Biological Oceanography of Larval Fish Diversity and Growth Off Eastern Australia

Home , Atypichthys, Pseudocaranx, Yellowtail scad

Biological oceanography of larval fish diversity and growth off eastern Australia

Augy Syahailatua

A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy University of New South Wales, AUSTRALIA

August 2005

ABSTRACT

Fish larvae in Australian waters have been studied progressively in the last 2-3 decades including the distribution and abundance of taxa, growth and age, their prey and predators. However, the effect of nutrient limitation on ichthyoplankton is unstudied, particularly in the oligotrophic Australian waters. My study was aimed to examine the effect of natural or anthropogenic nutrients on the abundance, distribution, growth and condition of fish larvae along-shore of the NSW coast (latitude 30-34ºS), where the East Australian Current departs the NSW coast and generates local upwelling of cool nutrient-rich water. This study shows no significant difference in the total abundance or diversity of either larval fishes amongst the 112 taxa (111 families and 1 order), among regions within or upstream of the upwelling. However in both months, there were distinctive ichthyoplankton assemblages at the family level. The Carangidae, Labridae, Lutjanidae, Microcanthidae, Myctophidae and Scombridae were more abundant in the EAC or oceanic water masses, while the Callionymidae, Clupeidae, Platycephalidae, Sillaginidae and Terapontidae were mostly found in the surface or deep upwelled/uplifted water masses. This pattern is observed in other ichthyoplankton studies and may be a general and useful method to determine mixing of water masses. Larvae of silver trevally (Pseudocaranx dentex) and yellowtail scad (Trachurus novaezelandiae) were generally larger and less abundant in the topographically induced upwelling region, than north of the region in pre-upwelled conditions of the East Australian Current. Both species were mostly at the preflexion stage (<4.3 mm in body length and <10 days old) in the pre-upwelled conditions, particularly during November, and proportionally more larger and older larvae in the upwelled waters (mostly post- flexion, >4.3 mm in body length and ≥10 days old). Ages from sagittal otoliths ranged from 2-25 increments (~days) and exhibited linear growth for both species and months over the size range (3-15 mm standard length). The otolith radius-length relationship and the growth rates were similar between species and months, despite the 3-4ºC difference between months. Overall growth rates of the younger larvae were uniform throughout the entire sampling area (0.5-0.6 mm.d-1), while older larvae grew significantly faster in the upwelled water (0.41±0.12 mm.d-1) compared to the non- upwelled conditions (0.34±0.11 mm.d-1). Both species tended to be depleted in δ13C in the upwelling region (from –18.5 to –19.0‰), consistent with expected ratios from iii

deeper water, whereas the δ15N composition tended to increase in Pseudocaranx, but decrease in Trachurus indicating different diets and possibly trophic level. The early life history of both species indicates spawning in pre-upwelled waters, but larval transport into upwelled waters is necessary for faster growth in the post-flexion stage. The assemblage of larval fishes did differ between the upwelled region and a region south of Sydney’s deepwater outfalls, but the difference was ascribed to a latitudinal effect and the EAC. Both larval carangids were enriched in 15N, possibly due to the enriched dissolved organic matter of primary treated sewage. In summary, this study found that the larval fish community can provide a biological means to trace water masses, and estimate their degree of mixing. Remarkably there was no significant effect of upwelling or sewage addition to the abundance or diversity of larval fish, in the nutrient poor waters of the East Australian Current. Larval carangids and pilchards were abundant in late spring off northern NSW, and their early life histories were inferred. Both larval carangid species seem to be spawned in the EAC waters, but as post-flexion larvae grew faster in the upwelled zone. Pre-flexion (<10 day old) larval carangids of both genera indicated spawning in the EAC, and the rarer post-flexion (>10 days old) carangids grew faster in the upwelled waters. Here, both genera had stable isotope signatures characteristic of upwelled waters for carbon, but had different nitrogen signatures, indicative of different diets and trophic level status. Larval pilchards actually grew more slowly in the upwelling region, as observed in coastal waters off Japan, and their nursery grounds may be further offshore in the Tasman Front, analogous to their early life history in the Kuroshio Extension. iv

Acknowledgment

I would like to thank and really appreciate my supervisor A/Prof. Iain Suthers for his time, great advice, guidance and encouragement throughout my study period. I also would like to acknowledge my co-supervisor Dr Jeff Leis (Australian Museum) for his suggestions and comments on my research proposal and the drafts of my thesis and manuscripts. Special thanks are directed to the Australian Development Scholarship of the Australian Agency for International Development and to the Research Centre for Oceanography of Indonesian Institute of Sciences for the trust to assign me to undertake a PhD program in Australia. I especially wish to thank Dr Tony Miskiewicz and Dr Kim Smith for their valuable guidance and assistance on larval fish identification and comments on the drafts of the manuscripts. I wish to acknowledge the Australian Research Council for funding this research, and I am most grateful to the Captain and crew of the RV ‘Franklin’, and particularly the CSIRO personal for maintaining the electronics, the nutrient analyses, and providing oceanographic data. I also thank the support of Prof. Jason Middleton, Mr Greg Nipard, Dr Moninya Roughan, Dr Jocelyn Dela-Cruz and Mr. Richard Piola during the field and laboratory works, and data analyses. Assistance of Dr Troy Gaston on stable isotope analyses is much appreciated. I also extend my appreciation to all former and current students at Suthers’ lab for their huge assistance, discussion and friendship, and always making me felling at home. My thanks are due to all friends and colleagues (specially Mr. Adisyahmeta) for their valuable support, cooperation and friendships throughout my life, particularly during my study in Australia. My truthful appreciation and thanks go also to the Moentaco and Pesik families for their knowledgeable and kind hospitality given to me during completing my thesis. I would like to express my sincere thanks to my mum and dad for their great love, prayers and strong encouragement. You both are incredible parents making my life is so brightly and beautifully. Last but not least, I would like to thanks my wife, Ella, for her endless love, patient, and kindness during my absence from home. v

Table of contents Page

Abstract i Acknowledgments iii List of Tables vii List of Figures x

Chapter 1: Introduction

1.1 Recruitment of fishes 1 1.2 Starvation and predation 2 1.3 Growth and condition of larvae 4 1.4 Stable isotope signatures 6 1.5 Biological oceanography of eastern Australia 8 1.6 Ichthyoplankton research in Australia 10 1.7 Aims of this study 16

Chapter 2: Is upwelling marked by signature ichthyoplankton? Larval assemblages as tracers of mixing

Abstract 19 2.1 Introduction 20 2.2 Methods 21 Study area 21 Ichthyoplankton sampling 22 Analysis 26 2.3 Results 27 Oceanographic features 27 Larval fish taxonomy and composition 32 Comparisons of the assemblages 40 Wind-induced upwelling event, 16 November 43 2.4 Discussion 46 vi

Ichthyoplankton assemblage structure 46

Chapter 3: A regional and stage-specific response of two larval carangids in an upwelling region of the East Australian Current

Abstract 51 3.1 Introduction 52 3.2 Methods 53 Study area 53 Sampling techniques 54 Laboratory procedure 57 Growth calculation 57 Stable Isotope Analysis (SIA) 59 Statistical analysis 59 3.3 Results 61 Abundance and size 61 Growth and age 66 Stable isotope signatures 75 3.4 Discussion 78 Early life history of two carangid larvae 78

Chapter 4: Larval fish diversity and growth in an upwelling and sewage impacted zones in spring 1998

Abstract 82 4.1 Introduction 83 4.2 Methods 84 Study area 84 vii

Sampling procedure 85 Laboratory procedure 86 4.3 Results 88 Oceanographic feature 88 Ichthyoplankton assemblage 93 Larval carangids 95 4.4 Discussion 104

Chapter 5: General Conclusion

5.1 Ichthyoplankton variability and dynamic in NSW coastal zone 108 5.2 Size structures, growth and condition 109 5.3 Upwelled water impacted versus sewage impacted zones in NSW coast 110 5.4 Further study and recommendation 110

Literature cited 111

Appendix 1. The recent growth rate of larval pilchards, Sardinops sagax in relation to their stable isotope composition, in an upwelling zone of the East Australian Current 131

viii

List of tables Page Table 1.1. Relative stable isotopes abundance (%) in marine and sewage particulate organic matter (after Spies et al. 1989). 8

Table 1.2. Summary of major finding on larval fish studies of species specific in Australian waters. 11

Table 2.1. Summary of sampling regions numbered from north to south, location name, stations (inshore, 50 m isobath; offshore, 100 m isobath), latitude, longitude and sampling date in November 1998 and January 1999. n-Nov and n-Jan are the numbers of samples in November and January respectively (surface, sub-surface). Sampling was conducted during the hours of darkness from 20:00-05:30. 25

Table 2.2. List of ichthyoplankton taxa (in alphabetical order), the total number of larvae caught over both cruises, their contribution (%) and rank based on individual number caught at surface and sub-surface in November 1998 and January 1999. 0, refers to <0.1%; -, refers to no larvae caught, numbers and families in bold refers to >1%, and were analysed for assemblage structure. If the family abundance is >5% or contains a commercial species, the dominant species are listed where possible. 33

Table 2.3. Summary table of analysis of variance, comparing a, c) the total abundance of larvae, and b, d) the Shannon-Wiener diversity indices among regions, stations and sampling depths (Fig. 2.5), of the cruises in a, b) November 1998 and c, d) January 1999. In January, only regions 2, 3 and 4 were compared with balanced data. In d), no transformation could stablise the variance so significance levels (**) are interpreted with caution. NS= not significant, *=0.01

Table 2.4. Analysis of similarity (ANOSIM, two-way crossed analysis) of the community to compare between stations inshore (IS) versus offshore (OS) stations, and among regions. P= significance level of sample statistic; NS= not significant, *=0.01

Table 2.5. The ichthyoplankton assemblages from the MDS ordinations (Fig. 2.6) and the average similarity (%) of each assemblage from the Similarity Percentages (SIMPER) procedure. The first five most influential families/taxa are listed. 41

Table 2.6. The average abundance (numbers. 100 m-3) in the EAC associated assemblages N1, J1, J2, J3) and the upwelling associated assemblages (N2, and J4, J5, as identified in Fig. 2.6). The table is derived from the dominant families identified in the SIMPER analysis (Table 2.4), with at ix

least one abundance >1.2 per standardized tow. The assemblage is classified as EAC or Upwelled on the relative abundance between the two groups. Classification of bothids, gobys and sparids is unclear. 45

Table 2.7. Pearson correlation coefficients (R) of the seven assemblages identified from the 14 taxa in Table 2.6. Coefficients in bold are significant (p<0.05). 45

Table 2.8. Numbers of families of ichthyoplankton recorded for shelf waters in western boundary current systems (excluding Anguilliformes). 50

Table 3.1. Summary table of the Region of sampling (see Fig. 3.1), Date and Station (inshore; 50 m isobath, offshore; 100 m isobath), SST (sea surface temperature from the thermosalinograph), number of tows (surface, neuston; deep, subsurface), number of Pseudocaranx dentex and Trachurus novaezelandiae, (surface, deep) sampled during the cruises in November 1998 and January 1999. Sample sizes for the November otolith analysis are Pd-oto and Tn-oto for Pseudocaranx and Trachurus respectively. 55

Table 3.2. Hatch length of Pseudocaranx and Trachurus larvae used in the growth back-calculation. 58

Table 3.3. Summary of Analysis of Similarity (ANOSIM) for the size structure of Pseudocaranx dentex and Trachurus novaezelandiae taken throughout an upwelling zone off NSW in November 1998 and January 1999. S, surface; D, deep or sub-surface; I, inshore station at the 50 m isobath; O, offshore at the 100 m isobath; P = significance level; ns=not significant at P = 0.05; R = Global R statistics. 60

Table 3.4. Summary of ANOVA of larval abundance (no. of individual.100m- ³) of Pseudocaranx and Trachurus sampled in November 1998 and January 1999. Data transformed to Ln(x+1); R=Region; S=Station (50 and 100 m isobath); D= Depth (surface and sub-surface); *, P<0.05; **, P<0.01. 61

Table 3.5. One-factor ANOVA table, for a) Pseudocaranx and b) Trachurus comparing November’s overall growth rate (G) amongst 12 treatments (3 R, regions * 2 S, stations * 2 size classes, Fig.3.8). 70

Table 3.6. Combined one-factor ANCOVA table for a) Pseudocaranx b) Trachurus (Fig. 3.9, significant difference among intercepts) comparing back-calculated recent growth (RG1&2) amongst regions (R; regions 1, 3 and 4) with standard length (SL) the covariate. NS=not significant different at P=0.05; *, P<0.05; **, P<0.01. 72

Table 3.7. ANOVA table for a) Pseudocaranx (Fig. 3.12a) and b) Trachurus (Fig. 3.12 b) comparing the recent growth index (RGI). *, P<0.05; **, P<0.01. 75 x

Table 3.8. Summary of the ANOVA results on stable carbon (δ13C) and nitrogen (δ15N) isotopes of Pseudocaranx dentex and Trachurus novaezelandiae from three regions (R, regions 1, 3 and 4), and two stations (S, 50m and 100m isobath). ; *= significant different at P<0.05. 78

Table 4.1. Three factor analysis of variance table of a) total larval abundance (transformed with √(x+1)) and b) Shannon-Weiner diversity index among the two regions (R), two Stations (S) and four depths (D). Significant effects are in bold. 93

Table 4.2. Percentage (%) of taxa individual number that collected during the study off Diamond Head and Port Hacking in November 1998, by averaging the vertical distribution. DH=Diamond Head; PH=Port Hacking; I=50 m, and O=100 m contour stations. 0, refers to <0.1%; -, refers to no larvae caught, numbers and families in bold refers to >1% of the total abundance, and were analysed for assemblage structure. 95

Table 4.3. The ichthyoplankton assemblages from the MDS ordinations (Fig. 4.5) and the average similarity (%) of each assemblage from the Similarity Percentages (SIMPER) procedure. The most influential families that contributed 90% of total individual numbers are listed. 97

Table 4.4. Result of Kolmogorov-Smirnov tests comparing length frequency distributions from Fig. 4.6. 98

Table 4.5. One-way ANCOVA table for a) Pseudocaranx and b) Trachurus comparing RG1&2 amongst four samples (R, S; 2 regions x 2 stations) with standard length (SL) as the covariate (back-calculated to the end of outermost complete increment; mm). 100

Table 4.6. One-way analysis of variance table of the recent growth index (RGI) for a) Pseudocaranx and b) Trachurus, comparing the effects of Region and Station (R, S) as four separate blocks (Fig. 4.9). 101

Table 4.7. Three factor analysis of variance table for a) 13Carbon and b) 15Nitrogen composition of Species (P. dentex, T. novaezelandiae), Region (Diamond Head and Port Hacking) and Station (at the 50 m or 100 m isobath), for comparison with Fig. 4.10. 103 xi

List of figures Page Figure 2.1. Contours of sea surface temperature (SST), derived from satellite images of the northern NSW coast taken by NOAA 14 on a) 21 November, 1998, and b) 19 January, 1999. The sampling regions are numbered 1-5, as referenced in the text and Table 2.1. Expanding symbols show the average total larval abundance for the surface and subsurface nets, at the approximate locations of the inshore and offshore stations. 24

Figure 2.2. Surface T-S plots from the vessel’s thermosalinograph, from 4 m depth, during the plankton tows at each of the regions and stations in a) November 1998 and b) January 1999. i=inshore; o=offshore). Error bars are ± standard error. In November, the theromosalinograph data for the wind-induced upwelling event is circled with a dashed line. 29

Figure 2.3. Temperature profiles along the inshore (50 m contour) and offshore (100 m contour) stations in a) November 1998, and b) January 1999. Dashed lines show the location of the CTD cast. The ichthyoplankton assemblages subsequently identified (N1, N2, J1 to J6, Fig. 2.6) are included. 30

Figure 2.4. Average concentrations in the upper 50 m of the water column in a) nitrate and nitrite (μM), b) silicate (μM), c) free reactive phosphorous (μM) and d) chlorophyll a biomass (μg.L-1), at inshore CTD stations (<50 m isobath) and at offshore stations (50-120 m isobaths), at each region in late January 1999. (1=Urunga; 3=Point Plomer; 4=Diamond Head & Crowdy Head; 5=Cape Hawke). 31

Figure 2.5. The average individual numbers of larval samples/100m3 and the average Shannon-Wiener diversity index at inshore and offshore stations of each region in a & c) surface, November 1998; b & d) sub- surface, November 1998; and e & g) surface, January 1999; f & h) sub-surface, January 1999. 38

Figure 2.6. Cluster dendrogram and two-dimensional MDS ordinations of Bray-Curtis similarities, based on standardised average family abundances of ichthyoplankton at each station/depth, that contributed ≥1% of the total abundances in a, b) November 1998 and c, d) January 1999. The assemblages (N1, N2, J1 to J5) are defined at the 50% Bray Curtis similarity level with a dashed line. The data labels refer to sampling depth (s, surface neuston net; d, deeper EZ net); region (1 to 5 as defined in Table 2.1), and distance offshore (i, inshore, at the 50 m isobath; o, offshore at the 100 m isobath). 42

Figure 2.7. Vertical profile of water temperature on 16 November 1998, showing the wind-induced upwelling. 44

xii

Figure 2.8. a) Cluster dendrogram and b) two-dimensional MDS ordinations of Bray-Curtis similarities of the opportunistic samples taken with a surface, neuston net only, during wind-induced upwelling 16-17 November off Urunga and Smoky Cape (labels in normal font, with a “w”). Labels in bold are the surface assemblages identified from Fig. 2.6. Asseblages are defined at the 50% similarity level with a dashed line. Note that the EAC assemblage N1 (Fig. 2.6, Region 1 and 2) is clearly separate from the new upwelling assemblage N2’ which also includes samples from Region 1 and 2. 44

Figure 3.1. Study area off eastern Australia, showing the spatio-temporal distribution and abundance (# 100m-3) of a) Pseudocaranx dentex and b) Trachurus novaezelandiae larvae at the surface (black circle) and in the sub-surface (grey circle) at the 50 m and 100 m isobath stations, during November 1998 and January 1999. Block arrows indicate the predominant flow of the East Australian Current observed with an Acoustic Doppler Current Profiler (1 to3 knots), and the approximate frontal line observed in SST images off Diamond Head, around the 153°E longitude, during our cruises is indicated for November (22 to 24°C) and for January (24 to 26°C). 56

3 Figure 3.2. Correlation in log10 abundance (no. per 100 m ) between Pseudocaranx and Trachurus, over November and January for the surface and deep nets. Legend shows the pearson correlation coefficients (bold if p<0.05, number of tows). 62

Figure 3.3. MDS plots based on the Bray-Curtis similarity matrix of the abundance in 0.5 mm length categories of Pseudocaranx (Pd) and Trachurus (Tn) sampled in a) November 1998 and b) January 1999. The groups are defined at the 75% Bray-Curtis similarity level, showing the average length (+/- SD). Samples are labeled by species (Pd, Tn), region number (as in Table 3.1, Fig. 3.1), and station, i, inshore (50 m station); o, offshore (100 m station). 64

Figure 3.4. Length frequency histograms of Pseudocaranx and Trachurus sampled in a) November 1998 and b) January 1999. The groups with their sample sizes are based on the MDS plots (Fig. 3.3). 65

Figure 3.5. Unpolished sagittal otoliths from a) Pseudocaranx, 5.4 mm SL, 11 Increments, and, b) Trachurus, SL=4.6 mm, 11 increments and showing the maximum radius typically used for increment width increment measurements; h, hatch check; ffc, first feeding check; p-ffc, pre-first feeding check. Scale bar is 12 μm. 66

Figure 3.6. Mean of increment width (µm) of (a) Pseudocaranx and b) Trachurus from the hatch check to the periphery. The dashed reference line is set a 4 μm. Standard error bars increase towards the periphery as sample size decreases (Fig. 3.7). 67

xiii

Figure 3.7. Body length (BL) on increment count (age) regression for Trachurus in November (BL= 1.05 + 0.31*age, n=231, r2=0.61), had a significantly different intercept to Pseudocaranx in November and January, and Trachurus in January (SL= 1.41 + 0.32*age, n=340, r2=0.70). For comparison, the dashed line is the average age-length regression of T. declivis off eastern Tasmania during summer (ln(SL)=0.051+1.139*age, where age is the days after first feeding, Jordan 1994). 69

Figure 3.8. Standard length divided by the increment count (i.e. mm.d-1) for a) Pseudocaranx and b) Trachurus, for regions 1, 3 and 4 in November 1998. For each region, the first two columns are for young larvae <10 d old at the 50 m and 100 m stations, and the last two columns are for older larvae ≥ 10 d old at the 50 m and 100 m stations. The sample size for each column is indicated (error bars are standard error). 71

Figure 3.9. Scatter plots between the back calculated recent growth over days 1 and 2 pre-capture RG-1&2 (mm. d-1) and SL (back-calculation to the end of outermost complete increments; mm) for Pseudocaranx (a-c) and Trachurus (d-f) for region 1, 3, and 4 at inshore station (50 m isobath). 73

Figure 3.10. Relationship between SL (mm) and recent growth index (RGI) of Pseudocaranx and Trachurus, confirming the size independence. 74

Figure 3.11. Mean recent growth index (RGI, ±SE) for each region and station for a) Pseudocaranx and b) Trachurus. Sample size is indicated above each column. 74

Figure 3.12. Scatterplot and regression statistics for Pseudocaranx and Trachurus of δ15N on the average length of larvae that were pooled for each sample (1-10 larvae per sample) for stable isotope analysis. 76

Figure 3.13. Average δ13C and δ15N from a) Pseudocaranx and b) Trachurus, at 50 m and 100 m station of regions 1, 3 and 4 (error bars are standard error). 77

Figure 4.1. Two sampling regions of ichthyoplankton survey in November 1998, Diamond Head and Port Hacking, on the east coast of Australia. Two stars ( ** ) indicate 50 m and 100 m contour stations. 89

Figure 4.2. Section plots of temperature from CTD casts of a transect across the continental shelf at a) Diamond Head and b) Port Hacking. Contour intervals are at 2ºC, except 21ºC indicated with a dashed line. Note the domed and upward sloping isotherms in a) and the flat isotherms in b). Arrowed stations mark the 50 m (inshore) and 100 m (offshore) stations. 90

xiv

Figure 4.3. Temperature-salinity plots (a, c) and the corresponding chlorophyll a depth distribution plots (c, d) during contrasting conditions at the 50 m and 100 m stations off Diamond Head (DH) and Port Hacking (PH) during a, b) early in the cruise (14 November, Port Hacking and 20 November, Diamond Head) and c,d) late in cruise (24 November, Diamond Head and 26 November, Port Hacking). The larval fish samples were taken and analysed from 24, 26 Nov. Boxes in d) indicate the depth range of net samples taken; S, surface; UML, upper mixed layer; LML, lower mixed layer; Cmax, chlorophyll maximum; DEEP, deep sub-thermocline layer. Dashed lines indicate the approximate location of the thermocline. 91

Figure 4.4. Bar charts for the inshore station (a, c, e) and offshore station (b, d, f), at Diamond Head (DH) and Port Hacking (PH), of a, b) the zooplankton displacement volume retained on 500 µm mesh (ml 100 m- 3); c, d) the total concentration of larval fish (per 100 m3, Table 4.1a) and e, f) the Shannon-Weiner diversity index (Table 4.1b). Error bars are standard error; significant SNK test post-hoc comparisons where possible are labelled with a different letter. Depth of sampling: S, surface; UML, upper mixed layer; LML, lower mixed layer; Cmax, chlorophyll maximum; DEEP, deep sub-thermocline layer. 92

Figure 4.5. a) cluster dendrogram and b) MDS ordination showing the similarity groups (dotted line and ellipses at 60% Bray-Curtis similarity index) of ichthyoplankton collected in late November. DH, Diamond Head; PH, Port Hacking; i, inshore (50 m station); o, offshore (100 m station), at S, surface; UML, upper mixed layer; LML, lower mixed layer; Cmax, chlorophyll maximum; DEEP, deep sub-thermocline layer. 96

Figure 4.6. Percentage length frequency distribution of a) Pseudocaranx dentex n=527 and b) Trachurus novaezelandiae n=342 at 50-m-isobath and 100-m-isobath stations off Diamond Head and Port Hacking in late November. SL (standard length in mm) and standard error are given for each sample group. Dashed lines at 3 and 5 mm are for comparison. Results of KS comparisons are in Table 4.4. 98

Figure 4.7. Relationships between SL (back-calculated to the end of outermost complete increments; mm) and the back calculated recent growth over days 1 and 2 pre-capture [RG-1&2 (mm. d-1)], for Pseudocaranx dentex for the surface and upper mixed layer nets combined, at a, c) 50-m-isobath and b, d) 100-m-isobath stations off a, b) Diamond Head and c, d) Port Hacking. See Table 4.5a for the ANCOVA. 99

Figure 4.8. Relationships between SL (back-calculated to the end of outermost complete increments; mm) and the back calculated recent growth over days 1 and 2 pre-capture [RG-1&2 (mm. d-1)], for Trachurus novaezelandiae for the surface and upper mixed layer nets combined, at a, c) 50-m-isobath and b, d) 100-m-isobath stations off a, xv

b) Diamond Head and c, d) Port Hacking. See Table 4.5b for the ANCOVA. 100

Figure 4.9. The average RGI a) Pseudocaranx dentex and b) Trachurus novaezelandiae at 50 m and 100 m contour stations off Diamond Head (DH) and Port Hacking (PH). Error bar is S.E. mean. Sample size is indicated; significant SNK test post-hoc comparisons are labeled with a different letter. 101

Figure 4.10. Average δ13C and δ15N from (a) Pseudocaranx dentex and (b) Trachurus novaezelandiae at i, inshore (50-m-isobath) and o, offshore (100-m-isobath) stations off Diamond Head (DH) and Port Hacking (PH) regions. Grey symbols have the average stable isotope signature recorded 20-60 km north of Diamond Head (Chapter 3). 102

Chapter 1

Introduction

1.1 Recruitment of fishes Fluctuation in the abundance of marine fish stocks due to variable recruitment has been a fundamental concern of fishery science for nearly 100 years. Dramatic examples of such fluctuations in the Pacific include the Japanese sardine (Watanabe 2002) and the Pacific hake, sardine, and northern anchovy in the northeast Pacific Ocean (Ware and Thomson 1991), coral reef fishes (Doherty and Fowler 1994), and the effect of the Leeuwin Current (Caputi et al. 1996). Variable recruitment is composed of many factors in the early life history of fish, including egg production and egg or larval mortality (e.g. Bailey and Houde 1989, Cushing 1996), the growth and condition of fish larvae (Houde 1987, Heath 1992), and oceanographic features (Leiby 1984, Kingsford 1990, Caputi et al. 1996), temperature (Theilacker 1986, Heath 1992), wind (Taggart and Leggett 1987, Cury and Roy 1989, Sundby and Fossum 1990), tide (Sinclair and Iles 1985, Muelbert et al. 1994), estuarine fronts and plumes (Bruce and Short 1992, Kingsford and Suthers 1994), turbulence (Rothchild and Obsorn 1988) and the contact rate of predators and prey (Rothchild and Obsorn 1988, Dower et al. 1997, Reiss et al. 2002). The recruitment mechanism may be summarised as three processes, composed of the starvation or predation of larval fish and the role of oceanographic events (Heath 1992). In recent decades, these three processes have been formalized by a number of testable hypotheses (or purportedly testable), which include: • The larval retention hypothesis, and the member/vagrant hypothesis (Iles and Sinclair 1981, Sinclair 1988), which suggests that oceanographic boundaries and not larval food limit the abundance and stock richness of fisheries; • The match/mis-match hypothesis (Cushing 1990), which argues that the occurrence of prey for larval fish may vary with respect to the timing of spawning and abundance of larvae; • The larval predation hypothesis (Bailey and Houde 1989, Oiestad 1985, Purcell 1984), which argues that the abundance of predators regulates the abundance of larvae; • The stable ocean hypothesis (Lasker 1981, Peterson and Wroblewski 1984), 2

which argues that excessive wind mixing can dilute patches of plankton, particularly around the thermocline. The related turbulence and encounter hypotheses argue that some wind mixing is necessary to enhance the contract rate between larvae and their prey (Rothschild and Osborn 1986, McKenzie and Leggett 1991); • The bigger-is-better hypothesis (Miller et al. 1992), and the related growth or stage duration hypotheses (Anderson 1988, Houde 1987, Leggett and Deblois 1994; Cowan and Shaw 2002) which argue that size at the end of the growing season and therefore growth during the high mortality larval phase enhances survival.

Clearly, the larval fish stage leads to the most influential period of the life cycle at the juvenile stage. Mortality during this stage is high, such that over 99% of eggs fail to survive to adulthood (May 1974). Prey or food availability and the initiation of feeding is important when the yolk supplies are completely absorbed. Potential predators include krill, gelatinous zooplankton (ctenophores and cnidarians), chaetognaths and other fish larvae and planktotrophic fish (e.g. Purcell 1981, 1984, 1985, Jenkins 1988, Houde 2002). Predation can be quick, and spatially or temporally patchy, making it a difficult process to quantify. Other biological factors contributing to the mortality rate are genetic influences, disease and parasites (May 1974, Houde 1987).

1.2 Starvation and predation

Starvation is an important factor causing high mortality rate of larval fishes (Hewitt et al. 1985, Lasker 1985, Anderson 1988). Immediately, after yolk-sac reabsorption, fish larvae face a 'critical feeding period' (Anderson 1988, Doherty and Williams 1988) when they have to find a suitable planktonic food and subsequently achieve feeding success. Larval fishes feed on zooplankton, especially copepods and their nauplii (Poulet and Williams 1991), and their abundance depends on the primary (phytoplankton) production which is related to light and nutrient levels (Tremblay and Roff 1983, Thomas and Emery 1986, Fernandez et al. 1993). Fish larvae actively feed during daytime and are inactive during the night (Blaxter and Hunter 1982), such as 3

tuna larvae (Young and Davis 1990), jack mackerel larvae (Young and Davis 1992), sardine, anchovy and mackerel of the Californian coast (Arthur 1976) and several larval fish species in the North Sea (Last 1980). Larger fish larvae can swim more rapidly and thus increase their searching ability (Houde and Schekter 1980), with the feeding ratio (mean number of prey organisms per stomach examined) increasing with size (Heath 1992). Larger larvae are in better condition than smaller larvae, due to their proportionally better swimming ability and therefore their ability in food searching and feeding is also enhanced.

Prey abundance and prey type can influence growth, condition and ultimately larval survival. Prey type varies with larval size, season, geography and species (de Mendiola 1974, Young and Davis 1992, Bailey et al. 1995, Fossum 1996, Rissik and Suthers 1996). For example, the main food of larval anchoveta, Engraulis ringens, was zooplankton, however, phytoplankton was also discovered in small amounts in the gut of larvae which had length less than 9 mm (de Mendiola 1974). Larval jack mackerel, Trachurus declivis, from eastern Tasmanian coastal waters, ate mostly crustacean microzooplankton during summer (Young and Davis 1992). The dominant diet included harpaticoids, cyclopoids, calanoids and the calyptopis stage of the euphausiid (Nyctiphanes australis). However, the variation in diet between years was positively correlated to the composition of plankton in the sampling area. Larvae of length ≤6 mm tended to feed on copepod nauplii, whereas all larvae consumed Microsetella rosea, cyclopoids and N. australis (in calyptopis stage) (Young and Davis 1992).

Two important studies were the first to show evidence of starvation mortality of larval fishes in marine ecosystems. O'Connell (1980) found that 8% of northern anchovy (Engraulis mordax) in the Southern California Bight showed signs of starvation. Theilacker (1986) found nearly 70% of jack mackerel (Trachurus symmetricus), collected 350 km off the coast California in the first feeding stage were found to be starving. Thus, apparently, starvation can be a major problem for larvae in early stage, but less so for late larvae or juveniles (Powell and Chester 1985). In later larval life (e.g. pelagic juvenile stage), starvation mortality may therefore be less prevalent, but slower growth from food limitation can increase the risk of predation mortality (see below). Starvation also varies between species (Powell and Chester 1985). A variety of factors such as larval size, the degree of parental investment (e.g. 4

size and quality of the yolk-sac) and other physiological aspects may influence variation in larval starvation (e.g. Purcell 1985, Houde, 1987, 2002).

There are several organisms which are identified as larval fish predators including copepods and isopods, chaetognaths, medusae, ctenophores and planktivorous fish (Hunter 1981, Jenkins 1984, Purcell 1981, 1985, Hewitt et al. 1985, Leak and Houde 1987), and are the main source of mortality in jack mackerel (Trachurus symmetricus, Hewitt et al. 1985). Another major source of predation is from cannibalism (Hewitt et al. 1985, Leak and Houde 1987, Anderson 1988). It is difficult to investigate predation due to a wide variety of predators, rapid digestion of the larvae, and difficulty in measuring predators in the field (Anderson 1988). In addition, predation pressure of fish larvae can decrease if alternate prey densities increased (Frank and Leggett 1981; Pepin et al. 1987, Anderson 1988). Therefore, the abundance of fish larvae together with alternative planktonic prey items can result in reduced mortality rates of fish larvae. Predation rates also decrease with size (Peterson and Wroblewski 1984, Folkvord and Hunter 1986, Anderson 1988). As larvae grow, they can improve their escape response, and thus avoid predators. Significantly, starved larvae would suffer increased predation induced mortality rates, due to being weaker (Shepherd and Cushing 1980, Hewitt et al. 1985).

1.3 Growth and condition of larvae

The condition of a fish larva indicates its health or nutritional status (Ferron and Leggett 1994, Suthers 1998), and can be used to predict the survival rate of larval fish and potential recruitment to the fishery (Ferron and Leggett 1994, Bailey et al. 1995, Houde 1996). It can also be applied to aquaculture and environmental monitoring (Suthers and Sundby 1996, Suthers et al. 2000). Condition indices can be classified ecologically rather than by technique into four broad groups; morphometric indices, storage indices, starvation indices and growth indices.

Morphometric characteristics are body part measurements including eye diameter, body depth at the pectoral fin, body depth at the anus, and ratios among them to remove the effect of size (Theilacker 1986). To use this technique, adjustment for shrinkage is often necessary since fish larvae shrink with handling and preservation 5

(Theilacker 1986). Morphometric indices are the least sensitive to food levels, exhibiting sensitivity to starvation of at least one week.

Storage indices usually refer to the concentration of lipids, representing an energy store and resistance to starvation (e.g. Zenitani 1995, Bell and Sargent 1996, Booth and Alquezar 2000). Storage lipids are typically triacylglycerol (TAG), which can be adjusted for larval size by the concentration of structural lipids (sterols; e.g. Suthers et al. 1992)

Growth is typically assessed by biochemical indices (RNA:DNA ratio), or by otolith growth. RNA (ribonucleic acid), the precursor to protein synthesis is adjusted amongst different sized larva by the number of cells, indicated the amount of DNA (deoxyribonucleic acid) (Buckley 1984, Bulow 1987, Clemmesen 1988, 1993). The RNA-DNA ratio can provide an indicator of growth rate of fish larvae, because protein synthesis is proportional to the amount of RNA in the cell (Clemmesen 1993). It has been shown that fed larvae have a higher RNA-DNA ratio than starved larvae within 1- 3 days of a starvation event (Clemmesen 1988), and that larger larvae will have a higher ratio as they are usually growing faster (i.e. the ratio is size dependent).

The otoliths or earstones of fish (sagittae or lapilli, but rarely the asterisci, Stevenson and Campana 1992), are widely used for daily or annual age determination. The daily growth increments in larval sagittae or lapillae can be used to back-calculate growth trajectories (e.g. Jones 1986, Suthers et al. 1989, Jenkins and Davis 1990, Jordan 1994, Jones 2002). The recent daily growth increments (1-3 days prior to capture) can be compared with the RNA:DNA ratio (Clemmesen and Doan 1992), and with environmental conditions at capture (Suthers 1996). Recent growth increments have the additional advantage in distinguishing whether an intermediate growth rate is on an increasing or decreasing trajectory (whereas an intermediate RNA:DNA ratio cannot be interpreted). Recent otolith growth (ROG) has been used in a number of studies to demonstrate the effect of temperature (Campana and Hurley 1989, Suthers and Sundby 1993, Meekan et al. 2003), water temperature and light intensity (Tandler and Mason 1983, Batty 1987, Suthers and Sundby 1996, Saka et al. 2002), genetic characteristic (Sogard 1991, Suthers et al. 1999), dissolved oxygen (Sepulveda 1994), food densities (Wyatt 1972, Houde 1975, Tandler and Mason 1983, Suthers 1996), and upwelling (Peterson 1996). The ROG and RNA:DNA of pelagic juvenile Diaphus 6

kapalae (Myctophidae), was enhanced in an island wake within the Coral Sea (Suthers 1996). There was no significant difference in ROG more than a week pre-capture, but ROG was correlated with prey abundance in the form of microzooplankton biomass (Suthers 1996).

Growth indices do not directly measure starvation – the only direct measure has been the height of midgut, mucosal cells in relation to nutritional status. Evidence from a rearing experiment on larval northern anchovies (Engraulis mordax) suggested that the midgut cells height declined from the fed to starved condition (Theilacker 1989). The advantages of this technique include the fact that mucosal cells are resistant to collection and handling process (Theilacker and Watanabe 1989), but require training and interpretation by the operator.

1.4 Stable isotope signatures

Tracing food sources of ichthyoplankton can be achieved by using stable isotope signatures (Owens 1987, Fry 1988, Hobson and Welch 1992, Hansson et al. 1997). Ratios of the heavy and light isotopes in larval tissue are examined using an isotope mass spectrometer, and compared to a standard. Stable carbon (13C/12C), nitrogen (15N/14N), and sulfur (34S/32S) isotopes are commonly applied to trace nutrient and organic matter sources (Owens 1987, Fry 1988, Hobson and Welch 1992, Hansson et al. 1997) as well as to determine polluted and unpolluted areas (Spies et al. 1989, van Dover et al. 1992, Thornton and McManus 1994).

