SYMPOSIUM

How Nemo Finds Home: The Neuroecology of Dispersal and ofPopulation Connectivity in Larvae of Marine FishesJeffrey M. Leis,1,* Ulrike Siebeck† and Danielle L. Dixson‡

*Ichthyology, Australian Museum, 6 College St, Sydney, NSW 2010, Australia; †ARC Centre of Excellence in Vision

Research, Sensory Neurobiology Group, School of Biomedical Science, University of Queensland, St Lucia, QLD 4072,

Australia; ‡ARC Centre of Excellence for Coral Reef Studies, School of Marine and Tropical Biology, James Cook

University, Townsville, QLD 4811, Australia

From the symposium ‘‘Neuroecology: Neural Determinants of Ecological Processes from Individuals to Ecosystems’’

presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3–7, 2011, at Salt Lake

City, Utah.

1E-mail: jeff.leis@austmus.gov.au

Synopsis Nearly all demersal teleost marine fishes have pelagic larval stages lasting from several days to several weeks,

during which time they are subject to dispersal. Fish larvae have considerable swimming abilities, and swim in an

oriented manner in the sea. Thus, they can influence their dispersal and thereby, the connectivity of their populations.

However, the sensory cues marine fish larvae use for orientation in the pelagic environment remain unclear. We review

current understanding of these cues and how sensory abilities of larvae develop and are used to achieve orientation with

particular emphasis on coral-reef fishes. The use of sound is best understood; it travels well underwater with little

attenuation, and is current-independent but location-dependent, so species that primarily utilize sound for orientation

will have location-dependent orientation. Larvae of many species and families can hear over a range of �100–1000 Hz,

and can distinguish among sounds. They can localize sources of sounds, but the means by which they do so is unclear.

Larvae can hear during much of their pelagic larval phase, and ontogenetically, hearing sensitivity, and frequency range

improve dramatically. Species differ in sensitivity to sound and in the rate of improvement in hearing during ontogeny.

Due to large differences among-species within families, no significant differences in hearing sensitivity among families

have been identified. Thus, distances over which larvae can detect a given sound vary among species and greatly increase

ontogenetically. Olfactory cues are current-dependent and location-dependent, so species that primarily utilize olfactory

cues will have location-dependent orientation, but must be able to swim upstream to locate sources of odor. Larvae can

detect odors (e.g., predators, conspecifics), during most of their pelagic phase, and at least on small scales, can localize

sources of odors in shallow water, although whether they can do this in pelagic environments is unknown. Little is known

of the ontogeny of olfactory ability or the range over which larvae can localize sources of odors. Imprinting on an odor

has been shown in one species of reef-fish. Celestial cues are current- and location-independent, so species that primarily

utilize them will have location-independent orientation that can apply over broad scales. Use of sun compass or polarized

light for orientation by fish larvae is implied by some behaviors, but has not been proven. Use of neither magnetic fields

nor direction of waves for orientation has been shown in marine fish larvae. We highlight research priorities in this area.

Introduction

Nemo, the cartoon clownfish had a problem in

common with most coral reef fishes—finding

home—but in the real ocean, it is the larval

life-history stage that is adrift in the big blue

ocean, and subject to dispersal. The challenges

larvae face in finding a reef home are greater than

those overcome by Nemo, and the consequences of

getting it wrong, more severe. For example, although

Sydney Harbour lacks coral reefs, larvae of coral-reef

fishes do settle there each summer, but unlike Nemo,

few survive the experience. How larvae avoid such

errors is still largely unknown, but swimming and

orientation abilities are part of the answer.

Integrative and Comparative Biology, volume 51, number 5, pp. 826–843

doi:10.1093/icb/icr004

Advanced Access publication May 11, 2011

� The Author 2011. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.

For permissions please email: journals.permissions@oup.com.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

Orientation in the pelagic environment is a chal-

lenge for any animal. The water column within

which the animal lives is itself moving, this move-

ment often varies with depth, and because of the

paucity of detectable reference points external to

the moving water, such movement is difficult for

the animal to detect. The pelagic environment is rel-

atively featureless and many cues that are useful for

orientation above water are diminished or modified

by seawater. These difficulties increase for larval

stages of marine animals that as adults live in close

association with the bottom (i.e., are demersal) due

to the very small size of the larvae and incomplete

development of sensory organs and swimming struc-

tures. Yet, the larvae must detect and move to de-

mersal settlement habitat from the pelagic

environment if they are to complete their life

cycles; there is no opportunity to learn from parents

or experience—each larva has only one chance to

settle.

Until the late 1990s, it was generally assumed that

larval dispersal of marine fishes is essentially a phys-

ical process (Roberts 1997), and that the tiny larvae

simply go where currents take them. Their behavioral

and sensory abilities were considered too feeble to

significantly alter dispersal outcomes (Williams

et al. 1984); i.e., they were regarded as passive plank-

ton. As a corollary, dispersal and population connec-

tivity would operate over very large spatial scales.

This traditional paradigm has been overturned by

demonstrations that larvae of warm-water marine

fishes have well-developed behavioral and sensory

abilities (reviewed by Leis 2006), and that dispersal

often operates at smaller scales than expected from

physical processes alone (Jones et al. 2009; Leis et al.

submitted for publication). Yet, much remains un-

known, especially about abilities of larvae to orient

during their pelagic stage and the sensory cues they

use (Arvedlund and Kavanagh 2009). At the end of

the pelagic stage, larvae must find appropriate habi-

tat, and make a dangerous ecological transition from

pelagic to demersal habitat (Kaufman et al. 1992).

Orientation behavior is thus intimately involved in

both dispersal and settlement (Montgomery et al.

2001; Montgomery et al. 2006). We now know fish

larvae are far from passive drifters. A wide taxonom-

ic range of larval fishes can swim at speeds similar to

the currents they occupy, and can do so for extended

periods and for most of their pelagic larval duration

(PLD) (Leis 2006; Fisher and Leis 2009; Leis 2010).

We also know that, somehow, these larvae can orient

within moving ocean waters so that they are not

simply swimming randomly. Research has concen-

trated on discovering and quantifying the abilities

of larvae, and the sensory cues that larvae use to

orient remain poorly understood (Kingsford et al.

2002; Arvedlund and Kavanagh 2009). We are just

beginning to understand what cues and what senses

larvae use to achieve the orientation documented in

laboratory and field studies. Further, how these sen-

sory systems operate to detect and use the cue is not

always entirely clear. So, when considering how

swimming abilities of larvae are applied, we need

to ask; what sensory abilities are brought to bear

by the larvae, how they are used, how they change

ontogenetically, and how they work. Nearly every

aspect of the behavior of fish larvae examined so

far has produced surprising evidence of the sophis-

tication and range of abilities, revolutionizing our

view of what larval fishes can do (Leis 2006), but

how dispersing larvae determine the direction they

will swim (orientation), how this varies spatially and

what cues they use, are all largely unknown.

The pelagic larval stage lasting from days to

months, found in almost all marine teleost fishes is

the single life-history trait that most distinguishes

marine fishes from terrestrial vertebrates. Larvae are

subject to dispersal (Fuiman and Werner 2002), and

in fish species that are demersal as adults, dispersal

of larvae typically sets the spatial scale of population

connectivity (Cowen and Sponaugle 2009). Larvae

and adults of marine fishes differ in appearance, hab-

itat, diet, and the challenges they face (Leis and

McCormick 2002). Not only must larvae survive

and grow from a few millimeters to a few centime-

ters, they must also develop morphologically and be-

haviorally. Behaviors of larvae enable survival in the

pelagic environment and facilitate arrival at appro-

priate locations during a narrow interval of time

when larvae are capable of settlement.

A major challenge in marine ecology today is de-

lineating the spatial scale of population connectivity,

and for coral-reef fishes this is largely determined by

dispersal of larvae (Cowen et al. 2007; Cowen and

Sponaugle 2009). Why is connectivity significant?

First, knowing the scale of connectivity and how it

is determined is essential to comprehending the key

issues of how populations of marine animals with

complex life cycles operate both demographically

and genetically. Second, the spatial scale of connec-

tivity is the natural geographic scale of population

units, and hence the scale over which fisheries must

be managed. Third, knowing the spatial scale of pop-

ulation connectivity is required for informed design

and management of marine protected areas. In short,

managers of living marine resources must know

about connectivity. Knowing how larvae orient is a

prerequisite for understanding, measuring and

How Nemo finds home 827

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

predicting demographic (ecological) connectivity,

and may also be relevant to genetic (evolutionary)

connectivity (Leis et al. submitted for publication).

Yet, how fish larvae orient and the sensory cues they

use for directing their considerable swimming abili-

ties constitute the largest gap in our knowledge of

the behavior of larvae.

Neuroecology synthesizes neuroethological and

ecological principles, including both the neural

basis for behavior and the role of that behavior in

establishing population dynamics, community struc-

ture and ecosystem function (Zimmer and Derby

2007). Dispersal of larvae in the sea integrates over

days to months swimming abilities and physiology,

sensory abilities, feeding (and associated growth and

survival), and the ontogeny of all of these, in com-

bination with physical oceanography. The outcomes

of dispersal directly impact both demographic and

genetic population connectivity, and affects species

composition and interactions in demersal ecosys-

tems. In this overlap of disciplines, dispersal of

larvae is a good example of what neuroecology

seeks to address.

We assess current knowledge of the ontogeny of

sensory abilities in larval fishes, with emphasis on

those used for orientation that is relevant to dispers-

al: sound, sight and smell. We also aim to point out

particular gaps and to stimulate further research.

Our focus is on teleost fishes that live as adults in

close association with the bottom with emphasis

on coral reefs. The patterns of orientation pro-

vide strong clues about the sensory systems and

cues used by larvae. During the day, larvae of

many pomacentridae damselfishes have location-

independent orientation (Leis and Carson-Ewart

2003; Leis et al. 2007), which implies the use of

location-independent stimuli, such as celestial cues

or the earth’s magnetic field. In contrast, orientation

by larvae of other species is location-dependent (Leis

and Carson-Ewart 2003), implying the use of

location-dependent cues, such as sound or dissolved

chemicals emanating from a reef or shoreline, or

wave direction.

Sound and hearing

Jacques Cousteau called his famous 1953 book about

life underwater ‘‘the Silent World’’, but the title re-

flects the limitations of the human ear underwater

rather than reality. The sea is actually a very noisy

place because sound travels so well under water

(faster and with less attenuation than in air) and

there are many sources of sound ranging from break-

ing waves to singing whales. In particular, many

underwater habitats such as reefs are the source of

a wide variety of physically and biologically generat-

ed sounds from infrasound to ultrasound; these

sounds are location-dependent. Further, the spread

of sound is essentially independent of currents, so

it spreads in all directions from a source. It is there-

fore an ideal cue for any animal capable of detecting

it. The ‘‘soundscapes’’ associated with different hab-

itats are significantly different (Kennedy et al. 2010;

Radford et al. 2010a) which opens the possibility that

larvae may be able to detect these differences and

find specific habitats using sound. Not only do un-

derwater soundscapes vary spatially, they also vary

temporally, with strong diel, lunar and seasonal dif-

ferences in quantity and quality of sound (Cato 1978;

Cato 1992; McCauley and Cato 2000). These tempo-

ral variations may provide larvae with information

on the appropriate time to approach or avoid a

site for settlement, but they may also make it

harder for a larva to detect it.

