Age and Growth of Larval Atlantic Bluefin Tuna, Thunnus Thynnus ...

University of Miami

Scholarly Repository Open Access Theses

Electronic Theses and Dissertations

2014-12-17

Age and Growth of Larval Atlantic Bluefin Tuna, Thunnus Thynnus from The Gulf of Mexico. Estrella Malca University of Miami, [email protected]

Follow this and additional works at: http://scholarlyrepository.miami.edu/oa_theses Recommended Citation Malca, Estrella, "Age and Growth of Larval Atlantic Bluefin Tuna, Thunnus Thynnus from The Gulf of Mexico." (2014). Open Access Theses. Paper 539.

This Open access is brought to you for free and open access by the Electronic Theses and Dissertations at Scholarly Repository. It has been accepted for inclusion in Open Access Theses by an authorized administrator of Scholarly Repository. For more information, please contact [email protected].

UNIVERSITY OF MIAMI

AGE AND GROWTH OF LARVAL ATLANTIC BLUEFIN TUNA, THUNNUS THYNNUS, FROM THE GULF OF MEXICO

By Estrella Malca A THESIS Submitted to the Faculty of the University of Miami in partial fulfillment of the requirements for the degree of Master of Science

Coral Gables, Florida December 2014

©2014 Estrella Malca All Rights Reserved

UNIVERSITY OF MIAMI

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science

AGE AND GROWTH OF LARVAL ATLANTIC BLUEFIN TUNA, THUNNUS THYNNUS, FROM THE GULF OF MEXICO

Estrella Malca Approved: ________________ Andrew Bakun, Ph.D. Professor of Marine Biology and Fisheries

_________________ Kenneth Broad, Ph.D. Chair and Professor, Dept. of Marine Affairs and Policy

________________ Maria L. Estevanez, M.B.A., M.A. Senior Lecturer of Marine Affairs and Policy

_________________ M. Brian Blake, Ph.D. Dean of the Graduate School

________________ Laura Carrillo Bibriezca, Ph.D. Professor of Sistemática y Ecología Acuática El Colegio de la Frontera Sur, Mexico

MALCA, ESTRELLA Age and Growth of Larval Atlantic Bluefin Tuna, Thunnus Thynnus from The Gulf of Mexico.

(M.S., Marine Affairs and Policy) (December 2014)

Abstract of a thesis at the University of Miami. Thesis supervised by Professors Andrew Bakun and Kenneth Broad. No. of pages in text. (53) Atlantic bluefin tuna (Thunnus thynnus) are the largest and most highly prized among the tuna family. They are highly pelagic, undertaking transoceanic migrations throughout the Atlantic, but the main spawning grounds are the Mediterranean Sea and the Gulf of Mexico (GOM). Given their mobility, management for this species is handled through international agreements under the auspices of the International Committee for the Conservation of Atlantic Tuna (ICCAT) and within the United States, through the National Oceanic and Atmospheric Administration (NOAA). Despite 30 years of ichthyoplankton surveys in the GOM, little is known about larval bluefin tuna (BFT) ecology with regards to growth and survival. Larvae were collected from plankton tows during the Spring Ichthyoplankton research survey using 1 x 2 m plankton nets (505 µm mesh) towed in the upper 10 m of the water column in the GOM in April-May, 2012. Otoliths (sagittae and lapilli) were dissected from 100 larvae, ranging from 2.4 to 8.4 mm (NL or SL) and larval daily age was determined by examining otolith microstructure. Estimated ages ranged from 4 to 18 days and new growth curves for the GOM were compared with existing age estimates for BFT larvae. Growth was significantly different when compared to similar studies in the Florida Keys and in the Mediterranean Sea, but always highly variable at any given length. Environmental parameters examined significantly influenced larval growth. Results will improve the larval index for BFT by

incorporating specimens collected from established spawning grounds. In addition, results will inform stock assessment and play a key role in developing predictive ecological models to enhance ecosystem based fisheries management in the region.

To my family and friends for their love and support and to Cristal Luza for her loving advice and spiritual strength over my lifetime

iii

I would like to thank my committee (Andy, Kenny, Laura and especially Maria), the Early Life History Lab at NOAA Southeast Fisheries Science Center (especially John Lamkin, Trika Gerard, Sarah Privoznik and the hardworking interns that have shared their valued time with us), the SEAMAP team (especially Tammy Cullins), and from the University of Southern Mississippi’s Gulf and Coast Research Laboratory, my sincere gratitude to Dr. Jim Franks and Jason Tilley for their early support and assistance with otolith ageing methods, and last but not least, to the Instituto Español de Oceanografía, Malaga (Alberto García and Raul Laiz-Carrión) for sharing their data, friendship and expertise. Nothing in my universe would be possible without the relentless and ongoing efforts made my beautiful and loving Mother throughout my life. I am grateful for the inspiration, laughter and encouragement of great friends and colleagues. Funding for this research was provided by the University of Miami’s Cooperative Institute for Marine and Atmospheric Studies and the National Oceanic and Atmospheric Administration. Thanks to the officers and crew of the NOAA Ship Gordon Gunter.

iv

TABLE OF CONTENTS Page LIST OF FIGURES .....................................................................................................

vi

LIST OF TABLES .......................................................................................................

vii

1

INTRODUCTION ......................................................................................... 1.1 Importance, ecosystem and recruitment of the Atlantic Bluefin tuna ...... 1.2 Management of Bluefin tuna .................................................................... 1.3 Determining BFT age and growth using otoliths…................................... 1.4 Environmental conditions of the Gulf of Mexico and BFT larvae ............ 1.5 Study objectives .........................................................................................

1 5 8 10 11

2

MATERIALS AND METHODS ................................................................... 2.1 Larval collections and ageing .................................................................... 2.2 Growth comparisons ................................................................................ 2.3 Environmental parameters .........................................................................

12 15 16

3

RESULTS ....................................................................................................... 3.1 Bluefin collections ..................................................................................... 3.2 Age, growth and comparisons.................................................................... 3.3 Environmental parameters ........................................................................

19 19 21

4

DISCUSSION ................................................................................................. 4.1 Bluefin tuna growth ...................................................................................

23

5

CONCLUSION ...............................................................................................

27

REFERENCES…………… ........................................................................................

30

FIGURES …………… ................................................................................................

37

TABLES …………… .................................................................................................

52

v

LIST OF FIGURES Page Figure 1

Map of sampling and density of bluefin tuna in the Gulf of Mexico..….

37

Figure 2

Model output for probability of larval occurrence in the study area…… . 38

Figure 3

Map of GOM, sampling locations, length distribution in 2012… ............. 39

Figure 4

Larval bluefin tuna and larval otolith …………….................................... 40

Figure 5

Mean daily increments and age distribution histogram …… .................... 41

Figure 6

Otolith age at length for 2012 larval bluefin tuna …………… ................. 42

Figure 7

Body length at otolith radius for 2012 larval bluefin tuna…………… ..... 43

Figure 8

Analysis of covariance for BFT growth curves in GOM and MED .......... 44

Figure 9

Scatterplot of sampled environmental parameters from CTD casts ......... 45

Figure 10 Temperature in the GOM during sampling period (5m, 100m)…………. 46 Figure 11 Oxygen in the GOM during sampling period (5m, 100m)…………… .... 47 Figure 12 Salinity in the GOM during sampling period (5m, 100m)……………..... 48 Figure 13 Fluorescence in the GOM during sampling period (5m, 100m)………… 49 Figure 14 Otolith growth measurements and body length………… ......................... 50 Figure 15 Distance based redundancy analysis and bubble plot indicating environmental parameters that influence growth ................................................................................. 51

vi

LIST OF TABLES Page Table 1: Growth for BFT larval studies ……………………………………………

52

Table 2: Environmental indicators analyzed and summary values ………………..

53

Table 3: Distance-Based Linear Model output..……………………..……………..

53

vii

Chapter 1: Introduction

The present study is part of an assessment of the distribution and abundance of the Western Atlantic bluefin tuna ecology with an aim to support an ecosystem-based management approach for BFT stock assessments. In particular, this study intends to provide an updated growth curve for the larval BFT population in the Gulf of Mexico, starting with the 2012 spawning stock.

1.1 Importance, ecosystem and recruitment of the Atlantic Bluefin tuna (BFT) Tunas (Scombridae) are a large family of pelagic fishes that inhabit various oceanic environments throughout the world and are heavily targeted for their high nutritional and economic value. Atlantic bluefin tuna (BFT), Thunnus thunnus, are the largest tuna, reaching up to 900 kg (Fromentin and Powers 2005), and are the species of highest economic value. BFT have been fished for thousands of years in the Mediterranean Sea and more recently throughout the North Atlantic Ocean, including U.S. waters (Scott et al., 1993). BFT have robust, elongated and streamlined bodies that scan the oceans forming large schools to quickly and swiftly satisfy their voracious appetites. In addition they ram ventilate thus they must swim constantly to fulfill their metabolic oxygen requirements. BFT are epipelagic as juveniles and mesopelagic as adults and inhabit the upper 200 meters of the water column in the North Atlantic Ocean, the Gulf of Mexico, and the Mediterranean Sea (Muhling et al., 2013). BFT are capable of thermoregulation which allows them to forage in colder (often more prey abundant) environments, they can also perform deep dives to avoid warmer waters that may result 1

