Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Age, growth, and spawning of Yellowfin tuna, (Thunnus albacares) Bonnaterre 1788, in the Southern Philippines Yamanaka, Kae Lynne 1989

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1989_A6_7 Y35.pdf [ 11.15MB ]
Metadata
JSON: 831-1.0097612.json
JSON-LD: 831-1.0097612-ld.json
RDF/XML (Pretty): 831-1.0097612-rdf.xml
RDF/JSON: 831-1.0097612-rdf.json
Turtle: 831-1.0097612-turtle.txt
N-Triples: 831-1.0097612-rdf-ntriples.txt
Original Record: 831-1.0097612-source.json
Full Text
831-1.0097612-fulltext.txt
Citation
831-1.0097612.ris

Full Text

A G E , GROWTH, AND SPAWNING OF YELLOWFIN TUNA, (Thunnus albacares) Bonnaterre 1788, IN THE SOUTHERN PHILIPPINES  by Kae Lynne Yamanaka B.Sc, University of British Columbia, 1980.  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Resource Management Science)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA September 1989 ©Kae Lynne Yamanaka, 1989.  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives.  It is understood  that  copying or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of R e s o u r c e Management S c i e n c e The University of British Columbia Vancouver, Canada D a t e  DE-6 (2/88)  13 O c t o b e r 1989  Abstract  Yellowfin tuna {Thunnus albacares)  Bonnaterre 1788, were studied between February 1987  and March 1988 from the landings of two fisheries active in the Moro Gulf and Celebes Sea area of the southern Philippines. The primary objective of the study was to estimate the age and growth of juvenile yellowfin tuna by microstructural growth increments in the sagittal otoliths. Other objectives were to examine seasonal effects on the spawning, recruitment, and growth of juvenile yellowfin tuna, and for adult yellowfin, to determine if the relationship between weight and fork length differed between sexes and between years, to estimate age and growth by length frequency analysis and to examine sex ratios in the landings. Juvenile yellowfin landings from the General Santos ringnet were sampled for fork lengths and otoliths. Acid etched otoliths viewed under a scanning electron microscope showed incremental and discontinuous growth zones analogous to those observed in Central and  Eastern  Pacific yellowfin otoliths. The daily increment formation rate for yellowfin tuna between 25 and 40 centimetres in fork length was validated by an oxytetracycline marking experiment carried out at the Kewalo Basic laboratory of the NMFS, SWFC/NOAA. Two hundred and seven otoliths were prepared for ageing by acid (EDTA) etching and whole mounting or by embedding in fibreglass resin, sectioning, polishing, acid etching, and mounting. Sectioning was done in a frontal plane containing both the primordium and post-rostral tip. Increments were counted along a straight line path between the primordium and the post-rostral tip using a phase contrast light microscope at 400X magnification. Growth in fork length and weight was best described by two linear stanzas.  The growth  rate of yellowfin between 15 and 35 cm in fork length was 2.50 mm/day and between 35 and 79 cm was 0.96 mm/day. The growth rate of yellowfin between 50 and 2350 g in whole wet weight was 9.15 g/day and between 2350 and 9200 g was 23.68 g/day. Philippine yellowfin attain a fork length of 57.46 cm and a weight of 3726 g in a year. Transition points between growth stanzas may indicate ontogenetic changes in the early life history of yellowfin tuna.  iii Estimates of age and growth were verified by length frequency analysis and back calculations done using daily increment measurements along the counting path of sectioned otoliths. Results showed that ageing by increments in otoliths is a precise and accurate technique. A reversed Lee's Phenomenon was apparent in back calculated lengths at previous ages, indicating that size selective mortality is adversely effecting the smaller fish of an age group. Consistent increment patterns were apparent in otoliths. Eight to ten evenly spaced and narrow increments surround the primordium, the next thirty increments increase in width to a maximum then slowly decrease in width over the next 120 and increments beyond 120 remain at a steady increment width. Monsoon seasons occur at five and seven month periodicity in the southern Philippines. The southwest monsoon begins in May and continues until September. The northeast monsoon begins in October and continues until April. Transitional tradewind seasons occur during April May and September - October. Yellowfin spawning peaks estimated by trends in a condition factor, gonadosomatic, and hepatosomatic indices, were apparent from April through June and October through November. Relative recruitment frequencies