Fish have well-developed ears, although there is

little external manifestation of them. Each fish ear

contains three otoliths (ear stones) in contact with

pads of sensory hair cells. The otolith is primarily

calcium carbonate, so is much denser than the rest

of the fish. Hence, when sound induces movement,

the fish itself moves differently from the otolith, and

the hair cells are stimulated. These, in turn stimulate

the auditory nerve. Larval fishes have otoliths within

a day or two of hatching, and as detailed below, they

can hear for much, if not all, of their PLD. This has

been determined by several approaches: Observation

of behavior of larvae in the field, use of electrophys-

iological techniques in the laboratory, and measure-

ment of responses of heart-rate to sound in the

laboratory. A detailed examination of hearing in

fishes is presented by Popper et al. (2003) and

Higgs et al. (2006), but it is worth noting the

caveat of Montgomery et al. (2006) that there may

be undiscovered means of hearing.

There is no consensus on the role of the gas-filled

swim bladder in so-called ‘‘hearing generalists’’, i.e.,

fish without a mechanical connection between the

swim bladder and the ear (Popper et al. 2003).

Such species constitute the large majority of demersal

teleost fishes. Sound pressure waves cause the swim-

bladder to vibrate and thus induce particle motion.

Some investigators argue such motion is detected by

the combination of otolith and hair cells in the ear,

whereas others point out that experimental deflation

of the swim bladder does not always lead to a de-

crease in sensitivity, contrary to expectations if the

swim bladder were involved in hearing (Yan et al.

2000; Lugi et al. 2003).

828 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

The lateral-line system is generally considered not

to be involved in hearing; rather, it detects vibrations

in the ‘‘near-field’’, or within a few body lengths of

the fish (Popper et al. 2003). However, if

sound-induced vibration of the swim bladder is de-

tectable by the ear, it may also be detectable by the

hair cells of the lateral line, even without a latero-

physic connection, a possibility that has not yet been

evaluated.

In a field experiment, settlement-stage pomacen-

tridae larvae changed their behavior in the presence

of broadcast reef sounds, and behaved differently

when reef sounds were broadcast than when a mix

of pure tones (a white noise) was broadcast (Leis

et al. 2002). This study confirmed that fish larvae

could hear and that they could distinguish a

biologically-generated sound with ecological signifi-

cance from an ecologically meaningless, artificial

sound. In fact, the larvae seemed to ignore the

white noise, although it was within the frequency

range that fish can detect.

The electrophysiological technique, auditory brain-

stem response (ABR), can determine what frequen-

cies and intensities of sound are detected by fishes,

and results are typically shown as ‘‘audiograms’’

(Higgs et al. 2006). For each frequency, audiograms

depict the lowest intensity (in decibels) that provides

an ABR response from a fish under test (for brevity,

we use decibels here, but all values are in fact deci-

bels re 1 mPa as is the convention for underwater

sound). Audiograms for larval marine fishes have

now been published for at least 12 demersal species

from six families of perciform fishes (Egner and

Mann 2005; Wright et al. 2005, 2008, 2010, 2011).

Typically, these show greatest sensitivity at low fre-

quencies (100 Hz), a second area of increased sensi-

tivity at 500–700 Hz, and a substantial decrease in

sensitivity at above 1000 Hz. So, the most relevant

sounds for larval fishes are apparently those between

100 and 1000 Hz. Sensitivity differed significantly

among larvae of different species; within families,

reef-associated species had more sensitive hearing

than did nonreef species (Wright et al. 2011).

Audiograms from ABR usually depict hearing that

is 10–30 dB less sensitive than hearing measured by

behavioral means, usually conditioning (Gorga et al.

1988; Kojima et al. 2005). The reasons for this ‘‘ABR

gap’’ are not entirely clear. It could be related to the

fact that ABR uses very short bursts of pure tones,

whereas behavioral conditioning tends to use sounds

of longer duration (D. Cato, personal communica-

tion), or it is possible that ABR is insufficiently sen-

sitive to detect very small nerve impulses (D. Higgs,

personal communication). Montgomery et al. (2006)

suggest the reason for the ABR gap is that ‘‘electro-

physiology typically relies on the response properties

of single nerve fibres, whereas the central nervous

system has access to a population of inputs and

can improve signal detection through ensemble av-

eraging’’. Regardless of the reason, it may be neces-

sary to ‘‘correct’’ ABR data by 10–30 dB, although

the lack of behaviorally-derived audiograms for

larvae of any marine fish makes determining the ap-

propriate magnitude of the correction problematical.

Further, any correction may vary among species or

ontogenetic stages. No behavior-based audiograms

for fish larvae exist, because by the time one can

train a fish larva via conventional conditioning tech-

niques, it will likely no longer be a larva, but rather a

metamorphosed juvenile. This also explains the em-

phasis on electrophysiological methods.

The measurements of hearing ability discussed so

far, were conducted on larvae ready to settle i.e.,

larvae of �15–25 mm in standard length (SL). It is

important to know when larvae can hear and how

hearing ability changes during development, yet mea-

suring hearing in smaller larvae is particularly diffi-

cult. To date, only one study has examined hearing

in larvae of marine demersal species smaller (and

younger) than settlement stage. Use of ABR technol-

ogy showed that larvae as small as could be tested

(ca 9 mm) responded to sound, and their audio-

grams were of similar shape to those of conspecific

settlement-stage larvae (Wright et al. 2011). As ex-

pected, hearing of smaller larvae was less sensitive

than that of larger larvae, by as much as 25 dB,

and in some species, the smaller larvae could not

hear the higher frequencies tested. For example, in

the serranid Epinephelus coioides, the maximum fre-

quency that elicited an ABR response increased from

600 Hz in larvae 9–13 mm SL to 2000 Hz in larvae

of 24–28 mm SL. Hearing sensitivity for 100 Hz

increased at a rate of 0.6–3.2 dB/mm increase in

size (mean¼ 1.48, SE¼ 0.59) in the four marine spe-

cies tested (families Carangidae, Polynemidae and

Serranidae). This unique study provides the only in-

formation on the development of hearing abilities in

larvae of marine fishes.

Using a different approach, Simpson et al. (2005b)

demonstrated a response to sound by embryos inside

the demersal eggs of a damselfish (Pomacentridae).

They noted an increase in heart rate when sounds

were played to the embryos. Whether this is hearing

in the normal sense is questionable, but it is clear

that sounds were detected by the embryos. This

result implies that fish larvae may be able to hear

for the entire PLD, and possibly before they hatch.

The latter opens up the possibility of acoustic

How Nemo finds home 829

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

imprinting—settlement-stage larvae may learn, and

later home in on, a natal acoustic signature in

much the same way as salmon home on an olfactory

signature of a natal stream. This possibility is

strengthened by the finding that settlement-stage

larvae of reef-fish are influenced by, and can retain

information from, recent acoustic experiences

(Simpson et al. 2010).

It is clear that fish larvae of several families of

tropical, perciform fishes can hear, but ability to

hear is only part of the story. To be detectable, a

sound must not only be within a certain frequency

range and of a minimum strength, but it must also

be louder than background ‘‘noise’’ within that fre-

quency range, which underwater can be considerable.

Many considerations of hearing in larval fishes

ignore this fact. Further, the ability to hear a

sound does not necessarily mean that a larva will

respond to the sound. There are likely to be ontoge-

netic factors involved in whether a larva responds

and, as detailed below, there are also temporal fac-

tors (perhaps lunar or diel) that influence the re-

sponse both qualitatively and quantitatively.

Field experiments have demonstrated that

settlement-stage larvae can use hearing for orienta-

tion and can find the source of the sound (localiza-

tion) (reviewed by Montgomery et al. 2006;

Arvedlund and Kavanagh 2009). The evidence

comes from three types of field approaches: (1)

Light-traps that broadcast sounds, (2) patch reefs

that broadcast sounds, and (3) binary-choice

chambers.

The first typically involves comparing catches of

larvae at night from light-traps with and without

speakers broadcasting what are thought to be

biologically-relevant sounds (usually recordings of

reef noises). In most cases and for larvae of most

demersal taxa tested, the ‘‘noisy’’ light traps captured

more larvae than did the silent ones (Tolimieri et al.

2000; Leis et al. 2003; Simpson et al. 2004), demon-

strating an attraction to sound, although alternative

interpretations of the results have been offered (e.g.,

Higgs 2005). These studies demonstrated the attrac-

tiveness of reef sounds to a wide range of tropical

species, and to a temperate species of tripterygiid

blenny, again mostly Perciform taxa, but including

some Tetradontiform species. Montgomery et al.

(2006) noted that although these field experiments

demonstrated an attraction to sound, they did not

determine if the larvae are able to sort out the 180o

ambiguity inherent in the problem of localizing a

source of sound. That is, half the larvae may have

swum away from the sound source, even if the other

half swam toward it and contributed to the higher

catches. Papers by Simpson et al. (2005a, 2008) used

a similar protocol with similar results, but instead of

light-traps, used patch reefs upon which larvae can

settle. This approach disposes of Higgs’ alternative

interpretation of the results from light-traps as no

lights were involved. In contrast, one light-trap

study found that larvae of six families of reef

fishes, most of which had been attracted by reef

sounds in previous work, were captured in larger

numbers by the silent light-traps (Heenan et al.

2008). This study confirms that settlement-stage

larvae can hear, but the authors could only speculate

about why larvae might have avoided the reef sounds

they broadcast. Clearly, we do not fully understand

the role of reef sounds in the settlement of reef-fish

larvae.

The actual frequencies to which the larvae respond

are unclear. One of the patch-reef studies (Simpson

et al. 2008) contrasted the attractiveness of

‘‘high-frequency’’ reef sounds (570–2000 Hz) to

‘‘low-frequency’’ reef sounds (5570 Hz), and found

most reef-fish families tested settled in higher num-

bers with high-frequency sounds. Combined with re-

sults of the ABR studies that showed larval-fish had

poor sensitivity to frequencies 41000 Hz, these re-

sults indicate that sounds between 570 and 1000 Hz

are likely to be the most attractive to settlement-stage

fish larvae.

