Abstract

The timing, sources and pathways for incorporation of larval fishes into a developing anti-cyclonic eddy off south-western Australia were investigated. Larval fish assemblages within the study region were structured by location (shelf, eddy and oceanic) and water mass. The larval fish assemblage within the eddy was significantly distinct from that characterizing the surrounding oceanic water. The eddy assemblage, which was comprised largely of larvae of oceanic meso-pelagic fishes (especially Diaphus spp. and Vinciguerria spp.) and less abundant neritic taxa, reflected its Leeuwin Current (LC), shelf and oceanic source waters. The occurrence of neritic taxa in the eddy confirmed the hypothesis that these larvae were incorporated as it developed in proximity to the shelf break. The significantly larger larval size of temperate neritic taxa (e.g. Sardinops sagax, Engraulis australis) in the eddy compared with the shelf suggests that these larvae were transported from the shelf adjacent to the developing eddy. The occurrence of tropical neritic taxa (e.g. Acanthuridae, Lutjanidae, Pomacentridae) highlighted the LC as an important transport route to higher latitudes. Coupling the sampling of larval fishes with the trajectories of Lagrangian drifters provided insight into how larval fish assemblages changed during development of the eddy.

INTRODUCTION

The influence of oceanographic processes upon larval distributions and transport is fundamental to recruitment dynamics and variability in marine populations (Nakata et al., 2000; Hare et al., 2002; Werner and Quinlan, 2002; Govoni, 2005). Eddies are an important mechanism driving shelf-slope exchange of water and constituent biomass as they often propagate in proximity of the shelf break. Remote observations of sea surface temperature and ocean colour have described the apparent incorporation of chlorophyll-rich shelf water into eddies (e.g. Flierl and Wroblewski, 1985; Myers and Drinkwater, 1989; Paterson et al., 2008). Field investigations of mature eddies have revealed the occurrence of elevated levels of coastal-origin nutrients and of modified shelf phytoplankton and zooplankton communities (e.g. Mackas and Galbraith, 2002; Mackas and Coyle, 2005; Waite et al., 2007a,b). In certain situations, fish populations are reliant upon eddy-induced exchange for the transport of larvae from spawning to nursery areas (e.g. Graber and Limouzy-Paris, 1997; Nakata et al., 2000; Hare et al., 2002; Sponagule et al., 2005). Conversely, a modelling study for the Gulf Stream found a loss of between 20 and 50% of neritic larvae resulted from eddy-induced offshore advection, which significantly affected year-class strength (Flierl and Wroblewski, 1985). In general, eddies, such as those of the Florida Current, are known to induce significant recruitment variability for fish and invertebrate populations (Graber and Limouzy-Paris, 1997).

Off Western Australia (WA) in the south-eastern Indian Ocean, the atypical eastern boundary Leeuwin Current (LC) attains its maximum poleward geostrophic flow during the austral autumn-winter (May–June) and there is distinct inter-annual variability associated with El Niño Southern Oscillation events (Cresswell and Golding, 1980; Cresswell, 1991; Pearce, 1991; Feng et al., 2003). Coincident with peak velocity, the LC is highly energetic and characterized by numerous perturbations in the form of meso-scale meanders, filaments and eddies. The LC is reported to have the highest eddy kinetic energy among all the mid-latitude eastern boundary currents (Feng et al., 2005).

Cross-shelf distributions of larval fish assemblages off south-WA display strong seasonal and inter-annual variation, which is influenced by the strength and trajectory of the LC (Muhling et al., 2008). The greatest variability in larval fish assemblages coincides with periods of meso-scale eddy formation during the austral autumn and winter (Muhling et al., 2008). With respect to marine teleosts and fisheries, Gaughan (Gaughan, 2007) suggested that the LC and its eddies could have a direct negative impact upon neritic fish populations due to deleterious offshore transport of larvae.

To date, most studies have described entrainment of shelf-origin biomass to the perimeter of mature eddies (e.g. Lobel and Robinson, 1986; Hare et al., 2002; Kasai et al., 2002; Ladd et al., 2005). There are few studies that have examined the timing of, and pathways for, the incorporation of biomass to eddies, specifically during eddy formation, although Atwood et al. (Atwood et al., 2010) studied larval fish assemblages during the formative stages of eddies in the Gulf of Alaska. They showed that eddy assemblages differed from those in adjacent shelf waters but were dynamic, changing as the eddies aged following detachment and seaward propagation. Off WA, larval fish assemblages of mature LC eddies located ∼500 km seaward from the coast were dominated by the larvae of meso-pelagic fishes and no neritic larvae were recorded (Muhling et al., 2007). The fact that a modified phytoplankton community of coastal origin was associated with the mature anti-cyclonic eddy (Waite et al., 2007b) suggested that larvae of neritic teleosts might also have been incorporated during eddy formation at the shelf break but were no longer present in the plankton (Muhling et al., 2007).

A detailed in situ examination of a developing anti-cyclonic LC eddy was conducted off south-WA in May 2006 (Paterson et al., 2008). In addition to physical oceanographic measurements, nutrients, primary production and zooplankton were examined in, and around, the eddy field. The water mass in the eddy was found to be largely modified LC water but there was no certainty regarding the timing and amount of coastally derived material that was incorporated into the eddy (Paterson et al., 2008). This study examines the distribution of larval fish assemblages in and around this developing anti-cyclonic eddy to identify the timing, sources and pathways for their incorporation into the eddy. In particular, because of their known source and planktonic duration, the larvae of neritic fishes were investigated as time-dependent tracers of the water incorporated into the developing eddy.

METHOD

Sampling protocol

Intensive oceanographic (Paterson et al., 2008) and biological sampling of a developing anti-cyclonic eddy and surrounding waters between 30 and 34°S and out to ∼112°E (350 km offshore) off the south-western coast of WA was conducted during the period 2–28 May 2006 (Fig. 1A and B). The locations of sampling stations where larval fishes were collected (Fig. 1B) were primarily determined by the trajectories of satellite-tracked, drifting buoys which were deployed during the voyage (Waite et al., unpublished data). The use of near-real time satellite data allowed for the strategic location of additional stations to sample the eddy (centre and perimeter), adjacent continental shelf, LC and surrounding oceanic waters (Fig. 1A). Plankton tows at each sampling station were concurrent with oceanographic measurements.

Fig. 1.

(A) MODIS Aqua SST for 11 May 2006, coinciding with the research voyage, showing the warm LC flowing southward along the shelf break and a developing anti-cyclonic eddy off south-WA. The distribution of mixed layer water mass (Paterson et al., 2008) is also indicated. (B) Position of sampling stations where larval fishes were collected from shelf and oceanic waters, including at the perimeter and centre of the developing anti-cyclonic eddy. The trajectories of three Lagrangian drifters are shown; the first (open triangle) was deployed in the LC meander at station 3 and retrieved/re-deployed at stations 5 and 7. A second drifter (open square) was deployed within the eddy perimeter at station 4, and was retrieved/re-deployed at stations 6 and 10. A third drifter (open diamond) was deployed at station 14 at the eddy centre and retrieved at station 17. The large arrow indicates the propagation trajectory of the eddy as it developed over the duration of the voyage and dashed ellipses indicate the approximate positions of the eddy roughly coinciding with 10 and 28 May, 2006.

