DOI: 10.1126/science.1060514, 102 (2001);293 Science

et al.Guido Dehnhardt)Phoca vitulinaHydrodynamic Trail-Following in Harbor Seals (

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): September 9, 2014 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/293/5527/102.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/suppl/2001/07/05/293.5527.102.DC1.html can be found at: Supporting Online Material

http://www.sciencemag.org/content/293/5527/102.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/293/5527/102.full.html#ref-list-1, 4 of which can be accessed free:cites 9 articlesThis article

46 article(s) on the ISI Web of Sciencecited by This article has been

http://www.sciencemag.org/content/293/5527/102.full.html#related-urls30 articles hosted by HighWire Press; see:cited by This article has been

http://www.sciencemag.org/cgi/collection/physiologyPhysiology

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2001 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Sep

tem

ber

9, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

9, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

9, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Sep

tem

ber

9, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

Hydrodynamic Trail-Followingin Harbor Seals (Phoca vitulina)

Guido Dehnhardt,1,2* Bjorn Mauck,1,2 Wolf Hanke,1

Horst Bleckmann1

Marine mammals often forage in dark or turbid waters. Whereas dolphins useecholocation under such conditions, pinnipeds apparently lack this sensoryability. For seals hunting in the dark, one source of sensory information mayconsist of fish-generated water movements, which seals can detect with theirhighly sensitive whiskers. Water movements in the wake of fishes persist forseveral minutes. Here we show that blindfolded seals can use their whiskers todetect and accurately follow hydrodynamic trails generated by a miniaturesubmarine. This shows that hydrodynamic information can be used for long-distance prey location.

Aquatic animals often have to cope with con-ditions where visibility is drastically reduced.Consequently, many aquatic species havesensory abilities that may supplement or evensubstitute vision, such as the active sonarsystem in toothed whales (1). However, al-though pinnipeds underlie similar ecologicaldemands, corresponding sensory abilitieshave remained unknown in these marinemammals. As suggested by the wide distri-bution of hydrodynamic receptor systems,water movements generated by prey, preda-tors, or conspecifics, as well as abiotic sourc-es such as tides and currents, provide impor-tant sensory information (2). Like the fishlateral line (3), the whiskers of harbor sealsare highly sensitive to water movements (4).However, the accepted view is that hydrody-namic object detection works only over shortdistances [for review, see (2)], as, for in-stance, during chemosensory mate search inmarine copepods (5) or during the final stageof prey pursuit of seals (6). Particle velocitiesattenuate rapidly with distance from the flowfield–generating source. Although this is truefor stationary vibrating sphere stimuli usuallyused in laboratory studies on hydrodynamicreceptor systems (3), the wake behind aswimming fish shows a vortex structure (7, 8)with particle velocities above threshold ofmost hydrodynamic receptors several min-utes after the fish has passed by (9). Thus, aswimming fish can leave a hydrodynamictrail of considerable length that piscivorouspredators might use for long-range preydetection.

To find out whether seals can locate dis-tant objects by hydrodynamic trail-following,trails were generated with a propeller-driven

miniature submarine (10, 11) and were visu-alized and measured using Particle ImageVelocimetry (PIV) (10–12). Interpolation ofoverlapping measurements showed that sim-ilar to fish-generated hydrodynamic trails, thesubmarine’s trail was a narrow street (widthafter ;20 s: ,60 cm) of turbulent watermovements. Compared with adjacent areas, itcontained a change in the main direction ofwater flow (11, 13) and higher water veloci-ties of .16 mm s21 after 20 s (11). This isthe same order of magnitude as velocitiescalculated for the wake of a fish of 30-cmbody length (extrapolated from goldfish-gen-erated trails) (9). The submarine’s trail dif-fered from fish trails, however, in that themain water flow was opposite to its swim-ming direction and it consisted of a rather

unstructured pattern of turbulent water move-ments that broke down after ;20 s (13). Incontrast, the fish-generated trails studied sofar consist of stable, ladder-like chains ofvortices (8, 14). Vortices in the wake of asmall goldfish persist for .30 s and watervelocities were, under laboratory conditions,significantly higher than background noisefor at least 3 min (9). Although decayingfaster, the submarine’s trail represents atrackable continuum of turbulence and veloc-ities deviating from the surrounding (11, 13).

The first block of experiments were con-ducted at the “Seehundstation Friedrich-skoog” (Germany) with an experimentallynaıve male harbor seal (“Henry”) kept in anoutdoor pool filled with naturally turbid sea-water. The seal was first trained to locate thesubmarine without any sensory restrictions.In the final experimental procedure, a trialstarted with blindfolding the seal with a vi-sually opaque stocking mask, supplying itwith head phones for acoustical masking, andplacing its head in a hoop station 40 cmabove the water surface (11). The submarinewas then started underwater from a platformclose to the station (Fig. 1A). The seallearned that the trail always could be found;1 m around its station. About 2 s after thesubmarine’s engine had turned off (runningtime 3 to 5 s), the headphones were removedfrom the seal’s head, which was the signal forthe animal to start its search.

