ArticlePDF Available

The Olfactory System in Fish: Structure, Function, and Role in Behavior

Authors:

Abstract and Figures

The paper reviews recent studies of the olfactory system in fishes. The anatomy and morphology of the olfactory organ in fish with different systematics and modes of life are considered, as well as the cellular composition of the olfactory epithelium, the structure of the olfactory bulb and the olfactory tract. The basic functional parameters of the olfactory system are presented: spectra, sensitivity, the speed of adaptation, the nature and the sources of olfactory signals, their modern classification, basic methodologies for the study of the olfactory function, and characteristics of ventilation of the olfactory cavity. Information about the mechanisms of olfactory orientation in fish and their ability to determine the individual, group, population, and species identity of individuals is systematized. The role of the olfactory reception in reproductive, parental, feeding, defensive, territorial, schooling, and migration behavior is considered. The time course of the development of the olfactory organ in ontogeny is traced, and the terms of the appearance and development of responses to different natural chemical stimuli are determined. The susceptibility of the olfactory organs to negative effects of chemical pollutants is shown. Data are presented on the dynamics of regeneration processes, especially regarding recovery of the olfactory system after acute effects of chemical pollutants.
Content may be subject to copyright.
S180
Journal of Ichthyology, Vol. 44, Suppl. 2, 2004, pp. S180–S223.
Original Russian Text Copyright © 2004 by Kasumyan.
English Translation Copyright © 2004 by MAIK “Nauka/Interperiodica” (Russia).
Chemoreception, the ability to perceive chemical
stimuli, is the oldest evolved way to obtain information
about the environment. It is thought to have appeared
more than 500 million years ago (Hara, 1992a). Not
only fish and other vertebrates, but such primitive
organisms as bacteria and protozoans have well-devel-
oped chemoreception. Chemocommunication is the
exchange of information between animals via chemical
signals. It is hypothesized (Sepp, 1959) that analysis of
the information incoming via chemoreceptors brought
about the evolution of the primary nervous centers and,
later, the central nervous system.
Three independent chemosensory systems are usu-
ally distinguished in fishes: olfaction, taste, and the
general chemical sense (Parker, 1912). All these sen-
sory systems have one thing in common: they perceive
chemical stimuli. Fish, like most other animals, per-
ceive the most diverse information stream with olfac-
tion. The perception of olfactory signals allows fish to
distinguish conspecific individuals, determine their
species and population identity, and brings about differ-
ent behavioral patterns: reproductive, schooling, defen-
sive, migration, parental, agonistic, territorial, etc.
(Malyukina
et al.
, 1969; Kleerkoper, 1969; Liley, 1982;
Døwing, 1986).
Special studies of the olfactory system in fish have
continued for more than one hundred years. The first
works on the morphology of the olfactory organ in fish
were conducted in the 19th century, also in Russia, on
sturgeon. The olfactory receptor cells were described,
and the ability of the fish to respond to olfactory stimuli
and determine their source was shown (Dogel, 1886;
Kleerkoper, 1969). Studies of the olfactory system in
fish became most active during the last 30–40 years,
when different electrophysiological methods, behav-
ioral recording methods, underwater observation, and
electronic microscopy became widely used.
Comprehensive studies of the olfactory system in
fish began in Russia in the mid-1940’s with the works
of A.P. Andriyashev (1944, 1945), followed by
D.S. Pavlov (1959, 1962), and M.P. Aronov (1961,
1962). Morphological studies of the olfactory organ in
fish were conducted from the middle of the 1950s
(Bakhtin and Filyushina, 1974; Pyatkina, 1975, 1976;
Bakhtin, 1977; Bronshtein, 1977). Diverse studies of
the olfactory system in fish were conducted in Moscow
State University from 1967, when a special laboratory
was founded in the Department of Ichthyology
(G.A. Malyukina, G.V. Devitsyna, E.A. Marusov,
A.O. Kasumyan, N.E. Lebedeva, L.S. Chervova,
T.V. Golovkina, N.I. Pashchenko, S.S. Sidorov). Stud-
ies of chemical signalization in fish were conducted at
the Severtsov Institute of Ecology and Evolution, Rus-
sian Academy of Sciences (Vinogradova and Manteifel,
1982, 1984). The results of this large and comprehen-
sive work have been summarized in numerous publica-
tions, including many surveys and analytic reviews (Fle-
rov, 1962; Shmalgauzen, 1962a; Malyukina
et al.
, 1969,
1980; Shparkovskii, 1980; Manteifel, 1987; Pavlov and
Kasumyan, 1994).
This paper reviews the basic results of recent studies
of the olfactory system in fish, considering characteris-
tics and the diversity of its structural organization, func-
The Olfactory System in Fish: Structure, Function,
and Role in Behavior
A. O. Kasumyan
Moscow State University, Vorob’evy gory, Moscow, 119899 Russia
Received March 19, 2004
Abstract
—The paper reviews recent studies of the olfactory system in fishes. The anatomy and morphology of
the olfactory organ in fish with different systematics and modes of life are considered, as well as the cellular
composition of the olfactory epithelium, the structure of the olfactory bulb and the olfactory tract. The basic
functional parameters of the olfactory system are presented: spectra, sensitivity, the speed of adaptation, the
nature and the sources of olfactory signals, their modern classification, basic methodologies for the study of the
olfactory function, and characteristics of ventilation of the olfactory cavity. Information about the mechanisms
of olfactory orientation in fish and their ability to determine the individual, group, population, and species iden-
tity of individuals is systematized. The role of the olfactory reception in reproductive, parental, feeding, defen-
sive, territorial, schooling, and migration behavior is considered. The time course of the development of the
olfactory organ in ontogeny is traced, and the terms of the appearance and development of responses to different
natural chemical stimuli are determined. The susceptibility of the olfactory organs to negative effects of chem-
ical pollutants is shown. Data are presented on the dynamics of regeneration processes, especially regarding
recovery of the olfactory system after acute effects of chemical pollutants.
JOURNAL OF ICHTHYOLOGY
Vol. 44
Suppl. 2
2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S181
tional parameters and the role in different behavioral
patterns, as well as its formation during ontogeny.
The Structure of the Olfactory Organ
The olfactory organ is a peripheral part of the olfac-
tory system. It is a paired structure. The olfactory organ
in most fish species is well developed and is located on
the dorsal surface of the head to the rostrum and medial
1 mm
(a)
(b)
Fig. 1.
Head of minnow
Phoxinus phoxinus.
(a) lateral
view; (b) view from above (from Pashchenko and
Kasumyan, 1983).
nostrilbridgenostril
Anterior Nasal Posterior
Olfactory
cavity
Raphe
Accessory
olfactory sac
Fig. 2.
Basic structural elements of the olfactory organ in
fish (sagittal section) (from Zeiske
et al.
, 1992).
2 mm
Nasopharyngeal duct
Olfactory lamellae
a-a
Anterior
Nasal duct Nasal valve
a-a
nostril
Fig. 3.
Head of hagfish
Myxine glutinosa
, sagittal sec-
tion.
a-a
olfactory organ, frontal section (from Zeiske
et al.
, 1992).
in en
(‡)
(b)
Olfactory organ
in en
Fig. 4.
Protopterus annectens
: (a) head (sagittal section),
(b) olfactory organ. “in” anterior olfactory opening, “en”
posterior olfactory opening (from Zeiske
et al.
, 1992).
S182
JOURNAL OF ICHTHYOLOGY
Vol. 44
Suppl. 2
2004
KASUMYAN
of the eyes (Fig. 1). Water gets into the olfactory organ,
the nasal or olfactory cavity via the anterior opening or
the anterior nostril, and gets out via the posterior olfac-
tory opening or posterior nostril. These openings are
divided by a skin nasal bridge or nasal bridge. The nasal
bridge of many fishes has a skinny nasal ridge directing
water into the anterior nostril. Olfactory lamellae form-
ing the olfactory rosette are located at the bottom of the
nasal cavity. The surface of the lamella is covered by
the olfactory epithelium. Certain fish have accessory or
ventilation olfactory cavities (Fig. 2).
This general scheme of the olfactory organ has
many different types and modifications (Zeiske
et al.
,
1992). The structure of the olfactory organ may signif-
icantly differ in representatives of different taxonomic
groups of fish and cyclostomata (hagfish and lamprey)
depending on their mode of life and behavior.
The enormous diversity of the olfactory organ, and
its size, position and structure in fish may be demon-
strated by many examples. For instance, in cyclosto-
mata,
Petromyzontes
lampreys and Myxini, unlike fish,
the olfactory organ represents an unpaired structure
(Fig. 3). Based on this morphological trait, lampreys
and myxini have earlier been included (by Haeckel)
into the group Monorhini, and all other vertebrates with
paired olfactory organ were collapsed into the group
Dirhini. But monorhiny in cyclostomata has a second-
ary origin because the olfactory nerve in lampreys and
myxini is paired and the olfactory organ is really split
into two parts by a partition (Tretyakov, 1915).
The olfactory cavities in most fish are not linked
with the oral cavity. However, this link was found in
myxini and some fish, e.g., chimaera
Chimaera mon-
strata
, dipnoans (Dipnoi), and certain teleosts such as
the
Astroscopus
stargazers,
Gymnodraco acuticeps
,
and in fish of the family Echelidae and Ochichthyidae
(Anguilliformes), so that water can pass into the oral
cavity from the nasal cavity during breathing (Fig. 4)
(Dahlgren, 1908; Mees, 1962; Holl, 1973; Jakubowski,
1975; Derivot, 1984). Myxini and stargazers have a
special valve in the nasopharingeal canal, directing the
water flow from the nasal cavity to the oral cavity
(Fig. 3) (Døwing and Holmbers, 1974). The nasophar-
ingeal canal is blind in lampreys. It does not open into
the oral cavity, but ends at the first gill openings
(Tretyakov, 1915, Kleerkoper and van Erkel, 1960;
Bermar, 1969; Theisen, 1976).
In such ancient fish as dipnoans and cartilaginous
Chondrichthys, the olfactory organs are located at the
ventral surface of the head (Fig. 5), whereas in repre-
sentatives of all other taxonomic groups, olfactory
organs are located at the dorsal surface of the head
(Shmalgauzen, 1962a; Kleerekoper, 1969; Bronshtein,
1977; Yamamoto, 1982).
The structural diversity of the olfactory organ is pro-
nounced in many other morphological characteristics,
such as the position of the inflow and outflow openings
leading to the olfactory organ, their size, the presence
of the nasal bridge dividing these openings, and the size
of the ridge. For example, in carps Cyprinidae, perches
Percidae, cods Gadidae, salmons Salmonidae, and
2 mm
Fig. 5.
Head of the lemon shark
Negaprion brevirostrus
, view from below. The box surrounds the inlet opening and the large skin
outgrowth nasal flap partially closes it (from Zeiske
et al.
, 1096, 1987).
(a)
(b)
Fig. 6.
Head of juvenile stellate sturgeon
Acipenser stella-
tus
: (a) normal olfactory organ; (b) abnormal olfactory
organ (without the nasal bridge) (from Kasumyan, 2002).
JOURNAL OF ICHTHYOLOGY
Vol. 44
Suppl. 2
2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S183
many other fish species, nostrils are well pronounced
and located close to each other and the nasal bridge
bears a large ridge, forming a kind of funnel which
directs water into the nasal cavity (Fig. 1) (Pashchenko
and Kasumyan, 1983; Døving, 1986; Zeitske
et al.
,
1992). Sturgeons Acipenseridae (Fig. 6) have a nasal
bridge, but the ridge is very small or absent (Kasumyan,
2002). When these fish are raised in artificial condi-
tions, this ridge may be partly or completely destroyed
by pathogenic bacteria. The proportion of artificially
raised sturgeon juveniles without the nasal ridge may
reach several dozen percent. It is accepted that such
abnormalities are associated with poor water quality
and the presence of pollutants (Shmalgauzen, 1957,
1962b; Podushka, 1999; Goryunova
et al.
, 2000).
The nasal bridge is not formed in sharks (Fig. 5), but
the lateral and medial edges of the olfactory opening of
these fish have well developed skin outgrowths (lamel-
lae) directed toward each other. They do not coalesce
but overlap each other. The sizes of these outgrowth are
so large that in many species of sharks and rays
(
Scyliorhinus canicula
,
Raja clavata
,
Trygon pastinaca
etc.) they cover a special nasal-oral groove, going from
the nasal to oral cavity and providing a link between
them. These outgrowths are motile in certain sharks
(Shmalgauzen, 1962a; Zeiske
et al.
, 1986, 1987).
in
in
in
in en
en
en
en
n
Moray eel,
Rhinomuraena ambonensis
Japanese lycodont,
Lycodontis javanicus
fifteen-spined stickleback,
Spinachia spinachia
tetradon,
Tetraodon nigropunctatus
Siphonostoma typhle
bedotia,
Bedotia geayi
Fig. 7.
Positioning and relative size of the olfactory organ in fish (from Zeiske
et al.
, 1992).
S184
JOURNAL OF ICHTHYOLOGY
Vol. 44
Suppl. 2
2004
KASUMYAN
The olfactory organ is not conspicuous in all spe-
cies. In certain species, the openings leading to the
olfactory organ are so small that they are quite difficult
to discern on the head of the fish, as in
Bedotia geayi.
In
Syphonostoma typhle
they are not only small but also
closely adjoin each other (Fig. 7) (Melinkat and Zeiske,
1979). The fifteen-spined stickleback
Spinachia spina-
chia
has only one olfactory opening, as in the garfish
Belone belone
, but, unlike the stickleback, it is wide
(Fig. 8) (Theisen
et al.
, 1980; Theisen, 1982). An uni-
tary olfactory opening not separated into anterior and
posterior nostrils can be found in other sticklebacks
Gasterosteidae, garfishes Belonidae, Hemirhamphidae,
Pomacentridae,
Zoarces viviparus
,
Liparis montaqui
,
pike
Esox lucius
, etc. (Pipping, 1926; Liermann, 1933;
Holl and Meinel, 1968; Theisen
et al.
, 1980). Only the
posterior opening, in the form of a narrow rudimentary
slit, is retained in greenlings
Hexagrammos.
An olfac-
tory organ with only one opening is called monoter-
mous, while the more typical ditermous has two olfac-
tory openings (Døving, 1986).
In some fish, mostly bottom-dwelling and inactive,
burrowing in the bottom substrate or hiding among
stones and weeds, nostrils have the form of short or
long pipes. Such nostrils are characteristic of many
moray eels, such as
Lycodontis javanicus
and
Rhino-
muraena ambonensis
(Fig. 7), and several species of
eels Anguillidae, catfishes
Siluridae
, snakehead
Channa
, Polypteridae,
Gaidropsarus
, etc. (Holl
et al.
,
1970). In some cases, these pipes are separated from
each other and have special valves, inflow in the ante-
rior pipe and outflow in the posterior pipe. In dipnoans,
posterior nostrils open into the oral cavity, but the fish
catch the air with the oral cavity rather than the olfac-
tory organ (Shmalgauzen, 1962a).
In Tetraodontidae,
Tetraodon nigropunctatus
, nos-
trils are completely absent and the olfactory organ is
located within tentacle-like outgrowths above the head
surface. The inner cavity of these outgrowths is covered
with olfactory epithelium (Fig. 7) (Wiedersheim, 1887,
cit in Zeiske
et al.
, 1992; Shmalgauzen, 1962a).
Fish differ by the presence or absence of accessory
olfactory sacs. These relatively large structures are
characteristic of relatively inactive bottom species or
species living in still water, dipnoans, Polypteridae,
flatfishes Pleuronectidae, eels Anguilliformes, perches
Perciformes and several other fish (Eaton, 1956; Holl
et al.
, 1970; Kapoor and Ojha, 1973; Jakubowski,
1975). But there are many exceptions from this general
rule. For example, accessory sacs are present in Clu-
peidae, Salmonidae, Mugilidae, Scombridae and sev-
eral other pelagic fish with high swimming activity
(Burne, 1909; Shmalgauzen, 1962a; Betmar, 1972).
Accessory sacs are primarily used for the ventilation
of the nasal cavity, therefore they are often called olfac-
tory ventilation sacs (Zeiske
et al.
, 1992). Another
function of olfactory sacs is the production of olfactory
mucus by numerous mucous cells in the epithelium of
the accessory sacs. Receptor cells are absent in this epi-
thelium (Shmalgauzen, 1962a; Kapoor and Ojha, 1972,
1973; Bronshtein, 1977). The olfactory organ may be
linked with the oral cavity when the wall at the distal
area of the ventral accessory sac is perforated, as is
observed in the American stargazer
Astroscopus gutta-
tus
or in
Gymnodraco acuticeps
(Atz, 1952a, 1952b;
Jakubowski, 1975).
The number of accessory sacs varies from one to
four. They may deviate from different parts of the basic
olfactory cavity, and are named depending on the scull
bones which they closely adjoin: lacrymal, ethmoidal,
suborbital, supraorbital, etc. For example, in flatfishes,
clupeids, and salmons, accessory sacs open to the nasal
cavity from the posterior part, in certain perches, in the
rostral part. Accessory sacs are usually smaller than the
basic nasal sacs, but in smelts
Osmerus
, the sole acces-
sory sac is several times larger than the basic.
There is a pronounced variability in the number of
olfactory lamellae (Table 1). In cyprinids, sturgeons,
salmonids, percids, and several other groups of fishes,
the number of lamellae in the olfactory organ is several
dozen, in some species, such as
Holopagus guentheri
from the family Lutjanidae, their number reaches 230
(Pfeiffer, 1964). However, fish exist in which olfactory
lamellae are few or are completely absent. For example,
from one to three lamellae are found in the olfactory
organ of the rainbow fish Melanotaeniidae,
Hemiram-
phus sajori
, in the fifteen-spined stickleback
Spinachia
spinachia
, and the Washington sand lance
Ammodytes
personatus.