Stable carbon (13C) and nitrogen (15N) isotopes are frequently used to determine the trophic level within a food chain (Owen 1987, Peterson and Howarth 1987, Fry 1988, Waldron et al. 2001). Generally, consumers are slightly more enriched in 13C and 15N than their diet by approximately 1-2‰ of δ13C and 3-5‰ of δ15N. Therefore, animals at higher trophic levels (predators) have an enriched stable isotope composition (de Niro and Epstein 1978, 1981, Rau et al. 1983, Fry and Sherr 1984, Migawa and Wada 1984). In addition, the δ13C composition in fish tissues of several different species may show a positive relationship (Kiriluk et al. 1995, Beaudoin et al. 1999) or a negative relationship (Sholto-Douglas et al. 1991) with their size, however there was no significant relationship with age (Hobson and Clark 1992). By using 15N composition, Japanese anchovy (Engraulis japonicus) changed in trophic position by half a trophic 7

level for larvae between 14-30 mm standard length, and again for larvae 30-80 mm (Lindsay et al. 1998). Japanese sardine (Sardinops melanostica) increased even more rapidly in trophic position than Japanese anchovy (Lindsay et al. 1998). δ15N is more variable than δ13C and may depend on environmental characteristics, such as salinity, depth and temperature (Jennings and Warr 2003). In addition, trophic shift in early life history stage of fishes may also correspond to their prey (van der Zanden 1998, Herzka and Holt 2000, Herzka et al. 2001). For instance, larvae of red drum (Sciaenops ocellatus) changed their isotopic composition in δ13C and δ15N in 15 days, after migrating from the larval, coastal area to settlement in shallow seagrass, indicating their dietary switch (Herzka et al. 2001). Larval of smallmouth bass (Micropterus dolomieu) from Lake Opeongo (Ontario) show a change in the values of δ15N from larval to juvenile stages indicating a dietary shift and an increase in trophic level of fish with size between 17 and 46 mm. The change in δ13C values from small larvae (<15 mm; -23.2 to –26.1‰) to adult fishes of smallmouth bass (-25.0 to –28.4‰) reflects a dietary shift from a mix of benthic and pelagic prey to benthic food items (Van der Zanden et al. 1998).

Stable isotopes may also determine nutrient or food sources, such as from polluted or unpolluted areas, and from upwelling or sewage impacted areas (Rau et al. 1981, Spies et al. 1989, Waldron et al. 2001). The δ13C values of plankton shows a significant difference between the sewage impacted and unimpacted areas, while the values of δ15N depended on the level of sewage treatment (Gaston and Suthers 2004). The relative ranges for stable isotope ratios of carbon and nitrogen (Table 1.1) in marine particulate matter are different to those in sewage impacted areas and can be exploited to determine the assimilation of anthropogenic nutrients.

Considerable challenges remain in the use of stable isotopes in marine fish larvae, for the reconstruction of food webs and particularly the calibration of turnover and depuration of stable isotope signatures. The prey species’ signatures are unknown, as obtaining sufficient quantities of nauplii or particular copepod species is difficult. Fractionation between diet and tissue composition is now recognized as being diet- and species-specific, making it difficult to relate signatures to pelagic food chains without a laboratory calibration (Bosley et al. 2002). 8

It is still difficult identifying tropic pathways from stable isotope analyses. For example, determining the relative importance of prey sources to the predators is obscured if the multiple food sources have the similar isotopic signature, or when multiple prey types are present, there is no unique solution indicated from the tracer information.

Table 1.1 Relative stable isotopes abundance (‰) in marine and sewage particulate organic matter (after Spies et al. 1989).

Isotope ratio Marine Sewage Difference

δ13C -19 to -21 -16.5 2.5 to 4.5

δ 15N 8 to 12 1.8 -6.2 to -10.2

1.5 Biological oceanography of eastern Australia

A range of oceanographic processes affect the life cycle of marine organisms, particularly fish larvae. Currents can take the larvae to and from the shore (Leiby 1984), or upwelling may take the nutrient-rich waters from the subsurface to the surface, and the nutrients can stimulate a phytoplankton. Therefore, nutrient supply is an important and limiting factor in the abundance and distribution of larval fishes, zooplankton and phytoplankton. Enhancement of nutrient levels could be determined by upwelling events (e.g. Rochford 1984, Creswell 1994), river flow (Govani and Grimes 1992, Grimes and Kingsford 1996) and urban discharge (Rendell and Pritchard 1996). The East Australian Current (EAC) flows poleward from the south Coral Sea along the NSW coast. The current is strong and narrow (Wyrtky 1960), but during winter, a countercurrent may flow northward further offshore due to frequently strong blowing of southerly winds. The average speed of the strongest EAC is about 50 cm s-1 which can be detected off Byron Bay (28.5°S) from December to April, whereas during the rest of the year, it is <30 cm s-1 (Wyrtky, 1960). The EAC tends to flow across the 9

Tasman Sea at 32°S, or dissipate as eddy system towards Tasmania (Cresswell 1983). The EAC eddy field or meanders off coast of NSW can take the water mass from the continental slope into the shelf (Tranter et al. 1986).

Upwelling events in NSW waters have been reported to occur in spring and summer (Rochford 1975), when the East Australian Current (EAC) flows stronger and departs the coast (Jeffrey et al. 1990, Griffin and Middleton 1992, Gibbs et al. 1998). The area off between Smoky Cape and Laurieton (31-32°S) on the northern NSW coast has been a focus for upwelling investigations (Rochford 1975, Godfrey et al. 1980, Cresswell 1983, Roughan and Middleton 2002). Several upwelling sites from Port Stephens to the Queensland’s border have also been detected in spring and early summer (Rochford 1984, Jeffrey et al. 1990).

When upwelling occurs, the nutrient-rich cold water from below the thermocline is transported to the surface, increasing near-surface nutrients, especially nitrates (Hutchings et al. 1994). Rochford (1975 ) reported that nutrient level off Laurieton during the upwelling events reached 6 μg atom L-1. The concentration of nutrients caused by the upwelling events can increase approximately 2-4 times higher than nutrients obtained from the river supply and from urban discharge (Anderson 1994). Nutrients from off Laurieton may affect the nutrient concentration in the Sydney region Cresswell (1994), and phytoplankton blooms generated here are certainly transported to off Sydney (Dela-Cruz et al. 2003).

The nutrient level off Sydney can also be influenced by the sewage discharge (Anderson 1994, Manly Hydraulics Laboratory 1997, Pritchard et al. 2001), when there is no upwelling event, because river flows supply a small amount of nutrients, particularly if rainfall is low (Rochford 1984, Sinden 1996). The Sydney region has an annual average of nitrate concentrations less than 1 μg-atom L-1 (14 μg nitrate-N L-1) (Rendell and Pritchard 1996). However, the data collected in March-June 1989 and 1990 show that the average total nitrogen of sewage outfalls was ranging 119-198 μg L-

1 (SPCC 1991). Sydney waters usually has relatively higher nutrient concentration than the other areas in NSW, because 12 of the 35 ocean outfalls are situated in the Sydney water area (Manly Hydraulics Laboratory 1997). Also, three major outfalls (North Head, Bondi and Malabar) can release nearly 1000 megalitres of primary treated sewage daily (Manly Hydraulics Laboratory 1997). Therefore, the Sydney region can 10

be a representative sewage impacted area, and the different level of nutrients between Sydney and the other regions in NSW may explain the variation in the quantity and quality of prey for fish larvae.

The nutrient-rich upwelled waters are positively correlated with a bloom of phytoplankton (Hallegraeff and Jeffrey 1993, Chaves and Smith 1994) and abundance of zooplankton (Hutchings et al. 1994). Generally, the response of phytoplankton to the nutrients is faster than the response of zooplankton to the phytoplankton (Hutchings et al. 1994, Mann and Lazier 1996). Also, the growth of nauplii is faster in warmer and low latitudes than in cooler and high latitudes (Steidinger and Walker 1984). The development and decrease of a phytoplankton bloom in an upwelling area can take between 6-12 days (Hutchings et al. 1994), whereas the growth of nauplii or early copepodites to adulthood at sea temperature between 11 and 18°C may occur within 14 to 40 days (Hutchings et al. 1994, Mann and Lazier 1996). Development from nauplii to adults in Acartia tonsa in temperate waters can take 7-12 days, whereas Pseudocalanus minutus takes 35 days at 8°C and 21 days at 22°C (Steidinger and Walker 1984).

1.6 Ichthyoplankton research in Australia

Ichthyoplankton including egg and larval fish has been studied progressively in Australian waters for the last two decades (Miskiewicz 1992, Table 1.2). The studies cover wide-ranging aspects of ichthyoplankton in the marine environment, from inshore (e.g. Jenkins 1986, Steffe and Pease 1987, Gray and Miskiewicz 2000) to offshore (e.g. Dempster et al. 1997, Smith et al. 1999), and from different habitats, such as estuaries (e.g. Neira et al. 1992, Newton 1996, Trnski 2001), to seagrass beds (e.g. Bell et al. 1987, Worthington et al. 1991, Smith and Suthers 2000), and to coral reefs (e.g. Leis and Goldman 1987, Williams et al. 1988, Milicich and Doherty 1994).

Table 1.2 Summary of major findings on larval fish studies of species in Australian waters.

Family Species Common Study Major findings Source name area Berycidae Centroberyx Nanygay The Larvae transported passively, and Smith 2000 affinis Sydney they distribute reflecting Shelf, hydrography. Spawning occurs in NSW the inner shelf. Carangidae Trachurus Jack mackerel Coastal The main food is crustacean micro- Young & declivis waters of zooplankton. Feeding occurs Davis 1992, eastern mostly during the daytime hours Jordan 1994 Tasmania with peaks in mid-morning and late afternoon. Otolith rings formed daily and larvae grew exponentially in 1989-1991. Spawning occurs during summer with peaks related to full and new moons cycles. Centrolophidae Seriolella Blue warehou South Larvae distribute from Kangaroo Bruce et al. brama eastern Island (SA) to southern NSW in 2001a Australia winter and spring within shelf and slope waters. Small larvae were none from western Tasmania to southern NSW. Western and eastern stocks of spawning were detected Seriolella Spotted South Larvae distribute from western Bruce et al. punctata warehou eastern Tasmania to southern NSW in 2001a Australia winter and spring within shelf and slope waters. Consistent low numbers of larvae were caught between western Tasmania and southern NSW. Unclear spawning stocks were suggested. 12

Table 1.2 (Continued)

Family Species Common Study Major findings Source name area Cheilodactylidae Nemadactylus Jackass The Larvae were found within Bruce et al. macropterus morwong southwest water masses originated 2001b ern from the East Australian Tasman Current and Sub-Tropical Sea Convergence Zone. Two major of the spawning stocks were detected, northern and southern regions. Clupeidae Hyperlophus Sandy sprat Off Spawning occurs near Gaughan et vittatus south- surface between may and al. 1996, western September with peaks in Tregonning Australia June and July. Eggs are et al. 1996. spherical with diameter ranging 0.83-0.95 mm. Eggs hatch in 2-3 days. The yolk-sac larvae are 2.6 mm in body length. Sardinops sagax Pilchard The shelf Larvae and eggs were Fletcher et region of spread out over the shelf of al. 1994, southern southern Australia region Fletcher & & SW due to the Leewin current. Tregonning Australia Spawned eggs were mostly 1992, abundant at depth 40-60m, Fletcher but older eggs were found 1999, abundantly near surface. Gaughan et Larval caught was in large al. 2001a, number at surface during 2001b; the daytime, while at night, larvae were in patches. Caputi et al. Larval up to 30 days had 1996 growth rate maximum, 0.63 µm day-1 with growth mean was 0.48 µm day-1. Gonorynchidae Gonorynchus Beaked The Larval transported Smith 2000 greyi salmon Sydney passively. Small larvae shelf, distribute partially, and NSW spawning occurs over the outer shelf. 13

Table 1.2 (Continued)

Family Species Common Study Major findings Source name area Merlucciidae Macruronus Blue grenadier Southern Variations in mean size and Thresher et novaezelandiae and age among sampling sites al. 1989, southeast of southern Australia. Bruce et al. ern Larvae transport passively 2001c Australia following alongshore current from the spawning area in the west coast to the southeast and east coast of Tasmania. Small numbers of larvae were found in eastern Victoria and southern NSW indicating a spawning area in southeastern Australia. Larvae of eastern Tasmania had wider otolith increment than from southeastern Australia Platycephalidae Platycephalus Flathead Tasmania Spawning occurs in coastal Jordan 2001 bassensis n waters zone of eastern Tasmania for up to six months. Larvae concentrated in mid waters and inshore due to the predominant subsurface current onshore. Pleuronectidae Ammotretis rosratus Long-snout Port The main food of small Jenkins flounder Phillip larvae was bivalve veliger, 1987a, Bay, VIC and large larvae fed mostly 1987b. paracalanid copepodids. Growth rate of larvae was 0.12-0.28 µm day-1. Rhombosolea Green back Port Otolith increment was Jenkins tapirina flounder Phillip deposited daily. Bivalve 1987a, Bay, vileger was the main prey 1987b, Victoria of small larvae, and Steward & Cladoceran Evadne Jenkins mordmanni loven was the 1991 main prey of large larvae. Growth rate of larvae (in pre-flexion) was 0.10-0.23 µm day-1. 14

Table 1.2 (Continued)

Family Species Common Study Major findings Source name area Pomacentridae Acanthochromis Damselfish One Tree Fed and unfed experiments Booth & polyacanthus Island of larvae over 20 days Alquezar (GBR) show that 60% and 48% in 2000 survival rate, respectively Larval density and their body size were independent factors to growth, condition and survival. Pomacentrus Damselfish Lizard Time of recruitment Milicich et amboinensis Island, correlates strongly to time al. 1992, northern of reproduction, however Meekan et GBR planktonic life determines al. 1993, the magnitude of these Kerrigan events. The leeward 1997, habitat was preferred to recruit. Pomatomidae Pomatomus Tailor Off Larvae were most abundant Miskiewicz saltatrix eastern on the northern NSW coast. et al. 1996 Australia Spawning may occur along the NSW coast from August to May. Scombridae Katsuwonus Skipjack tuna The east Greater larval numbers Davis et al. pelamis Indian were captured at night, but 1990, Ocean found abundantly in deeper Young & water during the day. The Davis 1990 main preys were appendicularians and fish larvae. Scomberomorus Spanish Great Larvaceans were the Jenkins et commerson mackerel Barrier dominant diet items of al. 1984 Reef small larvae. Fish larvae were also found significantly in the stomachs, while eggs and a copepods nauplius were found rarely. Scomberomorus School Great Larval prey of this species Jenkins et queenslandicus mackerel Barrier varied taxonomically, and al. 1984 Reef they were changed following the ontogenetic development. Larvaceans were found frequently in the stomachs. 15

Table 1.2 (Continued)

Family Species Common Study Major findings Source name area Scombridae Scomberomorus Broadbanded Great Larval diet was mostly on Jenkins et semifasciatus mackerel Barrier larval fish, and clupeid al. 1984 Reef larvae contributed 70% of identified larval fish prey. Thunnus Albacore The east Larvae move into surface Young & allalunga Indian layer during the day. Davis 1990, Ocean Feeding is only in the day Davis et al. with peaks in the early 1991a, morning and late afternoon. 1991b, Jenkins et al. 1991 Thunnus Southern The east Most abundance and large Davis et al. maccoyii bluefin tuna Indian larvae were found at night; 1990, Ocean Patch density reached 22 Davis et al. ind. m-3; Otolith 1991a, increments formed daily; 1991b, Growth rate varied; Food Jenkins et limitation was with a al. 1991, varying degree and density Young & dependent; Main preys Davis 1990 include copepod nauplii, calanoids, cyclopoids and cladocerans (all crustacean). Sillaginidae Sillaginodes King George Port Pre-settlement larvae were Bruce 1995, punctata whiting Phillip abundant near surface Jenkins et Bay, during the day and al. 1998, Australia scattered through the water 2000. column at night. Relationship between larvae and prey distribution is poor. Transport of pre- settlement larvae to seagrass beds could be related to diurnal and tidal migration. Otolith microstructure and hydrodynamic modeling suggested that predicted spawning area was about 400-5000 mile from recruitment sites in west and south of South Australia. Variability of larval supply may relate to variability in weather patterns. 16

Table 1.2 (Continued)

Family Species Common Study Major findings Source name area Sparidae Acanthopagrus Black bream Hopkins Calanoid copepod nauplii Willis et al. butcheri River were the main prey of small 1999 Estuary larvae (<9 mm), while large VIC larvae (>9mm) fed primarily calanoid copepodites and unidentified larvae.

Most studies of ichthyoplankton were in the coastal area (including estuaries) of western and south Australia, Great Barrier Reef, and south-eastern Australia regions, and fewer were conducted in Victorian and Tasmanian regions. In western and south Australia, most studies were carried out in estuaries (Gaughan et al. 1990, Neira et al. 1992, Neira and Potter 1992, 1994). On the Great Barrier Reef , larval studies tend to focus on larval distribution and settlement of post-larval stages of reef fish species (e.g. Leis and Goldman 1987, Ohman et al. 1998, Booth 2002a). In south-eastern region, observations on abundance and distribution of ichthyoplankton were mostly to be linked to the oceanographic features (e.g. upwelling, Smith and Suthers 1999; estuarine front, Kingsford and Suthers 1996) and sewage plumes (Gray et al. 1992, 1997, Gray and Kingsford 2003). In Victorian and Tasmanian regions, larval studies were on certain species, for example, King George whiting in Port Phillip Bay (Jenkins et al. 1998, 2000, Fowler et al. 2000), jack mackerel (Young and Davis 1992) blue grenadier, (Bruce 1988, Thresher et al. 1989, Bruce et al. 2001c) and sand flathead (Jordan 2001) in Tasmanian waters. In addition, feeding behaviour (e.g. Jenkins 1987, Shaw and Jenkins 1992, Young and Davis 1992, Rissik and Suthers 1996; Booth and Hixon 1999, Jordan 2001) and growth and condition of fish larvae (e.g. Jenkins 1987, Jenkins and Davis 1990, Jenkins et al. 1991) are documented in Australian waters. Recently larval swimming ability of coral reef and temperate fishes has come under increasing investigation (Leis and Carson-Ewart 1997, Fisher et al. 2000), in relation to marine parks and source-sinks of propagules.

Larval transport along and cross-shore is of international interest (e.g. Reiss 2000, Epifanio and Garvine 2001, Hare et al. 2001, 2002, Bailey and Piqculle 2002), and recent studies off the eastern Australia coast have dealt with similar issues (Dempster 1997, Smith et al. 1999, Smith and Suthers 1999). The EAC and upwelling 17

events are two major oceanographic processes that influence the larval fish community (Dempster et al. 1997, Smith et al. 1999, Smith and Suthers 1999). The study by Smith and Suthers (1999) found a dramatic effect of upwelling, but the comparison was over three months, and had no long-shore sampling. These two oceanographic features may have a significant contribution to other biological aspects, such as feeding and growth patterns of larval fish, and provide the rationale for this study. A significant gap in our knowledge of early life history of fish in Australia waters concerns that of the major pelagic fishes such as herring and carangids, as noted by Young and Davis (1992) and Jordan (1994). These taxa form as important component of coastal ecology, as revealed by the pilchard mass mortality observed off eastern Australia (Neira et al. 2000, Griffen et al. 2002), and because of their abundance in my samples became a focus of this study.

1.7 Aims of this study

Australia’s fisheries are remarkably unproductive, having the world’s 3rd largest fishing zone and yet the 55th largest fishery (by weight, Kailola et al. 1992). This thesis examines the central tenet of nutrient limitation from an ichthyoplankton perspective. There are at least three major nutrient sources along NSW coast; coastal upwelling events, sewage discharge and river runoff. Nutrients from river discharge are particularly sporadic and comparatively trivial. Upwelling events dominate the nutrient budget for the coast (particularly nitrate, Pritchard et al. 1998; 2001), but are sporadic in nature (2-3 times per summer). In contrast the nutrients from sewage are continuously discharged from a number of coastal and deep ocean outfalls off Sydney, with only a small effect on chlorophyll and nutrients evident within 10 km (Pritchard et al. 2001). Sewage is particularly enriched in ammonia which is preferentially assimilated by phytoplankton, although major algal blooms off Sydney are driven by upwelling events (Dela Cruz et al. 2002; 2003). Therefore upwelling and sewage discharge differ in their delivery and composition of nutrients, and could exhibit different effects on the abundance, distribution and growth of fish larvae that occupy the coastal waters of NSW, as a nursery or feeding areas. Thus, there were three objectives to this study.

1. To examine the spatial variation of the ichthyoplankton assemblage of an upwelling region from the East Australian Current. What families (possibly 18

species) are present? Are there patterns and associations amongst the families, and are some associated with upwelling events?

2. To compare the abundance, growth and stable isotope composition of two abundant carangids, Pseudocaranx dentex and Trachurus novaezelandiae between upwelled and non-upwelled regions. Is there evidence of a nutrient limited ecosystem off eastern Australia? What is the effect of upwelling compared to non-upwelled waters? What is the early life history of these two economically important fish?

3. To compare the ichthyoplankton assemblages, and the effects of nutrient injection, on the two larval carangids between an upwelling region (predominantly nitrate) and a region of sewage discharge (predominantly ammonia).

The thesis consists of three core chapters, with a general introduction and a general conclusion. Each chapter is linked to the other, based on the findings during two cruises on the Research Vessel ‘Franklin’ in November 1998 and January 1999.

Chapter 2 examines the larval fish assemblages off the northern coast of NSW, especially in the EAC and within and downstream of two upwelling events. In this chapter the abundance of certain families as tracers of topographically upwelled water is assessed, and the generality of these tracers is tested from an independent wind-induced upwelling event. The role of oceanographic features affecting the variation of ichthyoplankton assemblages is discussed.

Chapter 3 examines in detail the abundance and size gradient of two abundant species, the carangids Pseudocaranx dentex and Trachurus novaezelandiae, to infer their early life history. The impact of coastal upwelling on the growth and condition of these fish larvae is assessed, and their food source is inferred from stable isotope analysis.

Chapter 4 is a comparison of the ichthyoplankton assemblages from two nutrient enriched areas of the NSW coast: an upwelling region versus off the coast of Sydney, which receives over 1,000 ML d-1 of primary treated sewage. The study includes ichthyoplankton assemblage structure, and growth and condition of two carangids (P. dentex and T. novaezelandiae), and their 13C and 15N stable isotope signatures. 19

Chapter 2

Is upwelling marked by signature ichthyoplankton? Larval assemblages as tracers of mixing

Abstract Austral, summer ichthyoplankton assemblages were compared between regions dominated by the East Australian Current (EAC) and those from topographically induced upwelling regions, found on two cruises off eastern Australia in November 1998 and January 1999. Ichthyoplankton was sampled at the surface and in sub-surface waters, at 50 m and 100 m isobath stations. There was no significant difference in the total abundance nor diversity of larval fishes amongst the 111 families, between regions within or upstream of the upwelling. However in both months there were distinctive ichthyoplankton assemblages at the taxonomic level of family. A greater abundance of the Carangidae, Labridae, Lutjanidae, Microcanthidae, Myctophidae and Scombridae was associated with the EAC or oceanic water masses, while the Callionymidae, Clupeidae, Platycephalidae, Sillaginidae and Terapontidae were mostly found in the surface or deep upwelled/uplifted water masses. Correlation coefficients of the abundance of these signature families with their abundance from opportunistic sampling of a separate wind-induced upwelling event, from either month, were consistent with the conventional oceanographic interpretation. Without detailed oceanographic information, the correlation coefficients of family abundance are a biological alternative for identifying the water mass history and mixing. The generality of these family level distinctions can be tested.

Keyword: ichthyoplankton, coastal upwelling, East Australian Current, western boundary current, continental shelf, diversity, distribution, abundance. 20

2.1 Introduction Clear, oligotrophic waters are recognised as generally having low productivity. Such waters bathe northeastern Australia which are derived from the tropical Pacific, forming the Coral Sea and ultimately forming the poleward flowing East Australian Current (EAC). The passage of the EAC along the coast of New South Wales can generate upwelling of deep nutrient-rich waters by either topographic features and/or upwelling favourable winds (Middleton et al. 1997, Roughan and Middleton 2002). The behaviour of this western boundary current has many physical and biological similarities with the Kuroshio extension (Watanabe 2002), the Agulhas Current (Beckley 1993), the Brazil Current (Nonaka et al. 2000) and the Florida Current (Grothues and Cowen 1999, Pitts 1999). Upwellings can vary considerably in their mechanism, magnitude and biological response (Cury and Roy 1989). For example, persistent and strong upwelling can dilute larval abundance and diversity (off western Africa, Hamann et al. 1981; Olivar and Shelton 1993), or oceanographically displace ichthyoplankton assemblages (Frank and Leggett 1982), or where advection may uncouple the biological response (Peterson et al. 1988). In contrast, the EAC upwelling events are typically sporadic, with topographic complexity that retards advection in the nearshore zone (<50 m depth), serving to spatially confine the biological response. Phytoplankton blooms respond within a week to cold (14°C) nutrient-rich intrusions, and be transported southwards (Dela-Cruz 2002, Dela-Cruz et al. 2003). The contribution of estuarine nutrients in this area is less than 5% of that from natural upwelling events (Pritchard et al. 1998). There should therefore, be a dramatic influence of upwelled nutrients on the ichthyoplankton composition, its abundance and success in fish recruitment due to the increased food availability (Houde 1987, Cury and Roy 1989). It is possible that reproduction by some species may even be stimulated by sporadic upwelling events in oligotrophic waters. The response by ichthyoplankton to upwelling in the nutrient limited ecosystem off eastern Australia is unknown, but as coastal waters are physically dynamic, an unconfounded response by ichthyoplankton would be unusual. Local ichthyoplankton assemblages are indeed dynamic, being advected at least at the scales of days and kilometres across the Sydney shelf (Dempster et al. 1997, Smith et al. 1999). By using detailed current meter data to interpret the assemblage variability, these two studies 21

concluded that larval assemblages functioned as a passive tracer of the water mass (although their assumption of minimal along shore variability was untested). Cowen et al. (1993) found differences in hydrography (Salinity and Temperature) accounted for only 15% of the variability in ichthyoplankton assemblages, requiring the use of circulation models to interpret the high variability characteristic of ichthyoplankton studies. Larvae with developing swimming and sensory skills thus provide biological markers of the water mass beyond temperature and salinity, particularly when estimates of advection from a current meter mooring array are unavailable. Inferred currents from hydrography and dynamic height is novel (Reiss et al. 2002) but requires certain conditions and assumptions. This study investigates the use signature taxa, as another method to interpret ichthyoplankton variability. My aims were: 1) to describe the topographically induced upwelling observed on two summer cruises off eastern Australia; 2) to compare the abundance and diversity of ichthyoplankton among regions of upwelled and non-upwelled water; 3) to test the assumption of along-shore homogeneity in community structure (Cowen et al. 1993, Smith et al. 1999), in comparison with variability between the inner and mid-shelf community (i.e. is variance alongshore significantly greater than cross shore?); and, 4) to identify signature taxa of upwelled water. The generality of these signature upwelled taxa was assessed with opportunistic sampling of an earlier wind-induced upwelling event on the first cruise.

2.2 Methods

Study area Episodic upwelling of cold, nutrient-rich waters off eastern Australia is usually attributed to the flow of the East Australian Current (EAC) in conjunction with topographic features and with north-easterly wind stress (Jeffrey et al. 1990, Gibbs et al. 1998, Oke and Middleton 2000, 2001). Upwelling is typically observed between 30- 33°S at the separation point of northern New South Wales (NSW), where the EAC turns to the south east, forming the Tasman Front (Creswell et al. 1983, Cresswell 1994, Dela-Cruz et al. 2003). Several upwelling sites from Port Stephens to Queensland’s 22

border have been detected in spring and early summer (Rochford 1984), occurring episodically 2-3 times per summer, when the EAC flows stronger and upwelling favourable north-easterly winds are more frequent (Griffin and Middleton 1991). Two cruises were conducted from the Research Vessel ‘Franklin’ in November 1998 and January 1999 along a gradient of upwelling zone off the northern coast of NSW in the northwestern Tasman Sea (Fig. 2.1, Table 2.1). The sampling area was between 30.50ºS, 153.25ºE and 32.05ºS, 152.60ºE. Sampling was conducted at oceanographic regions off the town of Urunga, and off Smoky Cape, Point Plomer and Diamond Head (referred hereafter to Regions 1 to 4, numbered in a north to south direction). As the EAC flows from the wider shelf off Region 1 (approximately 30 km to the 200 m isobath), to the narrow shelf off Region 2 (~17 km), it accelerates except near the sea floor (Rochford 1975, Oke and Middleton 2000, 2001). Bottom stress perturbs the geostrophic balance, allowing deep nutrient rich water to encroach across the continental slope and shelf, as the shelf widens to ~22 km (Region 3) and ~32 km (Region 4). In January, additional sampling was also conducted further south off Crowdy Head and South Crowdy Head (close to Region 4) and Cape Hawke (Region 5, Fig. 2.1). Temporary reversal of this topographic upwelling occurred early in the November cruise (before 17 November), when strong upwelling favourable winds temporarily causing uplifting to be observed off Regions 1 and 2 (Roughan and Middleton 2002, Dela-Cruz 2002). We opportunistically sampled this event with the surface nets only (see below) to test our findings. Opportunistic samples were also collected in January in Region 4 off Crowdy Head with surface and deep nets.

Ichthyoplankton sampling Typically, physical oceanographic work was conducted during the day and biological work at night. Biological sampling at each region was conducted at two stations: at 50 m (inshore) and 100 m (offshore) bathymetric contours. These two stations were selected on the basis of the greatest difference in coastal ichthyoplankton observed off Sydney (Smith and Suthers 1999). A neuston net (0.75 x 0.75 m square mouth, 500 μm mesh) and an EZ-net (a multiple depth, opening/closing net with 1 m² mouth and 500 μm mesh, at three depth strata) were deployed to collect ichthyoplankton. Only the upper depth stratum of the EZ-net was sorted and identified: 10-20 m depth stratum at the 50 m station or 10-30 m at the 100 m station. The neuston and sub-surface nets are 23

hereafter referred to as “surface” and “deep” respectively. These nets were operated from the vessel at 2-3 knots for 5-10 minutes, and a General Oceanic flow meter indicated the volume of water filtered (200 to 400 m3). All sampling was done at night to minimise net avoidance. At each station, 4 surface and 2 deep tows were conducted (Table 2.1). Repeated within and between night sampling for some regions was also conducted depending on cruise schedule and weather (see below - Analysis). A regularly calibrated thermosalinograph recorded temperature, salinity and fluoresence continuously from the engine intake at 4 m depth. At each station before plankton tows, a CTD cast with a Neil-Brown Conductivity-Temperature-Depth was conducted, or due to logistical constraints, we used the CTD cast at the station earlier in the daytime. Fluoresence was measured with a WetLab fluorometer, which was converted to chlorophyll a concentration (Chl a, μg.L-1) using a regression determined by Dela Cruz (2002, Chl a = (0.1*fluoresence) + 0.2, r2=0.76, n=62). Niskin bottles were used to collect water samples for nutrient analyses of oxidised nitrogen (NOx), free reactive phosphorous (FRP) and dissolved reactive silicate (Si), using an autoanalyser (detection limits of 0.04, 0.02 and 0.12 μM for NOx, FRP and Si respectively). Due to logistical constraints, we do not have nutrient data for the latter part of the November cruise, when we were able to deploy both the plankton nets (current meter mooring gear on the deck precluded deploying the multiple opening-closing net). 24

a) November 1998 b) January 1999

1) Urunga surface 1) Urunga deep

2) Smoky Cape -31.0 -31.0 m m

0 0

0 0 2 2 3) Point Plomer 3) Point Plomer

Latitude °S

4) Diamond Head 4) Diamond Head, Crowdy Head

#ind./100m3 -32.0 1-20 -32.0 21-50 5) Cape 51-100 Hawke 101-200

> 200 152.0 153.0 152.0 153.0 Longitude °E Longitude °E

Figure 2.1. Contours of sea surface temperature (SST), derived from satellite images of the northern NSW coast taken by NOAA 14 on a) 21 November, 1998, and b) 19 January, 1999. The sampling regions are numbered 1-5, as referenced in the text and Table 2.1. Expanding symbols show the average total larval abundance for the surface and sub-surface nets, at the approximate locations of the inshore and offshore stations.

Table 2.1. Summary of sampling regions numbered from north to south, location name, stations (inshore, 50 m isobath; offshore, 100 m isobath), latitude, longitude and sampling date in November 1998 and January 1999. n-Nov and n-Jan is the number of samples in November and January respectively (surface, sub-surface). Sampling was conducted during the hours of darkness from 20:00-05:30.

Region Locality name Station Latitude (°S) Longitude Nov-98 Jan-99 # (°E) date n-Nov date n-Jan 1 Urunga Inshore 30°32.68’ 153°09.09’ 16/17a, (2, -) a 21/22 4, 2 22/23 4, 2 Offshore 30°33.79’ 153°15.34’ “ (2, -) a " 4, 1 4, 2 2 Smoky Cape Inshore 31°00.19’ 153°05.68’ 16/17a (2, -) a 22/23 4, 2 Offshore 30°59.97’ 153°09.90’ " (2, -) a " 4, 2 3 Point Plomer Inshore 31°19.31’ 153°01.21’ 23/24 4, 2 23/24 4, 2 Offshore 31°20.04’ 153°05.78’ " 4, 2 " 4, 2 4 Diamond Head Inshore 31°45.43’ 153°00.63’ 24/25 4, 2 24/25 4, 2 Offshore 31°44.62’ 152°53.45’ " 4, 2 " 4, 2 4 Crowdy Head Inshore 31°50.96’ 152°48.37’ - - 27/28 (2,1)b Offshore 31°52.48’ 152°57.06’ - - " (2,1)b 5 Cape Hawke Inshore 32°10.06’ 152°36.88’ - - 25/26 4, 2 Offshore 32°10.10’ 152°46.77’ - - " 4, 1 TOTAL 32, 12 44, 20 aOpportunistic samples with the surface net on November 16/17, during a time of wind- induced upwelling bOpportunistic samples on January 27/28 26

All ichthyoplankton samples were preserved into 5% formalin/seawater, buffered with sodium carbonate to preserve the otoliths. Some bleaching of melanophores occurred with this buffer, if not transferred to 95% alcohol within 2 months. In the laboratory, samples were gently rinsed before sorting to separate fish larvae from the zooplankton by using a binocular microscope. Fish larvae were then preserved in 95% ethyl alcohol and zooplankton were preserved in 5% formalin. Fish larvae were identified to as low a taxon as possible, following Leis and Carson-Ewart 2000, Neira et al. 1998, Moser et al. 1993, 1994). For analysis, I used the abundance of taxa identified to the family level, except the order Anguilliformes.

Analysis Data were converted to numbers per 100 m3, and were separately analysed for each cruise (November and January). Assemblages were determined using cluster dendrograms of the Bray-Curtis similarity matrix using standardized data of the station average, with taxa >1% of the total abundance, and non-metric Multi-Dimensional Scaling (MDS). Standardised (or compositional data) are calculated as the percentage composition of each family. Initially, I examined where possible, the magnitude of within night variation in the assemblage (before versus after midnight) from the Bray- Curtis similarity matrix, using ANOSIM (Clarke 1993). I found no significant difference in the ichthyoplankton assemblages between samples collected before and after midnight (P>0.3 for surface and subsurface samples in both months) and therefore samples at a particular station within a night, were treated as replicates. For each month I compared the total abundance of larvae and the Shannon- Wiener diversity index among the orthogonal and fixed effects of region (3), station (2) and depth (2) in a 3 factor ANOVA (n=2 replicate tows, using 2 averages of the 4 surface tows to produce balanced data). Three regions were sampled in November (regions 1, 3 and 4, Table 2.1), but in January our sampling of the 5 regions was not balanced (at regions 1 and 5 the final replicate was at dawn and not completely conducted during the night). Therefore in January only 3 regions were analysed by ANOVA (regions 2, 3, and 4), but for both cruises the analysis bracketed the upwelling region (Fig. 2.1). Homogeneity of variance was tested by Cochrans test, and if necessary was transformed successfully (except where noted). 27

I used ANOSIM to compare the assemblages among regions and between stations for each cruise and gear type. ANOSIM is a one-way comparison of the Bray- Curtis similiarity matrix by the R statistic - a degree of discrimination between samples (i.e. regions or stations) ranging between -1 and 1 (Clarke and Warwick 1994, chapter 6, p 6-3). Cluster and MDS plots were used to detect the oceanographic effects of region, station and depth differences in the ichthyoplankton assemblage structure, using the statistical package Primer (version 5.0). SIMPER (similarity percentage analysis) examined the contribution of taxa to the various clusters that the MDS identified (Clark and Warwick, 1994, p. 5-7). ANOSIM was attempted on the a priori determined Region, Station and Depth. It was evident that different larval fish assemblages were recognized from the different water masses distributed diagonally across these orthogonal factors. From a matrix of families abundance (rows) on assemblages (columns), I then examined the Pearson correlation coefficients between the assemblages and some opportunistically collected samples.

