disturbance and Migratory Bird Orientation: Is There an Effect?

Most workers now seem to find the evidence persuasive that some species of migratory birds and homing pigeons possess a magnetic compass, and behavioural effects of earth-strength magnetic fields have been reported in a variety of taxa (Ossenkopp & Barbeito 1978; Wiltschko 1983; Kirschvink et al. 1985). For migratory birds, the evidence supporting magnetic effects on orien- tation comes primarily from experiments per-

formed on birds in orientation cages. On the other hand, attempts to discover evidence of magnetic involvement in the orientation of free-flying migrants have yielded mostly negative and incon- sistent results and they provide the most notable inconsistency between orientation cage studies and field observations (Able & Cherry 1986).

Southern (1978 and references cited therein) first reported a relationship between the orientation behaviour of ring-billed gull, Larus delawarensis, chicks and the irregular natural disturbances in the earth's magnetic field (magnetic storms) caused by sunspot and solar flare activity. Subsequently, investigators have sought magnetic storm effects in other sets of orientation data (Keeton et al. 1974; Able 1974, 1982; Richardson 1974, 1976; Larkin & Keeton 1976; Moore 1977). In both spring and autumn, Moore (1977) found strong positive linear correlations between the dispersion in flight direc- tions of passerine nocturnal migrants sampled by portable ceilometer in the southeastern United States and the intensity of magnetic disturbance. Others have failed to observe similar effects in field studies of night migrants (Able 1974, 1982: Richardson 1974), although Richardson (1976) and Larkin & Sutherland (1977) reported other, possibly magnetic influences. Whether birds do indeed respond directly to these normal fluctua- tions in the earth's magnetic field is important in two regards: (1) they are the only data that suggest a magnetic influence on migrants in free flight: and (2) response to these minute fluctuations implies a remarkable detection sensitivity by birds to changes of the order of 10 4 gauss (=10 gam- mas = 10 nT) or about 0-02% of total field intensity (approximately 0-5 gauss = 50 000 nT).

Here I report the results of an analysis of a large data set of flight directions of passerine nocturnal migrants gathered over a number of years in both spring and autumn in the northeastern and south- eastern United States. The analysis revealed no relation between the dispersion of orientation directions and the intensity of natural magnetic disturbance.

The data consist of flight directions of migrants observed passing through the beam of two 100-W ceilometer lights (methods described in Able 1974: Gauthreaux 1980). Data were collected at three localities: Lake Charles, Cameron Parish, Loui- siana (LCH), during autumn 1969; Athens, Clark County, Georgia (AHN), during autumn 1970: Berne, Albany County, New York (ALB), during the autumns of 1972, 1973, 1976, 1978 1980, 1982 and during the springs of 1974, 1976, 1977, 1980. Only one sample per night was used and these were all obtained during the first 3 h after dark.

For each sample of visually observed flight