Attempts to condition homing pigeons to magnetic cues in an outdoor flight cage

Transcription

1 Animal Learning & Behavior (2), Attempts to condition homing pigeons to magnetic cues in an outdoor flight cage HUGH P. McISAAC and MELVIN L. KREITHEN University ofpittsburgh. Pittsburgh. Pennsylvania An attempt was made to train 9 homing pigeons to respond to the presence or absence of bar magnets by turning either left or right after flying the length ofa 20-ft outdoor flight cage. During initial training, color cues were placed in front offeeding stations on the left and right sides ofthe cage. The color cues were paired with magnetic cues by attaching either bar magnets or brass bars to the backs ofthe birds. The color cues were then deleted, leaving only the magnetic cues. Each pigeon received about 300 trials ofcolor training followed by about 200 trials of magnet testing. When only magnetic cues remained, none ofthe pigeons were able to choose the correct feeder at greater than chance levels of probability. The study of bird migration and homing in the last 20 years has revealed a variety of avian navigational mechanisms. Migratory birds, for example, have been found to have a celestial compass to guide their nocturnal flights, and homing pigeons have been shown to have a time-compensated sun compass to guide their diurnal flights on sunny days and an unknown, but not timedependent, guidance mechanism used on cloudy days. Orientation studies have suggested the possibility that the earth's magnetic field might be a factor in the birds' guidance system (Keeton, 1972), but unambiguous evidence that birds sense and respond to magnetic stimuli has proven difficult to obtain. Behavioral responses to magnetic fields have been reported for a number ofanimals, and magnetic particles have been found in an overlapping, but not identical, list ofanimals. Magnetic sensitivity has been established for magnetotactic mud bacteria and algae (Barros, Esquivel, Danon, & Oliveira, 1982; Blakemore, 1975); this sensitivity is due to linear chains of biogenic magnetite, which exert a torque on the bacterium, aligning the cell axis to the magnetic field of the earth (Kalmijn & Blakemore, 1978). The waggle dance of honeybees contains a residual misdirection trestmissweisung) that changes in response to artificial and natural magnetic fields, and the bees possess small superparamagnetic particles ofmagnetite; however, no clear connection has been established between the magnetite granules and the behaviors (Gould, Kirschvink, & Defeyes, 1978; Lindauer & Martin, 1968). The list of living organisms known to synthesize magnetic crystals has been increasing rapidly to include tuna, Portions of this work were submitted to Cornell University in partial fulfillment of the requirements for an undergraduate honors degree by H. McIsaac. The work was supported by NSF Grants BNS to M. Kreithen and W. T. Keeton and BNS to M. Kreithen, and by NIH Grant ROI NS22581-o2 to M. Kreithen. We thank Marian M. Kreithen for critical reading of the manuscript. Address reprint requests to M. L. Kreithen, Department ofbiological Sciences, Langley Hall, University of Pittsburgh, Pittsburgh, PA salmon, dolphins, bats, sea turtles, and others (reviewed by Gould, 1984). But with the exception ofthe magnetotactic bacteria, no behaviors have yet been linked experimentally to the presence of magnetic materials. The exact nature and location of magnetic particles (e.g., magnetite) in birds is far from resolved. Walcott, Gould, and Kirschvink (1979) reported 10-6 emu of magnetic materials in the heads ofpigeons, and Presti and Pettigrew (1980) reported similarquantities in the neck musculature of pigeons and sparrows; however, J. Crawford (personal communication, June 1985) was unable to locate these particles in pigeons, and Veda et al. (1982) were unable to locate magnetic particles in two species of migratory buntings and two nonmigratory passerine species. As is the case for most other animal groups, no causal link has been established between the behavior of birds in magnetic fields and the presence of magnetic materials in their bodies. Furthermore, until the nature ofthe behavioral responses of birds to magnetic fields