2.3 Results

Oceanographic features The East Australian Current (EAC) dominated the study area on both cruises (Fig. 2.1) with a warm, poleward flowing, tropical water mass. In November, the sea surface temperature only ranged between 21 to 23ºC (Fig. 2.1a, 2.2a), compared to that in January when it ranged between 22 to 28ºC (Fig. 2.1b, 2.2b). The sea surface temperature (SST) images showed the EAC flowing southward past Region 1 and 2 and departing the coast near Region 3. Intrusion of cooler, less saline Tasman Sea waters was detected near shore off Region 4 in November and 2 to 3°C cooler water in January off Regions 4 and 5 (22 to 24ºC). Upwelling was clearly evident off Region 4 (and to the south) in the along-shore section plots at the 100 m and especially 50 m stations in both months (Fig. 2.3). Isotherms shallowed with depth south of Region 2, but cooler water at the surface was only manifest by Region 4. The SST image also reveals a small patch of cooler water (22.5°C) still remaining off Region 1 on 21 November, 4 28

days after our opportunistic sampling of a wind induced upwelling event. Our main sampling of this region was on 22/23 November (Table 2.1). During January, the nutrient levels in the upper 50 m of the water column were clearly elevated at the inshore station off Region 3 and at the inshore and offshore stations off Region 4 and to the south (Fig. 2.4). The chlorophyll biomass had not responded as dramatically but had also increased off Regions 4 and 5 (Fig. 2.4d). During November we had no corresponding nutrient data (see Methods). However, using nutrient information collected earlier in the cruise (Roughan and Middleton 2002), and only data from the upper 50 m of the water column at our inshore and offshore CTD stations, I identified a significant negative correlation between water temperature and NOx, FRP and Si (r= -0.53, -0.62 and -0.53 respectively). Therefore for the November water temperatures 17-23°C and NOx concentrations 0-3 μM, I observed for every degree decrease an approximate increase of 25% in NOx concentration. 29

35.7 a) Nov. 1998 1o,w 2o,w 35.6 3o 1i,w 2i,w 1i 3i 1o 35.5 4o Salinity (‰) Salinity 4i 35.4

35.3 21.0 21.5 22.0 22.5 23.0 23.5 35.7 b) Jan. 1999

35.6 5o 1o 1i 4i 35.5 3i 5i 2i 2o

Salinity (‰) Salinity 4o 3o 35.4

35.3 22.0 23.0 24.0 25.0 26.0 27.0 28.0 Temperature ºC Figure 2.2. Surface T-S plots from the vessel’s thermosalinograph, from 4 m depth, during the plankton tows at each of the regions and stations in a) November 1998 and b) January 1999. i=inshore; o=offshore). Error bars are ± standard error. In November, the theromosalinograph data for the wind-induced upwelling event are circled with a dashed line. 30

d . e r a d Is p e e a e a m H e k n o H w to a C l d a h g y P n y H g n k t o d e u u o in m w p o r m o ia ro a r N1 U S PN1 D N2 C C B 0 a) Nov98, N1 N1 N2 20 Inshore

Depth (m) 40 0 b) Nov98, N1 N2 N2 offshore 20 N1 N1 N2

60 (m) Depth 80 100 0 c) Jan99, J4 J2 J2 J4 J4 inshore J1 J1 J5 J5 20 J1

Depth (m) 40

0 d) Jan99, J1 J3 J3 J1 J1 offshore 20 J3 J3 J1 J1 J2 40

60 Depth (m)

100 30°S 30’31° 30’32° 30’ 33°S

Latitude °S Figure 2.3. Temperature profiles along the inshore (50 m contour) and offshore (100 m contour) stations in a) November 1998, and b) January 1999. Dashed lines show the location of the CTD cast. The ichthyoplankton assemblages subsequently identified (N1, N2, J1 to J6, Fig. 2.6) are included. 31

3 0.4 a) c) Inshore/Offshore: 50 m 0.3 2 100 m

0.2

1 Phospate (µM) 0.1 Nitrate+Nitrite (µM)

0 0.0 4 3 b) d)

) 3 -1 2

Silicate (µM) 1 1 Chlorophyll a (µg. L

0 0 123 45 123 45 Region Region Figure 2.4. Average concentrations in the upper 50 m of the water column in a) nitrate and nitrite (μM), b) silicate (μM), c) free reactive phosphorous (μM) and d) chlorophyll a biomass (μg.L-1), at inshore CTD stations (<50 m isobath) and at offshore stations (50-120 m isobaths), at each region in late January 1999. (1=Urunga; 2=Smoky Cape; 3=Point Plomer; 4=Diamond Head and Crowdy Head; 5=Cape Hawke).

Larval fish taxonomy and composition A total of 25,438 individual larvae were identified comprising 112 taxa (111 families and 1 order, Table 2.2) over both cruises. The Anguilliformes consisted of at least three families including the Anguillidae, Congridae and Ophichthidae. Most larvae were at the flexion or postflexion stages, and larval length ranged from 1.6 mm (Carangidae) to more than 20 mm (Anguilliformes). In November, we observed an order of magnitude greater abundance of larvae and slightly less diversity than in January (Table 2.2, Fig. 2.5). In November we found fewer larvae in the surface nets than subsurface nets, although the effect was not as marked in January. In November, we found in both nets 101 taxa, while in January we found 103 taxa. In November, the most abundant 10 families accounted for 81% of the total abundance of larvae, whereas in January, the most abundant 10 families accounted for only 58%. The 10 dominant taxa in November were the carangids (trevally), clupeids (herrings), callionymids (dragonets), sillaginids (whiting), microcanthids (stripeys), gobies, platycephalids (flatheads), bothids, lutjanids (snapper), and sparids (bream). The 10 dominant taxa in January were gobies, carangids, clupeids, callionymids, labrids (wrasses), bothids, lutjanids, cynoglossids, sillaginids and scombrids (Table 2.2). In November there was no significant difference in the total abundance of larvae among regions, stations or depths (Fig. 2.5a, b, Table 2.3a, ANOVA, p>0.2), but for diversity there was a significant Region*Station interaction (Table 2.3b, ANOVA, p=0.05), where Region 4 offshore had a slight, but significantly lower diversity than the other regions and stations (Fig. 2.5c, d, SNK test). The total abundance in January was significantly greater at Region 2 compared to the other 2 southern regions (Table 2.3c, Fig. 2.5e, f, ANOVA, p=0.001). The diversity in January could not be satisfactorally transformed (Table 2.3d), but it tended to decline from north to south (Fig. 2.5g, h). 33

Table 2.2. List of ichthyoplankton taxa (in alphabetical order), the total number of larvae caught over both cruises (based on a constant catch per unit effort), their contribution (%) and rank based on individual number caught at surface and sub-surface in November 1998 and January 1999. 0, refers to <0.1%; -, refers to no larvae caught, numbers and families in bold refers to >1%, and were analysed for assemblage structure. If the family abundance is >5% or contains a commercial species, the dominant species are listed where identification was possible (at least for the larger individuals), but the species abundance is uncertain. Taxa November 1998 January 1999 surface sub-surface surface sub-surface n % n % n % n % Acanthuridae 9 0.2 31 0.4 14 0.3 87 0.9 Acropomatidae 2 0 13 0.2 - - 4 0 Ammodytidae 9 0.2 5 0.1 16 0.4 46 0.5 Anguilliformes 4 0.1 9 0.1 17 0.4 165 1.7 Apogonidae 50.1120.1 43 1.0 53 0.6 Argentinidae - - 2 0 - - -- Arripidae 24 0.5 12 0.1 18 0.4 11 0.1 Arripis trutta Atherinidae - - 9 0.1 2 0 -- Aulostomidae - - 2 0 - - 1 0 Balistidae 2 0 3 0 - - 21 0.2 Belonidae 1 0 - - - - 2 0 Berycidae 1 0 - - 2 0 8 0.1 Blenniidae 30 0.6 28 0.3 16 0.4 14 0.1 Bothidae 118 2.6 228 2.8 153 3.7 344 3.6 Pseudorhombus arsius Pseudorhombus jeynsii Bramidae - - 1 0 12 0.3 -- Bregmacerotidae - - 2 0 - - 12 0.1 Callanthiidae - - 3 0 - - -- Callionymidae 139 3.0 1035 12.6 100 2.4 454 4.8 Dactylopus dactylopus Foeteropus calauropomus Caproidae 1 0 7 0.1 1 0 18 0.2 Carangidae 768 16.6 2323 28.4 221 5.4 841 8.8 Pseudocaranx dentex Trachurus novaezelandiae Centrolophidae 2 0 10 0.1 4 0.1 1 0 Cepolidae - - 3 0 - - 66 0.7 Chaetodontidae 5 0.1 8 0.1 12 0.3 116 1.2 Chandidae 8 0.2 8 0.1 6 0.1 27 0.3 Cheilodactylidae 4 0.1 - - 1 0 -- Chironemidae 4 0.1 7 0.1 - - -- Cirrhitidae - - - - 1 0 4 0 34

Table 2.2 (Continued)

Taxa November 1998 January 1999 surface sub-surface surface sub-surface n % n % n % n % Clupeidae 1380 29.9 1725 21.1 368 8.9 255 2.7 Sardinops sagax Coryphaenidae 20 0.4 - - 5 0.1 -- Creediidae 4 0.1 16 0.2 3 0.1 9 0.1 Cynoglossidae 29 0.6 54 0.7 181 4.4 228 2.4 Paraglagusia unicolor Dactylopteridae 29 0.6 37 0.5 16 0.4 38 0.4 Diodontidae 1 0 4 0 - - 4 0 Engraulidae 26 0.6 86 1.1 27 0.7 25 0.3 Enoplosidae 3 0.1 4 0 6 0.1 3 0 Exocoetidae 15 0.3 - - 30 0.7 -- Fistulariidae - - 2 0 - - 1 0 Gempylidae 2 0 - - 2 0 1 0 Rexea solandri Gerreidae - - 37 0.5 13 0.3 36 0.4 Girellinae 12 0.3 18 0.2 1 0 6 0.1 Gobiesocidae 7 0.2 22 0.3 - - 7 0.1 Gobiidae 185 4.0 341 4.2 692 16.8 1127 11.9 Amblygobius phalaena Arenigobius bifrenatus Bathygobius krefftii Callogobius depressus Favonigobius lateralis Gnatholepis inconsequens Ptereleotris evides Ptereleotris microlepis Valenciennea immaculatus Gonorynchidae 3 0.1 5 0.1 5 0.1 54 0.6 Gonostomatidae - - 1 0 1 0 -- Haemulidae 3 0.1 9 0.1 1 0 10 0.1 Hemiramphidae 10 0.2 - - 2 0 1 0 Holocentridae - - - - 6 0.1 98 1.0 Kyphosidae 14 0.3 37 0.5 9 0.2 24 0.3 Labridae 29 0.6 182 2.2 309 7.5 412 4.3 Achoerodus viridis Bodianus axillaris Bodianus vulpinus Coris gaimard Hemigymnus fasciatus 35

Table 2.2 (Continued)

Taxa November 1998 January 1999 surface sub-surface surface sub-surface n % n % n % n % Hemigymnus melapterus Pictilabrus sp. Pseudolabrus gymnogenis Thalassoma purpureum Leiognathidae - - 8 0.1 1 0 -- Leptobramidae - - - - 2 0 3 0 Leptoscopidae 3 0.1 ------Lethrinidae 30 0.6 32 0.4 74 1.8 150 1.6 Lethrinus nebulosus Lophotidae - - 1 0 - - -- Lutjanidae 138 3.0 100 1.2 142 3.4 445 4.7 Lutjanus argentimaculatus Lutjanus erythropterus Lutjanus malabaricus Lutjanus sebae Paracaesio xanthurus Pristipoides multidens Pristipoides typus Malacosteinae - - 1 0 - - 1 0 Melamphaidae - - 2 0 - - -- Menidae - - 1 0 - - -- Microcanthidae 477 10.3 64 0.8 3 0.1 11 0.1 Atypichthys strigatus Microdesmidae 3 0.1 3 0.0 66 1.6 55 0.6 Monacanthidae 12 0.3 20 0.2 14 0.3 45 0.5 Monodactylidae - - - - 1 0 1 0 Mugilidae - - 17 0.2 3 0.1 4 0 Aldrichetta forsteri Liza argentea Mugil cephalus Myxus elongatus Mullidae 157 3.4 45 0.5 77 1.9 24 0.3 Myctophidae 16 0.3 38 0.5 355 8.6 1655 17.4 Nemipteridae 52 1.1 50 0.6 31 0.8 116 1.2 Notacanthidae - - 1 0 10 0.2 22 0.2 Odacidae 9 0.2 22 0.3 14 0.3 49 0.5 Ophidiidae 1 0 - - 1 0 13 0.1 Opistognathidae ------8 0.1 Ostraciidae 3 0.1 4 0 4 0.1 72 0.8 36

Table 2.2 (Continued)

Taxa November 1998 January 1999 surface sub-surface surface sub-surface n % n % n % n % Paralepididae - - 1 0 - - 6 0.1 Paralichthyidae 6 0.1 15 0.2 5 0.1 11 0.1 Pegasidae 2 0 - - 2 0 2 0 Pempheridae 1 0 - - 2 0 -- Percophidae 1 0 5 0.1 5 0.1 3 0 Pinguipedidae 27 0.6 31 0.4 23 0.6 55 0.6 Platycephalidae 72 1.6 377 4.6 47 1.1 226 2.4 Platycephalus arenarius Platycephalus caeruleopunctatus Platycephalus endrachtensis Platycephalus fuscus Platycephalus longispinis Platycephalus marmoratus Thysanophrys cirronasus Plesiopidae 1 0 12 0.1 2 0 11 0.1 Pleuronectidae 2 0 6 0.1 - - 8 0.1 Poecilopsettinae - - - - 1 0 -- Polynemidae 3 0.1 8 0.1 2 0 3 0 Pomacanthidae 14 0.3 13 0.2 8 0.2 32 0.3 Pomacentridae 39 0.8 6 0.1 43 1.0 98 1.0 Pomatomidae 8 0.2 19 0.2 5 0.1 9 0.1 Pomatomus saltrix Priacanthidae 1 0 17 0.2 2 0 21 0.2 Psettodidae - - 3 0 - - -- Pseudochromidae - - 2 0 1 0 22 0.2 Samaridae - - - - 3 0.1 -- Scaridae 7 0.2 15 0.2 9 0.2 1 0 Scatophagidae - - - - 6 0.1 -- Schindleriidae 9 0.2 19 0.2 42 1.0 50 0.5 Scianidae - - 13 0.2 32 0.8 46 0.5 Scomberesocidae 3 0.1 ------Scombridae 62 1.3 82 1.0 159 3.9 336 3.5 Scorpaenidae 37 0.8 52 0.6 29 0.7 271 2.8 Scorpidinae 20 0.4 10 0.1 - - 2 0 Serranidae 30 0.6 25 0.3 38 0.9 53 0.6 Siganidae ------1 0

Table 2.2 (Continued)

Taxa November 1998 January 1999 surface sub-surface surface sub-surface n % n % n % n % Sillaginidae 221 4.8 389 4.8 128 3.1 266 2.8 Sillago ciliata Sillago flindersi Sillago maculata Soleidae - - 3 0 - - 5 0.1 Sparidae 109 2.4 101 1.2 178 4.3 128 1.3 Acanthopagrus australis Chrysophrys auratus Rhabdosargus sarba Sphyraenidae 5 0.1 4 0 18 0.4 28 0.3 Synodontidae 22 0.5 27 0.3 56 1.4 177 1.9 Terapontidae 79 1.7 52 0.6 82 2.0 104 1.1 Tetraodontidae 1 0 3 0 - - 2 0 Toxotidae 1 0 1 0 - - 2 0 Trachichthyidae ------3 0 Trichiuridae 1 0 2 0 1 0 2 0 Trichonotidae ------2 0 Triglidae 1 0 31 0.4 12 0.3 82 0.9 Tripterygiidae - - 1 0 - - 1 0

Unidentified 22 0.4 88 1.1 70 1.7 132 1.4 Total no. of individual 4616 8183 4125 9509 Total no. of family 78 89 82 92 38

November 1998 January 1999 250 a surface e inshore 200 offshore 150

3 100

250 b sub-surface f 200

No. of individual/ 100 m No.individual/ of 150

100

0 1.6 c surface g

1.2

0.8

0.4

0 1.6 d sub-surface h 1.2

0.8

index diversity Shannon-Wiener

0.4

0 123 4 5125 123 4 5 Region Region

Figure 2.5. The average individual numbers of larval samples/100m3 and the average Shannon-Wiener diversity index at inshore and offshore stations of each region in a & c) surface, November 1998; b & d) sub-surface, November 1998; and e & g) surface, January 1999; f & h) sub-surface, January 1999. 39

Source SS DF MS F P SS MS F P a) Nov-98, total abundance b) Nov-98, ln(diversity index) Region (R) 5546 2 2772.9 0.41NS 0.30 0.15 2.50NS Station (S) 1449 1 1448.5 0.22NS 0.01 0.01 0.17NS Depth (D) 2099 1 2099.0 0.31NS 0.04 0.04 0.62NS R*S 11631 2 5815.5 0.87NS 0.48 0.24 3.99* R*D 10503 2 5251.3 0.78NS 0.19 0.10 1.59NS D*S 10891 110890.5 1.62NS 0.21 0.21 3.39NS R*S*D 8601 2 4300.3 0.64NS 0.09 0.04 0.73NS RES 80600 12 6716.6 0.73 0.06 TOT 131318 23 2.05 c) Jan-99, total abundance d) Jan-99, diversity index Region (R) 16020 2 8009.9 11.95** 0.02 0.01 0.43NS Station (S) 1039 1 1039.4 1.55NS 0.02 0.02 1.09NS Depth (D) 536 1 535.5 0.80NS 0.23 0.23 10.25(**) R*S 3347 21673.6 2.50NS 0.33 0.16 7.16(**) R*D 310 2 155.0 0.23NS 0.11 0.05 2.36NS D*S 50 150.3 0.08NS 0.03 0.03 1.52NS R*S*D 1231 2 615.6 0.92NS 0.02 0.01 0.49NS RES 8041 12 670.1 0.27 0.02 TOT 30574 23 1.04

Comparisons of the assemblages The ichthyoplankton assemblage was not significantly different between the 50 m and 100 m stations across all regions (except for the January surface samples), but was significantly different among regions over all stations for both cruises and both gear types (MDS, Table 2.4). The MDS ordinations of samples revealed two separate assemblages of samples in November, defined at the 50% similarity level (N1, N2, Fig. 2.6a, b). All the Region 4 samples (i.e. surface and deep, inshore and offshore off Diamond Head) and the surface Region 3 offshore were in N2. Except for surface Region 3 offshore, all N2 assemblages were found in water <21°C (Fig. 2.3a, b). The EAC assemblage (N1) was distinguished by SIMPER analysis by a greater representation of carangids, clupeids, callionymids, sillaginids and microcanthids (signature taxa, Table 2.5), and was found in water >20.5°C (Fig. 2.3c, d). The upwelled community (N2) was distinguished by callionymids, clupeids, gobiids, sillaginids and callionymids (Table 2.5). The two assemblages shared many taxa (particularly clupeids and carangids), but differed in the proportions of the signature taxa (Table 2.6). The MDS ordination was re-run using the station replicates (not the station averages), and there was little difference, with Groups N1 and N2 each broken into two close but separate groups at the 50% level. For clarity and consistency, the figure was left unchanged. There were five assemblages in January at the 50% similarity level, but two major ones were evident: J1 and J3 (the EAC assemblage) and J4 and J5 (from regions of upwelling, Fig. 2.6c, d). J1 and J3 were distinguished by SIMPER analysis by a greater representation of carangids, labrids, myctophids and scombrids (Table 2.5), and generally found >25°C, north of Region 4 (Fig. 2.3c, d). J4 and J5 were distinguished by clupeids, callionymids, platycephalids, terapontids and silliginids, and found inshore off regions 4 and 5, at temperatures <24°C (Fig. 2.3c). J5 was subsurface and J4 was surface, but J4 was also identified at the surface, Region 1 inshore at >26°C (Fig. 2.3c). J2 was characterized by a general paucity of families (Table 2.6). Again, the EAC and upwelled assemblages shared clupeids, callionymids and carangids but in different proportions (Table 2.6).

Table 2.4. Analysis of similarity (ANOSIM, two-way crossed analysis) of the community to compare between stations inshore (IS) versus offshore (OS), and among regions. P= significance level of sample statistic; NS= not significant, *=0.01

Depth-strata Cruise Comparison P (%) R Surface Nov-98 Regions 0.004 ** 0.42 Stations 0.185 NS 0.25 Jan-99 Regions 0.001 ** 0.45 Stations 0.001 ** 0.43 Sub-surface Nov-98 Regions 0.049 * 0.50 Stations 0.074 NS 0.67 Jan-99 Regions 0.001 ** 0.70 Stations 0.016 ** 0.68

Period Assemblage Average Dominant families similarity (%) Nov - 98 N1 60 Carangidae, Clupeidae, Callionymidae, Sillaginidae, Microcanthidae N2 64 Clupeidae, Carangidae, Gobiidae, Sillaginidae, Callionymidae Jan - 99 J1 61 Gobiidae, Carangidae, Labridae, Lutjanidae, Bothidae J2 64 Carangidae, Clupeidae, Gobiidae, Callionymidae, Bothidae J3 65 Myctophidae, Gobiidae, Labridae, Carangidae, Clupeidae, Scombridae J4 65 Clupeidae, Sparidae, Sillaginidae, Carangidae, Terapontidae J5 73 Callionymidae, Platycephalidae, Clupeidae, Sillaginidae, Carangidae

November 1998 January 1999

s4,o a) d1,i c) d1,i d3,i d4,o d2,i J1 s1,i d3,i N1 d3,o s1,o s2,i s3,i s1,o s5,o d1,o s3,i J2 d5,o d3,o s3,o s3,o s2,o J3 d1,o s4,o d2,o N2 s4,i J4 d4,o s1,i s4,i s5,i d5,i J5 d4,i d4,i 30 40 50 60 70 80 90 100 30 40 50 60 70 80 90 100 Bray-Curtis Similarity Index (%) Bray-Curtis similarity index (%) b) d) Stress: 0.08 J3 J2 Stress: 0.12 s1,i d2,o N1 J1 d5,o J5 d1,i s1,o d1,o d3,o s3,i d4,i d3,i N2 s3,o d1,i d5,i d4,i d3,is3,i s2,o d4,od2,i s4,o s2,i J4 d4,o s1,o s1,i s4,o s3,o d3,o s5,o s4,i d1,o

s4,i s5,i

Over both cruises, the SIMPER analysis revealed 14 families that were important in distinguishing assemblages (Table 2.6). We found N1 was significantly correlated with J2, and both were correlated with N2 (R>0.42, Table 2.7). N2 was significantly correlated with J5 and particularly J4, while J4 and J5 were not correlated (Table 2.7).

Wind-induced upwelling event, 16 November Early in the November cruise, upwelling favourable north-easterly winds (>15 kt) were experienced during mooring deployments for 2-3 days prior to the 16 November (Fig. 2.7). The 20°C isotherm was uplifted by >50 m, and cooler surface water (<22°C) was still apparent inshore even by 21 November (Fig. 2.1a). The surface net samples opportunistically deployed at this time off regions 1 and 2, revealed 3 larval assemblages at Region 1 inshore and Region 2 that grouped with N2 at the 50% similarity level, which was subsequently found off Region 4 (Fig. 2.8a, b). These 3 assemblages were significantly correlated with the upwelling assemblages N2 (r>0.78) and J4 (r>0.84), and negatively correlated with the EAC assemblages J1, J3. The unusual surface assemblage, Region 1 inshore (Fig. 2.8a, b) was significantly correlated with J5 only (R=0.55) – the January deep upwelled community off Region 4 and 5.

-50

-100

-150

-200 Depth (m) -250 Urunga, 16 November 98

-300 0 5 10 15 20 25 30 35 Distance offshore (km)

Figure 2.7. Vertical profile of water temperature on 16 November 1998, showing the wind-induced upwelling. 44

s3,o a) s4,o s2,o,w N2’ s4,i s1,o,w s2,i,w s1,o N1 s3,i s1,i s1,i,w 30 40 50 60 70 80 90 100 Bray-Curtis Similarity Index (%)

b) N1 Stress = 0.05 s1,o s3,o s3,i s4,o N2’

s4,i s1,i s2,o,w

s2,i,w s1,o,w

s1,i,w Figure 2.8. a) Cluster dendrogram and b) two-dimensional MDS ordinations of Bray- Curtis similarities of the opportunistic samples taken with a surface, neuston net only, during wind-induced upwelling 16-17 November off Urunga and Smoky Cape (labels in normal font, with a “w”). Labels in bold are the surface assemblages identified from Fig. 2.6. Asseblages are defined at the 50% similarity level with a dashed line. Note that the EAC assemblage N1 (Fig. 2.6, Region 1 and 2) is clearly separate from the new upwelling assemblage N2’ which also includes samples from Region 1 and 2. 45

Table 2.6. The average abundance (numbers 100 m-3) in the EAC associated assemblages N1, J1, J2, J3) and the upwelling associated assemblages (N2, and J4, J5, as identified in Fig. 2.6). The table is derived from the dominant families identified in the SIMPER analysis (Table 2.4), with at least one abundance >1.2 per standardized tow. The assemblage is classified as EAC or Upwelled on the relative abundance between the two groups. Classification of bothids, gobiids and sparids is unclear.

Family EAC/Upw N1 J1 J2 J3 N2 J4 J5 Carangidae E 27.7 5.0 4.8 4.7 12.2 1.6 2.5 Labridae E 0.0 3.6 0.7 6.2 0.0 0.0 0.0 Lutjanidae E 0.0 3.9 0.5 2.4 0.0 0.0 0.0 Microcanthidae E 6.3 0.0 0.0 0.0 1.2 0.0 0.0 Myctophidae E 0.0 1.5 0.0 31.8 0.0 0.0 0.0 Scombridae E 1.2 2.6 0.0 3.2 0.0 0.0 0.0 Callionymidae U 4.6 0.0 0.5 0.0 12.7 0.0 9.5 Clupeidae U 8.4 0.0 1.9 0.0 41.7 13.3 3.8 Platycephalidae U 2.2 0.0 0.0 0.0 4.3 0.0 4.2 Sillaginidae U 3.7 1.5 0.0 0.0 4.8 1.6 3.8 Terapontidae U 0.0 0.0 0.0 0.0 0.0 2.0 0.0 Bothidae - 2.1 3.6 0.4 0.0 3.1 2.2 1.6 Gobiidae - 2.5 7.6 1.5 11.3 5.6 1.2 1.3 Sparidae - 0.0 2.5 0.0 1.1 0.0 2.7 0.0

Table 2.7. Pearson correlation coefficients (R) of the seven assemblages identified from the 14 taxa in Table 2.6. Coefficients in bold are significant (p<0.05).

N1 J1 J2 J3 N2 J4 J5 N1 1.00 J1 0.18 1.00 J2 0.91 0.45 1.00 J3 -0.12 0.24 0.00 1.00 N2 0.42 -0.22 0.47 -0.20 1.00 J4 0.20 -0.21 0.29 -0.21 0.89 1.00 J5 0.24 -0.33 0.15 -0.29 0.49 0.15 1.00 46

2.4 Discussion

Upwelling off northern NSW

On both cruises during the austral summer of 1998/1999 off northern New South Wales, there was the typical EAC separation near Point Plomer (Region 3), as reported by Cresswell et al. (1983) and others, due to the topographically induced bottom stress off Smoky Cape (Oke and Middleton 2000, 2001). This was particularly evident in the SST images, revealing a near coastal water mass off Region 4 and to the south, that was approximately 2°C cooler than the EAC. Currents from Acoustic Doppler Current Profiles were strong offshore beyond the front, in excess of 2 m s-1 (Roughan and Middleton 2002), but inshore of the front currents were weakly to the north, consistent with that observed by Cresswell et al. (1983). Upward sloping isotherms indicated the upwelling was induced off Smoky Cape, and particularly in January, the levels of nutrients were 3-10 fold greater off Diamond Head than off Urunga and Smoky Cape. The comparatively low response in fluoresence may be due to grazing pressure, as the macrozooplankton in the southern regions appeared considerably larger than off Regions 1 to 3. A similar oceanographic event was apparent in late November, during my main plankton collections, but the temperature signal was not as obvious as in January, in part due to wind-induced upwelling early in the cruise (16/17 November) off Region 1, and weak downwelling off Region 4. A short gale on 18 November terminated this event, and by 21 November when the SST image was taken (Fig. 2.1a), the EAC separation was apparent. The wind-induced upwelling was used to assess my findings on signature taxa (below).

Effects of upwelling on the abundance and diversity of larval fish Despite the distinct change in nutrient levels due to the topographic upwelling, there was no consistent effect on the abundance, nor diversity of ichthyoplankton in either month. A comparison of larval abundance in Florida Current upwellings also found no significant difference in abundance or diversity (Pitts 1999), but that classification of species would have revealed signature taxa. There was however in both months two major larval assemblages at the 50% Bray-Curtis similarity level, such that assemblages were apparent as “upwelled”, or the “EAC”, or a mixture. In both 47

months the upwelled assemblage was distinguished by a greater abundance of clupeids, callionymids, platycephalids, sillaginids and teraponids, while the EAC assemblage was distinguished by carangids, labrids, lutjanids, microcanthids, myctophids, and scombrids. Not only were these familial signature communities similar between the two months, they were consistent with the wind-induced upwelling event on 16 November. Remarkably, an unusual surface assemblage inshore, Region 1 was found to be correlated with the deep upwelled water assemblage found in January. Clearly N1 and N2 assemblages had sustained some mixing between the wind- and topographically-induced upwelling, and were in fact positively correlated. The surface hydrographic data (Fig. 2.2) provide little assistance in interpreting the water masses and their history. In November range in temperature (2.5°C) and salinity (0.3) was narrow and did not distinguish the wind induced upwelling in November of region 1 and 2 offshore, while the surface Region 1 inshore TS signature did not indicate the deep upwelling, evident in the larval assemblage. The abundance of clupeids (dominated by larval pilchards Sardinops sagax, Table 2.2) in upwelling areas is well documented (Cole and McGlade 1998). Callionymids are abundant in coastal areas (Jenkins 1986, Leis 1991, Olivar and Sabates 1997, Gray and Miskiewicz 2000), but their larval presence due specifically to deeper water or upwelling events is to date, unknown. Similarly the early life history of platycephalids, sillaginids or teraponids as a family or specifically, is also unknown. In contrast the carangids, dominated by Trachurus novaezelandiae and Pseudocaranx dentex were more abundant in the non-upwelled waters. Santos et al. (2001) found carangids (Trachurus trachurus) had a positive relationship with the upwelling events. While this could indicate within family variability, I found different distributions of small (and most abundant) versus large carangids (Syahailatua and Suthers, submitted MS). The gobies varied in their response to upwelled or non- upwelled waters (Table 2.6), as this family is particularly specious with greater potential for species specific responses, and is distributed widely along-shore (e.g. Young et al. 1986, Leis 1991, Gray and Miskiewicz 2000). Therefore general signature taxa at the family level would be possible with families containing few local species.

Larval assemblages as signature taxa Are these 14 families useful in interpreting earlier ichthyoplankton studies off the East Australian coast, where modest upwelling events sporadically occur during the summer (Middleton et al. 1996), which lacked detailed oceanographic information? A recent study by Gray and Miskiewicz (2000) summarises the results of 14 periods of plankton collections over nearly 40 months off the coast of Sydney, beginning in late 1989. Briefly, plankton was collected with a 80 cm diameter net, 500 μm mesh at the surface and at 25 m depth at 6 locations each at nearshore (at the 30 m isobath) and offshore sites (at the 60 m isobath) – similar to this study. In all they collected over 78,000 larvae from 119 families. Using their total catch from their six summer collections (their Table 2.8), with the abundance of the 14 families listed in Table 2.6, I found that 4 of the 6 periods were significantly correlated with either the EAC assemblages (their Periods 5 and 9) or the Upwelling assemblages (their Periods 13 or 14). One period was correlated with both (#4) and one with neither (#1). The January data of another study off Sydney (Smith and Suthers 1999, their Table 2), during a non- upwelling event, were significantly correlated only with N1, J2 and J3 (R>0.51). Under moderate upwelling it is possible that by initially determining signature taxa of various water masses (e.g. Smith et al. 1999, Cowen et al. 1993, Hare et al. 2000), that the correlation coefficient of these taxa between two samples, could provide an index of water mass mixing, similar to the fickian diffusion of salt identified by Fortier and Leggett (1983). The composition of larval assemblages has naturally been ascribed to the spawning locality of adults (Doyle et al. 1993). However many ichthyoplankton surveys, using 500 μm mesh are sampling flexion and postflexion larvae (as in this study), which would have incurred larval mortality rates of 10-20% d-1 (Bailey and Houde 1989). Differential mortality of some taxa within the assemblage, by exposure to a new predator or prey field after a mixing event, could serve to restore the signature assemblage, or to a maturation of the assemblage. The samples in this study were collected at night, when ichthyoplankton is comparatively unstratified vertically, particularly above the thermocline (Leis 1991, Gray and Miskiewicz 2000). It is possible that the correlation analysis in this study is conservative compared to that obtained from day sampling. This study results show that along-shore variability, along a 200 km upwelling zone, exceeded the variability between the 50 and 100 m isobath stations. Despite 49

selecting the 50 and 100 m isobaths stations on the basis of a faunal step (Smith et al. 1999), these inshore and offshore stations were not necessarily any different to those at either end of the 200 km wide region. The cross shelf currents are typically an order of magnitude less than the along-shore currents (Smith et al. 1999). The growing ichthyoplankton literature now permits some general comparisons with other regions of the world (Table 2.8). The EAC separation zone (Smoky Cape- Cape Hawke) is clearly a transition zone between temperate and tropical systems, which causes the remarkable diversity (Smith and Suthers 1999). The eastern seaboard diversity is particularly enhanced during the summer by transport of larvae from the Great Barrier Reef. The importance of transition zones has also been recognised elsewhere (northern Benguela region, Olivar 1990; Cape Hatteras, Grothues and Cowen 1999; Abrolhos Bank region off eastern Brazil, Nonaka et al. 2000). Larval fish assemblages on the continental shelf and slope off the northern Benguela region differed latitudinally, and the most larval density occurred in the top of 50 m depth indicating high diversity in this region due to the intense intrusion of Angolan waters to this region (Olivar 1990). From Cape Hatteras, larval diversity corresponded to the mixed water mass between the Middle Atlantic Bight and South Atlantic Bight indicating larvae as a water mass tracer during spring (Grothues and Cowen 1999). In the Abrolhos Bank region off eastern Brazil, the mixed larval assemblages of coral-reef- associated fish and mesopelagic fishes was found mostly in the shelf break area indicating the strong influence of the Brazil current containing the tropical water mass meeting the south Atlantic Central water mass (Nonako et al. 2000). The origin and mixture of these different water masses would be difficult to determine from hydrographic traits and even current vectors. On the other hand, this study has shown that with a known affiliation between taxa and water masses, the correlation coefficient can provide a measure of origin and water mass mixing. To sum up, a high concentration and greater diversity of ichthyoplankton along- shore of NSW in spring-summer relate strongly to the EAC and upwelling events (Fig. 2.2, 2.3). These two oceanographic features may stimulate the mature fish to be spawn due to enhancement of the nutrient and primary productivity of seawater. Some key taxa, such as Carangidae, Labridae, Lutjanidae, Microcanthidae, Myctophidae and Scombridae occurred abundantly associated with the EAC or oceanic water masses (Table 2.6), while the Callionymidae, Clupeidae, Platycephalidae, Sillaginidae and 50

Terapontidae were mostly found in the surface or deep upwelled/uplifted water masses (Table 2.6, 2.7). Therefore, The EAC and upwelling episodes may play a significant role in transport and mixture of fish larvae and recruitment process of fishery in NSW waters.

Table 2.8. Numbers of families of ichthyoplankton recorded for shelf waters in western boundary current systems (excluding Anguilliformes); na=not available.

Latitude W. boundary current Notes No. Author region families 31.5ºS Northern NSW, Australia 2 cruises in 3 113 this study months 34ºS Off Sydney Australia 2 cruises in 4 111 Smith and Suthers months 1999 34ºS Off Sydney Australia over 4 years 119 Gray and Miskiewicz 2000 27ºN Gulf of Mexico over 33 days 100 Richards et al. 1993 Great Barrier Reef over 3 years 99 Leis & Goldman 1987 36ºN Cape Hatteras Spring 1996 78 Grothues & Cowen 1999 East coast United States 92 Fahay 1983 and Canada San-In coast, Sea of Japan na Minami & Tamaki 1980 6ºS NE Brazil 1 month 74 Ekau et al. 1999 20ºS Brazil Current 1978, 1995 77 Nonaka et al. 2000 30-34ºS Agulhas 1 year 139 Beckley 1993

Chapter 3

A regional and stage-specific response of two larval carangids in an upwelling region of the East Australian Current

Abstract The larvae of two carangid fishes, silver trevally (Pseudocaranx dentex) and yellowtail scad (Trachurus novaezelandiae), were generally larger and less abundant in a topographically induced upwelling region, than north of the region in pre-upwelled conditions of the East Australian Current, during two cruises in the austral summer of 1998-1999. The abundance of both species was significantly correlated for both cruises, with many small larvae (preflexion, <5 mm body length and <10 days old) in pre-upwelled regions, particularly during November, and proportionally more larger and older larvae in the upwelled regions (mostly postflexion, ≥5 mm and ≥10 days old). Estimated ages from sagittal otoliths ranged from 2-25 increments and exhibited linear growth for both species and months over the size range (3-15 mm standard length). The otolith radius-length relationship and the growth rates were similar between species and months, despite the 4ºC difference between months. Over their life, growth rates of the younger larvae were uniform throughout the entire sampling area (0.5-0.6 mm.d-1), while older larvae grew significantly faster in the upwelled regions (~0.4 mm.d-1) compared to the pre-upwelled regions (~0.3 mm.d-1). Using the greater abundance of larvae in November, it was found that the recent otolith growth during the two days before capture was significantly faster in the upwelling region, particularly inshore (<50 m isobath). Both species tended to be depleted in δ13C in the upwelling region (from – 18.5 to –19.0‰), consistent with deep water characteristics, whereas the δ15N composition tended to increase or decrease in Pseudocaranx or Trachurus respectively. The early life history of both species indicates spawning in pre-upwelled waters, but larval transport into upwelled waters is necessary for faster growth in the post-flexion stage.

