324
hearing in mormyrids (Stipetić 1939 ; Yan and Curtsinger 2000 ). The non-related
perciform suborder Anabantoidei (labyrinth fi shes) possesses a suprabranchial cavity (organ) dorsally of the gills for air-breathing. This labyrinth is anatomically in
direct contact with the saccule and enhances hearing, as has been shown experimentally (Schneider 1941 ; Yan 1998 ).
In fi shes lacking a direct connection between a gas bladder and inner ear, it is
assumed that the bladder has no auditory function and thus does not serve as a
peripheral structure for hearing (Fig. 1d ). This, however, cannot be concluded based
solely on anatomical considerations. In all species without a connection, the auditory
function needs to be assessed experimentally in order to be certain that this is not the
case (Popper and Fay 2011 ). It is experimentally diffi cult to prove that a fi sh with a
swim bladder does not respond to sound pressure changes in a sound fi eld. This can
either be demonstrated by decoupling sound pressure from particle motion in an
experimental tank (such as in standing wave tubes) or by eliminating the gas bladder
(e.g. removing the gas and/or fi lling it with fl uids) and then recording whether this
procedure decreases hearing sensitivities. Elimination experiments have shown that
swim bladders not connected to inner ears have no auditory function in toadfi sh and
labyrinth fi sh (Yan et al. 2000 ) but that they do have such a function in cods and
damselfi sh (Sand and Enger 1973 ; Myrberg and Spires 1980 ). In the latter species the
assumption is that the tissue between the bladder and the inner ear transmits bladder
vibrations and enables detection of sound pressure changes.
Finally, based on the above, peripheral auditory structures can be ruled out only
in species that completely lack any gas-fi lled cavities (swim bladders) such as fl atfi shes or sculpins (cottids) (Fig. 1e ). In all other species lacking a connection, it
needs to be shown experimentally that gas bladders play no role in hearing.
Accessory hearing structures may also differ in size within species either due to
ontogenetic development (see review Ladich 2015 ) or experimental manipulation
(Sand and Enger 1973 ).
The aim of the review is to go beyond our current knowledge, which is mainly
based on the observation that shorter distances between the bladder and inner ears
improve hearing. Comparative morphological and physiological studies show that
the size of the accessory gas bladders and the structure of the connection to the inner
ears affect hearing in fi shes as well. This is analyzed based on recent inter- and
intrafamilial fi ndings in catfi shes and cichlids.
2 Interfamilial Comparison in Catfi shes
All otophysines possess swim bladders and Weberian ossicles. Neither bladders nor
ossicles have been completely lost in any of the more than 8000 species known.
Otophysines inhabit freshwaters (except for 2 out of 36 families of catfi shes), comprising about two-thirds of all freshwater fi sh species. Peripheral structures for hearing improvement vary considerably among otophysines (Chranilov 1927 , 1929 ;
Alexander 1962 , 1964 ; Chardon 1968 ). This is particularly the case in the orders
Siluriformes and Cyriniformes.
F. Ladich
hearing in mormyrids (Stipetić 1939 ; Yan and Curtsinger 2000 ). The non-related
perciform suborder Anabantoidei (labyrinth fi shes) possesses a suprabranchial cavity (organ) dorsally of the gills for air-breathing. This labyrinth is anatomically in
direct contact with the saccule and enhances hearing, as has been shown experimentally (Schneider 1941 ; Yan 1998 ).
In fi shes lacking a direct connection between a gas bladder and inner ear, it is
assumed that the bladder has no auditory function and thus does not serve as a
peripheral structure for hearing (Fig. 1d ). This, however, cannot be concluded based
solely on anatomical considerations. In all species without a connection, the auditory
function needs to be assessed experimentally in order to be certain that this is not the
case (Popper and Fay 2011 ). It is experimentally diffi cult to prove that a fi sh with a
swim bladder does not respond to sound pressure changes in a sound fi eld. This can
either be demonstrated by decoupling sound pressure from particle motion in an
experimental tank (such as in standing wave tubes) or by eliminating the gas bladder
(e.g. removing the gas and/or fi lling it with fl uids) and then recording whether this
procedure decreases hearing sensitivities. Elimination experiments have shown that
swim bladders not connected to inner ears have no auditory function in toadfi sh and
labyrinth fi sh (Yan et al. 2000 ) but that they do have such a function in cods and
damselfi sh (Sand and Enger 1973 ; Myrberg and Spires 1980 ). In the latter species the
assumption is that the tissue between the bladder and the inner ear transmits bladder
vibrations and enables detection of sound pressure changes.
Finally, based on the above, peripheral auditory structures can be ruled out only
in species that completely lack any gas-fi lled cavities (swim bladders) such as fl atfi shes or sculpins (cottids) (Fig. 1e ). In all other species lacking a connection, it
needs to be shown experimentally that gas bladders play no role in hearing.
Accessory hearing structures may also differ in size within species either due to
ontogenetic development (see review Ladich 2015 ) or experimental manipulation
(Sand and Enger 1973 ).
The aim of the review is to go beyond our current knowledge, which is mainly
based on the observation that shorter distances between the bladder and inner ears
improve hearing. Comparative morphological and physiological studies show that
the size of the accessory gas bladders and the structure of the connection to the inner
ears affect hearing in fi shes as well. This is analyzed based on recent inter- and
intrafamilial fi ndings in catfi shes and cichlids.
2 Interfamilial Comparison in Catfi shes
All otophysines possess swim bladders and Weberian ossicles. Neither bladders nor
ossicles have been completely lost in any of the more than 8000 species known.
Otophysines inhabit freshwaters (except for 2 out of 36 families of catfi shes), comprising about two-thirds of all freshwater fi sh species. Peripheral structures for hearing improvement vary considerably among otophysines (Chranilov 1927 , 1929 ;
Alexander 1962 , 1964 ; Chardon 1968 ). This is particularly the case in the orders
Siluriformes and Cyriniformes.
F. Ladich