is well understood, it would be premature to speculate on the role of magnetite in their behavior, especially inasmuch as magnetite is not necessarily required for detection of and response to magnetic fields. Elasmobranch fishes, for example, have electroreceptors (ampullae of Lorenzini) that contain no magnetite but can nevertheless be used to sense the induced currents produced by relative motion between the fish and the earth's magnetic field (Kalmijn, 1982). The behavior ofmigratory birds in magnetic fields can be tested during their briefseasonal periods ofnocturnal migratory restlessness (Zugunruhe) by placing the birds in orientationcages. Whenpresented with artificial magnetic fields, either alone or in conjunction with planetarium star fields, birds jumping in these cages exhibit a weak bias in the expected migratory direction. Several aspects of the migratory birds' responses to applied magnetic fields indicate an unusual detection mechanism. (1) The birds respond to the dip angle, not the polarity, of the field to determine the magnetic north axis; that is, they appear Copyright 1987 Psychonomic Society, Inc. 118

2 MAGNETIC CUES OUTDOORS 119 to have an axial compass. (2) When the birds are acclimated to the earth's magnetic field, their responses are confined to fields between.34 and.68 G; that is, they appear to have an intensity-tuned compass. (3) The birds require several hours to several days to acclimate to altered levels in the strength ofbackground magnetic fields; that is, they exhibit very long time constants for shifting their intensity tuning to a new field strength (Able, 1980; Alerstam & Hogstedt, 1983; Emlen, Wiltschko, Demong, Wiltschko, & Bergman, 1976; Gwinner & Wiltschko, 1978; Merkel & Wiltschko, 1965; Moore, 1977; Wiltschko & Wiltschko, 1972). The unusual nature of these responses suggests that if a test of the birds' magnetic sense is to succeed, great care must be exercised to match any behavioral test to the biology of the birds and the characteristics of the magnetic stimuli involved. The behavior ofhoming pigeons in free flight has provided some of the best evidence to date for magnetic sensitivity in birds. The orientation of pigeons under overcast skies is independent oftheir sun compass, and when bar magnets are attached to the backs ofthe pigeons, disorientation occurs under overcast but not under sunny skies. These findings have created speculation about a magnetic compass that is invoked only under overcast skies (loale, 1984; Keeton, 1971, 1972). The ontogeny and experience of the birds are also important variables, because even under sunny skies magnets can disorient the first flight of young birds and new-to-site birds (Keeton, 1972). Magnetic storms have been shown to deflect pigeons' departure bearings to the left under sunny skies; furthermore, there is a linear correlation between the strength of the naturally fluctuating magnetic field of the earth and the magnitude ofthe leftward deflection of the departure bearings chosen by homing pigeons released from a familiar site under sunny skies (Keeton, Larkin, & Windsor, 1974). Larkin and Keeton (1976) found that bar magnets mask these effects ofgeomagnetic variation under sunny skies by shifting the departure bearings to the left and thereby swamping the variable leftward biases that are due to magnetic storms. Although the leftward shift in all cases was small, between 2 and 5, it was consistent; however, our attempt to replicate these results (McIsaac & Kreithen, 1987) was not successful. Walcott (1977) showed that under sunny skies, pigeons with electromagnetic coils attached to their heads depart slightly to the left and with a slight increase in scatter compared to control birds. but that under overcast skies a large fraction ofthe pigeons exhibit reversed departure orientation when the applied field from the. head coils points north up (NUP) (Walcott & Green. 