estimated from back dating age converted length frequencies showed recruitment peaks in May through June and October through November. Condition indices and growth estimated by otolith ageing, for the juvenile yellowfin, reach a minimum in April. Recruitment and growth for the juvenile yellowfin appears to be enhanced with the onset of the monsoonal winds. Using four years of weight and length data for adult yellowfin significant differences in weight-length relationships between sexes were found in 1982, 1983, 1984, and 1987. Significant differences were also found between years for females and between years for males. From the length frequency analyses, females were estimated to be 127 cm and males 136 cm at age three. Growth rates were estimated to be 2.90 cm/mo and 3.27 cm/mo for females and males respectively. The growth rate decreases at 0.07 cm/mo and 0.10 cm/mo for females and males every month after age three. Sex ratios showed a preponderance of males over all fork length intervals above 135 cm, between 30 to 44 months of age, and in every month of the year except November.  iv Table of Contents  Abstract  ii  Table of Contents  iv  List of Tables  ix  List of Figures  x  Acknowledgements  xvi  General Introduction  Chapter I.  1  Age and growth of juvenile yellowfin  General Introduction  19  Section A. Estimates of age and growth as determined by increments in sagittal otoliths.  Introduction  •  20  Methods and Materials  1. The collection of samples  21  1.1 Sampling for juvenile yellowfin  21  1.1.1 Conversion from thawed to fresh measurements  24  1.1.2 Relationship between weight and fork length  24  1.1.3 Change in body shape  24  1.2 Removal of sagittal otoliths  25  2. Scanning electron microscope examination of otolith microstructure .... 25 3. Oxytetracycline validation experiment  28  V  4. Ageing 4.1 Sample sizes: a priori estimate 4.2 Preparation of otoliths  30 '. . . 32  4.3 Counting increments  33  4.4 Sample sizes: posteriori adjustment  33  5. Growth 5.1 Growth in fork length and weight  34  5.2 Comparison of growth  34  Results  1. Otolith samples collected  35  1.1 Conversion between thawed and fresh measurements  35  1.2 Relationship between weight and fork length  35  1.3 Change in body form  38  2. Sagittal otolith microstructure 3. The rate of increment formation  38 42  4. Ageing  46  5. Growth  50  Discussion  1. Ageing by sagittal otolith increments  56  2. Growth  60  Section B.  Length frequency analysis and back calculations  Introduction  64  Methods and Materials  1. Length frequency analysis  64  vi 1.1 Data collection from the ringnet fishery  65  1.2 Data analyses  65  2. Back calculated fork lengths at previous ages  65  2.1 Relationship between otolith size and fish size  66  2.2 Measurements of increment widths  66  2.3 Back calculations  66  3. Growth patterns  67  Results  1. Length frequency analysis  67  2. Back calculations 2.1 Relationship between otolith size and fish size  70  2.2 Measurements of increment widths  73  2.3 Back calculations  73  3. Pattern of growth .  73  Discussion  1. Length frequency analysis  79  2. Back calculation  81  3. Growth comparisons  82  4. Growth patterns  84  Chapter II.  Monsoons, spawning, recruitment, and growth.  Introduction  •  86  Methods and Materials  1. Monsoon seasons  87  1.1 Sea surface temperatures  88  1.2 Prevailing wind and current  88  vii 2. Spawning  89  2.1 Data collection from the handline fishery  90  2.2 Condition index of adult yellowfin  90  2.3 Gonadosomatic and hepatosomatic indices  91  3. Recruitment patterns  91  4. Growth of juvenile yellowfin  93  4.1 Condition index of juvenile yellowfin  93  4.2 Growth of juvenile yellowfin over time  93  Results  1. Monsoon seasons  94  2. Spawning  96  3. Recruitment  96  4. Growth  100  Discussion  100  Chapter III.  Estimates of age and growth for adult yellowfin.  Introduction  105  Methods and Materials  1. Data collection from the handline fishery 1.1 Relationship between weight and fork length 2. Age and growth of adult yellowfin  106 106 107  3. Sex ratios 3.1 Sex ratio by fork length  108  3.3 month 3.2 Sex ratio by age  109 108  Results and Discussion  viii 1. Relationship between weight and fork length  109  2. Age and growth of adult yellowfin  114  3. Sex ratio 3.1 Sex ratio by fork length  125  3.2 Sex ratio by age  125  3.3 Sex ratio by month  125  General Summary  131  Literature Cited  135  Appendix 1  1 4 3  ix  List of Tables  Table 1.1  Differences between whole mounted and sectioned otolith increment counts done on left and right otoliths of a pair.  Table 1.2  Differences between first and second increment counts on 20 randomly selected otolith sections.  