2 in heat stress. BFT are well thus uniquely adapted to undertake extensive migrations between their feeding grounds in the cold waters of the North Atlantic Ocean and the warm but oligotrophic waters of their spawning grounds (Muhling et al., 2010). Minor spawning events have been recorded off the Bahamas (Lamkin, Pers. Comm) and the Mexican Caribbean (Muhling et al., 2010) but BFT spawning is regionally distinct, as nearly all Western stock spawning takes place in the Gulf of Mexico, while the eastern stock spawns principally in the western Mediterranean Sea with strong homing behavior driving the stock’s separation (Rooker et al., 2008). BFT spawning migrations were suspected but were first documented by conventional tagging and have been confirmed by numerous studies using electronic tags (Block et al., 2005) and chemical signatures (Rooker et al., 2008, Rooker et al., 2014). BFT migrate to spawn over a limited time period: April to May in the Gulf of Mexico and June to July in the Western Mediterranean Sea (Block et al., 2005; Brothers et al., 1983, Muhling et al., 2012). Spawning strategies for bony fishes are diverse and are coupled with mechanisms that favor dispersal (or retention) in different scales that can include short distances (meters) for some demersal species to tens of kilometers for reef fishes (Sponaugle et al., 2010) and even larger distances for highly migratory pelagics such as the members of the tuna family. BFT are oviparous, iteroparous, and are multiple batch spawners (Schaefer 2001, Medina et al., 2002), reaching maturity at a mean age between 8 and 12 years in the Western Atlantic Ocean. A fullygrown female may produce between 30 and 60 million eggs (Brothers et al., 1983, Mather et al., 1995). Warm waters with low chlorophyll levels characterize the BFT spawning grounds, as spawning takes place following the seasonal warming period, when temperatures reach approximately 24°C. Adult BFT physiology allows them to swim large distances and to target the edge of strong currents that may facilitate subsequent larval survival. The known aspects of the ecology of BFT indicate that although adult BFT are capable of transatlantic movements, it is likely that they target the Gulf of Mexico to improve the survival of their newly spawned larvae

3 by placement in a favorable habitat where conditions exist wherein prey sources are available. Catches of young-of-the-year (juvenile) BFT in the northwestern Atlantic Ocean support the theory that larval BFT remain in the Gulf of Mexico until they are able to swim and migrate northward likely using oceanic currents like the Florida Current or the Gulf Stream to facilitate reaching their feeding grounds in the northwestern Atlantic Ocean. Although much has been learned regarding BFT migrations to warm oligotrophic waters to deposit their larvae, little is known about their subsequent growth, mortality and recruitment during the early life stages. During the summer spawning season (April to May), newly hatched (pre-flexion) and older (flexion and post-flexion) larval BFT can be found in habitats with statistically predictable conditions in the GOM, as shown by habitat models developed by Muhling et al., (2010, 2012, 2013). Very few studies have examined the fate of wild BFT during the key but largely unstudied post-larval stages which encompasses until they reach about one year old (Brothers et al., 1983). Bakun (2013) noted that although BFT stocks are low compared to the more abundant decades (before 1980), the stock is unusually stable and refers to this as “the Bluefin conundrum.” Given the amount of eggs spawned each year, how is it that the population remains stable, and what factors act to control the population numbers and at what stage? Multiple theories address this conundrum associated with recruitment of fishes that can be applied to BFT. Recruitment research can be traced to Hjort’s “critical period hypothesis” which was proposed as a framework to understand the factors affecting early-life mortality of fish, particularly how early-life processes influence subsequent growth and survival (Hjort 1914). Cushing’s 1969 “match-mismatch hypothesis” suggests that variability in larval growth is determined by the degree that fish spawning is synchronized with food availability (plankton blooms) (Cushing 1990). Sinclair (1988) proposed the “member-vagrant hypothesis,” which states that larval retention in favorable habitats is key to successful recruitment. Houde (1987) highlighted the idea that variation in recruitment might be the result of subtle variations in daily

4 growth and mortality rates in eggs and larvae, and that since larval mortality due to predation is high, faster larval growth during the first days of life should lead to higher survival rates. Pitchford et al., (2005) documented that variance in growth among individuals influences overall survival. Munk (2007) showed that larval growth rates are enhanced at hydrodynamic fronts in the North Sea. Wexler et al., 2007 found temporal variation in growth for Yellowfin tuna and finally, Tanaka et al., (2014) found that fast growth during early larval stages for Pacific BFT influences late-larval growth with an earlier onset consumption of other fishes (piscivory) being the most important factor. All of the aforementioned studies point to the influence of abiotic and biotic factors on larval survival and ultimate recruitment to adult populations. Although the larval habitat preferences of BFT have been examined (Muhling et al., 2010; 2013), specific correlations between environmental variables and BFT larval growth remain largely unknown. In the Gulf of Mexico, BFT spawning takes place in offshore nutrient-poor waters and not in the more productive habitats found on the continental shelf (Muhling et al., 2010). Similarly, spawning activity in the Western Mediterranean Sea around the Balearic Islands is associated with oligotrophic regions characterized by very low chlorophyll concentrations (Laiz-Carrión et al., 2014; Muhling et al., 2013). While regions of high primary and secondary productivity overlap closely with spawning areas of some fish species (Cushing, 1990), this is clearly not the case with BFT. Additional factors appear to determine the suitability of spawning grounds and potential larval survival rates. The concentration of prey (and predators) by oceanographic features, such as the fronts and eddies that often take place in the GOM, combined with early onset of piscivory is likely an important piece of the BFT puzzle (Llopiz et al., 2014). Larval BFT feeding is currently being examined concurrently with this study (Llopiz, Pers. Comm.) and will provide insights into recruitment dynamics by taking into account growth variations as it directly reflects survival (Pepin et al., 2014).

5 1.2 Management of Bluefin Tuna BFT management is multifaceted and complex, encompassing multiagency and international collaborations. BFT are epipelagic oceanic fish that spend their lifetime traveling long distances that cross multiple international boundaries. Furthermore, BFT are highly prized and, after decades of increasing fishing pressure, are listed as an endangered species by the International Union for Conservation of Nature (Collete et al., 2011). Eastern and western BFT stocks are managed separately (Anonymous 2011) and international assessments are carried out to ascertain the yearly biomass for both stocks (Ingram et al., 2010). The eastern stock has been exploited for thousands of years. Harvest of the western stock started in the 1950s (Scott et al., 1993) and is currently considered to be overfished (NMFS 2014b). In accordance with agreements by the International Commission for the Conservation of Atlantic Tunas (ICCAT), the U.S. National Marine Fisheries Service, and more specifically the Atlantic Highly Migratory Species Management (HMS) Division (NMFS 2014) is responsible for BFT management in U.S. waters, including (i) the Atlantic Ocean, (ii) Gulf of Mexico, and (iii) US Caribbean. In these waters, BFT are subject to two regulations: the Magnuson-Stevens Fishery Conservation and Management Act (Magnuson-Stevens Act 1996), and the Atlantic Tunas Convention Act (ATCA 1975). ATCA authorizes the U.S. Secretary of Commerce to implement the binding recommendations of ICCAT. Arguably, the management of this species requires an integrative approach, such as ecosystem-based management, which integrates multiple factors into the study and management of fishery resources instead of focusing narrowly on a single target species (NMFS 2014). To be optimally effective, management of BFT must include a thorough understanding of their life history, environmental influences, and finally must aim to make sound scientific assessments that enable sustainable fishing practices in the ecosystem.

6 Overfishing in the 1970s and 1980s led to the decline of the western stock and since then the spawning stock biomass has remained relatively stable but low in abundance at approximately 15% to 18% of its pre-exploitation biomass (Collette et al., 2014). Despite continued declines in population size and high vulnerability associated with its long-lived ecology, successful measures to conserve BFT have not been implemented. As a result, the status of the western stock has not improved under the current fishing recovery plans (Collette et al., 2014). Tracking abundances of pelagic fishes over time presents a major challenge to managers. Data collected for fishery management usually includes fishery-dependent information that is provided by the fishers. The primary inputs into population models are indices derived from information that is fisheries-dependent, consisting of data from landings, on-board observers, and logbook reports from the fishing fleet (NMFS 2014). In addition, fishery scientists have collected independent but standardized assessments to provide objective information on various aspects of fishery data (Habtes et al., 2014). The only fishery independent assessment derived for BFT in the Gulf of Mexico is carried out as part of the Southeast Area Monitoring and Assessment Program (SEAMAP). SEAMAP is a multi-partnered program for the collection, management and dissemination of fishery-independent data and information in the southeastern United States (Habtes et al., 2014). “SEAMAP stations” are located in a fixed, systematic grid across the Gulf of Mexico at 0.5° intervals (Lyczkowski-Shultz et al., 2013) using multiple plankton gears depending on their target collections for monitoring during research cruises. The BFT larval survey from SEAMAP cruises is currently the only fishery-independent source of information that can be applied to annual population assessments. Using SEAMAP data, Scott et al., (1993) proposed a BFT spawning biomass index. This index uses average larval abundances of BFT collected from the Gulf of Mexico by NMFS

7 scientists using the long-term series of data from the SEAMAP sampling grid. This larval BFT index now takes into account the patchy nature of larval fishes (e.g. zero catches vs. highly abundant catches) and will be further improved once it is able to effectively incorporate information as to physical and biological interactions and mortality of larval BFT, as well as concurrent environmental parameters (Ingram et al., 2010). The Scott et al., (1993) index employs larval-length at day one, derived from otoliths to back-calculate from observed abundances to equivalent abundances (per 100 m2) of one-day old larvae collected from Bongo net tows (Scott et al., 1993; Ingram et al., 2010). Although this index represents the only fishery-independent metric to support stock assessments in the Gulf of Mexico ecosystem, it has a critical shortcoming when used for measuring larval growth, survival and recruitment. As mentioned earlier, the growth curve for larvae constructed from this index represents a very small sub-sample of the total BFT population taken during a relatively short sampling period, three days in May, and one day in June (Brothers et al., 1983), which were not sourced from the main spawning ground (the Gulf of Mexico). The index does not take into account inter-annual and geographical variability in larval growth. In contrast, recent research indicates that growth and survival of BFT larvae for the Mediterranean (Garcia et al., 2013) as well as for the Pacific Bluefin tuna is highly variable, both inter-annually and spatially (Tanaka et al., 2014). Although the SEAMAP database houses 30 years of larval surveys, the age and growth curve for BFT that is incorporated in present research and management procedures is entirely based upon a geographically restricted set of samples collected near Miami, Florida from 19801981. This study by Brothers et al., (1983) is the only published larval ageing study undertaken using BFT larvae from the western Atlantic stock. Their study examined larvae collected during one day near Pacific Light House and three days off Fowey Light House at the edge of the Gulf Stream. The oceanographic environment adjacent to the Miami area has significantly different

8 dynamics from those acting in the Gulf of Mexico. The larvae examined by Brothers et al., (1983) thus represent a very small sub-sample of the total larval population that collected hundreds of kilometers from the main spawning ground in the Gulf of Mexico. By restricting age and growth estimates to a small sample taken during four days during the spawning season, these estimated growth rates do not take into account likely differences. Moreover, these estimates do not consider oceanographic processes, even though these mechanisms are known to drive growth in larval fishes (Sponaugle 2010). A more complete understanding of the impacts of these confounding parameters is necessary to inform stock assessment and advance science-based fisheries management, and would be expected to play a key role in the development of credible predictive recruitment models.