1974). The reverse orientation of NUPs under overcast skies is a result consistent with the axial compass model for migratory birds and for migratory salamanders (Phillips, 1986). Walcott's results were replicated by Visalberghi and Alleva (1979), who reported that NUPs produced reverse orientation under overcast but not under sunny skies, and that bar magnets produced a shift to the left and an increase in scatter under sunny skies. Considering the complex nature of the behavioral responses in free flight, we felt it would be advantageous to have a laboratory method for isolating and evaluating the birds' responses to magnetic stimuli. Many laboratory studies ofpigeons' responses to magnetic stimuli had already been tried. Reille (1968) reported successful cardiac conditioning to magnetic stimuli, but attempts to replicate and extend these results have failed (Beaugrand, 1976, 1977; Kreithen & Keeton, 1974). Other researchers who have attempted laboratory conditioning of pigeons to magnetic fields have reported negative results (Alsop, 1987; Carman et al., 1987; Meyer &. Lambe, 1966; Moore et al., 1987; Orgel & Smith, 1954.) In addition, there have been many unpublished unsuccessful attempts to condition pigeons to magnetic fields, including attempts done as exercises in student psychology laboratories. The articles in this volume, although they represent a substantial collection, are but a small fraction of the total number of studies bearing negative results. Given the tantalizing results ofthe field work, it is not clear why so many laboratory tests have failed. One of the few successful laboratory attempts to condition birds to magnetic stimuli was a flight-cage experiment by Bookman (1977) in which pigeons chose left or right feeding boxes in response to vertical magnetic fields in a magneticallyshieldedroom. Bookman's flightcage tests were a compromise between the carefully controlled but consistently unsuccessful laboratory tests and the frequently successful open-flight tests, in which the stimuli are more difficult to control. This compromise, free flight in a limited space, produced responses in the laboratory. The objectives ofthe presentstudy wereto extend Bookman's (1977) flight-eage findings to responses to bar magnets, to use the birds' responses to develop a reliable laboratory testing method for exploring magnetosensitivity, and to probe the physiology and biophysics of the magnetoreceptors ofbirds. We succeeded only in adding to an already long list yet another study of pigeon magnetosensitivity that has produced negative results. METHOD The experimental task was a two-choice discrimination test with color cues or magnetic cues determining the correct side, left or right, to be chosen at the end of an outdoor flight cage in order to obtain water reinforcement from a feeding station. Apparatus The flight cage, shown in Figure I, was a wooden frame covered with a Y4-in(.6 em) plastic mesh; it was 4 ft wide x 6 ft high x 20 ft long (1.2 x 1.8 x6.1 rn), The direction of flight from the entry doorto the far end ofthe cage was alignedtoward290 magnetic. At the end of the cage were two feeding stations, one on the left wall and one on the right wall, with entrance perches located 4 ft (1.2 m) above ground level. The door to each feeding station was I x I ft (30 x 30 ern) and had standard one-way pigeon bobs to trap the birds in the feeder. The floor of the feeding station was I ft wide x 1.5 ft deep (30 x 46 ern); the roof sloped upward from a height of I ft (30 cm) at the bobs to a height of 1.25 ft (38 ern) in the back. In the rear of each feeding station was a water dispenser made of a lo-ml test tube inserted through the lid of a l-ib

3 120 McISAAC AND KREITHEN f4' + ""'- TOP VIEW... t.. 1.5' ' Water container ~ Feeding box END VIEW 20' T LI Entry door SIDE VIEW Figure 1. Flight cage used to test pigeons' responses to color cues and to bar magnets worn on the birds' backs. The pigeons, after being placed by band tbrough the entry door, Dew to the far end of the cage where they landed on one of the two perches and entered the feeding box to obtain water reinforcement. coffee can. The presence or absence of water was not visible to the pigeon until it had entered the feeding station and the bobs had closed and locked behind it. Water was the reinforcement, as it proved to be a more reliable reward than food. The birds received all of their water in the test apparatus and were given free access to food in their home cages. The color cues were perches painted red