A final class of studies, in contrast, clearly dem-

onstrated that larvae can resolve the 180o ambiguity

and truly localize the direction to a sound source,

and swim toward it. These studies utilize binary-

choice chambers, a linear arrangement of three seg-

ments with constricted return passages between

them. One end of the chamber points toward a

sound source (a reef or a speaker), while the other

end points away from the source. Larvae are intro-

duced into the central segment, and the chamber is

subsequently checked to see whether larvae have

moved from the central segment, and in which di-

rection. The initial study of this sort used a reef as

the ‘‘test target’’ and found that larvae of the two

tested families (Pomacentridae and Apogonidae)

swam toward the reef at night (Stobutzki and

Bellwood 1998). Stobutzki and Bellwood (1998)

argued that sound was the reef-based cue likely to

have been responsible for their results, but could not

demonstrate this as the reef was the source of other

cues as well. Subsequent work with binary-choice

chambers used a speaker to broadcast reef sounds,

and the taxa used were tripterygiids or pomacentrids

(Tolimieri et al. 2002; Tolimieri et al. 2004; Leis and

Lockett 2005). All three studies found that larvae

moved toward the sound at night, in contrast to

830 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

daytime, when larvae either moved away from the

sound or had no preference, thus clearly demonstrat-

ing the localization of underwater sound, and a diel

change in sound-related behavior. However, not all

pomacentridae species made a significant choice at

night in the study that looked at individual species

(Leis and Lockett 2005)—the other studies analyzed

results at the familial level. Unfortunately, no exper-

iments have attempted to assess ontogenetic influ-

ences on localization of sound, and it is likely that

both the ability to respond to sound and the moti-

vation to do so have a strong ontogenetic

component.

These field experiments show clearly that larval

fishes, like adult fishes, can localize underwater

sound sources, but how larvae, or even adult fish,

do this is unclear, although several possibilities

have been put forward. As noted by Higgs et al.

(2006), the acoustic cues used by terrestrial verte-

brates for localization of sound (acoustic differences

in timing and intensity, and pinna filtering) do not

apply to fishes. This is because the small distance

between the left and right ears in fishes (particularly

larvae) and the very high speed of sound underwater,

result in minimal binaural differences in fishes; also

fishes lack external ears for filtering by the pinna.

The most often mentioned means by which fish

larvae might be able to localize a sound source in-

volves the differential orientation of the pads of hair

cells in the ear—these are inherently directional in

their response to otolith movement, so, if they are

orientated differently, there is potential for detecting

direction of a sound source. Unfortunately, when

does the hair-cell orientation develop in larvae of

marine fishes is unknown (Higgs 2005). A study of

the structural development of the ear of larval

marine fishes may provide important clues about

the means and the ontogeny of the ability to localize

sound. For example, Gagliano et al. (2008) reported

evidence suggesting that pomacentridae larvae with

left–right asymmetry in otolith shape had lower sur-

vival and greater difficulty in detecting suitable hab-

itats for settlement than did larvae with symmetrical

otoliths, presumably because they would not have

been as able to localize reef sounds. Some models

for sound localization involve comparing sensory

inputs from the particle motion (assumed to be the

salient cue for otolithic hearing) and the pressure

induced by sound waves moving through water

(Higgs et al. 2006). It is frequently argued that the

gas bladder found in most fishes (including the

larvae of taxa in which it is absent in adults, e.g.,

tripterygiids and gobiids) is involved, particularly in

the context of converting pressure to particle motion,

and that fish can resolve the previously mentioned

180o ambiguity by comparing sound reradiated from

the swim bladder with sound directly reaching the

inner ear. If the swim bladder is involved in hearing,

sensitivity should decrease if the swim bladder is de-

flated. Yet, the expected decrease does not occur in

all fish species (Yan et al. 2000; Lugi et al. 2003),

raising questions about the role of the swim bladder

in hearing. An alternative model for localization in-

volves comparing inputs from the left and right ears

via interconnections of nerves in the dorsal zone of

the descending octaval nucleus of the brain

(Edds-Walton and Fay 2002). In part, this model

relies on the fact that the auditory organs on the

left side of the head are not parallel to those on

the right side. In contrast, Sand (2002) argued that

a fish need not be able to localize the source of a

sound from a distance, but instead, can be ‘‘guided’’

to it if the fish can maintain a constant angle be-

tween its body axis and the axis of the particle

motion. Opportunities for research in this area are

many.

A very important issue when considering the role

that behavior of larvae may play in dispersal is the

distance over which a larva can detect a sound cue.

This has been the subject of some debate. Most at-

tempts to answer this question have been theoretical

rather than experimental, and based on the assump-

tions that sound emanates from a point source and

that it spreads in an idealized way, either spherically

or cylindrically. These attempts have used ABR au-

diograms which, as noted above, is made problem-

atical by the ‘‘behavioral correction’’ issue. Egner and

Mann (2005) used ABR data from postsettlement

pomacentridae juveniles and assumed spherical

spreading of sound to estimate that a reef in the

Great Barrier Reef could be heard by larvae from

no more than 1 km. Mann et al. (2007) extended

the argument and concluded, based on the hearing

ability of an adult eleotrid that detection of reef

sounds by larvae 41 km from the reef was unlikely.

These attempts used uncorrected ABR data in their

calculations. Wright et al. (2005, 2008, 2010) took a

similar approach, but attempted to correct the ABR

data, and concluded that pomacentridae, serranid

and lutjanid larvae would be capable of detecting

sounds from a reef 5–6 km away. However, all

these attempts made the unrealistic assumptions

that a reef is a point source for sound, and that

sound spreads spherically. By examining the ratio

of sound levels at varying distances from the

source expected under spherical and cylindrical

models of sound spreading, Wright et al. (2011),

avoided many problematical assumptions and

How Nemo finds home 831

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

concluded that detection ranges increased ontogenet-

ically 1.5–10 times with spherical spreading, and an

impressive 2.4–100 times assuming cylindrical

spreading. Reefs, however, actually constitute a wall

of sound, rather than a point source, and because of

this, sound levels remain relatively constant within a

radius about equal to the length of the reef wall of

sound, after which cylindrical spreading seems to

apply, at least in relatively shallow water (Radford

et al. 2010b; D. Cato, personal communication).

This means that reef sounds should be detectable

by larvae from greater distances than had been

thought previously, and that large reefs should be

detectable from a greater range than small ones. It

also means that attempts to calculate the distance

from which sound from a reef can be detected, or

to depict changes in sound strength with distance

from a point source like a speaker (Leis et al.

2002), are unlikely to be relevant to sound emanat-

ing from a reef wall of sound.

The oceans are becoming increasingly noisy from

anthropogenic sound (Malakoff 2010). This increase

in background noise must make it increasingly diffi-

cult for marine animals to detect sounds generated

by conspecifics, or environmental sounds used for

orientation, because such sound cues may be insuf-

ficiently above background to be detected. Such

masking has until now been primarily of concern

for communication between marine mammals

(Clark et al. 2009) but, if fish larvae do use sound

to find settlement habitat, anthropogenic noise pol-

lution may potentially reduce recruitment of larvae

to reefs (Simpson et al. 2005a, 2008). Similarly, if

biogenic sound from a reef is used for orientation

by larvae, such sounds may be greatly decreased by

overfishing, coral-bleaching or the other means

whereby humans are damaging reef ecosystems.

This, too, has the potential to decrease recruitment

to reefs. On the positive side, it may be possible for

managers to broadcast reef sound to attract larvae to

damaged reefs (Simpson et al. 2008).

We have a greater empirical understanding of the

hearing abilities of larval fishes in a neuroecological

context than for any other sensory modality, yet

there are still huge gaps. To date, research, especially

field work, has concentrated on one family of reef

fishes—the Pomacentridae. We have little idea of

how hearing abilities of larvae in other families, let

alone orders other than Perciformes, compare. The

large majority of the field work on hearing in fish

larvae has been done at one location: Lizard Island

on the Great Barrier Reef. Is this location represen-

tative of what happens elsewhere? We do not fully

comprehend how fishes (adult or larval) localize

sources of sound, and this impedes our ability to

understand if, or how, larvae use sound to find

sites for settlement. No field experiments have

tested whether the intensity of sound detected by a

fish larva will influence its orientation. For example,

is a larva that is near its limits of acoustic detection

for reef sounds as likely to swim toward that reef as a

larva located much closer to a reef? Are there unsus-

pected sound receptors or transmitters that allow

larvae to use sound in ways we currently do not

think possible?

In the 1990s, when one of us first began to

wonder if fish larvae could use sounds from reefs

to find appropriate sites for settlement, he was ad-

vised by a physiologist to read a text book on fishes

in order to find out why this would not be possible.

Fish larvae continue to surprise us with their capa-

bilities and the way they do things. We should not

let theoretical arguments about the function of sen-

sory systems we do not fully understand deter us

from directly investigating just what these little

fishes can do. It is clear that we do not yet fully

understand how hearing works in fishes, and there

may be surprises in store as we try to find out. Nor

should we restrict our attention to fishes.

Crustaceans can respond to sounds (e.g., Lovell

et al. 2005) and larvae of marine crustaceans react

to sound in ways that indicate they, too, may be able

to use sound as a sensory cue for orientation

(Stanley et al. 2010).

Chemical signals and olfaction

Marine waters vary strongly in chemical composition

on both large and small spatial scales. Multiple

sources of chemical cues potentially used for naviga-

tional information by larvae exist in the seemingly

featureless pelagic environment (Kingsford et al.

2002). Controlled by physics, chemistry and biology,

chemical signals have a restricted lifetime. Once re-

leased into the environment, signals decline below

detectable levels due to mixing, diffusion, absorption,

photolysis, chemical transformation, and their uptake

and breakdown by organisms (Atema 1995). Thus,

olfactory cues are location-dependent. Temporal

scales tend to follow spatial scales, with small mole-

cules dissipating quickly as a result of molecular dif-

fusion, while larger molecules or compounds released

from a larger source can remain for days, weeks, or

months (e.g., an oil slick from a whale carcass, ter-

restrial runoff during a wet season) (Atema 1995).

This has been demonstrated in field studies in which

fish larvae in a current chose upstream sites for

832 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

settlement and ignored downstream sites (Elliott

et al. 1995; Lecchini et al. 2005).

Longevity and temporal patterns of signals are im-

portant aspects in use of chemical signals by larval

fishes. The speed at which larvae can recognize sig-

nals and process information depends on the fre-

quency of encounter with the patches of chemicals

and the concentration of the chemical cue

(Vergassola et al. 2007). In order to detect chemical

signals, larval olfactory receptors must recognize

them against the background of other chemical com-

pounds present in the water column. Recognition of

signals is a complex process requiring larvae to have

not only specificity and diversity in receptors, but

also ability to recognize intensity and time-course

of signals being detected; this would allow a receptor

to distinguish between a true signal and background

molecules (Atema 1995). The receptor organ deter-

mines, which chemicals will be detected and pro-

cessed as useful information, compared to those

ignored as background odors. As a result, the same

suite of dissolved chemicals can provide different in-

formation for different species due to differences in

the fine-tuning of the olfactory receptors.