First, “linear hydrodynamic trails” weregenerated (Fig. 1, A and B). The submarine’sheading was chosen pseudorandomly andcounterbalanced over all trials. Because of its

1Institut fur Zoologie, Universitat Bonn, PoppelsdorferSchloss, D-53115 Bonn, Germany. 2Ruhr-UniversitatBochum, Allgemeine Zoologie und Neurobiologie,D-44780 Bochum, Germany.

*To whom correspondence should be addressed. E-mail: dehnhardt@neurobiologie.ruhr-uni-bochum.de

Fig. 1. (A) Evaluation of the probabilitythat the seal finds the submarine bymere chance. After the submarine’s runin a pseudorandom direction (yellowline), it drifted somewhere on a semi-circle (blue line; circumference, 12.57m). With protracted vibrissae (span;25 cm) and slight lateral head move-ments, the seal should detect an objectin front of its head within a 50-cm span(1/25 length of the semicircle, whitearea). Thus, the seal’s probability offinding the submarine by mere chancewas P # 0.04 (13 out of 326 trials) (13)when swimming an arbitrarily chosensearch path (red line). (B) Single framestaken from video recordings showing atrial with a linear hydrodynamic trail.Yellow arrow, position of the subma-rine; red arrow, position of the seal.Colored lines are as in (A).

R E P O R T S

6 JULY 2001 VOL 293 SCIENCE www.sciencemag.org102

almost constant running distance, the subma-rine stopped somewhere on a semicircle (Fig.1A). After leaving the hoop station, the sealimmediately submerged and headed towardthe center of the pool. Searching for thehydrodynamic trail was characterized by theseal protracting the vibrissae to the most for-ward position and performing slight lateralhead movements. As soon as the seal inter-sected the hydrodynamic trail, the animalturned onto the submarine’s course, thus in-dicating its detection of the trail, and fol-lowed the trail at a swim speed of ;2.0 m s21

(estimated from video recordings). As frame-by-frame analysis of video recordings re-vealed, the seal clung exactly to the subma-rine’s trail. The seal located the submarine in256 of 326 trials (78.5%). Location accuracywas independent of the submarine’s position

on the semicircle. This location performanceis highly significant (binomial distribution,P , 0.001) (Fig. 1A) (15). In 59 of the 70trials during which the seal failed to find thesubmarine, the animal seemed to have missedthe beginning of the trail. In these cases, theseal stopped searching in the starting areaafter a few seconds. In only 11 trials (3.4%),failure to find the submarine was caused bythe seal deviating from the previously cor-rectly tracked hydrodynamic trail.

Increased delays between the start of thesubmarine and the start of the seal’s search didnot affect accuracy of trail detection and pur-suit. In 24 of 30 trials with delays of 10, 15, and20 s (10 trials each) the seal successfully locat-ed the submarine’s final position (three failuresafter a delay of 10 s, one failure after 15 s, andtwo failures after 20 s). After 20 s, aging of the

hydrodynamic trail in the seal’s search areacorresponded to that of a continuously runningsubmarine that already covered a distance of;40 m (swim speed ;2 m s21 for 20 s).

Following a fish certainly requires detec-tion of changes in the course of a trail. There-fore, we introduced an unpredictable changeof course in the submarine’s trail by usingtwo lateral steering propellers (10). The sealspontaneously turned onto the new coursewhen it met the submarine’s turning pointand successfully found the submarine in 26 of30 trials (Fig. 2A). These results unequivo-cally show that the seal’s location of thesubmarine was based on true trail-following.As in an echolocating animal, passive listen-ing or the use of vision should have resultedin a straight approach to the submarine (Fig.2A, yellow dashed line, last frame).

Under natural conditions, a seal mayencounter a hydrodynamic trail not at itsbeginning but must be able to determine theswimming direction of a fish at any givenpoint of the trail. Accordingly, we conduct-ed 50 trials with the submarine runningparallel to the long side of the pool, therebypassing the seal’s hoop station at a distanceof ;1 m. In a pseudorandom sequence, thenumber of trials with the submarine run-ning to the left or to the right was counter-balanced. In 46 trials the seal detected thetrail, sharply turned into the correct direc-tion, and followed the submarine’s trail.

Thirty trials were randomly interspersedin the course of the experiments with a stock-ing mask that also covered the seal’s muzzle.Whisker movements were impeded, but theseal was still able to open its mouth for theperception of potential chemosensory cues.In these trials, the seal started its search asusual but always failed to detect the hydro-dynamic trail even when its head exactly metthe submarine’s starting point. These testsshow that trail-following is only possiblewith unimpaired whiskers and is based on thedetection and analysis of hydrodynamicinformation.

