100
J.H. Choat
surgeonfi shes. There are also differences in reproductive patterns. Wrasses are
predominantly protogynous hermaphrodites (Sadovy de Mitcheson and Lui 2008 ) ,
while surgeonfi shes are gonochorists with a distinctive gonad morphology
(Montgomery and Galzin 1993 ; Montgomery et al. 1999 ) . A similar pattern is seen
in transient spawners. Groupers display complex patterns of sexual ontogeny with
protogyny being the dominant mode; snappers appear to be gonochorists as there
are no records of protogyny or protandry in this well-studied group (Sadovy de
Mitcheson and Liu 2008 ; Erisman et al. 2009 ) .
To identify ecological and life history features that might predict spawning mode,
representatives of these families were examined by ordination analysis (Fig. 4.3 ).
The fi rst analysis considered maximum size and morphological variables (body cross
sectional ratio and gape). Cross-sectional ratio is a proxy for internal body dimensions, habitat association and foraging mode (Fig. 4.3a ). Gape was estimated as the
ratio of the functional length of the upper jaw to head length. The use of external
Fig. 4.3 ( a ) Ordination of three morphological variables maximum length (Lmax), Body cross
sectional ratio (BW/Bd) and Functional Gape (HL/M) using Principal Component Analysis ( PCA )
showing the direction (eigenvectors) and contribution (relative length of eigenvectors) of the three
variables. Bubble size represents gradient from lowest to highest values of maximum size. Twenty
eight taxa with records of aggregate spawning from four perciform families were included in the
analysis. Acanthuridae Acanthurus lineatus Alin, A.nigrofuscus Anig, A.blochii Ablo, A.bahianus
Abah, Ctenochaetus striatus Cstr, Zebrasoma fl avescens Zfl a. Labridae (Scarines) Chlorurus
microrhinos Cmic, C.sordidus Csor, Scarus psittacus Spsi, S.rivulatus Sriv, Sparisoma viride Svir.
Lutjanidae Lutjanus apodus Lapo, L.adetti Lade, L.argentimaculatus Larg, L.bohar Lboh,
L.campechanus Lcam, L.griseus Lgri, L.synagris Lsyn, L.fulvus Lful. Serranidae Plectropomus
leopardus Pleo, P.laevi s Plae, P.areolatus Pare, Epinephelus fuscoguttatus Efus, E.polyphekadion
Epol, E.striatus Estr, E.adscensionus Eads, Mycteroperca bonaci Mbon, M.phenax Mphe. Body
cross-sectional ratios were estimated using the negative relationship between body depth and body
cross-sectional ratios (Fulton 2005 , Fig. 2.7). Fulton ( 2005 ) estimated body cross sectional ratios
as body width/body depth so that species with high body planes and lateral compression, for example acanthurids, generated low values for cross sectional ratios (0.15–0.25). Species with fusiform
bodies, for example serranids, generated higher values (0.4–0.5). HL is distance from snout tip to
posterior margin of operculum; M length of the maxillary margin of jaw. Morphometric data
extracted from digital images using Image J software. ( b ) Ordination of three demographic variables Maximum age (Tmax), Proportion of life span remaining when 50% of Linf is achieved
(50% Linf), Proportion of primary males (%1° males). Principal Component Analysis (PCA)
showing the direction (eigenvectors) and contribution (relative length of eigenvectors) of the three
variables. Bubble size represents gradient from lowest to highest values of maximum age. Twenty
eight taxa as in Fig. 4.2a . The metric used to compare growth profi les among the groups was an
estimate of the proportion of the life span remaining after a specifi ed size/developmental stage was
reached in the different species. Although the size/age at which sexual maturity occurs would be
the most appropriate reference point from which to estimate the remaining proportion of the life
span this information was not generally available from the literature. As Von Bertalanffy Growth
Functions were available for the relevant species a proxy value of the proportion of the life span
remaining after 50% of Linf was achieved was used as an alternative.