Fig. 1.

(A) MODIS Aqua SST for 11 May 2006, coinciding with the research voyage, showing the warm LC flowing southward along the shelf break and a developing anti-cyclonic eddy off south-WA. The distribution of mixed layer water mass (Paterson et al., 2008) is also indicated. (B) Position of sampling stations where larval fishes were collected from shelf and oceanic waters, including at the perimeter and centre of the developing anti-cyclonic eddy. The trajectories of three Lagrangian drifters are shown; the first (open triangle) was deployed in the LC meander at station 3 and retrieved/re-deployed at stations 5 and 7. A second drifter (open square) was deployed within the eddy perimeter at station 4, and was retrieved/re-deployed at stations 6 and 10. A third drifter (open diamond) was deployed at station 14 at the eddy centre and retrieved at station 17. The large arrow indicates the propagation trajectory of the eddy as it developed over the duration of the voyage and dashed ellipses indicate the approximate positions of the eddy roughly coinciding with 10 and 28 May, 2006.

Circulation of the LC meander/eddy perimeter and anti-cyclonic eddy was examined using Iridium satellite-tracked, surface-drifting buoys (Fastwave Communications Pty Ltd). Drifters were drogued at 245 m in order to reflect circulation within the upper mixed layer and decrease the influence of wind-driven surface transport (Nodder and Waite, 2001). The first drifter was deployed on 5 May 2006 near station 3 within the LC, north and slightly upstream of the meander (31°0.072′S, 114°39.97′E). The second drifter was deployed on 7 May near station 4 (32°50.46′S, 113°50.28′E) to define the rotational circulation within the eddy perimeter. A third drifter was deployed on 20 May near station 14 at the approximate position of the eddy centre (32°48.708′S, 113°32.60′E). The drifters were retrieved and re-deployed at 3-day intervals and full oceanographic and biological sampling was undertaken at these stations.

Vertical conductivity–temperature–depth–oxygen (CTD-O2) measurements were performed at each sampling station using a Seabird SBE 19+ instrument which was also equipped with a Chelsea TGI fluorometer. Casts were to 1000 m depth, or to ∼10 m above the bottom in shallower water. The mixed layer depth was defined where a change in potential density of 0.125 kg m−3 from 10 m depth occurred (Levitus, 1982; Feng et al., 2007).

Replicated depth-integrated bongo net samples (100 µm and 355 µm meshes, mouth area 0.196 m2, diameter 0.6 m) were obtained at night by towing the nets obliquely to the surface from a maximum depth of 150 m (or from 10 m off the bottom in shallow water) at a ship's speed of 2 knots for 15 min. A mechanical General Oceanics flowmeter was positioned centrally in each of the nets and they were linked electronically to an interface from which the volume of seawater filtered, depth of the nets and tow profile could be monitored. The samples from all stations were preserved in 5% formalin and data from the 355 µm net are presented here. The counts of larval fishes were standardized to number of larvae under 1 m2. Larval fishes were identified using relevant literature (Moser, 1984; Olivar and Fortuño, 1991; Neira et al., 1998; Olivar et al., 1999; Leis and Carson-Ewart, 2000; Moser and Watson, 2005; Richards, 2006).

In addition to the depth-integrated sampling of the water column, surface water was also sampled using a large 1 m2 net equipped with 1 mm mesh in order to collect late-stage larvae which may have avoided the smaller bongo net. Length frequency analyses of the larvae of three epipelagic, neritic species selected as tracers were undertaken to establish if there were significant differences between the size of larvae on the shelf and within the eddy.

Larval fish data analyses

Differences in the mean abundances (number m−2) of larval fishes among all sampling stations were examined using a one-way ANOVA in SPSS 14.0 upon square-root transformed data after using Levene's test for homogeneity of variance. Tukey's post hoc test was applied for pairwise comparisons to determine where significant differences existed.

Larval fish assemblage structure was examined using the Primer-6 software package (Clarke and Warwick, 2005). Prior to analysis, larval fish concentrations were log(x+1) transformed to reduce the weighting of dominant taxa and a Bray–Curtis resemblance matrix was constructed. Larval fish assemblages were grouped a priori according to the factors of location (shelf, LC meander/eddy perimeter, eddy centre, oceanic) and mixed layer water mass which was classified as LC, modified LC and oceanic subtropical surface water (STSW) according to Paterson et al. (Paterson et al., 2008).

Multidimensional scaling ordinations were used to display the similarities between larval fish assemblages based on the factors of location and water mass. One-way ANOSIM was applied to test the null hypotheses (Ho) of no difference in larval fish assemblages among locations or water masses. The SIMPER routine was applied to identify those taxa which were characteristic of each larval fish assemblage and those most responsible for differences among assemblages.

Larval fish assemblages were correlated with all available environmental variables (temperature, salinity, fluorescence, dissolved oxygen, turbidity, depth and surface chlorophyll a concentration) using the BVSTEP sub-routine. This is a step-wise permutation technique to identify the subset of environmental variables which best matches the reference (larval fish) matrix (Clarke and Gorley, 2006). Each environmental variable, excluding distance from shore, was expressed as a mean of the mixed layer for each sampling station. Prior to the analysis, correlations between all variables were assessed from draftsman plots and each variable was subjected to the appropriate transformation. The variables of salinity and dissolved oxygen were omitted from the analysis since they were strongly correlated with temperature (positively and negatively, respectively); the inclusion of temperature served as a surrogate for these two variables. The concentration of chlorophyll a had a strongly skewed relationship with distance from shore and was subjected to a log(x+1) transformation. Data were normalized and a resemblance matrix constructed based on Euclidean distance.

RESULTS

Oceanography

During May 2006, a meso-scale anti-cyclonic eddy developed from a meander of the LC seaward of the shelf break off south-WA (Fig. 1A). Synoptic data (SST and ocean colour) indicated that the developing eddy had a strong connection with the LC and shelf waters near 32°S, and that exchanges of varying magnitude occurred between the eddy and surrounding water masses which modified the physical, chemical and biological signature of the eddy (Paterson et al., 2008).

Three water masses were identified in the study area, namely, LC, modified LC and STSW. LC water (>23.0°C, < 35.4) was generally located upstream of the eddy (Figs 1 and 2). The majority of stations encompassing shelf, LC meander/eddy perimeter and eddy centre locations were characterized by temperature and salinity values ranging between 21.5–22.5°C and 34.4–35.7, respectively; this TS signature was classified as modified LC water (Paterson et al., 2008). Oceanic stations located outside of the eddy were characterized by cool (18.0–20.5°C), high salinity (35.8–35.9) STSW.

Fig. 2.

Mixed layer TS and water mass classification (LC, modified LC, STSW) determined for all stations representing shelf (x), eddy perimeter (open circles), eddy centre (black-shaded circles) and oceanic (grey-shaded circles) locations where larval fishes were collected off south-WA in the late austral autumn, May 2006.

Fig. 2.

Mixed layer TS and water mass classification (LC, modified LC, STSW) determined for all stations representing shelf (x), eddy perimeter (open circles), eddy centre (black-shaded circles) and oceanic (grey-shaded circles) locations where larval fishes were collected off south-WA in the late austral autumn, May 2006.