Keywords: Carangidae, Pseudocaranx, Trachurus, size structure, larval growth, stable isotope, East Australian Current. 52

3.1 Introduction

Nutrient limitation is one testable cause for the low yield of Australian fisheries, which despite having the third largest fishing zone, provides only 2% of the weight of global landings (Kailola et al. 1993). Both the eastern and west Australian coasts are dominated by tropical, low nutrient, poleward currents, each receiving comparatively little river run-off. Nevertheless, the East Australian Current can exceed 3 knots thus driving sporadic upwellings onto the shelf, particularly around 33°S where it departs the coast and traverses the Tasman Sea, forming the Tasman Front (Rochford 1984, Jeffrey et al. 1990, Cresswell 1994). Such upwelling events can have a significant impact on the phytoplankton composition and partly the cause of phytoplankton blooms to the south (Dela Cruz et al. 2002). The significance of the separation point on the early life history of fish is unknown. The mortality rate of larvae exponentially decreases from around 10% per day, to <0.1% per day by the size and age of metamorphosis to a juvenile. Clearly ample food and rapid growth during this larval phase will provide some escape from this decimation of larvae. Many studies have shown that “Bigger is Better” – the mechanism being by faster growth during the larval and particularly juvenile stages. The growth limitation or stage duration hypothesis unites many of the earlier hypotheses of recruitment limitation, and growth is the most consistent variable across oceans and species. To assess the effect of upwelling therefore is best assessed by larval growth, which is the basis of this chapter. It seems that on this point, the examiner may have missed the very fundamental point of the thesis. Larval growth is clearly the most fundamental and general process or ecosystem indicator. In the ocean, a larva’s motto may well be “Grow or die”. Recent otolith growth as a proxy for somatic growth, back-calculated from the outer (or peripheral) growth increments, is one way to assess the significance of oceanographic variability. Faster growth during the larval period may reduce the duration of the higher mortality generally experienced during this time (the “stage duration hypothesis”, Leggett and Deblois 1994, Bergenius et al. 2002). Small changes in growth during the late larval stage can have a dramatic effect on larval survival and numbers at recruitment (Houde 1987, Cushing and Horwood 1994). Larval otolith growth can respond to environmental conditions such as the prey field within 2-3 days 53

(review; Suthers 1998). The nutrient food source responsible for changes in the growth of larval fish could be identified using the stable carbon (13C) and nitrogen (15N) composition (van der Zanden et al. 1998, Rau et al. 2001), changing in larval fish within 2 days (Herzka and Holt 2000, Herzka et al. 2001). The relationships between upwelling, condition or health, and stable isotope composition may be complex (Rau et al. 2001), but no study has examined larvae from the open ocean. The trevallies, jacks and kingfish (Family Carangidae) are a diverse and abundant group of pelagic fishes of the Indo-Pacific, inhabiting tropical and often nutrient poor regions and yet supporting substantial artisanal fisheries (Caranx spp., Gunn 1990, Kailola et al. 1993, Sudekum et al. 1991). These small pelagic species provide the basis for some of the largest fisheries in the world, and provide prey for most top-level marine predators in the tropical and sub-tropical seas. Silver trevally Pseudocaranx dentex and yellowtail scad Trachurus novaezelandiae are two important carangids off south eastern Australia, yet nothing is known of their early life history. A companion study of the summer ichthyoplankton assemblages off eastern Australia (Chapter 2) found that larval carangids, predominantly these two species, constituted up to 28% of the total larval abundance. Landings of silver trevally doubled between 1970 and 1990, and it is a significant recreational species (Kailola et al. 1993). This fish schools in estuaries, bays and shallow continental shelf waters, spawning in summer and grows to a maximum total length of 76 cm in Australian waters (Kailola et al. 1993). Catches of yellowtail scad have risen for the last 10 years reaching a peak of 473 t in year 1997-1998 (Steward et al. 1999, Stewart and Ferrell 2001). Spawning occurs mostly in spring- summer, although off Sydney, its larvae were caught throughout the year (Neira et al. 1998). Recent studies on the early life history of jack mackerel, Trachurus declivis in the cooler Tasmanian waters (15 to 18ºC, Young and Davis 1992, Young et al. 1993, Jordan 1994), examined the larval feeding ecology, age, and growth of this species. There have been no similar studies of the larval ecology of silver trevally and yellowtail scad larvae. Nearly 30% of ichthyoplankton were carangids (Chapter 2), which were more abundant in the warmer waters of the EAC than upwelled waters. The aim of this study is firstly, to determine if the abundance, size and overall growth rates of two larval carangids, Pseudocaranx dentex and Trachurus novaezelandiae were altered by the effect of upwelling zone found during two cruises 54

off northern New South Wales. It was expected that nutrient enrichment could increase spawning activity, larval abundance and larval growth rates. The second aim was to determine if their recent otolith growth and their stable isotope composition were influenced by the upwelling (as a test of the nutrient limitation hypothesis), and compare the growth rates with other larval carangids.

3.2 Methods

Study Area. A 185 km long section of the continental shelf was sampled between the township of Urunga (30°32.68’S; 153°09.09’E) and Cape Hawke off eastern Australia (32°10.10’S; 152°46.77’E, Fig. 3.1). The continental shelf off Urunga (hereafter referred to as Region 1) is 30 km wide, and narrows to just 17 km off Smoky Cape (Region 2), before widening to 23 km off Point Plomer (Region 3), to 35 km off Diamond Head and Crowdy Head (Region 4) and to 50 km off Cape Hawke (Region 5, Fig. 3.1). The poleward flowing East Australian Current accelerates past Smoky Cape, generating an upwelling zone found off Diamond Head and Cape Hawke during both our research cruises in November 1998 and January 1999 (Roughan and Middleton 2002; Fig. 3.1). In November the surface water temperature ranged between 22 to 23°C, while in January, surface temperature ranged between 23 to 27°C (Table 3.1). A front separating cool upwelled water from the EAC was detected 10-15 km off Region 3, extending southward to Region 4, in the vicinity of the 153°E longitude (Fig. 3.1). In late November upwelling was detected to occur inshore (50 m station) off Region 4, and similarly in January off regions 4 and 5, and at offshore (100 m station) of Region 5. Opportunistic samples were collected at Crowdy Head, just south of Diamond Head (Region 4, Fig.3.1).

Sampling techniques Plankton was collected only at night, at two stations in each region, located either at the 50 m (referred to as “inshore”) or the 100 m depth contour (referred to as “offshore”, Fig. 3.1 and Table 3.1). Two types of plankton gear were used; a neuston net (0.75 x 0.75 m square opening, 500 µm mesh size) and a multiple opening and closing net (EZ- Net, 1 m square opening, 500 µm mesh size). In November, samples were collected off 55

regions 1, 3 and 4, and in January samples were collected off regions 1, 2, 3, 4 and 5. The EZ net was deployed over three depth ranges, but only data from the top depth bin are reported (at 10-20 m or 10-30 m from the surface at the 50 m or 100 m stations, respectively). Samples are therefore referred to as either “surface” or “deep”. Temperature and salinity at each sample site were recorded from the engine intake at 4 m depth by a regularly calibrated thermosalinograph as well as by a CTD cast at each station (the SeaBird Conductivity-Temperature-Depth, Table 3.1).

Region Date Station SST No. tows Pd Tn Pd-oto Tn-oto 1, Urunga 22-23 Nov. inshore 22.3 3, 2 16, 118 24, 44 31 69 offshore 22.2 4, 2 43, 228 157, 831 17 28 3, Pt. Plomer 23-24 Nov. inshore 22.7 4, 2 70, 108 85, 245 17 33 offshore 22.7 4, 2 111, 169 59, 174 16 66 4, Diamond Hd. 24-25 Nov. inshore 22.1 4, 2 8, 49 5, 35 34 19 offshore 22.7 4, 2 90, 140 24, 133 31 11 TOTAL 23, 12 338, 812 354, 1462 146 226 1, Urunga 21-22 Jan. inshore 26.6 4, 2 5, 46 7, 63 - - offshore 26.0 2, 1 4, 10 9, 2 - - 2, Smoky Cape 22-23 Jan. inshore 26.2 4, 2 10, 64 8, 27 - - offshore 27.1 2, 2 18, 124 18, 54 - - 3, Pt. Plomer 23-24 Jan. inshore 26.6 5, 2 20, 56 22, 36 - - offshore 27.2 4, 2 11, 3 0, 12 - - 4, Diamond Hd. 24-25 Jan. inshore 22.1 4, 2 3, 11 9, 25 - - offshore 26.9 4, 2 14, 58 5, 51 - - 5, Cape Hawke 25-26 Jan. inshore 22.9 4, 2 1, 6 1, 40 - - offshore 22.3 2, 2 12, 24 9, 25 - - TOTAL 35, 19 98, 402 88, 335 - -

152°30’E 153°E Nov98 Jan99 a) P. dentex in off 1, Urunga surface deep

2, Smoky Cape

m 31°S 31°S 0

0 2 3, Point Plomer

4, Diamond Head

Crowdy Head

32°S Tasman 32°S 5, Cape Sea Hawke b) T. novaezelandiae

1,Urunga

2, Smoky Cape m

31°S 0 31°S 0 2 3, Point Plomer

4, Diamond Head none <1 Crowdy Head 1-10 11-20 32°S 21-30 32°S 5, Cape Hawke >30

152°30’E 153°E 153°E

Figure 3.1. Study area off eastern Australia, showing the spatio-temporal distribution and abundance (# 100m-3) of a) Pseudocaranx dentex and b) Trachurus novaezelandiae larvae at the surface (black circle) and in the sub-surface (grey circle) at the 50 m and 100 m isobath stations, during November 1998 and January 1999. Block arrows indicate the predominant flow of the East Australian Current observed with an Acoustic Doppler Current Profiler (1 to 3 knots), and the approximate frontal line observed in SST images off Diamond Head, around the 153°E longitude, during our cruises is indicated for November (22 to 24°C) and for January (24 to 26°C).

Laboratory procedure Plankton samples were immediately preserved into 5% formalin buffered with sodium carbonate (NaCO3), to preserve the otoliths (which incurred some bleaching of the melanophores after 2-3 months). Within 2-3 months, samples were rinsed with freshwater, sorted to family level and stored in 95% alcohol. Carangids were particularly abundant, and were further identified to the species using (Neira et al., 1998). Larvae were sub-sampled approximately in proportion to the size range available (up to a maximum of 30) and measured for notochord or standard length (±0.01 mm), using a dissecting microscope attached to an image analysis system. Sagittal otoliths were identified and removed using a dissecting microscope with polarized light (Stevenson and Campana 1992), and the rest of the body was placed into a single vial for stable isotope analysis. The three pairs of otoliths were found in larvae >6 mm in length, and in smaller larvae only the sagittae and lapillus were found. Typically the left sagitta was mounted onto a glass slide with nail polish, and examined unpolished with a compound microscope using the 64x or 100x oil immersion lens coupled to an image analysis system. Polishing of the otolith did not improve its appearance in these young larvae. The total number of increments was counted at least twice, and the maximum radius was measured, before measuring the width of the daily growth increments along the maximum radius (see Results). Daily increment formation in these species was indicated by marginal increment formation of individuals that we caught just after dusk and or just before dawn (unpublished data), and we therefore assume that the total increment count is the age since hatch. Daily deposition of growth increments was established by Jordan (1994) in T. declivis.

Growth calculation

The overall mean growth rate (G) was calculated as standard length at capture (SLc) divided by age, with no correction for size at hatch. Recent otolith growth was determined only for the abundant November samples using the biological intercept method (Campana 1990, Campana and Jones 1992), by firstly determining a significant linear relationship between the otolith radius and standard length. Recent growth on days 1 and 2 precapture (RG1&2, mm d-1) was calculated from the width of the outer two most complete increments (W1&2, µm): -1 1) RG1&2= 0.5 * W1&2 * ((SLc – SLh) * (R – hatch check radius) 58

SLh is length at hatch which we determined as 1.6 mm for P. dentex and 2.1 mm for T. novaezelandiae (Table 3.2). The slope of RG1&2 varied with size and amongst regions (see Results) so we calculated an individual recent growth index (RGI), that was independent of size: 2) RGI = RG1&2 / Ln (SL).

Table 3.2. Hatch length of Pseudocaranx and Trachurus larvae used in the growth back-calculation. Hatch length Species Common name (mm) Sources James (1967); Neira et al. P. dentex Silver trevally 1.6 (1998) <3.6 mm, Neira et al. (1998); T. declivis Jack mackerel - Jordan (1994) T. japonicus Japanese mackerel 2.3-2.5 Ochiai et al.(1982) T. novaezelandiae Yellowtail scad - <3.3 mm, Neira et al. (1998) T. symmetricus Pacific jack mackerel 1.91-2.38 Matarese et al. (1989) T. symmetricus Pacific jack mackerel 2.45 Theilacker (1980) T. trachurus Atlantic horse mackerel 1.8-2.5 Russell (1976) 59

Stable Isotope Analysis (SIA) Only larvae from the abundant November samples (combining surface and deep nets), were subject to SIA. After extracting the otolith, the remainder of the body was freeze dried for at least two days, then cut in several parts, and weighed into a tin capsule on a microbalance (nearest 0.001 mg). To obtain sufficient material for SIA (ca. 1 mg), some samples required pooling several larvae (less the head) from the same plankton tow (from 1 to 12 extra larvae, n=3 samples per region and station). SIA for δ15N and δ13C analysis was conducted with an Automated Nitrogen Carbon Analysis-Mass Spectrometer (20-20 Europa Scientific) by CSIRO Land and Water Laboratory, Adelaide. The tin capsules were combusted and reaction products separated by gas- 15 chromotography to obtain pulses of pure N and CO2 for the composition of δ N and δ13C against standards of EDTA (Ethylene Diamine Tetra Acetic Acid) for nitrogen and Glycine for carbon which had been calibrated against the international standards (Gaston and Suthers 2004). The values of delta (δ) were calculated as

3) δX (ppt) = [(Rsample /Rstandard) – 1) x 1000, where X is 13C or 15N and R is 13C: 12C or 15N:14N, respectively. The precision of δ15N and δ13C values was ± 0.2 ‰ in the analyses.

Statistical analysis For each month’s data, the abundance of larvae (no. per 100m3) was compared amongst regions (3), stations (2) and depths (2, n=2 replicate tows) in an orthogonal and fixed factor ANOVA. In January there were only balanced data for regions 2, 3 and 4 (because the second offshore sample at regions 1 and 5 was not obtained before dawn, Table 1) and thus regions 1 and 5 are compared by inspection. Larval abundance was transformed to Ln(x+1) before analysis and homogeneity of variance was tested with a Cochrans test. Post hoc comparison of means was made by a SNK test. For all other analyses such as size and growth, larvae from the surface and deep nets were combined, as inspection or preliminary analysis showed no consistent trends. Multidimensional scaling analysis (MDS) and Analysis of Similarity (ANOSIM) were used to compare at each region and station the abundance of size groupings of larvae at 0.5 mm length intervals, where there were >20 larvae (<10 mm) in a sample (there was no significant difference between the surface and deep nets, Table 3.3, ANOSIM, p>0.10). The resultant length frequency data for each region, station and species were 60

clustered using the Bray-Curtis similarity index, and assemblages were defined at the 75% similarity level. I did not attempt to analyse this table at the level of replicate and a two-way crossed ANOSIM as the sample size for each species’ lengths was sometimes small, producing zeros which are highly influential in an MDS analysis. For the November data where I found greater abundance, I used for each species a one factor ANOVA to compare the overall mean growth rate of individual fish (G) among 12 levels or “treatments” composed of the 3 regions, 2 stations and 2 age classes, followed by a Tukey’s post-hoc test (the one factor ANOVA does not require balanced data). Length on age regressions were compared among species and months by ANCOVA. The November otolith analysis of RG1&2 for each species were also compared amongst regions by ANCOVA, using SL back-calculated less the marginal increment as the covariate. The individual RGI was compared as a one-factor ANOVA among 6 groups (3 regions x 2 stations).

Species Periods N Comparison P R P. dentex Nov. 6,6 S vs D 0.68 ns -0.13 6,6 I vs O 0.10 ns 0.00 Jan. 2,4 S vs D 0.25 ns 1.00 4,2 I vs O 0.67 ns 0.25 T. novaezelandiae Nov. 5,5 S vs D 0.82 ns -0.21 5,5 I vs O 0.10 ns -0.50 Jan. 2,2 S vs D 0.33 ns 1.00 2,2 I vs O 1.00 ns -0.50

3.3 Results

Abundance and size The abundance of the two species in November was higher than in January (Fig. 3.1, Table 3.1). In total 1650 larval Pseudocaranx were caught (average of 0.46 individuals 100 m-3), with 70% and 30% caught in November and January, respectively. In total 2240 larval Trachurus (average of 0.43 individuals 100 m-3), of which 81% and 19% were caught in November and January, respectively (Table 3.1). There was no significant difference in the concentration of Pseudocaranx or Trachurus among months, regions, stations or depths (Table 3.4, ANOVA, p>0.05), with the single exception of more Pseudocaranx at the 100 m station in November (Table 3.4a). The concentration of both species was significantly correlated in both months and gear types (Fig. 3.2).

Table 3.4. Summary of ANOVA of larval abundance (no. of individual/100m-³) of Pseudocaranx and Trachurus sampled in November 1998 and January 1999. Data transformed to Ln(x+1); R=Region; S=Station (50 and 100 m isobath); D= Depth (surface and sub-surface); *, P<0.05; **, P<0.01 November January Source SS DF MS F P SS MS F F a) P.dentex Region (R) 2.51 2 1.25 1.25 NS 1.12 0.56 0.11 NS Station (S) 3.05 1 3.05 3.05 NS 5.44 5.44 1.03 NS Depth (D) 5.68 1 5.68 5.67 * 0.20 0.20 0.04 NS R*S 1.75 2 0.88 0.88 NS 6.99 3.50 0.66 NS R*D 1.96 2 0.98 0.98 NS 17.52 8.76 1.66 NS S*D 0.80 1 0.80 0.80 NS 0.16 0.16 0.03 NS R*S*D 0.37 2 0.19 0.18 NS 3.19 1.60 0.30 NS RES 12.02 12 1.00 63.35 5.28 TOT 28.16 23 97.98 b) T. novaezelandiae Region (R) 1414.89 2 707.45 0.86 NS 0.16 0.08 0.27 NS Station (S) 1167.62 1 1167.62 1.42 NS 0.02 0.02 0.08 NS Depth (D) 1085.42 1 1085.42 1.32 NS 0.37 0.37 1.23 NS R*S 511.72 2 255.86 0.31 NS 0.15 0.08 0.25 NS R*D 1910.69 2 955.34 1.16 NS 0.83 0.41 1.38 NS S*D 507.84 1 507.84 0.62 NS 0.37 0.37 1.24 NS R*S*D 953.02 2 476.51 0.58 NS 0.05 0.03 0.09 NS RES 9862.19 12 821.85 3.60 0.30 TOT 17413.37 23 5.55

2.5 Nov., surface, R=0.66 (23)

Nov., deep, R=0.77 (12)

Jan., surface, R=0.53 (41) 2.0 Jan., deep, R=0.63 (22)

]+1)] 1.5

1.0

T. novaezelandiae

([ 10

Log 0.5

0.0 0.0 0.5 1.0 1.5 2.0

log10 ([P. dentex]+1)

-3 Figure 3.2. Correlation in log10 abundance (no. 100 m ) between Pseudocaranx and Trachurus, over November and January for the surface and deep nets. Legend shows the Pearson correlation coefficients (bold if p<0.05, number of tows). 63

Over 97% of Pseudocaranx and 42% of Trachurus were measured for body length. The range in body length of Pseudocaranx and Trachurus were similar, ranging from 1.6 to 15.8 mm, with most of the larval lengths between 1.6 to 10.0 mm. Only 3 of 1608 P. dentex and 10 of 948 T. novaezelandiae were >10.0 mm, and 17% and 14% of Pseudocaranx and Trachurus, respectively were >5.1 mm, the approximate size at flexion for Pseudocaranx and Trachurus (Neira et al. 1998). At 75% Bray-Curtis similarity level, the MDS plots of the size frequency structure from November, indicated smaller larvae of both species (average 3.8 mm) at regions 1 and 3, and a significantly different group in regions 3 and 4 (Fig. 3.3a). In January there was no discernable geographic pattern in the three different size groups (Fig. 3.3b). The different size groups identified at the 75% Bray-Curtis similarity level (Fig. 3), are shown as length-frequency histograms (Fig. 3.4) from size at hatch to 10 mm. Group I larvae in both months included recently hatched larvae <3 mm (group average <4 mm), whereas Groups II and III in both months had 92% larvae >3 mm and 25% >5 mm. 64

a) Nov.98 Tn-1i Stress=0.05

Group I (3.8±0.9, n=1038) Tn-4o Tn-4i

Group II (4.5±1.1, n=983) Pd-1i Pd-3i Pd-4i

Tn-3i Tn-3o Pd-4o Pd-3o Pd-1o

Tn-1o Group III (4.1±1.2, n=47)

b) Jan.99 Stress=0.01 Tn-3o

Group I (4.0±0.9, n=366)

Group II (5.5±1.8, n=22)

Pd-3i Tn-3i Pd-4o

Pd-1i

Group III (4.4±1.2, n=107)

Pd-2i

50 a) Nov 1998 Group II, n=983

30 Group I, Group III, n=47 n=1038 20

0 50 b) Jan 1999 40

(%) distribution Frequency Group III, n=107 30 Group I, n=366 20 Group II, n=22

0 2 4 6 8 10 Standard length (mm)

Figure 3.4. Length frequency histograms of Pseudocaranx and Trachurus sampled in a) November 1998 and b) January 1999. The groups with their sample sizes are based on the MDS plots (Fig. 3.3). 66

Growth and Age The otoliths of Pseudocaranx exhibited clearly defined growth increments, with the hatch diameter at approximately 18.4 μm (Fig. 3.5a). The otoliths of Trachurus were similar in appearance (Fig. 3.5b) with the hatch diameter at around 18.2 μm (Fig. 3.5b). Typically in both species the initial 2-3 rings (after hatching) were 2-3 μm wide, after which increments increased over the next 20 d to ca 6 μm wide (Fig. 3.6). The averaged increment width series were not significantly different between species (KS test, p=0.25).

a)a) h p-ffc ffc

h ffc

b) d)

8 a) Pd

0 0 5 10 15 20 25 30 8 b) Tn 6

(µm) of increment width Mean 4

0 5 10 15 20 25 30 Increment count

Figure 3.6. Mean of increment width (µm) of (a) Pseudocaranx and b) Trachurus from the hatch check to the periphery. The dashed reference line is set at 4 μm. Standard error bars increase towards the periphery as sample size decreases (Fig. 3.7). 68

There was a linear relationship between the otolith radius (R, µm) and standard length (mm) that was not significantly different between months or species: 4) Pseudocaranx, SL = 1.80 + 0.05 * R, r2=0.86, n=146 5) Trachurus; SL = 1.73 + 0.05* R, r2=0.86, n=226 For larvae 2 to 15 mm body length, total increment counts ranged from 1 to 25 (Fig. 3.7). There was a strong linear relationship between length and total increment counts, with no apparent pattern remaining in the resultant residual plots. There was no significant difference in the length-increment regression slopes of both species between months (ANCOVA, p=0.12), while Trachurus in November had a significantly smaller intercept, compared to that in January and Pseudocaranx in either month (Fig. 3.7). 69

15 Pd Nov98 Pd, Jan99 Tn, Nov98 Tn, Jan99 10

Standard length (mm) 5

0 0 5 10 15 20 25

Increment count

Figure 3.7. Body length (BL) on increment count (age) regression for Trachurus in November (BL= 1.05 + 0.31*age, n=231, r2=0.61), had a significantly different intercept to Pseudocaranx in November and January, and Trachurus in January (SL= 1.41 + 0.32*age, n=340, r2=0.70). For comparison, the dashed line is the average age- length regression of T. declivis off eastern Tasmania during summer (ln(SL)=0.051+1.139*age, where age is the days after first feeding, Jordan 1994). 70

Using greater abundance data from November, I classified the larvae as “young” or “old” larvae (<10 increments and ≥10 increments), which I based on their overall average increment count and size (Pseudocaranx; 11.1 increments, 4.7 mm SL, Trachurus; 10.3 increments, 4.3 mm SL). The overall mean growth rate (G) of “young” larvae of both species did not differ significantly among regions or stations (approximately 0.50 mm increment-1, Fig. 3.8a, b, Table 3.5, ANOVA, p<0.05). The growth rate of “old” larvae of both species was significantly less off Region 1 [0.30- 0.35 mm increment-1], compared to off Region 4 [0.40-0.45 mm increment-1]. Growth of older larval Pseudocaranx was not significantly different between the 50 and 100 m station at any region, but larval Trachurus off Region 3 and 4 grew significantly faster inshore than offshore (SNK test, Fig. 3.8).

Source SS DF MS F P a) P.dentex R/S/size 0.72 11 0.066 5.90 ** Residual 1.42 128 0.011 b) T. novaezelandiae R/S/size 1.08 10 0.11 12.02 ** Residual 1.941 216 0.01

0.6 a) Pd 15 7 13

4 10

0.5 4 24 22

9 0.4 13 12

Standard length per day per length Standard

0.3 0.6 b) Tn 13 Yg,100 m Yg, 50 m

Old,100 m 18 8 40 13 Old, 50 m 24 0.5

16 13

0.4 30 15 41

Standard length per day per length Standard

0.3 1234 Region

The back-calculated recent growth over the previous 2 days before capture (RG1&2, using only November samples at inshore station) was linearly related to SL in both species (Fig. 3.9). For Pseudocaranx the slopes among regions were not significantly different (ANCOVA, p>0.1), and the intercepts were not significantly different among regions (ANCOVA, Table 3.6, Fig, 3.9). For Trachurus, the slopes were significantly different among regions (Table 3.6). Inspection of the scatterplots showed that the slopes were steeper at Region 3 (Fig. 3.9). The individual growth index (RGI, Equation 2) which was independent of size (Fig. 3.10) was significantly greater for Pseudocaranx, at Regions 3i and 4o than other regions and stations (Fig. 3.11a, Table 3.7), while for Trachurus, RGI was significantly greater at Regions 3i and 4i than other regions and stations (Fig. 3.11b, Table 3.7).

Table 3.6. Combined one-factor ANCOVA table for a) Pseudocaranx b) Trachurus shown in Fig. 3.9 comparing back-calculated recent growth (RG1&2) at the inshore station amongst regions (R; regions 1, 3 and 4) with standard length (SL) the covariate. NS=not significant different at P=0.05; *, P<0.05; **, P<0.01.

Source SS DF MS F P a) P. dentex Region 0.133 2 0.067 1.786 NS SL 1.098 1 1.098 29.461 ** Residual 2.908 78 0.037 b) T. novaezelandiae Region 0.536 2 0.268 10.549 ** SL 1.799 1 1.799 70.869 ** Residual 2.970 117 0.025

P. dentex T. novaezelandiae

0.75 a) Region 1 d)

0.5

0.25

0.0

) 0.75 -1 b) Region 3 e)

0.5

RG-1&2 (mm.RG-1&2 d 0.25

0.0

0.75 c) Region 4 f)

0.5

0.25

0.0 0 2 4 6 8 10 0 2 4 6 8 10 SL (mm) SL (mm)

0.7 Pd Tn 0.6

0.5

0.4 RGI 0.3

0.2

0.1

0.0 0 5 10 15 20 Standard length

Figure 3.10. Relationship between SL (mm) and recent growth index (RGI) of Pseudocaranx and Trachurus, confirming the size independence.

0.50 a a) Pd Station 17 50 m 31a 100 m 0.45 31a,b

16a,b 34a,b 0.40 RGI 17b 0.35

0.30 0 1 2 3 4 0.50 b) Tn 33b 19b 0.45

28a,b

0.40 11a,b RGI a 69 66a 0.35

0.30 0 1 2 3 4 Regions Figure 3.11. Mean recent growth index (RGI, ±SE) for each region and station for a) Pseudocaranx and b) Trachurus. Different letters signify differences by Tukey’s pairwise post- hoc comparions p<0.03). Sample size is indicated above each column. 75

Table 3.7. ANOVA table comparing the recent growth index (RGI) for a) Pseudocaranx (Fig. 3.11a) and b) Trachurus (Fig. 3.11b). *, P<0.05; **, P<0.01

Source SS DF MS F a) Region/station 0.205 5 0.041 3.124** residual 1.836 140 0.013 b) Region/station 0.374 5 0.075 5.876** residual 2.805 220 0.013

Stable isotope signatures The δ15N content of both species was significantly related to average length of larvae that were pooled for each sample (Fig. 3.12), with a similar decreasing slope and a significantly different intercept between species (ANCOVA p<0.01). Trachurus tended to be ca 0.5‰ more enriched in δ15N at length than Pseudocaranx. There was also a similar but non-significant trend for δ13C on size. A scatterplot of the two stable isotopes (Fig. 3.13), on which this size trend is superimposed, reveals that the larvae tend to converge to a similar stable isotope signature by around 8 mm SL. Pseudocaranx larvae tended to be more enriched in δ13C (p<0.13) and particularly in δ15N (p=0.01, Table 3.8) at Region 4i and Region 1o (Fig. 3.13a). Trachurus larvae were significantly more enriched in δ13C at Region 4 (inshore and offshore, Fig. 3.13b, Table 3.8). 76

10 Tn, δ15N=9.10-0.16*Sl, r2=0.19 Pd, δ15N=8.81-0.21*Sl, r2=0.11

N 15

δ 8

2 3 4 5 6 7 8 9 10

Average length (mm)

9.0 a) Pd

4i 8.5

1o 8.0

1i 7.5 4o

3i 3o

N 7.0 15

δ 9.0 b) Tn 1i 3o 8.5 3i 8.0 4i 4o 1o 7.5

7.0 -20.0 -19.5 -19.0 -18.5 -18.0 -17.5 δ13C

Figure 3.13. Average δ13C and δ15N from a) Pseudocaranx and b) Trachurus, at 50 m and 100 m station of regions 1, 3 and 4 (error bars are standard error). 78

Stable Isotope Pseudocaranx SS DF MS F P Trachurus SS MS F P δ13C R 1.27 2 0.64 3.73 0.06 R 2.49 1.25 5.80 0.02* S 0.31 1 0.31 1.84 0.20 S 0.14 0.14 0.65 0.44 RXS 0.82 2 0.41 2.40 0.13 RXS 0.06 0.03 0.14 0.87 Residual 2.05 12 0.17 Residual 2.58 0.21 Total 4.46 17 Total 5.27 δ15N R 1.29 2 0.65 3.33 0.07 R 0.21 0.11 0.28 0.76 S 0.05 1 0.05 0.25 0.63 S 0.25 0.25 0.67 0.43 RXS 2.68 2 1.34 6.88 0.01* RXS 1.21 0.61 1.59 0.24 Residual 2.33 12 0.19 Residual 4.55 0.38 Total 6.35 17 Total 6.23

3.4 Discussion

Early life history of the two larval carangids The larvae of Pseudocaranx dentex and Trachurus novaezelandiae were two to three fold more abundant in November than in January, and in subsurface than surface nets, also tending to be more abundant at the 50 m station rather than 100 m station. The abundance of both species was highly correlated across both gears and months. Most larvae of both species were at flexion – at 4 to 6 mm in length. Therefore both species appear to have protracted spawning in spring and early summer off the east Australian coast (e.g. Kailola et al. 1993, Neira et al. 1998). The adults of both genera occur in estuaries and coastal regions, and both have similar maximum sizes and growth coefficients. The early summer, inshore larval abundance of these two species is consistent with that observed off Sydney shelf waters (Smith and Suthers 1999, Smith 2003), from the southeastern Brazilian Bight (Katsuragawa and Matsuura 1992), and from the Celtic sea and west of Ireland (Fives et al. 2001). The multivariate analysis of the abundance of 10 size groups of both species did not initially distinguish the species, but grouped them with respect to a north-south gradient (November) or haphazardly (January). Remarkably there was no statistically significant regional effect in abundance, contrary to our expectation of a nutrient limited ecosystem. In fact as a component of the ichthyoplankton assemblage (Chapter 2), the 79

abundance of carangids in particular, was indicative of East Australian Current water. As a generality this finding is accurate, but with low power (n=2 and high variance of replicates), the ANOVA found few significant differences in abundance. It was initially expected that some form of classical niche partitioning would separate the larvae in time and space. This study shows it surprising that there were either very small or no significant differences in their growth curve, or in their hatch check diameter, or in their otolith-fish size relationship, or in the increment width trajectory. The growth curves were only slightly or not significantly different between months, despite the 3-4°C difference in temperature. Small larvae of both species, <10 days old grew at around 0.5 mm d-1 across all regions and stations. At the 100 m station it is likely that larvae were being transported southwards at 1 to 3 knots, such that in 10 days larvae at Region 1 could be advected to Region 4. However at the 50 m station flow was observed to be weaker and sometimes to the north (Roughan and Middleton 2002), such that mixing of larvae among regions is less likely. It is possible that predation rates may have been particularly high in these oligotrophic waters and selectively removed slow growing larvae, generating the uniform growth distribution, as food for early larvae is not believed to be limiting (Cushing 1983, except when larvae are unusually abundant, Duffy-Anderson et al. 2002). Consistent with the expectation however, was that large larvae >10 d old, of both species, had significantly greater growth rates in the upwelling region (region 3 and/or 4), compared to region 1. This comparison was only made in November when larvae were abundant and represented in all regions. Overall both species grew <0.4 mm d-1 in Region 1 inshore and offshore since hatch, but grew nearly 0.5 mm d-1 in the upwelling regions, despite these larvae being 10-20 d old and possibly subjected to mixing and advection. Therefore, it was examined the RGI over the previous two days prior to capture, assuming that the 2 d growth is representative of the region where they were caught. RGI confirmed the overall comparison, and the inspection of the slopes of recent growth (RG1&2, Fig. 3.9) on size, being significantly greater in the upwelling (Fig. 3.11). Growth was faster at the 50 m station in the core of the upwelling regions, but at region 4 Pseudocaranx had greater RGI offshore – possibly near the front (Fig. 3.1). It is acknowledged that the trends ascribed to enhanced growth could be, as in any study of larval growth, a result of differential mortality. For example predation on 80

smaller larvae could be more intense in the upwelling regions, and removing the slower growing larger larvae. Increment counts in this study were unvalidated, but the relative widths of the marginal increment were wider for fish caught in the evening compared to the early morning (personal observation). Jordan (1994) used marginal increment formation in Trachurus declivis, to establish daily increment formation, and our two species are probably the same. However Jordan (1994) also observed that the formation of the first clear increment was in response to first feeding, 2 days after hatch, and thus he corrected all his increment counts. I usually observed (and counted) an initial narrow increment inside the first clear increment, which was apparent from careful observation in the perinuclear region and altering the microscope’s fine focus. I therefore speculate that my increment counts are one day (possibly two days) less than the actual age. In comparison with the curvilinear growth curve of Trachurus declivis (Jordan 1994), a linear fit was not improved by the curvilinear model, possibly as a function of the narrow length (2-14 mm) and age (1-25 d) range. Despite T. declivis occurring in water >4°C cooler off eastern Tasmania (Jordan 1994), they were not smaller at age (Fig. 3.7), consistent with there being little difference in size-at-age between November (ca. 22°C) and January (ca. 26°C). Water temperature at capture did not improve my growth model fit. The resilience in growth to the effect of temperature in these three species is surprising, considering that temperature is the most important and general growth variable after size in temperate (Campana 1992) and tropical environments (McCormick and Molony 1995). It is possible that the high larval mortality rate (>99%, Houde 1987) may be even greater in nutrient poor waters off eastern Australia, particularly during the early larval period, selectively removing slow growers. The δ15N declined by 1-1.5‰ from hatch to post-flexion in both species, similar to the decrease observed for larval smallmouth bass in a freshwater lake (Van der Zanden et al. 1998). The decline may be due to an enriched parental and egg signature and the commencement of exogenous feeding by the larvae on comparatively depleted crustacean microzooplankton (Young and Davis 1992). Over 90% of the stomach content from adult T. declivis off eastern Tasmania, in summer and winter, was comprised of the euphausid Nyctiphanes australis. Other planktivorous fish – albeit temperate rocky reef residents – was found to be enriched in both stable isotopes relative to zooplankton (Gaston and Suthers 2004), consistent with this interpretation. It 81

is possible that the adults of both species may have different diets and stable isotope signatures, or different stable isotope fractionation rates, as the smaller larvae of Trachurus were 15N enriched compared to Pseudocaranx (Fig. 3.14). The two species had different stable isotope signatures amongst regions. Both species were significantly depleted in δ13C at Region 4i, and Pseudocaranx was also significantly enriched in δ15N at 4i (where RGI in larger larvae was greater for both species). There are few comparative data to interpret these trends. The stable isotope characteristics of upwelled water are complex, depending upon the season and phytoplankton growth (e.g. Altabet and McCarthy 1986; Wu et al. 1999). The δ15N content of two local zooplankters, Penillia and Temora, was detected to decrease in upwelled water (opposite to the carangid larvae), while the δ13C component was found to either decrease or increase (Suthers, unpublished data). The variable response at Region 1 may be related to a wind induced upwelling recorded one week earlier (Chapter 2). Does the size related δ15N content influence these findings, in view of the north to south size gradient in larval carangids (Fig. 3.3, 3.4)? I therefore calculated two size independent indices of δ15N (a ratio and a residual based), but found no substantial difference in the relationship among regions and stations. In conclusion, this study results suggest a different diet in upwelled waters, which in larger larvae led to faster growth, but further studies of the larval gut content and size related trends are needed. Both carangids appear to spawn in coastal waters of northern NSW. More post flexion carangids were found off Diamond Head (region 4), indicating along-shore displacement in the East Australian Current. Uplifting, driven by the narrow continental shelf off Smoky Cape is manifest near the surface by Diamond Head and maybe a necessary but limiting stage in the productivity of local carangid fishes. The size or stage specific importance of the upwelling for carangid early life history is a new finding that needs to be investigated off other coasts.

Chapter 4

Larval fish diversity and growth in upwelling and sewage impacted zones in spring 1998

Abstract The ichthyoplankton was compared between regions with natural versus anthropogenic nutrients in the near shore (<50 m isobath) waters off southeastern Australia, to address the hypothesis of nutrient limitation to larval survival. The upwelling region off Diamond Head (31°S), and downwelling region off Port Hacking (34°S) which was downstream of Sydney’s deep ocean outfalls releasing ca 900 ML day-1 of primary treated sewage – exhibited similar vertical profiles of chlorophyll concentration. Larval fish assemblages nearshore were not significantly different between regions, and the night-time macro-zooplankton biomass concentrations and the total abundance of ichthyoplankton also showed no consistent differences between the two regions. Ichthyoplankton diversity indices were significantly greater at Diamond Head, and distinctive regional larval assemblages were identified due to the offshore stations within the East Australian Current. The growth rates of two larval carangids <5 mm standard length, Pseudocaranx dentex and Trachurus novaezelandiae, showed no significant differences between regions. Despite the similarities, the δ15N composition of both species was significantly enriched by ~1‰ at Port Hacking compared to Diamond Head, while δ13C differed amongst species, regions and stations. The δ15N composition indicates that sewage dissolved organic matter (ammonia) could supplement the planktonic food chain at local spatial and temporal scales.

Keyword: larval fish, diversity, growth, Carangidae, stable isotopic, upwelling, sewage impacted area.