Table 1.3  Estimates of growth for yellowfin aged by daily otolith increments.  Table 1.4  Back calculated fork lengths by monthly intervals.  Table 1.5  Comparison of growth rates for juvenile yellowfin.  Table 2.1  Condition indices for adult yellowfin.  Table 2.2a  Hepatosomatic indices for adult yellowfin.  Table 2.2b  Gonadosomatic indices for adult yellowfin.  Table 2.3a  Condition factors for juvenile yellowfin between 26 and 28.5 cm.  Table 2.3b  Growth rates for juvenile yellowfin between 19 and 21 cm.  Table 3.1  Growth rates for adult yellowfin.  X  List of Figures  Figure 1a  Annual total landings (MT) of tunas in the Philippines from 19761986.  Figure 1 b  Annual total landings (MT) of yellowfin by sector and region in 1984.  Figure 2  Map of General Santos City.  Figure 3  Map of the Philippines.  Figure 4a  Photograph of a payao float  Figure 4b  Diagram of a payao.  Figure 5  Diagram of a ringnet.  Figure 6a  Photograph of the "Timbogan", a 30 M ringnet fishing vessel.  Figure 6b  Photograph of a light vessel (note the lamps on the top of the cabin).  Figure 7  Photograph of a carrier vessel approaching a payao float.  Figure 8  Photograph of a ringnet vessel brailling fish into a carrier vessel.  Figure 9a  Photograph of a municipal handline motorized banca.  Figure 9b  Photograph of a deep sea tuna handline.  Figure 10  Photograph of yellowfin tuna at Lion Beach.  Figure 1.1a  Photograph of juvenile yellowfin tunas (15 cm ruler).  Figure 1.1b  Photograph of juvenile yellowfin tunas showing vertical banding.  Figure 1.2  Photograph of juvenile yellowfin stomach and liver.  Figure 1.3a  Photograph of juvenile yellowfin head with brain exposed.  Figure 1.3b  Photograph of otolith sampling set up.  Figure 1.4  Electronmicrophotograph of a yellowfin otolith.  Figure 1.5  Diagram of a right sagittal otolith.  Figure 1.6a  Relationship of fresh against thawed fork lengths.  Figure 1.6b  Relationship of fresh against thawed weights.  Figure 1.7  Relationship between weight and fork length.  XI  Figure 1.8  Comparison of the weight - fork length relationship between the Philippines and Hawaii.  Figure 1.9  Condition factor at fork length for juvenile yellowfin.  Figure 1.10a  Electronmicrophotograph of discontinuous and incremental growth zones in etched otolith.  Figure 1.10b  Electronmicrophotograph of otolith growth zones.  Figure 1.11a  Electronmicrophotograph of narrow otolith growth zones along the edge of the otolith.  Figure 1.11b  Electronmicrophotograph of wide otolith growth zones near the primordium.  Figure 1.12a  Otolith from a tetracycline injected yellowfin viewed under white light.  Figure 1.12b  Otolith viewed under phorescent light, showing the OTC mark.  Figure 1.13  Regression of increments against days for the OTC marking experiment.  Figure 1.14  Plot of increment counts by fork length interval showing means, and + ,- one standard deviation.  Figure 1.15a  Fork length at age data and fitted von Bertalanffy growth function.  Figure 1.15b  Scatter plot of fork length residuals.  Figure 1.16a  Fork length at age data and fitted linear growth stanzas.  Figure 1.16b  Scatter plot of fork length residuals.  Figure 1.17a  Weight at age data and fitted Gompertz growth function.  Figure 1.17b  Scatter plot of weight residuals.  Figure 1.18a  Weight at age data and linear growth stanzas.  Figure 1.18b  Scatter plot of weight residuals.  Figure 1.19a  Comparison of fork lengths at age for the Western Pacific and the Eastern Pacific.  Figure 1.19b  Comparison of linear growth equations for the Western, Eastern, and Central Pacific.  Figure 1.20  Fork length frequency distributions for yellowfin caught in the ringnet fishery from February to August 1987.  Figure 1.21  Fork length frequency distributions for yellowfin caught in the ringnet fishery from September 1987 to March 1988.  Figure .22  Modal progression of prominent length frequencies.  xii Figure 1.23  Relationship between fork length and counting path length.  Figure 1.24a  Photograph of a sectioned otolith showing the increments surrounding the primordium, 400X magnification.  Figure 1.24b  Photograph of a sectioned otolith showing the narrow increments near to the post-rostral tip, 400X magnification.  Figure 1.25a  Photograph of a sectioned otolith showing the increment patterns, 400 X magnification.  Figure 1.25b  Photograph of a sectioned otolith showing the increment patterns, 1000X magnification.  Figure 1.26a  Comparison of back calculated fork lengths at age and the linear growth stanzas derived from otolith ageing.  Figure 1.26b  Comparison of the mean back calculated fork lengths at age and the linear growth stanzas derived from otolith ageing.  Figure 1.27  Patterns of increment width measurements.  