In larval fishes, the olfactory system functions by

binding the odorant substance to a receptor in the

olfactory epithelium. The epithelium then relays the

information to the central nervous system (Hino

et al. 2009). The degree of olfactory discrimination

fish are able to process is high. In most teleosts, the

olfactory epithelium lies in a pit that opens to the

external environment via the incurrent nare, with

water exiting through the excurrent nare. Water con-

taining an olfactory stimulus is taken through the

incurrent nare and released via the excurrent nare

in distinct pulses that correlate to jaw movements

(Atema et al. 2002). The olfactory epithelium con-

tains several types of cells, but only three are consid-

ered to be receptors: (1) Ciliated receptor cells, (2)

mircovillous receptor cells, and (3) crypt cells.

Thommesen (1983) provided evidence that in adult

salmonoid fishes, microvillous cells are specific to

amino acid recognition whereas ciliated receptor

cells are specific to bile salts. A correlation between

density of receptor cells and olfactory acuity in larval

fishes has not been experimentally confirmed (Zeiske

et al. 1992), but it is plausible to assume that a

higher density of receptor cells is correlated with a

higher olfactory ability (Lara 2008). Due to the small

size of larval fishes, no lesion experiments have been

conducted to determine whether responses to odors

are based on olfaction rather than on taste. The fully

developed olfactory system found in most larval

coral-reef fish at settlement suggests that selection

of settlement sites would be mitigated by olfac-

tion rather than gustation which has been experi-

mentally shown in adult fish to be used primarily

in food detection and in assessing palatability

(Atema 1980).

Unlike the wave-like transmission of auditory and

visual signals, chemical signals disperse through the

environment by molecular diffusion and advection.

The factors influencing the distribution of chemical

stimuli vary both spatially and temporally (Atema

1988). At small spatial scales (less than millimeters)

and for brief periods (seconds), passive diffusion

governs dispersal; however, at larger scales, dispersal

of a chemical cue requires fluid motion, and is

current-dependent. Although diffusion gradients are

largely predictable, the behavior of chemical stimuli

in currents is less so (Atema 1988). Variation in the

concentration of chemical cues has the potential to

influence the number of larvae that successfully

settle. This variation may be temporal (e.g., between

years or seasons), or caused by stochastic events

(such as storms), or spatial, and it will also have a

species-dependent component depending on the bi-

ology of the organisms releasing odors.

Because the distribution of chemical cues is de-

pendent on movement of water, oceanographic fea-

tures such as currents, upwelling, and tides will not

only influence the concentrations of chemical stimuli

but also the directional information that the chemi-

cals provide, as well as, differences in temperature

and salinity, oceanographic features are likely to con-

tain unique chemical signatures that include

by-products of organisms (nutrients, amino acids

and monosaccharides) and the lipids and proteins

arising from decomposition (Kingsford et al. 2002).

The odor cues from the organisms that characterize

specific communities can be distributed via oceano-

graphic features. For example, the lagoons of coral

reefs contain high concentrations of mucous and dis-

solved organic compounds from corals and other

associated organisms (Davies and Hughes 1983).

This material can be transported out of the lagoon

in turbid plumes that are tens of meters–kilometers

long (Booth et al. 2000). Atema et al. (2002) sug-

gested that in larval apogonids, the recognition and

preference for odors emanating from lagoons, could

be used to retain the larvae within a reef’s tidal

odors, or olfactory halo during the PLD. It is also

plausible that larval reef fish use chemical gradients

of lagoon odor to locate reefs for settlement.

Wolanski (1994) demonstrated that ebb-tide waters

from the continental shelf of the Great Barrier Reef

and lagoons in other coral reef systems may generate

gradients in chemical concentration that larvae in the

How Nemo finds home 833

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

Coral sea may detect and to which they may behav-

iorally respond. Plumes generated by rivers and es-

tuaries may also produce a clear chemical signature

to which larvae may respond. These features may

disperse kilometers and even tens of kilometers

into mainstream currents (Grimes and Kingsford

1996).

The exact mechanism used by larvae for orienta-

tion or settlement-site selection with olfactory cues is

unknown, especially whilst in the pelagic environ-

ment where there are few features to assist in orien-

tation. As olfactory cues are current-dependent,

locating the source of a cue would require movement

upstream. In terrestrial systems and in shallow aquat-

ic systems the earth’s surface can be used as a refer-

ence, providing directional information and

simplifying the search for an upstream source of a

cue. In contrast, it is difficult for a larva in the pe-

lagic environment to determine it is even in a cur-

rent, let alone which direction is upstream, without

an external reference (e.g., a view of the bottom, or

the ability to detect the earth’s magnetic field). This

makes location of the source of any odor more chal-

lenging for a pelagic animal than for one living in

shallow water. Most experiments on orientation of

larval fish using olfactory cues have taken place in

shallow water over a reef or in the laboratory, and

may not be particularly relevant to orientation in the

pelagic environment. Concentration gradients have

been identified as a plausible means of navigating

through the environment. Odor plumes do not pro-

duce simple concentration gradients; instead they are

turbulent and extremely patchy (Moore and Atema

1991). However, the detection of a particular odor

might trigger changes in behavior, for example a

change in speed or in search pattern that could in-

crease the likelihood of encountering the source of

the odor without tracking a gradient or swimming

upstream.

Most marine organisms are thought to use a com-

bination of chemical information and flow to locate

distant sources of odors. A great deal of research on

tracking plumes has been conducted on marine in-

vertebrates and elasmobranchs. Sharks use chemotro-

potaxis, tracking plumes by comparing gradients in

chemical concentration through their bilateral olfac-

tory receptors (Sheldon 1911; Fraenkel and Gunn

1940), chemically stimulated rheotaxis, i.e., the reac-

tion to swim upstream regardless of the actual loca-

tion of the odor (Hodgson and Mathewson 1971),

and the inability to locate an odor source in stagnant

water (Kleerekoper et al. 1975) have all been

recorded. Gardiner and Atema (2007) showed the

importance of other senses in conjunction with the

olfactory stimulus in the location of sources. To

locate a food source, the smooth dogfish shark,

Mustelus canis, requires the lateral line to detect

flow fields, and olfactory sensors. Although no sim-

ilar mechanisms for orientation have been discovered

in larval fishes, it is likely that the same principles

apply, and that a suite of cues are required for

orientation.

Alternatively, the use of ‘‘infotaxis’’ has been sug-

gested as a means of tracking odors in dilute, patchy

media, where there would be infrequent, sporadic

encounters with the odor cue followed by no detect-

able signal (Vergassola et al. 2007). Vergassola et al.

(2007) developed an algorithm that explains how or-

ganisms may locate sources in dilute media or when

signals are weak. In this type of search pattern, in-

formation plays the same role that concentration

plays in chemotaxis. The searcher uses the history

of its encounters with odors to make informed de-

cisions on the direction to move. The searcher

chooses the direction of movement that locally max-

imizes the rate of information to be gained. Patterns

produced are similar to those naturally shown in

moths and birds, but the context of the search pat-

tern can be applied more broadly to any olfactory

search situation in which the demands of exploring

the environment and maximizing the information

received need to be balanced. This is clearly the

case in the settling of larval fish which have a

finite time in the pelagic environment.

Larval fishes possess the necessary sensory organs

for chemotaxis and behave in ways indicating that

chemotaxis is used for orientation. Traditionally,

olfactory cues were thought to provide information

on a micro-habitat (millimeters–meters) settlement

selection scale (Kingsford et al. 2002). These cues

play an important role for larval fish in the recogni-

tion of microhabitats for settlement (Sweatman 1983;

Elliott et al. 1995; Atema et al. 2002), of food

(Dempsey 1978; Knutsen 1992; Døving et al. 1994;

Batty and Hoyt 1995; Kolkovski et al. 1997), of con-

specifics (Sweatman 1983, Sweatman 1988, Ben-Tzvi

et al. 2010) and of chemical cues associated with

recognition of predators (Dixson et al. 2010).

Sweatman (1988) showed that not only are damsel-

fish, Dascyllus aruanus, attracted to the olfactory cues

of conspecifics, but that other species use chemical

cues to avoid settlement sites occupied by

D. aruanus. Chromis viridis choose settlement sites

based on the olfactory cues of corals known to

have positive settlement histories, indicating that se-

lection of settlement sites could be based on

water-borne cues originating from the coral itself,

and that these varied among individual coral

834 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

colonies (Ben-Tzvi et al. 2010). Microhabitat selec-

tion in anemonefishes is well established; larvae are

attracted to species-specific anemone odor (Elliott

et al. 1995; Arvedlund and Nielsen 1996; Arvedlund

et al. 1999; Dixson et al. 2008). A field experiment

with C. viridis larvae suggested that larvae 4.8 m from

potential settlement sites relied more on olfactory

cues for selecting a coral patch upon which to

settle, than did larvae at smaller distances (Lecchini

et al. 2005).

Settlement habitats are often characterized by spe-

cific organisms. For example, corals characterize

coral reefs; like-wise seagrasses for seagrass beds.

Due to differences in characteristic organisms, it is

likely that specific chemical signatures of habitat

types could allow the detection of chemical cues

for habitat recognition at distances greater than

olfactory cues are traditionally expected to be

useful (Kingsford et al. 2002). Arvedlund and

Takemura (2006) found that not only do

settlement-stage spangled emperor fish, Lethrinus

nebulosus, possess the necessary sensory-organ mor-

phology for using olfactory cues in selection of set-

tlement sites, but laboratory-based behavioral tests

confirmed that olfactory cues were used by these

larvae to choose seagrass beds rather than rubble.

Apogonid larvae in a choice chamber were attracted

to the olfactory cues of the location (lagoon water or

ocean water) to which they were acclimated (Atema

et al. 2002), which suggests that once settlement

stage is reached larvae may be able to return to

natal reefs by actively remaining within the reef-odor

plume. Both apogonid and pomacentrid larvae could

distinguish between chemical cues in water from the

reef where they were collected from, cues in water

from other reefs within a 10 km radius of the

‘‘home’’ reef (Gerlach et al. 2007). This indicates

that chemical cues may be useful for settlement on

a much larger scale than previously expected, with

reefs themselves showing distinct chemical signatures

to which larvae can recognize and respond. Further,

evidence for reef-based signatures was provided by

choice-chamber experiments demonstrating that

newly settled anemonefish, Amphiprion percula

could distinguish between the reefs upon which

they settled from other reefs where suitable habitat

was present (Dixson et al. 2008). In addition, juve-

nile A. percula, which are closely associated with reefs

on vegetated off-shore islands in Papua New Guinea,

used olfactory cues produced by the islands to iden-

tify potential settlement sites (Dixson et al. 2008);

they were able to distinguish water from reefs

which surround islands, from water from reefs

where no islands were present. Islands themselves

are a potential source of many olfactory water-borne

cues that would not emanate from reefs without is-

lands. Elevated concentrations of organic material

from the lush tropical rainforest vegetation on the

islands could extend some distance, and naive larval

A. percula strongly and positively responded to the

olfactory cues produced by terrestrial vegetation

(Dixson et al. 2008). This research provides further

evidence that olfactory cues are potentially useful in

identification of settlement sites at distances greater

than previously expected.