or blue. The perches were 6 in. wide x 3 in. deep x 1/2 in. thick (15 x 8 x 1.3 em) and were mounted by means of wooden pegs that fitted into holes in front of each feeding station. The perches served as landing platforms for the pigeon prior to its walking through the bobs. For the magnet test (Task 3), the red and blue perches were replaced with white perches. The magnet/no-magnet cues were a bar magnet or a brass bar of the same size and weight; the bars were carried in Velcro backpacks, which attached to strips of Velcro glued to the birds' backs between their wings. The bar magnets were the type used by Keeton and his colleagues (Keeton, 1971, 1972; Larkin & Keeton, 1976) to produce disorientation in overcast homing flights; they were, in fact, from the same batch of magnets. Each bar magnet was positioned so that the north-seeking end pointed toward the pigeon's head. The magnets were 2.56 em long,.64 wide, and.33 em thick, and had a pole strength of 85 cgs units and a field strength of0.45 Oe 10 ern from the center. The weight of each magnet (or brass bar) was approximately 4 g. Subjects The subjects were 9 homing pigeons taken at 30 days of age from the Cornell University homing pigeon breeding colony to an outdoor home loft approximately 4 miles from the breeding colony. The pigeons were housed communally and had free access to food but not to water. The birds were allowed several weeks of freeflight experience around the home loft so that they would be able to return on their own if they escaped accidentally from either the home loft or the nearby flight cage. Each bird was taken daily from the communal cage and tested in the flight cage, where it received all its water from the feeders in the testing apparatus. There were 20 trials per bird per day. Training Initially each bird was presented with only one perch of one color, the color (either red or blue) it was to associate with the reinforce-

4 MAGNETIC CUES OUTDOORS 121 ment. The feeding station on the opposite side of the cage had no perch, and the bobs were locked to prevent an incorrect response. To obtain water, the bird had to land on the painted perch and enter the feeding station. The painted perch, with its accompanying reinforcements, was switched back and forth between the left and right sides of the flight cage in a restricted random sequence throughout training. After each bird was familiar with the flight cage, the perch, and the feeder, a graded sequence of three tasks was begun. Task 1: Asymmetric color cues. For the first task, both feeding stations had perches, one red and one blue. Ifblue was the color that had been associated with water during a bird's initial training, then the bird had to enter the feeding station with the blue perch in order to receive water; the feeding station with the red perch contained no water. If red was the color that had been associated with water during a bird's initial training, then the bird had to enter the feeding station with the red perch in order to receive water; the feeding station with the blue perch contained no water. To minimize possible biases due to color preferences, the groups in this test, as in all other tests, were either equal in size or as close to equal as possible, given an odd number of birds. The birds did not wear magnets or brass bars for Task 1. Task 2: Symmetric color cues. The second task employed symmetric color cues; that is, the two feeding station perches were the same color, and left-right choices to obtain water depended on the shared color of the two perches. If, for example, the bird saw two blue perches, the correct choice was to turn left; but if it saw two red perches, the correct choice was to turn right. To avoid biases, pigeons trained to blue-left were alternated with pigeons trained to red-left. The symmetry ofthis task was chosen to resemble the final magnet task in that the right-left choices were not directly labeled by the cues; the birds had to translate symmetric cues into a right or left choice. Throughout Task 2, the pigeons wore brass bars or magnets in conjunction with the colors. The blue-left birds were divided into two groups: those that wore magnets for left, brass bars for right, and those that