Figure 2.1  Sea surface temperatures, prevailing wind and current by month for the Celebes Sea plotted along with relative recruitment frequency.  Figure 2.2  Adult yellowfin condition, hepatosomatic, and gonadosomatic indices by sex plotted over time along with the relative recruitment frequency.  Figure 2.3  Prevailing wind, juvenile yellowfin growth rate, condition factors, and relative recruitment frequency.  Figure 3.1a  Relationship between log weight and log fork length for female adult yellowfin.  Figure 3.1b  Relationship between log weight and log fork length for male adult yellowfin.  Figure 3.2  Relationships between log weight and log fork length for males and females for 1982, 1983, 1984, and 1987.  Figure 3.3  Comparison between the relationship between weight and fork length for yellowfin from the Philippines, Eastern, and Central Pacific.  Figure 3.4  Fork length frequencies for male yellowfin caught in the handline fishery between February and July 1987.  Figure 3.5  Fork length frequencies for male yellowfin caught in the handline fishery between August 1987 and January 1988.  Figure 3.6  Fork length frequencies for female yellowfin caught in the handline fishery between February and July 1987.  Figure 3.7  Fork length frequencies for female yellowfin caught in the handline fishery between August 1987 and January 1988.  Figure 3.8a  Combined fork length frequencies from the ringnet and handline fisheries for the months of May and October 1987, for females.  Figure 3.8b  Combined fork length frequencies from the ringnet and handline fisheries for the months of May and October 1987, for males.  Figure 3.9  Growth curves for yellowfin derived from fork length frequencies.  Figure 3.10  Comparison of growth curves for yellowfin from the Philippines and the Eastern Pacific.  Figure 3.11  Comparison of growth curves for yellowfin from the Philippines, Western, and Central Pacific.  Figure 3.12a  Percent males by fork length for adult yellowfin from the handline fishery.  Figure 3.12b  Fork length frequencies for females from the handline fishery.  Figure 3.12c  Fork length frequencies for males from the handline fishery.  Figure 3.13a  Percent males by age for adult yellowfin from the handline fishery.  Figure 3.13b  Age frequencies for females from the handline fishery.  Figure 3.13c  Age frequencies for males from the handline fishery.  Figure 3.14  Percent males by month from the handline fishery.  Appendix i  Yellowfin measurements, mean increment counts, and standard deviations.  xiv Acknowledgements  Thanks are extended to Mr. Gordon A. McFarlane of the Pacific Biological Station for arranging space and the use of equipment in the ageing lab and Ms. Shayne MacLellan for her assistance in photographing phosphorescent and sectioned otoliths, Director Richard S. Shomura, Dr. Christofer H. Boggs, and the technicians at the Kewalo Basin lab facility of the Southwest Fisheries Center, National Marine Fisheries Service for their persistence and hard work in carrying out the oxytetracycline marking experiment, Mr. Robert Nishimoto, Dr. Alex Wild, Mr. Terry J. Foreman, and Mr. James H. Uchiyama for sharing their knowledge, and instructing me on their otolith techniques, Dr. Milton Ordillas for his assistance with the scanning electron microscope, Brother Robert McGovern and the Notre Dame of Dadiangus College in General Santos City for providing a lab space for me, San Andres, RD, Dole Philippines, Delapina, and Amadeo Fishing Companies for permission to sample their catches and for donating samples for this study, Mr. Madin Tan for allowing me aboard a carrier vessel, Ms. Divina Labao for her unfailing assistance in collecting the length frequencies, Brother Robert McGovern, the late Mr. D. Wong and the employees of Dole Philippines for their interest in the project and concern for my safety, Ms. Edwina Plecis and the Plecis family for their warmth and hospitality throughout my stay in General Santos City, Any. Reuben Ganaden and Mr. Noel Barut for their assistance in all aspects of the field work, their permission to use three years of length frequency data collected on their Tuna Sampling Program, and their hospitality during my stay in the Philippines. Special thanks are extended to Dr. Norman J. Wilimovsky and Mr. Mitsuo Yesaki for their assistance in obtaining funding, their encouragement and guidance, and most of all their patience over the duration of this project. Special thanks are also extended to my parents for their support, understanding, and durability over the years. While doing the field work for this project I am certain there were times when the news from the Philippines must have caused them much concern. A great many friends have contributed