4.1 Introduction

Australia has the world’s third largest Exclusive Economic Zone but the annual fisheries landings are among the world’s lowest (~200,000 tonnes, Kialola et al. 1993). Reasons for this low productivity can be ascribed to “nutrient limitation” due to the small volume of river discharge, the warm poleward flowing currents on both the east and west coasts, and the small regions of sporadic upwelling. Some upwelling events can last a week and cover over 400 km of coastline off northern NSW and western Victoria, raising the concentration of nutrients by 2-4 fold (Rochford 1972, 1975, 1984) and their biological effects are only recently being appreciated (e.g. Hallegraeff 1993; Ajani et al. 2001, Dela Cruz et al. 2002, 2003). The ichthyoplankton assemblage of such upwellings – at the taxonomic level of family - does differ significantly from non- upwelled conditions (Chapter 2). The mechanisms for these separate assemblages are probably oceanographic and water mass specific (Frank and Leggett 1982, Smith & Suthers 1999), rather than any opportunistic response by spawning fish. However larval survival could have altered the ichthyoplankton community as growth by two larval carangids was significantly enhanced in an upwelling region (Chapter 3), whereas growth of larval pilchards was significantly reduced (Appendix 1). These findings are important as larval survival and growth are conceptually linked (the stage duration hypothesis, Houde 1987, Cowan and Shaw 2002) and in practice (e.g. Bailey et al. 1995, Campana 1996). In comparison, the effects of nutrients from sewage discharge into an oligotrophic ocean are less understood. Sydney region has the largest sewage outfall system in Australia that contributes significant nutrient loads into coastal waters (Pritchard et al. 1996, 1998, 2001). The biological or nutritional effects of rivers or of primary treated sewage would be expected to be small and undetectable compared to upwelling. Surprisingly, the stable isotope composition of a planktivorous, rocky reef fish from around Sydney did consistently reveal the signature of primary treated discharge (Gaston & Suthers 2004). The nutritional effects of sewage were probably detected due to the sedentary nature of rocky reef fish, the relatively slow tissue turnover of muscle, and also due to the continuous discharge that nurtured an anthropogenic food chain. Sewage discharge is enriched with various chemical forms 84

of reduced nitrogen (ammonia), which is preferentially taken up by phytoplankton (Handley & Scrimbeour 1997), compared to nitrate, which is the hallmark of upwelled water. The effects of nutrient limitation on larval fish could be tested using their stable isotope composition, in combination with larval growth. Larval fish may directly benefit from nutrient addition from upwelling or primary treated sewage, at the time scale of weeks. The particulate organic matter (POM) of Sydney’s primary treated sewage is depleted in δ15N (0-2‰), compared to upwelled water (4-6‰) and the dissolved organic matter (7-9‰, Gaston 2003). Larval fish 15N values can be adjusted, to indicate source composition, by their approximate trophic position as δ15N increases by 3.5‰ for each trophic level. The aims of this study were to further test the hypothesis of nutrient limitation on fisheries productivity during the early life history of fish. I compared the biological oceanography between two regions that receive nutrients from massive, natural sources (upwelling) versus small but continuous anthropogenic sources from off Sydney (during downwelling). To put this comparison into an oceanographic perspective, I include oceanographic data (no biological sampling) obtained within the same month when the upwelling/downwelling was reversed at each region. Specifically my aims were to compare the ichthyoplankton community and to compare the growth and stable isotope composition of two larval carangids between the two regions.

4.2 Methods

Study areas The northern New South Wales continental shelf narrows by half off Smoky Cape (~31ºS) to just 16 km wide, in <0.5º latitude, generating marked upwelling by the poleward flowing East Australian Current (EAC, Rochford 1975, Creswell et al. 1983, Cresswell 1994). Upwelling is usually manifest at the surface nearly 100 km south, off Diamond Head, from distinctive patterns in Sea Surface Temperature (SST) and chlorophyll a (Roughan and Middleton 2002). Diamond Head, at 32°S is 300 km north of Sydney (Fig. 4.1). Two cruises by the Research Vessel Franklin in the summer of 1998/99 (FR14/98 and FR01/99) identified three upwelling processes (Roughan and Middleton 2002) and consequent stimulation of phytoplankton populations (Dela Cruz 85

2002). Upwelling is responsible for many of Sydney’s red tides (Dela Cruz et al. 2003), as well as stimulating growth in larval trevally (Chapter 3). Port Hacking (34°S) is 20 km south of Sydney Harbour, and the two stations at the 50 and 100 m isobaths have been the subject of long-term studies of physical and biological oceanography since the 1940s (Humphrey 1963, Ajani et al. 2001, Lee et al. 2001). North of Sydney the shelf is 45 km wide, narrowing again to 20 km off Jervis Bay (35ºS). The dominant current in the region is the EAC and associated eddies (Nilsson and Cresswell 1981), with regular reversals due to wind and coastally trapped waves. In 1991 Sydney began discharging over 9 x 108 liters d-1 of primary treated sewage from three deep ocean outfalls at ~60 m deep and 3-4 km from land (Pritchard et al. 2002), 10-20 km north of our study area. During the summer, the sewage plume largely remains trapped beneath the thermocline and is dispersed (Fagan et al. 1993). Adjacent to the 50 m station is an additional coastal outfall releasing nearly 5 x 107 liters d-1 at Potter Point. The distribution of ammonium is an indicator of sewage discharge and has been detected in coastal surveys (Pritchard et al. 2002), and yet can be overwhelmed by the nutrient contribution of sporadic upwelling events, driven by the interplay of the EAC and northeasterly winds (Griffin and Middleton 1992).

Sampling procedure Larvae were collected on cruise FR14/98 at two stations (50 m and 100 m isobath) in each region on the 24-25 November 1998 off Diamond Head (DH), and on the 26-27 November 1998 off Port Hacking (PH). The two stations were approximately located at the 50 m and 100 m isobaths, corresponding to the largest persistent difference in coastal ichthyoplankton observed on earlier cross-shelf cruises (Smith and Suthers 1999). The four stations were therefore labeled as DH50, DH100, PH50 and PH100. Additionally, some oceanographic data were collected during different oceanographic conditions off Port Hacking on 14 November as the cruise began, and off Diamond Head on 20 November. Ichthyoplankton were only collected at night using a metered neuston net (75 x 75 cm2 mouth area, 500 µm mesh), towed 2-3 m from the side of the vessel and a multiple opening and closing net (an “EZ” net, 1m2 mouth area and 500 µm mesh), that was used to sample three depth strata. The EZ net supported an electronic package that remotely triggered a net for each depth stratum, as well as recording depth, temperature 86

and volume filtered. At the 50 m station we sampled at 10-20 m, 20-30 m, and 30-40 m from the surface, and at the 100 m contour station we sampled at 10-30 m, 30-50 m, and 50-80 m from the surface. With respect to vertical profiles of temperature and chlorophyll (see Results), these three depth strata are hereafter referred to as the upper mixed depth layer (UML), the chlorophyll a maximum layer (Cmax) and/or the lower mixed layer (LML), and the deep sub-thermocline layer (DEEP). Before each station was sampled, a calibrated Conductivity-Temperature-Depth meter (a Neil Brown WOCE standard), supporting a calibrated fluorometer was deployed. Unfortunately for logistical reasons, no nutrient analyses could be conducted during the latter stages of the cruise (after 23 November).

Laboratory procedure In the laboratory, larvae were sorted from other planktonic organisms and transferred into 95% alcohol. Zooplankton settled volumes were determined 1 h after the entire sample had been transferred to a 100 ml volumetric cylinder. The zooplankton was converted to a biomass concentration using the sampled volume of each net, and assuming that zooplankton had a density of 1.0 kg m-3. Larvae were identified to the family level (Moser et al. 1993, Moser et al. 1994, Neira et al. 1998, Leis and Carson-Ewart 2000). Two carangids species, Pseudocaranx dentex and Trachurus novaezelandiae, constituted 16 and 28% of the ichthyoplankton found in a companion study (Chapters 2 and 3), and were thus used for our comparison on the different nutrient sources between Diamond Head and Port Hacking. The preserved larvae were measured for standard length, and the birefringent sagittal otoliths were identified under polarsised light and removed. The otoliths were glued to microscope slides with nail polish and a total increment count was made (usually the left sagitta), before measuring the radius and the width of the daily growth increment history. Increments are assumed to be formed daily in both species, based on diel patterns in the width of the marginal increment between dusk and dawn, and upon previous research on related species (Jordan 1994). Growth rates were calculated over the two days prior to capture from the back-calculated lengths using the biological intercept method (Campana 1990, Campana and Jones 1992, Chapter 3), by firstly determining a significant linear relationship between the otolith radius and standard length. Recent 87

growth on days 1 and 2 precapture (RG1&2, mm d-1) was calculated from the width of the outer two most complete increments (W1&2, µm): -1 RG1&2= 0.5 * W1&2 * ((SLc – SLh) * (R – hatch check radius) , where SLh is length at hatch which we determined as 1.6 mm for P. dentex and 2.1 mm for T. novaezelandiae (Chapter 3). The slope of RG1&2 varied with size and amongst regions (see Results, Chapter 3), so I calculated an individual recent growth index (RGI), that was independent of size [RGI = 10*(RG1&2 / Ln (SL))]. The remaining body of each larva (i.e. less the head) was freeze dried for 1-2 days, weighed and transferred to a tin capsule for stable carbon and nitrogen isotope analysis. To achieve a minimum weight of 1.1 to 1.2 mg, some samples consisted of up to 12 larvae (n=3 samples per region and station). Stable isotope analysis (SIA) for δ15N and δ13C analysis was conducted with an Automated Nitrogen Carbon Analysis- Mass Spectrometer (20-20 Europa Scientific) by CSIRO Land and Water Laboratory, Adelaide. The tin capsules were combusted and reaction products separated by gas- 15 chromotography to obtain pulses of pure N and CO2 for the composition of δ N and δ13C against standards of EDTA (Ethylene Diamine Tetra Acetic Acid for nitrogen and glycine for carbon which had been calibrated against the international standards(Gaston and Suthers, 2004). The values of delta (δ) were calculated as

δX (‰) = [(Rsample /Rstandard) – 1) x 1000, where X is 13C or 15N and R is 13C: 12C or 15N:14N, respectively. The precision of δ15N and δ13C values was ± 0.2 ‰ in the analyses. Contour section plots were produced using krigging algorithm using the software package SURFER. Abundance and diversity of larval fish were compared among regions (2), stations (2) and depths (3; neuston, UML and Cmax strata, with n=2 replicate samples) by analysis of variance (ANOVA), preceded by Cochrans test for homogeneity of variance, and completed with SNK or Tukey’s post hoc tests (using SYSTAT or GMAV5, Underwood et al. 1998). While 4 depth strata were sampled, with n=2 replicates, some of the deeper strata samples were lost because of preservation problems. To attain a balanced ANOVA design between the 2 regions we dropped the bottom depth and analysed statistically 3 depth strata, and compared the bottom depth by inspection of the graph. One replicate sample of the Cmax stratum at PH50 was missing, so to produce balanced data for the ANOVA we used one LML sample at PH50, although they were graphed as actually sampled. 88

The larval assemblage was compared among regions and stations by multidimensional scaling (MDS) using the PRIMER 5 statistical package PRIMER 5 (Clarke and Warwick 1994). For this analysis the average of the two replicate samples was used, and only families with >1% of the total abundance were included. By removing families with <1% of the total abundance, the influence of rare taxa was minimized. Rare taxa can dominate an analysis, when their presence or absence at very low levels is proportional to the tow volume (i.e. the influence of rare taxa is artificial, driven by duration of the tow). Data were standardized using percentage data of each taxon found for each sample (2 regions x 2 stations x 4 depths, totaling 16 samples). Larval length frequency distributions were compared amongst regions and stations with a Komogorov Smirnov test, with all depth strata combined. Few carangids were caught in the two deepest nets, and inspection of the sizes revealed no consistent trend in size between the neuston and UML samples. Growth rates were also compared amongst regions and stations by combining surface and UML samples. Recent growth was compared among regions and stations with ANCOVA, using standard length back- calculated at the last complete increment as the covariate. The RGI was compared amongst regions and stations, labeled as 4 separate samples, with a one-factor ANOVA as there were unequal sample sizes.

4.3 Results

Oceanographic feature A conspicuous upwelling event was evident off Diamond Head on the 24 November (Roughan & Middleton 2002, Chapter 3), with the 14ºC isotherm elevated by nearly 100 m between 20 and 40 km offshore (Fig. 4.2a). At DH100 therefore there was marked stratification of nearly 8°C between the surface and bottom (Fig. 4.3c), and the inshore waters around DH50 were more mixed with temperatures ranging between 19 to 21ºC (Fig. 4.3c). To put this upwelling event in perspective, we observed one week earlier a 2 day southerly wind event leading up to an earlier collection of CTD data on 20 Nov. off Diamond Head, which revealed a downwelling scenario (Fig. 4.3a). Here we observed surface water temperatures 2°C cooler and waters were less stratified, being completely mixed at DH50 and ranging <5°C between surface and bottom at DH100 (Fig. 4.3a). During this downwelling, chlorophyll a biomass was uniformly 89

mixed throughout the water column (~1 mg l-1, Fig. 4.3b), compared to a distinctive subsurface chlorophyll maximum (Cmax) near the thermocline during the upwelling (Fig.

4.3d). During the upwelling the Cmax was double that found in the upper mixed layer at -1 DH50, and at DH100 the Cmax was nearly 6 fold greater (<3 mg l ) than the upper mixed layer. Conversely, we observed a static or downwelling event during the ichthyoplankton sampling off Port Hacking on 26 Nov., where the isotherms were parallel with the surface (Fig. 4.2b). The water temperature was typically 1-2°C cooler and more stratified at PH50 and PH100 (Fig. 4.3c) than at DH (Fig. 4.3a), with small sub-surface chlorophyll maxima at 30 m (Fig. 4.3d). During an upwelling event observed off PH in mid November, the waters were particularly stratified between the surface and bottom (range of 4 and 7°C at PH50 and PH100 respectively, Fig. 4.3a), and up to 2 mg l-1 was recorded throughout the upper mixed layer (Fig. 4.3b). Despite the difference in chlorophyll biomass, there was no apparent difference in zooplankton biomass between DH and PH in late November (Fig. 4.4a, b).

151˚E 152˚E 153˚E Diamond Head 32˚S 32˚S Study area

m 0 0 2

33˚S 33˚S

Tasman Sea

34˚S N 34˚S 50 km Port Hacking 151˚E 152˚E 153˚E Figure 4.1. Two sampling regions of ichthyoplankton survey in November 1998, Diamond Head and Port Hacking, on the east coast of Australia. Two stars ( ** ) indicate 50 m and 100 m contour stations. 91

) -100

(

h t

p e

D -200 a) Diamond Head, 24 Nov 98, upwelling

-300 5 1015202530354045 0

-100

)

(

D -200 b) Port Hacking, 26 Nov 98,

downwelling -300 5 101520253035 Distance offshore (km)

Figure 4.2. Section plots of temperature from CTD casts of a transect across the continental shelf at a) Diamond Head and b) Port Hacking. Contour intervals are at 2°C, except 21°C indicated with a dashed line in b). Note the domed and upward sloping isotherms in a) and the flat isotherms in b). Arrowed stations mark the 50 m (inshore) and 100 m (offshore) stations.

mid November late November 35.7 thermocline a) c)

35.6 DH100 DH50 (downwelling, PH50 (upwelling, PH50 20 Nov) (downwelling, 24 Nov) 35.5 (upwelling, 26 Nov) DH50 Salinity 14 Nov) DH100

35.4 PH100, To 35.1 & 12.3°C PH100 35.3 10 15 20 25 10 15 20 25 Temperature °C Temperature °C SS 0 b) d) PH50 50 m 100 m DH50 UML -20 PH50 UML LML C max Cmax -40 DH50

PH100 Depth (m) -60 DH100 DEEP

-80 PH100 DH100

-100 0 1 2 3 0 1 2 3 Chlorophyll a (mg l-1) Chlorophyll a (mg l-1)

inshore (50 m isotbath) offshore (100 m isobath) 80 a) DH ) 2 b) -3 60 PH 40 2 2 2 2 2 2

100m (ml. 1 2 20 1 1 1 1 1 1 Plankton biomass 0 400 c)1 d) 300 2b 2b 2b

200 2b Total a b 2b 2a 2 2 100 a a 2a 2 1 1 2a 2 2 0 1.2 2 2 e) 2b 1 1 2b f) 2 b 2 0.9 a 2 2 1b 2 2 2 0.6 a 2 2

diversity index diversity 0.3

Shannon-Weiner 0.0 SUMLLMLCmax S UML Cmax DEEP Depth of net

Figure 4.4. Bar charts for the inshore station (a, c, e) and offshore station (b, d, f), at Diamond Head (DH) and Port Hacking (PH), of a, b) the zooplankton displacement volume retained on 500 µm mesh (ml 100 m-3); c, d) the total concentration of larval fish (per 100 m3, Table 4.1a) and e, f) the Shannon-Weiner diversity index (Table 4.1b). Error bars are standard error; significant SNK test post-hoc comparisons for within a region (where possible) are labelled with a different letter. Depth of sampling: S, surface; UML, upper mixed layer; LML, lower mixed layer; Cmax, chlorophyll maximum; DEEP, deep sub-thermocline layer.

Source SS DF MS F P a) total concentration Region (R) 31.3 1 31.34 6.13 0.03 Station (S) 13.1 1 13.09 2.56 0.14 Depth (D) 7.9 2 3.95 0.77 0.48 R*S 3.3 1 3.32 0.65 0.44 R*D 187.5 2 93.72 18.34 0.00 S*D 0.7 2 0.35 0.07 0.93 R*S*D 25.6 2 12.81 2.51 0.12 Residual 61.3 12 5.11 TOTAL 330.7 23 b) Diversity index Region (R) 0.64 1 0.64 36.69 0.00 Station (S) 0.01 1 0.01 0.03 0.86 Depth (D) 0.19 2 0.09 5.31 0.02 R*S 0.05 1 0.05 2.58 0.13 R*D 0.07 2 0.04 2.11 0.16 S*D 0.15 2 0.08 4.44 0.03 R*S*D 0.05 2 0.02 1.36 0.29 Residual 0.21 12 0.02 TOTAL 1.36 23

Ichthyoplankton assemblage There was a significantly greater abundance of larval fish at DH50 and DH100 at all depths than at PH, except at the surface where the pattern was reversed (Fig. 4.4c, d, Table 4.1a). There was also significantly greater diversity at DH at all depths than at PH (Fig. 4.4e, f, Table 4.1b). There were 95 families and one order (Anguilliformes) identified from the ichthyoplankton (Table 4.2). There were 67 and 70 families at DH50 and DH100 respectively and 53 and 58 families at PH50 and PH100 respectively, but the majority of these families were <0.1% of the total abundance. Eleven families were each represented by at least 1% of the total abundance, with just three families accounting for >50% of the total abundance: Callionymidae, Carangidae and Clupeidae. The Callionymids and Carangids were predominant at DH, while the clupeids predominated (>70%) at PH (Table 4.2). Four assemblages were defined by MDS ordination (Fig. 95

4.5a) at the 60% Bray-Curtis similarity index, which could be attributed to the relative mix of the dominant three families (Table 4.3). Most of the DH samples were in assemblages A and B, due in part to all surface samples being in C and all of the PH samples except one were in assemblages C and D (Fig. 4.5b). Overall assemblages of the two regions were significantly different (ANOSIM, Global R=0.322, p<0.01), but due to the offshore assemblages (i.e. the nearshore assemblages were not significantly different, p=0.17).

Larval carangids Length frequency distributions were similar for the two species (Fig. 4.6a, b). They were significantly different between regions, but within a region there was no significant difference between the inshore and offshore (KS test, Table 4.4). Carangid larvae off DH were larger (2-10 mm, average ~4.5 mm) than off PH (2-5 mm, average ~3.6 mm). The recent growth of the larval carangids over the two days prior to capture ranged between 0 and 0.5 mm d-1, but was significantly correlated with length. The ANCOVA found that recent growth in P. dentex as a function of length, varied significantly among regions and stations (Fig. 4.7, ANCOVA; Table 4.5a). In contrast the recent growth of larval T. novaezelandiae showed no significant difference among regions or stations (Fig. 4.8, Table 4.5b). As the ANCOVA may have been confounded by the smaller size range of larvae at PH, we compared a size independent growth index among regions and stations by a one-factor ANOVA. The growth index of P. dentex was significantly greater at DH100 than PH50, with DH50 and PH100 in between (Fig. 4.9a, Table 4.6a). The growth index of T. novaezelandiae showed no significant difference among regions and stations (Fig. 4.9b, Table 4.6b). The stable isotope composition in carbon and nitrogen of the larval carangids varied significantly amongst species, regions and stations (Table 4.7), but some general patterns were evident. In both species δ15N was less at DH50 and DH100 (generally ≤8‰, except P. dentex at DH50 ~8.5‰), compared to PH50 and PH100 (>9.5‰). In particular, δ15N ratios at PH were greater than data collected north of DH (Chapter 3). δ13C ratios tended to be enriched at DH but this was not consistent between species, particularly for T. novaezelandiae at PH100. 96

Taxon DH-I DH-0 PH-I PH-O Taxon DH-I DH-0 PH-I PH-O Acropomatidae 0 0.1 - - Microcanthinae 0.1 0.5 0.2 0.8 Ambassidae 0 0.2 - 0 Microdesmidae - 0.1 0.1 - Ammodytidae - 0.1 0 - Monacanthidae 0.3 0.2 0.1 0.2 Anguilliformes 0 0.3 0 0.1 Mugilidae 0.2 - 0.1 - Apogonidae 0 0.1 0.1 0 Mullidae 1.1 0.2 0.6 0.9 Argentinidae - - 0 - Myctophidae 0.1 0.6 0.1 0.2 Arripidae 0 0.7 0 0 Nemipteridae 0.4 0.1 0 0 Atherinidae - - - 0 Notacanthidae - 0 - - Aulostomidae - 0 - - Odacidae 0.2 0.2 - 0.2 Balistidae - 0 - 0 Ophidiidae - - - 0 Berycidae 0 - - - Ostraciidae - 0.1 - - Blenniidae 0.3 0.3 0.4 0 Paralichthyidae 0.3 0.1 0.1 0 Bothidae 2.3 1.7 0.9 0.9 Pegasidae - 0 - 0 Bregmacerotidae 0 - - - Pempheridae - - - 0 Callionymidae 24.0 9.5 2.1 2.5 Percophidae 0 0 - 0 Caproidae - 0 - - Pinguipedidae 0.4 0.3 0 0.3 Carangidae 6.4 26.4 2.7 4.8 Platycephalidae 7.9 5.5 1.4 1.6 Centrolophidae 0 0.03 - - Plesiopidae 0.1 0.2 0.2 0 Cepolidae 0.5 2.0 1.9 4.55 Pleuronectidae - 0.1 0 - Chaetodontidae 0 0 0 0 Poecilopsettinae 0 - - - Cheilodactylidae - - - 0 Polynemidae 0 0.2 - - Chironemidae - 0.03 - - Pomacanthidae 0.2 0.1 - 0 Clupeidae 33.0 29.4 71.3 72.0 Pomacentridae 0.3 0.5 0.1 0.1 Coryphaenidae - 0.3 - 0 Pomatomidae 0.1 0.2 0 - Creediidae 0.2 0.5 0.1 0.1 Priacanthidae - 0 - - Cynoglossidae 0.8 0.4 0.3 0.2 Scaridae 0.1 0.1 - 0 Dactylopteridae - - 0.1 0 Schindleriidae 0 0.1 - 0 Engraulidae 1.3 0.2 1.4 0.1 Sciaenidae 0.6 0.1 - 0 Enoplosidae 0 0.1 - - Scomberesocidae 0 - 7.8 4.5 Exocoetidae - 0.1 0 0.1 Scombridae 0.1 0.5 0.1 - Gempylidae 0 - - - Scorpaenidae 1.5 2.2 0.6 1.0 Gerreidae 0 0.5 0.1 0 Scorpidinae 0 - 0 - Glaucosomatidae - 0.1 - - Serranidae 0 0.4 0 0 Gobiesocidae 0.1 - 0 0 Siganidae - - - 0 Gobiidae 2.5 3.6 0.8 0.5 Sillaginidae 7.9 1.9 0.9 0.8 Gonorhynchidae 0 0 - 0 Soleidae 0 0.1 - - Gonostomatidae 0 - 0.04 - Sparidae 0.9 1.4 0.45 0.2 Haemulidae 0.1 0.3 - - Sphyraenidae 0 0.1 - - Hemiramphidae - - 1.2 0.4 Synodontidae 0.3 0.5 0 0.1 Kyphosidae 0.3 1.0 0 0.1 Terapontidae 0.2 0.7 0.2 0.2 Labridae 2.4 0.7 1.7 0.4 Tetraodontidae - 0 - Latinae - 0.1 - - Toxotidae - 0 - - Leiognathidae - - - 0 Trachichthyidae 0 - - - Leptobramidae 0 - - - Trichonotidae - 0.2 0 - Leptoscopidae 0 - - - Triglidae 0.7 1.9 0.1 0.5 Lethrinidae 0.1 0.2 - - Uranoscopidae - - 0 - Lutjanidae 0.5 1.3 0.1 0.5 Zanclidae - - - 0 Macrorhamphosinae - - 0 - Malacanthidae 0.1 - - - Unidentified 0.7 0.5 0.6 0.1

a) DH100-Cmax A DH100-Deep DH50-LML DH50-C max B DH50-UML PH50-LML PH100-UML PH50-UML DH100-S DH100-UML C PH100-Cmax DH50-S PH100-S PH50-S PH100-Deep PH50-C D max 40 60 80 100

Bray-Curtis similarity index (%) b)

C Stress: 0.07 PH50-S PH100-S DH50-S DH100-S

PH100-CmaxPH50-UML DH100-UML PH100-UML

A D B

PH50-Cmax DH50-LML DH100-C DH50-UML max DH50-C max

PH50-LML PH100-Deep DH100-Deep

Table 4. 3. The ichthyoplankton assemblages from the MDS ordinations (Fig. 4.5) and the average similarity (%) of each assemblage from the Similarity Percentages (SIMPER) procedure. The most influential families that contributed 90% of total individual numbers are listed.

Assemblage Average Dominant families Similarity (%) A 63.4 Carangidae, Callionymidae, Clupeidae, Platycephalidae, Cepolidae B 81.9 Clupeidae, Callionymidae, Platycephalidae, Sillaginidae Carangidae C 74.3 Clupeidae, Carangidae, Callionymidae D 80.3 Clupeidae, Callionymidae

60 a) P. dentex PH-o PH-o, n=76, SL=3.4(±0.1) PH-i 40 DH-o PH-i, n=84, DH-i SL=3.0(±0.1) DH-i, n=146, SL=4.4(±0.1) 20 DH-o, n=221, SL=4.5(±0.1)

60 b) T. novaezelandiae

PH-i, n=28, SL=4.2(±0.5) 40 Frequency distribution (%) PH-o, n=21, DH-i, n=109, SL=4.4(±0.2) SL=3.6(±0.2) 20 DH-o, n=194, SL=4.5(±0.1)

0 012345678910

SL (mm)

Table 4.4. Result of Kolmogorov-Smirnov tests comparing length frequency distributions from Fig. 4.6 a) P. dentex DH-i DH-o PH-i PH-o DH-i - DH-o NS - PH-i ** ** PH-o ** ** ** - b) T. novaezelandiae DH-i - DH-o NS - PH-i ** * - PH-o ** ** NS -

100

P. dentex 0.75 a) DH-50 b) DH-100

0.5

0.25

)

-1 0.0

0.75

c) PH-50 d) PH-100

RG-1&2 (mm. d 0.5

0.25

0.0 0 2 4 6 8 10 0 2 4 6 8 10 SL (mm) SL (mm)

Figure 4.7. Relationships between SL (back-calculated to the end of outermost complete increments; mm) and the back calculated recent growth over days 1 and 2 precapture [RG-1&2 (mm. d-1)], for Pseudocaranx dentex for the surface and upper mixed layer nets combined, at a, c) 50-m-isobath and b, d) 100-m-isobath stations off a, b) Diamond Head and c, d) Port Hacking. See Table 4.5a for the ANCOVA.

101

T. novaezelandiae

0.75 a) DH-50 b) DH-100

0.5

0.25

)

-1 0.0

0.75 c) PH-50 d) PH-100

RG-1&2 (mm. d 0.5

0.25

0.0

02468100246810 SL (mm) SL (mm)

Figure 4.8. Relationships between SL (back-calculated to the end of outermost complete increments; mm) and the back calculated recent growth over days 1 and 2 precapture [RG-1&2 (mm. d-1)], for Trachurus novaezelandiae for the surface and upper mixed layer nets combined, at a, c) 50-m-isobath and b, d) 100-m-isobath stations off a, b) Diamond Head and c, d) Port Hacking. See Table 4.5b for the ANCOVA.

Species Source SS df MS F P a) Pseudocaranx SL0 1.93 1 1.93 92.13 <0.001 R,S 0.69 3 0.23 10.95 <0.001 R,S*SL0 0.68 3 0.23 10.88 <0.001 Error 3.79 181 0.02 b) Trachurus SL0 1.40 1 1.40 81.27 <0.001 R,S 0.02 3 0.01 0.43 0.735 Error 1.09 63 0.02 102

a) Pd a 31 Station 4.5 100 49a,b 50 34a,b 77b

RGI 3.5

2.5

b) Tn 4.5

20 RGI 3.5 11 18 24

2.5 DH PH Region Figure 4.9. The average RGI a) Pseudocaranx dentex and b) Trachurus novaezelandiae at 50 m and 100 m contour stations off Diamond Head (DH) and Port Hacking (PH). Error bar is S.E. mean. Sample size is indicated; significant SNK test post-hoc comparisons are labelled with a different letter.

Source SS df MS F P a) Pd R, S 13.174 3 4.391 3.718 0.013 Error 218.512 185 1.181 b) Tn R, S 3.047 3 1.016 1.178 0.325 Error 55.18 64 0.862

103

11.0 a) Pd

10.0 PH-o PH-i

9.0 DH-i

8.0 Ur-o Ur-i PP-i DH-o PP-o 7.0 N (‰)

15 11.0 δ b) Tn PH-o

10.0 PH-i

9.0 Ur-i PP-o

PP-i DH-i PP-i 8.0 DH-o Ur-o

7.0 -20.5 -19.5 -18.5 -17.5 δ13C (‰)

104

Stable Isotope Source SS DF MS F P a) 13C Species (S) 0.98 1 0.98 6.30 0.02 Regions (R) 0.02 1 0.02 0.11 0.75 Stations (N) 0.34 1 0.34 2.17 0.16 S x R 2.69 1 2.69 17.26 <0.01 S x N 0.35 1 0.35 2.27 0.15 R x N 0.71 1 0.71 4.57 0.05 S x R x N 1.49 1 1.49 9.54 0.01 Residual 2.49 16 0.16 Total 9.06 23 b) 15N Species (S) 0.69 1 0.69 6.85 0.02 Regions (R) 24.46 1 24.46 242.82 <0.01 Stations (N) 0.19 1 0.19 1.84 0.19 S x R 0.32 1 0.32 3.22 0.09 S x N 0.93 1 0.93 9.25 0.01 R x N 0.34 1 0.34 3.36 0.09 S x R x N 0.50 1 0.50 4.98 0.04 Residual 1.61 16 0.10 total 29.05 23

105

4.4 Discussion

Nutrient profiles between upwelling and discharge areas The strong upwelling off Diamond Head in late November and associated chlorophyll a biomass was surprisingly similar to the vertical distribution in chlorophyll a concentration observed during the weak downwelling event off Port Hacking. Elevated chlorophyll a has been reported as a significant local feature off Sydney in conjunction with elevated ammonia (Pritchard et al. 1996). Sewage plumes from Sydney’s deepwater outfalls are typically dispersed to the south into a field 1-2 km wide by ~10 km south (Pritchard et al. 2001). There is a wide variance in the flow field due to a coastal counter current to the north that can act to retain nutrients and chlorophyll. Current reversals, tidal currents and boundary shear retain some nutrients and plankton in the Port Hacking area, coupled with the cliff face outfall at Potters Point (~46 ML d- 1). Retention of sewage nutrients from Potters Point and phytoplankton in Bate Bay, at the mouth of Port Hacking, was found to be responsible for at least one red tide (Dela Cruz et al. 2003). Therefore evidence for a planktonic food chain influenced by Sydney’s sewage discharge is not surprising. This albeit local and persistent Sydney effect on the nutrient dynamics is evident in the stable isotope composition of planktivorous fish (Gaston and Suthers 2004) and is consistent with our comparison of vertical structure in chlorophyll concentrations when the oceanographic conditions were reversed. Downwelling at Diamond Head in mid November resulted in lower and uniformly distributed chlorophyll concentrations, while upwelling at Port Hacking produced greater concentrations throughout the mixed layer. The significant enrichment of δ15N of both larval carangids off Port Hacking was surprising, considering that Sydney’s discharge is a comparatively small amount in coastal waters and would be subject to dispersal. In the vicinity of Sydney’s discharge higher than expected concentrations of chlorophyll were observed. Despite downwelling during the collections at this region, there was a sub surface chlorophyll maximum layer consistent with that observed in other studies and ascribed to nutrient enrichment in Sydney coastal waters (Hallegraeff 1993, Pritchard et al. 1996, 2001).

Stable Isotope composition of larval fish between upwelling and discharge regions Stable isotope analysis reveals that the nitrate from upwelled waters is moderately enriched in the naturally occurring heavy isotope of nitrogen, 15N (δ15N; 4- 106

6‰). The particulate organic matter (POM) of primary treated sewage contains over half of the nitrogen, and is depleted in 15N (0-2‰, Gaston 2003). A planktonic food chain based on sewage POM is presumably long, complex and would involve bacteria. On the other hand, dissolved organic matter (DOM) of primary treated sewage has <50% of the total nitrogen and is enriched in 15N to levels of 7-9‰ (similar to that from secondary or tertiary treated discharge, Gaston 2003) and is readily assimilated by phytoplankton. Consequently the 15N enriched larvae found off Sydney may be in response to the enriched 15N of the sewage DOM, although the exact links between 15 NH3, phytoplankton, nauplii and larval fish are unknown, and do not conform to the 3.5‰ enrichment per trophic level. Despite sewage dispersal predominantly to the south, retention of nutrients is evident on at least some occasions (e.g. Dela Cruz et al. 2003), and larval fish have rapid growth rates (0.3 mm d-1, or 10% d-1) such that 15N is rapidly assimilated as protein. This nutrient enrichment hypothesis needs to be further tested, as only a single night of larval fish sampling was conducted off both regions. Furthermore, recent studies of the stable isotope composition of particulate organic matter (POM) and planktivorous fish along the NSW coast indicate a possible north to south increasing gradient in 15N composition (Gaston 2003). Another potentially confounding factor is that δ15N in larval carangids was found to decline by ~1‰ from 3 to 8 mm SL (Chapter 3), although the larvae off Port Hacking were only 1-2 mm smaller than off Diamond Head.

Zooplankton -ichthyoplankton comparisons between upwelling and discharge regions Does natural versus anthropogenic nutrients influence the ichthyoplankton assemblage, or growth rates? Despite the enhanced chlorophyll, there was no difference in the macrozooplankton concentrations (>500 μm), due to the short duration of these upwelling and downwelling events. Microzooplankton (nauplii) populations may have increased at this time, but the mesh was too coarse to make this comparison. Consequently the growth rates of the two larval carangids showed no consistent effect between species, regions or stations. Other investigators could find no significant difference between the ichthyoplankton assemblage of coastal sewage plumes versus control sites (Gray et al. 1992), but this comparison could have been confounded by oceanographic variation at 107

the control sites during the study (see Smith et al. 1999). I found a high diversity of ichthyoplankton, comparable to previous studies in the coastal region of NSW (Gray 1993, Dempster et al. 1997, Miskiewicz and Dixon 1997, Smith and Suthers 1999, Gray and Miskiewicz 2000), and also in another coastal waters of Australia (Young et al. 1986; Leis 1991). Ichthyoplankton in Australian waters is typically more diverse than other, larger ichthyoplankton surveys of western boundary currents such as off northeast Brazil (Ekau et al. 1999), and off coast of eastern US (the Cape Hatteras; Grothues and Cowen 1999). Reasons for the high diversity of larval fish in our shorter term study, and the significantly higher diversity at Diamond Head include transport and mixing of tropical taxa along the NSW coast, as well as the uplift and mixing of deeper taxa (Smith and Suthers 1999). Diamond Head region had more numbers of taxa and higher diversity index than Port Hacking, due to the proximity of the Great Barrier Reef and the impact of frequent upwelling during spring-summer months that may transport more taxa from the sub- surface to the surface. The larval abundance of several economic species may indicate that the region in between Diamond Head and Port Hacking would be an area for spawning or nursery of commercial fishes during the spring months. This circumstance may be related to nutrient supply and larval food availability in these regions. For instance, the abundance of clupeid larvae (e.g. Sardinops neopilchardus) could be associated with the upwelling nutrient enrichment. In many upwelling areas of the world, the similar situation was reported that a fluctuation in abundance of small pelagic species, like sardines and anchovies, have a correlation with the upwelling events (e.g . Cole and McGlade 1998, Beckley and Hewitson 1994, Cury and Roy 1989, Cury et al. 2000, Santos et al. 2001). 108

Chapter 5

General Conclusion

5.1 Ichthyoplankton variability and dynamic in NSW coastal zone

The essence of the problem is that Australia is bathed by a nutrient poor ocean, and the effects of nutrient enrichment are unknown. • This is the first study of ichthyoplankton off northern NSW in an area of persistent upwelling. The essence of my findings are that signature taxa of upwelled water were identified on both cruises, and on a separate wind-induced upwelling event. • Upwelling did enhance the growth of larval carangids, but enrichment from sewage did not, despite evidence from stable isotopes of sewage nutrient assimilation. • Remarkably upwelling did not enhance the growth of larval pilchard, and further sampling out into the Tasman Front is needed. • The early life history of two carangids and pilchards for eastern Australia was inferred from this study, and synthesized with existing anecdotal information.

Studies on ichthyoplankton in Australian waters are not only for biological and oceanographic reasons, but also these are providing valuable information in fisheries management (e.g Smith 2003). These aspects confirm to the larval observations along- shore of NSW coastal zone during the two cruises on the RV Franklin in November 1998 and January 1999. The study findings indicate that a high richness (111 families and 1 order) of fish larvae in the sampling sites, and most of abundant larvae are commercial valuables species, such as members of Carangidae, Lutjanidae, Clupeidae, Sillaginidae, and Scombridae. It suggests that these areas could be potential spawning or nursery habitats for the important fish species, especially during the spring-summer months. The East Australian Current (EAC) and an upwelling event as two main oceanographic structures greatly influence the distribution and abundance of larval fish in NSW coastal waters. The EAC flows poleward from tropical to temperate regions of Australia and transport fish larvae of the tropics further south to the study area. Mixed 109

taxa originated from tropics and temperate were found along the NSW coast as reported from this study and others (e.g. Miskiewicz 1987, Gray 1993, 1996, Smith and Suthers 1999, Smith et al. 1999). While, the upwelling event that occur mostly in spring- summer months (e.g. Rochford 1975, Jeffrey et al. 1990) off Smoky Cape region (31ºC) brings the water mass from the sub-surface to surface, and simultaneously the demersal species are moved to surface water column. In other word, the mixture of fish larvae is not only between tropic and temperate taxa, but also between demersal (e.g. member of Scorpanidae, Myctophidae and Platycephalidae) and pelagic taxa (e.g. Clupeidae, Carangidae and Lutjanidae). The concentration of key taxa, particularly the Callionymidae, may be used as indicators of water mass origins, and relative mixing. Consistent of oceanographic dynamics (e.g. EAC and upwelling) during the study period may reflect similar, distinctive ichthyoplankton assemblages at the taxonomic level of family. The relative abundance of Carangidae, Labridae, Lutjanidae, Microcanthidae, Myctophidae and Scombridae may be associated with the EAC or oceanic water masses, whereas the Callionymidae, Clupeidae, Platycephalidae, Sillaginidae and Terapontidae were mostly abundant in the surface or deep upwelled/uplifted water masses. Correlation coefficients of the abundance of these signature families with their abundance from opportunistic sampling of a separate wind- induced upwelling event, from either month, provided a basis for assessing the history of these coastal water masses.