Physical properties of the water column for each sampling station showed considerable variability throughout the eddy field (Table I). Stations representative of each of the four locations in the eddy field are presented to demonstrate the water column structure (Fig. 3). Shelf and eddy centre stations showed more homogenous vertical profiles in both salinity and temperature compared with the western perimeter of the meander and oceanic stations which had LC and STSW characteristics, respectively. The mixed layer at the centre of the eddy deepened with time and was consistent with strengthening of the eddy over the period of the voyage.

Table I:

Mixed layer TS properties and water mass classification for stations where larval fishes were collected off south-WA, May 2006

Station Depth of water (m) Location Mixed layer depth (m) Mean mixed layer temperature (°C) Mean mixed layer salinity Mixed layer water mass 
935 Perimeter 150 21.93 35.61 Modified LC 
945 LC upstream 31 23.73 35.32 LC 
3a 1197 LC upstream 38 23.94 35.32 LC 
4b 2817 Perimeter 43 22.38 35.45 Modified LC 
5a 4641 Perimeter 28 23.65 35.36 Modified LC 
6b 4660 Perimeter 72 22.11 35.49 Modified LC 
7a 855 Perimeter 46 22.72 35.42 Modified LC 
104 Shelf 67 22.13 35.50 Modified LC 
280 Shelf 77 22.47 35.46 Modified LC 
107 Shelf 56 23.15 35.34 LC 
109 Shelf 50 23.19 35.34 LC 
10b 1405 Perimeter 73 22.51 35.48 Modified LC 
11 104 Shelf 21 22.70 35.40 Modified LC 
12 4201 Perimeter 70 22.09 35.49 Modified LC 
13 4028 Oceanic 40 20.21 35.88 STSW 
2931 Eddy centre 59 22.35 35.45 Modified LC 
3296 Eddy centre 85 22.44 35.44 Modified LC 
3652 Eddy centre 77 22.39 35.45 Modified LC 
14c 3324 Eddy centre 94 22.51 35.45 Modified LC 
15 3064 Oceanic 93 18.93 35.87 STSW 
16 4493 Oceanic 54 20.40 35.79 STSW 
17c 4076 Eddy centre 119 22.25 35.47 Modified LC 
1099 Perimeter 49 22.57 35.42 Modified LC 
Station Depth of water (m) Location Mixed layer depth (m) Mean mixed layer temperature (°C) Mean mixed layer salinity Mixed layer water mass 
935 Perimeter 150 21.93 35.61 Modified LC 
945 LC upstream 31 23.73 35.32 LC 
3a 1197 LC upstream 38 23.94 35.32 LC 
4b 2817 Perimeter 43 22.38 35.45 Modified LC 
5a 4641 Perimeter 28 23.65 35.36 Modified LC 
6b 4660 Perimeter 72 22.11 35.49 Modified LC 
7a 855 Perimeter 46 22.72 35.42 Modified LC 
104 Shelf 67 22.13 35.50 Modified LC 
280 Shelf 77 22.47 35.46 Modified LC 
107 Shelf 56 23.15 35.34 LC 
109 Shelf 50 23.19 35.34 LC 
10b 1405 Perimeter 73 22.51 35.48 Modified LC 
11 104 Shelf 21 22.70 35.40 Modified LC 
12 4201 Perimeter 70 22.09 35.49 Modified LC 
13 4028 Oceanic 40 20.21 35.88 STSW 
2931 Eddy centre 59 22.35 35.45 Modified LC 
3296 Eddy centre 85 22.44 35.44 Modified LC 
3652 Eddy centre 77 22.39 35.45 Modified LC 
14c 3324 Eddy centre 94 22.51 35.45 Modified LC 
15 3064 Oceanic 93 18.93 35.87 STSW 
16 4493 Oceanic 54 20.40 35.79 STSW 
17c 4076 Eddy centre 119 22.25 35.47 Modified LC 
1099 Perimeter 49 22.57 35.42 Modified LC 

Stations are listed in the order in which they were sampled during the voyage. Superscripts denote stations which corresponded with drifter deployment/retrievals; (a) for stations in the LC meander, (b) for stations located within the eddy perimeter and (c) for stations at the centre of the eddy.

Fig. 3.

Temperature (solid line) and salinity (dotted line) profiles representative of stations located over the shelf, at the western LC meander/eddy perimeter, eddy centre and oceanic waters off south-WA, May 2006.

Fig. 3.

Temperature (solid line) and salinity (dotted line) profiles representative of stations located over the shelf, at the western LC meander/eddy perimeter, eddy centre and oceanic waters off south-WA, May 2006.

Larval assemblages

Numbers of neritic and oceanic larval fishes in the depth-integrated bongo samples were highly variable within the eddy field (range: 14–639 larvae m−2), but were greatest in shelf waters (range: 96–639 larvae m−2; mean 392 larvae m−2) (Fig. 4). Overall, the mean abundance of larvae differed significantly between locations (P < 0.001). The mean abundance of larvae in shelf waters was significantly higher (P = 0.001) compared to eddy and oceanic stations (Fig. 5A). Significantly greater abundance of larval fishes occurred at the perimeter of, and within the eddy compared to the surrounding oceanic water (P = 0.001), but there was no significant difference between meander/perimeter and eddy centre locations (P = 0.8) (Fig. 5A).

Fig. 4.

Mean abundance of larval fishes (number m−2) per sampling station from depth-integrated bongo tows (355 µm mesh) at stations located on the shelf, offshore and in a developing anti-cyclonic eddy off south-WA, May 2006.

Fig. 4.

Mean abundance of larval fishes (number m−2) per sampling station from depth-integrated bongo tows (355 µm mesh) at stations located on the shelf, offshore and in a developing anti-cyclonic eddy off south-WA, May 2006.

Fig. 5.

(A) Mean abundance of larval fishes (number m−2) and (B) total number of teleost families per location showing the proportions of neritic and oceanic meso-pelagic taxa for the respective assemblages. Standard error bars are for the mean abundances of larval fishes per sampling location.

Fig. 5.

(A) Mean abundance of larval fishes (number m−2) and (B) total number of teleost families per location showing the proportions of neritic and oceanic meso-pelagic taxa for the respective assemblages. Standard error bars are for the mean abundances of larval fishes per sampling location.

From the bongo net samples, 7392 larval fishes comprising 57 genera (30 species) from 68 families were identified (Supplementary data, Online Appendix 1). There was a conspicuous difference between the number of families represented in shelf (49) and oceanic (9) stations largely due to the contribution by neritic families on the shelf (Fig. 5B). Eddy centre and meander/perimeter stations had markedly higher numbers of neritic families (23 and 27 families, respectively) than would have been expected relative to their geographic location seaward of the shelf break (Fig. 5B). All samples (shelf, meander/perimeter, eddy centre and oceanic) were dominated by larvae of the oceanic, meso-pelagic family Myctophidae, particularly Diaphus species (slender and deep morphotypes) (Fig. 6).

Fig. 6.

Proportional contribution by the dominant teleost families to total larval fish abundance for shelf, meander/eddy perimeter, eddy centre and oceanic assemblages off south-WA, May 2006.