J.H. Choat
surgeonfi shes. There are also differences in reproductive patterns. Wrasses are
predominantly protogynous hermaphrodites (Sadovy de Mitcheson and Lui 2008 ) ,
while surgeonfi shes are gonochorists with a distinctive gonad morphology
(Montgomery and Galzin 1993 ; Montgomery et al. 1999 ) . A similar pattern is seen
in transient spawners. Groupers display complex patterns of sexual ontogeny with
protogyny being the dominant mode; snappers appear to be gonochorists as there
are no records of protogyny or protandry in this well-studied group (Sadovy de
Mitcheson and Liu 2008 ; Erisman et al. 2009 ) .
To identify ecological and life history features that might predict spawning mode,
representatives of these families were examined by ordination analysis (Fig. 4.3 ).
The fi rst analysis considered maximum size and morphological variables (body cross
sectional ratio and gape). Cross-sectional ratio is a proxy for internal body dimensions, habitat association and foraging mode (Fig. 4.3a ). Gape was estimated as the
ratio of the functional length of the upper jaw to head length. The use of external
Fig. 4.3 ( a ) Ordination of three morphological variables maximum length (Lmax), Body cross
sectional ratio (BW/Bd) and Functional Gape (HL/M) using Principal Component Analysis ( PCA )
showing the direction (eigenvectors) and contribution (relative length of eigenvectors) of the three
variables. Bubble size represents gradient from lowest to highest values of maximum size. Twenty
eight taxa with records of aggregate spawning from four perciform families were included in the
analysis. Acanthuridae Acanthurus lineatus Alin, A.nigrofuscus Anig, A.blochii Ablo, A.bahianus
Abah, Ctenochaetus striatus Cstr, Zebrasoma fl avescens Zfl a. Labridae (Scarines) Chlorurus
microrhinos Cmic, C.sordidus Csor, Scarus psittacus Spsi, S.rivulatus Sriv, Sparisoma viride Svir.
Lutjanidae Lutjanus apodus Lapo, L.adetti Lade, L.argentimaculatus Larg, L.bohar Lboh,
L.campechanus Lcam, L.griseus Lgri, L.synagris Lsyn, L.fulvus Lful. Serranidae Plectropomus
leopardus Pleo, P.laevi s Plae, P.areolatus Pare, Epinephelus fuscoguttatus Efus, E.polyphekadion
Epol, E.striatus Estr, E.adscensionus Eads, Mycteroperca bonaci Mbon, M.phenax Mphe. Body
cross-sectional ratios were estimated using the negative relationship between body depth and body
cross-sectional ratios (Fulton 2005 , Fig. 2.7). Fulton ( 2005 ) estimated body cross sectional ratios
as body width/body depth so that species with high body planes and lateral compression, for example acanthurids, generated low values for cross sectional ratios (0.15–0.25). Species with fusiform
bodies, for example serranids, generated higher values (0.4–0.5). HL is distance from snout tip to
posterior margin of operculum; M length of the maxillary margin of jaw. Morphometric data
extracted from digital images using Image J software. ( b ) Ordination of three demographic variables Maximum age (Tmax), Proportion of life span remaining when 50% of Linf is achieved
(50% Linf), Proportion of primary males (%1° males). Principal Component Analysis (PCA)
showing the direction (eigenvectors) and contribution (relative length of eigenvectors) of the three
variables. Bubble size represents gradient from lowest to highest values of maximum age. Twenty
eight taxa as in Fig. 4.2a . The metric used to compare growth profi les among the groups was an
estimate of the proportion of the life span remaining after a specifi ed size/developmental stage was
reached in the different species. Although the size/age at which sexual maturity occurs would be
the most appropriate reference point from which to estimate the remaining proportion of the life
span this information was not generally available from the literature. As Von Bertalanffy Growth
Functions were available for the relevant species a proxy value of the proportion of the life span
remaining after 50% of Linf was achieved was used as an alternative.