Experiments with hydrodynamic trails con-taining a change of course have been repro-duced with a second experimentally naıve maleharbor seal (“Nick”). These tests were conduct-ed in an outdoor pool at “Zoo Koln” (Germany)filled with clear fresh water. Trail-followingbehavior of this seal corresponded to that de-scribed for the animal tested before (see Fig.2B). The seal located the submarine in 37 out of45 trials (82.2%). However, when the whiskerswere covered by the stocking mask the sealnever detected the trail. With this seal, we alsotested the hypothesis that the use of acousticcues should result in a straight approach to thesubmarine. During 10 trials randomly inter-spersed in the experiments, the seal was al-lowed to start its search right before the engineof the submarine stopped running, thus receiv-

Fig. 2. (A) Sequence of single frames taken from video recordings showing a trial with ahydrodynamic trail containing a sharp change of course. (B) Sequence of single frames showing atrial with the submarine running a right-hand curve. Solid yellow lines indicate hydrodynamic trails.Arrows are as in Fig. 1. Broken yellow lines indicate the short-cuts the seal would have taken if usingacoustic cues.

R E P O R T S

www.sciencemag.org SCIENCE VOL 293 6 JULY 2001 103

ing a short acoustic cue from the submarine’sfinal position. In these trials the seal directlyapproached the submarine’s position in an un-usually fast reaction (see Fig. 2B, last frame).

What might be the detection range of atrail-following seal for prey fish? Even after.3 min, the wake behind a swimming gold-fish contains water velocities that are signif-icantly higher than background noise (9) andexceed the sensitivity threshold of the whis-kers of harbor seals (4). Given that a herringswimming at a sustained speed of ;1 m s21

(16) leaves a hydrodynamic trail just as sta-ble as that of a goldfish, it might be detectablefor a seal even when the herring is more than180 m away. However, for a reliable estimateof the maximum detection range, we need tolearn more about background noise in thewild as well as the aging of fish-generatedtrails under natural conditions.

Because a swimming seal itself produc-es considerable water movements that cer-tainly affect the whiskers, the detection offish-generated water movements wasthought to be hardly possible (17 ). Howev-er, preliminary results from our laboratorysuggest that seals may have overcome thisproblem by a simple mechanism. As a func-tion of swim speed and their biomechanicalproperties, the whiskers of a swimming sealvibrate with characteristic frequencies. Ahydrodynamic trail intersected by the sealwill cause a modulation of this characteristicvibration that might be sensed by the seal.

Our results describe a system for spatialorientation in the aquatic environment thatcan explain successful feeding of pinnipedsin dark and turbid waters. The sensory abilityof hydrodynamic trail-following may be alsoimportant to other species equipped with hy-drodynamic receptor systems.

References and Notes1. W. W. L. Au, The Sonar of Dolphins (Springer, New

York, 1993).2. H. Bleckmann, Reception of Hydrodynamic Stimuli in

Aquatic and Semiaquatic Animals (Fischer-Verlag,Stuttgart, Jena, New York, 1994).

3. S. Coombs, P. Gorner, H. Munz, The MechanosensoryLateral Line. Neurobiology and Evolution (Springer,New York, 1989).

4. G. Dehnhardt, B. Mauck, H. Bleckmann, Nature 394,235 (1998).

5. M. J. Weissburg, M. H. Doall, J. Yen, Philos. Trans. R.Soc. London B 353, 701 (1998).

6. R. W. Davis et al., Science 283, 993 (1999).7. H. Bleckmann, T. Breithaupt, R. Blickhahn, J. Tautz,

J. Comp. Physiol. A 168, 749 (1991).8. R. Blickhan, C. Krick, D. Zehren, W. Nachtigall, Natur-

wissenschaften 79, 220 (1992).9. W. Hanke, C. Brucker, H. Bleckmann, J. Exp. Biol. 203,

1193 (2000).10. For the generation of linear hydrodynamic trails, the

independently moving submarine was powered by asingle propeller only. Two lateral steering propellersallowed the generation of curved trails. The submarinewas always started with the steering propellers in thevertical plane, but rotated once around its longitudinalaxis while running. Depending on the submarine’s list,four inclination contacts switched on the steering pro-pellers when these were in the almost horizontal planeafter some meters of straight run. The new course was