Although olfactory capabilities during the latter

portion of the larval phase have been researched in

laboratory settings (Arvedlund et al. 1999; Atema

et al. 2002; Wright et al. 2005; Arvedlund and

Takemura 2006; Gerlach et al. 2007; Wright et al.

2008; Dixson et al. 2010), few studies have been con-

ducted in the field (Sweatman 1988; Elliott et al.

1995, Lecchini et al. 2005; Ben-Tzvi et al. 2010).

Sensory information available to larvae for behavioral

choices, such as the specific chemical cues or the

concentrations of them required for larvae to recog-

nize and respond is poorly understood. Fish larvae

are fragile, and only a limited number of marine fish

species can be reared in captivity, so research has

concentrated on the latter part of the larval period.

It is clear that late-stage larvae of a few species have

well-developed olfactory systems capable of distin-

guishing between different associated habitats (e.g.,

live coral, conspecifics, and symbionts) and that

this does influence the location of their final settle-

ment site. Little research has been conducted on che-

motaxis throughout the entire larval stage; the small

amount that has examined early larvae has tended to

document morphological development (Arvedlund

et al. 1999; Lara 2008) rather than behavior and abil-

ity. The development of the olfactory system of 14

species of parrotfishes (Scaridae) and wrasses

(Labridae) have been described (Lara 2008).

Settlement-stage scarids had incomplete formation

of the olfactory pit and low density of receptor

cells, whereas all labrid species examined had com-

plete and well developed olfactory organs at settle-

ment (Lara 2008). This morphological variation has

not been correlated with olfactory ability or behav-

ior. In the only study on the behavioral response of

newly hatched larvae of coral-reef fish larvae to an

olfactory cue, newly hatched A. percula larvae in a

two-channel choice-chamber were capable of innate

recognition of predators through olfactory cues alone

(Dixson et al. 2010). This demonstrates that use of

olfactory cues is likely throughout the entire pelagic

larval stage, at least for species that hatch relatively

well-developed from demersal eggs. However,

How Nemo finds home 835

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

additional research is required to identify the chem-

ical cues used and the behavioral responses that their

recognition generates. Larvae from pelagic eggs are

generally less developed at hatching, and might not

have functioning olfactory systems soon after hatch-

ing. As the larvae of demersal fishes are in the water

column from weeks to months, a full understanding

of marine connectivity will not be achieved until re-

searchers consider the ontogeny of olfactory abilities

throughout the larval period.

Variation in the presence and concentrations of

chemical stimuli and the likelihood that natal envi-

ronments have unique chemical signatures could

allow pre-settlement larvae to use chemical cues to

find specific habitats. Although, never proven in

coral reef systems, the use of chemical cues in rec-

ognition of natal settlement sites has been suggested,

particularly if a greater proportion of recruits return

to natal reefs than would be expected if larvae were

passively drifting (Jones et al. 2005; Almany et al.

2007). Recognition of natal settlement sites through

olfactory cues requires larvae to imprint on the odor

associated with their natal reefs. Chemical imprinting

by anemonefish in the laboratory fulfils all of the

requirements for ecological imprinting (Arvedlund

et al. 1999); however, this mechanism of learning

has not been demonstrated to occur in the sea. For

imprinting to occur, the ‘‘sensitive phase’’ would

need to take place during the embryonic period

and to be stimulated by chemical cues diffusing

across the membrane of an egg being incubated on

the reef, or immediately after hatching, in the brief

time before larvae are exported. Settling anemonefish

larvae, A. melanopus, which had been incubated near

the anemone Entacmaea quadricolor, used olfactory

rather than visual cues to recognize anemones at set-

tlement (Arvedlund et al. 1999). Further, the larvae

significantly preferred this anemone, compared to

other host anemone species. This provides evidence

that embryos are capable of learning a chemical cue

and responding to it at a later stage (Arvedlund and

Nielsen 1996). Many fish have strong microhabitat

associations and these have a complex influence on

recruitment patterns (Sale et al. 1984; McCormick

and Makey 1997). Imprinting on chemical cues is

one mechanism which could help to explain strong

site attachments and self-recruitment in many spe-

cies. Although the possibility of imprinting on chem-

ical cues exists, the influence it plays in natural

systems remains unknown.

One of the biggest challenges in chemical ecology

is gaining comprehension of the complete chemical

picture; this includes not only the characterization of

the specific chemicals and the behavioral reaction

they induce in the larval fishes, but also an under-

standing of the ecologically relevant concentrations

at which the chemicals are useful as cues. Although

the use of olfactory cues for orientation and for lo-

cating sites for settlement is well documented in

larval fishes, in no case is it known what substance

is being used by the larvae. Further, the research thus

far has been conducted in the laboratory or in shal-

low water over reefs, and nothing is known of the

possible use of olfactory cues by larvae in the pelagic

environment where most of the larval phase takes

place.

Light and vision

Vision clearly plays a role in orientation of fish

larvae, but direct use of vision is limited by the clar-

ity of sea water to a few tens of meters, and is there-

fore not of great use for orientation at larger scales.

However, if fish larvae can utilize cues like polarized

light or a solar compass, then vision can be used for

orientation over very large scales, and the cues would

be location-independent.

Many animals use polarized light or a solar com-

pass for orientation and recent studies suggest that

such visual cues may also be used by larval damsel-

fish (e.g., Chromis atripectoralis) at a time when they

must find suitable habitat for settlement. (Leis and

Carson-Ewart 2003; C.B. Paris et al. manuscript in

preparation). This idea is supported by the fact that

juveniles of damselfishes, including the sister species

of C. atripectoralis, possess ‘‘the most complex polar-

ization sensitivity recorded for any vertebrate’’

(Hawryshyn et al. 2003). Although polarized light

will not assist larvae to locate a particular reef, it

can enable them to orient in a particular direction,

and this is helpful in counteracting a prevailing cur-

rent, or, e.g., assisting a larva in the Coral Sea to find

the Great Barrier Reef to the west. Research is re-

quired to directly examine the visual cues available to

larval fish and to determine how these cues are de-

tected and used for orientation. Here, we briefly de-

scribe the underwater light field to show what visual

cues are available to dispersing fish larvae in the

pelagic environment; then we review current knowl-

edge on polarization detection in fishes.

The underwater light spectrum is variable and de-

pends on the depth and quality (amount and type of

particulate matter and dissolved organic material) of

the water (Jerlov 1976). Attenuation of light in water

depends on wavelength-specific scattering and ab-

sorption processes. As a consequence, water acts as

a monochromator and, depending on the type of

water, light may penetrate to depths of a few to

836 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

several 100 m (e.g., turbid waters transmit green or

even yellow light to just a few meters at most while

the clearest oceanic waters best transmit blue light to

several 100 m). However, the visibility of any object

(e.g., landmark) is limited to a maximum of tens of

meters as detection is only possible if sufficient con-

trast exists between the object and background. With

increasing distance from an object, the amount of

scattered light (veiling light) increases and thus, the

contrast between object and background decreases

until it falls below the detection threshold of the

observer (Jerlov 1976). Orientation via visual land-

mark is thus very limited underwater and most likely

not available to larval fish during their pelagic phase,

except as they approach settlement habitat.

The sun radiates unpolarized light; however,

during its passage, light becomes partially linearly

polarized due to scattering in the atmosphere and

water column, and also due to refraction at the wa-

ter’s surface (Horvath and Varju 2004; Sabbah et al.

2005). The overall polarization pattern is variable

and depends on a range of variables, such as

depth, quality of the water and, importantly, the po-

sition of the sun (Horvath and Varju 2004; Sabbah

et al. 2005; Waterman 2006). Seen from below the

water’s surface, the world above is compressed into

Snell’s window (cone with 978). As a consequence,

the overall underwater polarization pattern is com-

plex and consists, at least at relatively shallow depths,

of two components, the distorted polarization pat-

tern of the sky seen through Snell’s window and the

polarization pattern outside Snell’s window caused

by underwater scattering of sunlight. The resulting

overall polarization pattern is ubiquitous and forms

a sphere around the observer (Hawryshyn 1992).

With increasing depth, the visibility of the sky’s

polarization pattern decreases until at 10–15 m

(depending on the water quality) the two patterns

become indistinguishable (Shashar et al. 2004).

Navigation based on the sky’s polarization pattern,

well-known for terrestrial invertebrates, is thus pre-

dicted to be restricted to relatively shallow depths.

However, any animal sensitive to e-vector (electric

field component of light) orientation is potentially

able to use the overall underwater polarization pat-

tern as a sun compass even when a direct view of the

sun is obscured (for detailed reviews of underwater

polarization, see Jerlov 1976; Horvath and Varju

2004; Sabbah et al. 2005; Waterman 2006). The

depth at which larval fishes swim during their pelagic

phase depends on the species and time of day, but

the larvae of many species (e.g., damselfish) are

found at relatively shallow depths (Leis 1991) so

that orientation using sky polarization, and with a

sun compass, is possible.

Sensitivity to polarization has been shown for

adults or juveniles of several fish species of at least

three orders (e.g., Clupeiformes: Novales Flamarique

and Hawryshyn 1998; Novales Flamarique and

Harosi 2002; Perciformes: Hawryshyn et al. 2002;

Hawryshyn et al. 2003; Mussi et al. 2005; and

Salmoniformes: Kawamura et al. 1981; Hawryshyn

et al. 1990) using a range of techniques including

behavioral conditioning, electrophysiology and

innate orientation behavior. Thus, polarization sen-

sitivity is wide-spread in fishes, and given that the

retinal structures involved differ among taxa (see

below), it seems it has evolved in teleosts more

than once.

Many animals, including fishes, orient themselves

using polarization cues; however, the biological func-

tion of this behavior is largely unclear. Exceptions

are the shore-escape responses shown to be depen-

dent on polarization cues for Daphnia (Schwind

1999) and grass-shrimp (Goddard and Forward

1991). Also, there is evidence from studies of fish

migrations for the use of celestial cues for orienta-

tion. The anadromous sockeye salmon fry (Quinn

1980) and smolts (Quinn and Brannon 1982) use

both a celestial and a magnetic compass during

their fresh-water migrations. There is some evidence

for sun-compass orientation in the white bass,

Roccus chrysops (Moronidae), two species of

sunfish (Lepomis macrochirus and L. gibbosus,

Centrarchidae), and the carachid, Cheirodon pulcher

(Hasler et al. 1958; Levin et al. 1992). As mentioned

above, the swimming behavior of some reef-fish

larvae is also highly suggestive of the use of celestial

cues for orientation (Leis and Carson-Ewart 2003;

Leis et al. 2007) but, aside from salmon, the nature

of the visual cues used by the fish was not investi-

gated further.