to the completion of this project and to the many who have not been acknowledged here, I thank you for your encouragements, kindness, and companionship. Funding was provided by a Canadian International Development Agency scholarship and assistance in sampling was sponsored by the Indo-Pacific Tuna Programme/UNDP of the FAO.  1 General Introduction  Research Problem  Increased foreign interest in the Philippine tuna fisheries has resulted in the expansion of both the Philippine commercial and municipal fishing sectors. Concerns of over-exploitation of the tuna resources have been raised (Aprieto, 1980; White, 1982) but little information is available on the tuna fishery or biology and population dynamics of the local stocks. A data gathering system was established in 1979 by the Bureau of Fisheries and Aquatic Resources (BFAR) with the assistance of the FAO/UNDP South China Sea Fisheries Development and Coordinating Programme (SCSP) (White, 1982). Four locations in Mindanao in the southern Philippines were chosen to obtain basic information on catch, effort, species, and size composition of landings in the tuna fishery. The data gathering system is continuing however some basic questions have yet to be addressed and are central to investigations on the biology and population dynamics of the tuna stock in the southern Philippines. In an assessment of the data gathering system White (1982) points out the uncertainty of stock and yield assessments due to the assumptions made on the age and growth of these small tunas. One of the recommendations from the initial study was to "better document the growth and spawning periodicity of the Philippine tunas". With better estimates of juvenile growth of yellowfin tuna, mortalities could be estimated thus enabling yield estimates to be made with more certainty and the question of over-exploitation examined. The focus of this project is to investigate the early life history of the yellowfin tuna caught in the ringnet fishery of the southern Philippines. The primary topic of research is to estimate age and growth for these young yellowfin tuna from the examination of sagittal otolith microincrements. The unusual fishery for juvenile yellowfin tuna allows a unique opportunity to sample very young fish which have not yet been obtained for ageing studies. Early life history information is extracted from the examination of increment patterns and back calculations done on otoliths.  2 With the information from otolith ageing and sampling of ringnet and handline landings, seasonal variation in spawning, recruitment, and growth are examined. A ringnet fishery targeting on the juvenile yellowfin tuna and a handline fishery targeting on the adult yellowfin tuna exist side by side in the Moro Gulf and Celebes Sea. The presence of both these fisheries provides the opportunity to sample and compare estimates of spawning derived from the adult yellowfin tuna to estimates of recruiment derived from the juvenile yellowfin tuna. Seasonal monsoonal events are characterized by environmental factors and trends are derived. Patterns in yellowfin tuna spawning, recruitment, and growth are then compared with seasonal trends. Finally, general aspects of population biology are investigated for the adult yellowfin tuna landed from the handline fishery. Estimates of age and growth of adult yellowfin tuna is derived by combining length frequency data from both the ringnet and handline fisheries.  Specific Objectives  1.  Estimate age and growth of juvenile yellowfin tuna by microstructural growth increments on sagittal otoliths.  2.  Examine seasonal effects on the spawning, recruitment, and growth of yellowfin tuna.  3.  Estimate age and growth of adult yellowfin tuna by length frequency analysis.  This thesis is divided into three chapters. Chapter I has two sections. The first section deals with estimating age and growth of juvenile yellowfin tuna by the examination of sagittal otolith microincrements and the second section verifies those age and growth estimates by comparisons with estimates derived from length frequency analysis and back calculations. In Chapter II seasonal influences on spawning, recruitment, and early growth of yellowfin tuna are investigated. Seasons are defined in terms of sea surface temperature, prevailing wind, and prevailing current trends collected from the literature.  Spawning peaks for yellowfin tuna are determined by the examination  of condition, hepatosomatic, and gonadosomatic indices over a period of a year. Recruitment frequencies were derived by back dating length frequencies. Condition factors and growth rates of  3 juvenile yellowfin tuna were determined from sampling and otolith ageing. Chapter III deals with biological aspects of the adult population. Weight and length relationships were determined by sex for 1982, 1983, 1984, (data collected by BFAR) and 1987. Age and growth of adults were estimated using Petersen and modal progression methods. Sex ratios were determined by fork length, age, and monthly intervals.  