The mosquito fish
Gambusia agiffi
, guppy
Poecilia reticulata
, sand lance
Ammodytes lancea
,
Syphonostoma typhle
have no olfactory lamellae (Lier-
mann, 1933; Holl, 1965). Garfish
Belone belone
also
has no olfactory lamellae. The olfactory organ of these
rapidly swimming fish represents a depression with a
large club-shaped outgrowth on the bottom. Its rostral
surface, directed to the water current, has shallow
0.5 mm
Fig. 8.
Olfactory organ of garfish
Belone belone.
Arrow
points the rostro-caudal direction (from Theisen
et al.
,
1980).
JOURNAL OF ICHTHYOLOGY
Vol. 44
Suppl. 2
2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S185
lamellae and ridges which only approximately resemble
true olfactory lamellae (Fig. 8) (Theisen
et al.
, 1980).
The olfactory organs of flatfishes and most other
representatives of Pleuronectiformes located at the
eyed and blind sides of the body significantly differ in
size and number of lamellae in the olfactory rosette. In
Pleuronectes platessa
,
Solea solea
, and many other
representatives of this genus, the number of lamellae in
the olfactory organ at the eyed side is 1.5–2.0 times
greater than at the blind side (Table 1) (Harvey, 1996).
In fish with symmetrical body shape, rosettes in the left
and right organs have the same number of lamellae
(Table 2).
The basic function of olfactory lamellae is increase
the area of the olfactory epithelium bearing the receptor
cells which perceive chemical stimuli. It was counted
that the overall area of the olfactory rosette in
Calam-
oichhys sp.
, with a length of 41 cm, is 3200 mm
2
(Pfe-
iffer, 1968). The same function is realized by secondary
lamellae on the surface of basic olfactory lamellae. The
secondary lamellae are well pronounced in salmonids,
garpikes, sturgeons, burbot
Lota lota
, pike
Esox lucius
(Pfeiffer, 1963; Holl, 1965; Devitsyna, 1972; Zeiske
et al.
, 1987; Devitsyna and Kazhlayev, 1992, 1993;
Hara
et al.
, 1993; Chen and Arratia, 1994). Secondary
lamellae may be oriented parallel or perpendicular to
the base of the primary lamella, and in some cases,
along the diagonal (Chen and Arratia, 1994). In sharks,
there are also tertiary lamellae (Fig. 9) (Zeiske
et al.
,
1987). The lateral surface of olfactory lamellae in the
striped eel catfish Plotosus lineatus represents a system
of areas isolated from each other by ascending ridges
(Fig. 10) (Theisen et al., 1991).
Olfactory lamellae form the olfactory rosette (Fig.
11). There exist several types of olfactory rosettes dif-
fering by the position of lamellae: radial (Acipens-
eridae, Esocidae), arrow-like (Cyprinidae, Salmo-
nidae), parallel (Pleuronectidae, Lepisosteus, Lophii-
formes), and bilateral (Anguilliformes, Siluriformes,
Dipnoi, Amia, Chondrychthyes). There exist numerous
transitory forms between these basic types (Fig 12).
Closely related species of fish, for example, belonging to
the same family, usually have olfactory rosettes of the
same or similar type (Kleerekoper, 1969; Yamamoto,
1982; Zeiske et al., 1992; Chen and Arratia, 1994).
In spite of significant morphological variability, sex
dimorphism in the structure of the olfactory organ is
Table 1. Number of lamellae in the olfactory organ of fish of
different systematic groups (from Yamamoto, 1982)
Orders and species Number of lamellae
Anguilliformes:
Anguilla japonica 108
Conger myriaster 120
Microdonophis erado 112
Gymnothorax kikado 102
Pleuronectiformes:
Paralichthys olivaccus 22/19
Pleuronichthys cornutus 15/9
Zebrias zebra 40/30
Areliscus joyneri 64/34
Clupeiformes:
Clupandon punctatus 26
Sardinops melanosticta 24
Engraulis japonica 28
Scorpaeniformes:
Sebastes inermis 16
Minous monodactylis 23
Hypodytes rubripinnis 7
Hexagrammus otakii 21
Platycephalus indicus 20
Vellitor centopomus 12
Tetraodotiformes:
Triacanthus brevirostris 22
Fugu niphobles 6
Navodon modestus 1
Rudarius arcodes 0
Ostracion tuberculatus 0
Atheriniformes:
Atherion elymus 3
Iso hawaiiensis 3
Hemiramphus sajori 1
Cheilopogon agoo 1
Cololabis saira 1
Orysias latipes 0
Gambusia affinis 0
Note: For flatfishes, the number of lamellae in the olfactory organs
on the eyed and blind sides of the body is presented.
Table 2. The number of lamellae in the olfactory rosette of left
and right olfactory organs of Claris batrachus (from Goel, 1980)
Body length, cm Number of lamellae
left organ right organ
9.5 26 27
12.0 33 33
12.6 34 34
13.8 43 43
14.3 45 45
15.0 50 50
16.1 51 52
16.5 55 55
S186
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
rather uncommon in fish. The exception from this rule
is only such extremely specialized species of fish as
deep-sea anglers Lophiiformes and a few other bathy-
pelagic and mesopelagic fishes: luminous anchovy
Myctophiformes, Saccopharyngidae, Monognathidae,
Gonostomatidae (Marshall et al., 1967, 1979; Caruso,
1975; Bertelsen, 1980; Nielsen and Bertelsen, 1985; Ber-
telsen and Nielsen, 1987; Baird et al., 1990). The olfactory
organ and other parts of the olfactory system in males of
many of these fish are much more developed than in
females (Fig. 13). This is associated with the extremely
important role of olfaction in the search by males for ripe
females of these species during the reproductive period.
Olfactory Epithelium
The olfactory epithelium covers the olfactory lamel-
lae. Its width in different fish ranges from 20 to 130 µm
(Bronshtein, 1977). Cells of different types are
included into the epithelium, mostly receptor, support,
mucous, and basal (Fig. 14).
Receptor cells. Olfactory receptor cells are the pri-
mary sensitive cells, that is, they represent nervous cells
(sensory neurons). Receptor cells have a thick dendrite,
which goes out from the central (nuclear) part of the
neuron and reaches the surface of the epithelium. The
apical part of the dendrite terminates at a rounded tip
(knob) which projects over the surface of the epithe-
lium. In fish, receptor cells in the olfactory epithelium
are represented by two different types: ciliated and
0.5mm
1 mm
(a)
(b)
Fig. 9. Secondary lamellae: the olfactory rosette of the Arctic
char Salvelinus alpinus (a) and the olfactory lamella of the lemon
shark Negaprion brevirostrus (b) (from Zeiske et al., 1987).
0.1 mm
0.02 mm
(a)
(b)
Fig. 10. Lateral surface of the olfactory lamella in the
striped eel catfish Plotosus lineatus at different magnifica-
tion (from Theisen et al., 1991).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S187
microvillous receptor cells (Bernshtein, 1977; Yama-
moto, 1982; Zeiske et al., 1992). Recently, a third type
of cells has been found in the olfactory epithelium of
different fish species: crypt cells. They are relatively
rare and their knob does not reach the surface of the epi-
thelium but opens in a small cavity located directly
under the external surface of it (Hansen and Zeiske,
1998; Hansen and Finger, 2000; Zeiske et al., 2003).
The knob of receptor ciliated cells bears weakly
motile cilia. Their number ranges from 4 to 20 and the
length reaches 5–10 µm, and diameter, 0.25–0.30 µm.
The number of microvilli in microvillous receptor cells
is significantly larger, from 30 to 80. Their length
ranges from 0.5 to 3–5 µm, diameter about 0.1 µm
(Bannister, 1965; Gemne and Døving, 1969; Schulte,
1972; Bronshtein, 1977; Sorensen and Caprio, 1998).
Numerous ciliated–microvillous receptor cells can be
found in the olfactory epithelium of certain fishes (e.g.
in sturgeons), the knobs of these cells bear both cilia
and microvilli (Pyatkina, 1975, 1976). Cilia of receptor
cells move slowly and asynchronously and are usually
stretched along the surface of the epithelium. A long
central process, the axon, goes out from the opposite
side of the receptor cell (Fig. 14).
Special protein receptor molecules are contained in
the cellular membrane of cilia and microvilli. When the
receptor protein contacts the olfactory substance, a
complicated cascade of molecular transformations
occurs in the cell, which is called the signal transduc-
tion process. This process changes the functioning of
the ion channels in the cell, generation of receptor
potential, and its distribution as an electric nervous
impulse from the generation location to the primary and
secondary olfactory centers (Brand and Bruch, 1992).
The central (nuclear) part of the receptor cell is
located in the olfactory epithelium at different levels.
Five such layers are usually distinguished in the epithe-
lium. The lower the layer with the nucleus of the recep-
tor cell, the longer its dendrite, which reaches the outer
surface of the epithelium. It was found that the nuclear
part of ciliated receptor cells is usually located in
deeper layers of the olfactory epithelium than in
microvillous. Therefore, its dendrites are longer. The
shortest dendrites are characteristic of crypt receptor
cells, which are located in the upper layer of the epithe-
lium (Hansen and Finger, 2000; Hamdani et al., 2001a;
Hamdani and Døving, 2002).
Receptor cells occur in the olfactory epithelium in
areas covering the side surface of olfactory lamellae.
They are absent on ridges of lamellae, and on vaults of
the olfactory cavity. Areas of the olfactory epithelium
bearing receptor cells are named receptor zones or
receptor (sensory) epithelium. These zones alternate
Fig. 11. Olfactory rosette of the goldfish Carassius auratus.
Gambusia affinis, Poecilia reticulata
Ammodytes personatus
Acanthogobius, Odontamblyopsys, Navodon
Gasterosteus aculeatus
Acipenseridae, Esox
Salmonidae, Clupeidae
Cyprinidae
Anguilliformes, Amia, Lota
Plotosus, Pleuronectes, Lepisosteus
Lophiiformes, Fugu niphobles
Cololabis, Cheilopogon, Hemiramphus
Fig. 12. Types of olfactory rosettes in fish. Arrow points the
rostral (R) and caudal (C) directions.
S188
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
with zones of non-receptor epithelium. The distribution
of receptor cells at the lateral surface of lamellae in dif-
ferent fish species may differ. Continuous, large-zone,
fine-zone, and irregular types of sensory epithelium
distribution are usually distinguished (Fig. 15) (Yama-
moto and Ueda, 1979; Yamamoto, 1982).
The average density of receptor cells in the receptor
epithelium ranges from 20–30 thousand per mm2 to
400–500 thousand per mm2, and their total number
ranges between several dozens to a hundred thousand to
several dozen million or more (Holl, 1965; Holl et al.,
1970; Zeiske et al., 1976, 1979; Thommesen, 1983;
Devitsyna and Attar, 1988). The density of ciliated and
microvillous cells may differ in different areas of olfac-
tory lamellae but the total number of ciliated cells is
two times more than microvillous (Thommesen, 1983).
The largest number of receptor cells has been found in
the common wels Silurus glanis which approach cer-
tain mammals (dog) with well-developed olfaction
(Table 3). It should be noted that linear growth in fish
and the associated increase of the number of receptor
cells occurs throughout their whole life, therefore, large
individuals have larger olfactory organs and more
receptor cells (Pashchenko, 1986). All types of receptor
cells are present in most groups of fish, teleosts, carti-
laginous, Chondrostei (sturgeons), and in myxini
(Theissen, 1976; Yamamoto, 1982; Eistein, 1982;
Zeiske et al., 1992). But Elasmobranchs and some dip-
1
2
3
(a) (b)
(c)
1
Fig. 13. Olfactory organ and brain of Cyclothone microdon
(Gnatostomidae): (a) and (b) male, view from above and lat-
eral, (c) female, from above; (1) olfactory rosette, (2) olfac-
tory bulb, (3) telencephalon (from Marshall, 1979).
mc
k
m
c
f
sc
crc
aon mrc cc bc
Fig. 14. Basic cell elements in the olfactory epithelium of
fish: a. axons of receptor cells; (c) club; (bc) basal cells;
(f) flagellae; (crc) ciliated receptor cell; (k) kinocilia;
(m) microvilli; (cc) ciliate cell; (mrc) microvillous receptor
cell; (sc) supporting cells; (on) olfactory nerve;
(mc) mucous cell (from Zeiske et al., 1992).
(‡) (b) (c) (d)
Fig. 15. Types of the distribution of the receptor epithelium
on lateral side of olfactory lamellae: (a) continuous; (b)
large-zonal; (c) irregular; (d) small-zonal (from Yamamoto,
1982).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S189
noans (Neoceratodus fosteri) have only microvillous
receptor cells, whereas Polypteridae, coelacanths,
Crosspterigi, the river lamprey Lampetra fluviatilis,
marine lamprey Pteromyzon marinus and probably other
lampreys have only ciliated cells (Thornhill, 1967; Reese
and Brightman, 1970; Theissen, 1972; Bronstein, 1976;
Zeiske et al., 1987; Van Denbossche et al., 1995).
Ciliated, microvillous and crypt cells specialize in
perception of different groups of chemical stimuli
(Thommesen, 1982, 1983; Hamdani et al., 2000,
2001a, 2001b; Sato and Suzuki, 2001; Hamdani and
Døving, 2002, 2003; Weltzen et al., 2003). Within all
these groups, the cells probably have further functional
specialization. Unlike amphibians, reptiles and other
vertebrates, this specialization in fish is not associated
with zonality in the distribution of different receptor
types in the olfactory epithelium (Yamamoto, 1982;
Ngai et al., 1993).
Supporting cells. Supporting cells surrounding the
receptor cells are represented in the olfactory epithe-
lium by two different types: the supporting cells, i.e.,
cells with a free surface not bearing any specialized
structures, and ciliated cells, i.e., cells bearing numer-
ous kinocilia on the surface facing the cavity of the
nasal sac. The number of kinocilia in one ciliate cell
may reach several dozen, they have a length of up to
5 µm and a diameter 0.2–0.3 µm. Kinocilia of the cili-
ated cells move synchronously, and their activity is
responsible for the ventilation of the olfactory cavity in
areas located between close lamellae of the olfactory
rosette, and in many fish species, ventilation of the
whole olfactory organ (Døving, 1986). The function of
these cells is very important because they provide for
the transport of water which carries the signal to the
receptor cells.
Goblet cells. Goblet cells are very numerous in the
olfactory epithelium. They are located in its upper layer
and may differ in size. Small and large mucous cells are
usually distinguished. The function of these cells is
associated with the production of a special olfactory
mucus, which covers the surface of the olfactory epi-
thelium. This mucous layer or glycocalix completely
covers cilia and microvilli of the receptor cells and the
bases of kinocilia of the ciliated cells. The olfactory
mucus includes mucopolysaccharides, proteins, lipids,
and different ions (Doroshenko and Pinchuk, 1974;
Bronshtein, 1977).
It is hypothesized that glycocalix plays an important
role in chemoreception and that molecules of signal
substances are first introduced into the mucus, interact
with its components, and only then reach the receptor
areas in the cellular membrane of cilia and microvilli.
The mucus creates optimal conditions for the molecular
processes involved into the interaction between the sig-
Table 3. The number of receptor cells in the olfactory organ of some species of cyclostomata, fish, and other vertebrates
Species Number of receptor cells Body length or age Source
Cyclostomata
river lamprey Petromyzon marinus 800 Kleerokoper, Van Erkel, 1960
Fishes
verkhovka Leucaspius delineatus 35 5.5 cm Devitsyna and Attar, 1988
sand lance Ammodytes lancea 200–300 Holl, 1965
minnow Phoxinus phoxinus 900 –
river perch Perca fluviatilis 2.000–3.000 –
trigla Trigla corax 4.000 –
pike Esox lucius 6.000 54 cm Devitsyna and Attar, 1988
blue bream Abramis ballerus 6.000 30 cm
burbot Lota lota 11.000 Gemne, Døving, 1969
bream Abramis brama 27.000 40 cm Devitsyna and Attar, 1988
wels Silurus glanis 155.000 62 cm
Amphibians
fire-bellied toad Bombina bombina 500 adult Khmelevskaya, 1972
gray toad Bufo bufo 2.400 ″″
marsh frog Rana ridibunda 2.650 ″″
Mammals
humans Homo sapiens 10.000 Brunn, 1892
rabbit Oryctolagus cuniculus 100.000 Allison, Warwick, 1949
dog Canis familiaris 224.000 Neyhaus, Müller, 1954
S190
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
nal substance (ligand) and the receptor proteins
(Sorensen and Caprio, 1998).
It is thought that the olfactory mucus has important
protective functions, protecting flagella and microvilli
of receptor cells from mechanical disturbances by
small particles transmitted by water currents. The
important protective function of the mucous is
revealed by the fact that secretion of olfactory mucus
significantly increases when the water is contaminated
by various chemical agents (detergents, heavy met-
als). The production of olfactory mucus in fish of the
same family is most pronounced in marine (cod
Gadus morhua, navaga Eleginus navaga) than in
freshwater (burbot Lota lota) representatives
(Devitsyna, 1972).
Basal cells. Basal cells are located in the deepest
layer of the olfactory epithelium at the basal membrane.
New cells of the olfactory epithelium, receptor, sup-
porting, and mucous, having limited longevity, develop
and differentiate from basal cells. Receptor cells are
renewed every 7–10 days. Old cells of the epithelium
degenerate and their content is either secreted into the
nasal cavity or is phagocytized by lymphatic cells (Gra-
ziadei and Monti-Graziadei, 1978; Cancalon, 1982;
Pashchenko and Kasumyan, 1984).