1.3 Determining age and growth using otoliths Otoliths are calcium carbonate structures found in all fishes. Daily and annual increments in the otoliths form as a result of the circadian rhythm and seasonal patterns reflected by the surrounding environment allow a historical analysis of the life history of a fish (Campana 1992). A banding pattern can be observed on whole otoliths from adult fishes when examined under at least 10x magnifications. In most cases, sectioning is necessary to accurately estimate annual ages of fishes. Daily increments can also be distinguished on whole larval otoliths as they form with daily frequency and can be observed using higher magnifications ( >40 – 400 x) that can be measured using image analysis software (Campana and Jones 1992; Sponaugle 2009). Repeated measurements (e.g. increment width and otolith radius) of sufficient sample sizes can reveal patterns in otolith growth (e.g. fast or slow) and reflect spatial and temporal variability. Increments represent a continuous timeline in which one day’s growth is not independent of the previous day.

9 The age of larval BFT in days can be determined directly by examining their otolith microstructure (Brothers et al., 1983; García et al., 2013). Larval BFT otoliths are easily viewed under high magnification (100 x -1000 x) and like most fishes, have recognizable patterns that reflect continuous formation of daily increments (Secor 1995). BFT daily increments were validated for Pacific bluefin tuna by Itoh et al., (2000) in laboratory-reared fish from day 5 to day 71 after fertilization. In BFT otoliths, the primordium is encircled by one or more diffuse zones followed by alternating bands (daily increments) formed by L-zones and D-zones (Brothers et al., 1983, Franks, Pers. Comm.). Previously, otoliths of Atlantic BFT have been aged from larval specimens collected near the Florida Straits (Brothers et al., 1983) and also from the Baleares Sea in the Mediterranean Sea (García et al., 2013). Besides extensive larval work on Pacific bluefin tuna Thunnus orientalis (Tanaka et al., 2006, 2012, 2014; Miyashita et al., 2001), other tuna previously aged include (i) yellowfin tuna Thunnus albacares (Wexler et al., 2007; Lang et al., 1994), (ii) skipjack tuna, Katsuwonus pelamis (Tanabe et al., 2003), and, (iii) bullet tuna Auxis Rochei, collected in the Mediterranean Sea (Laiz-Carrión et al., 2013). Pelagics from the Atlantic Ocean that have been aged during their larval or early life history stages include Atlantic blue marlin Makaira nigricans (Sponaugle et al., 2005), Atlantic sailfish Istiophorus platypterus (Luthy et al., 2005) and swordfish, Xiphias Gladius (Govoni et al., 2003). Recent research reported by García et al., (2013) showed that warm anomalies in the western Mediterranean during 2003 resulted in increased growth rates, which may translate into stronger recruitment and subsequent large year classes. Tanaka et al., (2006, 2014) reported similar results for Pacific Bluefin tuna. Tanaka et al., (2014) observed larval growth was greater for those individuals that had larger sized otoliths. However, research of this nature has not been completed for the western stock of BFT tuna to date. Evaluating the Western BFT larval tuna growth may resolve large recruitment differences found between the East and West stocks.

10 1.4 Environmental conditions of the Gulf of Mexico and BFT larvae Although biotic factors influence growth, abiotic factors in biophysical environment in the Gulf of Mexico strongly influenced by oceanographic circulation patterns. The main mesoscale feature in this ecosystem is the Loop Current, a northward extension of the Yucatan Current, which exits the Gulf through the Straits of Florida to become the major source for the Gulf Stream (Lindo-Atichati et al., 2012). The Loop Current forms an intense anti-cyclonic flow in the eastern Gulf, and may extend northwards as far as 28°N (Hurlburt and Thompson 1980). These northward extensions may last for several months, until the current ultimately becomes unstable and sheds a large warm-core anti-cyclonic ring (Maul and Vukovich 1993). The rings drift westward, where they spin up smaller cyclonic rings along the shelf, resulting in complex and dynamic oceanographic conditions in the northern Gulf of Mexico (Lindo-Atichati et al., 2012). In the Gulf of Mexico’s northern shelf, cyclonic circulation is driven by along-shore wind stress, a strong coastal current within the inner shelf and a weak broad current over the outer shelf (Zavala-Hidalgo et al., 2003). The lower salinity waters that flow into the Gulf of Mexico from the Mississippi and Atchafalaya Rivers is advected westward along the shelf that in turn develops along-shelf fronts. Moreover, this advected water mass enhances local productivity as a result of transporting high nutrient and sediments as observed in the higher concentrations of chlorophyll-a found in this area (Biggs and Muller-Karger, 1994; Nababan et al., 2011). These various oceanographic features structure the Gulf into habitats with different physical and biological characteristics, with associated surface fronts delineating boundaries between different surface water features. Eddies and fronts act as mechanisms of enrichment, concentration and retention, which could in turn benefit larval growth and survival (Bakun 2006). BFT larvae are distributed across these features and larvae spawned are likely exposed to

11 different ecological regimes that include hydrodynamic regimes over the several weeks and months before they grow into the juvenile stage. The juvenile stage begins when these fishes have developed sub adult characteristics that include fast swimming capacity and low temperature tolerances. 1.5 Study objectives This study aims to demonstrate that the updated growth curve is statistically different from the previously available estimates from the Brothers et al., (1983) study. Moreover, this study will search for evidence to disentangle the influence of environmental indicators may have on BFT growth. Updating the larval BFT growth curve, together with a deeper understanding of the influence of environmental variables, would be valuable in defining a more accurate larval index of abundance. Improving the ecology of larval BFT will contribute to ongoing efforts that examine ecosystem and population oriented studies, such as feeding success, prey types, habitat modeling, and climate change scenarios.

This study has four specific objectives: a) To estimate the daily age from Bluefin tuna otoliths from the 2012 Gulf of Mexico spawning season, b) To develop new growth curves using age at larval length (mm) and otolith radius (μm) c) To compare age at length estimates with those proposed by Brothers et al. (1983) and Garcia et al. (2013) d) To identify the apparent influence of environmental variables on Bluefin tuna aged larvae from the Gulf of Mexico

Chapter 2: Materials and methods 2.1 Larval collections and ageing Plankton surveys have been completed by National Marine Fisheries Service in the Gulf of Mexico since 1982 to characterize spawning periods and determine larval fish abundances concurrently with oceanographic observations with the goal to establish a long-time plankton reference series (Habtes et al., 2014; Richards and Potthoff 1980; McGowan and Richards 1986; Lyczkowski-Shutlz et al., 2013). For this study, BFT larvae were collected for age estimation during the summer, the known spawning period for BFT (Scot et al 1993, Richards 1976). Sampling took place continuously (24 hours operations) from April 30 through May 2012 targeting historical stations within the US exclusive economic zone of the Gulf of Mexico outside the 200m depth contour that form part of the SEAMAP spring plankton survey (Millet et al., 2012). Additional stations were sampled if large numbers of Bluefin tuna (n > 50) were recorded during preliminary identification of plankton collections by trained taxonomists on board. Two 1 x 2 m plankton nets were towed at approximately 2 knots from the NOAA Ship Gordon Gunter. Nets were towed for 10 minutes portside as the ship made a wide circular pattern and remained within 1 nm of the station coordinates. The surface neuston (NN) net had a mesh size of 0.947 mm and collected the uppermost layer of the water column, approximately half (0.5 m) of the net’s mouth remained at the air-sea interphase and the rest (0.5 m) remained submerged. Often large patches of sargassum and other seagrasses were also collected, thus all large (> 5 cm) debris were removed, recorded, and discarded. The subsurface Neuston (S10) net had a mesh size of 0.505 mm and was towed in an undulating fashion “yo-yo” manner from 0-10m as described in Habtes et al., 2014. The S10 had a mechanical flowmeter (model 2030R, General Oceanics, Inc) mounted at the center of the net to record the volume filtered by the net. The flowmeter counts were recorded prior to each tow and immediately upon recovery of net. Plankton samples were 12

13 preserved in 95% Ethanol to conserve tissues and otolith structures. After 24 hours, fresh preservative (95% Ethanol) was changed to ensure proper fixation of samples. Wet plankton volumes for S10 samples were calculated as a measure of overall zooplankton abundance. The majority of larval fishes were subsequently sorted and identified at the Sea Fisheries Institute, Plankton Sorting and Identification Center in Szczecin, Poland under a collaborative agreement with NMFS following taxonomic keys and SEAMAP protocols (Richards 2006; Lyczowski et al., 2013). Tuna (Scombrid) larvae are challenging to identify to species level given similarities in pigmentation patterns especially within the Thunnus genera; thus identifications were confirmed at the Southeast Fisheries Science Center by Dr. William Richards, Dr. John Lamkin and myself. All S10 net samples and a subset of NN net samples that were positive for BFT were subsampled for length measurements and assigned a nine-digit larval ID sequential number to remove bias. Up to ten BFT were measured per station, if more than 10 individuals were collected, then 40% of the total rounded to the nearest whole number were measured to represent the range of collected sizes and larval stages. Standard length (SL, nearest 0.1 mm) was measured following Brothers et al., 1983 by starting at the dorsal tip of the jaw or mouth to the proximally to either the notochord (if pre-flexion) or to the developing hypural plates (flexion, post-flexion) using a dissecting microscope (Leica M205C) equipped with a digital camera (Leica DFC290HD) and previously calibrated image analysis software (Image Pro Plus 7). Shrinkage is known to occur when preserving fish larvae (Brothers et al., 1983), but it was assumed to be constant and proportional for all fish in this study. A total of 100 BFT larvae were selected for ageing, covering post yolk-sac to post-flexion stages. Larvae were selected from a subset of samples covering the spatial and temporal extent of the 2012 Gulf of Mexico within the US Exclusive Economic Zone (Figure 3). Sample collection often damages larvae, so all larvae were first examined for any apparent damage which could confound length measurements. Larval fish were then submerged in a distilled water bath for one