The Philippine tuna fishery  The Philippine fishery for tuna and tuna-like species has expanded in the last fifteen years to one of the country's largest and most valuable marine fisheries. Total production has increased from 9,000 MT in 1971 to 216,000 MT in 1977, and reached 266,000 MT in 1986 (Bureau of Fisheries and Aquatic Resources (BFAR), 1987)(Figure 1a). Approximately twenty percent of the overall tuna production is made up of yellowfin tuna but in the southern areas surrounding the Moro Gulf and Celebes Sea, yellowfin tuna account for almost ninety percent of the tuna production (Figure 1b). In 1986, the South Cotobato BFAR region XI, and the Zamboanga region IXB were the highest producers of yellowfin tuna in the Philippines. Zamboanga City in region IXB is home port to a large multinational commercial tuna purse seine fleet which land large quantities of tuna to supply a local cannery market. General Santos City in region XI is home to a large municipal tuna handline fleet which targets on large tunas for the Japanese sashimi market. In region XI, both municipal and commercial fishing sectors operate a variety of gears in combination with artificial fish aggregating devices, to supply both domestic and export markets. The municipal handline fleet is centred in the Lion Beach and Bula areas of General Santos City and the commercial ringnet fleet is centred in Calumpang, a nearby village (Figure 2). Unlike major tuna fisheries of the world, the Philippine commercial ringnet fishery is based on a catch consisting almost exclusively of small sized juvenile fish. The Moro Gulf and Celebes Sea is known to be a major tuna spawning ground as larval scombroids are ubiquitous in these waters (Wade, 1950) and young tunas are fished year round from this region. During 1986, the General Santos ringnet  4  Figure 1 a  Annual total landings (MT) of tunas in the Philippines from 19761986.  300  76  Figure 1b  77  78  79  '80  '81  '82  '83  '84  '85  '86  Annual total landings (MT) of yellowfin by sector and region in 1984.  15 municipal <u rO  commercial  2 C  o  <A 3 Q  5 +  X 0  X X  l=xl  III N R C  X X IV  X X V  VI  X X  VII Vlll  Region  X X  X X X  AM  IXA IXB X  X X X X X X X X X X X X X XI  X X XII  Figure 2  Map of General Santos City.  6 fishery landed an estimated 4,700 MT of yellowfin tuna of which over ninety percent were less than 70 centimetres in fork length. In the same year, 5,700 MT of large adult yellowfin tuna was landed from the municipal handline fishery (BFAR, 1987). Approximately sixty percent of the commercial ringnet catch is canned locally and exported almost exclusively to the United States. As much as ninety-five percent of the municipal handline catch is exported fresh to Japan for the sashimi market and about four percent is frozen and exported to Japan and Italy for the canning market. Frozen fish exports account for 20 percent of the total local exports from the port of General Santos (BRC, 1989). Clearly the tuna fishery is a major industry in General Santos City creating employment and earning foreign capital.  An  estimated 20,000 fishermen, 1,200 cannery workers, 3,000 shore workers, and countless buyers and sellers are employed directly from the fishing industry in General Santos where the population in 1986 was 180,000. Exports of canned, frozen and chilled tuna in the first quarter of 1987 were estimated at 1,760 metric tonnes for a value of 2.37 million U.S. dollars (BRC, 1988).  Background  A brief overview of the two major tuna fisheries operating from General Santos City and Calumpang is given here to acquaint the reader with the general type of fishing gears referred to throughout the text. The use of the terms "commercial" and "municipal" were arbitrarily set by the Bureau of Fisheries and Aquatic Resources.  