The ability of the olfactory epithelium, primarily the
receptor cells, for constant renewal is an extremely
important adaptation, allowing the fish to retain the
capacity to obtain olfactory information when the prob-
ability of receptor cell destruction and disturbances is
very high. The olfactory epithelium, including the
receptor cells, directly opens into the environment and
is weakly protected from its negative effects. Particles
suspended in water may damage the epithelium (abra-
sive effects). Iridioviruses and bacteria, as well as dif-
ferent parasitic organisms, microsporidia, fungi, and
small crustaceans may also cause significant distur-
bances of the olfactory epithelium. Receptor cells may
be disturbed after a sharp change of water salinity in
anadromous and catadromous migrants. Deep struc-
tural disturbances and complete degeneration of many
elements of the olfactory epithelium may be caused by
chemical contaminants (detergents, heavy metals, low
pH) (Bertmar, 1972; Devitsyna, 1977; Doroshenko and
Pinchuk, 1978; Pashchenko and Kasumyan, 1984;
Baatrup and Døving, 1990; Klaprat et al., 1992; Morri-
son and Plumb, 1994; Watson et al., 1998).
The Olfactory Nerve
and the Olfactory Tract
Axons of all receptor cells after the basal membrane
of the olfactory epithelium are joined into the single
nerve, the n. olfactorius (pair I of brain nerves). The
olfactory nerve links the olfactory organ with the pri-
mary olfactory centers: paired olfactory bulbs (bulbus
olfactorius) to which the olfactory information comes
as electric impulses (Fig. 16).
Nervous fibers, i.e., axons of receptor cells, branch
in the olfactory bulb and form a surface fibrous layer.
The glomerular layer of the olfactory bulb, formed by
glomerules, is located under it. Glomerules represent a
complex interlacement of myriad branches of the olfac-
tory nerve and dendrites of mitral cells interacting with
synaptic contacts (Døving, 1986; Satou, 1992;
Sorensen and Caprio, 1998). The relationships between
the number of axons of receptor cells and the number of
mitral cells ranges from 40 : 1 in verkhovka Leucaspius
delineatus to 1000/4000 : 1 in cod Gadus morhua, blue
bream Abramis ballerus, and bream Abramis brama. In
the common wels Silurus glanis, the relationship
between these cells exceeds 12000 : 1 (Døving and
Gemne, 1965; Devitsyna, 1977; Devitsyna and Attar,
1988). It is thought that the higher the ratio, the more
developed are the functional capacities of the olfactory
system (Døving, 1986).
The layer of mitral and ruffed cells is located
below the layer of glomerulas. Each intercalary cell
connects several mitral cells and each mitral cell, in
turn, may have contacts with different intercalary
cells, which causes the formation of large neuronal
assemblages.
An aggregation of granular cells forming the ante-
rior olfactory nucleus (nucleus olfactorius anterior) is
situated under the layer of mitral and intercalary cells.
Granular cells have synaptic contacts with mitral and
intercalary cells. Centrifugal effects from the secondary
olfactory center (nucleus olfactorius posterior), located
in the telencephalon and involved in the central regula-
tion of the olfactory bulb, affect granular cells. The pri-
mary and secondary olfactory centers are connected by
a tract formed by axons of mitral and intercalary cells,
along with centrifugal fibers incoming to granular cells
1
2
3
4
5
6
7
Fig. 16. Scheme of the olfactory bulb in fish: (1) axons of
olfactory receptor cells; (2) glomerula; (3) mitral cell;
(4) ruffed cell; (5) granule cell; (6) axons forming the olfac-
tory tract; (7) centrifugal axons (from Satou, 1992).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S191
(they belong to the system of the 0 pair of brain nerves,
terminal nerve n. terminalis) (Sorensen and Caprio,
1998). Lateral and medial fibers are distinguished in the
olfactory tract, each of which, in turn, is separated into
smaller segments (Sheldon, 1912; Holmgren, 1920;
Døving and Gemne, 1965). Olfactory bulbs are con-
nected by interbulbar links passing via the forebrain
(Satou, 1992).
In most fish species, olfactory bulbs are located
adjacent to the forebrain and the olfactory nerve is quite
long. Such bulbs are called sessille. But in cyprinids,
catfishes, cods, and sharks, holocephali and Mormy-
ridae, olfactory bulbs are remote from the forebrain and
are adjacent to the olfactory organ (Døving, 1986).
Olfactory nerves in these fish are short but the olfactory
tracts, going from olfactory bulbs to olfactory centers in
the forebrain, are of significant length. Such bulbs are
called stalked (Fig. 17). In Characidae and some other
fish, olfactory bulbs are located in the middle between
the olfactory organ and the forebrain.
Morphological and functional zonality has been
found in the olfactory bulb of fish in recent years (Hara
and Zang, 1996, 1998; Laberge and Hara, 2001;
Nikonov and Caprio, 2001). For example, the axons of
flagellated receptor cells come to the ventral surface of
the olfactory bulb, whereas axons of microvillous
receptor cells arrive at the dorsal surface (Morita and
Finger, 1998). When the olfactory epithelium is stimu-
lated by different chemical substances, the electric
activity is recorded in different areas of the olfactory
bulb. For example, in the channel catfish Ictalurus
punctatus, electric activity goes to the medial part of the
olfactory bulb via axons of ciliated cells sensitive to
fatty acids, and via axons of cells sensitive to free
amino acids, electric activity goes to the ventral part of
the olfactory bulb (Hansen et al., 2003). Signals caus-
ing different behavioral patterns come to secondary
centers via different fibers of the olfactory tract (Døv-
ing and Selset, 1980). For example, olfactory signals
bringing about food search behavior in the silver carp
Carassius carassius come via the lateral fiber of the
olfactory tract whereas signals causing defensive
behavior, via the medial segment of the medial fiber of
the olfactory tract. The lateral segment of the medial
fiber may be involved in the mediation of chemical sig-
nals responsible for the stimulation of sexual behavior
(Døving and Selset, 1980; Stacey and Kyle, 1983;
Hamdani et al., 2000, 2001b; Hamdani and Døving,
2002, 2003; Weltzein et al., 2003). Studies of topo-
graphic and functional specialization in the olfactory
system of fish are continued quite intensely. They are
very important for understanding the evolutionary path-
ways of the secondary olfactory system: the vomerona-
sal (Jacobian) organ, lacking in fish (Eistein, 1992). In
mammals, this highly specialized system is involved in
the perception of sexual signals (Meredith and Fernan-
dez-Fewell, 1994; Døving and Trotier, 1998).
Ventilation of the Olfactory Organ
The water, bearing olfactory stimuli, must enter the
olfactory organ to reach the receptor cells. The different
information which is hidden in chemical stimuli will
only be obtained by the fish in this case, allowing them
to perform the necessary behavioral response. The ven-
tilation of the nasal cavity, i.e., the inflow of water, is
conducted by several ways (Pipping, 1927).
The simplest and the most widespread way is asso-
ciated with the directed flow of water created either
when the fish swims or due to the current. This method
is called pressure ventilation. The water flows into the
olfactory organ due to anatomical characteristics of the
olfactory organ, mostly nostrils. The anterior nostril,
via which the water enters the olfactory cavity, usually
is a kind of short pipe, round in section, with quite hard
and widened walls (Fig. 18). Therefore, the walls of the
12
34
5
4(a)
(b)
Fig. 17. Types of olfactory bulbs in fish. Brain and structural
elements of the olfactory system in the Arctic char Salveli-
nus alpinus (a) and catfish Silurus glanis (b): (1) telenceph-
alon; (2) olfactory bulb; (3) olfactory tract; (4) olfactory
nerve; (5) olfactory organ (from Døving, 1989 and
Jakubowski and Kunysz, 1979, respectively).
S192
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
pipe do not occlude by the force of the current, and the
aperture of the nostril is retained (Pashchenko and
Kasumyan, 1986). The posterior nostril in most fish
species is larger in size, usually has an elongated shape
and its walls are thicker. The olfactory rosette located at
the bottom of the olfactory sac may be seen via the wide
aperture of the posterior nostril in such fish as stur-
geons, many cods, etc. (Kasumyan, 2002). In fish with
a monotremous olfactory organ, pressure ventilation is
also possible, but it has a certain specificity. If the single
olfactory opening is large, and the olfactory cavity rep-
resents a depression, water exchange occurs even after
very small movements of the fish (garfish Belone
belone and pike Esox lucius). When the single olfactory
opening is small in size, as in fifteen-spined stickleback
Spinachia spinachia (Fig. 7), the directed movement of
water and water exchange (absence of recirculation)
depend on the specific internal architecture of the organ
(Fig. 18) (Theisen, 1982).
The second method for water flow into the olfactory
organ is active ventilation, depending on the work of
accessory sacs or ciliated nonsensory cells of the olfac-
tory epithelium. The fish in which the ventilation of the
olfactory sac occurs primarily because of the work of
accessory sacs are called cyclosmats (Døving et al.,
1977). The volume of these sacs, which are located
between the bones of the skull, changes in correlation
with the rhythmical breathing movements of the fish.
Due to characteristics of hydrodynamics of the olfac-
tory organ and sometimes also because of special
valves, the intake of a new portion of water after the
reduction of pressure in accessory sacs occurs via the
olfactory opening, and the release, via the posterior
opening. Cyclosmats may regulate the ventilation of
the olfactory cavity by increasing the gill and jaw mot-
orics. Cyclosmats include numerous representatives of
flatfishes, wrasses Labridae, some cyprinids, and such
species as Sebastodes melanops, Lampris quattus,
Bedotia geayi and other (Johnson and Brown, 1962;
Kapoor and Ojha, 1972; Oelschlãger, 1976; Døving et
al., 1977; Melinkat and Zeiske, 1979). For example, a
sharp increase of the volume of the accessory sacs (and
therefore rapid inhalation of large water volume into
the nasal cavity) in flatfishes and perhaps in other fish
occurs during “coughing,” characteristic sharp move-
ments of the gular membrane, opercula and jaws (Nev-
itt, 1992). These movements become more intense
when the fish is presented with an olfactory stimulus
(food extract), which allows us to consider coughing as
an analogue of smelling, characteristic of terrestrial
mammals.
The fish which do not have accessory sacs, in which
active ventilation is performed by ciliated nonsensory
cells, are called isosmats (Døving et al., 1977). Such
fish include the European wels Silurus glanis, Lepisos-
teus osseus, Clarias batrachus, Labeo rohita, Rhino-
muraena ambonensis, Polypterus ornatipinnis, and
many other species having no accessory nasal sacs
(Pfeiffer, 1968; Holl et al., 1970; Schulte and Holl,
1971; Ojha and Kapoor, 1972; Bashor et al., 1974;
Jakubowski and Kunysz, 1979; Goel, 1980). The ante-
rior nostrils in isosmat fish often have the form of pipes,
which helps to suck the water layers at a distance from
the fish body (Døving, 1986).
The speed of water intake into the olfactory organ
due to ciliated nonsensory cells is, while insignificant,
still obvious in experiments. When a suspension of
Indian Ink is introduced in proximity of the olfactory
organ of fish immobilized by myorelaxant, it is possible
to determine the time needed for particles to pass via
the olfactory organ, which usually ranges from 6–7 to
9–10 s (Parker, 1910; Teichmann, 1959). In Lepisosteus
(a)
(b)
(c)
Fig. 18. Scheme of pressure ventilation of the olfactory
organ in ditremous and monotremous fishes, minnow Phox-
inus phoxinus (a) and fifteen-spined stickleback Pungitius
pungitius (b, c). Arrow points to the basic directions of
water flow (from Pashchenko, 1986 and Theisen, 1982,
respectively).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S193
osseus, water is pumped with a rate of 2.2 mm/s. At this
speed 400–600 ms is enough for water to pass over the
lateral surface of the lamella (Bashor et al., 1974). The
maximum speed of the current in water inhaled into the
olfactory organ in small cyprinodontid fish is 12–
17 mm/s, at the release, 30–35 mm/s (Kux et al., 1988).
On the whole, water exchange in the olfactory tract in
cyclosmats occurs faster than in isosmats (Døving,
1986). During pressure ventilation, the speed of water
exchange does not increase significantly and even in
such fast swimming fish as tuna and scombers, the
olfactory organ which has accessory sacs and a poste-
rior nostril has a valve. However, constant movement of
the fish provides for a constant intake of water into the
olfactory organ (Døving, 1986).
Ventilation of the olfactory organ in most fish
involves both methods: pressure ventilation and the
work of accessory olfactory sacs allows rapid water
exchange in the olfactory organ, and ciliated nonsen-
sory cells allow the transfer of water with smells into
the narrow spaces between the olfactory lamellae
where receptor cells are located. In some fish (wels Sil-
urus glanis), the distance between the lamellae is only
30–50 µm (Devitsyna and Attar, 1988).
Functional Characteristics
of the Olfactory System
Olfactory spectra. The olfactory stimuli for fish
include numerous chemical substances (amino acids,
bile acids, their salts and their derivatives, prostogland-
ines, steroids, nucleotides, amines and polyamines,
etc.), different extracts, wash-offs, exometabolites of
plants and animals, etc. (Fig. 19) (Kleerkoper, 1969;
Malyukina et al., 1969; Døving, 1986). The fish char-
acterized by high sensitivity to a wide spectrum of
olfactory stimuli (mostly natural in origin, and evoked
different behavioral patterns) belong to the group of
macrosmatics. Microsmatics respond to only a limited
range of odors. However, the sensitivity to these smells
may be as high as in macrosmatic fish. An intermediate
group of mediosmatics is also sometimes distin-
guished. The concepts of micro and macrosmatics are
used not only with fish but also many other animals
(Frisch, 1941b).
Olfactory spectra are characterized by specificity
when species different in systematic position and life
mode are compared. For example, the amino acids cys-
teine, leucine, glutamine and aspartic acids are
extremely efficient olfactory stimuli for one–year old
juveniles of the Atlantic salmon Salmo salar and juve-
niles of some other salmonid fish but have no effect on
sturgeons, Fundulus heteroclitus, Menidia menidia
(Sutterlin, 1975; Mearns, 1985, 1986). However, sig-
nificant similarity of olfactory spectra is observed in
closely related species. For example, the olfactorily
efficient amino acids almost completely coincide in
four representatives of the order Siluriformes, in the
channel catfishes Ictalurus punctatus, I. catus, I. serra-
canthus from the family Ictaluridae and the catfish
Arius felis of the family Ariidae (electrophysiological
recording) (Caprio, 1980, 1982). The same results were
obtained by behavioral tests conducted on five species
of sturgeon: from twenty free amino acids only glycine
(in some species, also L-alanine) released the food
search response in the Russian sturgeon Acipenser gul-
denstaedtii, Siberian sturgeon A. baerii, stellate stur-
geon A. stellatus, green sturgeon A. medirostris, and in
the beluga Huso huso (Table 4) (Kasumyan and Taufik,
1994; Kasumyan, 1994, 1999). Based on phylogenetic
relationships of sturgeon revealed using karyological
analysis, one can conclude that the olfactory spectra of
fish are extremely stable and are retained without
changes for millions of years. (Fig. 20). In channel cat-
fish, this number is 75–80 million years (Caprio, 1980).
Olfactory sensitivity. Fish have extremely high
sensitivity to olfactory stimuli and do not differ from
terrestrial animals, such as mammals, in this respect.
The threshold or the threshold concentration of solution
causing significant electrophysiological responses in
the olfactory system may reach very low values, up to
10–9–10–13 M (Table 5) (Hara, 1992b). Different ways
to record electric activity have been used in experi-
ments: from the whole olfactory epithelium, from sep-
arate receptor cells, or from the olfactory nerve, olfac-
tory tract or individual nervous fibers (Fig. 21). Calcu-
lations showed that when the threshold concentration is
equal to 10–11 M, electric responses in the olfactory sys-
tem of salmonid fish will be recorded even when the
stimulus is the water from a basin 25 × 10 × 2 m
(500000 l in volume) in which only one drop (50 µl) of
gall is dissolved (Døving, 1989). Also, calculations of
the active olfactory space (in which the concentration
of natural olfactory substances exceeds the threshold
level) created by the tilapia Oreochromis mossambicus
are interesting (Frade et al., 2002) (Table 6).
Threshold concentrations of chemical substances
which could evoke behavioral responses in fish are
Clean Stimulus
Recorder
Output
Electrodes
Olfactory
water solution
device
rosette
Fig. 19. Setting for the recording of electroolfactogram in
fish (scheme).
S194
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
Table 4. The intensity of food search response (points) of sturgeon to amino acids solutions (10–4 M) (from Kasumyan and
Taufik, 1994)
Amino acid Russian sturgeon
Acipenser gueldenstaedtii Siberian sturgeon
Acipenser baerii Stellate sturgeon
Acipenser stellatus Green sturgeon
Acipenser medirostris Beluga
Huso huso
Alanine 0.8* 0.5* 1.7 0
Arginine 0 0 0 – –
Asparagine 0 0 0 – –
Aspartic acid 0 0 0
Valine 0 0 0 – –
Histidine 0 0 0 – –
Glycine 2.5* 2.6 1.2 1.4 0.5
Glutamine 0 0 0 – –
Glutamic acid 0 0 0
Isoleucine 0 0 0 – –
Leucine 0 0 0 – –
Lysine 0 0 0 – –
Methionine 0 0 0 – –
Norvaline 0 0 0 – –
Proline 0 0 0 – –
Serine 0 0 0 – –
Tyrosine 0 0 0 – –
Threonine 0 0 0 – –
Phenylalanine 0 0 0
Cysteine 0 0 0 – –
* Responses of fish to solutions with the concentration 10–5 M.