14 minute to soften the body, followed by dissection under a stereomicroscope on a glass microscope slide previously inscribed with the assigned larval ID number. Up to four otoliths (sagittae and lapilli) were dissected from the cranial cavity of each larva using minutien pins and sharpened glass probes. Extraneous tissue was carefully removed and otoliths were allowed to air dry overnight. Otoliths were transferred into one drop of mounting medium, Flo-TexxTM, ensuring that all otoliths sank to the bottom of the microscope slide with the distal side up, and dried overnight. Sagittal otoliths are usually the largest structures amongst the three pairs (Itoh et al., 2000), however, for very small larvae, all otoliths resembled each other in shape and sizes, thus the otolith diameters were measured at widest diameters and the largest observed distance was assigned as the sagittal otolith. Previous studies have found no differences between ageing estimates derived from the left or right otolith (Campana 1999), thus one randomly chosen sagittal otolith was selected from each larva. Each otolith was blind read twice using digital micrographs captured with an Olympus BH2 compound microscope (40x-1000x) with a drop of immersion oil under transmitted light. The reading axis was determined using the Image Pro Plus 7.0 caliper tool to mark and measure otolith features. After the two reads, mean age values were assigned to each larvae. The coefficient of variation (CV) was calculated to measure the precision of multiple reads for each larval fish (Chang 1982): 2 �∑𝑅 (𝑋𝑖𝑖 − 𝑋𝑗 ) 𝑖=1 𝑅−1 𝐶𝐶 = 100% × 𝑋𝑗

The otolith radius (OR) indicated the longest axis and for consistency, was chosen as the reading axis for all age estimates. Following Brothers et al., l983 the OR was measured from the center of the primordium to the otolith margin (edge) (Figure 4b). Sub-daily increments were avoided by only counting D-zones which were continuous for at least 50% of the individual

15 increment’s circumference. Each daily increment was enumerated and subsequent individual increment widths were recorded at the edge of each D-zone using the Otolith MacroTM (Media Cybernetics, Inc.) within Image Pro Plus 7.0. Otolith size increases daily with length, thus increment widths from the 100 individual otolith images were measured at the edge of each Dzone and the mean values were plotted with larval body length (standard length, mm). It takes approximately 72 to 96 hours for BFT hatchlings to absorb their yolk salc, mouth development to be completed and thus the onset of feeding begins with the first increment formation (Brothers et al., 1983, Itoh et al., 2000). Previous age estimates for larval BFT added four days to the final increment count in order to calculate a total age in days (Brothers et al., 1983, Itoh et al., 2000, García et al., 2013). The growth curve (mean age at length) from the 100 larvae examined in this study was applied to all measured BFT from the 2012 cruise (n = 479), and a histogram was created to estimate the age distribution of the 2012 spawning BFT in the GOM. Approximate spawning dates were calculated by subtracting the age (days) from the collection dates following Lang et al., (1994).

2.2 Growth comparisons Published ageing data from other studies on larval Atlantic BFT were compared with results from this study. As part of an ongoing collaboration, the larval BFT otolith data published in García et al., (2013) was provided by the lead author (García, Pers. Comm). In addition the Brothers et al., (1983) dataset was digitized by extracting figures 3 & 4 from the publication and using Surfer 9 to carefully extract all data points into a spreadsheet. Graphical comparisons were generated among all studies, including the current one, for otolith growth units (days) vs. BFT length (SL, mm) and for BFT length (SL, mm) vs. otolith radius (µm). Least squares regressions were performed for best fit to all curves).

16 Individual growth rates (mm-1) were calculated by dividing the observed length (SL, mm at capture) with the assigned age (days). To compare observed differences among growth curves from Garcia et al., 2013, Brothers et al., 1983 and from this study, length at age relationships were tested among studies using an analysis of covariance (Proc GLM, SAS Inc.), with length utilized as a continuous covariate (Figure 8). To compare overall growth, the slopes of each curve were utilized.

2.3 Environmental parameters At each sampled station in the Gulf of Mexico, a CTD (Seabird SBE 9/11 plus CTD) was used to collect oceanographic data. CTD casts were processed using SBE Data Processing software following SEAMAP protocols, and binned to 1 m depth intervals (Lyczkowski-Shultz et al, 2013). The following variables were examined in this study: i) temperature (°C), ii) salinity, iii) dissolved oxygen (mg/L), and iv) fluorescence (µg/l) including maximum values and the depth at which they occurred during each cast. Temperature, salinity and fluorescence were extracted for analysis at two depths: near surface (mean 0 – 5 m) and at 100m. Conditions at the sea surface were selected to characterize the larval environment, as BFT are usually collected in the upper water column (Habtes et al., 2014). Temperature, salinity and fluorescence at 100 m depth trace water mass structure, and can be used to identify water of Loop Current origin, including warm-core eddies (Muhling et al., 2010). Dissolved oxygen at 100m was included for the same reason. The maximum fluorescence value throughout the water column was also included as a potential explanatory variable, as was the water depth at which the maximum value occurred at each station. To visualize spatial patterns in the oceanographic environment during the 2012 cruise, contour plots of temperature, salinity, oxygen and fluorescence at near-surface and 100m depth

17 were constructed in Surfer 9 (Golden Software, Inc., Golden Co), using the kriging method of interpolation. Locations of aged BFT larvae were overlaid to investigate patterns of positive aggregations or avoidance of the habitat (Figure 9,10,11,12). Multivariate statistical models were then used to test the influence of environmental characteristics on the observed growth of BFT larvae. In addition to environmental variables extracted from CTD casts, the following parameters were also included: vi) longitude, vii) latitude, viii) date of collection, and iv) scombrid density (total abundance or the total number of tuna larvae per m3). A Pearson product moment correlation matrix among all environmental factors was used to highlight strongly correlated predictor variables. “Oxygen at 100m” and “Date Spawned” were strongly correlated with other variables (p > 0.5), and were excluded from further analyses. A Kolmogorov-Smirnov test ran using SigmaPlot 12.3 showed that all environmental variables were not normally distributed; therefore non-parametric statistical procedures were used instead, in Primer-E with Permanova+ software (Clarke and Gorley, 2006). Larvae aged in this study ranged from very young yolk-sac larvae to older post-flexion larvae . Growth rates from other studies show a strong effect of larval age on growth, with older larvae growing faster, in terms of both SL and otolith radius (Tanaka et al., 2014). To account for this effect in our analyses, the mean recent otolith growth was calculated by measuring the mean increment width (μm) of the last 3 increments of each larva, representing recent larval growth. Mean recent growth has been utilized as a tool to examine potential differences caused by feeding regimes on otoliths (Aguilera et al., 2009), the influence of the environment on otolith growth (Shulzitski, 2012) and most recently on body condition of larvae (Zenteno et al., 2014). Mean recent growth was used to examine potential differences caused by the influence of the environment on otolith growth. This metric was then regressed on larval length, and an

18 exponential line of best fit applied. Residual values from this line were considered to represent unexplained variability in recent growth, which may have been caused by ambient environmental conditions. To assess how environmental conditions influenced recent larval BFT tuna growth in the Gulf of Mexico, a Distance-Based Linear Model (DISTLM) analysis in Primer was used (McArdle and Anderson, 2001). This routine uses a Euclidean-distance matrix of environmental variables to explain variability in an additional target matrix, which in this case describes recent growth variability. The analysis is based on permutation, and is therefore distribution-free, removing the requirement for normally-distributed data. As environmental parameters are typically measured on different scales, first variables were plotted using Primer-E’s Draftsman plot tool to ensure that no variables were skewed or had extreme outliers all variables were normalized before analysis, by subtracting the mean and dividing by the standard deviation within each variable (Clarke and Gorley, 2006). Within the DISTLM routine, the “best” selection procedure was specified to examine all possible combinations for defining the optimal subset environmental variables to include in the final model with the AIC (Akaike’s Information Criteria) as the selection criteria specified. The contribution of the final model to explaining variability in recent larval BFT was defined using the calculated R2 parameter and important variables were identified by statistically significant p values (p < 0.05) within the marginal tests performed by the DISTLM routine.

Chapter 3: Results 3.1 Bluefin collections A total of 1,346 bluefin tuna larvae were identified from 208 stations sampled in the GOM during the Spring SEAMAP 2012 cruise from all plankton gear deployed during the research survey legs 2 and 3 (30 April -27 May 2012). Tuna catches are patchy and rare, as a result, positive BFT catches (n ≥ 1) encompassed only (35%) of the stations sampled in the GOM. BFT densities (m3) peaked in after 16 May, 2012. BFT were not collected in waters within the Loop Current during our research survey (Figure 1). Although only a subsample of all BFT catches was measured (38%), BFT larvae represented the observed lengths and geographical locations of the most 2012 BFT collections (Figure 3). Mean standard length was 5.09 ± 0.05 and ranged in sizes from 2.13 to 8.76 mm SL (Table 1) with just one larval BFT measuring 10.88 mm SL as the largest (and oldest) individual measured during the 2012 survey. A histogram of the observed lengths for the 2012 BFT population was generated and the observed growth curve was utilized to estimate the ages of the subset of the collected larval BFT in 2012.