wore magnets for right, brass bars for left. The red-left group was similarly divided. Deletion of color cues. After the pigeons were able to perform Task 2 satisfactorily, they received 40 trials of transitional training. The pigeons continuedto wear the appropriate magnet or brass bar for each trial, but the color cues for right and left were removed for every other trial. That is, trials in which two red or two blue perches indicated right or left were alternated with trials in which the perches were both white, a color having no directional significance. Task 3: Magnetic cues only. In Task 3, the only directional cues available to the birds were the bar magnets or brass bars carried in their backpacks. There were no longerany directional colorcues accompanying the magnet or brass bar cues; both feeding station perches were white for every trial. To minimize possible side biases, pigeons trained to magnet-left were alternated with pigeons trained to magnet-right. During all three tasks, the sides (right and left) producing reinforcement and no reinforcement were transposed in a restricted random sequence that avoided sampling biases for short runs of 20 trials, so that each sequence contained an equal number oflefts and rights and no sequence had more than three lefts or three rights in succession. RESULTS Eight of the 9 pigeons learned to differentiate between the color cues in Task I or Task 2. Correct responses from cumulative graphs for Tasks 1 and 2 averaged just under 70%. In Task 2, 7 of the 9 pigeons had cumulative responses significantly greater than 50% at the 1% con- fidence level and one of the pigeons (3309) had a cumulative response significant at the 5% level. Of the 7 pigeons that went on to Task 3, 6 had performedwell on the color training tasks. None of the 7 pigeons in Task 3 (magnetic stimuli only) performed well. Six pigeons produced responses that were not significantly different from chance. One pigeon (3222) had a response mean of.572, which yields a 5% lower boundary to the confidence interval of.503; this is just barely statistically significant and is presumably of little behavioral significance (see Table 1 and Figure 2). Although, as individuals, 6 of the 7 pigeons produced proportions of correct responses slightly greater than.50, when examined collectively the seven proportions ofcorrect responses did not differ significantly from.50 [1(6) = 1.549, P =.17 (two-tailed test)]. The slight, nonsignificant response bias may have resulted from the presentation ofthe stimuli in a restricted random sequence, which can allow a subject to perform consistently above or below chance values. DISCUSSION The outcome of this study was that the birds did not respond to the magnetic stimuli. The fact that the birds did, however, respond to the colorcues suggests that the apparatus, training procedures, and shaping methods were viable and might have been expected to work for other stimuli, including magnetic fields, had the birds been capable of detecting them. Assuming, then, that the apparatus and training were adequate, we must address other questions about our tests, for although they were similar in many ways to Bookman's (1977) successful flight-cage tests, there were some differences. Were the magnetic stimuli appropriate? Bookman (1977) used a magnetically shielded room (0.02 G) and added a 0.5-G vertical field produced by three pairs of Helmholtz coils. The pigeons in our tests wore bar magnets in an outdoor flight cage. We chose to use the bar magnets because they were the magnets that, in field experiments, had produced disorientation under overcast Table 1 Responses to Stimuli: Proportion of Correct Responsesrrotal Trials (n) Bird Task 1 Task * (85) * (113) (95) * (138):f:.693* (231) t (232).784* (232) * (248).680* (222) * (247).711* (246) * (220).607* (150) * (247).58Ot (176) Task 3.572t (229).530 (215).547 (190).514 (175).528 (235).557 (235).435 (184) Note-Statistical significance computed using the normal approximation to confidenceintervals of the mean of binomial proportions for large sample sizes, where p=.50. *Statistically significant at p=.ol. tstatistically significant at p =.05. :f:preceded by.445 (101) during initial shaping trials.

