101
Vehrencamp 1998; Davies et al. 2012 ). An exception to this association between
egg-clustering and sound production occurs in the Boleosoma clade, where eggclustering species were found to lack sound production ( Etheostoma olmstedi and
Etheostoma nigrum , CEJ unpublished data). Another distinguishing characteristic
of the soniferous Catonotus is that they are virtually the only group lacking bright
male breeding coloration within Etheostoma (Page 1985 ). Additionally, the silent
barcheek group within Catonotus lacks sound production but has bright breeding
coloration (Page 1985 ).
Since all soniferous darters documented to date produce all three sound types, a
comparative analysis on the evolutionary origin of particular sound types within
Catonotus is not possible. However, one notable phylogenetic association warranting further investigation is the divergent signals of Etheostoma fl abellare , which are
in higher dominant frequencies and have a more complex pulse repetition rate than
the soniferous darters belonging to the more phylogenetically distant Etheostoma
squamiceps clade (Fig. 1a ; Speares et al. 2011 ) (see Page et al. 2003 for a recent
phylogeny).
2.2 Sculpins
The sound descriptions for marine sculpins are mostly anecdotal. Myoxocephalus
“growls” are sustained, harmonic sounds of variable duration with a fundamental
frequency near 60 Hz (Fig. 1c ; Fish and Mowbray 1970 ) (Table 1 ). Fish and
Mowbray ( 1970 ) collected Myoxocephalus aenaeus and M. octodecemspinosus
sounds passively from several adults together in an aquarium, and M. octodecemspinosus also produced sounds during feeding and handling. The neurophysiology of
sound production in Myoxocephalus scorpius and Leptocottus armatus has been
investigated without description of behaviours or acoustic parameters (Bass and
Baker 1991 ).
All Cottus species produce pulsed “knocks” (~40 ms pulse duration, peak frequencies: 50–300 Hz), some species produce structured “knock trains”, and at least
one species produces a tonal “moan” ( Cottus carolinae , Fig. 1d ). The acoustic
Fig. 1 (continued) ( c ) An uninterrupted sequence of Myoxocephalus octodemospinosus “grunts”
(from the CD archive companion to Fish and Mowbray 1970 ). ( d ) An uninterrupted sequence of
Cottus carolinae distinct pulsatile knock and tonal “moan” sounds. ( e ) An uninterrupted sequence
of Padogobius bonelli breeding sounds recorded in stream Stirone and emitted by a male (with
eggs) from the nest hollow (under a stone) toward a caged ripe female placed in front of the nest.
The P. bonelli sounds grade from purely tonal (1,7,8) to purely grunt-like (5,6), three of them
(2,3,4) being a mix of the two types. This sequence shows how modulation of the pulse rate can
change the spectral representation of the sound on the spectrogram. ( f ) An uninterrupted sequence
of the mudskipper Periophthalmodon septemradiatus , indicating pulsatile, tonal, and complex
sound types. All spectrograms were computed with the following parameters: window length: 0.05
s, maximum frequency: 1500 Hz, time step: 0.002 s, frequency step: 20 Hz, window: Hamming
Convergent Aspects of Acoustic Communication in Darters, Sculpins, and Gobies
Vehrencamp 1998; Davies et al. 2012 ). An exception to this association between
egg-clustering and sound production occurs in the Boleosoma clade, where eggclustering species were found to lack sound production ( Etheostoma olmstedi and
Etheostoma nigrum , CEJ unpublished data). Another distinguishing characteristic
of the soniferous Catonotus is that they are virtually the only group lacking bright
male breeding coloration within Etheostoma (Page 1985 ). Additionally, the silent
barcheek group within Catonotus lacks sound production but has bright breeding
coloration (Page 1985 ).
Since all soniferous darters documented to date produce all three sound types, a
comparative analysis on the evolutionary origin of particular sound types within
Catonotus is not possible. However, one notable phylogenetic association warranting further investigation is the divergent signals of Etheostoma fl abellare , which are
in higher dominant frequencies and have a more complex pulse repetition rate than
the soniferous darters belonging to the more phylogenetically distant Etheostoma
squamiceps clade (Fig. 1a ; Speares et al. 2011 ) (see Page et al. 2003 for a recent
phylogeny).
2.2 Sculpins
The sound descriptions for marine sculpins are mostly anecdotal. Myoxocephalus
“growls” are sustained, harmonic sounds of variable duration with a fundamental
frequency near 60 Hz (Fig. 1c ; Fish and Mowbray 1970 ) (Table 1 ). Fish and
Mowbray ( 1970 ) collected Myoxocephalus aenaeus and M. octodecemspinosus
sounds passively from several adults together in an aquarium, and M. octodecemspinosus also produced sounds during feeding and handling. The neurophysiology of
sound production in Myoxocephalus scorpius and Leptocottus armatus has been
investigated without description of behaviours or acoustic parameters (Bass and
Baker 1991 ).
All Cottus species produce pulsed “knocks” (~40 ms pulse duration, peak frequencies: 50–300 Hz), some species produce structured “knock trains”, and at least
one species produces a tonal “moan” ( Cottus carolinae , Fig. 1d ). The acoustic
Fig. 1 (continued) ( c ) An uninterrupted sequence of Myoxocephalus octodemospinosus “grunts”
(from the CD archive companion to Fish and Mowbray 1970 ). ( d ) An uninterrupted sequence of
Cottus carolinae distinct pulsatile knock and tonal “moan” sounds. ( e ) An uninterrupted sequence
of Padogobius bonelli breeding sounds recorded in stream Stirone and emitted by a male (with
eggs) from the nest hollow (under a stone) toward a caged ripe female placed in front of the nest.
The P. bonelli sounds grade from purely tonal (1,7,8) to purely grunt-like (5,6), three of them
(2,3,4) being a mix of the two types. This sequence shows how modulation of the pulse rate can
change the spectral representation of the sound on the spectrogram. ( f ) An uninterrupted sequence
of the mudskipper Periophthalmodon septemradiatus , indicating pulsatile, tonal, and complex
sound types. All spectrograms were computed with the following parameters: window length: 0.05
s, maximum frequency: 1500 Hz, time step: 0.002 s, frequency step: 20 Hz, window: Hamming
Convergent Aspects of Acoustic Communication in Darters, Sculpins, and Gobies