Table 5. Olfactory sensitivity of cyclostomata and fish to some types of chemical substances (based on electrophysiological
data) (from Hara, 1992b)
Species L-amino acids, M:
L-amino acids steroids F-prostaglandins
Myxine glutinosa 10–5–10–6 10–5–10–6
Negaprion brevirostris 10–7–10–8
Salmo salar 10–5–10–9
Salvelinus fontinalis 10–7–10–8
Salvelinus namaycush 10–7–10–9 10–8–10–9
Salvelinus alpinus 10–7–10–9 10–8–10–9 10–9–10–11
Oncorhynchus mykiss 10–7–10–8 10–9–10–10
Oncorhynchus nerka 10–6–10–7
Oncorhynchus kisutch 10–6–10–7
Ictalurus catus 10–7–10–9
Clarias gariepinus 10–10–10–11
Cyrpinys carpio 10–7–10–9
Carassius auratus 10–8–10–9 10–12–10–13 10–12
Misgurnus anguillicaudatus 10–10–10–13
Mugil cephalus 10–7
Seriola quinqueradiata 10–7
Pagrus major 10–6–10–7
Conger myriaster 10–8–10–9
Anguilla rostrata 10–7–10–9
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S195
higher than in electrophysiological experiments and
usually range from 10–6 to 10–9 M (Table 7). It is possi-
ble to record behavioral responses to lower concentra-
tions than in electrophysiological experiments in only
very rare cases (Murphy et al., 2001). Methods for the
recording of behavioral responses and the types of
aquariums used are also very diverse (Kasumyan,
1990a, 1990b; Marusov, 1990). The most important
requirement is adequacy, i.e., the degree they reflect
biological and behavioral characteristics of the species,
the fish age, motivational state, and the behavioral
response recorded. Many of the devices represent
mazes with different numbers of compartments, one of
which contains the smell stimulus (Fig. 22). When the
researcher uses the methods based on prior training and
conditioned reflex to the substance, the threshold con-
centrations may approach or even exceed those
recorded electrophysiologically. For example, using
this methodology, a record fish sensitivity to smells,
(a)
(b)
(c)
(d)
Fig. 20. Aquariums for the study of behavioral responses of fish to olfactory stimuli: (a) aquarium with two opposite flows; (b) Y-
like maze with compartment traps; (c) aquarium with four parallel flows; (d) flow-through olfactometer aquarium.
S196
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
0.5 × 10–18 mg/l of β-phenylethyl alcohol was found in
the eel Anguilla anguilla (Teichmann, 1959).
It should be noted that the above threshold concen-
trations represent fish sensitivity to artificial chemical
substances. It was hypothesized (Malyukina et al.,
1974; Devitsyna and Malyukina, 1977) that the thresh-
olds for natural substances may be even lower. How-
ever, there is no experimental support for this hypothe-
sis because the exact nature of natural chemical stimuli
for fish is still unknown.
There exist no patterns or relationships between the
structure of the stimulus molecule and its olfactory effi-
ciency. It is only known that aliphatic α amino acids
with short carbon chains are more efficient among free
amino acids. Also, L-stereoisomeres of amino acids are
always more efficient than their D-enantiomers (Suzuki
and Tucker, 1971; Hara, 1982, 1994; Caprio, 1982,
1984). Minor structural modifications of the stimulus
molecule, such as alteration of the position of some
functional group (e.g., the transition of the amino-
group from the α to β-position in the case of amino
acids), inclusion of an additional functional group or
removal of an existing one may cause significant alter-
ations of olfactory efficiency of the substance (Hara,
1977; Caprio, 1978, 1984). The simplest dipeptides
formed of two highly active amino acids may be ineffi-
cient (Hara, 1977, 1982; Caprio, 1978). It was also
found that mixtures of substances may have a very dif-
ferent efficiency than their separate components. Such
interactions may include effects of synergism, additism
or antagonism (Kang and Caprio, 1991; Hara, 1994).
CH3
CH3
HO
CH3
CONH(CH2)2SO3H
HO
HOH
OH
HOCH
CH3
CH3
CH3
O
OH
HO
COOH
OH
HOOC SH
NH2
Amino acid
Bile salt
Prostoglandin
Steroid hormone
Fig. 21. Types of chemical substances representing efficient olfactory stimuli for fish (from Hara, 1992b).
Table 6. Threshold concentrations of urine, bile, and feces
for the olfactory system of the Mossambique tilapia Oreo-
chromis mossambicus, the rate of extraction of these excre-
tions and creation of the active olfactory space
Source
of olfactory
stimuli
Threshold
solution
Rate of excre-
tion, ml/(h kg)
(body weight)
Olfactory
space, l/(h kg)
(body weight)
Urine 1 : 10–6.9 4.26 616
Bile 1 : 10–5.2 0.0008 2.2
Feces 1 : 10–4.9 0.5 18
Table 7. Olfactory sensitivity of fish to L-isomers of free
amino acids (based on behavioral experiments)
Species Threshold con-
centration, M Source
Anguilla anguilla 10–8–10–9 Sola et al., 1993
Salmo salar 5 × 10–9 Mearns, 1989
Salmo trutta 5 × 10–6
Salvelinus alpinus 4.6 × 10–7 Olsén, 1986
Oncorhynchus kisutch 10–7 Rehnberg,
Schreck, 1986
Ictalurus punctatus 10–7–10–9*Holland, Tee-
ter, 1981
Cyprinus carpio 2 × 10–7 Kruzhalov,
1986
Acipenser
gueldenstaedtii 10–6 Kasumyan,
Taufik, 1994
Acipenser baerii 10–6
Coregonus clupeaformis 10–8–10–10 Jones, Hara,
1985
Thunnus albacares 10–11 Atema et al.,
1980
Carassius carassius 10–7 Kruzhalov,
1990
* Data obtained with the conditioned reflex methodology.
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S197
Adaptation. The olfactory system is characterized
by slow adaptation to the acting olfactory stimulus (Ott-
son, 1956; Døving, 1966; Hara and Gorbman, 1967;
Malyukina et al., 1969, 1974; Devitsyna and Maly-
ukina, 1977). Because of this property, olfactory stim-
uli retain their signal value for a prolonged time, which
is important for the orientation of fish to the source of
the smell or directed movement along the smell corri-
dor as often occurs during fish migrations. Prior adap-
tation of the olfactory system to a smell may differently
affect the sensitivity to other substances (cross-adapta-
tion) (Caprio and Bird, 1984; Hara, 1994).
The pattern of smell distribution in the water media
enables retention of the signal properties of a substance
and the possibility for the fish to find its source. The
olfactory trace, even in a relatively laminar current, has
a very complex internal structure because of chaotic
alteration of areas with significantly different substance
concentrations (Atema, 1988). Therefore, when the fish
moves along the olfactory corridor, the intensity of the
stimulus affecting the olfactory receptors may vary in a
wide range. Special studies showed that adaptation of
the receptors to the substance in such conditions is
much slower.
Distance and mechanisms of olfactory orienta-
tion. The ability to perceive and respond to trace con-
centrations of olfactory substances and slow adaptation
of the olfactory system allows fish to obtain informa-
tion with a very remote source. According to some
assessments, olfaction is characterized by the highest
distance in comparison with other sensory systems
(Table 8) (Pavlov and Kasumyan, 1990; Wilson and
Smith, 1984). Longevity of the olfactory trace depends
on such characteristics of the water media as stable cur-
rent direction and pronounced vertical stratification.
Observations indicated that fish could determine the
presence and even identify and find the location of an
olfactory lure at a distance of dozens or even hundreds
of meters. The lemon shark Negaprion brevirostrus in
an environment similar to natural (in a large lagoon iso-
lated from the ocean) followed along an olfactory trace
of curvilinear trajectory if the lure was at a distance
exceeding 17 m (Gilbert, 1961; Hobson, 1963). Pacific
salmon (Oncorhynchus) identify spawning rivers by the
smell of the spawning grounds they were born in, even
when these are many dozens or hundreds of kilometers
upstream (Hasler and Scholz, 1983; Stabell, 1992;
Døving and Stabell, 2003).
The ability of fish to rapidly find the source of a
smell depends not only on the level of olfactory sensi-
tivity and the duration of adaptation, but also on the
hydrodynamic characteristics of the media. In a cur-
rent, fish, in most cases, can find the source of the
attracting stimulus, such as food, more easily. In the
absence of current this may require much more time
(Kleerekoper et al., 1975). Observations of the behav-
ior of deep-sea fish in natural environments indicated
that the higher the current speed, the more rapidly and
more precisely fish could locate an olfactory lure (Wil-
son and Smith, 1984). However, for fish inhabiting
slow-moving waters, searching for an olfactory lure in
a current is more difficult (Sherman and Moore, 2001).
Another characteristic of many natural water bodies
which enables the olfactory orientation of fish is verti-
cal stratification, that is, the presence of stable horizon-
tal layers of water with different temperature, salinity,
or current direction or speed (Fig. 23) (Døving et al.,
1994; Døving and Stabell, 2003).
It is accepted that the mechanism of fish orientation
to the source of smell in currents is based on klinotaxis,
when the fish moves against the current towards the
source of the stimulus, due to comparisons of the stim-
ulus concentration over a certain time interval. This
mechanism accounts for the directed movements of fish
Table 8. The possibility of finding remote food objects with
different sense systems (from Pavlov and Kasumyan, 1990)
Distance, m Sensory systems
More than 100 m olfaction
100–25 olfaction, audition
25–5 olfaction, vision, audition
5–1 vision, olfaction, audition
1–0.25 vision, audition, lateral line, olfaction
Less than 0.25 vision, lateral line, electroreception,
external gustatory sensitivity, general
chemical sense, tactile reception
0external and intraoral gustatory
sensitivity, tactile reception
Note: The sensory systems are listed in an order which corresponds
to their significance during the search for remote food.
10 million
> 80 million
H.huso
2n = 118 ± 2 A. stellatus
2n = 118 ± 2 A. baerii
2n = 240 – 260
A. gueldenstaedtii
years ago
years ago
Fig. 22. Scheme of phylogenetic relatioships of sturgeon
fishes (from Vasilyev, 1985).
S198
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
along a smell gradient, from areas with low concentra-
tions to those with high concentrations. Orientation in
stalled water occurs due to tropotaxis, i.e., due to com-
parison of concentrations of the attracting stimulus in
the left and right olfactory organs. The fish can proba-
bly actively use both mechanisms in natural environ-
ments. This is supported by experiments in bonnethead
sharks Sphyrna tiburo conducted in a round basin. A
special device (2.4 × 0.8 × 1.0 cm) was attached to the
head of the fish, allowing delivery of the appropriate
quantity of Callinectes sapidus crab extract separately
to the left and right nostrils via thin tubes (Fig. 24). The
Current speed
Depth
Fig. 23. Scheme of possible horizontal currents in a water body, their direction and relative speed.
(a)
(b) (c) (d) (e)
Fig. 24. The behavior of the bonnethead sharks Sphyrna tiburo due to different methods of stimulation with food smells, extract of
the blue crab Callinectes sapidus. (a) Device for delivery of smell to the olfactory organ; (b) and (c) and (d) and (f) fish movements,
respectively, in stalled water and in the presence of circular current; (b) and (d) after delivery of the stimulus solution into the tank;
(c) and (e) after delivery of the stimulus solution into the olfactory organ (uni- and bilateral stimulation). Arrows point to the fish
position at the moment of olfactory stimulus delivery (from Johnsen and Teeter, 1985).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S199
fish constantly moved around the circle along the walls
both in the presence of a circular current and in its
absence. When 10 ml of the extract was released into
the basin in the absence of a current, the fish lagged in
the olfactory zone and begun short searching move-
ments. The same response was observed when 0.5 ml of
the stimulus was released to both nostrils simulta-
neously. In the presence of a circular current (15 cm/s),
the response to 10 ml of the extract was more compli-
cated. When the fish found itself in the stimulus zone
(which itself moved in the current as a continuously
eroded spot), it continued to follow forward and there-
fore rapidly left the scent. Turning back, the shark
began to swim along the current until it again reached
the smell spot. Moving more rapidly than the current,
the subject left the scent zone, turned back and begun
swimming against the current until it again entered the
smell spot. Such behavior repeated until the concentra-
tion of the stimulus reduced to below the threshold
value. The responses of the sharks to the administration
of 0.5 ml of the extract to both nostrils simultaneously
was very similar. When the stimulus was released to
only one nostril, the fish always tended to turn toward
the side in which the stimulus was released rather than
toward clear sea water, or to the side to which the stim-
ulus was administered in a higher concentration (the
differences in concentrations was equal to 50%)
(Johnsen and Teeter, 1985).
Fish can determine the correct direction toward the
source of a smell during prolonged migrations along
the olfactory corridor, based on certain characteristics
of their behavior. Detailed observations of salmon
migrating to their spawning sites indicated that when
approaching the river mouth, the fish keep at a certain
water layer, periodically moving outside it (Fig. 25)
(Døving et al., 1985; Tanaka et al., 2001). The same
behavior is characteristic of adult American eels
Anguilla rostrata entering the estuary during catadro-
mous migration (Barbin, 1998). It was hypothesized
that the fish, in this way, could control their spatial posi-
tion as to not lose the zone with the maximum smell
concentration. One possible orientation mechanism in
such cases is tropotaxis. However, unilateral olfactory
deprivation does not cause any disturbances in the
behavior and orientation of migrating salmon. Only
complete dysfunction of the olfactory system can cause
a total loss of orientation (Døving et al., 1985). Move-
ments along curved or zigzag trajectories circular, or
S-like, or -like are characteristic of many fish search-
ing for the source of the smell (Parker, 1912; Bardach
and Case, 1965; Johnsen and Teeter, 1985; Kasumyan,
1999, 2002).
Horizontal zigzag movements of migrating salmon
are often observed in rivers, especially in places where
large tributaries inflow. Such swimming allows the
migrating salmon to remain in those areas of the current
which bear the smell of the home spawning site.
Chemical Signals
Classification of chemical signals. Chemical sig-
nals of fish are very diverse. According to the classifi-
cation used for all animals, chemical signals regulating
interactions between conspecific individuals are named
pheromones. Interactions between representatives of
different species are regulated by kairomones and
allomones. Kairomones are such interspecific odors
which bear useful information for the recipient of the
stimulus. When the chemical signal is useful for the
producing individual rather than the recipient, this sig-
nal is called an allomone (Fig. 26) (Karlson and Lüsher,
1959). Chemical signals within these groups are sepa-
rated into sex pheromones, alarm pheromones and
Fig. 25. Vertical movements of the Atlantic salmon Salmo
salar in nearshore zone after approaching the mouth of the
home river. The stripped lines show isotherms with the step
0.2° between 13 and 14°C (from Døving et al., 1985).
Fig. 26. Types of chemical signals regulating intra- and
interspecific relationships in animals. A and B species
linked by chemocommunication; the darkened circle
depicts the participant of the communication pair, for which
this link is favorable.
Time, min
Depth, m
14°C
13°C
AA
A A
A B
BA
Pheromones
Kairomones
Allomones
Producer Recipient
of the chemical
stimulus of the chemical
stimulus
S200
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
kairomones, stress (fear) pheromones, territory mark-
ing pheromones, etc.
Releasing signals and priming signals are distin-
guished among chemical signals as well as among sig-
nals of other modalities (Wilson and Bossert, 1963).
The releaser signals after reception cause rapid behav-
ioral response with relatively short latency. Priming
signals first release complex endocrine processes, caus-
ing the secretion of physiologically active substances,
in turn initiating the appropriate alterations of meta-
bolic and regulatory processes. This causes shifts in
metabolism, respiration intensity, alterations of body
pigmentation, stress development, etc.
Sources of smell substances. In terrestrial animals,
smells are usually secreted by special glands. In most
fishes, the glands responsible for the secretion of chem-
ical stimuli are absent. There exist only a small number
of exceptions from this rule among male fish. For
example, males of Claris gariepinus secrete sex phero-
mones into the seminiferous vesicles, special structures
of the urogenital system of these fish, and the phero-
mones are then released simultaneously with the secre-
tions of the seminiferous vesicles (Resnik et al., 1987,
1989). The secretion of sex pheromones in males of the
blenny Blennius pavo occurs in preanal blind processes
(anal appendices) (Laumen et al., 1974), in the goby
Gobius jozo, in the mesorchial gland (Colombo et al.,
1980).
Most smell substances represent so-called exome-
tabolites or external metabolites. They are released into
the environment as a result of the living activities of the
fish with urine or feces, or are excreted via the skin, first
being excreted to the skin mucus and then to the envi-
ronment (Stabell et al., 1982; Malyukina et al., 1983;
Hara et al., 1984; Canario, Scott, 1989, 1994; Sorensen,
1992; Moore et al., 1994; Moore et al., 1994; Courte-
nay et al., 1997; Vermeirssen et al., 1997; Yambe et al.,
1999). It is proposed that the excretion of some smell
substances may occur via the gill epithelium (Sorensen
et al., 2000; Li et al., 2002). Exometabolites include
signal substances revealing the species, population,
sex, and individual identity of the individual, the stage
of gonad development, possible stress, and social sta-
tus. Exometabolites may be used for chemical marking
of the territory. Sex pheromones, stress and alarm pher-
omones also belong to exometabolites.
The source of some sex pheromones are gonadal liq-
uids, such as the ovarial and spermatic liquids released
in water during spawning, when the gonadal products
are released (Van Den Hurk and Lambert, 1983; Van
Den Hurk et al., 1987; Sorensen, 1992; Van Den Hurk
and Resnik, 1992).
There exist smells (alarm substance) which are kept
in special epithelial cells. These substances can be
released in the water only when the skin is damaged by
the predator (Frisch, 1941a; Schutz, 1956; Pfeiffer,
1963). In dead fish, these substances are released in
water spontaneously, most intensely during the first
hours after the death (Kasumyan, 1989; Kasumyan and
Tuvikene, 2004).
Representatives of the genus Pardachirus (fam.