The mechanism for detection of polarization is

well understood for invertebrates, but it is still

unclear how sensitivity to polarization is achieved

in vertebrates. At least two receptors sensitive to dif-

ferent e-vector orientations are required to detect

polarization patterns. In invertebrates, the

visual-pigment molecules responsible for the absorp-

tion of light lie parallel to each other within micro-

villi that are oriented perpendicular to incident light.

The microvilli of different photoreceptors have dif-

ferent orientations (e.g., perpendicular to each other

in the bee) and thus, sensitivity to polarization is

achieved when the signals of the different photore-

ceptors are compared and translated into different

intensities. In vertebrates, on the other hand, the

How Nemo finds home 837

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

pigment molecules appear to be oriented randomly

relative to the incident axial light so that it is unclear

how sensitivity is achieved. Three mechanisms are

currently discussed in the literature (for detailed

review see Hawryshyn 2010):

(1) The mechanism in anchovies (Engraulidae) is

best understood. Anchovies have a specialized

retina in which the disc membranes of the

cone outer segments are tilted so that mem-

brane stacks parallel the long axis of the pho-

toreceptors (Fineran and Nicol 1978). In

anchovies, bilobed and long cones (both are

maximally sensitive to 500 nm light) have mem-

brane stacks perpendicular to each other and

are sensitive to vertically and horizontally po-

larized light, respectively (for details see Novales

Flamarique and Hawryshyn 1998).

(2) Sensitivity to polarization may also be achieved

by internal reflection within double cones (red/

green sensitive) onto neighboring single cones

that are UV-sensitive, as has been suggested for

some cyprinids and salmonids (Novales

Flamarique et al. 1998). In these fishes, double

cones are arranged in a square mosaic with UV

corner cones and a blue,

polarization-insensitive, central cone.

(3) More recently, studies using microspectropho-

tometry suggested that individual cones may be

polarization-sensitive based on the tilted optical

geometry of the cones (but not rods) of coho

salmon, Onchorhynchus kisutch (Roberts et al.

2004) and goldfish, Cassius auratus

(Cyprinidae) (Roberts and Needham 2007).

Possibly, the latter two mechanisms work to-

gether in maximizing polarization contrast

(Hawryshyn 2010).

The retina of larval fish changes rapidly during

early development. At hatching, the eyes are generally

relatively undeveloped and lack visual pigment.

However, by the time ‘‘typical’’ teleost larvae feed

for the first time (�2 days after hatching) their

eyes are functional. Their retinae initially contain

cones of a single morphological photoreceptor type

(Blaxter 1986; Shand et al. 1999) with rods and other

cone types developing later, at about the time of

settlement (Evans and Fernald 1990; Shand 1993).

In spotted sandbass, Paralabrax maculatofasciatus

(Serranidae) larvae, e.g., double cones do not differ-

entiate until 14 days after hatching, or about

mid-way through the pelagic phase (Pena and

Dumas 2007). If double cones, arranged in a

square mosaic, are indeed required for sensitivity to

polarization, then it would be expected that larvae

are only able to use this cue as a compass relatively

late during their pelagic phase, which may coincide

with the start of the search for a reef to settle upon.

In salmon, ontogenetic changes take place in the dis-

tribution of retinal cells sensitive to polarized light

(Hawryshyn 2010), which serves as a warning about

the possibility of such changes in reef fishes.

Integration and future directions

In most animals, including larval marine fishes, ori-

entation depends on more than one sensory mode

(Kingsford et al. 2002). We need to move toward an

understanding of the role of all sensory modes in

orientation by larval fish, and how they might inter-

act with and complement each other. It is important

to do this for a range of larval developmental stages

and also to address the neglected areas of spatial

variation in, and social effects on, orientation. A

broad, integrated view of the role of different sensory

modalities in orientation of larval fish will lead to a

better understanding of dispersal and population

connectivity and to improved ability to model

these processes.

How does Nemo find home? Almost certainly,

marine fish larvae use a variety of cues and senses

that come into play during different portions of the

pelagic phase, and at different distances from the reef

home. Hypothetically, celestial visual cues might be

used early in development to maintain a general po-

sition or counteract an average current direction.

Reef-based sound and hearing may be used as settle-

ment competency approaches, and at greater dis-

tances from the reef, whereas dissolved scents and

olfaction might come into play after sound has as-

sisted the larva to reach the general vicinity of the

reef. It is also possible that cues and senses interact

more directly, e.g., detection of a particular scent

may induce the larva to respond to sound in a dif-

ferent way, or, in the absence of appropriate sound

cues, celestial cues may take precedence. We have

not yet begun to scratch the surface of these possible

interactions.

We have not included consideration of a magnetic

sense here simply because nothing is known about it

in larvae of marine fishes. However, as a magnetic

sense is widely distributed in fishes (e.g., rockfish,

tuna, and salmon), there is no reason to expect it

will not exist in marine fish larvae. This is clearly a

fruitful area for research.

Future research in all areas must emphasize just

how the cues that the larvae can demonstrably detect

are actually used for orientation. The fact that we are

838 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

not entirely sure how sound is localized by fishes in

the ocean, or how a pelagic animal can use olfaction

to locate a reef, makes it difficult to investigate these

senses in the context of orientation in larval stages.

Some priority research areas include answering the

following questions:

(1) Over what spatial scales can location-dependent

cues be detected?

(2) What taxonomic similarities and differences

exist in sensory abilities?

(3) What portion of the variety of

location-dependent and location-independent

sensory cues available in the marine environ-

ment do larvae respond to physiologically and

behaviorally and how does this change ontoge-

netically over the full period of larval

development?

(4) What level of cue strength above background

‘‘noise’’ (or smell or light) is required for de-

tectability and response?

(5) What relationships are there between structure

and function of sensory organs? Can function

be predicted by structure and state of

development?

(6) How do larvae localize sound and olfactory cues

in the pelagic environment?

(7) Does knowledge of sensory sensitivity enable

prediction of behavioral response?

(8) Do habitat-unique sounds or odors enable

homing to natal locations, and if such homing

based on location-dependent cues exists, is it

hard-wired, or capable of being imprinted? If

natal homing exists, is it present only in species

that incubate their eggs on the reef, or can it

occur in species that spawn pelagic eggs?

(9) Can field experiments test for differences in

location-independent orientation behavior of

larvae between cloudy and sunny days? This

could indicate what portions of the underwater

light field are used for orientation, and

(10) If location-independent cues are used for orien-

tation, how are they applied, and what are the

selective advantages? Further, over just what

spatial scales are these apparently location-

independent cues actually independent of

location?

At present, we know by far the most about sound

as an orientation cue for marine larval fishes, but

there are still large gaps that are important to fill.

We know a little about olfaction as a sensory cue,

mostly in shallow-water situations and at small

scales. We know almost nothing about the use of

polarized light as a cue for orientation. The use of

other cues, such as wave direction or magnetic fields

has been speculated upon, but there are no hard

data. It will be essential to test in the sea the

things learned in the laboratory, and this is especially

challenging with small organisms such as larvae that

live in pelagic environments.

Acknowledgments

We thank Chuck Derby and Dick Zimmer for invit-

ing our participation in the symposium, and sup-

porting travel. Dennis Higgs and Bruce Munday

constructively criticized portions of the manuscript,

Jackie Webb provided insightful discussions, and

Suzanne Bullock provided editorial assistance. We

also thank the National Science Foundation, Biologi-

cal Bulletin, the Society for Integrative and Com-

parative Biology and its Divisions of Neurobiology,

Animal Behavior, and Ecology and Evolution, and

the American Microscopical Society for support of

the symposium.

Funding

The Australian Museum; the Australian Research

Commission Discovery grant (DP110100695) to

J.M.L. and U.S.; National Science Foundation

grant (IOS-1036742, IOS-0614685) to C.D. Derby.

References

Almany GR, Berumen ML, Thorrold SR, Planes S, Jones GP.

2007. Local replenishment of coral reef fish populations in

a marine reserve. Science 316:742–44.

Arvedlund M, Kavanagh K. 2009. The senses and environ-

mental cues used by marine larvae of fish and decapod

crustaceans to find tropical coastal ecosystems.

In: Nagelkerken L, editor. Ecological connectivity among

tropical coastal ecosystems. Dordrecht: Springer. p. 135–84.

Arvedlund M, McCormick MI, Fautin DG, Bildsoe M. 1999.

Host recognition and possible imprinting in the anemone-

fish Amphiprion melanopus. Mar Ecol Prog Ser 188:207–18.

Arvedlund M, Nielsen LE. 1996. Do the anemone fish

Amphiprion ocellaris (Pisces : Pomacentridae) imprint

themselves to their host sea anemone Heteractis magnifica

(Anthozoa : Actinidae)? Ethology 102:197–211.

Arvedlund M, Takemura A. 2006. The importance of

chemical environmental cues for juvenile Lethrinus

nebulosus Forsskal (Lethrinidae, Teleostei) when settling

into their first benthic habitat. J Exp Mar Biol Ecol

338:112–22.

Atema J. 1980. Smelling and tasting underwater. Oceanus

23:4–19.

How Nemo finds home 839

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

Atema J. 1995. Eddy chemotaxis and odor landscapes:

exploration of nature with animal sensors. Biol Bull

191:129–38.

Atema J. 1988. Distribution of chemical stimuli. In: Atema J,

Fay RR, Popper AN, Tavolga WN, editors. Sensory

Biology of Aquatic Organisms. Berlin: Springer-Verlag.

p. 29–56.

Atema J, Kingsford MJ, Gerlach G. 2002. Larval fish could use

odour for detection, retention and orientation to reefs. Mar

Ecol Prog Ser 241:151–60.

Batty RS, Hoyt RD. 1995. The role of sense organs in the

feeding behaviour of juvenile sole and plaice. J Fish Biol

47:931–9.

Ben-Tzvi O, Tchernov D, Kiflawi M. 2010. Role of

coral-derived chemical cues in microhabitat selection by

settling Chromis viridis. Mar Ecol Prog Ser 409:181–7.

Blaxter JHS. 1986. Development of sense organs and behavior

of teleost larvae with special reference to feeding and pred-

ator avoidance. Trans Am Fish Soc 115:98–114.

Cato DH. 1978. Marine biological choruses in tropical waters

near Australia. J Acoust Soc Am 64:736–43.