5.2 Size structures, growth and condition Larval size structure may be related to larval growth and condition, and their environmental condition (e.g. Campana and Hurley 1989, Sogard 1991, Sepulveda 1994, Peterson 1996, Meekan 2003). Carangid larvae, silver trevally (Pseudocaranx dentex) and yellowtail scad (Trachurus novaezelandiae) appeared generally larger and less abundant in a topographically induced upwelling region, than north of the region in pre-upwelled conditions of the East Australian Current. The abundance of both species was significantly correlated for both months, with many small larvae (preflexion, <4.3 mm body length and <10 days old) in the pre-upwelled conditions, particularly during November, and proportionally more larger and older larvae in the upwelled waters (mostly postflexion, >4.3 mm and ≥10 days old). The early life history of both species 110

indicates spawning in pre-upwelled waters, but larval transport into upwelled waters is necessary for faster growth in the post-flexion stage. Larval growth reflects their condition, nutritional or health states (Suthers 1991, 1998, Jenkins 1991). Estimated ages from sagittal otoliths ranged from 2-25 increments and exhibited linear growth for both species and months over the size range (3-15 mm standard length). It indicates that small larvae (<15mm) are probably in the poor condition (Jenkins 1991) for up to 15-20 days due to their ability to find suitable food. The same result was also reported form the study on larvae of T. symmetricus (Hewitt et al. 1985) caught in wild, but not from the laboratory experiment (Theilacker 1978). In addition, study on larval growth of T. declivis in Tasmania waters did not support this study’s finding, however the growth rate from this study was higher than T. declivis (Jordan 1984). The overall growth rates of the younger larvae were uniform throughout the entire sampling area (0.5-0.6 mm.d-1), while older larvae grew significantly faster in the upwelled water (<0.5 mm.d-1) compared to the non-upwelled conditions (<0.4 mm.d-1), while the growth rate of T. declivis ranged 0.24-0.27 mm. day-1 (Jordan 1994). Therefore, the different in the water temperature of their habitats may take into account in the larval growth rate as Tasmanian waters are cooler. Diets of fish larvae (in terms of type and size) may differ between species and larval stages (e.g. Jenkins 1986, Young and Davis 1990, 1992). Although, there was no food types or sizes examination in this study, the stable isotope of Carbon (13C) and Nitrogen (15N) was used to detect the original nutrient source. The results of 13C and 15N showed that both species of carangid larvae tended to be enriched in δ13C in the upwelling region (from –18.5 to –19.0‰), whereas the δ15N composition tended to increase or decrease in Pseudocaranx or Trachurus respectively. The effects of larval size and the spatial effects were probably confounded, however these stable isotope signatures indicated that sources of larval diets during this study originated from marine environment.

5.3 Upwelled-water impacted versus sewage impacted zones in NSW coast The two zones, Diamond Head and Port Hacking, represented upwelled-waters and sewage impacted zones, respectively. During the short-term observation in spring 1998, a relatively high diversity of larval fish was detected (a total of 96 taxa), and at least 10 families contributed significantly (more than 1%) on the total number of larval samples. 111

They include the Bothidae, Callionymidae, Carangidae, Cepolidae, Clupeidae, Gobiidae, Labridae, Platcephalidae, Scomberesocidae, Scorpaenidae, and Sillaginidae. However, larvae in Diamond Head had more abundance and higher diversity than in Port Hacking region. This indicates that the area between these two regions (32-34ºS) may use as a spawning or nursery habitat of fishes in Australian waters including those commercially valuable species (e.g. Bothidae, Carangidae, Clupeidae, Labridae, Platycephalidae and Sillaginidae). So, along-shore of NSW waters could be one important region for fishery stock recruitment in Australian waters, particularly during spring-summer months. Geographical differences of two regions (in term of latitudinal and oceanographic features) also impact on biological and ecological condition of larval fish. A little variation in abundance of two carangid species, Pseudocaranx dentex and Trachurus novaezelandiae, was found between 50m and 100m isobath station, however more small larvae (SL<5mm) were caught in Port Hacking than in Diamond Head. It suggests that larvae of carangids grew fast in the habitat with high temperature and nutrient enrichment of upwelling (e.g. Diamond Head). In addition, this geographical effect appeared also on the relationship between otolith and somatic growth, which was linear but tended to be slightly different in their growth curves. Moreover, different signatures of stable carbon and nitrogen isotope of the two carangid species between regions may reflect differences in their prey and larval stage.

5.4 Further study and recommendation This ichthyoplankton study was conducted in November 1998 and January 1999 along- shore of NSW coast. Although, several previous observation on this field were already done, some further observation are required, especially to repeat this study for along- shore and cross-shelf sampling. Therefore, the long-term observations on ichthyoplankton is needed to show an inter-annual variability, particularly during the spring-summer months when most fishes spawning in Australian waters. This study is more concerned on growth and condition of two important carangids species, Pseudocaranx dentex and Trachurus novaezelandiae, however a lot of commercially valuable taxa were collected during the study including families of Clupeidae, Platycephalidae and Sillaginidae. The future study may focus on one or two species of these families, although larva of the Australian pilchard (Sardinops sagax) 112

along the NSW coast was initiated in this study. In addition, re-doing the growth observation on P. dentex and T. novaezelandiae including using the different techniques on larval condition (RNA/DNA, lipid, morphometric and histology) is suggested to find variability annually. Extended study on fishery management and conservation in NSW coastal waters may be recommended due to many larvae of commercial species sampled during this study. The studies on larval feeding behaviour (e.g. Gaughan 1992, Young 1992), mortality rates (e.g. Houde 1987, Heath 1992) and related to recruitment processes (e.g. Zeldis 1992, Cushing 1996) are mainly to be carried out. Thus, program on fishery management and conservation in the NSW marine area can be established by using results of this study to understand the dynamics of fishery stocks. 113

Literature cited

Ajani, P., R. Lee, T.R. Pritchard, and M. Krogh. 2001. Phytoplankton patterns at CSIRO’s longterm coastal station off Sydney. Journal of Coastal Research 34:60-73. Altabet, M.A., and J.J. McCarthy. 1986 Vertical patterns in 15N natural abundance in PON from the surface waters of warm-core rings. Journal of Marine Research 44:185-201. Anderson, J.T. 1988. A review of size-dependent survival during pre-recruit stages of fish in relation to recruitment. Journal of Northwest Atlantic Fisheries Science 8:55-66. Anderson, J.T. 1994. Sydney Ocean Studies: Nutrient Assessment for Sydney Coastal Waters. AWACS Technical Report 94/12. Arthur, D.K. 1976. Food and feeding of larvae of three fishes occurring in the California Current, (Sardinops sagax), (Engraulis mordax), and (Trachurus symmetricus). U.S. NMFS Fishery Bulletin 74:517-530. Bailey, K.M and E.D. Houde. 1989. Predation on eggs and larvae of marine fishes and the recruitment problem. Advances in Marine Biology 25:1-83 Bailey, K.M. and S.J. Picquelle. 2002. Larval distribution of offshore spawning flatfish in the Gulf of Alaska: Potential transport pathways and enhanced onshore transport during ENSO events. Marine Ecology Progress Series 236:205-217. Bailey, K.M., M.F. Canino, J.F. Napp, S.M. Spring and A.L. Brown. 1995. Contrasting years of prey levels, feeding conditions and mortality of larval walleye pollock Theragra chalcogramma in the western Gulf of Alaska. Marine Ecology Progress Series 119:11-23. Batty, R. S. 1987. Effect of light intensity on activity and food-searching of larval herring, (Clupea harengus): a laboratory study. Marine Biology 94:323-328. Beaudoin, C.P., W.M. Tonn, E.E. Prepas, and L.I. Wassenaar. 1999. Individual specialisation and tropic adaptability of northern pike (Exos lucius) : an isotope and dietary analysis. Oecologia 120:386-396. Beckley, L.E. 1993. Linefish larvae and Agulhas Current. Pages 57-63. In: Beckley, L.E. and R.P. van der Elst (eds.), Fish, fishes and fisheries proceedings of the 2nd Linefish Symposium, Durham. Bell, J.D., M. Westoby and Steffe, A.S. 1987. Fish larvae settling in seagrass: Do they discriminate between beds of different leaf density. Journal of Experimental Marine Biology and Ecology 111(2):133-144. Bergenius, M.I., M.I. Meekan, R. Robertson, and M.I. McCormick. 2002. Larval growth predicts the recruitment success of a coral reef fish. Oecologia 131(4):521-525. Blaxter, J. H. S., and J. R. Hunter. 1982. The biology of the clupeoid fishes. Pages 1- 223. In: J. H. S. Blaxter, F. S. Russel, and M. Yonge, editors. Advances in Marine Biology. Academic Press, London. Bone, Q., N.B. Marshall and J.H.S. Blaxter. 1995. Biology of fishes. Chapman and Hall, Glasgow. Booth, D.J. 2002a. Distribution changes after settlement in 6 species of damselfish (Pomacentridae) in One Tree Island lagoon, Great Barrier Reef. Marine Ecology Progress Series 226:157-164. Booth, D.J. 2002b. Larval supply, condition and persistence of the coral reef fish, Pomacentrus moluccensis. Pages 463–468. In: M.K. Moosa, S. Soemodihardjo, 114

A. Nontji, A. Soegiarto, K. Romimohtarto, Sukarno and Suharsono (Eds.) Proceedings of the Ninth International Coral Reef Symposium, Bali, Indonesia, October 23-27 2000. Published by the Ministry of Environment, the Indonesian Institute of Sciences and the International Society for Reef Studies. Booth, D.J. and M.A. Hixon. 1999. Food ration and condition affect early survival of the coral reef damselfish, Stegastes partitus. Oecologia 121:364-368 Booth, D.J. and R.A. Alquezar. 2002. Food supplementation increases larval growth, condition and survival of the brooding damselfish Acanthochromis polyacanthus (Pomacentridae). Journal of Fish Biology 60:1126-1133. Bruce, B. D. 1988. Larval development of blue grenadier (Macruronus novaezelandiae) (Hector), in Tasmanian waters. Fishery Bulletin U.S. 86:199-128. Bruce, B. D., and D. A. Short. 1992. Observations on the distribution of larval fish in relation to a frontal zone at the mouth of Spencer Gulf, South Australia. Pages 124-137. In: D.A Hancock (ed), Larval Biology, Bureau of Rural Resources Proceedings 15. Australian Government Publishing Service, Canberra. Bruce, B.D, F.J. Neira and R.W. Bradford. 2001a. Larval distribution and abundance of blue and spotted warehous (Seriolella brama and S. punctata: Centrolophidae) in south-eastern Australia. Marine and Freshwater Research 52(4):631-636. Bruce, B.D. 1995. Larval development of King George whiting, Sillaginodes punctata, school whiting, Sillago bassensis, and yellow fin whiting, Sillago schomburgkii (Percoidei: Sillaginidae), from South Australian waters. Fishery Bulletin 93(1): 27-43. Bruce, B.D., K. Evans, C.A. Sutton, J.W. Young and D.M Furlani. 2001b. Influence of mesoscale oceanographic processes on larval distribution and stock structure in jackass morwong (Nemadactylus macropterus: Cheilodactylidae). ICES Journal of Marine Science 58(5):1072-1080. Bruce, B.D., S.A. Condie and C.A. Sutton. 2001c. Larval distribution of blue grenadier (Macruronus novaezelandiae Hector) in south-eastern Australia: further evidence for a second spawning area. Marine and Freshwater Research 52(4):603-610. Buckley, L.J. 1979. Status of the RNA/DNA ratio technique for the evaluation of physiological condition and growth rate in larval fish. NOAA, NMFS Narragansett Lab. Ref. No.79-52. Buckley, L.J. 1984. RNA-DNA ratio: an index of larval fish growth in the sea. Marine Biology 80:291-198. Buckley, L.J. and F.J. Bulow. 1987. Techniques for the estimation of RNA, DNA, and protein in fish. Pages 345-354. In: R.C Summerfelt and G.E. Hall (eds.), Age and growth of fish, Iowa State University Press. Bulow, F.J. 1987. RNA-DNA ratios as indicators of growth in fish: a review. Pages 45- 64. In: R.C Summerfelt and G.E. Hall (eds.), Age and growth of fish, Iowa State University Press. Campana, S.E. 1983. Feeding periodicity and the production of daily growth increments in otoliths of steelhead trout (Salmo gairneri) and starry flounder (Platichthys stellatus). Canadian Journal of Zoology 61:1591-1597. Campana, S.E. 1992. Measurement and interpretation of the microstructure of fish otoliths. Canadian Special Publication of Fisheries and Aquatic Sciences 117:59-71. Campana, S.E. and J.D. Neilson. 1985. Microstructure of fish otoliths. Canadian Journal of Fisheries and Aquatic Sciences 40:1014-1032 115

Campana, S.E., and C.M. Jones. 1992. Analysis of otolith microstructure data. In: Stevenson DK, Campana SE (Eds), Otolith Microstructure Examination and Analysis. Canadian Special Publication Fisheries and Aquatic Sciences, Deptament of Fisheriws and Oceans, Ottawa. Campana, S.E., and P.C.F. Hurley. 1989. An age- and temperature-mediated growth model for cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) larvae in the Gulf of Maine. Canadian Journal of Fisheries and Aquatic Sciences 46: 603-613. Campana, S.E. 1990. How reliable are growth back-calculations based on otoliths?. Canadian Journal of Fisheries and Aquatic Sciences 47(11):2219-2227. Canino, M.F., and E.M. Caldarone. 1995. Modification and comparison of two fluorometric techniques for determining nucleic acid contents of fish larvae. Fishery Bulletin 93:158-165. Caputi, N., W. J. Fletcher, A. Pearce, and C. F. Chubb. 1996. Effect of the Leeuwin Current on the recruitment of fish and invertebrates along the Western Australian coast. Marine and Freshwater Research 47:147-155. Chaves, F.P., and S.L. Smith. 1994. Biological and chemical consequences of open ocean upwelling. Pages 149-170. In: Summerhayes, et al. (eds.), Upwelling in the ocean: modern processes and ancient records. John Wiley and Sons, Chichester. Clarke, K.R. 1993. Non-parametric multivariate analyses of change in community structure. Australian Journal of Ecology l18:117-143 Clarke, K.R., and R.M. Warwick. 1994. Change in marine communities: An approach to statistical analysis and interpretation. Natural Environment Council, UK Clemmesen, C. 1988. A RNA and DNA fluorescence technique to evaluate the nutritional condition of individual marine fish larvae. Meeresforch 32:134-143. Clemmesen, C. 1993. Improvement in the fluorometric determination of the RNA and DNA content of individual marine fish larvae. Marine Ecology Progress Series 100:177-183. Clemmesen, C. 1996. Importance and limits of RNA/DNA ratios as a measure of nutritional condition in fish larvae. Pages 67-82. In: Y. Watanabe, Y. Yamashita and Y. Oozeki (eds), Survival Strategies in early life stages of marine resources, AA Balkema, Rotterdam. Clemmesen, C. and T. Doan. 1996. Does otolith structure reflect the nutritional condition of a fish larva? Comparison of otolith structure and biochemical (RNA/DNA ratio) determined on cod larvae. Marine Ecology Progress Series 138:33-39. Cloern, J.E., E.A. Canuel, E.A., and D. Harris. 2002. Stable carbon and nitrogen isotope composition of aquatic and terrestrial plants of the San Francisco Bay estuarine system. Limnology and Oceanography 47(3):713-729. Cole, J., and J. McGlade. 1998. Clupeoid population variability, the environment and satellite imaginary in coastal upwelling system. Review on Fish Biology and fisheries 8:445-471. Cowan Jr, J.H., and R.F. Shaw. 2002. Recruitment. Pages 88-111. In: L.A. Fuiman and R.G. Werner, Fishery Science: The unique contributions of early life stages, Blackwell, Oxford (UK). Cowen, R.K., J.A. Hare, and M.P. Fahay. 1993. Beyond hydrography: Can physical processes explain larval fish assemblages within the middle Atlantic Bight? Bulletin of Marine Science 53(2):567-587. 116

Cresswell, G.R. 1983. Physical evolution of warm-core eddy. Australian Journal of Marine and Freshwater Research 34:495-513. Cresswell, G.R. 1994. Nutrient enrichment of the Sydney Continental Shelf. Australian Journal of Marine and Freshwater Research 45:677-691. Cresswell, G.R., C. Ellyett, R. Legeckis, A.F. Pearce. 1983. Nearshore features of the East Australian Current system. Australian Journal of Marine Freshwater Research 34:105-114. Creutzberg, F., A. T. G. Eltink, and G. J. van Noort. 1978. The migration of plaice larvae, (Pleuronectes platessa) into the western Wadden Sea. Pages 243-251. In: D.S. McLusky and A.J. Berry (eds.) Physiology and behavior of marine organisms. Pergamon Press, Oxford (UK). Cury P., and C. Roy. 1989. Optimal environmental window and pelagic fish recruitment success in upwelling areas. Canadian Journal of Fisheries and Aquatic Sciences 46:670-680. Cushing, D.H. 1975. Marine ecology and fisheries. Cambridge University Press, London. Cushing, D.H. 1983. Are fish larvae too dilute to affect the density of their food organisms? Journal of Plankton Research 5:847-854. Cushing, D.H. 1990. Plankton production and year-class strength in fish populations: An update of the match/mismatch hypothesis. Advances in Marine Biology 26:250-293. Cushing, D.H. 1996. Towards a science of recruitment in fish populations. Excellence in ecology Vol 7, Ecology Instute, Oldendorf/Luhe. Cushing, D.H., and J.W. Horwood. 1994. The growth and death of fish larvae. Journal of Plankton Research 16:291-300. Davis, T.L.O., G.P. Jenkins and J.W. Young. 1990a. Patterns of horizontal distribution of the larvae of southern bluefin (Thunnus maccoyii) and other tuna in the Indian Ocean. Journal of Plankton Research 12:1295-1314. Davis, T.L.O., G.P. Jenkins, J.W. Young. 1990b. Diel patterns of vertical distribution in larvae of southern bluefin Thunnus maccoyii, and other tuna in the east Indian Ocean. Marine Ecology Progress Series 59:63-74. Davis, T.L.O., V. Lyne and G.P. Jenkins. 1991. Advection, dispersion and mortality of a patch of southern bluefin tuna larvae Thunnus maccoyii in the east Indian Ocean. Marine Ecology Progress Series 73:33-45. de Mendiola, B.R. 1974. Food of larval anchoveta, Engraulis ringens J. Pages 277-285. In: J.H.S. Blaxter (ed). The early life history of fish. Springer-Verlag, Berlin. De Niro, M.J., and S. Epstein. 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42:495-506. De Niro, M.J., and S. Epstein. 1981. Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta 45:341-351. Dela-Cruz J, P. Ajani, R. Lee, T. Pritchard and I.M. Suthers. 2002. Temporal abundance patterns of the red tide dinoflagellate Noctiluca scintillans along the southeast coast of Australia. Marine Ecology Progress Series 236:75-88 Dela-Cruz J. 2002. Population dynamics and ecology of the red tide dinoflagllate Noctiluca scintillans. PhD thesis, University of New South Wales, Sydney. Dela-Cruz, J., J.H. Middleton and I.M. Suthers. 2003. Population growth and transport the red tide dinoflagellate, Noctiluca scintillans, in the coastal waters off Sydney Australia, using cell diameter as a tracer. Limnology and Oceanography 48(2): 656-674 117

Dempster, T., M. T. Gibbs, D. Rissik and I. M. Suthers. 1997. Beyond hydrography: daily ichthyoplankton variability and short term oceanographic events on the Sydney continental shelf. Continental Shelf Research 17:1461-1481. Doherty, P. and T. Fowler. 1994. An empirical test of recruitment limitation in a coral reef fish. Science-(Washington-D-C) 263(5149):935-939. Doherty, P.J. and D.Mc.B. Williams. 1988. The replenishment of coral reef fish populations. Oceanography and Marine Biology Annual Review 26:487-551. Dower, J.F., T.J. Miller and W.C. Leggett. 1997. The role of microscale turbulence in the feeding ecology of larval fish. Advances in Marine Biology 30:169-220. Doyle, M.J., W.W. Morse and A.W. Kendall. 1993. A comparison of larval fish assemblages in temperate zone and the northest Pacific and northwest Atlantic Oceans. Bulletin of Marine Science 52:588-644. Duffy-Anderson, J.T., K.M. Bailey, and L Ciannelli. 2002. Consequences of a superabundance of larval walleye Pollock Theragra chalcogramma in the Gulf of Alaska in 1981. Marine Ecology Progress Series 243:179-190. Eckmann, R. and P. Rey. 1987. Daily increments on the otoliths of larval and juvenile Coregonus spp., and their modification by environmental factors. Hydrobiologia 148:137-143. Ehrlich, K. F. 1974. Chemical changes during growth and starvation of herring larvae. Pages 301-324. In: J. H. S. Blaxter, editor. The early life history of fish. Springer-Verlag, Berlin. Ekau, W., P. Westhaus-Ekau and C. Medeiros. 1999. Large scale distribution of fish larvae in the continental shelf waters off NE-Brazil. Archive of Fishery and Marine Research 47(2/3):183-200. EPA. 1995. Sydney Deepwater Outfalls. Final report Series, Volume 3 (Water Quality). Environmental Protection Authority of New South Wales. Epifanio, C.E. and R.W. Garvine. 2001 Larval transport on the Atlantic continental shelf of north America: A review. Estuarine, Coastal and Shelf Science 52(1):51-77. Fagan, P., A.G. Miskiewicz, and P.M. Tate. 1992. An approach to monitoring sewage outfalls. A case study on the Sydney deepwater sewage outfalls. Marine Pollution Bulletin, 25(5-8):172-180. Fahay, M.P. 1983. Guide to the early stages of marine fishes occurring in the western North Atlantic Ocean, Cape Hatteras to the southern Scotian Shelf. Journal of Northwest Atlantic Fishery Science 4:1-423. Fernandez, E., J. Cabal, J.L. Acuna, A. Bode, A. Botas, C. Garcia-Soto. 1993. Plankton distribution across a slope current induced front in the southern Bay of Biscay. Journal of Plankton Research 15:619-641. Ferron, A. and W.C. Leggett. 1994. An appraisal of condition measures for marine fish larvae. Advances in Marine Biology 30:217-303. Fisher, R. and D.R. Belwood 2002. The influence of swimming speed on sustained swimming performance of late-stage reef fish larvae. Marine Biology 140(4): 801-807. Fisher, R., D.R. Bellwood and R. Job. 2000. Development of swimming abilities in reef fish larvae. Marine Ecology Progress Series 202:163-173. Fives, J.M., S. Acevedo, M. Lloves, A. Whitaker, M. Robinson and P.A. King. 2001. The distribution and abundance of larval mackerel, Scomber scombrus L., horse mackerel, Trachurus trachurus (L.), hake, Merluccius merluccius (L.), and blue 118

whiting, Micromesistius poutassou (Risso, 1826) in the Celtic Sea and west of Ireland during the years 1986, 1989 and 1992. Fisheries Research 50:17-26. Fletcher, W.J. 1999. Vertical distribution of pilchard (Sardinops sagax) eggs and larvae off southern Australia. Marine and Freshwater Research 50(2):117-122. Fletcher, W.J., and N.R. Sumner. 1999. Spatial distribution of sardine (Sardinops sagax) eggs and larvae: an application of geostatistics and resampling to survey data. Canadian Journal of Fisheries and Aquatic Sciences 56(6):907-914. Fletcher, W.J., R.J. Tregonning, and G.J. Sant. 1994. Interseasonal variation in the transport of pilchard eggs and larvae off southern Western-Australia. Marine Ecology Progress Series 111:209-224. Folkvord, A., and J.R. Hunter. 1986. Size-specific vulnerability of northern anchovy, (Engraulis mordax), larvae to predation by fishes. Fishery Bulletin 84(4):859- 870. Fortier L, and W.C. Leggett. 1982. Fickian transport and the dispersal of fish larvae in estuaries. Canadian Journal of Fisheries and Aquatic Sciences 39:1150-1163 Fortier, L. and W.C. Leggett. 1983. Vertical migrations and transport of larval fish in a partially mixed estuary. Canadian Journal of Fisheries and Aquatic Sciences 40: 1543-1555. Fortier, L., M. E. Levasseur, R. Drolet, and J. C. Therriault. 1992. Export production and the distribution of fish larvae and their prey in a coastal jet frontal region. Marine Ecology-Progress Series 85:203-218. Fossum, P. 1996. A study of first-feeding herring (Clupea harengus L) larvae during the period 1985-1993. ICES Journal of Marine Sciences 53:51-59. Fowler, A.J., K.P. Black and G.P. Jenkins. 2000. Determination of spawning areas and larval advection pathways for king george whiting in southern Australia using otolith microstructure and hydrodynamic modeling. II. South Australia. Marine Ecology Progress Series 199:243-254. France, R.L., 1995. Carbon-13 enrichment in bentic compared to planktonic algae: foodweb implications. Marine Ecology Progress Series 124:307-312. Frank, K.T. and W.C. Leggett. 1982. Coastal water mass replacement: Its effect on zooplankton dynamics and the predator-prey complex associated with larval capelin (Mallotus villosus). Canadian Journal of Fisheries and Aquatic Sciences 39:991-1003 Fry, B. 1988. Food web structure on Georges Bank from C, N and S isotopic compositions. Limnology and Oceanography 35:1182-1190. Fry, B. 1999. Using stable isotopes to monitor watershed influences on aquatic trophdynamics. Canadian Journal of Fisheries and Aquatic Sciences 56:2167- 2171. Fry, B., and E.B. Sherr. 1984. δ13C measurements as indicators of carbon flow in marine and freshwater ecosystems. Contributions in Marine Science 27:13-47 Gaston, T.F., and I.M. Suthers. 2004. Spatial variation in the δ13C and δ15N of liver, muscle and bone in a rocky reef planktivorous fish: the relative contribution of sewage. Journal of Experimental Marine Biology and Ecology 304(1):17-33. Gaughan, .J. 1992. Feeding by estuarine and marine larvae? In: D.A Hancock (ed.), Larval biology: Australian Society for Fish Biology Workshop. Bureau of Rural Resources Proceeding 15:37-40. Gaughan, D.J., F.J. Neira, L.E. Beckley and I.C. Potter. 1990. Composition, seasonality and distribution of the ichthyoplankton in the lower Swan Estuary, south- 119

western Australia. Australian Journal of Marine and Freshwater Research 41(4):529-543. Gaughan, D.J., W.R. Fletcher and R.J. Tregonning. 1996. Spatial and seasonal distribution of eggs and larvae of sandy sprat, Hyperlopus vittatus (Clupeidae). Marine and Freshwater Research 47(8):971-979. Gibbs, M.T., J.H. Middleton and P. Marchesiello. 1998. Baroclinic response of Sydney Shelf waters to local and wind and deep ocean forcing. Journal of Physical Oceanography 28:178-190. Godfrey, J.S., G.R. Cresswell, T.J. Golding, and A.F. Pearce. 1980. The separation of the East Australian Current. Journal of Physical Oceanography 10:430-440. Govoni, J.J., and C.B. Grimes. 1992. Surface accumulation of larval fishes by hydrodynamic convergence within the Mississippi River plume front. Continental Shelf Research 12(11):1265-1276. Gray, C.A. and A.G. Miskiewicz. 2000. Larval fish assemblages in South-east Australia coastal waters: Seasonal and Spatial Structure. Estuarine, Coastal and Shelf Science 50:549-570. Gray, C.A. and M.J. Kingsford. 2003. Variability in thermocline depth and strength, and relationship with vertical distributions of fish larvae and mesozooplankton in dynamic coastal waters. Marine Ecology Progress Series 247:211-224. Gray, C.A., N.M. Otway and A.G. Miskiewicz. 1997. Numerical responses of larval fishes to deepwater sewage disposal: A field assessment. Marine Pollution Bulletin 33(7-12):190-200 Gray, C.A., N.M. Otway, F.A. Laurenson, A.G. Miskiewicz and R.L. Pethebridge. 1992. Distribution and abundance of marine fish larvae in relation to effluent plumes from sewage outfalls and depth of water. Marine Biology 113:549-559. Green, B.S and M.I. McCormick. 1999. Influence of larval feeding history on the body condition of Amphiprion melanopus. Journal of Fish Biology 6:1273-1289 Griffin, D.A. and J.H. Middleton. 1991. Local and remote wind forcing of New South Wales inner shelf currents and sea level. Journal of geophysical Research 97: 14389-14405. Grimes, C.B., and M.J. Kingsford. 1996. How do estuarine and riverine plumes influence fish larvae, do they enhance recruitment? Journal of Marine and Freshwater Research 47:191-208. Grothues, T.M, and R.K. Cowen. 1999. Larval fish assemblages and water mass history in a major faunal transition zone. Continental Shelf Research 19:1171-1198. Gunn, J.S. 1990. A revision of selected genera of the family Carangidae (Pisces) from Australian waters. Record of Australian Museum Supplemen 12, Sydney. Hallegraeff, G.M. and S.W. Jeffrey. 1993. Annually recurrent diatom blooms in spring along the New South Wales coast of Australia. Australian Journal of Marine and Freshwater Research 44:325-334. Hamann, I., H-C. John and E. Mittelstaedt. 1981. Hydrography and its effect on fish larvae in the Mauritanian upwelling area. Deep Sea Research 28A(6):561-575. Hansson, S., J.E. Hobbie, R. Elmgren, U. Larrson, B. Fry, and S. Johannsson. 1997. The stable isotope ratio as a marker of food-web interactions and fish migration. Ecology 78:2249-2257. Hare, J. A., M. P. Fahay and R. K. Cowen. 2001. Springtime ichthyoplankton of the slope region off the north-eastern United States of America: larval assemblages, relation to hydrography and implications for larval transport. Fisheries Oceanography 10(2):164-192. 120

Hare, J.A., J.H. Churchill, R.K. R.K. Cowen, T.J. Berger, P.C. Cornillon, P. Dragos, S.M. Glenn, J.J. Govoni and T.N. Lee. 2002. Routes and rates of larval fish transport from the southeast to the northeast United States continental shelf. Limnology and Oceanography 47(6):1774-1789. Heath, M.R. 1992. Field Investigations of the early life stages of marine fish. Advances in Marine Biology 28:1-174. Herzka, S.Z,, S.A. Holt, and G.J. Holt. 2001. Documenting the settlement history of individual fish larvae using stable isotope ratios: model development and validation. Journal of Experimental Marine Biology and Ecology 265(1):49-74. Herzka, S.Z., and G.J. Holt. 2000. Changes in isotopic composition of red drum (Sciaenops ocellatus) larvae in response to dietary shifts: potential applications to settlement studies. Canadian Journal Fisheries and Aquatic Sciences 1:137- 147. Herzka, S.Z., Holt, S.A., and Holt, G.J. 2002. Characterization of settlement patterns of red drum Sciaenops ocellatus larvae to estuarine nursery habitat: a stable isotope approach. Marine Ecology Progress Series 226:143-156. Hewitt, R.P., G.H. Theilacker, and N.C.H. Lo. 1985. Causes of mortality in young jack mackerel. Marine Ecology Progress Series 26:1-10. Hobson, K.A., and H.E. Welch. 1992. Determination of trophic relationships within a high Arctic marine food web using δ13C and δ15N analysis. Marine Ecology Progress Series 84:9-18. Hobson, K.A., and R.G. Clark. 1992. Assessing avian diets using stable isotopes I: Turnover of 13C in tissues. The Condor 94:181-188. Houde, E. D. 1975. Effects of stocking density and food density on survival, growth, and yield of laboratory-reared larvae of the sea bream, Archosargus rhomboidalis (L.) (Sparidae). Journal of Fish Biology 7(1):115-127. Houde, E.D. 1987. Fish early life dynamics and recruitment variability. American Fisheries Society Symposium 2:17-29. Houde, E.D. 1996. Evaluating stage-specific survival during the early life of fish. Pages 51-66. In: Y. Watanabe, Y. Yamashita and Y. Oozeki (eds.), Survival Strategies in early life stages of marine resources, AA Balkema, Rotterdam. Houde, E.D. and E.C. Schekter. 1978. Simulated food patches and survival of larval bay anchovy, Anchoa mitchelli, and sea bream, Archosargus rhomboidalis. Fishery Bulletin NOAA (U.S.) 76(2):483-487. Hunter, J.R. 1981. Feeding ecology and predation of marine fish larvae. Pages 33-77. In: R. Lasker (ed). Marine fish larvae: morphology, ecology, and relation to fisheries. Washington Sea Grant Program, University of Washington Press, Seattle. Hutchings, L., G.C. Pitcher, T.A. Probyn, and G.W. Bailey. 1994. The chemical and biological consequences of coastal upwelling. Pages 65-82. In: Summerhayes, et al. (eds.), Upwelling in the ocean: modern processes and ancient records. John Wiley and Sons, Chichester. James, G.D. 1976. Eggs and larvae of the trevally Caranx georgianus (Teleostei: Carangidae). New Zealand Journal of Marine and Freshwater Research 10(2):301-310 Jeffrey, S.W., D.J. Rochford and G.R. Cresswell. 1990. Oceanography of the Australian Region. Pages 244-265. In: Clayton, M.N. and R.J. King (eds.), Biology of Marine Plants, Longman, Melbourne. 121

Jenkins, G.P. 1986. Composition, seasonality and distribution of ichthyoplankton in Port Phillip Bay, Victoria. Australian Journal of Marine and Freshwater Research 37:507-520. Jenkins, G.P. 1987. Age and growth of co-occurring larvae of two flounder species, Rhombosolea tapirina and Ammotretis rostratis. Marine Biology 95:157-166. Jenkins, G.P. 1991. Comparative diets prey selection and predatory impact of co- occurring larvae of two flounders species. Journal of Experimental Marine Biology and Ecology 110(2):147-170. Jenkins, G.P. 1992. What can growth trajectory tell us about the nutritional state of fish larvae? In: D.A Hancock (ed.), Larval biology: Australian Society for Fish Biology Workshop. Bureau of Rural Resources Proceeding 15:44-48. Jenkins, G.P. and T.L.O. Davis. 1990. Age, growth rate, and growth trajectory determined from otolith microstructure of southern bluefin tuna Thunnus maccoyii larvae. Marine Ecolology Progress Series 63:93-104. Jenkins, G.P., J.W. Young and T.L.O. Davis. 1991. Density dependence of larval growth of a marine fish, the southern bluefin tuna, Thunnus maccoyii. Canadian Journal of Fisheries and Aquatic Sciences 48:1358-1363. Jenkins, G.P., K.P. Black and P.A. Hamer. 2000. Determination of spawning areas and larval advection pathways for king george whiting in southestern Australia using otolith microstructure and hydrodynamic modeling. I. Victoria. Marine Ecology Progress Series 199:231-242. Jenkins, G.P., M.J. Keough and P.A. Hamer. 1998. The contributions of habitat structure and larval supply to broad-scale recruitment variability in a temperate zone, seagrass-associated fish. Journal of Experimental Marine Biology and Ecology 226(2):259-278. Jennings, S., and K.J. Warr. 2003. Environmental correlates of large-scale spatial variation in the delta 15N of marine animals. Marine Biology 142:1131-1140. Jones, C. 1986. Determining age of larval fish with the otolith increment technique. Fishery Bulletin 84(1):91-104. Jones, C.M. 2002. Age and growth. Pages 33-63. In: L.A. Fuiman and R.G. Werner, Fishery Science: The unique contributions of early life stages, Blackwell, Oxford (UK). Jordan, A.R. 2001. Reproductive biology, early life-history and settlement distribution of sand flathead (Platycephalus bassensis) in Tasmania. Marine and Freshwater Research 52(4):589-601 Jordan, A.R., 1994. Age, growth and back-calculated birthdate distributions of larval jack mackerel, Trachurus declivis (Pisces: Carangidae), from eastern Tasmanian coastal waters. Australian Journal of Marine and Freshwater Research 45(1):19- 33. Jordan, A.R., and B.D. Bruce. 1993. Larval development of three roughy species complexes (Pisces: Trachichthyidae) from southern Australian waters, with comments on the occurrence of orange roughy larvae Hoplostethus atlanticus. Fishery Bulletin 91:76-86. Kailola, P., M. Williams, P. Stewart, R. Reichelt, A. McNee and C. Grieve. 1993. Australian Fisheries Resources. Bureau of Resources Sciences, Canberra. Katsuragawa, M. and Y. Matsuura. 1992. Distribution and abundance of carangid larvae in the southern Brazilian Bight during 1975-1981. Boletim do Instituto Oceanografico, Universidade de Sao Paulo, Sao Paulo 40(1/2): 55-78. 122