Soleidae) have special glands located at the base of the
dorsal and anal fins with external opening by numerous
pores. The secretion developed by these glands has
repellent properties for piscivorous fish feeding on
small nearshore fish (Clark, 1981, 1983). On the whole,
the presence of special glands secreting smell sub-
stances is rather unusual for fish.
The nature of chemical signals. Whereas there
exists a limitation on the size of a distant smell sub-
stance in terrestrial animals required by the volatility
(the molecular weight of volatile substances must not
exceed 300–400 Da), to fish and other aquatic animals
smell signals have no such limitation. Therefore, chem-
ical signals for fish may include substances with small
molecular size and large high-molecular substances
with a molecular weight up to 10000 Da or higher
(Kasumyan and Ponomarev, 1987, 1991; Carr et al.,
1996). The only requirement for such substances is sol-
ubility in water, which does not need to be high,
because the fish can perceive very low trace concentra-
tions of signal substances.
Very different chemical substances satisfy these
requirements. This makes the identification of the
chemical structure of natural chemical signals of fish
and other aquatic organisms very difficult, much more
difficult than in terrestrial animals. For example, the
chemical structure of several hundred pheromones has
been identified in insects. Many of them have been syn-
thesized artificially and are applied by humans in indus-
try. A. Butenandt was awarded the Nobel Prize in 1959
for studies associated with the identification of the first
insect pheromone, the sex pheromone of the silkworm
Bombyx mori, which turned out to be the multi-atom
alcohol bombicol.
Natural smell stimuli in fish, as in higher animals,
represent multicomponent mixtures of substances,
making the problem of identification even more com-
plicated. The chemical nature of no single natural sig-
nal of fish has been identified up to the present time.
Attempts to determine the chemical nature of the alarm
pheromone of cyprinid fish have been made over the
past 60 years by specialists from Germany, Russia,
Switzerland, Canada, Norway, and the USA (Døving
et al., 2004). The most successful studies were of the
sex pheromones of fish. The basic components of sex
pheromones in teleosts and possibly in cyclostomata
are different steroids and their derivatives as well as
prostaglandins (Adams et al., 1987; Sorensen and Sta-
cey, 1990; Sorensen, 1992).
The Role of Olfaction in Fish Behavior
Olfaction is involved in regulation of almost all
behavioral forms (Kleerekoper, 1969; Malyukina et al.,
1969, 1980; Døving, 1986). This is mainly associated
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S201
with the fact that they live in the aquatic environment,
in which the perception of other stimulus modalities,
such as visual, is often difficult. Diffusion of chemical
substances in water is slow, which makes olfactory
traces relatively long-lasting. With currents, the turbu-
lence of which is not as pronounced as in air, smells
may spread to large distances (Atema, 1980). Another
advantage of olfactory communication is the possibility
to transmit an almost unlimited volume of information
with chemical stimuli, from general to highly specific.
This can be reached by the ease with which endogenous
chemical substances may form different combinations.
Smells regulate such forms of behavior as feeding,
spawning, parental, defensive, migration, social, and
territorial. In many cases, olfaction is the leading sense,
determining the possibility of a particular behavioral
pattern or repertoire of the species. There are many
experimental studies showing that olfactory deprivation
(blocking the olfactory sensitivity by closing the water
intake into the olfactory organ, cauterizing the olfactory
rosette, treatment of the olfactory epithelium by deter-
gents or salts of heavy metals, cutting or removal of the
olfactory nerve section, extirpation of the olfactory
bulb, etc.) makes spawning very difficult or impossible,
as well as locating food, homing, maintenance of the
group hierarchical structure, etc. (Parker, 1910; Göz,
1941; Frisch, 1941a; Todd et al., 1967; Zeiske, 1968;
Kleerekoper, 1969; Pfeiffer, 1969; Pavlov et al., 1970;
Pollack et al., 1978; Atema, 1980; Honda, 1980a,
1980b; Kasumyan and Pashchenko, 1982; Stacey and
Kyle, 1983; Hasler and Scholz, 1983; Pashchenko and
Kasumyan, 1984; Døving et al., 1984, 1985;
Kasumyan and Pashchenko, 1985; Stabell and Refstie,
1985; Hellstrøm and Døving, 1986; Olsén et al., 1986,
1998; Sorensen, 1992; Liley et al., 1993; Kasumyan
and Kazhlayev, 1993; Zippel, 1993; Zippel et al.,
1993a; Kasumyan and Devitsyna, 1997; Kasumyan,
1999, 2002; Hamdani et al., 2001b; Kasumyan and
Marusov, 2002, 2003). The most diverse is the partici-
pation of the olfactory reception in intraspecific rela-
tionships in fish: intraspecific communications. The
role of olfaction in fish life is supported by the ability,
found in many species, to memorize olfactory stimuli
and retain the conditioned associations for a very long
time (Vinogradova and Manteifel, 1984; Zippel et al.,
1993a, 1993b; Rekowski et al., 1994).
The smell of the species, population, and individ-
ual. Even early studies revealed that fish extract sub-
stances create a so-called smell of the species and allow
them to distinguish conspecific individuals from other
species, even systematically related ones (Wrede,
1932; Göz, 1941; McLennan and Ryan, 1997, 1999).
The smells produced by the fish are characterized not
only by species but also population and individual spec-
ificity, allowing the fish to distinguish representatives
of their own population or group from other conspecif-
ics (Døving et al., 1974; Selset and Døving, 1980; De
Fraipont and Thinés, 1986; Groot et al., 1986; Olsén,
1986; Stabell, 1987, 1992; Berti et al., 1989; Courtenay
et al., 1997, 2001; Hiscock and Brown, 2000). Fish
could also discriminate conspecific individuals (Wrede,
1932; Todd et al., 1967; Carr and Carr, 1985). It has
been shown in territorial fish and fish with well devel-
oped hierarchical relationships that individual chemical
signals, the source of which is mucus and urine, could
provide information about the hierarchical status of the
partner, its physiological state (stress), sex, and ripe-
ness of gonads, age and body size. Olfactory depriva-
tion makes the fish unable to obtain such information
(Todd et al., 1967; Bardach and Todd, 1970; Todd,
1971). Species smells have an attractive effect for many
fish species, especially schooling (Hemmings, 1966;
Pavlov and Kasumyan, 2000). But such an effect is not
always observed and in some cases, as has been found
in the cod Gadus morhua, the response may range from
attractive to repellent, depending on the concentration
of the smell or the life mode of the fish during that par-
ticular stage of the life cycle (Malyukina et al., 1980,
1983). Certain cave fish exhibit an avoidance response
to the smell of conspecific individuals and other species
(Berti and Zorn, 2001).
Reproductive behavior. The involvement of chem-
ical signalization in fish reproductive behavior has been
found in all fish studied. Fish use olfaction to find a
partner ready for spawning and determine the readiness
of their gonads. Smell also stimulates the ripening of
gonads and promotes synchronization of spawning.
Sexual smells are mostly used by male fish; involve-
ment of chemical signalization in the behavior of
females is less pronounced (Colombo et al., 1980;
Liley and Stacey, 1983; Adams et al., 1987; Sorensen,
1992; Olsén and Liley, 1993; McLennan and Ryan,
1997).
The sensitivity of fish to sex pheromones has been
found in species with different systematics and life
modes. Sex pheromones are very efficient olfactory
stimuli even for microsmatic fish, which are character-
ized by a narrow olfactory spectrum and do not respond
to many chemical stimuli, such as food odor (pike,
sticklebacks) (Devitsina and Malyukina, 1977; Gol-
ubev and Marusov, 1979, 1984).
Recent studies indicated that a whole complex of
pheromones with both priming and releaser effects is
involved in the chemical regulation of spawning and
sexual behavior. Each sexual pheromone is directed at
specific chains of a complex behavioral unit and physi-
ological alterations associated with the final stages of
ripening and spawning in fish (Sorensen, 1992;
Sorensen and Stacey, 1999).
The goldfish Carassius auratus is the most studied
object in this respect. This species is a very convenient
laboratory model for the study of fish pheromones
(Dulka et al., 1987; Sorensen, 1992; Sorensen and Sta-
cey, 1999; Polling et al., 2001). Most sex pheromones
regulating the reproductive behavior in the goldfish
enter the water with urine, portions of which are regu-
larly excreted by females (Fig. 27). Before ovulation,
S202
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
females of the goldfish secrete 17α, 20β-dihydro-
progesterone. When excreted in water, it works as a
sexual preovulatory pheromone with a priming effect.
17α, 20β-dihydroprogesterone increases the level of
gonadotropin in male blood and, as a consequence,
increases sperm production (Fig. 28).
Another sexual pheromone is excreted by females of
the goldfish later, during ovulation, and brings about
prespawning behavior of males. This pheromone is a
chemical signal with a releaser effect. It is hypothesized
that a postovulatory releasing pheromone exists, which
appears in the water at the moment of egg release by the
female or immediately before it, and causes sperm
release in males. The function of this pheromone is to
synchronize the spawning of both partners. Fish sex
pheromones represent complex mixtures, the most
active components of which include steroid hormones
or their derivatives and prostaglandins (Fig. 21)
(Sorensen, 1992).
Sex pheromones are characterized by species speci-
ficity and probably participate in the reproductive isola-
tion of related species or even populations of the same
species (McLennan and Ryan, 1997). For example, it
was found in the anadromous and resident forms of the
masu salmon Oncorhynchus masou with similar repro-
ductive ecology that only the smell stimuli produced by
ripe females of the same form are highly efficient for
males (Bloom and Perlmutter, 1978; Honda, 1982;
Liley, 1982).
However, sex pheromones may have significant
interspecific efficiency, as has been found in sturgeon.
A solution of the ovarial liquid taken from female Rus-
sian sturgeon Acipenser gueldenstaedtii evoked repro-
ductive behavior not only in conspecific males but also
in males of the stellate sturgeon A. stellatus
(Kasumyan, 1993, 1999).
As noted above, pheromone regulation of sex
behavior has been found in all fish species studied. But
it is especially important for those species, in which the
function of non-olfactory systems, mostly vision, is
limited because of lifestyle or ecology. This was dem-
1
Consecutive 5-min intervals
Number of urine portions excreted
123456789101112
2
3
4
by six females for 5 min
Fig. 27. Time course of urine excretion by ripe females of
the goldfish Carassius auratus (from Appetl and Sorensen,
1999).
17α, 20βP
1200 2000 Sexual
Excretion in water
Spawning
Sexual
sperm
PGFs
ovulation
Maturation
GfH
GfH
17α, 20βP
Males
Females
1200
1200 2000 1200
Time of day, h
behavior
of females
synchronization
behavior
of male
Fig. 28. Participation of sex pheromones in regulation of reproductive behavior in the goldfish Carassius auratus (scheme) (from
Sorensen, 1992).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S203
onstrated in a small short-cycle mesopelagic fish Argy-
ropelecus hemigymnus (family Sternoptychidae, order
Stomiiformes) (Jumper and Baird, 1991). Ripe individ-
uals have a length of about 30 mm and inhabit oceanic
depths from 200 to 600 m. The olfactory organ in males
of these fish is morphologically developed significantly
more than in females. Complicated mathematical com-
putations determined the amount of time the male fish
would need to find a ripe female using vision and olfac-
tion. The computations were based on the behavior and
biology of the species and the data known for sex pher-
omones of fish. The following was accepted: (1) the
average population density of Argyropelecus hemigym-
nus is about 3 × 10–5 /m3; (2) the swimming speed of
males before and after receiving the signal of female
presence is, respectively, 1.5 and 3.0 cm/s; (3) the
female takes a stationary position during the emission
of the pheromone and produces about 2 ×
1014 molecules; (4) the sensitivity of males to the pher-
omone is 2 × 1010 /m3; (5) the smell spreads homoge-
neously within 10 cm of the horizontal water layer.
Computations indicated that with such parameters,
males, in 90% of the cases, could find the presence of
the female 1–2 h after the female begun emission of the
sexual pheromone, and they could find the female dur-
ing the subsequent 30 min, if the smell was found at the
distance 20 m from the female. If the male of Argyro-
pelecus hemigymnus relied only on vision (the visual
radius was taken equal to 1 m), it required eight days to
find the inactive female or 5–6 days of searching if the
female moved randomly (non-vectorized movements)
(Jumper and Baird, 1991).
Experimental data suggest that smells excreted in
water by ripe females of the mosquitofish Gambusia
affinis cause the retardation of growth of juveniles and
development of smaller ovaries in females (Luthesky
and Adkins, 2003).
Relationships between parents and offspring.
Chemical signalization is involved in the regulation of
another behavior, relationships between the parents and
the offspring. Many fish species exhibit different forms
of such relationships. In cichlids defending the brood,
the water from their own brood stimulated behavioral
units associated with parental care. The signal value of
the smell of juveniles is retained during the whole dura-
tion of parental care until the family group is broken.
Defensive behavior. The olfactory regulation of this
behavior involves several chemical signals of danger.
On of the most interested is the alarm pheromone,
which is contained in special epidermal cells of the skin
(club cells) and appears in water when the skin cover of
the fish is damaged by the predator. The behavioral
response to this signal of danger involves the fright
response and then quick avoidance and hiding. For a
long time, the fish avoid areas of the habitat where they
encounter this smell (Frisch, 1941a; Schutz, 1956; Pfe-
iffer, 1962, 1977; Smith, 1982, 1992; Wiesenden,
2003). In addition to the releaser effect, the alarm pher-
omone has a priming effect, causing stress responses
(Malyukina et al., 1982; Lebedeva and Golovkina,
1988) and alterations of the body proportions of the fish
(Stabell and Lwin, 1997).
A defense behavioral mechanism based on the
alarm substance is characteristic of most small freshwa-
ter fishes with different systematics (Pfeiffer, 1977;
Chivers and Smith, 1998). The response to the alarm
pheromone is innate, it does not require prior learning.
The effect of the alarm substance has a pronounced
interspecific and intra-specific efficiency, but the
strength of the behavioral response of the fish to the
skin extract of related species is always lower then to
skin extract of conspecifics (Schutz, 1956; Pashchenko
and Kasumyan, 1983, 1986; Kasumyan and Pono-
marev, 1986b). The adaptive function of the defense
behavior caused by the alarm pheromone is the increase
of the relative protection of prey fish after one member
of the group is dead.
Detection of predators and transition of fish to safe
areas is reached due to another signal of danger: the
alarm kairomone. The alarm kairomone is constantly
excreted in water by predatory fish. The basic source of
the substance is skin and skin mucus (Lebedeva and
Chernyakov, 1978). The behavioral response of non-
predatory fish to the predator smell resembles the
response to the alarm pheromone but differs by smaller
latency (Marusov, 1976; Malyukina et al., 1990). Per-
ceiving this smell, non-predatory fish obtain informa-
tion about the presence of the predator and avoid the
dangerous area of the water body. It was found that the
smell of ambush predators has a stronger repellent
effect than pursuit predators. In addition to the releaser
effect, the alarm kairomone also has a priming effect,
altering the pigmentation of the fish, as was found in the
minnow Phoxinus phoxinus presented with the smell of
the pike Esox luceus. A longitudinal black stripe
appears at the lateral side of the body of this fish in such
a case (Malyukina et al., 1974, 1990; Marusov, 1976;
Lebedeva and Chernyakov, 1978; Kasumyan and Pash-
chenko, 1985). Having perceived the smell of the pred-
ator, many fish reduce their swimming activity, termi-
nate feeding, and tend to hide in shelters, and subse-
quently avoid the place where the danger was detected
(Magurran, 1989; Keefe, 1993; Mathis and Smith,
1993; Peterson and Brönmark, 1993; Jacher, 1995;
Boyer et al., 2001; Utne and Bacchi, 1997). The joint
presentation of the alarm pheromone and the predator’s
smell increases the repellent effect (Boyer et al., 2001).
Along with the alarm pheromone, a prolonged effect of
the predator’s smell brings about an increase of the
body depth, which reduces the vulnerability of the prey
fish (Brönmark and Miner, 1992; Holopainen et al.,
1997; Petterson et al., 2000). Not all fish capable of
determining the presence of the predator by its smell
have the alarm pheromone in their skin (Utne and Bac-
chi, 1997).
S204
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
Recent studies revealed the repellent involvement of
another type of smell stimuli in the chemical signaliza-
tion of danger, the so-called stress or fear pheromone.
Fish excrete the stress pheromone in water after the fear
or fright associated with the appearance of a real pred-
ator or demonstration of its visual model (Malyukina
et al., 1983; Lebedeva et al., 1993; Wiesenden et al.,
1995; Jordão and Volpato, 2000; Bryer et al., 2001).
Different species of fugu (Takifugu pardalis, T. poecil-
onotus, T. vermicularis, T. niphobles) secret tetrodot-
oxin which may have a repellent effect on predators
(Kodama et al., 1985). The same role may be played by
certain substances contained in the skin mucus secreted
by Arius thalassinus (Al-Hassan et al., 1982).
Skin wash-out of mammals, especially bears, has a
strong repellent effect on migrating salmon. This chem-
ical factor received the name “the beast skin factor.”
According to the classification of chemical signals, it
belongs to the group of kairomones. The presence of this
smell in water may retard the spawning migration of
Pacific salmon for a prolonged time, and in some cases is
used for temporary retardation of fish migration. It is
hypothesized that the main active component of this
smell is a small peptide containing L-serine amino acid
(Brett and McKinnon, 1954; Idler et al., 1956).
Natural repellents belonging to the group of
allomones, i.e., interspecific signals favoring the pro-
ducer include the special secret of the Red Sea Moses
sole Pardachirus marmoratus. A white milky-colored
secretion in these small nearshore fish is secreted by the
gland located at the base of the dorsal and anal fins. The
total volume of secretions in large sole (20 cm and more
in length) can reach 2 ml. The pores via which they are
excreted to the environment are located in the rostral
parts of these fins. The total number of pores exceeds
200. The secretions appear in water after the fish is
frightened, and have a repellent effect on sharks catch-
ing prey. They never tried to catch live soles, whereas
other fish serving as lure on a throat line were swal-
lowed almost immediately. The sole could be caught
only several hours after death or if the mucus was
removed from the body. The repellent effect was pro-
nounced both in artificial conditions (basins) and in
nature. The repellent response was observed in several
species of shark and a few moray eels (Clark, 1981,
1983).