Cato DH. 1992. The biological contribution to the ambient

noise in waters near Australia. Acoust Aust 20:76–80.

Clark CW, Ellison WT, Southall BL, Hatch L, van Parijis SM,

Frankel A, Ponirakis D. 2009. Acoustic masking in marine

ecosystems: intuitions, analysis and implication. Mar Ecol

Prog Ser 395:201–22.

Cowen RK, Gawarkiewicz G, Pineda J, Thorrold SR,

Werner FE. 2007. Population connectivity in marine sys-

tems: an overview. Oceanography 20:14–21.

Cowen RK, Sponaugle S. 2009. Larval dispersal and

marine population connectivity. Annu Rev Marine Sci

1:443–66.

Davies PJ, Hughes H. 1983. High energy reef and terrigenous

sedimentation, boulder reef, Great Barrier Reef. BMR

Journal of Australian Geology & Geophysics 8:201–9.

Dempsey CH. 1978. Chemical stimuli as a factor in feeding

and intraspecific behaviour of herring larvae. J Mar Biol

Assoc UK 58:739–47.

Dixson DL, Jones GP, Munday PL, Planes S, Pratchett MS,

Srinivasan M, Syms C, Thorrold SR. 2008. Coral reef fish

smell leaves to find island homes. Proc Royal SocB Biol Sci

275:2831–9.

Dixson DL, Munday PL, Jones GP. 2010. Ocean acidification

disrupts innate ability of fish to detect predator olfactory

cues. Ecol Lett 13:68–75.

Døving KB, Marstøl M, Andersen JR, Knutsen JA. 1994.

Experimental evidence of chemokinesis in newly hatched

cod larvae (Gadus morhua L.). Mar Biol 120:351–8.

Edds-Walton PL, Fay RR. 2002. Directional auditory proses-

sing in the oyster toadfish Opsanus tau. Bioacoustics

12:197–217.

Egner SA, Mann DA. 2005. Auditory sensitivity of sergeant

major damselfish Abudefduf saxatilis from post-settlement

juvenile to adult. Mar Ecol Prog Ser 285:213–22.

Elliott JK, Elliott JM, Mariscal RN. 1995. Host selection,

location, and association behaviours of anemonefishes in

field selection experiments. Mar Biol 122:377–89.

Evans BI, Fernald RD. 1990. Metamorphosis and fish vision.

J Neurobiol 21:1037–52.

Fineran BA, Nicol JAC. 1978. Studies on the photoreceptors

of Anchoa mitchilli and A. hepsetus (Engraulidae) with par-

ticular reference to the cones. Philos Trans R Soc Lond, Ser

B: Biol Sci 283:25–60.

Fisher R, Leis JM. 2009. Swimming performance in larval

fishes: from escaping predators to the potential for long

distance migration. In: Domenici P, Kapoor BG, editors.

Fish locomotion: an etho-ecological approach. NH, USA:

Science Publishers, Enfield. p. 333–73.

Fraenkel GS, Gunn DI. 1940. The orientation of animals:

kineses, taxes and compass reactions. Oxford: Clarendon

Press.

Fuiman LA, Werner RG. 2002. Fishery Science: the unique

contributions of early life stages. Oxford: Blackwell.

Gagliano M, Depczynski M, Simpson SD, Moore JAY. 2008.

Dispersal without errors: symmetrical ears tune into the

right frequency for survival. Proc Royal Soc B Biol Sci

275:527–34.

Gardiner JM, Atema J. 2007. Sharks need the lateral line to

locate odor sources: rheotaxis and eddy chemotaxis. J Exp

Biol 210:1925–34.

Gerlach G, Atema J, Kingsford MJ, Black KP, Miller-Sims V.

2007. Smelling home can prevent dispersal of reef fish

larvae. Proc Natl Acad Sci USA 104:858–63.

Goddard SM, Forward RB. 1991. The role of the underwater

polarized light pattern in sun compass navigation of the

grass shrimp, Palaemonetes vulgaris. J Comparative

Physiology A 169:479–91.

Gorga MP, Kaminski JR, Beauchaine KA, Jesteadt W. 1988.

Auditory brainsteam response to tone bursts in nor-

mally hearing subjects. J Speech, Language Hearing Res

31:87–97.

Grimes CB, Kingsford MJ. 1996. How do riverine plumes of

different sizes influence fish larvae: Do they enhance re-

cruitment? Mar Freshwat Res 47:191–208.

Hasler AD, Horrall RM, Wisby WJ, Braemer W. 1958.

Sun-orientation and homing fishes. Limnol Oceanogr

3:353–61.

Hawryshyn CW. 1992. Polarization vision in fish. Am Sci

80:164–75.

Hawryshyn CW. 2010. Ultraviolet polarization vision and vi-

sually guided behavior in fishes. Brain Behav Evol

75:186–94.

Hawryshyn CW, Arnold MG, Bowering E, Cole RL. 1990.

Spatial orientation of Rainbow-trout to plane-polarized

light - the ontogeny of E-vector discrimination and spectral

sensitivity characteristics. J Comp Physiol a-Sensory Neural

Behavioral Physiol 166:565–74.

Hawryshyn CW, Moyer H, Allison WT, Haimberger TJ,

McFarland WN. 2002. Polarization vision in coral reef

840 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

fishes: Multichannel detection capabilities it Damselfishes.

Invest Ophthalmol Vis Sci 43:U1298.

Hawryshyn CW, Moyer HD, Allison WT, Haimberger TJ,

McFarland WN. 2003. Multidimensional polarization sen-

sitivity in damselfishes. J Comp Physiol A 189:213–20.

Heenan A, Simpson SD, Braithwaite VA. 2008. Testing

the generality of acoustic cue use at settlement in lar-

val coral reef fish. Proceedings of the 11th

International Coral Reef Symposium. Fort Lauderdale,

Florida. p. 554–8.

Higgs DM. 2005. Auditory cues as ecological signals for

marine fishes. Mar Ecol Prog Ser 287:278–81.

Higgs DM, Lu Z, Mann DA. 2006. Hearing and mechanore-

ception. In: Evans DH, Claiborne JB, editors. The

physiology of fishes. 3rd edn. New York: CRC Press.

p. 389–427.

Hino H, Miles NG, Bandoh H, Ueda H. 2009. Molecular

biological research on olfactory chemoreception in fishes.

J Biol 75:945–59.

Hodgson ES, Mathewson RF. 1971. Chemosensory orientation

in sharks. Annals of the New York Academy of Science

188:175–82.

Horvath G, Varju D. 2004. Polarized light in animal vision:

polarization patterns in nature. Berlin: Springer.

Jerlov NG. 1976. Marine optics. Amsterdam: Elsevier.

Jones GP, Almany GR, Russ GR, Sale PF, Steneck RS,

van Oppen MJH, Willis BL. 2009. Larval retention and

connectivity among populations of corals and reef fishes:

history, advances and challenges. Coral Reefs 28:307–26.

Jones GP, Planes S, Thorrold SR. 2005. Coral reef fish larvae

settle close to home. Curr Biol 15:1314–8.

Kaufman L, Ebersole J, Beets J, McIvor CC. 1992. A key phase

in the recruitment dynamics of coral reef fishes:

post-settlement transition. Environ Biol Fishes 34:109–18.

Kawamura G, Shibata A, Yonemori T. 1981. Response of tel-

eosts to the plane of polarized light as determined by the

heart beat rate. Bull Jap Soc Sci Fish 47:727–9.

Kennedy EV, Holderied MW, Mair JM, Guzman HM,

Simpson SD. 2010. Spatial patterns in reef-generated

noise relate to habitats and communities: Evidence from

a Panamanian case study. J Exp Mar Biol Ecol 395:85–92.

Kingsford MJ, Leis JM, Shanks A, Lindeman K, Morgan S,

Pineda J. 2002. Sensory environments, larval abilities and

local self-recruitment. Bull Mar Sci 70:309–40.

Kleerekoper H, Gruber D, Matis J. 1975. Accuracy of locali-

zation of a chemical stimulus in flowing and stagnant water

by the nurse shark Ginglymostoma cirratum. J Comp

Physiol A 42:79–84.

Knutsen JA. 1992. Feeding behaviour of North Sea turbot

(Scophthalmus maximus) and Dover sole (Solea solea)

larvae elicited by chemical stimuli. Mar Biol 113:543–8.

Kojima T, Ito H, Komada T, Taniuchi T, Akamatsu T. 2005.

Measurements of auditory sensitivity in common carp

Cyprinus carpio by the auditory brainstem response technique

and cardiac conditioning technique. Fishery Sci 71:95–100.

Kolkovski S, Arieli A, Tandler A. 1997. Visual and chemical

cues stimulate microdiet ingestion of sea bream larvae.

Aquacult Int 5:527–36.

Lara MR. 2008. Development of the nasal olfactory organs

in the larvae, settlement-stages and some adults of

14 species of Caribbean reef fishes (Labridae, Scaridae,

Pomacentridae). Mar Biol 154:51–64.

Lecchini D, Shima J, Banaigs B, Galzin R. 2005. Larval sensory

abilities and mechanisms of habitat selection of a coral reef

fish during settlement. Oecologia 143:326–34.

Leis JM. 1991. Vertical distribution of fish larvae in the Great

Barrier Reef Lagoon, Australia. Mar Biol 109:157–66.

Leis JM. 2006. Are larvae of demersal fishes plankton or

nekton? Adv Mar Biol 51:59–141.

Leis JM. 2010. Ontogeny of behaviour in larvae of marine

demersal fishes. Ichthyol Res 57:325–42.

Leis JM, Carson-Ewart BM. 2003. Orientation of pelagic

larvae of coral-reef fishes in the ocean. Mar Ecol Prog Ser

252:239–53.

Leis JM, Carson-Ewart BM, Cato DH. 2002. Sound detection

in situ by the larvae of a coral-reef damselfish

(Pomacentridae). Mar Ecol Prog Ser 232:259–68.

Leis JM, Carson-Ewart BM, Hay AC, Cato DH. 2003.

Coral-reef sounds enable nocturnal navigation by some

reef-fish larvae in some places and at some times. J Fish

Biol 63:724–37.

Leis JM, Lockett MM. 2005. Localization of reef sounds by

settlement-stage larvae of coral-reef fishes (Pomacentridae).

Bull Mar Sci 76:715–24.

Leis JM, McCormick MI. 2002. The biology, behaviour and

ecology of the pelagic, larval stage of coral-reef fishes.

In: Sale PF, editor. Coral reef fishes: dynamics and diversity

in a complex ecosystem. San Diego: Academic Press.

p. 171–99.

Leis JM, van Herwerden L, Patterson HM. Estimating con-

nectivity in marine fish populations: what works best?