Kerrigan, B.A. 1997. Variability in larval development of the tropical reef fish Pomacentrus amboinensis (Pomacentridae): The parental legacy. Marine Biology 127(3):395-402. Kingsford, M.J. 1990. Linear oceanographic features: A focus for research on recruitment processes. Australian Journal of Ecology 15:391-401. Kingsford, M.J. and I.M. Suthers. 1994. Dynamic estuarine plumes and fronts: importance to small fish and plankton in coastal waters of N.S.W., Australia. Continental Shelf Research 14(6):655-672. Kingsford, M.J. and I.M. Suthers. 1996. The influence of tidal phase on patterns of ichthyoplankton abundance in the vicinity of an estuarine front, Botany Bay, Australia. Estuarine, Coastal and Shelf Science 43:33-54. Kiriluk, R.M., M.R. Servos, D.M. Whittle, G. Cabana, and J.B. Rasmussen. 1995. Using ratios of stable nitrogen and carbon isotopes to characterize the biomagnification of DDE, Mirex, and PCB in a Lake Ontario pelagic food web. Canadian Journal of Fisheries and Aquatic Sciences 52:2660-2674. Lasker, R. 1981. The role of a stable ocean in larval fish survival and subsequent recruitment. Pages 80-87. In: R. Lasker, editor. Morphology, ecology and relation to fisheries. Univ. Washington Press, Seattle, WA. Lasker, R. 1985. An egg production method for estimating spawning biomass of pelagic fish: application to the northern anchovy, Engraulis mordax. NOAA Technical Report, NMFS 36, 99 pp. Last, J. M. 1980. The food of twenty species of fish larvae in the west central North Sea. Fishery Research Technical Report, MAFF Direct., Fish. Res. Lowestoft 60. 44 pp. Leak, J.C., and E.D. Houde. 1987. Cohort growth and survival of bay anchovy (Anchoa mitchilli) larvae in Biscayne Bay, Florida. Marine Ecology Progress Series 37:109-122. Lee, R.S.; Ajani, P.; Pritchard, T. and Black, K. (2000) Resolving climatic variances in the context of retrospective phytoplankton pattern investigations off the east coast of Australia. International Coastal Symposium 2000, Rotorua, April 24 – 28, 2000, pp 116 – 117. (Abstract) Leggett, W.C. 1986. The dependence of fish larval survival on foods and predator densities. In: S. Skreslet, (ed.), The role of freshwater outflow in coastal marine ecosystems. Bodo, 21-25 May 1985, NATO ASI Series, Series G: Ecological Sciences, Berlin, 7:107-117. Leggett, W.C., and E. Deblois. 1994. Recruitment in marine fishes: Is it regulated by starvation and predation in the egg and larval stages? Netherland Journal of Sea Research 32(2):119-134 Leiby, M.M. 1984. Life history and ecology of pelagic fish eggs and larvae. Pages 121- 140. In: K.A. Steidinger and L.M. Wakler (eds.), Marine Plankton Life Cycle Strategies. CRC Press, Boca Raton, Florida. Leis, J.M. 1991. Vertical distribution of fish larvae in the Great Barrier Reef lagoon, Australia. Marine Biology 109:157-166 Leis, J.M. and B. Carson-Ewart. 2000. The larvae of Indo-Pacific coastal fishes: An identification guide to marine fish larvae. Brill, Leiden. Leis, J.M. and B. Goldman. 1987. Composition and distribution of larval fish assemblages in the Great Barrier Reef Lagoon, near Lizard Island. Australia Journal of Marine and Freshwater Research 38:211-224. 123

Leis, J.M. and B.M. Carson-Ewart. 1997. In situ swimming speeds of the late pelagic larvae of some Indo-Pacific coral-reef fishes. Marine Ecology Progress Series 159(1):175-174. Lindsay, D.J., M. Minagawa, I. Mitani, and K. Kawaguchi. 1998. Trophic shift in the Japanese anchovy Engraulis japonicus in its early life history stage as detected by stable isotope ratios in Sagami Bay, Central Japan. Fishery Science 64(3):403-410. MacKenzie, B.R. and W.C. Leggett. 1991. Quatiying the contribution of small scale turbulence to the encounter rates between larval fish and their zooplankton prey effects of wind and tide. Marine Ecology Progress Series 73 (2-3):149-160. Manly Hydraulic Laboratory. 1997. Review of sewage effluent ocean outfall performance in NSW. Department of Land and Water Conservation, NSW. Mann, K.H. and J.R.N. Lazier. 1996. Dynamics of marine ecosystems: Biological- physical interactions in the oceans. 2nd Edition. Blackwell Science, Inc, Massachusetts. May, R.C. 1974. Larval mortality in marine fishes and the critical period concept. In: J.H.S. Blaxter (ed.), The early life history of fish, Springer-Verlag Berlin. pp. 3- 20. McCormick, M.I., and B.W. Molony. 1995. Influence of water temperature during the larval stage on size, age and body condition of a tropical reef fish at settlement. Marine Ecology Progress Series 118(1-3):59-68. Meekan, M.G., J.H. Carleton, A.D. McKinno, K. Flynn and M. Furnas. 2003. What determines the growth of tropical reef fish larvae in the plankton: Food or temperature? Marine Ecology Progress Series 256:193-204. Meekan, M.G., M.J. Milicich, and P.J. Doherty. 1993. Larval production drives temporal patterns of larval supply and recruitment of a coral reef damselfish. Marine Ecology Progress Series 93:217-225. Middleton, J.H., D. Cox and P. Tate. 1997. The oceanography of the Sydney region. Marine Pollution Bulletin 33(7-12):124-131. Migawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: Further evidence and the relation between δ15N and animal age. Gechimica et Cosmochimica Acta 48:1135-1140 Milicich, M.J. and P.J. Doherty. 1994. Larval supply of coral reef fish populations - magnitude and synchrony of replenishment to Lizard Island, Great Barrier Reef. Marine Ecology Progress Series 110:121-134. Miller, T. J., L. B.Crowder, J. A. Rice, and F. P. Binkowski. 1992. Body size and the ontogeny of the functional response in fishes. Canadian Journal of Fisheries and Aquatic Sciences 49:805-812. Minami, T. and T. Tamaki. 1980. Offshore and nearshore distribution patterns of fish larvae along the San-In Coast, the Sea of Japan. Japanese Journal of Ichthyology 27(2):156-164 Miskiewicz, A.G. 1987. Taxonomy and ecology of fish larvae in Lake Macquarie and New South Wales coastal waters. PhD thesis. University of New South Wales, Sydney. Miskiewicz, A.G., B.D. Bruce and P. Dixon. 1996. Distribution of tailor (Pomatomus saltatrix) larvae along the coast of New South Wales, Australia. Marine and Freshwater Research 47(2):331-336. Moser H, R. Charter, P. Smith, D. Ambrose, S. Charter, C. Myer, E. Sandknop and W. Watson. 1993. Distributional atlas of fish larvae and eggs in the Californian 124

Current region: Taxa with 1000 or more total larvae, 1951 through (1984) (Atlas no 31). California Cooperative Oceanic Fisheries Investigations, California. Moser H, R. Charter, P. Smith, D. Ambrose, S. Charter, C. Myer, E. Sandknop and W. Watson. 1994. Distributional atlas of fish larvae in the Californian Current region: Taxa with less than 1000 total larvae, 1951 through (1984) (atlas no 31). California Cooperative Oceanic Fisheries Investigations, California. Muelbert, J.H., M.R. Lewis and D.E. Kelley. 1994. The importance of small-scale turbulence in the feeding of herring larvae. Journal of Plankton Research 16(8): 927-944. Neira, F.J. and I.C. Potter. 1992. The ichthyoplankton of a seasonally closed estuary in temperate Australia: Does an extended period of opening influence species composition? Journal of Fish Biology 41:935-953. Neira, F.J. and I.C. Potter. 1994. The larval fish assemblage of the Nornalup-Walpole Estuary, a permanently open estuary on the southern coast of Western Australia. Australian Journal of Marine and Freshwater Research 45:1193-1207. Neira, F.J., A.G. Miskiewicz and T. Trinski. 1998. Larvae of temperate Australian fishes; Laboratory guide for larval fish identification. University of Western Australian Press, Nedlands. Neira, F.J., I.C. Potter and J.S. Bradley. 1992. Seasonal and spatial changes in the larval fish fauna within a large temperate Australian estuary. Marine Biology 112:1- 16. Neira, F.J., M.I. Sporcic and A.R. Longmore. 1999. Biology and fishery of pilchard, Sardinops sagax (Clupeidae), within a large south-eastern Australian bay. Marine and Freshwater Research 50(1): 43-55. Newton, G.M. 1996. Estuarine ichthyoplankton ecology in relation to hydrology and zooplankton dynamics in a salt-wedge estuary. Marine and Freshwater Research 47(2): 99-111. Nilsson, C.S., G.R. Cresswell. 1980. Formation and evolution of EastAustralian Current warm core eddies. Progress in Oceanography 9(3):133-184. Nonaka, R.H., Y. Matsuura and K. Suzuki. 2000. Seasonal variation in larval fish assemblages in relation to oceanographic condition in the Abrolhos Bank region off eastern Brazil. Fishery Bulletin 98:767-784. Ochiai, A., K. Mutsutani, S. Umeda. 1982. Development of eggs, larvae and juveniles of jack mackerel, Trachurus japonicus. Japanese Journal of Ichthyology, 29(1): 86-92. O'Connell, C.P. 1980. Percentage of starving northern anchovy, Engraulis mordax, larvae in the sea as estimated by histological methods. U.S. NMFS Fishery Bulletin 78:475-489. Ohman, M.C., P.L. Munday, G.P. Jones and M.J. Caley. 1998. Settlement strategies and distribution patterns of coral-reef fishes. Journal of Experimental Marine Biology and Ecology 225(2): 219-238. Oiestad, V. 1985. Predation on fish larvae as a regulatory force, illustrated in mesocosm studies with large groups of larvae. NAFO SCI. COUNC. STUD, 8:25-32. Oke, P.R. and J.H. Middleton. 2000. Topographically induced upwelling off eastern Australia. Journal of Physical Oceanography 30:512-531. Oke, P.R. and J.H. Middleton. 2001. Nutrient enrichment off Port Stephens: The role of the East Australian Current. Continental Shelf Research 21:587-606. 125

Olivar, M.P. 1990. Spatial patterns of ichthyoplankton distribution in relation to hydrographic features in the northern Benguela region. Marine Biology 106:39- 48. Olivar, M.P. and A. Sabates. 1997. Vertical distribution of fish larvae in the north-west Mediterranean Sea in spring. Marine Biology 129:289-300. Olivar, M.P. and P.A. Shelton 1993. Larval fish assemblages of the Benguela Current. Bulletin of Marine Science 53(2): 450-474 Owen, N.J.P. 1987 Natural variations in 15N in the marine environment. Advances in Marine Biology 24:389-451. Parsons, T.R., Y. Maita and C.M. Lalli. 1984. Effect of nutrient loading in seawater as observed in experimental ecosystems. Umi/la mer. Tokyo 22(3-4): 90-94 Pepin, P., S. Pearre Jr. and J.A. Koslow. 1987. Predation on larval fish by Atlantic mackerel, (Scomber scomberus), with a comparison of predation by zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 44:2012-2018. Peterson, B.J., and R.W. Howarth. 1987. Sulfur, carbon, and nitrogen isotopes used to trace organic matter flow in the salt-marsh estuaries of Sapelo Island, Georgia. Limnology and Oceanography 32:1195-1213. Peterson, B.J., R.W. Howarth, and R.H. Garritt. 1985. Multiple stable isotope used to trace the flowof organic matter in estuarine food webs. Science 227:1361-1363. Peterson, I. and J.S. Wroblewski. 1984. Mortality rate of fishes in the pelagic ecosystem. Canadian Journal of Fisheries and Aquatic Sciences 41(7):1117- 1120. Peterson, K.1996. Effect of prey abundance and temperature on the otolith microstructure of larval fish. Honours thesis, School of Biological Science, the University of New South Wales, Sydney. Peterson, W.T., D.F. Arcos, G.B. McManus, H. Dam, D. Bellantoni, T. Johnson and P. Tiselius. 1988. The nearshore zone during coastal upwelling: Daily variability and coupling between primary and secondary production off central Chile. Progress in Oceanography 20:1-20. Pitts, P.A. 1999. Effects of summer upwelling in the abundance and vertical distribution of fish and crustacean larvae off central Florida's Atlantic coast. Journal of Experimental Marine Biology and Ecology 235:135-146. Poulet, S.A., and R. Williams. 1991. Characteristics and properties of copepods affecting the recruitment of fish larvae. Bulletin of Plankton Society of Japan. Special Publication, pp 271-290. Powell, A.B. and A.J. Chester. 1985. Morphometric indices of nutritional condition and sensitivity to starvation of spot larvae. Transactions of the American Fisheries Society 114(3):338-347. Primer-E Ltd. 2001. Primer 5 for windows. Plymouth Marine Laboratory, Plymouth (UK). Pritchard, T., P. Ajani, D. Andrew, M. Calfas, C. Holden, R. Lee, S. Linforth, and P. Rendell. 1998. Relative significance of slope water, estuarine discharges ans sewage outfalls for nutrients in offshore coastal waters. Coastal nutrients workshop. Australian Water and Wastewater Association Incorporated, Sydney. Pritchard, T., P. Rendell, P. Scanes, and N. Phillip. 1996. Assessment of the deepwater outfalls. Vol. 1. EMP Final Report Series. Environmental Protection Authority, Sydney, 194 p. 126

Pritchard, T., R. Lee, P. Ajani, P. Rendell and K. Black. 2001. How do ocean outfalls affect nutrient patterns in coastal waters of New South Wales, Australia? Journal of Coastal Research 34:60-73. Purcell, J. E. 1981. Feeding ecology of Rhixophysa eysenhardti, siphonophore predation of fish larvae. Limnology and Oceanography 26:424-432. Purcell, J. E. 1984. Predation on fish larvae by Physalia physalia, the Portugese-man- of-war. Mar. Ecol. Prog. Series 19(1-2):189-191. Purcell, J. E. 1985. Predation on fish eggs and larvae by pelagic cnidarians and ctenophores. Bulletin of Marine Science 37:739-755. Rau, G.H., A.J. Mearns, D.R. Young, R.J. Olson, H.A. Schafer, and I.R. Kaplan. 1983. Animal 13C/12C correlates with trophic level in pelagic food webs. Ecology 64:1314-1318. Rau, G.H., R.E. Sweeney, I.R. Kaplan, A.J. Mearns, and D.R. Young. 1981. Diffrences in animals 13C, 15N, D abundances between a polluted and an polutted coastal site : likely indicators of sewage uptake by a marine food web. Estuarine, Coastal and Shelf Science 13:701-707. Rau, G.H., S. Ralston, J.R. Southon and F.P. Chavez. 2001. Upwelling and the condition and diet of juvenile rockfish: A study using 14C, 13C, and 15N natural abundances. Limnology and Oceanography 46(6):1565-1570. Reiss, A.A., C.T. Taggart, J.F. Dower, and B. Ruddick. 2002. Relationships among vertically structured in situ measures of turbulence, larval fish abundance and feeding success and copepods on Western Bank, Scotian Shelf. Fisheries Oceanography 11(3):156-174. Reiss, C.T., G. Panteleev, C.T. Taggart, J. Sheng and B. de Young. 2000. Observations on larval fish transport and retention on the Scotian Shelf in relation to geostrophic circulation. Fisheries Oceanography 9(3): 195-213. Rendell, P.S. and T.R. Pritchard, 1996. Physiochemical conditions in coastal waters off Sydney, Central NSW, Australia. Marine Pollution Bulletin 33:132-139. Richards, W.J., M.G. McGowan, J. Leming, T. Lamkin, S. Kelly. 1993. Larval fish assemblages at the loop current boundary in the Gulf of Mexico. Bulletin Marine Science 53(2):475-537. Rissik, D. and I.M. Suthers. 1996. Feeding in a larval fish assemblage: the nutritional significance of an estuarine plume front. Marine Biology 125:233-240. Rochford, D.J. 1975. Nutrient enrichment of East Australian Coastal Waters. II* Laurieton upwelling Australian Journal of Marine and Freshwater Research 26: 233-243. Rochford, D.J. 1984. Nitrates in Eastern Australia Coastal Waters. Australian Journal of Marine and Freshwater Research 35:385-397. Rothchild, B.J. and T.R. Osbron. 1988. Small-scale turbulence and plankton contact rates. Journal of Plankton Research 10:465-174. Roughan, M. and J.H. Middleton. 2002. A comparison of observed upwelling mechanisms off the east coast of Australia. Continental Shelf Research 22:2551- 2572. Saka, S., K. Firat, and C. Suzer. 2002. Effects of light intensity on early life development of gilthead sea bream larvae (Sparus aurata). Israeli Journal of Aquaculture - Bamidgeh 53(3-4):139-146. Santos, A.M.P, M.F. Borges and S. Groom. 2001. Sardine and horse mackerel recruitment and upwelling off Portugal. ICES Journal of Marine Science 58:589-596. 127

Sepulveda, A. 1994. Daily growth increments in the otoliths of European smelt Osmerus eperlanus larvae. Marine Ecology Progress Series 108: 33-42. Shaw, M. and G.P. Jenkins. 1992. Spatial variation in feeding, prey distribution and food limitation of juvenile flounder Rhombosolea tapirina Guenther. Journal of Experimental Marine Biology and Ecology 165(1):1-21. Shepherd, J.G. and D.H. Cushing. 1980. A mechanism for density-dependent survival of larval fish as a basis of a stock-recruitment relationship. Journal du Conseil International pour l’Exploration de la Mer 39:160-167. Sholto-Douglas, A.D., J.G. Field, A.G. James, N.J. van der Merwe. 1991. 13C/12C and 15N/14N isotope ratios in the Southern Benguela ecosystem: Indicators of food web relationships among different size-classes of plankton and pelagic fish; differences between fish muscle and bone collagen tissues. Marine Ecology Progress Series 78(1):23-32. Sinclair, M. and T.D. Iles. 1985. Atlantic herring (Clupea harengus) distributions in the Gulf of Maine-Scotian Shelf area in relation to oceanographic features. Canadian Journal of Fisheries and Aquatic Sciences 42:880-887. Sinden, G. 1996. The northern rivers: A water quality assessment. Environment Protection Authority, NSW. Smith, K.A, M. Gibbs, J.H. Middleton, and I.M. Suthers. 1999 Short term dynamics of coastal ichthyoplankton off the Sydney coast. Marine Ecology Progress Series 178:1-15. Smith, K.A. 2000. Active and passive dispersal of Centroberyx affinis (Berycidae) and Gonorynchus greyi (Gonorynchidae) larvae on the Sydney shelf. Marine and Freshwater Research 51(3):229-234. Smith, K.A. 2003. Larval distribution of some commercially valuable fish species over the Sydney continental shelf. Proceedings of the Linnean Society of New South Wales 124:1-11. Smith, K.A. and I.M. Suthers 2000. Consistent timing of juvenile fish recruitment to seagrass beds within two Sydney estuaries. Marine and Freshwater Research 51(8):765-776. Smith, K.A. and Suthers, I.M. 1999. Displacement of Sydney shelf ichthyoplankton by a coastal upwelling event. Marine Ecology Progress Series 176:49-62 Sogard, S. M. 1991. Interpretation of otolith microstructure in juvenile winter flounder (Pseudopleuronectes americanus)-ontogenetic development, daily increment validation, and somatic growth relationships. Canadian Journal of Fisheries and Aquatic Sciences 48:1862-1871. SPCC. 1991. Sydney Deepwater Outfalls: Pre commisioning Phase Progress Report, September 1990. State Pollution Control Commission. pp 121-122. Spies, R.B., H. Kruger, R. Ireland, and D.W. Rice Jr. 1989. Stable isotopes ratios and contaminant concentrations in a sewage-distorted food web. Marine Ecology Progress Series 54:157-170 SPSS Inc. 2000. Systat 10. SPSS Science Marketing Department, Chicago. Steffe, A.S. and B.S. Pease. 1987. Diurnal survey of ichthyoplankton abundance distribution and seasonality in Botany Bay New South Wales. Proceedings of the Linnean Society of New South Wales. 110(1-2):1-10. Steidinger, K.A. and L.M. Walker. 1984. Marine plankton life cycle strategies. CRC Press, Inc, Florida. 128

Stephenson, R.L., and M.J. Power. 1988. Semidiel vertical movements in Atlantic herring (Clupea harengus) larvae: A mechanism for larval retention? Marine Ecology Progress Series 50:3-11. Stevenson, D.K., and S.E. Campana. 1992. Otolith microstructure examination and analysis. Canadian Special Publication of Fisheries and Aquatic Sciences 117, Departement of Fisheries and Oceans, Ottawa. Stewart, J. and D.J. Ferrell. 2001. Age, growth, and commercial landings of yellowtail scad (Trachurus novaezelandiae) and blue mackerel (Scomber australasicus) off the coast of New South Wales, Australia. New Zealand Journal of Marine and Freshwater Research 35(3):541-551. Stewart, J., D.J. Ferrell and N.L. Andrew 1999. Validation of the formation and appearance of annual marks in the otoliths of yellowtail (Trachurus novaezelandiae) and blue mackerel (Scomber australis) In New South Wales. Marine and Freshwater Research 50:389-395. Sudekum, A.E., J.D. Parrish, R.L. Radtke, and S. Ralston. 1991. Life history and ecology of large jacks in undisturbed, shallow, oceanic communities. US Fishery Bulletin 89(3): 493-513. Sundby, S. and P. Fossum. 1990. Feeding conditions of Arcto-Norwegian cod larvae compared to the Rothschild-Osborn theory on small-scale turbulence and plankton contact rates. Journal of Plankton Research 12:1153-1162. Suthers, I.M. 1996. Spatial variability of recent otolith growth and RNA indices in juvenile Diaphus kapalae (Myctophidae): An effect of flow disturbance near an island? Marine and Freshwater Research 47:273-282. Suthers, I.M. 1998. Bigger? Fatter? or is faster growth better? Considerations on condition in larval and juveniles coral-reef fishes. Australian Journal of Ecology 23: 265-273. Suthers, I.M. and S. Sundby. 1993. Dispersal and growth of pelagic juvenile Arcto- Norwegian cod (Gadus morhua), inferred from otolith microstructure and water temperature. ICES Journal of Marine Science 50:261-270. Suthers, I.M. and S. Sundby. 1996. Role of the midnight sun: comparative growth of pelagic juvenile cod (Gadus morhua) from the Arcto-Norwegian and a Nova Scotian stock. ICES Journal of Marine Science 53:827-836. Suthers, I.M., J.J. Cleary, S.C. Battaglene, and R. Evans. 1996. Relative RNA content as a measure of condition in larval and juvenile fish. Marine and Freshwater Research 47:301-307. Suthers, I.M., K.T. Frank and S.C. Campana. 1989. Spatial comparison of recent growth in post-larval cod (Gadus morhua) off southwestern Nova Scotia: inferior growth in a presumed nursery area. Canadian Journal of Fisheries and Aquatic Sciences (Supplement 1) 46:113-124. Suthers, I.M., T. van de Meeren, and K.E. Jørstad. 1999. Growth histories of three Norwegian cod stocks co-reared in mesocosms, derived from otolith microstructure; the effect of prey size. ICES Journal of Marine Science 56:658- 672. Taggart, C.T. and W.C. Leggett. 1987. Wind-forced hydrodynamics and their interactions with larval fish and plankton abundance: A time-series analysis of physical-bilogical data. Canadian Journal of Fisheries and Aquatic Sciences 44:438-451 Tandler, A. and C. Mason. 1983. Light and food density effects on growth and survival of larval gilthead seabream (Sparus aurata L.; Sparidae). Pages 103-116. In: R. 129

R. Stickney and S. P. Meyers, editors. Proc. Warmwater fish culture workshop, World Mariculture Soc. Spec. Publ. 3, Louisiana State Univ., Baton Rouge, LA. Theilacker, G.H. 1980. Rearing container size affects morphology and nutritional condition of larval jack mackerel, Trachurus symmetricus. Fishery Bulletin 78(3):789-791. Theilacker, G.H. 1986. Starvation-induced mortality of young sea-caught jack mackerel, Trachurus symmetricus, determined with histological and morphometrical methods. Fishery Bulletin U.S. 84(1):1-17. Theilacker, G.H. and Y. Watanabe. 1989. Midgut cell height defines nutritional status of laboratory raised larval northern anchovy, Engraulis mordax. Fishery Bulletin U.S. 87: 457-469. Thomas, A.C. and W.J. Emery. 1986. Winter hydrography and plankton distribution on the southern British Columbia continental shelf. Canadian Journal of Fisheries and Aquatic Sciences 43:1249-1258. Thornton, S.F., and J. McManus. 1994. Applications of organic carbon and nitrogen stable isotope and C/N ratios as sources indicators of organic matter provenance in estuarine systems: Evidence from the Tray Estuary, Scotland. Estuarine, Coastal and Shelf Science 38:219-233. Thresher, R.E., B.D. Bruce, D.M. Furlani and J.S. Gunn. 1988. Distribution, advection and growth of larvae of the southern temperate gadoid (Macruronus novaezelandiae) in Australian coastal waters. Fishery Bulletin U.S. 87:29-48. Tranter, D.J., D.J. Carpenter and G.S. Leech. 1986. The coastal enrichment effect of the East Australian Current eddy field. Deep-Sea Research 33(11/12):1705-1728. Tremblay, M.J. and J.C. Roff. 1983. Community gradients in the Scotia shelf zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 40:598-611. Trnski, T. 1998. Carangidae: Trevallys, jacks. Pages 192-203. In: Neira, F.J, G.A. Miskiewicz, and T. Trinski (eds), Larvae of temperate Australian fishes, University of Western Australia Press, Nedlands. Trnski, T. 2001. Diel and tidal abundance of fish larvae in a barrier-estuary channel in New South Wales. Marine and Freshwater Research 52:995-1006. Underwood, A.G., M.G. Chapman, S.A. Richards and M.B. Sage. 1997. GMAV 5. Institute of Marine Ecology, University of Sydney, Sydney. Van der Zanden, M.J., M. Hulshof, M.S. Ridgway, and J.B. Rasmussen. 1998. Application of stable isotope techniques to tropic studies of age-0 smallmouth bass. Transaction of American Fishery Society 127:729-739. van Dover, C.L., J.F. Grassle, B. Fry, R.H. Garritt, and V.R. Starczak. 1992. Stable isotope evidence for entry of sewage-derived organic material into a deep-sea food web. Nature 360:153-156. Waldron, S., P. Tatner, I. Jack, and C. Arnott. The impact of sewage discharge in a marine embayment: A stable isotope reconnaissance. Estuarine, Coastal and Shelf Science 52(1):111-115. Ware, D.M. and R.E. Thomson. 1991. Link between long-term variability in upwelling and fish production in the Northeast Pacific Ocean. Canadian Journal of Fisheries and Aquatic Sciences 48:2296-2306. Watanabe, Y. 2002. Resurgence and decline of the Japanese sardine population. Pages 243-256. In: L.A. Fuiman and R.G. Werner, Fishery Science: The unique contributions of early life stages, Blackwell, Oxford (UK). 130

Williams, D.McB., P. Dixon, and S. English. 1988. Cross-shelf distribution of copepods and fish larvae across the central Great Barrier Reef. Marine Biology 99:577- 589. Willis, S.E., L.J.B. Laurenson, B.D. Mitchell and D.J. Harrington. 1999. Diet of larval and juvenile black bream, Acanthopagrus butcheri, in the Hopkins River Estuary, Victoria, Australia. Proceedings of the Royal Society of Victoria 111(2):283-295. Worthington, D.G., M. Westoby, and J. D. Bell. 1991. Fish larvae settling in seagrass effects of leaf density and an epiphytic alga. Australian Journal of Ecology 16: 289-293. Wu, J.P., S.E. Calvert, and C.S. Wong. 1999. Carbon and nitrogen isotope ratios in sedimenting particulate organic matter at an upwelling site off Vancouver Island. Estuarine, Coastal and Shelf Science 48:193-203. Wyatt, T. 1972. Some effects of food density on the growth and behavior of plaice larvae. Marine Biololgy 14(3):210-216. Wyrtki, K. 1960: Surface circulation in the Coral and Tasman Seas. CSIRO Division of Fishery Oceanography, Technical Paper 8, 44 pp. Young, J.W. 1992. Feeding ecology of marine fish larvae: An Australian perspective. In: D.A Hancock (ed.), Larval biology: Australian Society for Fish Biology Workshop. Bureau of Rural Resources Proceeding 15:30-36. Young, J.W. and T.L.O. Davis. 1990. Feeding ecology of larvae of southern bluefin, albacore and skipjack tunas (Pisces, Scombridae) in the eastern Indian Ocean. Marine Ecology Progress Series 61:17-29 Young, J.W. and T.L.O. Davis. 1992. Feeding ecology and interannual variations in diet of larval jack mackerel, Trachurus declivis (Pisces, Carangidae), from coastal waters of eastern Tasmania. Marine Biology 113:11-20. Young, J.W., A.R. Jordan, C. Bobbi, R.E. Johannes, K. Haskard and G. Pullen. 1993. Seasonal and interannual variability in krill (Nyctiphanes australis) stocks and their relationships to their fishery for jack mackerel (Trachurus declivis) off eastern Tasmania, Australia. Marine Biology 116(1):9-18. Young, P.C., J.M. Leis and H.F. Hausfeld. 1986. Seasonal and spatial distribution of fish larvae in waters over the north west continental shelf of Western Australia. Marine Ecology Progress Series 31:209-222. Zeldis, J.R. 1992. Ichthyoplankton studies for fisheries research. ? In: D.A Hancock (ed.), Larval biology: Australian Society for Fish Biology Workshop. Bureau of Rural Resources Proceeding 15:19-26. 131

APPENDIX 1

The recent growth rate of larval pilchards, Sardinops sagax in relation to their stable isotope composition, in an upwelling zone of the East Australian Current

By Shinji UeharaA,B, Augy Syahailatua, and Iain M. SuthersC

ASchool of Biological, Earth and Environmental Sciences, University of New South Wales, Sydney. 2052 Australia. B Coastal Fisheries and Aquaculture Division, Tohoku National Fisheries Research Institute (TNFRI),Shinhamacho, Shiogama, Miyagi 985-0001, Japan

Marine & Freshwater Research (accepted) 132

Abstract. The recent growth rate and stable isotope composition of larval pilchards, (Sardinops sagax, 6-29 mm standard length), were examined in coastal upwelling and non-upwelling regions of the East Australian Current over two cruises during the austral summer of 1998/1999. Pilchards were growing significantly slower in early summer (0.64 mm d-1) than in mid summer (0.98 mm d-1), which could partly be attributed to microzooplankton abundance but not water temperature. Compared to the non-upwelled regions, larvae were larger in the upwelling regions, and yet the back- calculated recent growth over two days prior to capture was significantly less on both cruises. This surprising result is consistent with slower larval growth of this species near coastal Japan and California, where strong year classes may form in offshore waters. δ15N ratios were significantly correlated with larval length. On one cruise, slower growers in upwelled waters were enriched in δ15N and depleted in δ13C indicating dietary differences. The pilchard’s early life history off the eastern Australia is proposed and compared to that off eastern Japan.

Extra keywords: age, sagitta, recent otolith growth, fish larvae, condition, back- calculation, western boundary current, daily growth increments, Kuroshio, Tasman Front 133

Introduction

Year class strengths of planktivores such as herring, pilchard or anchovy are frequently stimulated by coastal upwelling during the spawning season, in association with favourable winds (Cury and Roy 1989; Ware and Thompson 1991). The actual process that enhances recruitment is unclear, and can vary non-linearly with the strength of winds (Cury and Roy 1989) and with the locality. Response to upwelling is season- specific for sardine (Sardina pilchardus) and horse mackerel (Trachurus trachurus) off Portugal, having a negative influence during winter spawning but positive influence during spring-summer juvenile phase (Santos et al. 2001). It would appear the response to nutrient supply is species and location specific. The Pacific sardine, or pilchard Sardinops sagax is a cosmopolitan species with important fisheries in Australia, Japan, South Africa, Chile and California. As a consequence, their larval ecology has been studied and their growth rates in particular, well researched (overview in Gaughan et al. 2001). Pilchards were not appreciated as being particularly abundant off eastern Australia, until widespread deaths of pilchards were observed in 1995 and again 1998 (Griffen et al. 1997; Gaughan et al. 2001). Pilchards are important for the coastal ecosystem, as the effects of this die-off were particularly severe for local marine birds (Bunce and Norman 2000, Dann et al. 2000), and for the west coast fishery. Nevertheless, our knowledge of the pilchard on the east coast is very limited, but as the genus is now considered to be monospecific (Grant et al. 1998), there are useful parallels with Sardinops studies off Western Australia (Gaughan et al. 2001), California (Logerwell and Smith 2001), and Japan (Watanabe and Kuroki 1997). The slower growth parameters of Western Australia larval and adult pilchard was argued to be the result of nutrient limitation in the characteristically oligotrophic waters, compared to elsewhere in the world (Gaughan et al. 2001). If growth is actually limited by nutrient availability, then the source of the assimilated nutrients supporting either faster or slower larva growth could be identified by their stable isotope composition. Stable isotope analysis (SIA) of carbon and nitrogen (δ13C and δ15N) in the tissues of planktivorous fish has proved useful in identifying the contribution of upwelled nutrients, which have distinctive stable isotope signatures (Gaston and Suthers 2004). SIA has the advantage of indicating the C and N that is actually assimilated 134

compared to traditional gut content analyses. The proportion of the rarer, heavier, stable isotope is assessed by comparison to an international standard, and expressed as parts per thousand (‰). There are very few studies of stable isotopes and larval fish, due in part to the amount of tissue necessary (0.5-1 mg dry weight, equivalent to nearly 10 individual larvae). The stable isotope composition of larval fish varies with size, from hatch (containing the maternal contribution) to transformation when the diets may suddenly change (Lindsay et al. 1998; Vander Zandin et al. 1998). Similar changes in composition are observed at settlement to seagrass beds in juvenile red drum (Herzka and Holt 2000; 2001). The SIA of larvae reveals rapid changes, over 2-3 days at 18ºC due to both tissue turnover and particularly to growth (Bosley et al. 2002). The change in stable isotope composition of fish larvae thus approximates the time scale of response of daily growth from otolith microstructure. The recent daily growth increments of the otolith provide a robust method of assessing the response of growth rate to a variety of local environmental conditions (summarized in Suthers 1998). In particular, the daily growth increments of pilchards are amenable to this analysis, and larval Atlantic menhaden were found to respond to a stress within 2 days (Maillet and Checkley 1990). The relationship between larval growth and stable isotope composition was first documented for a larval gadoid (Macruronus novaezelandiae, Thresher et al. 1989; 1992). Small larvae retained near the spawning area off western Tasmania had faster growth compared to similarly small larvae that were advected south (Thresher et al. 1989). The faster growing larvae near the spawning area had a distinctively depleted δ13C composition from their tintinnid diet, which was derived from the offshore advection of terrestrially depleted seagrass detritus (Thresher et al. 1992). No other study has simultaneously compared larval growth rates and the stable isotope composition. The East Australian Current (EAC) transports clear oligotrophic waters from the Coral Sea poleward, but can stimulate a variety of upwellings at the separation point off northern New South Wales (NSW), where it forms the Tasman Front. There should be dramatic influence on larval growth and their composition, but no study has made the simultaneous comparison of growth rates and stable isotope composition. Therefore our first aim was to test if recent otolith growth (ROG) of larval pilchards was significantly different between pre-upwelled conditions of the EAC and topographically 135

induced upwelling conditions. Our second aim was to compare the larval ROG with the corresponding stable isotope composition, to determine if the source of C and N between faster and slower growing larval pilchards was different. We compare our findings with other larval pilchard studies to infer the early life history of pilchards off eastern Australia.

Materials and methods Study area The continental shelf bounded by the 200 m isobath off northern New South Wales, narrows from 33 km wide off the town of Urunga (referred to herein as Region 1), to just 16 km wide off Smoky Cape (Region 2), in a distance of 43 km. Smoky Cape is located approximately 430 km north of Sydney (Fig. 1). In response, the poleward East Australian Current accelerates through this region, forcing deep water up into near- surface coastal regions, 100 km south of Smoky Cape at Diamond Head (Region 4) and Cape Hawke (Region 5, Creswell 1994; Roughan and Middleton 2002). Here, the continental shelf extends to 22 km wide at Point Plomer (Region 3), to 32 km wide at Diamond Head (Region 4) and 47 km wide off Cape Hawke (Region 5, Fig. 1). While the currents offshore of 100 m were >1 ms-1, at the 50 m station where we collected larvae, currents were usually south and <0.3 ms-1, such that the distance between Region 1 and 4 (120 km) was greater than any drift distance during our 2 day growth back-calculation (Fig. 1). Upwelling in the region has been well studied (Creswell 1994; Oke and Middleton 2000; 2001; Roughan and Middleton 2002), and has a significant effect on phytoplankton and algal blooms (Dela Cruz et al. 2002, 2003).

Plankton sampling and design Two summer cruises by the Research Vessel Franklin were undertaken in November 1998 and January 1999, taking replicated samples near the 50 m isobath off Regions 1, 3, and 4 in late November, and off Regions 1, 2, 3, 4, and 5 in January 1999 (Fig. 1). Sampling was also conducted at the 100 m isobath (Syahailatua 2004), but few pilchard larvae were caught at the 100 m station in January and were not further analysed. Larval fish were collected at night by simultaneous tows with a neuston net (75 x 75 cm, 500 μm mesh) and a subsurface opening and closing net (“EZNET”, 1 m2 136

mouth opening, 500 μm mesh) at 10-20 m from the surface at the 50 m station. The EZNET was also deployed over deeper depth strata but few pilchard larvae were obtained in these and were not further analysed. All nets had a flow meter to calculate volume filtered. Larvae from the two gear types were combined for this study to attain sample size. Replicated tows were conducted with the neuston net and in part with the multiple opening and closing net and were combined for each night at each region (Table 1). Larvae were compared from Regions 1 to 3 (pre-upwelled waters of the EAC) with those from the topographically induced upwelled waters (Regions 4 and 5). Similar events were observed on both cruises. During the cruise, a calibrated thermosalinograph and a calibrated fluorometer operated continuously. At each station a depth profile was made using a Neil-Brown conductivity-temperature-depth (CTD) meter. Inside the surface and sub-surface nets, a 20 cm diameter 100 µm mesh net was mounted, to obtain microzooplankton abundance. Plankton samples were immediately preserved in 5% buffered formalin (to preserve the otoliths, samples were buffered using sodium carbonate which caused some bleaching of the melanophore pattern after 2-3 months). Back at the laboratory the larvae were sorted out from the plankton, using the identification key of Neira et al. (1999), and larvae were stored in 95% ethanol until otolith analysis. Too few larvae were sampled in January at Region 2 and 3 for otolith analysis. All larvae were measured for standard length (up to a maximum of 100 larvae per sample, Table 2), using image analysis. The microzooplankton samples were gently rinsed with seawater through a 90 µm mesh sieve and concentrated to 0.1 L, and zooplankton counts were made on two replicate subsamples and converted to numbers L-1.