The active component of this secretion, which was
called pardaxine (from Pardachirus) represents a small
peptide composed of 162 amino acids with a molecular
weight of 17 kDa (Primor et al., 1978). Subsequently, it
was found in another sole species, Pardachirus pavoni-
nus, that the secretion may contain three different par-
daxines having the same number of amino acid resi-
dues. All have active surfactant properties and by their
pharmacological properties resemble melittine, the
basic component of bee poison. Several steroid-like
monoglucosides called pavonines were also found in
the secretion, which not only had repellent but also
toxic and hemolytic effects (Tachibana et al., 1984,
1985; Thompson et al., 1986).
The finding of a natural shark repellent attracted a
lot of attention because multiple attempts to create an
anti-shark repellent had not been successful, in spite of
great effort and means directed to this task (Baldridge,
1990). The need to create or find a shark repellent first
appeared in the US immediately after participation in
World War II. The US Government, the Navy and the
Air Force directed much attention to rapid solution of
this problem. The fear of being attacked by sharks dur-
ing accidents and military actions on the sea was very
significant among the personnel even though the real
threat was minimal. For example, the number of shark
attacks on humans in 2000, the record year for all
recorded history, was less than one hundred, and not all
these attacks resulted in human death.
The first experiments directed toward finding a
shark repellent were conducted in the Woods Hole
Institute of Oceanography, and then on a wider scale, in
Hawaii and the Bahamas in conditions closely resem-
bling natural. A variety of artificial substances were
tested, along with numerous natural extracts, wash-
outs, biologically active substances, poisons, and irri-
tants. But all these substances either had no effect or
were too inefficient. The strongest repellent effect was
obtained from an extract of rotten shark meat (after 4–
6 days at 20°C) (Gilbert and Gilbert, 1973). Chemical
analysis identified a large quantity of acetic acid and
copper ions. This became the base for the “Shark
Chaser” repellent, representing a mixture of copper
acetate (20%) and nigrosine stain (80%). Unfortu-
nately, the efficiency of this repellent was too low, and
in some cases, e.g., during food excitation of the shark,
it was completely absent. All attempts to create a more
reliable shark repellent were unsuccessful and these
projects were continuously closed. The history of these
studies is described in several publications (Gilbert and
Springer, 1963; Tester, 1963; Hodson and Mathewson,
1981; Sisneros and Nelson, 2001).
Feeding behavior. The long-distance nature of
olfactory reception makes this sensory system
extremely important in obtaining information about
food availability and presence and the selection of the
most appropriate direction to search in. The olfactory
system is the most important for both close and remote
food searching in many fish species, especially those
with poor vision (night, bottom, deep-sea and some
other fishes) (Groot, 1971).
The behavioral response of fish to food smells is
very diverse. In less active fish, which spend most of
their time in shelters waiting for prey, the tendency to
seek out food is weakly pronounced or absent. The per-
ception of the food smell by these fish is clear from the
alteration of the fish posture, fin movements, eye move-
ments, increased breath rate, jaw movements, cough,
etc. (Fig. 29) (Atema, 1980; Jones, 1992). In non-terri-
torial freely swimming fish, the food search is well pro-
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S205
nounced and differs in species with different life strat-
egy and food behavior pattern (Fig. 30) (Atema, 1980;
Kasumyan and Ponomarev, 1989; Pavlov and
Kasumyan, 1998). The sensitivity to food smells is
associated with the life mode of the fish: the lowest con-
centration threshold is characteristic of benthivorous
fish, and the highest threshold is characteristic of
pelagic planktivorous fish, the feeding behavior of
which is primarily based on visual reception (Table 9).
It was found, regarding food smells, that fish form
an olfactory image of the food objects, i.e., the food
smell provides information not on the presence of some
abstract food object but on a certain specific prey organ-
ism with specific behavioral and behavioral character-
istics (McBridge et al., 1962; Atema et al., 1980;
Kasumyan and Ponomarev, 1986a). The ability to form
the olfactory images optimizes the food search and
makes it more efficient, pointing to an important role of
olfaction in the regulation of feeding behavior in fish.
The experimental support for this conclusion includes
experiments on the study of feeding behavior of fish to
smells of known and unknown food objects and smells
of different food types (Fig. 31).
Migrations. Olfactory reception is an important
sensory component of the migratory behavior in many
fish species. Orientation using smell during prolonged
migrations is most pronounced in salmonids, adult indi-
viduals of which return from the areas of feeding in the
sea to their home rivers and their tributaries for spawn-
ing (Hasler, 1966; Stabell, 1992; Døving and Stabell,
2003). This phenomenon, returning to the home area or
the territory, is called homing, whereas erroneous
choice of the spawning place and wandering in differ-
ent locations is called straying.
It is determined that homing in salmonids is based
on olfactory imprinting, an important biological char-
acteristic of many animals representing a specific form
of learning (Wisby and Hasler, 1954; Hasler and
Scholz, 1983). Imprinting was discovered by renowned
Austrian zoologist, the 1973 Noble Prize winner Kon-
rad Lorenz. Imprinting occurs in animals during the so-
called sensitive period, a limited period of time when it
is possible.
Imprinting of the home smell in salmonid fish
occurs during the short period of time when juveniles
terminate their period of life in the fresh water and pre-
pare for downstream migration to the sea (Hasler and
Scholz, 1983). This period is called smoltification, and
is marked by a complex of physiological and biochem-
ical processes preparing the fish for marine life. Smolt-
ification is controlled by the hormonal system (thyroid
hormones) (Scholz et al., 1985; Yamauchi et al., 1985).
Technically very complex, prolonged studies (includ-
ing observation of homing and straying in anosmic and
intact salmons, imprinting of juveniles to an artificial
substance morpholone, transition of the migrating juve-
niles from one river system to another, etc.) indicated
that the sensitive period in salmon smolts may last from
several days to a few hours. It was hypothesized that not
only the smell of the home area of the river is imprinted,
but also the whole pathway of the migrating fish from
the home area to the river mouth (Harden Jones, 1968;
Hansen et al., 1987; Quinn et al., 1989). These smells
serve as points of orientation for the returning adults
not only in the river but also in the open sea (Wisby and
Hasler, 1954; Groves et al., 1968; Døving et al., 1985;
Stabell, 1992; Døving and Stabell, 2003). It was
hypothesized that the smell of the home river is
imprinted by juvenile sturgeon as well (Boiko et al.,
1993; Kasumyan, 1999).
Chemical markers of the home spawning site
include specific smells of this area. It was proposed that
the smell is created by substances appearing in water by
wash-outs or drains from adjacent land areas, along
with extracts of local aquatic plants and animals
(Hasler and Scholz, 1983). Electrophysiological exper-
iments indicated that samples of fresh water taken from
different water bodies cause responses of different
intensities in the olfactory system of salmonid fish, not
only in adults (Sato et al., 2000). There is also a phero-
Unfolding Small shifts and
Increased respiration
Barbel
Mouth
Rapid eye
movements
and moving fins posture alterations
frequency
movements
openings, cough
Fig. 29. Characteristics of the behavioral response to food
smells in inactive fish or ambush predators.
Table 9. Olfactory sensitivity of fish with different life
modes to water extract of food (Chironomidae larvae)
Species Threshold
concentration, g/l
carp Cyrpinys carpio 10–4–10–5
goldfish Carassius auratus 10–4
Russian sturgeon
Acipenser gueldenstaedtii 10–4
Stellate sturgeon Acipenser stellatus 10–4
Minnow Phoxinus phoxinus 10–4
Gudgeon Gobio gobio 10–4
grass carp Ctenopharyngodon idella 10–3
verkhovka Leucaspius delineatus 10–2
Beluga Huso huso 10–1–10–2
paddlefish Polyodon shathula 10–1–10–2
S206
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
mone homing hypothesis proposed by Norwegian zool-
ogist Hans Nording, that the smell of the home popula-
tion living in the river is the orientation marker for
migrating adults (Nording, 1971, 1977). This hypothe-
sis is probably not universal and cannot be applied to
cases when young leave the river long before the begin-
ning of the spawning migration of adults, as, for exam-
ple, in the pink salmon Oncorhynchus gorbuscha.
(a)
(b) (c)
(e)(d)
(f) (g)
Fig. 30. Feeding behavior of fish with different ecology after stimulation with food smells: (a) carp Cyprinus carpio, goldfish Car-
assius auratus, (c) minnow Phoxinus phoxinus, (d) zebrafish Brachydanio rerio, (e) verkhovka Leucaspius delineatus, (f) chum
salmon Oncorhynchus keta, (g) grayling Thymallus thymnallus, (h) and (i) paddlefish Polyodon spathula. The light silhouette is the
behavior of the fish in the pre-stimulus period, dark silhouette, during the search response.
Bloodworms extract Duckweed extract
Fig. 31. Food search behavior of the grass carp Ctenopharyngodon idella stimulated by extract of bloodworms (Chironomidae lar-
vae) and duckweed Lemna minor.
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S207
However, the orientation markers for such fish may be
bile acids or their derivatives, which could be retained
for a long time in feces left in the river by juveniles
which had previously left for the sea for feeding (Døv-
ing, personal communication).
Olfaction plays an important role in the migration
behavior of the marine lamprey Petromyzon marinus. It
was found that for spawning, adult lampreys enter only
the rivers or their tributaries inhabited by conspecific
ammocete larvae. Feeding larvae constantly excrete
several specific derivates of bile acids. The chemical
markers for adult lampreys include petromyzonole
(3α7α12α24-tetrahydroxy-5α-cholan-24-sulphate) and
allocholic acid (3α7α12α-trihydroxy-5α-cholan-24-ic
acid). These substances are excreted only by larvae and
cannot be found in excretes of lampreys after metamor-
phosis. Petromyzole sulphate and allocholic acid are
perceived with olfaction (Bjerselius et al., 2000; Li
et al., 1995, 2002). It was found that each ammocete
larva on average excretes these substances with a rate of
16 and 5 ng/h, respectively, which is quite enough for
supra-threshold concentrations of the pheromone
(Polkinghorne et al., 2001). However, the characteristic
sexual behavior in female lampreys is observed only in
response to conspecific males ready for spawning,
excreting sexual pheromone 7α12α24-tetrahydroxy-
5α-cholan-3-on-24-sulphate (3-ketopteromyzonol sul-
phate) (Li et al., 2002; Yun et al., 2002; Siefkes et al.,
2003).
The species smell produced by adult conspecifics is
the chemical marker for the catadromous fish Galaxis
fasciatus returning to the home river. This fish migrates
to the sea after hatching (Baker and Montgomery,
2001b). The attracting stimulus for the glass eels
Anguilla anguilla migrating from the sea to fresh water
may include such natural substances as geosmine and
different pyrazines or thyazoles, which are excreted by
different microorganisms. The sensitivity of fish to
them reaches 10–9–10–13 mg/l (Tsoi and Sola, 1993;
Sola, 1995).
Territorial behavior. Many fish are resident and
keep to their home ranges. They are often characterized
by aggressiveness to conspecifics approaching the resi-
dent’s territory. Some territorial fish use smells to mark
their home territory. Such smell marks bear the infor-
mation that this territory is occupied. This reduces the
intensity of aggressive contacts within the population.
Additionally, smell marks help fish not only to protect
their territory from intruders, but also to return if they,
for some reason, need to leave due to feeding migra-
tion, fright, or being moved by the current. Resident
juvenile salmon can mark their territories in rivers by
substances (bile acids and their derivates) which enter
the water with feces (Døving et al., 1980; Seleset,
1980; Seleset and Døving, 1980; Foster, 1985; Stabell,
1987; Courtenay et al., 1997). Removal of the olfactory
organ significantly reduces the ability of the fish (trout
Salmo trutta) to return to their home range (Halvorsen
and Stabell, 1990). In some marine nearshore fish, ter-
ritorial marking could also depend on bile acids or their
derivatives excreted with feces (Døving, personal com-
munication) or substances from the skin mucus left by
the fish on stones and other objects.
Olfaction plays the most important role in choosing
the host actinia in clownfish Amphiprion and related
representatives of the genus Premnas (Pomacentridae).
These fish are territorial and spend most of the time
defended by their actinia. The eggs develop in proxim-
ity to the actinia, but early juveniles do not stay there
and are distributed by currents immediately after hatch-
ing. Reaching a certain age, juveniles return to the reef
zone and easily find host actinias. For example,
Amphiprion melanopus usually lives among the tenta-
cles of Entacmaea quadricolor, more rarely, two other
species of actinias, Heteractic crispa and H. magnifica.
The errorless choice of the host actinia occurs due to an
innate preference for a specific actinia smell, which is
further increased by imprinting when the fish emerge
(Arvedlund and Nielsen, 1996; Arvedlund et al., 1999).
Attempts to artificially imprint the fish to the smell of
an actinia which the fish never encounter in nature were
not successful. The substances causing the actinia smell
which is distinguished by the fish have been synthe-
sized (amphicumines and their analogues, Murata
et al., 1986; Konno et al., 1990). These chemical sig-
nals with interspecific effects belong to the category of
allomones.
The Ontogeny of the Olfactory System
Structure development. The development of the
olfactory system begins early in the mid-embryonic
period when the olfactory placode, from which the
future olfactory organ develops, becomes easily dis-
cernible (Zeiske et al., 2003). The olfactory depression
appears in the central part of the placode, either because
of invagination of the outer cell layer or by first the
appearance of a cavity and then widening as the cover-
ing cells flake away. The depth and size of the olfactory
depression rapidly increases, and the first specialized
cells, including receptor, cells, ciliated nonsensory
cells, supporting, and mucous cells, form (Fig. 32). Up
to the moment of hatching, the number of cells in the
olfactory organ significantly increases. Further increase
of the size of the olfactory depression occurs, the first
olfactory lamella (raphe) appear, then their number and
size increase, which also increases the number of olfac-
tory receptor cells. The olfactory rosette is formed, and
then the unitary olfactory opening is split into the ante-
rior and posterior nostrils (Teichmann, 1954; Pfeiffer,
1963, 1964; Evans et al., 1982; Zelinski and Hara,
1988; Hansen and Zeiske, 1993; Werner and Lannoo,
1994; Appelbaum and Reihl, 1997; Arvedlund et al.,
2000). In cyprinid fishes, for example, the first olfac-
tory raphe appears in the middle of the larval period of
development (stage D1 in the minnow Phoxinus phoxi-
nus and shemaya Chalcalburnus chalcoides, D2 in the
S208
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
grass carp Ctenopharyngodon idella, silver carp
Hypophthalmichthys molitrix and carp Cyprinus car-
pio, E in bitterling Rhodeus sericeus. The merging of
the superficial outgrowths and the development of the
nasal split occurs at the end of the larval or at the begin-
ning of the fry period (stages E and F). The first side
lamellae, symmetrical with respect to the central raphe,
appear at this time. In this way, the first olfactory rosette
consisting of three lamellae appears. Further develop-
ment of new lamellae also occurs in pairs, one from
each side of the central raphe. During the end of the
second fry period of development, the nasal ridge
develops in the nasal split, the number of lamellae in
the rosette becomes even more, seven, and begins to
resemble the rosette of adult individuals by its struc-
ture. The olfactory organ takes a definitive form (Pash-
chenko and Kasumyan, 1983, 1986).
With further growth of the fish, the size of the olfac-
tory organ also increases, as does the number of lamel-
lae in the rosette (Fig. 33) (Halama, 1982).The receptor
structures become stronger with the development of the
olfactory organ and its basic structures, especially rap-
idly after the development of the first lamellae (Fig.
34). In flatfishes, the olfactory organ at the eyed side of
the body develops more rapidly than on the blind side
(Harvey, 1996).
Function development. The terms of development
of the olfactory functions have been quite thoroughly
studied. According to these data, the ability to respond
to olfactory stimuli develops in the olfactory organ
when the first receptor cells appear.
It has been found that even before transition to
external feeding, prolarvae of several marine fishes
(cod Gadus morhua, Scophtalmus maximus) shows
nonspecialized behavioral responses to such chemical
stimuli as food extracts and free amino acids. The
responses of prolarvae to these stimuli is expressed in
slower swimming, reduction of discrete swimming
movements, reduction of the distance moved per motor
act, and reduction of the distance between the position
of the fish in three-dimensional space during observa-
tion (Fig. 35). It is hypothesized that such alterations of
locomotor activity cause prolarvae that should soon
switch to external feeding to lag in the zone of the smell
created by potential food objects. The adaptive sense of
this behavior includes an increase in the probability of
(a) (b)
(c) (d)
(e) (f)
Fig. 32. Development of the olfactory organ in the Siberian
sturgeon Acipenser baerii: (a) and (b) head part of the
embryo and the olfactory depression (age, 110 h); (c) and
(d) olfactory depression (age 132 and 156 h); (e) and (f) cil-
iated receptor cells and ciliated nonsensory cells (age 120 h).
30
10 Body length, cm
20 40 50 60 70
10
40
60
Number of olfactory lamellae
70
50
20
30 80
0
Fig. 33. Increase of the number of olfactory lamellae in the
ontogeny of the grass carp Ctenopharyngodon idella (from
Pashchenko, 1986).
2500
60 Body length, mm
20 100 140 180
500
3500
5500
Number of receptor cells, thousands
4500
1500
Fig. 34. Increase of the number of receptor cells in the
ontogeny of the grass carp Ctenopharyngodon idella (from
Pashchenko, 1986).