Oceanogr Mar Biol Annu Rev. Forthcoming 2011.

Leis JM, Wright KJ, Johnson RN. 2007. Behaviour that influ-

ences dispersal and connectivity in the small, young larvae

a reef fish. Mar Biol 153:103–17.

Levin LE, Belmonte P, Gonzalez O. 1992. Sun-compass ori-

entation in the characid Cheirodon pulcher. Environ Biol

Fishes 35:321–5.

Lovell JM, Findlay MM, Moate RM, Yan HY. 2005. The hear-

ing abilities of the prawn Palaemon serratus. Comp

Biochem Physiol, A: Mol Integr Physiol 140:89–100.

Lugi M, Yan HY, Fine ML. 2003. Acoustic communication in

two freshwater gobies: the relationship between ambient

noise, hearing thresholds and sound spectrum. J Comp

Physiol A 189:309–20.

Malakoff D. 2010. A push for quieter ships. Science

328:1502–3.

Mann DA, Casper BM, Boyle KS, Tricas TC. 2007. On the

attraction of larval fishes to reef sounds. Mar Ecol Prog Ser

338:307–10.

How Nemo finds home 841

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

McCauley RD, Cato DH. 2000. Patterns of fish calling in a

nearshore environment in the Great Barrier Reef. Philos

Trans R Soc London B 355:1289–93.

McCormick MI, Makey LJ. 1997. Post-settlement transition in

coral reef fishes: overlooked complexity in niche shifts. Mar

Ecol Prog Ser 153:247–57.

Montgomery JC, Jeffs A, Simpson SD, Meekan M, Tindle C.

2006. Sound as an orientation cue for the pelagic larvae

of reef fishes and decapod crustaceans. Adv Mar Biol

51:143–96.

Montgomery JC, Tolimieri N, Haine OS. 2001. Active habi-

tat selection by pre-settlement reef fishes. Fish Fish

2:261–77.

Moore P, Atema J. 1991. Spatial information contained in

three dimensional fine structure of an aquatic odor

plume. Biol Bull 181:408–18.

Mussi M, Haimberger TJ, Hawryshyn CW. 2005. Behavioral

discrimination of polarized light in the damselfish Chromis

viridis (family Pomacentridae). J Exp Biol 208:3037–3046.

Novales Flamarique I, Harosi FI. 2002. Visual pigments and

dichronism of anchovy cones: a model system for polari-

zation detection. Visual Neurosci 19:467–73.

Novales Flamarique I, Hawryshyn CW. 1998. Photoreceptor

types and their relation to the spectral and polarization

sensitivities of clupeid fishes. J Comp Physiol A

182:793–803.

Novales Flamarique I, Hawryshyn CW, Harosi FI. 1998.

Double-cone internal reflection as a basis for polarizarion

detection in fish. J Opt Soc Am A 15:349–58.

Pena R, Dumas S. 2007. Desarrollo de la tetina dn larvas de

cabrilla arenera Paralabrax maculatofasciatus (percoidei:

Serranidae) bajo condiciones de cultivo [Development of

the retina in spotted sand bass Paralabrax maculatofasciatus

(Percoidei: Serranidae) larvae under culture conditions].

Cienc Mar 33:293–300.

Popper AN, Fay RR, Platt C, Sand O. 2003. Sound detection

mechanisms and capabilities of teleost fishes. In: Collin SP,

Marshall NJ, editors. Sensory processing in aquatic environ-

ments. New York: Springer. p. 3–38.

Quinn TP. 1980. Evidence for celestial and magnetic compass

orientation in lake migrating sockeye salmon fry. J Comp

Physiol 137:243–8.

Quinn TP, Brannon EL. 1982. The use of celestial and mag-

netic cues by orienting sockeye salmon smolts. J Comp

Physiol 147:547–52.

Radford CA, Stanley JA, Tindle CT, Montgomery JC,

Jeffs AG. 2010a. Localized coastal habitats have dis-

tinct underwater sound signatures. Mar Ecol Prog Ser

401:21–9.

Radford CA, Tindle CT, Montgomery JC, Jeffs AG. 2010b.

Are reefs acting as larval loud hailers? Paper presented at

The 9th Larval Biology Symposium. Wellington, NZ.

Roberts CM. 1997. Connectivity and management of

Caribbean coral reefs. Science 278:1454–6.

Roberts NW, Gleeson HE, Temple SE, Haimberger TJ,

Hawryshyn CW. 2004. Differences in the optical properties

of vertebrate photoreceptor classes leading to axial polari-

zation sensitivity. J Opt Soc Am A 21:335–45.

Roberts NW, Needham MG. 2007. A mechanism of polarized

light sensitivity in cone photoreceptors of the goldfish,

Crassius auratus. Biophys J 93:3241–8.

Sabbah S, Lerner A, Erlick C, Shashar N. 2005. Underwater

polarization vision: a physical examination.

In: Pandalai SG, editor. Recent research developments in

experimental and theoretical biology. Trivandrum:

Transworld Research Network. p. 123–77.

Sale PF, Douglas WA, Doherty PJ. 1984. Choice of micro-

habitats by coral reef fishes at settlement. Coral Reefs

3:91–9.

Sand O. 2002. Sound source localisation: an historical assess-

ment. Bioacoustics 12:199–201.

Schwind R. 1999. Daphnia pulex swims toward the most

strongly polarized light - a response that leads to ’shore

flight’. J Exp Biol 202:3631–5.

Shand J. 1993. Changes in the spectral absorption of cone

visual pigments during the settlement of the goatfish

Upeneus tragula: the loss of red sensitivity as a benthic

existence begins. J Comp Physiol A 173:115–21.

Shand J, Archer MA, Colin SP. 1999. Ontogenetic changes in

the retinal photoreceptor mosaic in a fish, the black bream,

Acanthopagrus butcheri. J Comp Neurol 412:203–17.

Shashar N, Sabbah S, Cronin TW. 2004. Transmission of lin-

early polarized light in seawater: implications for polariza-

tion signaling. J Exp Biol 207:3619–28.

Sheldon RE. 1911. The sense of smell in selachians. Journal of

Experimental Zoology 10:51–62.

Simpson SD, Meekan M, Montgomery J, McCauley R, Jeffs A.

2005a. Homeward sound. Science 308:221.

Simpson SD, Meekan MG, Jeffs A, Montgomery JC,

McCauley RD. 2008. Settlement-stage coral reef fishes

prefer the higher frequency invertebrate-generated audible

component of reef noise. Anim Behav 75:1861–8.

Simpson SD, Meekan MG, Larsen NJ, McCauley RD, Jeffs A.

2010. Behavioural plasticity in larval reef fish: orientation is

influenced by recent acoustic experiences. Behavioural

Ecology 21:1098–105.

Simpson SD, Meekan MG, McCauley RD, Jeffs A. 2004.

Attraction of settlement-stage coral reef fishes to reef

noise. Mar Ecol Prog Ser 276:263–8.

Simpson SD, Yan HY, Wittenrich ML, Meekan MG. 2005b.

Response of embryonic coral reef fishes (Pomacentridae:

Amphiprion spp.) to noise. Mar Ecol Prog Ser 287:201–8.

Stanley JA, Radford CA, Jeffs AG. 2010. Induction of settle-

ment in crab megalopae by ambient underwater reef sound.

Behav Ecol 21:113–20.

Stobutzki IC, Bellwood DR. 1998. Nocturnal orientation to

reefs by late pelagic stage coral reef fishes. Coral Reefs

17:103–10.

Sweatman HPA. 1983. Influence of conspecifics on choice of

settelment sites by larvae of two pomacentrid fishes

(Dascyllus aruanus and D. reticulatus) on coral reef. Mar

Biol 75:225–9.

842 J. M. Leis et al.

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from

Sweatman HPA. 1988. Field evidence that settling coral reef

fish larvae detect resident fishes using dissolved chemical

cues. J Exp Mar Biol Ecol 124:163–74.

Thommesen G. 1983. Morphology, distribution and specifi-

city of olfactory receptor cells in salmonid fishes. Acta

Physiol Scand 117:241–9.

Tolimieri N, Haine O, Jeffs A, McCauley R, Montgomery J.

2004. Directional orientation of pomacentrid larvae to am-

bient reef sound. Coral Reefs 23:184–94.

Tolimieri N, Haine O, Montgomery JC, Jeffs A. 2002.

Ambient sound as a navigational cue for larval reef fish.

Bioacoustics 12:214–7.

Tolimieri N, Jeffs A, Montgomery JC. 2000. Ambient sound

as a cue for navigation by the pelagic larvae of reef fishes.

Mar Ecol Prog Ser 207:219–24.

Vergassola M, Villermaux E, Shraiman BI. 2007. ‘Infotaxis’ as a

strategy for searching without gradients. Nature 445:406–9.

Waterman TH. 2006. Reviving a neglected celestial underwa-

ter polarization compass for aquatic animals. Biol Rev

Camb Philos Soc 81:111–5.

Williams DM, Wolanski E, Andrews JC. 1984. Transport

mechanisms and the potential movement of planktonic

larvae in the central region of the Great Barrier Reef.

Coral Reefs 3:229–36.

Wolanski E. 1994. Physical oceanographic processes of the

Great Barrier Reef Boca Raton. FL: CRC Press.

Wright KJ, Higgs DM, Belanger AJ, Leis JM. 2005. Auditory

and olfactory abilities of pre-settlement larvae and

post-settlement juveniles of a coral reef damselfish

(Pisces: Pomacentridae). Mar Biol 147:1425–34.

Wright KJ, Higgs DM, Belanger AJ, Leis JM. 2008. Auditory

and olfactory abilities of larvae of the Indo-Pacific Coral

Trout Plectropomus leopardus (Lacepede) at settlement.

J Fish Biol 72:2543–56.

Wright KJ, Higgs DM, Cato DH, Leis JM. 2010. Auditory

sensitivity in settlement-stage larvae of coral reef fishes.

Coral Reefs 29:235–43.

Wright KJ, Higgs DM, Leis JM. 2011. Ontogenetic and inter-

specific variation in hearing ability in marine fish larvae.

Mar Ecol Prog Ser 424:1–13.

Yan HY, Fine ML, Horn NS, Colon WE. 2000. Variability in

the role of the gasbladder in fish audition. J Comparative

Physiol A 186:435–45.

Zeiske E, Theisen B, Breuker H. 1992. Structure, development

and evolutionary aspects of the peripheral olfactory system.

In: Hara TJ, editor. Chemoreception in fishes. Amsterdam:

Elsevier. p. 13–39.

Zimmer RK, Derby CD. 2007. Biological Bulletin virtual sym-

posium: the neuroecology of chemical defense. Biol Bull

213:205–7.

How Nemo finds home 843

at University of Sydney on O

ctober 26, 2011http://icb.oxfordjournals.org/

Dow

nloaded from