5 122 McISAAC AND KREITHEN 100 _ o COLOR CUES MAGNETIC CUES 75 cf. en w en 50 Z 0 n, en w 25 a: a BIRDNUMBER Figure 2. Bar graphs of the responses of 7 pigeons to color cues paired with magnetic cues (Task 2, solid bars) and to the magnetic cues alone (Task 3, open bars). Although magnets had been worn during Task 2, deletion ofthe color cues in Task 3 caused the responses to drop to cbance values. See Table 1 for additional details. skies and left shifts and increased scatter under sunny skies (Keeton, 1972). Because these bar magnets had been observed to produce an effect on pigeons in the field, it seemed reasonable to test pigeons wearing them in a flight cage, on the assumption that the magnetic cues available to the pigeons in free flight in the field would be available to the pigeons in free flight in the outdoor cage. It should be noted, however, that the pigeons wearing bar magnets carried the applied magnetic field with them, whereas the pigeons in Bookman's tests flew through the magnetic field produced by the Helmholtz coils. If the pigeons' mechanism for detection of magnetic fields is an induced electrical potential caused by the relative motion ofthe magnetic field and some conductor in the bird, then this difference is critical. Yet induction as a mechanism of magnetic field detection in pigeons had been studied with negative results (Kreithen & Keeton, 1974), and the bar magnets had appeared to have an effect in Keeton's field tests (Keeton, 1971, 1972; Larkin & Keeton, 1976). Therefore, if bar magnets were the cause of the negative results in our study, the way in which they were the cause is a puzzle. Was our testing of one pigeon at a time appropriate? Bookman (1977) flew pairs of birds because he found that pairs were more active than single birds. He noticed that during their social interactions some of the birds exhibited fluttering (hovering) flight. Only those birds that spent more than 3 sec in fluttering flight appeared to train to the magnetic stimuli. Our pigeons were quite active in the flight cage, so the extra complexity of flying pairs of pigeons did not seem necessary. Locomotion has been an important element in successful magnetic tests (reviewed in Kreithen & Keeton, 1974), and the birds in our flight cage seemed to exhibit an appropriate level of locomotion. We attempted to encourage fluttering flight by placing obstacles in the flight cage, but because the arrangement proved too cumbersome for the pigeons, we removed the barriers. Because our birds flew straight to the end ofthe cage without hovering, fluttering flight is a variable that is not resolved by this replication study. Were the pigeons appropriate? Bookman (1977) used homing pigeons from three lofts, one of which was the Cornell University loft. His Cornell birds performed successfully. All ofour birds were homing pigeons from the Cornell University loft, and all had had flight experience. These same birds could have been expected to respond to the bar magnets in outdoor field tests; therefore, it seemed reasonable to expect them to respond in our outdoor flight cage. Weare left with no satisfactory explanation for the failure ofthe tests. Our methods would have worked for most sound and light stimuli. Magnetic stimuli, if available to the pigeons at all, seem not to be available in any conventional way; and the pigeons' responses to magnetic stimuli, if they exist at all, seem not to be available to researchers in any conventional way. An unambiguous demonstration that homing pigeons respond to magnetic fields remains a challenge for the future. REFERENCES ABLE, K. (1980). Mechanisms oforientation, navigation, and homing. In S. Gauthreaux (Ed.), Orientation, navigation, andhoming (pp ). New York: Academic Press. ALERSTAM, T., & HOGSTEDT, G. (1983). The role ofthe geomagnetic field in the development of birds' compass sense. Nature, 306, BARROS, H. G. P. LINSDE, ESQUIVEL, D. M. S., DANON, J., & Ouv EIRA, L. P. H. DE. (1982). Magnetotactic algae. Academia Brasileria CBPF Notas Fisicas, 48,