3.2 Age, growth and comparisons A total of one hundred larvae from 29 stations within the US EEZ were aged throughout the GOM collected in May 1-28, 2012. The majority of larvae (n=96) were collected using the S10, and only four were collected using the NN net (Figure 2). Despite having a larger mesh size than the S10, the Neuston net did not collect larger larvae as we had expected. All otoliths examined with transmitted light displayed a primordium encircled by either one or two diffuse zones followed by daily increments formed by bipartite structures composed of 19

20 a transparent layer (continuous L-zones that appear white) and a darker often-wider layer (discontinuous D-zones that appears dark), (Secor 1995) as seen in Figure 4b. Difficulty in distinguishing the first increment as well as large variability in increment widths made some otoliths difficult to read, however no otoliths were excluded for ageing. The mean CV was 4.2 ± 0.4 % (mean ± SE) and ranged from 0-15.7%. These measures of precision are the first reported values for Atlantic bluefin tuna from the western Gulf of Mexico. Accepted CV values fall less than 10%, however differences in interpreting otolith structure is observed in most otolith studies (Campana and Jones 1992). Similar to Itoh et al., (2000) individual increment measurements from 100 otoliths showed that growth was relatively uniform during the first week in life (1-2 μm), then increasing exponentially (>3 μm) and for older fish, having higher variability in the calculated SE (standard error) as the fish laid down more daily increments and became older aged (Figure 5a). This study shows a linear fit for age at length curves with similar slopes (growth rates) to the linear regressions published for larval BFT from the Mediterranean (Garcia et al., 2013). Ages for BFT ranged from 4-18 days post-fertilization are shown in Figure 6. The least squares regression in the linear form was fitted to the age at length data resulting in: y = mx + b (n = 100, m = 0.45, b= 0.53), r2 = 0.88 (Figure 6). Calculated spawning dates for this data set started on 16 April 2012 and continued almost daily through 21 May 2012 implying daily spawning events in the GOM during 2012. Overall, otolith radius (OR), μm had a positive relationship with body length similar to Brothers et al., (1983) as shown in Figure 7. Mean OR was 31.10 μm ± 1.64 (mean ± SE) and ranged from 10.90 - 98.90 μm (Table 1). The least squares regression in the exponential form was fitted to the observed OR measurements resulting in: y=a℮bx (n =100, a = 5.04, b = 0.33, r2 = 0.93). Results from this study are within previously reported values for daily increment values by

21 Itoh et al., (2000) and Brothers et al., (1983). The OR for the first increment measured 11.20 µm ± 0.09 (mean ± SE) corresponding to a 4 day old larval BFT. In comparison, the smallest OR reported in Brothers et al., (1983) was 19 μm corresponded to a 4.5 mm SL fish. It is important to note that in Brothers et al., (1983) a coarser mesh size on all of their plankton collections likely missing smaller larval BFT in their growth curves. Garcia’s smallest OR was 9.90 μm corresponding to a 3 mm SL larva. In contrast to Brothers et al., (1983), Garcia et al., (2013) included smaller and (likely younger larvae) in their study. Comparisons between the digitized data points from Figure 3 in Brothers et al., (1983) and raw data from Garcia et al., (2013) for the years 2003, 2004 and 2005 were examined with the current study for evaluation and analysis of covariance showed significant differences between this study and each of the other BFT growth curves (p < 0.001), see Figure 8. Mean values for size (SL, mm) at age (days), OR (μm), and growth rates (mm day-1) are reported and compared in Table 1. Growth was highly variable for the larvae examined. The observed mean growth rate for the 100 BFT larvae was 0.51 ± 0.6 mm day-1 (mean ± SD) with an overall slope of 0.45. These values are lower than Brothers et al., (1983) and within the variability found in the growth rates for the Garcia et al., (2013) study. In our dataset, larvae smaller than 2.13mm SL are not collected (and thus not aged) likely due to net extrusion for newly hatched BFT nor did it include larvae larger than 8.10 mm SL likely due to net avoidance by post-larval BFT’s excellent swimming behavior. 3.3 Environmental parameters Contour maps using Surfer 9, (Golden Software, Inc) were made for to include all available CTD casts (n = 205) of the following environmental parameters at 5 m and 100 m whenever available: i) temperature (°C), ii) salinity, iii) dissolved oxygen (mg/L); and iv)

22 fluorescence (µg/l) (see Figures 9-12). A summary of the values measured for each parameter including means and ranges (minimum and maximum) are shown in Table 2 including the GOM during the survey. Larval fish collected in the Gulf of Mexico had variable growth rates as observed in both lengths at age and with individual otolith increment measurements. Values for recent otolith growth for the last 3 days was averaged 4.41 µm ± 0.24 (mean ± SE) with a range of 1.17 µm10.87 µm. BFT aged represented a broad spectrum of the early young stages for BFT. In stations that had more than one aged BFT, no correlation was found to associate size with location. Statistically significant evidence of an association between growth and environmental parameters were found. The distance-based linear models (DISTLM) modeled the relationship between the multivariate environmental data as described by a resemblance matrix and the residuals of recent otolith. The DISTLM model output is shown in Table 3. The overall best solution selected five parameters (F100 m, S100 m, F5 m, O5 m, volume filtered) explaining 20.35% of the variability. However, the low value for R2 reflects weakness in explaining the variability of the growth rates. A subset of the values was found to be statistically significant: temperature at100m (p = 0.011), fluorescence at 100m (p = 0.026) and the depth of the fluorescence max (p = 0.011).

Chapter 4: Discussion 4.1 Bluefin tuna growth Despite being the only western Atlantic spawning ground for bluefin tuna, larval growth and mortality studies have not been undertaken to date within the GOM. During the 2012 survey, larvae were distributed throughout the GOM, however, collections were patchy. Calculated spawning dates from this study were almost continuous, possibly reflecting clumped but frequent spawning during the spawning season for the western stock of bluefin tuna. The mean observed age was 10.17 days ±0.33 (mean ± SE) and the majority of fish aged (90%) were older than 8 days old. The 1981 collections from Brothers et al., (1983) included a larger size range of larvae; however our results indicate slower growth than Brothers et al. (1983). In the GOM, a combination of younger and older conspecific larvae was found. Considering that piscivory has been found early when compared to other tuna species (Llopiz et al., 2014), having different life stages and size ranges within one given location is likely nutritiously advantageous for the older larvae and removes younger individuals from the equation as a result of predation. Although sampling took place in May for both this study and for Brothers et al., (1983), a direct comparison is difficult given the disparity in sampling durations and geographic coverage. The larvae aged in Brothers et al., (1983) were collected at the inshore edge of the Gulf Stream (Fig.1) as it passes Miami on two separate sampling efforts in May (Brothers et al., 1983). These larvae represent a small subset of bluefin in the GOM and would not reflect the wide range of temperatures and feeding conditions that BFT larvae may be exposed to in the rest of their spawning grounds. For instance, SST ranged from 21° to 29°C during the sampling period for the 2012 survey, and many of the larvae may be entrained in cyclonic or anti-cyclonic eddies with

23

24 differing temperatures and prey availability. The presence of features (such as eddies) and the LC within the GOM creates zones of convergence that can in turn accumulate prey items (zooplankton and pelagic larval fishes) and predators (Lindo-Atichati, et al., 2012). The SEAMAP plankton spring survey samples the Gulf of Mexico during the majority of the bluefin tuna spawning season. This maximizes opportunities to collect larvae exposed to a variety of environmental conditions which will likely be reflected in otolith microstructure. Future efforts will focus on determining growth rates on a weekly basis during the spawning season in the GOM. In addition, we will investigate the contribution of mesoscale oceanographic features to variability in growth rates. Recent research indicates that growth and survival of bluefin tuna larvae is highly variable, both inter-annually and spatially (Garcia et al., 2013). Observed growth rates have similar slopes indicating comparable growth for larvae spawned in the GOM for 2012 and the Mediterranean from 2004 and 2005. Upon completion of this work, additional otoliths will be aged to include a wider range of sizes, collection locations and same years to compare GOM larvae to Mediterranean spawned larvae. In addition, further ageing of available larvae can determine growth and mortality rates in different environments and larval transport through oceanographic features to greatly improve our understanding of yearly fluctuations in growth. These data are critical to improving our understanding of factors affecting recruitment which are essential for parameterization of stock assessment models, and thus effective management. Limitations of this study include formulating estimates for BFT age and growth with only 100 specimens. Although the study collected fish from a wide geographic area in the US Gulf of Mexico, additional samples would provide better estimates of the natural variability in otolith growth for the 2012 spawning season. In addition, adding BFT from the Mexican territory would

25 enhance this study by providing an even larger geographic scope. Ageing larval fishes involves providing a best estimate of the number of increments in each otolith, nevertheless the growth curves calculated include measured fish length which carries measurement error as well as shrinkage that occurs from the saltwater environment to preservation in 90% Ethanol. Another issue that the error involved in determining growth has consequences to influence mortality rates. Mortality estimates are highest during the larval stage and estimates of instantaneous mortality will be calculated and addressed in a subsequent study utilizing larval fish increment data from this 2012 spawning season. With the current study, there is not enough data to formulate a reliable calculation. Another limitation of this study relates to otolith growth data when interpreted with environmental data. There is a lag time between the actual environmental interaction of a larval fish and the increment formation because it has to assimilate into the metabolism. Increments form following the natural circadian rhythm near 24 h mainly dictated by the amount of light. The larval fishes collected in this study were found in the top 10 m, thus they are constantly exposed to natural light (as opposed to larval reef fishes that can occupy depths of up to 100m). The Gulf of Mexico is a dynamic environment with respect to the presence of mesoscale and sub-mesoscale features (Lindo-Atichati et al., 2012). During the sampling period in 2012, the northern edge of the LC extended northward to 26-26.5° which is located at the southernmost edge of the US EEZ, however our survey was limited to sampling in US waters and measurements south of 26° were not permitted. In addition, although a cyclonic eddy was located western GOM, it was also south of US EEZ and we purely sampled its northern edge (visible in figure 9a). Thus, all aged BFT larvae were collected in the outer shelf at the top of the water column (depths 0-10 m) which is well within the mixed layer (25-35 m) in May (Mendoza et al., 2005). Values for surface temperature, salinity and saturated oxygen at stations positive for larval BFT catches were similar across the GOM surveyed area. Surface temperature, salinity and

26 saturated oxygen fields showed low spatial variability, the latter with values higher than 5.2 mg/l. Surface temperature ranged between 24.4 - 27.8 °C and surface salinity ranged 33.57 - 36.63. Lower salinity values were observed close to the northeastern coast. The spatial distribution of these surface (5 m) parameters contrasted with the same values at 100 m, where temperature values dropped to up to 17.81 °C far away from the Loop Current’s influence. A good visual correlation can be observed with an increase the salinity values and a reduction of the oxygen concentration.