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S209
food detection during the critically important period of
transition to external feeding (Døving et al., 1994;
Kasumyan et al., 1998).
The development of specialized behavioral
responses to olfactory stimuli begins later, during the
larval period of development. The age of the appear-
ance and development of the olfactory sensitivity to dif-
ferent smells differs. For example, it has been shown in
many cyprinid fish species that the young begin to
respond to the alarm pheromone at the beginning of the
larval period, in the period of mixed feeding (Fig. 36).
Sensitivity to the alarm pheromone (as well as to the
alarm kairomone) develops slowly and finishes only in
fingerlings or yearlings. The defensive response to nat-
ural chemical signals of danger forms more slowly than
the ability to detect threshold concentrations of these
smells. Even more time is required for the fish to
acquire the ability to distinguish the alarm pheromones
of different species (Fig. 37) (Pashchenko and
Kasumyan, 1983, 1986; Døving et al., 2004).
Juveniles begin to respond to food at the same age,
during the prolarval stage. It is clear by the example of
chemical stimuli that development of the olfactory sen-
sitivity occurs in close relationship with the ecology of
juveniles and their general sensory equipment. For
example, responses to food stimuli develop earlier and
more quickly in fish to which olfaction has a major sig-
nificance in feeding behavior (Fig. 38). For example,
the response to the food odor in sturgeons, whose feed-
ing behavior is significantly based on olfaction, appears
just after the transition of larvae to external feeding, and
the sensitivity reaches the definitive level in the middle
60
0
Control
50
40
30
20
10
End
Start
60
010
10–5 M
20 30 40 50 60
50
40
30
20
10
End
Start
10–9 M End
Start
10–8 M
End
Start
10
10–4 M
20 30 40 50 60
End
Start
10
10–3 M
20 30 40 50 60
End
Start
Distance, mm
Distance, mm
Fig. 35. Trajectory of cod Gadus morhua prolarvae stimulated by different concentrations of L-arginine amino acid solutions. Con-
trol: movement of prolarvae in artificial sea water; End and Start: beginning and termination of observation period, the duration of
recording 1 min. Points depict short-term stops of prolarvae (from Døving et al., 1994).
10
10–1
30 50 70 90
Body length, mm
Larval Fry period
10–3
10–5
10–7
10–9
1
2
34
5
6
Concentration of skin extract, g/l
period
Fig. 36. Change of sensitivity to the alarm pheromone in the
ontogeny of the grass carp Ctenopharyngodon idella (1),
bitterling Rhodeus sericeus (2), kutum Rutilus frisii kutum
(3), zherekh Aspius aspius (4), shemaya Chalcalburnus
chalcoides (5), silver carp Hypophthalmichthys molitrix (6)
(from Kasumyan, 1982).
S210
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
of the second month of the juveniles’ life (Kasumyan
and Kazhlayev, 1993; Kasumyan and Taufik, 1994;
Kasumyan, 1999). The sensitivity to food odor devel-
ops less rapidly in cyprinid bentivorous fish (carp Cyp-
rinus carpio) and even slower in pelagic planktivorous
fish, for which visual reception is of greater value not
only in juveniles but also in adults (zebrafish Brachy-
danio rerio) (Kasumyan and Ponomarev, 1990).
Effect of Pollutants and Other Destructive Factors
to the Olfactory System
Olfactory receptors are open to the environment and
are susceptible to the effects of different unfavorable
environmental factors, both biological and abiotic. Dif-
ferent organisms may live in the olfactory organ (bacte-
ria, microsporidia, crustaceans) and disturb or elimi-
nate a significant part of the olfactory epithelium
(Devitsyna, 1977; Doroshenko and Pinchuk, 1978;
Pashchenko and Kasumyan, 1984; Baatrup and Døv-
ing, 1990; Klaprat at al., 1992; Morrison and Plumb,
1994; Watson et al., 1998). Various chemical pollut-
ants, such as heavy metals and detergents, are very dan-
gerous (Baatrup et al., 1990; Klaprat et al., 1992; Døv-
ing, 1991; Winberg et al., 1992; Kasumyan, 2001).
Even short-term (for few minutes or seconds) effects of
such substances cause rapid degeneration of the olfac-
0
1
2
10–1 10–3 10–5
Larvae, 8–10 mm
intensity, points
Fright reaction
0
1
2
10–1 10–3 10–5
Fry, 16–18 mm
0
1
2
10–1 10–3 10–5
Fry, 25–30 mm
3
4
***
0
1
2
10–1 10–3 10–5
Adults, 35–40 mm
3
4
***
**
Bitterling skin
Spined loach skin
Concentration of skin extract, g/l
Fig. 37. The intensity of the fright reaction in the bitterling Rhodeus sericeus amarus of different age groups to the extract of con-
specific skin and skin of spined loach Cobitis sp. (from Kasumyan and Ponomarev, 1986b). *, **, *** Significant differences from
the control, respectively, p < 0.05; p < 0.01; p < 0.001.
2
1
2Body length, cm
4 6 8 10 12
1
3
42
Threshold concentration of food extract, g/l, –lg
Fig. 38. Increase of the olfactory sensitivity to chemical
food signals in the ontogeny of carp (1) and sturgeon (2).
intensity, points
Fright reaction
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S211
tory epithelium, as well as the destruction of cilia and
microvilli of receptor cells and kinocilia of ciliated
nonsensory cells, and, as a consequence, the complete
loss of the ability to perceive chemical stimuli by the
fish (baker and Montgomery, 2001; Beyers and Farmer,
2001).
The loss of sensitivity caused by pollutants is revers-
ible and olfaction may be restored after several hours.
Regeneration occurs due to the formation of new recep-
tor cells and other cell types from the basal cells. The
restoration of the olfactory function may be so rapid as
to outstrip the regeneration of the cellular structures of
the olfactory epithelium (Fig. 39). It was found that
only 1/3 of the overall number of receptor elements
found in intact fish is enough for normal sensitivity to
smells (Kasumyan and Pashchenko, 1982; Pashchenko
and Kasumyan, 1984). The ability to regenerate and the
ability to restore a normal level of sensitivity due to the
receptor cells pool are extremely important adaptive
mechanisms increasing the functional reliability of the
olfactory system, allowing the fish to obtain diverse
information from the environment (Kasumyan, 1991).
ACKNOWLEDGMENTS
The author is very grateful to E.A. Marusov for
valuable criticism on the manuscript. This work was
partly supported by the program “Leading scientific
schools” (project NSh-1334.2003.4) and the program
“Universities of Russia” (topic UR.07.03.011).
REFERENCES
1. M. A. Adams, J. H. Teeter, Y. Katz, and P. B. Johnsen,
Sex Pheromone of the Sea Lamprey (Petromyzon mari-
nus): Steroid Studies, J. Chem. Ecol. 13, 387–395
(1987).
2. J. M. Al Hassan, M. Thomson, and R. S. Criddle, “Com-
position of Proteinaceous Gel Secretion from the Skin
of the Arabian Gulf catfish (Arius thalassinus),” Mar.
Biol. 70, 27–33 (1982).
3. A. C. Allison and R. T. T. Warwick, “Quantitative
Observations on the Olfactory System of the Rabbit,
Brain. 72, 186–197 (1949).
4. A. P. Andriyashev, “Role of Sensory Organs in the Food
Search of the Silver Rockling,” Zh. Obshch. Biol. 5,
123–127 (1944).
5. A. P. Andriyashev, “Role of the Sensory Organs in Food
Search of Fish,” Proc. Workshop on Methods for the
Study of the Food Resources of Fish Stocks (Acad.
Nauk SSSR, Moscow, 1955), pp. 135–142.
6. S. Appelbaum and R. Riehl, “Scanning Electron Micro-
scopic Observations of the Chemo-and Mechanorecep-
tors of Carp Larvae (Cyprinus carpio) and their Rela-
tionship to Early Behaviour,” Aquat. Living Resour. 10,
1–12 (1997).
7. C. W. Appelt and P. W. Sorensen, “Freshwater Fish
Release Urinary Pheromones in a Pulsatile Manner,” in
Advances in Chemical Signals in Vertebrates, Ed. by
R. E. Johnston, D. Muller-Schwartze, and P. W. Sorensen,
(Kluwer, New York, 1999), pp. 247–256.
8. M. P. Aronov, Role of Sensory Organs in Food Search
of the Drum Corvina umbra, Vopr. Ikhtiol. 1, 194–199
(1962).
9. M. P. Aronov, “Role of Sensory Organs in Obtaining
Food in Fish,” Usp. Sovrem. Biol. 54, 128–145 (1962).
10. M. Arvedlund, K. Larsen, and H. Winsor, “The Embry-
onic Development of the Olfactory System in Amphip-
rion melanops (Perciformes: Pomacentridae) Related to
the Host Imprinting Hypothesis,” J. Mar. Biol. Ass. U.
K. 80, 1103–1110 (2000).
11. M. Arvedlund, M. I. McCormick, D. G. Fautin, and
M. Bildsee, “Host Recognition and Possible Imprinting in
the Anemonefish Amphiprion melanopus (Pisces: Poma-
centridae),” Mar. Ecol. Prog. Ser. 188, 207–218 (1999).
12. M. Arvedlund and L. E. Nielsen, “Do the Anemonefish
Amphipnon ocellaris (Pisces: Pomacentridae) Imprint
Themselves to Their Host Sea Anemone Heteractis
magnificii (Anthozon: Actinidae)?” Ethology 102, 197–
211 (1996).
13. J. Atema, “Chemical Senses, Chemical Signals and
Feeding Behaviour in Fishes,” in Fish Behaviour and its
Use in the Capture and Culture of Fishes, (Manila,
1980), pp. 57–101.
14. J. Atema, “Distribution of Chemical Stimuli,” in Sen-
sory Biology of Aquatic Animals Ed. by J. Atema,
A. N. Popper, R. R. Fay, and W. N. Tavolga, (Springer,
New York, 1988), pp. 29–56.
15. J. Atema, K. Holland, and W. Ikehara, “Olfactory
Responses of Yellowfin Tuna (Thunnus albacares) to
Prey Odors: Chemical Search Image,” J. Chem. Ecol. 6,
457–465 (1980).
16. J. W. Atz, “Internal Nares in the Teleost Astroscopus,
Anat. Rec. 113, 105–115 (1952a).
17. J. W. Atz, “Natural Breathing in Fishes and the Evolu-
tion of Internal Nares,” Q. Rev. Biol. 27, 366–377
(1952b).
18. E. Baatrup and K. B. Døving, “Histochemical Demon-
stration of Mecury in the Olfactory System of Salmon
123456
2
3
4
5
6
7
8
9
10
0
25
50
75
100
Number of receptor cells, %
Concentration of pheromone –lg/l
1
2
Time, days
Fig. 39. Increase of the sensitivity to the alarm pheromone
(1) and the number of olfactory receptor cells (2) in the
grass carp Ctenopharyngodon idella after treatment of the
olfactory epithelium with the detergent Triton X-100 (from
Pashchenko and Kasumyan, 1984).
S212
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
(Salmo salar L.) Following Treatments with Dietary
Methylmercuric Chloride and Dissolved Mercuric
Chloride,” Ecotoxicol. Environ. Safety 20, 277–289
(1990).
19. E. Baatrup, K.B. Døving, and S. Winberg, “Differential
Effects of Mercurial Compounds on the Electroolfacto-
gram (EOG) of Salmon (Salmo salar L.),” Ecotoxicol.
Environ. Safety 20, 269–276 (1990).
20. R. C. Baird, G. Y. Jumper, and E. E. Gallaher, “Sexual
Dimorphism and Demography in two Species of Oce-
anic Midwater Fishes (Stomiiformes: Stemoptychidae)
from the Eastern Gulf of Mexico,” Bull. Mar. Sci. 47,
561–566 (1990)
21. C. F. Baker and J. C. Montgomery, Sensory Deficits
Induced by Cadmium in Banded Kokopu, Galaxias fas-
ciatus, Juveniles, Environm. Biol. Fishaa. 62, 455–464
(2001a).
22. C. F. Baker and J. C. Montgomery, “Species-Specific
Attraction of Migratory Banded Kokopu Juveniles to
Adult Pheromones,” J. Fish Biol. 58, 1221–1229
(2001b).
23. E. K. Bakhtin, “Characteristics of Microstructure of the
Olfactory Organin the Dogfish Squalus acanthias,” Tsi-
tologiya 19, 725–731 (1977).
24. E. K. Bakhtin and E. E. Filyushina, “Ultrastructural
Characteristics of Sectret Formation in Minor Secret
Cellsof the Olfactory Epithelium of Zander and Perch,
Tsitologiya 16, 90–93 (1974).
25. H. D. Baldridge, “Shark Repellent: Not Yet, Maybe
Never,” Military Medicine 155, 358–361 (1990).
26. L. H. Bannister, The Fine Structure of the Olfactory
Surface of Teleostean Fishes, Quart J. Micr. Sci. 106,
333–342 (1965).
27. G. P. Barbin, “The Role of Olfaction in Homing and
Estuarine Migratory Behavior of Yellow-Phase Ameri-
can Eels,” Can. J. Fish. Aquat. Sci. 55, 564–575 (1998).
28. J. E. Bardach and J. Case, Sensory Capabilities of the
Modified Fins of Squirrel Hake (Urophycis chuss) and
Searobins (Prionotus carolinus and P. evolans), Copeia,
No. 2, 194–206 (1965).
29. J. E. Bardach and J. H. Todd, “Chemical Communica-
tion in Fish,” in Communication by Chemical Signals,
Ed. by J. W. Johnston jun., D. G. Moulton R.A., and
A. Turk (Appelton-Century-Crofts., New York, 1970),
pp. 205-240.
30. D. P. Bashor, R. W. Beuerman, and D. M. Easton, “Cil-
iary Action and Normal Movement of Odorant Wave
Fronts in Garfish Nasal Capsule of Lepisosteus asseus,”
Experientia 30, 777–779 (1974).
31. E. Bertelsen, “Notes on Linophrynidae V: A Revision of
the Deepsea Anglerfishes of the Linophryne arbrifera
Group (Pisccs, Ceralioidei),” Steenstrupia 6, 29–70
(1980).
32. E. Bertelsen and J. G. Nielsen, “The Deep Sea Eel Fam-
ily Monognathidae (Pisces, AnguillifonDes),” Steen-
strupia 13, 141–198 (1987).
33. R. Berti, R. Vezzosi, and A. Ercolmi, “Locomotory
Response of Phreatichthys andruzzii Vinciguerra
(Pisces, Cyprinidae) to Chemical Signals of Conspecif-
ics and Closely Related Species,” Experientia 45, 205–
207 (1989).
34. R. Berti and L. Zom, “Locomotory Responses of the
Cave Cyprinid Phreatichthys andruzzii to Chemical sig-
nals from Conspecifics and Species: New Findings,
Environai. Biol. Fisher. 62, 107–114 (2001).
35. G. Bertmar, “The Vertebrate Nose, Remarks on Its
Structural and Functional Adaptation and Evolution,”
Evolution. 23, 131–152 (1969).
36. G. Bertmar, “Ecostructural Studies on Olfactory Organ
in Young and Adult Sea Trout (Osteichthyes, Salmo-
nidae),” Z. Morphol. Okol. Tiere. 72, 307–330 (1972).
37. D. W. Beyers and M. S. Fanner, “Effects of Copper on
Olfaction of Colorado Pikeminnow,” Environm. Toxi-
col. Chem. 20, 907–912 (2001).
38. R. Bjerselius, W. Li, J. H. Teeter, J. G. Seely, et al.,
“Direct Behavioral Evidence that Unique Bile Acids
Released by Larval Sea Lamprey Function as a Migra-
tory Pheromone,” Can. J. Fish. Aquat. Sci. 57, 1–14
(2000).
39. H. D. Bloom and A. Perlmutter, “Possible Pheromone
Mediated Reproductive Isolation in two Species of
Cyprinid Fishes of the Genus Brachydanio,” J. Fish
Biol. 13, 47–50 (1978).
40. N. E. Boiko, R. A. Grigoryan, and A. S. Chikhachev,
Olafctory Imprinting in Juveniles of the Russian Stur-
geon Acipenser guldenstadti, Zh. Evol. Biokhim. Fiz-
iol. 29, 509–513 (1993).
41. J. G. Brand and R. C. Bruch, “Molecular Mechanisms
of Chemosensory Transduction: Gestation and Olfac-
tion,” in Fish Chemoreception, Ed. by T. J. Hara (Chap-
man and Hall, London, 1992), pp. 126–149.
42. J. R. Brett and D. McKinnon, “Some Aspects of Olfac-
tory Perception in Migrating Adult Coho and Spring
Salmon,” J. Fish. Res. Bd. Can. 11, 310–318 (1954)
43. C. Brönmark and L. B. Miner, “Predator-Induced Phe-
notypical Change in Body Morphology in Crucian
Carp,” Science 258, 1348–1350 (1992).
44. A. A. Bronstein Olfactory Receptors of Vertebrates
(Nauka, Leningrad, 1977) [in Russian].
45. A. A. Bronstein, “Some Peculiarities of Fine Structure
of the Olfactory Organ in Elasmobranches,” J. Evol.
Biochem. Physiol. 12, 63–67 (1976).
46. A. Brunn, von, “Beitrage zur mikroskopischen Anat-
omie der menschlichen Nasenhöhle,” Arch. Mikr. Anat
39, 632–651 (1892).
47. P. J. Bryer, R. S. Mirza, and D. P. Chivers, “Chemosen-
sory Assessment of Predation Risk by Slimy Sculpins
(Cottus cognatus): Responses to Alarm, Disturbance
and Predator Cues,” J. Chem. Ecol, 27, 533–546 (2001).