Fig. 6.

Proportional contribution by the dominant teleost families to total larval fish abundance for shelf, meander/eddy perimeter, eddy centre and oceanic assemblages off south-WA, May 2006.

There were a high number of neritic fish taxa, albeit at low concentrations, within the LC meander and eddy perimeter and at the eddy centre compared to oceanic samples outside of the eddy. Larval Diaphus spp. (Myctophidae) were numerically dominant and ubiquitous across the eddy field (Supplementary data, Online Appendix 1) possibly related to the pervasive shoreward influence of the LC. Likewise, several oceanic Vinciguerria spp. (Phosichthyidae) were common across all locations (Supplementary data, Online Appendix 1). At shelf stations, neritic taxa such as Sardinops sagax and Bregmaceros spp. were abundant but also present at the eddy centre and perimeter indicating seaward transport (Fig. 6 and Supplementary data, Online Appendix 1). Despite the relatively high latitude of this study (31–34°S), several taxa were of tropical origin (e.g. Acanthuridae, Bothidae, Champsodontidae, Lutjanidae and Pomacentridae) indicative of southward transport of larvae by the LC.

Ordination produced a good representation of the structuring of larval fish assemblages grouped a priori according to the factor location (Fig. 7A). There was a spatial progression of assemblages extending from the shelf to oceanic waters and a clear separation between the oceanic assemblage and those associated with the eddy centre and meander/perimeter. Overall, larval fish assemblages were structured significantly in relation to location (R = 0.37, P = 0.001). Differences between the oceanic assemblage and those from all other locations were highly significant as indicated by the higher R-statistics in the ANOSIM analysis (Table II). The shelf assemblage differed significantly from those at the eddy perimeter and within the eddy, although these differences were weak as indicated by the low R-statistics (Table II). There was no significant difference between assemblages at the meander/perimeter and eddy centre.

Table II:

One-way ANOSIM and SIMPER analyses between larval fish assemblages classified according to the a priori factor of location

 Shelf LC meander/Eddy perimeter Eddy centre Oceanic 
Shelf Avg. Dissim. = 49%; Diaphus spp. (slender); Vinciguerria spp.; Bregmaceros spp.; M. asperum; V. nimbaria    
Eddy perimeter *R = 0.2, P = 0.001; Avg. Dissim. = 52%; Bregmaceros spp.(S); S. sagax(S) Avg. Dissim. = 46%; Diaphus spp. (slender); Vinciguerria spp.; V. nimbaria; L. alatus; Diaphus spp. (deep); H. hygomii; Bregmaceros spp.   
Eddy centre *R = 0.16, P = 0.007; Avg. Dissim. = 48%; S. sagax(S); Diaphus spp. (deep)(S) R = 0.15, P = 0. 05; Avg. Dissim. = 46%; Avg. Dissim. = 40%; Diaphus spp. (slender); Bregmaceros spp.; Vinciguerria spp.; V. nimbaria; Notosudidae spp.; Diaphus spp.(deep); M. asperum; H. hygomii  
Oceanic *R = 0.9, P = 0.001; Avg. Dissim. = 77%; Bregmaceros spp.(S); Diaphus spp. (deep)(S); Vinciguerria spp.(S); M. asperum(S); Diaphus spp. (slender)(S) *R = 0.8, P = 0.001; Avg. Dissim. = 71%; L. alatus(P); Vinciguerria spp.(P); Diaphus spp. (deep)(S); V. nimbaria(P); Diaphus spp.; (slender) (P) *R = 0.8, P = 0.001; Avg. Dissim. = 72%; Bregmaceros spp.(E); Notosudidae spp.(E); Diaphus spp. (slender)(E); Vinciguerria spp.(E); Diaphus spp. (deep)(E); V. nimbaria(E); L. alatus(E); M. asperum(E) Avg. Dissim. = 66%; Diaphus spp. (slender) 
 Shelf LC meander/Eddy perimeter Eddy centre Oceanic 
Shelf Avg. Dissim. = 49%; Diaphus spp. (slender); Vinciguerria spp.; Bregmaceros spp.; M. asperum; V. nimbaria    
Eddy perimeter *R = 0.2, P = 0.001; Avg. Dissim. = 52%; Bregmaceros spp.(S); S. sagax(S) Avg. Dissim. = 46%; Diaphus spp. (slender); Vinciguerria spp.; V. nimbaria; L. alatus; Diaphus spp. (deep); H. hygomii; Bregmaceros spp.   
Eddy centre *R = 0.16, P = 0.007; Avg. Dissim. = 48%; S. sagax(S); Diaphus spp. (deep)(S) R = 0.15, P = 0. 05; Avg. Dissim. = 46%; Avg. Dissim. = 40%; Diaphus spp. (slender); Bregmaceros spp.; Vinciguerria spp.; V. nimbaria; Notosudidae spp.; Diaphus spp.(deep); M. asperum; H. hygomii  
Oceanic *R = 0.9, P = 0.001; Avg. Dissim. = 77%; Bregmaceros spp.(S); Diaphus spp. (deep)(S); Vinciguerria spp.(S); M. asperum(S); Diaphus spp. (slender)(S) *R = 0.8, P = 0.001; Avg. Dissim. = 71%; L. alatus(P); Vinciguerria spp.(P); Diaphus spp. (deep)(S); V. nimbaria(P); Diaphus spp.; (slender) (P) *R = 0.8, P = 0.001; Avg. Dissim. = 72%; Bregmaceros spp.(E); Notosudidae spp.(E); Diaphus spp. (slender)(E); Vinciguerria spp.(E); Diaphus spp. (deep)(E); V. nimbaria(E); L. alatus(E); M. asperum(E) Avg. Dissim. = 66%; Diaphus spp. (slender) 

The species listed are those which contributed highest and consistently to the characterizations. The location given in superscript brackets denotes the greater concentration of the given characterizing species; S (shelf), P (perimeter), E (eddy) and O (oceanic). Asterisks denote significant differences.

Fig. 7.

Multi-dimensional scaling (MDS) ordinations showing larval fish assemblage structure in relation to the a priori factors of (A) location and (B) mixed layer water mass. Sampling stations corresponding with the trajectories of three oceanographic drifters are shown in (A) and represent deployments in the LC meander (stations 3, 5 and 7), eddy perimeter (stations 4, 6 and 10) and eddy centre (stations 14 and 17).

Fig. 7.

Multi-dimensional scaling (MDS) ordinations showing larval fish assemblage structure in relation to the a priori factors of (A) location and (B) mixed layer water mass. Sampling stations corresponding with the trajectories of three oceanographic drifters are shown in (A) and represent deployments in the LC meander (stations 3, 5 and 7), eddy perimeter (stations 4, 6 and 10) and eddy centre (stations 14 and 17).

Overall, larval fish assemblages were also structured significantly based on water mass (R = 0.6, P = 0.001), although, due to the regional dominance of modified LC water within the eddy field, this was not reflected in the ordination (Fig. 7B). The STSW assemblage was significantly different from both LC and modified LC assemblages (Table III). There was also a weak significant difference between LC and modified LC water assemblages (Table III). Overall, the factors of location and water mass were reasonably good descriptors of larval fish assemblages though there is a strong likelihood of an interactive effect between these two factors which could not be tested due to insufficient replication at each factor level.