Chapter 5: Conclusion Simply put: in order for an individual to eventually recruit into adulthood, fish larvae must find food, avoid predation and avoid advection to unsuitable habitats. Recent studies have begun to investigate these processes in larval fishes by examining diets, feeding success and relating them to trophodynamics (Pepin et al., 2014, Llopiz et al., 2014; Laiz-Carrión et al., 2014) and by backtracking BFT larvae from capture locations to presumed spawning locations using known age (from otoliths) at length (Turner, Pers. comm.). More specifically, Pepin et al., (2014), provided an insight into the variability of growth with feeding success across different taxa that showed that faster growth was achieved in larval fish with more variable growth rates as reflected in larval otolith increments. The next step of this BFT ageing study is to incorporate the feeding success of individual larvae to adequately incorporate known larval growth. The coupled known age-length, locations of capture as well as feeding success (or failure) can provide an assessment of how larvae collected in the Gulf of Mexico ecosystem are surviving. Even further, prey preferences (and incidence of piscivory) along with how long they remain in the small size stages most vulnerable to predation, and how far they have traveled from spawning locations, the age (days) of each larva must be known. Examination of otoliths of surviving juveniles suggests fast growth as indicative of favorable conditions and higher survival rates (Brothers et al., 1983, Tanaka et al., 2014). A future step is to incorporate known age of individual larval BFT into transport estimates and their interaction with observed oceanographic features. Despite the limitations of this study, enhancing the known BFT growth will greatly improve our ecological understanding of yearly fluctuations in growth, and will lead to stock specific growth estimates for BFT. Stock assessments can influence decisions made by regional and international managers (ie. ICCAT)

27

28 which can in turn have an impact on the fishing quotas and other conservation strategies to manage the stock. Providing adequate ecological data on important fisheries resources such as Bluefin tuna stocks can enhance ecosystem based fisheries management in the Gulf of Mexico. Including larval abundance and growth rates using otoliths will improve the larval index that is used in the only fisheries independent assessment for this species. The central questions which must be addressed in order to improve the western Atlantic BFT stock assessment therefore relate to drivers of recruitment. If less than 1% of larvae typically survive to become juveniles, then changes in larval mortality can cause large fluctuations in recruitment. These processes may result in a de-coupling of recruitment from spawning stock biomass, depending on the factors which are most influential in determining larval survival. Large scale events such as the Deep Water Horizon spill in 2010 highlight the importance of understanding ocean circulation on a larger scale as well as meso and small scale processes (Gyory et al. 2010). This connectivity is important not only for biological particles, but also for water quality, for pollution and invasive species such as the relatively recent invasion of the lionfish into the Caribbean and East coast of the USA. Likely yearly ageing of the larval bluefin tuna cohorts will be necessary to build a predictive model for larval growth in the GOM. Although environmental variables were not strongly influential for explaining recent otolith growth, temporal variability must be assessed before reaching further conclusions. Currently, the Brothers et al., (1983) growth curve uses older larvae from Florida and this can inflate the already highly variable larval BFT index by overestimating abundances for the BFT population. Improving the data that develops the growth curve that is utilized in the larval index has the potential to reduce current uncertainty in the prediction of relationships between spawning stock biomass and recruitment in the GOM (Ingram, Pers. Comm). Stock assessments for fish populations provide estimates of abundance but also track temporal variation that can reveal trends (increases vs decreases) over time and can

29 highlight strong recruitment year classes to be harvested (or protected) as needed. Larval ages derived from otoliths are critical to interpret the factors affecting BFT recruitment, which are essential for supporting stock assessment models, and thus effective ecosystem based management for the Western and Eastern Bluefin tuna stock.

29

REFERENCES Anonymous. (2011) Report of the 2010 Atlantic bluefin tuna assessment session. Collective Volume of Scientific Papers, International Commission for the Conservation of Atlantic Tunas, 66: 505-714. Bakun, A. (2006) Fronts and eddies as key structures in the habitat of marine fish larvae: opportunity, adaptive response and competitive advantage. Scientia Marina, 70: 105-122. Bakun, A. (2013) Ocean eddies, predator pits and bluefin tuna: implications of an inferred ‘low risk–limited payoff’ reproductive scheme of a (former) archetypical top predator. Fish and Fisheries, 14: 424-438. doi:10.1111/faf.12002. Aguilera, B.; Catalan, I., Palomera, I., Olivar, M.P. (2009) Otolith growth of European sea bass (Dicentrarchus labrax L.) larvae fed with constant or varying food levels. Scientia Marina (Barcelona) 73.1: 173-182. Biggs, D.C. and Muller-Karger, F.E. (1994) Ship and satellite observations of chlorophyll-a stocks in interacting cyclone-anticyclone pairs in the western Gulf of Mexico. Journal Geophysical Research, 99: 7371–7384. Block, B.A., Teo, S.L.H., Walli, A., Boustany, A., Stokesbury, M.J.W., Farwell, C.J., Weng, K.C., Dewar, H., Williams, T.D. (2005) Electronic tagging and population structure of Atlantic bluefin tuna. Nature, 434: 1121-1127. Brothers, E.B., Prince, E.D. Lee, D.W. (1983) Age and growth of young-of-the-year blue fin tuna, Thunnus thynnus, from otolith microstructure. Pages 49-60 in E. D. Prince and L. M. Pulos, eds. Proceedings of the international workshop on age determination of oceanic pelagic fishes: tunas, billfishes, and sharks. NOAA Technical Repository National Marine Fisheries Service 8, 211 pp. Campana, S.E. (1992). Measurement and interpretation of the microstructure of fish otoliths. Pages 59-71 in D.K. Stevenson and S.E. Campana, eds. Otolith microstructure examination and analysis. Canadian Special Publication of Fisheries and Aquatic Sciences, 117. Campana, S.E. and Jones, C. (1992). Analysis of otolith microstructure data. In: Stevenson D.K., Campana, S.E., eds. Otolith microstructure examination and analysis. Canadian Special Publication of Fisheries and Aquatic Sciences, 117: 73-100.

30

31

Campana, S.E. (1999) Chemistry and composition of fish otoliths: pathways, mechanisms and applications. Marine Ecology Progress Series, 188: 263-297. Clarke, K.R. and Gorley, R.N. (2001 & 2006) PRIMER v6: User manual/tutorial, PRIMER-E, Plymouth UK, 91pp & 192pp. Collette, B., Amorim, A.F., Boustany, A., Carpenter, K.E., de Oliveira Leite Jr., N., Di Natale, A., Die, D., Fox, W., Fredou, F.L., Graves, J., Viera Hazin, F.H., Hinton, M., Juan Jorda, M., Kada, O., Minte Vera, C., Miyabe, N., Nelson, R., Oxenford, H., Pollard, D., Restrepo, V., Schratwieser, J., Teixeira Lessa, R.P., Pires Ferreira Travassos, P.E., Uozumi, Y. (2014) Thunnus thynnus. The IUCN Red List of Threatened Species. Version 2014.2. <www.iucnredlist.org>. Accessed on September 16, 2014. Cushing, D.H. (1969) The regularity of the spawning season of some fishes. Journal of Conseil International pour l'Exploration de la Mer, 33: 81-92. 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: 249-293. Fromentin, J.M., Powers, J.E. (2005) Atlantic bluefin tuna: population dynamics, ecology, fisheries and management. Fish and Fisheries, 6: 281-306. García, A., Cortés, D., Quintanilla, J., Rámirez, T., Quintanilla, L., Rodríguez, J.M., Alemany, F. (2013) Climate‐induced environmental conditions influencing interannual variability of Mediterranean bluefin (Thunnus thynnus) larval growth. Fisheries Oceanography, 22(4): 273-287. Govoni, J.F., Laban, E.H., Hare, J.A. (2003) The early life history of swordfish (Xiphias gladius) in the western North Atlantic. Fishery Bulletin, 101:778-789. Habtes S., Muller-Karger, F., Roffer, M., Lamkin, J.T., Muhling, B.A. (2014) A comparison of sampling methods for larvae of medium and large epipelagic fish species during spring SEAMAP ichthyoplankton surveys in the Gulf of Mexico Limnology and Oceanography Methods, 12: 86-101. Hjort, J. (1914) Fluctuations in the great fisheries of northern Europe viewed in the light of biological research. Rapports et Procès-Verbaux des Réunions du Conseil International pour l’ Exploration de la Mer, 20: 1-228.