48. R. H. Bume, “The Anatomy of the Olfactory Organ of
Teleostean Fishes,” Proc. Zool. Soc. London. 2, 610–
663 (1909).
49. A. V. M. Canario and A. P. Scott, “Conjugates of Ova-
rian Steroids, Including 17α,20β-dihydroxy-4-pregnen-
3-one (Maturation-Inducing Steroid), Accumulate in
the Urine of a Marine Teleost (Plaice, Pleuronectes pla-
tessa),” J. Endocrinol. 123, 1–4 (1989).
50. P. Cancalon, “Degeneration and Regeneration of the
Olfactory Cells Induced by ZnSO4 and Other Chemi-
cals,” Tissue and Cell. 14, 717–733 (1982).
51. J. Caprio, “Olfaction and Taste in the Channel Catfish:
an Electrophysiological Study of the Responses to
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S213
Amino Acids and Derivates,” J. Comp. Physiol. 123,
357–371 (1978).
52. J. Caprio, “Similarity of Olfactory Receptor Responses
(EOG) of Freshwater and Marine Catfish to Amino
Acids,” Can. J. Zool. 58, 1778–1784 (1980).
53. J. Caprio, “High Sensitivity and Specificity of Olfactory
and Gustatory Receptors of Catfish to Amino Acids,” in
Chemoreception in Fishes, Ed. by T. J. Hara (Elsevier,
Amsterdam, 1982), pp. 109–134.
54. J. Caprio, “Olfaction and Taste in Fish,” in Comparative
Physiology of Sensory systems, Ed. by L. Bolis, R. D. Key-
nes, and S. H. P. Maddrell (Cambridge Univ. Press,
Cambridge, 1984), pp. 257–283.
55. J. Caprio and R. Bird, Jr., “Electrophysiological Evi-
dence Far Acidic, Basic and Natural Amino Acid Olfac-
tory Receptor Sites in the Catfish,” J. Gen. Physiol. 84,
403–422 (1984).
56. M. G. Carr and J. E. Carr, “Individual Recognition in
the Jnvenue Brown Bullhead (Ictalums nebulosus),”
Copeia, No. 4, 1060–1062 (1985)
57. W. E. S. Carr, J. C. III. Netherton, R. A. Gleeson, and
C. D. Derby, “Stimulants of Feeding Behavior in Fish:
Analyses of Tissues of Diverse Marine Organisms,”
Biol. Bull. 190, 149–160 (1996).
58. J. H. Caruso, “Sexual Dimorphism of the Olfactory
Organs of Lophiids,” Copeia, 380–381 (1975).
59. X.-Y.Chen and G. Arratia, “Olfactory Organ of Aci-
penseriformes and Comparison with other Actinoptery-
gians: Patterns of Diversity,” J. Morphol. 222, 241–267
(1994).
60. D. P. Chivers and R. J. F. Smith, “Chemical Alarm Sig-
nalling in Aquatic Predator-Prey Systems: A Review
and Prospectus,” Ecoscience 5, 338–352 (1998).
61. E. Clark, “Shark Repellent Effect of the Red Sea Moses
Sole, Pardachirus mannoratus,” Nat. Geogr. Soc. Res.
Rep. 13, 177–186 (1981).
62. E. Clark, “Shark Repellent Effect of the Red Sea Moses
Sole,” in Shark Repellents from the Sea: New Perspec-
tives, Ed. by B. J. Zahuranec (Westview Press, Boulder,
1983), pp. 135–150.
63. L. Colombo, A. Muconato, B. P. Colombo, and C. Friso,
“Endocrinology of Teleost Reproduction: A Testicular
Steroid Pheromone in the Black Goby Gobius jozo L.,”
Bol. Zool. 47, 355–364 (1980).
64. S. C. Courtenay, T. P. Quinn, H. M. C. Dupuis, et al.,
“Factors Affecting the Recognition of Population-Spe-
cific Odours by Juvenile Coho Salmon,” J. Fish Biol.
50, 1042–1060 (1997).
65. S. C. Courtenay, T. P. Quinn, H. M. C.Dupuis, et al.,
“Discrimination of Familty-Specific Odours by Juve-
nile Coho Salmon: Roles of Learning and Odour Con-
centration,” J. Fish Biol. 58, 107–125 (2001).
66. U. Dahlgren, “The Oral Opening of the Nasal Cavity in
Astroscopus,” Science 27, 993–994 (1908).
67. M. De Fraipont and G. Thines, “Responses of the Cave-
fish Astyanax mexicanus (Anoptichthys antrobius) to
the Odor of Known and Unknown Conspecifics,” Expe-
rientia 42, 1053–1054 (1986).
68. J. H. Derivot, “Functional Anatomy of the Peripheral
Olfactory System of the African Lungfish Protopterus
annectens Owen: Macroscopic, Microscopic, and Mor-
phometric Aspects,” Am. J. Anat. 169, 177–192 (1984).
69. G. V. Devitsyna, “Morphology of the Olfactory Organ
in Cods (fam. Gadidae),” Vopr. Ikhtiol. 12, 1094–1103
(1972).
70. G. V. Devitsyna, “Comparative Study of Olfactory Ana-
lyzer Morphology in Fish,” Vopr. Ikhtiol. 17, 129–133
(1977).
71. G. V. Devitsyna and A. A. Kazhlayev, “Development of
Chemosensory Organs in the Siberian Sturgeon Aci-
penser baeri and Stellate Sturgeon A. stellatus,” Vopr.
Ikhtiol. 32, 167–176 (1992).
72. G. V. Devitsyna and A. A. Kazhlayev, “Chemoreception
Organs in Early Juveniles of Paddlefish Polyodon
spathula,” Vopr. Ikhtiol. 33, 142–146 (1993).
73. G. V. Devitsyna and G.A. Malyukina, Functional Orga-
nization of the Olfactory Organ in Fish Micro and Mac-
rosmatics, Vopr. Ikhtiol. 17, 493–502 (1977).
74. G. V. Devitsyna and A. B. Al Attar el Saied, “Cytoarchi-
tectonics and Morphometry of the Olfactory Analyzer
of Fish Micro- and Macrosmatics,” Vopr. Ikhtiol. 28,
837–845 (1988).
75. A. S. Dogel, The Structure of the Olfactory Organ in
Ganoids, Teleost Fish, and Amphibians, Tr. Obshch.
Estestvoispyt pri Kazan Univ. 16, 3–82 (1886).
76. M. A. Doroshenko and L. E. Pinchuk, “New Data on
Secretory Elements of the Olfactory Epithelium of
Some Marine Fish,” Dokl. Akad. Nauk SSSR, 219,
756–758 (1974).
77. M. A. Doroshenko and L. E. Pinchuk, “Characteristics
of Morphology of the Olfactory Epithelium of Marine
Fish Disturbed by Microsporidias,” Biol. Morya, No. 3,
88 (1978).
78. K. B. Døving, “The Influence of Olfactory Stimuli upon
the Activity of Secondary Neurons in the Burbot (Lota
lota L.),” Acta Physiol. Scand. 66, 290–299 (1966).
79. K. B. Døving, “Functional Properties of the Fish Olfac-
tory System,” in Progress in Sensory Physiology, Vol. 6,
Ed. by D. Ottoson (Springer, Berlin, 1986), pp. 39–104.
80. K. B. Døving, “Molecular Cues in Salmonid Migra-
tion,” in Molecules in Physics, Chemistry and Biology,
Ed. By Jean Maruani. Vol. IV (Academic Press, 1989),
pp. 299–329.
81. K. B. Døving, “Assessment of Animal Behaviour as a
Method to Indicate Environmental Toxocoty,” Compar.
Biochem. and Physiol. 100, 247–252 (1991)
82. K. B. Døving, M. Dubois-Dauphin, A. Holley, and
F. Jourdan, “Functional Anatomy of the Olfactory
Organ of Fish and the Ciliary Mechanism of Water
Transport,” Acta Zool., Stockholm 58, 245–255 (1977).
83. K. B. Døving, G. Gemne, “Electrophysiological and
Histological Properties of the Olfactory Tract of the
Burbot (Lota lota L.),” J. Neurophysiol. 28, 139–153
(1965).
84. K. B. Døving, E. H. Hamdani, E. Hoglund, et al., “A
Review on the Chemical and Physiological Basis of
Alarm Reactions in Cyprinids,” in Fish Chemosenses,
Ed. by K. Reutter and B. G. Kapoor (Sci. Publishers,
Enfield, 2004).
S214
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
KASUMYAN
85. K. B. Døving and K. Holmberg, “A Note on the Func-
tion of the Olfactory Organ of the Hagfish Myxine glu-
tinosa,” Acta Physiol. Scand. 91, 430–432 (1974).
86. K. B. Døving, B. Jonsson, and L. P. Hansen, “The Effect
of Anosmia on the Migration of Atlantic Salmon Smolts
(Saimo salar L.) in Freshwater,” Aquaculture 38, 383–386
(1984).
87. K. B. Døving, M. Mårstøl, J.R. Andersen, and
J. A. Knutsen. “Experimental Evidence of Chemokine-
sis in Newly Hatched Cod Larvae (Gadus morhua L.),
Mar. Biol. 120, 351–358 (1994).
88. K. B. Døving K.B., H. Nordeng, and B. Oakley, “Single
Unit Discrimination of Fish Odours Released Char
(Saimo alpinus L.) Populations,” Comp. Biochem.
Physiol. 47A, 10511063 (1974).
89. K. B. Døving and R. Selset. “Behavior Patterns in Cod
Released by Electrical Stimulation of Olfactory Tract
Bundles,” Science 207, 559–560 (1980).
90. K. B. Døving, V. L. Selset, and G. Tommesen, “Olfac-
tory Sensitivity to Bile Acids in Salmomd Fishes,” Acta
Physiol. Scand. 108, 123–131 (1980).
91. K. B. Døving and O. B. Stabell. “Trails in Open Waters:
Sensory Cues in Salmon Migration,” in Sensory Pro-
cessing in Aquatic Environments, Ed. by S. P. Collin and
N. J. Marshall (Springer, New York, 2003), pp. 39–52.
92. K. B. Døving and D. Trotier, “Structure and Function of
the Vomeronasal Organ,” J. Exp. Biol. 201, 2913–2925
(1998).
93. K. B. Døving K.B., H. Westerberg, and P.B.Johnsen,
“Role of Olfaction in the Behavioural and Neural
Responses of Atlantic Salmon, Saimo salar, to Hydro-
graphic Stratification,” Can. J. Fish. Aquat. Sci. 42,
1658–1667 (1985).
94. J. G. Dulka, N. E. Stacey, P. W. Sorensen, and G. J. Van
Der Kraak, “A Steroid Sex Pheromone Synchronizes
Male-Female Spawning Readiness in Goldfish,” Nature
325, 251–253 (1987).
95. R. Eaton, “Notes on the Olfactory Organs in Cen-
trarchidae,” Copeia, No. 3, 196–199 (1956).
96. H. L. Eistein, “Phytogeny of the Vomeronasal System
and Receptor Cell Types in the Olfactory and Vomero-
nasal Epithelia of Vertebrates,” Microscopy Res. Tech.
23, 1–21 (1992).
97. R. E. Evans R.E., B. Zielinski, and T. J. Hara, “Devel-
opment and Regeneration of the Olfactory Organ in
Rainbow Trout,in Chemoreception in Fishes, Ed. by
T.J.Hara (Elsevier, Amsterdam, 1982), pp. 15–37.
98. B. A. Flerov, “Olfaction in Fish,” Vopr. Ikhtiol. 2, 517–
529 (1962).
99. N. R. Foster N.R, Lake Trout Reproductive Behaviour:
Influence of Chemosensory Cues from Young of the
Year By-Products, Trans. Amer. Fish. Soc. 114, 794–
803 (1985).
100. P. Frade, P. C. Hubbard, E. N. Barata, V. M. Canário,
“Olfactory Sensitivity of the Mozambique Tilapia to Con-
specific Odours,” J. Fish Biol. 61, 1239–1254 (2002).
101. K. Frisch, “Über einen Schreckstoff der Fishhaut und
seine biologische Bedeutung,” Z. vergl. Physiol. 29, 46–
145 (1941a).
102. K. Frisch von, “Die bedeutung der Geruchsinnes in Leben
des Fisches,” Nalurwissenschaften 29, 321–333 (1941b).
103. G. Gemne G., K. B. Døving, “Ultrastructural Properties
of Primary Olfactory Neurons in fish (Lota lota L.),”
Amer. J. Anat. 126, 457–475 (1969).
104. P. W. Gilbert P.W., “The Visual Apparatus of Sharks and
its Probable role in Predation,” Tenth Pacific Sci. Con-
gress, Abstr. Symp. Papers. Honolulu, Hawaii, pp. 176–
177 (1961).
105. P. W. Gilbert and C. Gilbert, “Sharks and Sharks Deter-
rents,” J. Underwater 5, 69–79 (1973).
106. P. W. Gilbert and S. Springer, “Testing Shark Repel-
lents,” in Sharks and Survival, Ed. by P. W. Gilbert
(Health and Company, Boston, 1963), pp. 477–494.
107. H. R. Goel, Functional Anatomy of the Olfactory
Organs in the Fresh Water Indian Catfish Clarias batra-
chus (Linn.), Zool. Poloniae. 28, 5–14 (1980).
108. A. V. Golubev and E. A. Marusov, “Studies of Chemical
Reception by the White Sea Three-Spined Stickleback
(Gasterosteus aculeatus L) during the Reproductive
Period,” Nauch. Dokl. Vyssh. Shkoly. Biol. Nauki,
No. 9, 33–35 (1979).
109. A. V. Golubev and E. A. Marusov, “Some Characteristics of
Chemical Communication in the Three-Spined Stickleback
Gasterosteus aculeatus L (Gasterosteidae),” Vopr. Ikhtiol.
24, 829–835 (1984).
110. V. B. Goryunova, V. G. Shagaeva, and N. F. Shevchenko,
An Analysis of Developmental Abnormalities in Mor-
phology of the Sturgeon Larvae and Juveniles in the
Volga-Caspian Basin under Conditions of Artificial
Reproduction,” Vopr. Ikhtiol. 40, 804–809 (2000).
111. H. Göz, “Uber den Art- und Individualgeruch bei Fis-
chen,” Z. vergl. Physiol. 29, 1–45 (1941).
112. P. P. C. Graziadei and G. A. Monti-Graziadei, “Contin-
uous Nerve Cell Renewal in the Olfactory System,” in
Handbook of Sensory Physiology. IX. Development of
Sensory System, Ed. by M. Jacobson (Springer, Heidel-
berg, 1978), pp. 55–83.
113. C. Groot, T. P. Quinn, and T. J. Hara, “Responses of
Migrating Adult Sockeye Salmon (Oncorhynchus
nerka) to Population-Specific Odours,” Can. J. Zool. 64,
926–932 (1986).
114. S. J. Groot, de, “Only the Inter-Relationships between
Morphology of the Alimentary Tract, Food and Feeding
Behaviour in Flatfishes (Pisces: PIeuronectiformes),
Neth. J. Roy. Soc. B. 5, 121–196 (1971).
115. A. B. Groves A.B., G. B. Collins, and P. S. Trefethen,
“Roles of Olfaction and Vision on Choice of Spawning
Site of Homing Adult Chinook Salmon (O. tshaw-
ytscha),” J. Fish. Res. Bd. Can. 25, 867–876 (1968).
116. L. Halama, “Morphometry of the Olfactory Organs of
the Rainbow Trout (Salrno gairdneri Rich.),” Acta Biol.
Cracoviensia. Ser. Zool. 24, 105–111 (1982).
117. M. Halvorsen and O. B. Stabell, “Homing Behavior of
Displaced Stream-Dwelling Brown Trout,” Anim.
Behav. 39, 1089–1097 (1990).
118. E. H. Hamdani, G. Alexander, and K. B. Døving, “Pro-
jection of Sensory Neurons with Microvilli to the Lat-
eral Olfactory Tract Indicates their Participation in
Feeding Behaviour in Crucian Carp,” Chem. Senses 26,
1139–1144 (200la).
119. E. H. Hamdani and K. B. Døving, “The Alarm Reaction
in Crucian Carp is Mediated by Olfactory Neurones
JOURNAL OF ICHTHYOLOGY Vol. 44 Suppl. 2 2004
THE OLFACTORY SYSTEM IN FISH: STRUCTURE, FUNCTION, AND ROLE IN BEHAVIOR S215
with Long Dendrites,” Chem. Senses. 27, 395–398
(2002).
120. E. H. Hamdani and K. B. Døving, “Sensitivity and
Selectivity of Neurons in the Medial Region of the
Olfactory Bulb to Ssin Extract from Conspecifics in
Crucian Carp, Carassius carassius,” Chem. Senses 28,
181–189 (2003).
121. E. H. Hamdani, A. Kasumyan, and K. B. Døving, “Is
Feeding Behaviour in Crucian Carp Mediated by the
Lateral Olfactory Tract?,” Chem. Senses 26, 1133–1138
(2001b).
122. E. H. Hamdani, O. B. Stabell, G. Alexander, and
K. B. Døving, “Alarm Reaction in the Crucian Carp is
Mediated by the Medial Bundle of the Medial Olfactory
Tract,” Chem. Senses 25, 103–109 (2000).
123. L. P. Hansen, K. B. Døving, and B. Jonsson, “Migration
of Farmed Adult Atlantic Salmon With and Without
Olfactory Sense Released on the Norwegian Coast,” J.
Fish Biol. 30, 713–721 (1987).
124. A. Hansen and T. E. Finger, “Phyletic Distribution of
Crypt-Type Olfactory Receptor Neurons in Fishes,”
Brain Behav. Evol. 55, 100–110 (2000).
125. A. Hansen, S. H. Rolen, K. Anderson et al., “Correla-
tion Between Olfactory Receptor Cell T