Table III:

One-way ANOSIM and SIMPER analyses between larval fish assemblages classified according to the a priori factor water mass

 Leeuwin Current Modified Leeuwin Current Sub-tropical surface water 
Leeuwin Current Avg. Dissim. = 62%; Diaphus spp.(slender); Lampanyctus sp.; Vinciguerria spp.; V. nimbaria; Diaphus spp. (deep); L. alatus; M. asperum; H. hygomii; Bregmaceros spp.   
Modified Leeuwin Current *R = 0.3, P = 0.001; Avg. Dissim. = 51%; low and inconsistent contribution of several taxa Avg. Dissim. = 45%; Diaphus spp.(slender); Vinciguerria spp.; Bregmaceros spp.; V. nimbaria; Diaphus spp. (deep); Notosudidae spp.; M. asperum; L. alatus  
Sub-tropical surface water *R = 0.7, P = 0.002; Avg. Dissim. = 75%; Diaphus spp. (deep) (LC); L. alatus(LC); Vinciguerria spp. (LC); M. asperum(LC); V. nimbaria(LC) *R = 0.9, P = 0.001; Avg. Dissim. = 73%; Bregmaceros spp. (mod. LC); Vinciguerria spp. (mod. LC); Diaphus spp. (slender) (mod. LC); Diaphus spp.(deep)(mod. LC); Notosudidae spp. (mod. LC); V. nimbaria(mod.LC); L. alatus(mod. LC); M. asperum(mod. LC) Avg. Dissim. = 66%; Diaphus spp.(slender) 
 Leeuwin Current Modified Leeuwin Current Sub-tropical surface water 
Leeuwin Current Avg. Dissim. = 62%; Diaphus spp.(slender); Lampanyctus sp.; Vinciguerria spp.; V. nimbaria; Diaphus spp. (deep); L. alatus; M. asperum; H. hygomii; Bregmaceros spp.   
Modified Leeuwin Current *R = 0.3, P = 0.001; Avg. Dissim. = 51%; low and inconsistent contribution of several taxa Avg. Dissim. = 45%; Diaphus spp.(slender); Vinciguerria spp.; Bregmaceros spp.; V. nimbaria; Diaphus spp. (deep); Notosudidae spp.; M. asperum; L. alatus  
Sub-tropical surface water *R = 0.7, P = 0.002; Avg. Dissim. = 75%; Diaphus spp. (deep) (LC); L. alatus(LC); Vinciguerria spp. (LC); M. asperum(LC); V. nimbaria(LC) *R = 0.9, P = 0.001; Avg. Dissim. = 73%; Bregmaceros spp. (mod. LC); Vinciguerria spp. (mod. LC); Diaphus spp. (slender) (mod. LC); Diaphus spp.(deep)(mod. LC); Notosudidae spp. (mod. LC); V. nimbaria(mod.LC); L. alatus(mod. LC); M. asperum(mod. LC) Avg. Dissim. = 66%; Diaphus spp.(slender) 

The species listed are those which contributed highest and consistently to the characterizations. The water mass given in superscript brackets denotes the greater concentration of the given characterizing species; S (shelf), P (perimeter), E (eddy) and O (oceanic). Asterisks denote significant differences.

Each location and water mass assemblage was characterized by moderate-to-high within-group variability as indicated by the high average dissimilarity values (Tables II and III). Generally, larval fish assemblages representing locations and water masses were delineated based largely upon varying abundances of the dominant Myctophidae (Diaphus spp.) and Phosichthyidae (Vinciguerria spp.). Although the shelf assemblage comprised a number of neritic taxa, including S. sagax and the anchovy E. australis, the low abundances of these larvae, which were also spatially variable, resulted in these having a low and inconsistent contribution to characterization of the shelf assemblage. Nevertheless, assemblages within the meander/eddy perimeter and at the eddy centre displayed an affinity with the shelf assemblage based on the occurrence of larvae of a number of neritic taxa, including S. sagax. Likewise, differentiation between LC and modified LC assemblages was based upon varying abundances of similarly occurring larvae of meso-pelagic and neritic taxa.

Of the environmental factors tested using the BVSTEP routine, water temperature had the highest correlation (ρ = 0.66, P = 0.001) with larval fish assemblages. The combination of temperature and surface chlorophyll a provided the next highest correlation (ρ = 0.57).

Larval pathways

The abundance of neritic and meso-pelagic larval fishes within the LC meander did not change significantly along the trajectory of the first drifter (stations 3, 5 and 7) over the 7-day period (P = 0.78) (Fig. 8A). The assemblages showed weak spatial separation in the ordinations relating to both location and water mass (Fig. 7A and B). There was a significant increase in the abundance of larvae of oceanic meso-pelagic taxa along the trajectory of the second drifter within the eddy perimeter (Fig. 8B) between stations 4 (southern perimeter) and 6 (north perimeter (P = 0.03). This was followed by a significant decrease in abundance between stations 6 and 10 (southern perimeter) (P = 0.02). In contrast, the trajectory of the eddy centre assemblage tended more towards oceanic/STSW over the 6 days that the third drifter was followed (Fig. 7A and B; stations 14 and 17). This was driven by an observable decrease in larval abundance, in particular, the dominant Diaphus spp. (slender morphotype), although this difference was not significant (P = 0.09) (Fig. 8B).

Fig. 8.

Mean abundance (number m−2) of larval fishes for sampling stations corresponding with the trajectories of three oceanographic drifters for (A) LC meander stations from northeast to southeast, (B) eddy perimeter and (C) eddy centre. The proportions of oceanic meso-pelagic (black) and neritic (white) taxa to the mean abundances are shown. Standard error bars are for the mean abundance of larval fishes per sampling station.

Fig. 8.

Mean abundance (number m−2) of larval fishes for sampling stations corresponding with the trajectories of three oceanographic drifters for (A) LC meander stations from northeast to southeast, (B) eddy perimeter and (C) eddy centre. The proportions of oceanic meso-pelagic (black) and neritic (white) taxa to the mean abundances are shown. Standard error bars are for the mean abundance of larval fishes per sampling station.

Length frequency analysis of abundant larvae of epipelagic, neritic species collected using surface samples (Fig. 9) indicated that larvae of E. australis from shelf stations were mainly preflexion with a mean size of 8 mm SL. In contrast, at the eddy centre, larvae of this species had a significantly larger mean size of 14 mm SL (P < 0.0001). For S. sagax, although larvae were fewer, the same pattern was evident; mean size was 9.3 mm SL and 13.3 mm SL for the shelf and eddy, respectively (P < 0.00001). For the neustonic larvae of Gonorhynchus greyii, the mean size was also significantly larger in the eddy centre (17.9 mm SL) when compared with that on the shelf (13.8 mm SL) (P < 0.0004, t-test).

Fig. 9.

Length frequency distributions of E. australis, S. sagax and G. greyii collected in surface waters for shelf and eddy centre stations off south-WA, May 2006 (n = number of larvae collected).

Fig. 9.