32 Houde, E.D. (1987) Fish early life dynamics and recruitment variability. Pages17-29, in R.D. Hoyt, ed. American Fisheries Society Symposium, 2. Hurlburt, H.E. and Thompson, J.D. (1980) A numerical study of Loop Current intrusions and eddy shedding. Journal of Physical Oceanography, 10: 1611–1651. ICCAT (2012) Report of the 2012 Atlantic bluefin tuna stock assessment session. International Commission for the Conservation of Atlantic Tunas, SCI-033/2012, 124 pp. Ingram, G.W., Richards, W.J., Lamkin, J.T., Muhling, B.A. (2010) Annual indices of Atlantic bluefin tuna (Thunnus thynnus) larvae in the GOM developed using delta-lognormal and multivariate models. Aquatic Living Resources, 23: 35-47. Itoh, T., Shiina, Y., Tsuji, S., Endo, F., Tezuka, N. (2000). Otolith daily increment formation in laboratory reared larval and juvenile bluefin tuna Thunnus thynnus. Fisheries Science, 66(5): 834-839. Shulzitski, K. (2012) Population connectivity in a dynamic coastal system: effects of mesoscale eddy on distribution, growth, survival, and transport of larval reef fishes. PhD Dissertation, University of Miami, paper 835. http://scholarlyrepository.miami.edu/oa_dissertations/835 McArdle B.H., Anderson, M.J. (2001) Fitting multivariate models to community data: a comment on distance-based redundancy analysis. Ecology, 82: 290-297 Mather, F.J., Mason, J.M., Jones, A.C. (1995). Historical document: life history and fisheries of Atlantic bluefin tuna (No. 8986). US Dept. of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service. Maul, G.A. and Vukovich, F.M. (1993) The relationship between variations in the GOM Loop Current and straits of Florida volume and transport. Journal of Physical Oceanography, 23: 785-796. Medina, A., Abascal, F.J., Megina, C., Garcia, A. (2002) Stereological assessment of the reproductive status of female Atlantic northern bluefin tuna during migration to Mediterranean spawning grounds through the Strait of Gibraltar. Journal of Fish Biology, 60: 230-217. Mendoza V.M., Villanueva, E.E., Adem, J. (2005). On the annual cycle of the sea surface temperature and the mixed layer depth in the Gulf of Mexico. Atmósfera, 18: 127-148.

33 Millet, A. (2012). Southeast Area Monitoring and Assessment Program 2012 Spring Plankton Survey. Cruise Report GU-12-01(66), 50 pp. Miyashita, S., Sawada, Y., Okada, T., Murata, O., Kumai, H. (2001) Morphological development and growth of laboratory-reared larval and juvenile Thunnus thynnus (Pisces: Scombridae). Fishery Bulletin, 99: 601 - 616. Muhling, B.A., Lamkin, J.T., Roffer, M.A. (2010) Predicting the occurrence of bluefin tuna (Thunnus thynnus) larvae in the northern GOM: Building a classification model from archival data. Fisheries Oceanography, 19: 526-539. Muhling, B.A., Roffer, M.A., Lamkin, J.T., Ingram Jr, G.W., Upton, M.A., Gawlikowski, G., Richards, W.J. (2012). Overlap between Atlantic bluefin tuna spawning grounds and observed Deepwater Horizon surface oil in the northern Gulf of Mexico. Marine Pollution Bulletin, 64: 679-687. Muhling, B.A., Reglero, P., Ciannelli, L., Alvarez-Berastegui, D., Alemany, F., Lamkin, J.T., Roffer, M.A. (2013) Comparison between environmental characteristics of larval bluefin tuna Thunnus thynnus habitat in the GOM and western Mediterranean Sea. Marine Ecology Progress Series, 486: 257-276. Munk, P. (2007) Cross-frontal variation in growth rate and prey availability of larval North Sea cod, Gadus morhua. Marine Ecology Progress Series, 334: 225–235. Nababan, B., Muller-Karger, F.E., Hu, C., Biggs, D.C. (2011) Chlorophyll variability in the northeastern Gulf of Mexico. International Journal of Remote Sensing, DOI:10.1080/01431161.2010.542192. Lang, K.L., Grimes, C.B., Shaw, R.F. (1994) Variations in the age and growth of yellowfin tuna larvae, Thunnus albacares, collected about the Mississippi River plume. Environmental Biology of Fishes, 39: 259-270. Laiz-Carrión, R., Quintanilla, J.M., Torres, A.P., Alemany, F., García, A. (2013) Hydrographic patterns conditioning variable trophic pathways and early life dynamics of bullet tuna Auxis rochei larvae in the Balearic Sea. Marine Ecology Progress Series, 475: 203-212. Laiz-Carrión, R., Gerard, T., Uriarte, A., Malca, E., Quintanilla, J.M., Muhling, B., Alemany, F., Privoznik, S., Shiroza, A., Lamkin, J.T., García, A. (2014) Larval bluefin tuna (Thunnus thynnus) trophodynamics from the Balearic Sea (WM) and Gulf of Mexico spawning ecosystem by stable isotope. ICCAT SCRS-103/2014, 15 pp.

34 Lindo-Atichati, D.; Bringas, F.; Goni, G.; Muhling, B.; Muller-Karger, F.E.; Habtes, S. 2012.Varying mesoscale structures influence larval fish distribution in the northern Gulf of Mexico. Marine Ecology Progress Series, 463: 245-257. Llopiz, J.K., Muhling, B.A., Lamkin, J.T. (2014) Feeding dynamics of Atlantic Bluefin tuna (Thunnus thynnus) larvae in the Gulf of Mexico. ICCAT SCRS-173/2014, 6 pp. Luthy, S.A., Serafy, J.E., Cowen, R.K., Denit, K.L., Sponaugle, S. (2005). Age and growth of larval Atlantic sailfish, Istiophorus platypterus. Marine and Freshwater Research, 56:1027-1035. National Marine Fisheries Service (2014). Atlantic HMS management-based research needs and priorities. Highly Migratory Management Division, NMFS, NOAA, July 2, 2014 , accessed August 14, 2014. * National Marine Fisheries Service (2014a). National Marine Fisheries Service 2nd Quarter 2014 Update: Summary of Stock Status for FSSI Stocks. , accessed August 14, 2014. National Marine Fisheries Service (2014b). FishWatch Western Atlantic bluefin tuna. , accessed October 30, 2014. Pepin, P., Robert, D., Bouchard, C., Dower, J.F., Falardeau, M., Fortier, L., Jenkins, G.P., Leclerc, V., Levesque, K., Llopiz, J.K., Meekan, M.G., Murphy, H.M., Ringuette, M., Sirois, P., Sponaugle, S. (2014). Once upon a larva: revisiting the relationship between feeding success and growth in fish larvae. ICES Journal of Marine Science, doi: 10.1093/icesjms/fsu201 Pitchford, J.W., James, A., Brindley, J. (2005). Quantifying the effects of individual and environmental variability in fish recruitment. Fisheries Oceanography, 14(2): 156-160. Richards, W.J. (2006). Early Stages of Atlantic Fishes. CRS Press, Boca Raton, FL, 2640 pp. Rooker, J.R., Secor, D.H., De Metrio, G., Schloesser, R. (2008) Natal homing and connectivity in Atlantic bluefin tuna populations. Science, 322(5902): 742-744. Scott, G.P., Turner, S.C., Grimes, C.B., Richards, W.J., Brothers, E.B. (1993) Indices of larval bluefin tuna, Thunnus thynnus, abundance in the GOM: Modeling variability in growth, mortality, and gear selectivity: Ichthyoplankton methods for estimating fish biomass. Bulletin of Marine Science, 53: 912-929.

35 Schaefer, K.M. (2001) Reproductive biology of tunas. B.A. Block and E.D. Stevens, eds. In: Tuna. Physiology, Ecology, and Evolution. Academic Press, San Diego, pp. 225–270. Secor, D.H., Dean, J.M., Campana, S.E. (1995). Recent developments in fish otolith research. Univ. of South Carolina Press. Columbia, S.C. 730 pp. Sinclair, M. and Iles, D. (1988). Population richness of marine fish species. Aquatic Living Resources, 1(1): 71-83. Sponaugle, S., Denit, K.L., Luthy, S.A., Serafy, J.E., Cowen, R.K. (2005). Growth variation in larval Makaira nigricans. Journal of Fish Biology, 66: 822-835. Sponaugle, S. (2009). Daily otolith increments in the early stages of tropical fishes. Pages 93-132 in: B. Green, B. Mapstone, G. Carlos, and G. Begg, eds. Gathering information from otoliths of tropical fishes. Methods and Technologies in Fish Biology and Fisheries 11. Springer Science + Business Media B.V. Sponaugle, S. (2010). Otolith microstructure reveals ecological and oceanographic processes important to ecosystem-based management. Environmental Biology of Fishes, Doi:10.1007/s10641-010-9676-z. Tanabe T., Kayama S., Ogura M., 2003, An outline of the growth study on skipjack tuna (Katsuwonus pelamis) in the Western Pacific. Doc. IOTC WPTT-03-17. Tanaka, Y., Satoh, K., Iwahashi, M., Yamada, H. (2006). Growth-dependent recruitment of Pacific bluefin tuna Thunnus orientalis in the northwestern Pacific Ocean. Marine Ecology Progress Series, 319: 225-235. Tanaka, Y., Minami, H., Ishihi, Y., Kumon, K. Higuchi, K., Eba, T., Nishi, A., Nikaido, H. Shiozawa, S. (2014) Differential growth rates related to initiation of piscivory by hatchery-reared larval Pacific bluefin tuna Thunnus orientalis. Fishery Science, Doi: 10.1007/s12562-014-0798-7. Wexler, J.B., Chow, S., Wakabayashi, T., Nohara, K., Margulies, D. (2007) Temporal variation in growth of yellowfin tuna (Thunnus albacares) larvae in the Panama Bight 1990-1997. Fishery Bulletin, 105: 1-18. Zavala-Hidalgo, J., Morey, S.L., O’Brien, J.J. (2003) Seasonal circulation on the western shelf of the Gulf of Mexico. Journal of Geophysical Research, 108, Doi:10.1029/2003JC001879.

36 Zenteno, J.; Bustos, C. Landaeta, M. (2014) Larval growth, condition and fluctuating asymmetry in the otoliths of a mesopelagic fish in an area influenced by a large Patagonian glacier. Marine Biology Research, 10.5: 504-514.

a) 30

129

102 101 128

208 207

28

174

175

177

176

178

179

180

195 130 131 194

182

173

145 162

165 172

166

171

167

161 164

143 142 160

154

159

155

153

152

146

151

147

141

184 140

163

118

170

169

168

158

157

191 192

185

137

190 188 119 117

150

149

148

198 133 134

138

116

193

205 206 127 202

200

187

201

100 104

204

203 99

105

126

125

106

98

124

107

97

120

139

156

26

132

183

144

199

197

196

181

103

115

186 189 136

135 123

114

108

96

109

95

94

81

80

91

85

86

90

89

87

88

122

121 113

112

93

92

110 83 82

111

84

24 -95

-90

-85

-80

b) N↑

Figure 1. a) Stations sampled (+) during current study’s ichthyoplankton research survey (April , 2012). Locations of previous aged larval bluefin tuna in the Western North Atlantic (◊) are also shown. b) Larval bluefin tuna density (m3) calculated for the upper 10 m of the water column. Density of larval bluefin tuna across the Gulf of Mexico is shown as bubble plots with larger densities having larger bubble sizes.