6 MAGNETIC CUES OUTDOORS 123 BEAUGRAND, J. P. (1976). An attempt to confirm magnetic sensitivity in the pigeon, Columba livia. Journal ofcomparative Physiology, 110, BEAUGRAND, J. P. (1977). Test of magnetic sensitivity in seven species ofeuropean birds using a cardiac nociceptive conditioning procedure. Behavioural Processes, 2, BLAKEMORE, R. P. (1975). Magnetotactic bacteria. Science, 190, BOOKMAN, M. A. (1977). Sensitivity ofthe homing pigeon to an earthstrength magnetic field. Nature, 267, EMLEN, S. T., WILTSCHKO, W., DEMONG, N. J., WILTSCHKO, R., & BERGMAN, S. (1976). Magnetic direction finding: Evidence for its use in migratory indigo buntings. Science, 193, GOULD, J. L. (1984). Magnetic field sensitivity in animals. Annual Review ofphysiology, 4(), GoULD, J. L., KiRSCHVINK, J. L., & DEFEYES, K. S. (1978). Bees have magnetic remanence. Science, 202, GWINNER, E., & WILTSCHKO, W. (1978). Endogenously controlled changes in migratory direction ofthe garden warbler. Journal ofcomparative Physiology, 125, IOALE, P. (1984). Magnets and pigeon orientation. Monitore Zoologico Italiano, 18, KALMIJN, A. J. (1982). Electric and magnetic field detection in elasmobranch fishes. Science, 218, KALMIJN, A. J., & BLAKEMORE, R. P. (1978). The magnetic behavior of mud bacteria. In K. Schmidt-Koenig & W. T. Keeton (Eds.), Animal migration, navigation, and homing (pp ). Berlin: Springer. KEETON, W. T. (1971). Magnets interfere with pigeon homing. Proceedings of the National Academy of Sciences (Washington, DC), 68, KEETON, W. T. (1972). Effects of magnets on pigeon homing. In S. R. Galler, K. Schmidt-Koenig, G. J. Jacobs, & R. E. Belleville (Eds.), Animal orientation and navigation (NASA SP-262, pp ). Washington, DC: U. S. Government Printing Office. KEETON, W. T., LARKIN, T. S., & WINDSOR, D. M. (1974). Normal fluctuations in the earth's magnetic field influence pigeon orientation. Journal ofcomparative Physiology, 95, KREITHEN, M. L., & KEETON, W. T. (1974). Attempts to condition homing pigeons to magnetic stimuli. Journal ofcomparative Physiology, 91, LARKIN, T. S., & KEETON, W. T. (1976). Bar magnets mask the effect of normal magnetic disturbances on pigeon orientation. Journal of Comparative Physiology, 110, LINDAUER, M., & MARTIN, H. (1968). Die Schwereorientierung der Bienen unter dem Einfluss des Erdmagnetfeldes. Zeitschrift fur vergleichende Physiologie, 60, McISAAC, H. P., & KREITHEN, M. L. (1987). Bar magnets do not significantly alter pigeon orientation from a familiar release site. Manuscript in preparation. MERKEL, F. W., & WILTSCHKO, W. (1965). Magnetismus und Richtungsfinden zugunruhiger RotkehIcben (Erithacus rubecula). Die Vogelwarte, 23, MEYER, M. E., & LAMBE, D. R. (1966). Sensitivity ofthe pigeon to changes in the magnetic field. Psychonomic Science,S, MOORE, F. (1977). Geomagnetic disturbanceand the orientationofnoc turnally migrating birds. Science, 196, ORGEL, A. R., & SMITH, J. C. (1954). Test of the magnetic theory of homing. Science, 120, PHILUPS, J. B. (1986). Two magnetoreception pathways in a migratory salamander. Science, 233, PREsTI, D., & PETTIGREW, J. D. (1980). Ferromagnetic coupling to muscle receptors as a basis for geomagnetic field sensitivity in animals. Nature, 285, REILLE, A. (1968). Essai de mise en evidence d'une sensibilitedu pigeon au champ magnetique ai'aide d'un conditionnement nociceptif. Journal de Physiologie (Paris), 60, VEDA, K., KUSUNOKI, M., KATO, M., KAKIZAWA, R., NAKAMURA, T., YASKAWA, K., KOYAMA, M., & MAEDA, Y. (1982). Magnetic remanences in migratory birds. Journal ofthe Yamashina Institutefor Ornithology, 14, VISALBERGHI, E., & ALLEVA, E. (1979). Magnetic influences on pigeon homing. Biological Bulletin, 156, WALCOTT, C. (1977). Magnetic fields and the orientation of homing pigeons under sun. Journal ofexperimental Biology, 70, WALCOTT, C., GoULD,J., & KiRSCHVINK, J. (1979). Pigeons have magnets. Science, 205, WALCOTT, C., & GREEN, R. (1974). Orientation ofhorning pigeons altered by a change in the direction of an applied magnetic field. Science, 184, WILTSCHKO, W., & WILTSCHKO, R. (1972). Magnetic compass ofeuropean robins. Science, 176, (Manuscript received February 28, 1986; revision accepted for publication October 27, 1986.)