not predictable, but if the submarine rotated fast, itchanged its course sharply (Fig. 2A), if the rotation wasslowly the resulting course was a left-hand or right-hand curve (Fig. 2B). The speed of the submarine was;2 m s21 and ;1.5 m s21, for linear and curved trails,respectively. Experiments were recorded by a cam-corder installed ;5 m above the pool. Video recordingswere analyzed off-line frame by frame. The straight partof the hydrodynamic trail was described by measuringthe direction and velocity of the particle movements(PIV ) (9). A thin horizontal layer of laser light (l 5 650nm) was laid about 30 cm below the water surface. ACCD-camera was installed 1 m from the edge of thepool, and neutrally buoyant seeding particles (Vetosint1101, Huls AG, Germany) were put into the water. Thesubmarine was started in the depth of the laser plane ata distance of about 3 m. Several runs were performedwith the submarine passing the camera in increasinglateral distances. Before each measurement, the directcurrent (DC) components of the water velocities werebelow 10 mm s21; velocities of about 5 mm s21 weretypical. Particle movements were analyzed manually(Scion Image; Scion, Frederick, MD) or with a cross-correlation technique (MatLab; MathWorks, Natick, MA)

in order to describe direction, form and velocity of thewater flow.

11. Web fig. 1 is available on Science Online at www.sciencemag.org/cgi/content/full/293/5527/102/DC1.

12. R. J. Adrian, Annu. Rev. Fluid Mech. 23, 261 (1991).13. Web fig. 2 is available on Science Online (11).14. U. K. Muller, B. L. E. Van Den Heuvel, E. J. Stamhuis,

J. J. Videler, J. Exp. Biol. 200, 2893 (1997).15. W. J. Bell, Searching Behaviour (Chapman & Hall,

London, 1991).16. J. J. Videler, Ed., Fish Swimming (Chapman & Hall,

London, 1993), pp. 1–22.17. D. H. Levenson, R. J. Schusterman, Mar. Mamm. Sci.

15, 1303 (1999).18. The experimental animals were treated in accord

with the official German regulations for research onanimals. We thank D. Adelung, G. Nogge, S. Prange,and I. Robbecke for their support during this study.Financed by grants of the Deutsche Forschungsge-meinschaft (G.D.).

8 March 2001; accepted 24 May 2001

Human Chromosome 19 andRelated Regions in Mouse:

Conservative andLineage-Specific EvolutionParamvir Dehal,1,2,4 Paul Predki,1,5 Anne S. Olsen,1,2

Art Kobayashi,1,2 Peg Folta,1,2 Susan Lucas,1,2 Miriam Land,1,8

Astrid Terry,1,2 Carol L. Ecale Zhou,1,2 Sam Rash,1,2

Qing Zhang,1,5 Laurie Gordon,1,2 Joomyeong Kim,1,2

Christopher Elkin,1,3 Martin J. Pollard,1,6 Paul Richardson,1,5

Dan Rokhsar,1,7 Ed Uberbacher,1,8 Trevor Hawkins,1,5

Elbert Branscomb,1,2 Lisa Stubbs1,2*

To illuminate the function and evolutionary history of both genomes, wesequenced mouse DNA related to human chromosome 19. Comparative se-quence alignments yielded confirmatory evidence for hypothetical genes andidentified exons, regulatory elements, and candidate genes that were missed byother predictive methods. Chromosome-wide comparisons revealed a differ-ence between single-copy HSA19 genes, which are overwhelmingly conservedin mouse, and genes residing in tandem familial clusters, which differ exten-sively in number, coding capacity, and organization between the two species.Finally, we sequenced breakpoints of all 15 evolutionary rearrangements, pro-viding a view of the forces that drive chromosome evolution in mammals.

Spanning 65 to 70 Mb and estimated to con-tain 1100 genes, human chromosome 19(HSA19) is one of the smallest and most

gene-dense of human chromosomes (1, 2). Aclone-based physical map spanning all butthe centromeric regions of the chromosomewith seven gaps (3) has provided the frame-work for HSA19 sequence, which to dateincludes 35 Mb of finished sequence and 22Mb of high-quality draft. The solidly an-chored clone framework and high percentageof ordered and oriented contigs generatedthrough application of a plasmid paired-endsequencing strategy (4) have rendered unfin-ished portions of HSA19 draft sequence par-ticularly amenable to annotation and analysis.Comparing human DNA sequence with thatof other species has proved to be an especial-

1DOE Joint Genome Institute, Walnut Creek, CA94598, USA. 2Genomics and 3Engineering Divisions,Lawrence Livermore National Laboratory, Livermore,CA 94550, USA. 4Department of Genetics, Universityof California Davis, Davis, CA 95616, USA. 5Genomicsand 6Engineering Divisions, Lawrence Berkeley Na-tional Laboratory, Berkeley, CA 94720, USA. 7PhysicsDepartment, University of California Berkeley, Berke-ley, CA 94720, USA. 8Life Sciences Division, Oak RidgeNational Laboratory, Oak Ridge, TN 37831, USA.

*To whom correspondence should be addressed. E-mail: stubbs5@llnl.gov

R E P O R T S

6 JULY 2001 VOL 293 SCIENCE www.sciencemag.org104