37

38

Figure 2. Model output for week of May 13 to 19, 2012 showing the probability of larval bluefin tuna occurrence in Gulf of Mexico and northwestern Caribbean. Warmer colors indicate higher probability of encountering a larval bluefin tuna. Model output courtesy of B. Muhling.

39

30

Standard Length (mm) 2 to 4 4 to 7 7 to 12

28

26

24 -95

-90

-85

Figure 3. Gulf of Mexico sampling grid (+) in summer 2012 overlaid with bathymetry contours acquired from Geophysical Data system (GEODAS) and binned at 500m intervals for the Gulf of Mexico. Bubble plot shows all measured bluefin tuna larvae (n=467) across the GOM (Standard Length, mm) are shown. Aged larvae (n=100) are plotted (x) across the GOM.

-80

40

a)

b)

Figure 4. a) Larval bluefin tuna b) Larval bluefin tuna otolith at 1000x. Daily growth units (increments ●) originate at the primordium (P) and are enumerated along the otolith radius (OR) towards the otolith margin (OM). Scale indicates 0.020 mm.

41

a) 20 18 16

Frequency

14 12 10 8 6 4 2 0 3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 Age (days)

b)

Figure 5. Histogram showing age distribution a) for aged larval bluefin tuna from the Gulf of Mexico (n=100) and b) predicted ages for 38% (n=479) of collected bluefin tuna larvae using age and growth linear regression equation from this study.

42

10 9

Length (mm) , Standard or Notochord

8 7 6 5 4

y = 0.4486x + 0.5345 R² = 0.8749

3 2 1 0

3

4

5

6

7

8

9 10 11 Otolith growth units (days))

12

13

14

15

16

Figure 6. Closed circles (●) show otolith growth units (days) vs bluefin tuna length (SL or NL, mm). Least squares regression in the linear form is y=mx + b (n=100, m=0.45, b=0.53), r2=0.8749.

43

100 90

Expon. (2012 Bluefin larval ages, n=100) y = 5.0413e0.3296x R² = 0.9317

Otolith radius, μm (longest radius)

80 70 60 50 40 30 20 10 0

0

1

2

3

4

5

6

7

8

9

10

11

Length (mm SL or NL)

Figure 7. Scatter plot of T. thynnus length (SL or NL) mm vs otolith radius (μm). The line is the least squares regression exponential form y=a℮bx (N=100, a=5.04, b=0.33, r2=0.93).

Body Length (SL, mm)

44

Age (days)

Figure 8. Analysis of covariance for larval bluefin tuna growth curves. From the Western Atlantic: Brothers et al., 1983 (∆) collected from Florida Straits 1981 and current study (□) from Gulf of Mexico 2012. Growth curves from the Mediterranean Sea, growth curves for Garcia et al., 2013, T. Thynnus from 2003 (●), 2004 (+), and 2005 (x). Age and growth values from the Mediterranean provided by A. Garcia.

45 Temperature 29

Temperature, °C

27 25 Temperature 5m

23

Temperature 100m

21

Aged

19

Aged

17 15 75

80

85

90

95 100 105 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 205 210 215 Station #

Fluorescence 1.4

Fluoresence

1.2 1 0.8

Fluorescence 5m

0.6

Fluorescence 100m

0.4

Aged 5 Aged 100

0.2 0 75

80

85

90

95 100 105 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 205 210 215 Station #

Salinity 38 37

Salinity psu

36 35

Salinity 5m

34

Salinity 100m Aged 5

33

Aged 100

32 31 75

80

85

90

95 100 105 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 205 210 215 Station #

Oxygen 8 7.5 7

Oxygen mg

6.5 6

O2 (mg) 5m

5.5

O2 (mg) 100m

5

Aged 5

4.5

Aged 100

4 3.5 3 75

80

85

90

95 100 105 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 205 210 215 Station #

Figure 9. Scatter plot overview of environmental parameters measured from vertical water profiles (CTD casts) at 5 m (black) and 100 m (purple) for Temperature, Fluorescence, Salinity and Oxygen. Shaded symbols (●) indicate that at least one larval bluefin tuna were aged from any particular station.

46

a)

28

26

24

-95

-90

-85

-95

-90

-85

b)

28

26

24

Figure 10. Gridded contour map shows measured temperature from survey’s vertical casts at each station sampled (+) at a) 100 m and b) 5 m, °C. Aged larvae (x) are plotted. Scale is the same for both depths.

47

a)

28

26

24

-95

-90

-85

-95

-90

-85

b)

28

26

24

Figure 11. Gridded contour map shows measured oxygen values from survey’s vertical casts at each station sampled (+). Contours are plotted for depths at a) 100 m and b) 5 m, (mg/L) Aged larvae (x) are also plotted. Scale is the same for both depths.

48

a)

28

26

24

-95

-90

-95

-90

-85

b)

28

26

24

-85

Figure 12. Gridded contour map shows measured salinity values from survey’s vertical casts at each station sampled (+) at a) 100 m and b) 5 m. Aged larvae (x) are also plotted. Scale is the same for both depths.

49

a)

28

26

24

-95

-90

-85

-95

-90

-85

b)

28

26

24

Figure 13. Gridded contour map for Gulf of Mexico fluorescence (µg/L) measured at depths a) 100 m and b) 5 m. Values are taken from survey’s vertical casts at each station sampled (+). Aged larvae (x) are also plotted. Scale is the same for both depths.

50

Mean daily increment width (µm)

a) 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

y = 0.6508e0.1807x R² = 0.988

3

4

5

6

7

8

9

10 11 12 Age (days)

13

14

15

16

17

18

b) 12.00

Recent increment width (μm)

10.00

8.00 Small (2-3mm SL)

6.00

Intermediate (4-6mm) Largest (7-8mm)

4.00

2.00

0.00 0.00

2.00

4.00 6.00 Body Length (SL mm)

8.00

10.00

Figure 14. Scatter plot for recent increment growth (mean width of last 3 increments) and body length for small (♦), intermediate (■) and large (▲) larval bluefin tuna.

51

a) Resemblance: D1 Euclidean distance

dbRDA1 (100% of fitted, 17.7% of total variation)

1.0

0.5

0

-0.5

-1.0

-1.5 -4

-2

0 T100

2

4

120 110

Depth of Fluorescence max (m)

100 90 80 70 60 50 40 16

17

18

19

20

21

22

23

Temperature at 100 m, °C

Figure 15. a) Distance-based redundancy analysis (dbRDA) to visualize model output from distance linear based model indicating significantly different variables, Temperature at 100 m is shown with the most significant value (p = 0.009) b) Bubble plot shows temperature at 100 m and the depth of fluorescence max (m) with recent otolith growth (calculated from the mean of last 3 increments, µm).

Table 1. Summary of age and growth studies for larval bluefin tuna (Thunnus thynnus) and Thunnus sp. within the Western Atlantic (Miami and Gulf of Mexico) and the Mediterranean Sea. Current study is bolded for emphasis.

Location

Gulf of Mexico Florida Straits Mediterranean Sea

Year

Study

2012

Current

1981

Brothers et al., 1983

2003 2004 2005

Garcia et al., 2013

Length (SL, mm)

Otolith radius, µm

Growth rate, mm day-1

mean

min

max

mean

min

max

mean

min

max

5.71

2.13

8.10

31.11

10.9

98.90

0.51

0.32

0.82

3.81

8.91

28.82

19.0

88.0

0.69

0.52

1.38

2.83 3.71 3.0

8.7 7.83 8.34

25.75 28.40 25.6

9.9 13.2 11.5

80.9 52.1 54.5

0.81 0.61 0.64

0.43 0.44 0.22

1.48 0.78 1.16

5.42 6.03 5.29

52

53

Table 2. Summary (ranges and mean values) of the environmental indicators included in the Distance linear based model for recent growth of bluefin tuna larvae. Environmental indicators Longitude Latitude Date collected Temperature, 100 m Fluorescence, 100 m Salinity, 100 m Temperature, 5 m Fluorescence, 5 m Salinity, 5 m Oxygen, 5 m Fluorescence maximum Depth Fluorescence max (m) Tuna density (m-3)

mean 90.43 27.13 5/19/2012 20.13 0.64 36.47 26.44 0.23 36.01 6.37 1.28 77.56 0.08

minimum 84.01 24.50 5/1/2012 17.81 0.35 36.18 24.44 0.14 33.57 6.20 0.84 54.00 0.00

maximum 96.01 29.50 5/28/2012 21.66 1.07 36.59 27.79 0.67 36.63 6.58 2.12 108.00 0.32

Table 3. Summary table for marginal tests results of a distance based linear model from in situ measurements. Significant variables indicated in bold, p< 0.05. Variable Longitude Latitude Date caught Temperature, 100 m Fluorescence, 100 m Salinity, 100 m Temperature, 5 m Fluorescence, 5 m Salinity, 5 m Oxygen, 5 m Fluorescence maximum Depth Fluorescence max (m) Total tuna (Scombridae) density

SS (trace) 0.6346 0.6673 1.2329 8.4373 5.8342 2.4487 2.6196 1.4352 2.1263 2.3162 2.6627 7.6888 0.3675

Pseudo-F 0.5175 0.5443 1.0123 7.5571 5.0595 2.0392 2.1859 1.1811 1.7641 1.9258 2.223 6.8223 0.2987

P 0.477 0.474 0.324 0.009 0.028 0.145 0.158 0.275 0.210 0.173 0.138 0.012 0.572

Proportion 0.00724 0.00761 0.01407 0.09619 0.06652 0.02792 0.02987 0.01636 0.02424 0.02641 0.03036 0.08767 0.00419