Length frequency distributions of E. australis, S. sagax and G. greyii collected in surface waters for shelf and eddy centre stations off south-WA, May 2006 (n = number of larvae collected).

DISCUSSION

Meso-scale eddies associated with the LC are well-known oceanographic phenomena of the south-east Indian Ocean (Feng et al., 2005; Fieux et al., 2005). There are detailed biological studies of mature, detached eddies which suggest that, during their development, neritic, planktonic biota is incorporated (Strzelecki et al., 2007; Waite et al., 2007a). The developing eddy studied here allowed examination of the incorporation of planktonic biota, particularly neritic fish larvae, which have a limited pelagic larval phase (circa 1 month) during which time they can act as biological tracers of specific water masses.

The May 2006 eddy arose from a meander of the LC, but results, in terms of water properties, indicated that considerable exchange occurred between the eddy and adjacent shelf waters which altered the physical, chemical and biological signature of the eddy (Paterson et al., 2008). Furthermore, satellite observations showed upstream incursions of the LC onto the shelf north of the developing eddy which were probably important for generation of modified LC water and initial entrainment of planktonic stages of neritic biota. Overall, the oceanography throughout the eddy field was characterized by strong spatial and temporal variability which was reflected in the physical properties (e.g. TS, horizontal velocities) of the water column (Paterson et al., 2008; Waite et al., unpublished data).

The abundance of larval fishes within the May 2006 developing eddy was significantly higher when compared with that of the surrounding oceanic waters and is consistent with the accumulation of larvae and other planktonic biota reported for eddies off Alaska (Mackas and Galbraith, 2002; Mackas et al., 2005), the Canary Islands (Rodriguez et al., 2004), California (Nishimoto and Washburn, 2002) and Japan (Okazaki and Nakata, 2007). Earlier studies of larval fish assemblages off south-WA (Muhling and Beckley, 2007; Muhling et al., 2008) indicated a distinct cross-shelf gradient with lower abundance and diversity in oceanic waters.

Although there was a wide diversity of larvae representing 68 families, numerically the larvae of the oceanic, meso-pelagic families Myctophidae and Phosichthyidae dominated across the eddy field. The high abundances of early stage larvae of these fishes, particularly, Diaphus spp., indicates that adult spawning occurred in proximity to the eddy. Many of these oceanic meso-pelagic larvae occurred on the shelf, consistent with the earlier description by Muhling et al. (Muhling et al., 2008) and are indicative of the pervasive shoreward influence of the LC in the austral autumn/winter (Pearce et al., 2006). Such ocean-shelf advective transport of Myctophidae larvae has been documented elsewhere including in the south-western Indian Ocean (Olivar and Beckley, 1994), western Pacific (Sánchez-Velasco et al., 2004) and Mediterranean Sea (Sabatés and Masó, 1992; Olivar et al., 2010).

Within the LC meander/eddy perimeter and at the eddy centre, numerous neritic taxa were captured indicating seaward transport and confirming the hypothesis that these larvae are incorporated into anti-cyclonic LC eddies as they develop in proximity to the shelf break. These larvae represented demersal (e.g. Bothidae), pelagic (e.g. Clupeidae, Engraulidae, Carangidae) and reef-dwelling (e.g. Lutjanidae, Pomacentridae) families and this contrasted with their absence in samples from the surrounding oceanic water. Furthermore, the occurrence of larvae of tropical neritic fish taxa is indicative of their transport south via the LC. Upstream incursions (at lower latitudes) of the current onto the shelf north of the developing eddy is a process that is considered to be important for mixing and the initial entrainment of tropical neritic larvae into the LC (Pearce et al., 2006). Advective loss of neritic larvae due to eddy entrainment has been shown to result in the reduction of recruitment of marine fish populations, particularly for eddies of the Gulf Stream (Flierl and Wroblewski, 1985; Myers and Drinkwater, 1989).

Larval fish assemblages showed clear spatial structuring corresponding to shelf, eddy and oceanic locations although assemblages occurring within the LC meander/eddy perimeter and at the eddy centre shared some affinity with the shelf assemblage. The larval fish assemblages within the eddy field were not as strongly structured by water mass as previously reported for the region by Muhling et al. (Muhling et al. 2008) which was largely reflective of the regional dominance of modified LC water over the shelf and in the eddy during the study period. It was apparent that the ubiquitous distributions of several highly abundant oceanic meso-pelagic taxa (e.g. Diaphus spp., Vinciguerria spp.), which were associated with all locations and water masses, restricted the predictive capacity of water mass as a driving factor. Indeed, water mass history is an important factor to be considered (e.g. Cowen et al., 1993; Dempster et al., 1997). Gray and Miskiewicz (Gray and Miskiewicz, 2000) suggested that variability of larval fish assemblages associated with the East Australian Current on the continental shelf off Sydney was due to the circulation history of the current, particularly onshore and offshore meandering. Nevertheless, structuring of larval fish assemblages was correlated with average temperature of the mixed layer (>60%), and was probably driven by the lower diversity and abundance of larval fishes in the cooler STSW.

This study, which coupled Lagrangian drifters with larval fish sampling, did not identify any significant change in larval fish abundances along a trajectory around the LC meander. As the same water mass was repeatedly sampled, it would be expected that the assemblage would be similar over time, although larvae would be subject to natural mortality and, in the case of oceanic species, possible entrainment from the surrounding ocean. A significantly higher abundance of larval fishes at the northern eddy perimeter contrasted with lower abundances at the southern perimeter, suggesting that the northern perimeter was the region of entrainment of larvae to the eddy. At the eddy centre, there was also an observable decrease in the abundance of larval fishes over the 6-day period of drifter tracking, particularly for the larvae of oceanic meso-pelagic fishes, although this difference was not significant. While the exact reasons for these decreases in abundance are not known, it is possible that many of these larvae suffered natural mortality and there was no replenishment of early-stage larvae to the eddy. In addition to this, it is also possible that larvae were transported deeper due to downwelling as the eddy strengthened and that the larvae avoided the net or migrated deeper (>150 m for bongo tows) with larval ontogeny (Ropke, 1993; Sassa et al., 2007).

In the study by Muhling et al. (Muhling et al., 2007) of larval fish assemblages of a mature anti-cyclonic eddy that had detached from the LC 5 months previously, there were no neritic larvae and the assemblage was comprised of meso-pelagic oceanic species. However, this was not the case in the May 2006 developing eddy where larvae of many taxa of neritic fishes were recorded in the bongo samples, albeit in low numbers. These neritic larvae thus can serve as tracers reflecting incorporation of water from the shelf, and in particular from lower latitudes via the LC, into the developing eddy. Further, the comparison of the length frequencies of dominant neritic larvae collected near the surface with a large net (1 mm mesh) over the shelf and within the eddy showed that for species such as E. australis, G. greyi and S. sagax, those occurring in the eddy were significantly larger than those in shelf waters. This would support the hypothesis that neritic larvae were subject to cross-shelf transport and incorporated early on in the evolution of the eddy and had subsequently developed therein.