Otolith analyses. Between 8-10 larvae per sample were selected for otolith analyses (in proportion to the size range, Table 1). Using a dissecting microscope and polarized light we removed both sagittae and mounted them onto separate glass slides with nail polish. The otolith radii and increment width series along the longest axis were measured using a compound microscope (64 x oil immersion lens) and by image analysis. Daily increment counts of all fish were checked by independent counts from a random, blind-labeled subsample (n=27) by a second reader. Counts were highly 137

correlated (r=0.92), with a mean difference of 0.7 (range of differences from -3 to +1). Since the initial daily growth ring is deposited at 2-3 days after hatching (Hayashi et al. 1989), we added two days to the total increment count, to estimate the post-hatching age of pilchard larvae. Back-calculation of length at 2, 4, 6 and 8 days pre-capture was calculated using the non-linear method of Watanabe and Kuroki (1997), which incorporates the biological intercept method (Campana and Jones 1992, Equation 27). Daily increment formation in this species was demonstrated by Hayashi et al. (1989), and was re- examined in our study by marginal increment formation (see Results). Following the rationale of Watanabe and Kuroki (1997), we used the preserved length=5.0 mm at first feeding when daily increments begin to be deposited at age 3 d (Hayashi et al. 1989). We then calculated the change in length during Days 1&2 precapture, Days 3&4, Days 5&6 and Days 7&8. We did not back-calculate to within 6 increments of the first feeding ring (8 d old), to avoid back-calculation to the size range less than we caught. We initially used the Campana and Jones (1992, Equation 27) biological intercept method of back-calculation, as the length-radius (L-R) relationship appeared linear, with minimal improvement in the R2 statistic for various non-linear fits (power curve, L-ln(R), or ln(L)-ln(R). However R2 is insensitive to departures from linearity in small and large fish, as revealed by inspection of the residuals plot. Watanabe and Kuroki (1997) use a power curve L=a*Rb to back-calculate size, incorporating the biological intercept concept, by solving the parameters a and b for each larva for two points in time: at hatch and at capture. Watanabe and Kuroki (1997) set the size at hatch (actually the first feeding check, for pilchard larvae caught in a plankton net and preserved in formalin) at 5 mm, and used each individual's hatch check diameter. Solving for a and b gives: b a = Lh / (Rh ) ……………………………… (1)

b = ln(Lh/Lc)/ln(Rh/Rc)…………………… (2)

where Lh is the size at hatch (set at 5 mm), Rh is the hatch check diameter (or in the case of S. sagax is the first feeding check), Lc is the length at capture and Rc is the otolith radius at capture. Using the individually determined parameters a and b, the length at various radii (ages) could then be back-calculated. Interestingly, an almost identical back-calculation was obtained using the sample level estimates (i.e., not individual level) of the modified Fry equation (Vigliola et al. 2000). 138

Linear growth models and, Laird-Gompertz growth models (only November 1998 data) were fitted to the length on age data. We used the following form of the Laird-Gompertz (Equation 15, Campana and Jones 1992):

Lt = L∞ exp[-exp(-G{age-X0})]………………..(3) where L∞ is the asymptotic length, G is the instantaneous rate of growth at age X0 and

X0 is the inflection point of the curve. Laird-Gompertz growth functions were fitted iteratively to the length-age data for comparison of parameter values with the literature. The back-calculated recent otolith growth (RG) was compared amongst regions by ANCOVA and tested with temperature and prey abundance by linear regression. Length frequency distributions off each region, with each gear type (neuston or EZ net) were compared by a pairwise KS test, with an adjusted P-value of significance set at P=0.01. Larval density (no per 100 m3) and stable isotope ratios were tested for homogeneity by Cochrans test and compared amongst regions by ANOVA.

Stable isotope analyses. The larval body remaining from the otolith dissection (i.e. less the head) was then freeze-dried to a constant weight cut up into small pieces and approximately1.5-1.7 mg of the tail musculature was sealed into foil capsules. Stable isotope analysis on each capsule was performed at the CSIRO Land and Water Adelaide Laboratory on an Automated Nitrogen Carbon Analysis – Mass Spectrometer (20-20 Europa Scientific). Capsules were combusted and the reaction products separated by GC (Gas Chromatography) to give pulses of pure CO2 and N2 for analysis of total C and N, and δ13C and δ15N. Isotope values were expressed in del (δ) notation, δ13C and δ15N, relative to standards (Glycine for carbon and EDTA for nitrogen) that had been calibrated against international standards (Gaston and Suthers 2004). The precision in δ13C and δ15N analyses was ±0.2‰.

Results

Oceanographic conditions and larval distribution. Synoptic sea surface temperature images on both cruises revealed a front and separation of the EAC from the coast, near Point Plomer (Fig. 1, Roughan and 139

Middleton 2002). Consequently a strong temperature gradient was evident off Diamond Head, with a horizontal difference of 1-2ºC in November and 3-4ºC in January. At the 50 m station, cooler water was evident at Diamond Head, with an elevation of isotherms by 40 m (Fig. 2). Maximum currents were observed beyond the 100 m isobath off Smokey Cape of >1.5 ms-1, but were <0.5 ms-1 near the 50 m station and sometimes even reversed to the north, off Diamond Head (Fig. 1, Roughan and Middleton 2002). Pilchard larvae were 10 fold more abundant in November than later in summer during January (Fig. 3a, b). Overall, 2,628 pilchard larvae were collected in November 1998 (over all nets and regions, ~24 larvae 100 m-3) and 143 larvae collected in January 1999 (overall ~2 larvae 100 m-3). During November we observed similar abundances in the neuston and subsurface nets, but comparisons across regions by ANOVA were confounded by non-homogeneity of variance. During November the average log concentration of larval pilchards was 3 to 10 fold greater in the upwelling regions off Region 4 (Diamond Head) compared to Region 1 (Fig. 3a), and in January there was no consistent alongshore pattern (Fig. 3b). The standard length of larvae was unimodally distributed between 5-25 mm, with the average SL in November (14.8 mm, SE=0.1) greater than in January (11.7 mm, SE=0.2). There were no consistent trends in length between sampling gear (Fig. 3 c,d), with the exception of the subsurface net in November (Fig 3c). In both months the length distribution from Region 1 and 3 to 4 or 5 was significantly shifted to the larger sizes (Fig. 3 c, d, KS test, p<0.01, except the subsurface net, November).

Otolith microstructure and growth The pilchard otoliths had an average hatch check diameter of 11.6 μm, with 2- 4 faint perinuclear increments before an established and definite pattern was evident (Fig. 4). The average proportion of the marginal increment was almost 100% of the previous increment by early in the evening (20:00-21:00, unpublished data). This proportion declined during the night such that around 02:00 the average proportion was approximately 40%, consistent with daily increment formation. A linear model adequately described the relationship between size and age, with a similar R2 statistic between months, and a similar corrected R2 statistic for the Laird-Gompertz model, calculated for November data only (Fig. 5, Table 2). There were insufficient older larvae in January to determine a non-linear fit. The linear growth curves were: November 1998, SL = 4.224 + 0.625*age, (R2=0.64, n=215) and January 1999, SL=0.551 + 0.976*age, (R2=0.66, n=110). 140

The slope of the linear regression for November (0.63, and 0.98 for January, Fig. 5) is similar to the range of daily growth estimated for larvae between 5-25 days old, from the Laird-Gompertz model for November (average 0.64 mm d-1, range 0.45- 0.72). The growth rate in January was over 33% greater compared to November (slopes were significantly different between months, ANCOVA, p<0.001). A nearly identical result was obtained by restricting the ages to a common range for each month (<17 days). The correlation between size-at-age and temperature and zooplankton (ln transformed), for Regions 1-5 over both months was not significant (r<0.23, Table 3).

The larval length-otolith radius relationship (Fig. 6) could be described by the power functions: L = 2.016*R0.477 (r2= 0.83, November 1998), L = 1.654*R0.514 (r2=0.80, January 1999), where L is the standard length (mm) and R is the maximum sagittal otolith radius (μm). The two months were significantly different using the linear log-log equivalent form of the power function, with a significantly greater intercept for the November 1998 data (ANCOVA, p<0.001).

To assess our back-calculation, the average daily change in back-calculated standard length of the two outer growth increments ([RG1&2]÷2) for each daily age class was compared with the daily growth independently estimated from November’s Laird-Gompertz model (Fig. 7). In general, the two daily growth estimations were similar and declined with age. As a function of size, the back-calculated growth for days 1 and 2 precapture increased as a function of length or remained constant at around 1-2 mm growth (Fig. 8). The growth for previous two days prior to capture exhibited significant differences in the slopes among regions, in both months (ANCOVA, p<0.001, Table 3). In November the slope of the recent growth, as a function of length was significantly greater off Region 1 and 3 than off the upwelling areas (Region 4, 5, Table 3, Fig. 8). In January, the slope was also significantly greater at Region 1, than at Region 4 and 5 (Table 3, Fig. 8). To interpret these differences among regions we compared the explanatory power (R2 statistic) between the ANCOVA model and a multiple regression model that did not include Region, but with water temperature and micro-zooplankton concentration. Temperature was not significantly correlated with recent growth in either month, and zooplankton was positive in November, but negative in January 141

(Table 3). The back-calculated recent growth in standard length for days 1 and 2 precapture (RG1&2) was highly correlated with that calculated at days 3-4 precapture (Table 4). However the correlation declined with increasing separation in time, such that RG1&2 was less correlated with RG7&8 (Table 4).

Stable isotope composition and growth rate. There was no significant correlation between δ13C and length in either month, but there was between δ15N and length (Fig. 9, r=0.47, 0.52 over all regions in November and January respectively). The overall positive correlation was also apparent within each month and region combination (Fig. 9). The size independent recent growth index (RGI=RG1&2÷ln(SL)) was significantly related to the stable isotope composition in November (Fig. 10a, RGI=1.488-0.104*δ15N, N=69, P<<0.001, R2=0.29, Fig. 10b, RGI=2.300+0.096*δ13C, N=70, P<<0.001, R2=0.21), such that faster growing pilchard larvae were depleted in δ15N and enriched in δ13C. There was no significant correlation in January. In November, pilchard larvae from the pre-upwelled, EAC conditions off Regions 1 and 3 were significantly depleted in δ15N and enriched in δ13C, compared to those off the upwelling regions (i.e. Region 4, Fig. 11, Tukey’s test p<0.05). In January, Region 1 was also significantly depleted in δ15N, while Regions 1 and 5 were significantly depleted in δ13C compared to Region 4 (Tukey’s test, p<0.05). Faster average recent growth for each region (shown as expanding symbols in Fig. 11) was associated with the depleted δ15N and enriched δ13C of the pre-upwelled EAC waters.

Discussion

Growth rates in different regions and months Our yardstick for comparing the significance of upwelling on larval pilchards was the back-calculated growth over the previous two complete daily growth increments, prior to capture. Our back-calculation seems to be a robust measure of recent growth, as shown by the two different back-calculation methods (Watanabe and Kuroki 1997; Vigliola et al. 2000), and by the comparison with the daily growth rate from the Laird-Gompertz relationship. A 2-day response in growth (rather than some lag effect) is justified by the documented 2-3 d response of larval menhaden (Maillet 142

and Checkley 1990), and by the declining correlation of RG1&2 with RG3&4, RG5&6 and RG7&8. However the recent growth rate varied with larval size, such that we had to compare this regression slope among regions. The slope on both cruises was steeper in the northern, pre-upwelled waters of the East Australian Current off the town of Urunga, than the southern upwelled waters off Diamond Head. The decline in slope was due to the decline in the growth rate of larger pilchards (>10 mm) in upwelled water, rather than an increase in growth of smaller larvae. Thus, contrary to our expectations, we found that larger pilchard larvae grew significantly slower during the 2 days prior to capture in the upwelling regions (Regions 4 and 5), compared to those from the pre-upwelled waters. A clue to our surprising result may be found in the stable isotope signature, which showed that the slower growing larvae in the upwelled waters were comparatively enriched in δ15N and depleted in δ13C relative to those faster growing larvae from Regions 1 and 3 (Urunga and Point Plomer). Therefore, contrary to general expectations (e.g. Ware and Thompson 1991), larval pilchard had a different diet in upwelled conditions that did not stimulate their growth. Nitrate is the hallmark of upwelled waters which are typically enriched in 15N (4-6‰, Altabet and McCarthy 1986; Boutton 1991) and depleted in 13C (Gaston and Suthers 2004), consistent with the stable isotope composition of larvae from Region 4 and 5. The upwelled waters were slightly cooler by 0.5-1ºC (Syahailatua 2004), but water temperature was not a significant covariate in an expanded regression model of recent growth, for either cruise. Microzooplankton <100 µm abundance was significantly correlated with recent growth in November, but the data were unfortunately incomplete in November, compared to January when it was negatively correlated. Therefore the cause of the slower short-term growth is still unclear, and could involve turbulence, disruption of the thermocline by upwelling, or the quality or size structure of zooplankton. At the scale of months, the faster growing larvae in January was clearly due to the 5-6ºC warmer water compared to November. The allometric relationship of fish size-otolith size in our study was not as marked as found for S. melanosticus (Oozeki and Zenitani 1996, Fig. 6). The cause could be due to possible genetic differences among stocks or sub-species, or due to more rapid growth in the Kuroshio producing proportionally smaller otoliths. The effects of this non-linearity was addressed by Watanabe and Kuroki (1997), who 143

determined a novel non-linear back-calculation of recent growth based on the biological intercept method. We found identical back-calculations between this method and the modified Fry equation (Vigliola et al. 2000), which was also consistent with the daily growth expected from the overall Laird-Gompertz growth model. The variance in RG1&2 compared to the growth model (Fig. 7) was due in part to the unequal representation of larvae, with ages from different regions.

Larval growth and early life history of pilchards The slower growth of larval pilchard was perplexing, but may be interpreted in the light of Watanabe’s (2002) explanation of the pilchard life history in the Kuroshio Extension (equivalent of the EAC’s Tasman front). They too found slower growth inshore, but found enhanced growth offshore in the frontal waters of the Kuroshio Extension. Larval pilchards are advected away from the coast by the Kuroshio, where larvae grow and then migrate across the Transition region to the northern summer feeding grounds off the island of Hokkaido (Watanabe 2002). Adults then migrate south to spawn off Kyushu and Shikoku. Slower growth by larval Sardinops nearshore was also observed off the coast of California (Loggerwell and Smith 2001). This study used the extensive CalCOFI (California Cooperative Fisheries Investigations) dataset to determine the location of “survivors’ habitat”, where patches of ≥18 d old larvae were found, which corresponded to SST images of mesoscale eddies in offshore regions. Sardinops eggs, chlorophyll biomass and zooplankton volume were the greatest inshore. The life cycle of the east coast pilchards could mirror that observed for the Kuroshio, with adults migrating north and spawning in coastal areas of northern NSW and southern Queensland in late winter (Ward et al. 2003). Eggs and larvae are advected southwards by the EAC, some along the coast, but larval growth and survival is enhanced for those transported past the separation zone and east into the Tasman Front, where they grow and migrate back to the coastal summer feeding grounds near Jervis Bay (southern New South Wales). The tendency for larger larvae to be found at the southern end of our study area is consistent with larval transport south, from the northern spawning areas. During our two cruises we did not continue to sample the front out into the Tasman Sea to test this hypothesis, although a recent cruise in 144

September 2004 found pilchard larvae to be abundant in the Tasman Front, east of 153ºE and 300 km from eastern Australia (unpublished data). The parameter estimates for the Laird-Gompertz model (Table 2) provide an average growth rate of 0.64 mm d-1 (10-25 mm SL), consistent with those calculated off Japan, California, South Africa, South America (as reviewed in Gaughan et al. 2001), but double the larval growth found in the Leeuwin Current (Gaughan et al. 2001). It is therefore possible that larval pilchard in the EAC are not food limited, in comparison with those from the Leeuwin Current.

Larval fish ecology of stable isotopes Larval pilchards increased in δ15N with length, presumably as larvae consumed more higher trophic order plankton. A positive relationship between δ15N and size was demonstrated in larval anchovy (10-30 mm, Lindsay et al. 1998) and a negative relationship in larval smallmouth bass (4-15 mm, Vander Zanden et al. 1998), as the larvae lose the maternal signature and select larger and isotopically different prey. A weak positive relationship between δ13C and size was reported in both these studies, and was observed in this study as well. Therefore the δ15N composition of larval fish can provide an index of assimilation, and contrast of tissue turnover versus growth (Vander Zanden et al. 1998; Bosley et al. 2002), but the general utility is more suited to the pelagic-benthic transition in the early life history (e.g. Herzka and Holt 2001). Considerable challenges remain in the use of stable isotopes in marine fish larvae, for the reconstruction of food webs and particularly the calibration of turnover and depuration of stable isotope signatures. The prey species’ signatures are unknown, as obtaining sufficient quantities of nauplii or particular copepod species is difficult. Fractionation between diet and tissue composition is now recognized as being diet- and species-specific, making it difficult to relate signatures to pelagic food chains without a laboratory calibration (Bosley et al. 2002). For example two typical prey items, a copepod (Temora) and cladoceran (Penillia) exhibited a strong response in stable isotope composition to upwelling off the New South Wales coast, particularly in δ13C (unpublished data). Rearing of larval pilchards, and sampling further out along the Tasman Front is needed to understand the conundrum posed by this study: what conditions reduced the growth rate of larval pilchards in an upwelling zone? The 145

relationship between clupeid production and upwelling is not a simple one, and seems to be species and location specific.

Acknowledgements We acknowledge the support of the Australian Research Council, and the National Research Institute of Fisheries Science (NRIFS), Japan. We thank the excellent seamanship of the captain and crew of the R.V. Franklin and the remarkable CSIRO support staff. Richard Piola provided considerable technical expertise during the cruise, Matt Taylor provided laboratory support and Jocelyn Dela Cruz kindly provided the microzooplankton data. Laurent Vigliola kindly assisted with the modified Fry back- calculation.

References Altabet, M.A., and J.J. McCarthy. 1986 Vertical patterns in 15N natural abundance in PON from the surface waters of warm-core rings. Journal of Marine Research 44,185-201. Bosley, K.L, Witting, D.A., Chambers, R.C. and Wainright, S.C. 2002. Estimating turnover rates of carbon and nitrogen in recently metamorphosed winter flounder Pseudopleuronectes americanus with stable isotopes. Marine Ecology Progress Series 236, 233-240 Boutton, T. W. 1991. Stable carbon isotopic ratios of natural materials: II. Atmospheric, terrestrial, marine and freshwater environments. Ed. Coleman, D. C., Fry, B. Carbon isotope techniques. San Diego, Academic Press. Pp 173 – 185 Bunce, A. and Norman, F.I. 2000. Changes in the diet of the Australasian gannet (Morus serrator) in response to the 1998 mortality of pilchards (Sardinops sagax). Marine and Freshwater Research 51, 349-353 Campana, S.E., and C.M. Jones. 1992. Analysis of otolith microstructure data. In: Stevenson DK, Campana SE (Eds), Otolith Microstructure Examination and Analysis. Canadian Special Publication Fisheries and Aquatic Sciences, Deptament of Fisheriws and Oceans, Ottawa. Cresswell, G.R. 1994. Nutrient enrichment of the Sydney Continental Shelf. Australian Journal of Marine and Freshwater Research 45, 677-691. Cury P., and C. Roy. 1989. Optimal environmental window and pelagic fish recruitment success in upwelling areas. Canadian Journal of Fisheries and Aquatic Sciences 46, 670-680. Dann, P., Norman, E.I., Cullen, J.M., Neira, E.J., and Chiaradia, A. 2000. Mortality and breeding failure of little penquins, Eudyptula minor, in Victoria, 1995-96, following a widespread mortality of pilchard, Sardinops sagax. Marine and Freshwater Research 51, 355-362. Dela-Cruz J, P. Ajani, R. Lee, T. Pritchard and I.M. Suthers. 2002. Temporal abundance patterns of the red tide dinoflagellate Noctiluca scintillans along the southeast coast of Australia. Marine Ecology Progress Series 236, 75-88 146

Dela-Cruz, J., J.H. Middleton and I.M. Suthers. 2003. Population growth and transport the red tide dinoflagellate, Noctiluca scintillans, in the coastal waters off Sydney Australia, using cell diameter as a tracer. Limnology and Oceanography 48, 656- 674 Gaston, T.F., and I.M. Suthers. 2004. Spatial variation in the δ13C and δ15N of liver, muscle and bone in a rocky reef planktivorous fish: the relative contribution of sewage. Journal of Experimental Marine Biology and Ecology 304, 17-33. Gaston, T.F. and Suthers I.M.. 2005. The 13C, 15N and 34S signature of a rocky reef planktivorous fish does indicate different coastal discharges of sewage. Marine and Freshwater Research 55, 689-699 Gaughan, D.J., Fletcher, W.J., and White, K.V. 2001. Growth rate of larval Sardinops sagax from ecosystems with different levels of productivity. Marine Biology 139: 831-837. Grant, W.S., Clark, A.M., and Bowen, B.W. 1998. Why restriction fragment length polymorphism analysis of mitochondrial DNA failed to resolve sardine (Sardinops) biogeography: insights from mitochondrial DNA cytochrome b sequences. Canadian Journal of Fisheries and Aquatic Sciences 55, 2599-2547. Griffin, D.A., P.A. Thompson, N.J. Bax, R.W. Bradford and G.M. Hallegraeff (1997). The 1995 mass mortality of pilchard: no role found for physical or biological oceanographic factors in Australia. Marine and Freshwater Research 48, 27-42. Hayashi, A, Y. Yamashita, K. Kawaguchi and T. Ishii. 1989. Rearing method and daily otolith ring of Japanese sardine larvae. Nippon Suisan Gakkaishi 55, 997- 1000. Herzka, S.Z,, S.A. Holt, and G.J. Holt. 2001. Documenting the settlement history of individual fish larvae using stable isotope ratios: model development and validation. Journal of Experimental Marine Biology and Ecology 265, 49-74. Herzka, S.Z., and G.J. Holt. 2000. Changes in isotopic composition of red drum (Sciaenops ocellatus) larvae in response to dietary shifts: potential applications to settlement studies. Canadian Journal Fisheries and Aquatic Sciences 57, 137- 147. Kailola, P., M. Williams, P. Stewart, R. Reichelt, A. McNee and C. Grieve. 1993. Australian Fisheries Resources. Bureau of Resources Sciences, Canberra. Lindsay, D. J., M. Minagawa, I. Mitani, K. Kawaguchi (1998). Life history of Japanese anchovy, Engraulis japonicus, in Sagami Bay: stable isotope analysis. Monthly Kaiyo, 29, 418-424 (in Japanese). Logerwell, E., and Smith P.E. 2001. Mesoscale eddies and survival of late stage Pacific sardine (Sardinops sagax) larvae. Fisheries Oceanography 10, 13-25. Maillet, G. L., and Checkley, D. M. (1990). Effects of starvation on the frequency of formation and width of growth increments in sagittae of laboratory-reared Atlantic menhaden Brevoortia tyrannus larvae. Fishery Bulletin, U.S. 88, 155- 165. Neira, F.J., A.G. Miskiewicz and T. Trinski. 1998. Larvae of temperate Australian fishes; Laboratory guide for larval fish identification. University of Western Australian Press, Nedlands. Oke, P.R. and J.H. Middleton. 2000. Topographically induced upwelling off eastern Australia. Journal of Physical Oceanography 30, 512-531. Oke, P.R. and J.H. Middleton. 2001. Nutrient enrichment off Port Stephens: The role of the East Australian Current. Continental Shelf Research 21, 587-606. 147

Oozeki Y. and Zenitani H. 1996. Factors affecting the recent growth of Japanese sardine larvae (Sardinops melanosticus) in the Kuroshio Current. Pages 95-105. In: Y. Watanabe, Y. Yamashita and Y. Oozeki (eds), Survival Strategies in early life stages of marine resources, AA Balkema, Rotterdam. Roughan, M. and J.H. Middleton. 2002. A comparison of observed upwelling mechanisms off the east coast of Australia. Continental Shelf Research 22, 2551-2572. Santos, A.M.P, M.F. Borges and S. Groom. 2001. Sardine and horse mackerel recruitment and upwelling off Portugal. ICES Journal of Marine Science 58, 589-596. Suthers, I.M. 1998. Bigger? Fatter? or is faster growth better ? Considerations on condition in larval and juveniles coral-reef fishes. Australian Journal of Ecology 23, 265-273. Syahailatua, A. 2004. Biological oceanography of larval fish diversity and growth off eastern Australia. PhD thesis, University of New South Wales, Sydney, Australia. 128 pp. Thresher, R.E., B.D. Bruce, D.M. Furlani and J.S. Gunn. 1989. Distribution, advection and growth of larvae of the southern temperate gadoid (Macruronus novaezelandiae) in Australian coastal waters. Fishery Bulletin U.S. 87, 29-48. Thresher, R. E., P. D. Nichols, Gunn, J. S. Bruce, B. D., and Furlani, D. M. (1992). Seagrass detritus as the basis of a coastal planktonic food chain. Limnology and Oceanography 37, 1754-1758. Vander Zanden, M.J., M. Hulshof, M.S. Ridgway, and J.B. Rasmussen. 1998. Application of stable isotope techniques to tropic studies of age-0 smallmouth bass. Transaction of American Fishery Society 127, 729-739. Vigliola, L., Harmelin-Vivien, M. and Meekan, M.G. 2000. Comparison of techniques of back-calculation of growth and settlement marks from the otoliths of three species of Diplodus from the Mediterranean Sea. Canadian Journal of Fisheries and Aquatic Sciences 57, 1291-1299. Ware, D.M. and R.E. Thomson. 1991. Link between long-term variability in upwelling and fish production in the Northeast Pacific Ocean. Canadian Journal of Fisheries and Aquatic Sciences 48, 2296-2306. Ward, T.M., Staunton-Smith, J., Hoyle, S., and Halliday, I.A. 2003. Spawning patterns of four species of predominantly temperate pelagic fishes in the sub-tropical waters of southern Queensland. Estuarine and Coastal Shelf Science 56, 1125- 1140. Watanabe, Y. and Kuroki, T. 1997. Asymptotic growth trajectories of larval sardine (Sardinops melanostictus) in coastal waters off western Japan. Marine Biology 127, 369-378. Watanabe, Y. 2002 Resurgence and decline of the Japanese sardine population. Fishery Science. Pages 243-256. In: L.A. Fuiman and R.G. Werner, The unique contributions of early life stages, Blackwell, Oxford (UK).

148

Table 1. Summary of samples collected on the two cruises showing the date in November 1998 and January 1999, local time of collection, the type and depth of net (0- 1 m, neuston net; 10-20 m, EZ subsurface net), total no. of larvae collected; N, no. of individuals; SIA, no. used for stable isotope analysis; SL, Standard length (mm).

Region Date Time Net Total Morphology Otolith SIA N SL range N SL range N 1) Urunga 21-Nov 23:34 0-1 40 40 9.2-16.1 9 11.4-15.0 9 “ 23:47 0-1 18 16 10.7-17.0 9 12.5-17.0 9 “ 22-Nov 3:09 0-1 2 2 11.2-13.7 2 11.2-13.7 2 “ 3:26 0-1 4 4 8.4-11.8 4 8.4-11.8 3 “ 23-Nov 1:30 10-20 15 15 7.5-17.0 14 7.5-17.0 14 3) Point Plomer 23-Nov 21:38 0-1 12 12 8.3-16.1 9 9.5-16.1 - “ 23-Nov 21:38 10-20 68 66 6.8-16.4 9 10.4-16.1 9 “ 24-Nov 3:40 10-20 59 47 6.8-15.8 9 11.3-15.8 9 4)Diamond 24-Nov 23:44 0-1 80 77 13.6-18.5 9 16.7-18.3 10 Head “ 23:58 0-1 199 199 9.2-20.3 9 16.3-18.4 9 “ 24-Nov 23:25 10-20 98 81 8.5-14.3 9 9.3-14.3 9 “ 25-Nov 3:00 10-20 152 103 8.6-17.6 9 10.1-15.9 - SUBTOTAL 2230 1116 156 83

Urunga 21-Jan 22:25 0-1 35 26 8.6-13.5 22 8.6-13.5 10 “ 22-Jan 3:43 0-1 2 2 10.1-12.9 2 10.8-12.9 - “ 21-Jan 22:00 10-20 3 3 5.0-9.8 - - - “ 22-Jan 3:30 10-20 20 14 7.8-11.2 18 7.8-11.2 20 Diamond Head 25-Jan 0:55 0-1 7 6 9.5-15.1 6 9.5-15.1 6 “ 25-Jan 1:07 0-1 4 4 11.8-15.9 4 11.8-15.9 4 “ 24-Jan 21:00 10-20 16 13 7.3-14.9 13 7.3-14.9 16 “ 25-Jan 1:00 10-20 18 18 9.3-15.7 17 9.3-15.7 18 Cape Hawke 25-Jan 21:50 0-1 2 2 12.7-15.5 2 12.7-15.5 2 “ 26-Jan 2:25 0-1 2 2 13.6-16.6 2 13.6-16.6 2 26-Jan 2:32 0-1 2 2 13.6-17.9 2 13.6-17.9 2 Cape Hawke 28-Jan 21:54 0-1 7 7 12.9-16.7 7 12.9-16.7 7 “ 28-Jan 22:16 0-1 12 12 8.2-15.9 12 8.2-15.9 7 “ 25-Jan 21:43 10-20 6 5 10.9-14.2 6 10.9-14.2 6 “ 26-Jan 2:30 10-20 7 6 10.0-14.0 6 10.4-13.2 7 SUBTOTAL 143 122 119 107

149

Table 2. Parameter estimates for the fitted Laird-Gompertz growth function for S.sagax 2 from eastern Australia in November 1998, shown in Fig. 7 (n=215, R =0.65). L∞, asymptotic length, G, instantaneous rate of growth at age X0 and X0, is the inflection point (age) of the curve.

Parameter estimate (SE) 95% CI L∞ 25.760 (1.944) 16.48-35.04 G 0.079 (0.024) 0.03-0.12 X0 9.055 (0.610) 4.84-13.27

Table 3. Summary table of ANCOVA and multiple regression models of the dependent variable RG (recent growth) – the back-calculated growth over the previous 2 days prior to capture – as a function of Lc (standard length at capture, back-calculated to the first complete increment), Region, Temp (water temperature) and Zoopl, micro-zooplankton concentration (ln transformation). The sample size (N) is reduced in November multiple regression model as microzooplankton was only collected in the neuston net. Those variables in bold are statistically significant (p<0.05). Note that Temperature is not significantly correlated with recent growth, and zooplankton is positive in November, but negative in January. [piloto4.syd]

Month Regions Model N R2 November 1, 3, 4, RG= Lc + Region + Region* Lc 148 0.47 1, 3, 4 RG= -8.79 + 0.09*Lc + 0.38*Temp + 0.87*ln(Zoopl) 50 0.40

January 1, 4, 5 RG = Lc + Region + Region* Lc 95 0.42 1, 4, 5 RG = -0.47 + 0.06*Lc + 0.06*Temp - 0.24*ln(Zoopl) 95 0.22

Table 4. The Pearson correlation matrix for the back-calculated recent growth of all larvae for days 1-2 precapture (RG1&2), days 3-4 precapture (RG3&4), days 5-6 precapture (RG5&6) and days 7-8 (RG7&8). N=105, **p<0.01, *p<0.05

Interval RG1&2 RG3&4 RG5&6 RG7&8 RG1&2 1 RG3&4 0.81** 1 RG5&6 0.63** 0.74** 1 RG7&8 0.27* 0.43** 0.66** 1

150

List of Figures.

Fig. 1. Contours of sea surface temperature (SST), derived from NOAA 14 satellite images of the northern NSW coast on a) 21 November 1998, and b) 19 January 1999. The highlighted sampling regions were subsampled for age and stable isotope analyses. Representative vectors show currents measured underway using the shipboard Acoustic Doppler Current Profiler.

Fig. 2 Alongshore temperature profiles derived from CTD casts at the 50 m station during a) November 1998 and b) January 1999. Upwelling off Diamond Head is evident by the uplifted isotherms.

Fig. 3. Larval density of pilchards (no. per 100 m3) across regions in a) November 1998 and b) January 1999, and corresponding box and whisker plots of the length frequency distribution of larval pilchards c) November 1998 and d) January 1999. Significant differences in the cumulative length frequency distribution shown in c), d) by a KS test (P<0.01) are identified with a different letter.

Fig. 4. Image of a 11.6 mm SL larval S. sagax (10 growth increments), showing, the maximum radius, the location of growth increment widths with the last outer two complete growth increments bracketed.

Fig. 5. Standard length (SL) on age (determined as increment count+2) for both cruises. The Laird-Gompertz fit for only November 1998 is shown (Table 2), but excluding the 3 circled data points.

Fig. 6. Non-linear relationship of standard length (SL, mm) on the maximum otolith radius (μm). The curves for the two months were slightly, but significantly different. The curve labeled O&Z96 is that for S. melanosticus, from Oozeki and Zenitani (1996).

Fig. 7. Estimated growth in mm d-1 from the fitted Laird-Gompertz growth equation for November 1998 as a function of larval age (circles), plotted with the back-calculated growth for each daily age class for November 1998 (square symbols, mm d-1, ± standard error).

Fig. 8. Back-calculated larval growth over the previous 2 complete days prior to capture (mm), as a function of length, back-calculated for the last complete increment, plotted for each region. Left column; November 1998, right column; January 1999.

Fig. 9. Scatterplot of the δ15N on standard length of larval S. sagax in a) November 1998 and b) January 1999. Region (Reg) are labelled with different symbols, with Region 4 experiencing uplifting in November and Region 4, 5 experience upwelling in January.

Fig. 10. Recent growth index (size adjusted) of larval S. sagax for November (solid line) and January (dashed line), plotted on a) δ15N and b) δ13C. 151

Fig. 11. Average δ15N and δ13C (±SE) for larval S. sagax at each region (Reg) in November 1998 (Nov) and January 1999 (Jan), with the expanding symbols proportional to recent growth over the 2 days prior to capture.

152

a) November 1998 b) January 1999

1) Urunga 1) Urunga

2) Smoky Cape 2) Smoky Cape 31.0 31.0 m m

0 0

2 2 3) Point Plomer 3) Point Plomer Latitude °S Latitude

4) Diamond Head 4) Diamond Head

32.0 32.0

5) Cape -1 1 ms Hawke

152.0 153.0 152.0 153.0 Longitude °E Longitude °E

Fig. 1, Uehara et al. 153

d . e r a d s p e e a e I a m H e k n C o H w to a l d a h g y P n y H g n k t o d e u u o in m w p o r o ia o a r U Sm P r C B 0 D C a) Nov98

20 Depth (m) Depth

0 b) Jan99

Depth (m) 40

0 30 30.531 31.532 32.5 33 Latitude °S

Fig. 2. Alongshore temperature profiles derived from CTD casts at the 50 m station during a) November 1998 and b) January 1999. Upwelling off Diamond Head is evident by the uplifted isotherms.

Fig. 2, Uehara et al 154

600 100 a) Nov98 b) Jan99

400 60

40 200 Larval density 20

0 0 1 3 4 1 2 3 4 5 20 d 20 c) Nov98 d) Jan99 b,c g a b a,b,c c g e,g 15 15 g e

Standard lengthStandard (mm) 10 10

Subsurface net Neuston net 5 5 1 3 4 1 4 5 Region Region

Fig. 3, Uehara et al. 155

Fig. 4, Uehara et al. 156

20 Nov98 Jan99

Standard length (mm) Standard 10

5 0 10 20 30 Age

Fig. 5. Standard length (SL) on age (determined as increment count+2) for both cruises. The Laird-Gompertz fit for only November 1998 is shown (Table 2), but excluding the 3 circled data points.

Fig. 5, Uehara et al. (piloto4.syd) 157

O&Z96

Standard length (mm) length Standard 10

MONTH Nov98 Jan99 5 20 40 60 80 100 Sagitta radius

Fig. 6 Uehara et al. 158

0.9

0.8 ) -1

0.7

0.6 daily growth (mm.d

0.5

0.4 0 5 10 15 20 25 age

Fig. 7. Estimated growth in mm d-1 from the fitted Laird-Gompertz growth equation for November 1998 as a function of larval age, plotted with the back-calculated growth for each daily age class for November 1998 (mm d-1, ± standard error).

Fig. 7 Uehara et al. Laird-Gompertz.xls 159

3 Urunga – Nov. 21 3 Urunga – Jan 20

) Reg.1 Reg.1 -2d 2 2

1 1 RG (mm.

05 10 15 20 05 10 15 20 3 P.Plomer- Nov 23 ) Reg.3 -2d 2

1 RG (mm.

05 10 15 20 3 Diamond Hd – Nov 24 3 Diamond Hd – Jan 25

) Reg.4 Reg.4 -2d 2 2

1 1 RG (mm.

05 10 15 20 05 10 15 20 3 C.Hawke – Jan 26 Reg.5 2

05 10 15 20 Standard length (mm)

Fig. 8. Back-calculated larval growth over the previous 2 complete days prior to capture (mm), as a function of length, backcalculated for the last complete increment, plotted for each region. Left columnm; November 1998, right column; January 1999.

Fig. 8 Uehara et al. Piloto4.syd 160

11 11 a) Nov98 b) Jan99

10 10

N 9

15 9 d

Reg Reg 8 8 4 5 3 4 1 1 7 7 5 10 15 20 5 10 15 20 Standard length Standard length

Fig. 9. Scatterplot of the d15N on standard length of larval S. sagax in a) November 1998 and b) January 1999. Region (Reg) are labelled with different symbols, with Region 4 experiencing uplifting in November and Region 4, 5 experience upwelling in January.

Fig. 9 Uehara et al. Pilsia5b.syd, sia-car.syd 161

1.0 1.0 a) b)

0.8 0.8

0.6 0.6 RG-index 0.4 0.4 Nov98 Jan99 0.2 0.2 7 8 9 10 11 -20 -19 -18 -17 -16 d15N d13C

Fig. 10. Recent growth index (size adjusted) of larval S. sagax for November (solid line) and January (dashed line), plotted on a) d15N and b) d13C.

Fig. 10 Uehara et al. Pilsia4.syd 162

10.0

Reg 4, Nov

9.5

Reg 5, Jan Reg 4, Jan

N 9.0 15 d Reg 1, Nov

RG (mm.-2d) Reg 1, Jan 8.5 Reg 3, Nov 1.7

1.5 1.3 8.0 -19.0 -18.5 -18.0 -17.5 d13C

Fig. 11. Average d15N and d13C (±SE) for larval S. sagax at each region (Reg) in November 1998 (Nov) and January 1999 (Jan), with the expanding symbols proportional to recent growth over the 2 days prior to capture.

Fig. 11 Uehara et al. Mpilsia5.xls