In summary, larval fishes were more abundant within a developing anti-cyclonic warm-core eddy of the LC compared with the surrounding ocean. The composition of the eddy assemblage was complex and reflects the origin of the waters that were incorporated into the eddy. The assemblage was comprised largely of larvae of meso-pelagic fishes from the regional ocean and LC, possibly supplemented by localized spawning in proximity to the eddy. In addition, larvae of temperate, neritic species that had been subjected to cross-shelf transport during early development of the eddy, as well as some tropical neritic species transported southwards by the LC, contributed to the assemblage.

SUPPLEMENTARY DATA

Supplementary data can be found online at http://plankt.oxfordjournals.org.

FUNDING

D.H. was supported by a Murdoch University PhD scholarship and funding from the Western Australian Marine Science Institution.

ACKNOWLEDGEMENTS

This research was conducted under Murdoch University Animal Ethics permit number W1181/06. We acknowledge the Australian Marine National Facility for ship's time, an Australian Research Council Discovery Grant to Anya Waite and a Murdoch University grant to Lynnath Beckley. We thank Anya Waite for making the oceanographic data available and Ming Feng and Peter Thompson for helpful discussions. The captain, crew and scientific complement of the R.V. Southern Surveyor are thanked for their contributions during the voyage.

REFERENCES

Atwood
E.
Duffy-Anderson
J. T.
Horne
J. K.
, et al.  . 
Influence of mesoscale eddies on ichthyoplankton assemblages in the Gulf of Alaska
Fish. Oceanogr.
 , 
2010
, vol. 
19
 (pg. 
493
-
507
)
Clarke
K. R.
Gorley
R. N.
Primer v6: User Manual/Tutorial
 , 
2006
Plymouth
PRIMER-E Ltd
Clarke
K. R.
Warwick
R. M.
Primer-6 Computer Program
 , 
2005
Plymouth
Natural Environment Research Council
Cowen
R. K.
Hare
J. A.
Fahay
M. P.
Beyond hydrography: can physical processes explain larval fish assemblages within the middle Atlantic Bight?
Bull. Mar. Sci.
 , 
1993
, vol. 
53
 (pg. 
567
-
587
)
Cresswell
G. R.
The Leeuwin Current—observations and recent models
J. Roy. Soc. W. A.
 , 
1991
, vol. 
74
 (pg. 
1
-
14
)
Cresswell
G. R.
Golding
T. J.
Observations of a south-flowing current in the south-eastern Indian Ocean
Deep Sea Res.
 , 
1980
, vol. 
27
 (pg. 
449
-
466
)
Dempster
T.
Gibbs
M. T.
Rissik
D.
, et al.  . 
Beyond hydrography: daily ichthyoplankton variability and short term oceanographic events on the Sydney continental shelf
Cont. Shelf Res.
 , 
1997
, vol. 
17
 (pg. 
1461
-
1481
)
Feng
M.
Majewski
L.
Fandry
C. B.
, et al.  . 
Characteristics of two counter-rotating eddies in the Leeuwin Current system off the Western Australian coast
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
961
-
980
)
Feng
M.
Meyers
G.
Pearce
A.
, et al.  . 
Annual and interannual variations of the Leeuwin Current at 32°S
J. Geophys. Res.
 , 
2003
, vol. 
108
 pg. 
3355
 
Feng
M.
Wijffels
S.
Godfrey
J. S.
, et al.  . 
Do eddies play a role in the momentum balance of the Leeuwin Current?
J. Phys. Oceanogr.
 , 
2005
, vol. 
35
 (pg. 
964
-
975
)
Fieux
M.
Molcard
R.
Morrow
R.
Water properties and transport of the Leeuwin Current and eddies off Western Australia
Deep-Sea Res. I
 , 
2005
, vol. 
52
 (pg. 
1617
-
1635
)
Flierl
G. R.
Wroblewski
J. S.
The possible influence of warm core Gulf Stream rings upon shelf water larval fish distribution
Fish. Bull.
 , 
1985
, vol. 
88
 (pg. 
313
-
330
)
Gaughan
D. J.
Management implications of the Leeuwin Current eddy system for teleosts of the Western Australian continental shelf
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
1129
-
1140
)
Govoni
J. J.
Fisheries oceanography and the ecology of early life histories of fishes: a perspective over fifty years
Sci. Mar.
 , 
2005
, vol. 
69
 (pg. 
125
-
137
)
Graber
H. C.
Limouzy-Paris
C. B.
Transport patterns of tropical reef fish larvae by spin-off eddies in the Straits of Florida
Oceanography
 , 
1997
, vol. 
10
 (pg. 
68
-
71
)
Gray
C. A.
Miskiewicz
A. G.
Larval fish assemblages in south-east Australian coastal waters: seasonal and spatial structure
Estuarine Coastal Shelf Sci.
 , 
2000
, vol. 
50
 (pg. 
549
-
570
)
Hare
J. A.
Churchill
J. H.
Cowen
R. K.
, et al.  . 
Routes and rates of larval fish transport from the southeast to the northeast United States continental shelf
Limnol. Oceanogr.
 , 
2002
, vol. 
47
 (pg. 
1774
-
1789
)
Kasai
A.
Kimura
S.
Nakata
H.
, et al.  . 
Entrainment of coastal water into a frontal eddy of the Kuroshio, and its biological significance
J. Mar. Syst.
 , 
2002
, vol. 
37
 (pg. 
185
-
198
)
Ladd
C.
Stabeno
P.
Cokelet
E. D.
A note on cross-shelf exchange in the northern Gulf of Alaska
Deep-Sea Res. II
 , 
2005
, vol. 
52
 (pg. 
667
-
669
)
Leis
J. M.
Carson-Ewart
B. M.
The Larvae of Indo-Pacific Coastal Fishes: A Guide to Identification
 , 
2000
Leiden
Brill
pg. 
850
 
Levitus
S.
Climatological atlas of the world ocean
 , 
1982
Washington, D.C
U.S. Govt. Printing Office
 
NOAA Professional. Paper, 13
Lobel
P. S.
Robinson
A. R.
Transport and entrapment of fish larvae by ocean mesoscale eddies and currents in Hawaiian waters
Deep-Sea Res.
 , 
1986
, vol. 
33
 (pg. 
483
-
500
)
Mackas
D. L.
Coyle
K. O.
Shelf-offshore exchange processes and their effects on meso-zooplankton biomass and community composition patterns in the northeast Pacific
Deep-Sea Res. II
 , 
2005
, vol. 
52
 (pg. 
707
-
725
)
Mackas
D. L.
Galbraith
M. D.
Zooplankton distribution and dynamics in a north Pacific eddy of coastal origin: 1. Transport and loss of continental margin species
J. Oceanogr.
 , 
2002
, vol. 
58
 (pg. 
725
-
738
)
Mackas
D. L.
Tsurumi
M.
Galbraith
M. D.
, et al.  . 
Zooplankton distribution and dynamics in a North Pacific eddy of coastal origin: II. Mechanisms of eddy colonization by and retention of offshore species
Deep-Sea Res. II
 , 
2005
, vol. 
52
 (pg. 
1011
-
1135
)
Moser
H. G.
Ontogeny and systematics of fishes
1984
pg. 
760
  
American Society for Ichthyologists and Herpetologists Special Publication 1
Moser
H. G.
Watson
W.
Richards
W. J.
Myctophidae: Lanternfishes
Early Stages of Atlantic Fishes: An Identification Guide for the Western Central North Atlantic
 , 
2005
, vol. 
Vol. 1
 
USA
Taylor and Francis Group
Muhling
B. A.
Beckley
L. E.
Seasonal variation in horizontal and vertical structure of larval fish assemblages off south-western Australia, implications for larval transport
J. Plankton Res.
 , 
2007
, vol. 
29
 (pg. 
967
-
983
)
Muhling
B. A.
Beckley
L. E.
Koslow
J. A.
, et al.  . 
Larval fish assemblages and water mass structure off the oligotrophic south-western Australian coast
Fish. Oceanogr.
 , 
2008
, vol. 
17
 (pg. 
16
-
31
)
Muhling
B. A.
Beckley
L. E.
Olivar
M. P.
Ichthyoplankton assemblage structure in two meso-scale Leeuwin Current eddies
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
1113
-
1128
)
Myers
R. A.
Drinkwater
K. F.
The influence of Gulf Stream warm core rings on recruitment of fish in the northwest Atlantic
J. Mar. Res.
 , 
1989
, vol. 
47
 (pg. 
635
-
656
)
Nakata
H.
Kimura
S.
Okazaki
Y.
, et al.  . 
Implications of meso-scale eddies caused by frontal disturbances of the Kuroshio Current for anchovy recruitment
ICES J. Mar. Sci.
 , 
2000
, vol. 
57
 (pg. 
143
-
152
)
Neira
F. J.
Miskiewicz
A. G.
Trnski
T.
Larvae of Temperate Australian Fishes
 , 
1998
University of Western Australia Press
pg. 
474
 
Nishimoto
M. M.
Washburn
L.
Patterns of coastal eddy circulation and abundance of pelagic juvenile fish in the Santa Barbara Channel, California, USA
Mar. Ecol. Prog. Ser.
 , 
2002
, vol. 
241
 (pg. 
183
-
199
)
Nodder
S. D.
Waite
A. M.
Is Southern Ocean organic carbon and biogenic silica export enhanced by iron-stimulated increases in biological production? Sediment trap results from SOIREE
Deep-Sea Res. II
 , 
2001
, vol. 
48
 (pg. 
2681
-
2701
)
Okazaki
Y.
Nakata
H.
Effect of mesoscale hydrographic features on larval fish distribution across the shelf break of East China Sea
Cont. Shelf Res.
 , 
2007
, vol. 
27
 (pg. 
1616
-
1628
)
Olivar
M. P.
Beckley
L. E.
Influence of the Agulhas Current on the distribution of lanternfish larvae off the southeast coast of Africa
J. Plankton Res.
 , 
1994
, vol. 
16
 (pg. 
1759
-
1780
)
Olivar
M. P.
Emelianov
M.
Villate
F.
, et al.  . 
The role of oceanographic conditions and plankton availability in larval fish assemblages off the Catalan coast (NW Mediterranean)
Fish. Oceanogr.
 , 
2010
, vol. 
19
 (pg. 
209
-
229
)
Olivar
M. P.
Fortuño
J. M.
Guide to ichthyoplankton of the Southeast Atlantic (Benguela Current region)
Sci. Mar.
 , 
1991
, vol. 
55
 (pg. 
1
-
383
)
Olivar
M. P.
Moser
H. G.
Beckley
L. E.
Lanternfish larvae from the Agulhas Current (SW Indian Ocean)
Sci. Mar.
 , 
1999
, vol. 
63
 (pg. 
101
-
120
)
Paterson
H. L.
Feng
M.
Waite
A. M.
, et al.  . 
Physical and chemical signatures of a developing anticyclonic eddy of the Leeuwin Current, Eastern Indian Ocean
J. Geophys. Res.
 , 
2008
, vol. 
113
 pg. 
C07049
 
Pearce
A. F.
Eastern Boundary Currents of the Southern Hemisphere
J. Roy. Soc. W. A.
 , 
1991
, vol. 
74
 (pg. 
35
-
45
)
Pearce
A. F.
Lynch
M. J.
Hanson
C. E.
The Hillarys transect (1): seasonal and cross-shelf variability of the physical and chemical properties off Perth, Western Australia, 1996–98
Cont. Shelf Res.
 , 
2006
, vol. 
26
 (pg. 
1689
-
1729
)
Richards
W. J.
Early Stages of Atlantic fishes: An Identification Guide for the Western Central North Atlantic
 , 
2006
, vol. 
Vol. 1
 
USA
Taylor and Francis Group
pg. 
2640
 
Rodriguez
J. M.
Barton
E. D.
Hernandez-Leon
S.
, et al.  . 
The influence of mesoscale physical processes on the larval fish community in the Canaries CTZ in summer
Prog. Oceanogr.
 , 
2004
, vol. 
62
 (pg. 
171
-
188
)
Ropke
A.
Do larvae of mesopelagic fishes in the Arabian Sea adjust their vertical distribution to physical and biological gradients?
Mar. Ecol. Prog. Ser.
 , 
1993
, vol. 
101
 (pg. 
223
-
235
)
Sabatés
A.
Masó
M.
Unusual larval fish distribution pattern in a coastal zone of the western Mediterranean
Limnol. Oceanogr.
 , 
1992
, vol. 
37
 (pg. 
1252
-
1260
)
Sánchez-Velasco
L.
Avalos-García
C.
Rentería-Cano
M.
, et al.  . 
Fish larvae abundance and distribution in the central Gulf of California during strong environmental changes (1997–1998 El Niño and 1998–1999 La Niña)
Deep-Sea Res. II
 , 
2004
, vol. 
51
 (pg. 
711
-
722
)
Sassa
C.
Kawaguchi
K.
Hirota
Y.
, et al.  . 
Distribution depth of the transforming stage larvae of myctophid fishes in the subtropical–tropical Waters of the western North Pacific
Deep-Sea Res. I
 , 
2007
, vol. 
54
 (pg. 
2181
-
2193
)
Sponagule
S.
Lee
T.
Kourafalou
V.
, et al.  . 
Florida Current frontal eddies and the settlement of coral reef fishes
Limnol. Oceanogr.
 , 
2005
, vol. 
50
 (pg. 
1033
-
1048
)
Strzelecki
J.
Koslow
J. A.
Waite
A.
Comparison of mesozooplankton communities from a pair of warm- and cold-core eddies off the coast of Western Australia
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
1103
-
1112
)
Waite
A. M.
Pesant
S.
Griffin
D. A.
, et al.  . 
Oceanography, primary production and dissolved inorganic nitrogen uptake in two Leeuwin Current eddies
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
981
-
1002
)
Waite
A. M.
Thompson
P. A.
Pesant
S.
, et al.  . 
The Leeuwin Current and its eddies: an introductory overview
Deep-Sea Res. II
 , 
2007
, vol. 
54
 (pg. 
789
-
796
)
Werner
F. E.
Quinlan
J. A.
Fluctuations in marine populations: physical processes and numerical modelling
ICES Mar. Sci. Sym.
 , 
2002
, vol. 
215
 (pg. 
264
-
278
)

Author notes

Corresponding editor: Mark J. Gibbons

Supplementary data