Commercial refers to vessels over three gross tonnes  and municipal refers to vessels less than three gross tonnes. Commercial vessels are usually company owned with crews being employees of the fishing companies. Municipal vessels are usually owned and operated by an individual or family. The fishing area covered by the fishermen from General Santos and Calumpang extends from Sarangani Bay to the Moro Gulf in the north, and the entire Celebes sea along the Sulu Archipelago to the west, Indonesian waters to the south, and the islands bordering the Pacific on the east (Figure 3). It is not uncommon for fishing vessels to travel 240 nautical miles to reach the  8 grounds. The tuna fishery in this region is based solely on fish aggregating devices or payaoes. The number of deep sea payaoes on the fishing grounds is not known but an estimated 3,000 payaoes are deployed from commercial operators in Calumpang. Payaoes are fabricated and deployed by every ringnet fishing vessel operator. Payaoes consist of an anchor made from nine, forty gallon drums filled with cement, ten coils (660 M) of 2.5 cm nylon rope, wire cable (6 meters), a swivel, and a steel pontoon type float (1 M diameter, 3 M long)(Figure 4b). The floats are fitted with a marker and painted with company colours and identification numbers (Figure 4a). Coco palm branches are strung together and hung underneath the floats. Fishermen maintain that about four nautical miles seems to be the maximum radius within which a single payao has influence over fish abundance. Therefore payaoes are deployed no closer than eight nautical miles from each other. The location of each payao is pinpointed at the time of deployment using a satellite navigator.  The Ringnet Fishery  Deep sea ringnet vessels operate in association with accessory vessels and payaoes in the Moro Gulf and Celebes Sea area (120° - 124° E, 2° - 7° N). Ringnet vessels are called 'mother, "catcher', or "fishing" vessels as they carry the ringnet (Figure 5), the manpower (35 - 50  men)  required to operate the ringnet, and do the actual catching of the fish. Each ringnet vessel is accompanied by one or two "light" vessels and a "guard" or "auxiliary" vessel, which assist in the fishing operation, and two "service" or "carrier" vessels which replenish supplies and transport the catch back to company landing sites in Calumpang. The fishing vessels remain on the fishing grounds year round and fish the same pattern of 20 to 40 payaoes every month. Vessels return to Calumpang for dry docking and a general repair every 3 to 4 months. Between dry dockings, crew members rotate for 3 to 4 day holidays about once a month, travelling to and from Calumpang on the carrier vessels.  9 Figure 4 a  Photograph of a payao float  10  Figure 5  Diagram of a ringnet.  700 M  11 There are approximately 73 licensed commercial ringnet vessels in General Santos City, with ten private companies owning 86 percent of these.  Ringnet vessels range in size from 4 to 122  gross tonnes and average about 40 gross tonnes (Figure 6a). Light and auxiliary vessels on average are less than 3 gross tonnes and therefore are not required to register and obtain a commercial licence (Figure 6b). There are approximately 105 licensed commercial carrier vessels, 85 percent of these are owned by the same ten private companies. Carrier boats range from 4 to 66 gross tonnes and average around 24 gross tonnes (Figure 7).  Fishing operations  Fishing begins in the early morning (0400 hrs) with light vessels shining bright mercury lights onto the surface of the water around a payao. A crew member is sent into the water and free dives under the payao to assess the size and composition of the fish school and monitor the depth of the school. The fish are attracted to the surface by the light. Once the school is near the surface, the light vessel slowly moves away from the payao and in doing so directs the fish into position so that the fishing vessel can set the ringnet. The fishing vessel attaches a sea anchor to the end of the ringnet and sets the end of the net with a small lifeboat or auxiliary vessel. The ringnet is then payed off the stern of the fishing boat as it moves in a semi-circle around the fish school. When the net is set, divers are again sent into the water to monitor the fish school as the sea anchor end of the net is winched in from the fishing boat or dragged in by the auxiliary vessel. When the net is pursed, the light vessel is guided over the net floats, then the ringnet is hauled in by hand over the starboard side of the fishing vessel. The auxiliary and light vessels then manoeuvre the fishing vessel and the net so that the net does not collapse while it is being-hauled in. Once the net is hauled in and all that is left in the water is a small section of net containing the fish, the carrier vessel comes along the starboard side of the fishing vessel and the ringnet floats are pulled onto the carrier vessel's port side and the fish brailled into the carrier vessels holds  12  Figure 6a  Photograph of the "Timbogan", a 30 M ringnet fishing vessel.  Figure 7  Photograph of a carrier vessel approaching a payao float.  14 (Figure 8). The fishing vessel and the auxiliary vessels then head off in the same general direction and locate the next payao in the series. Once a payao is located, a diver is sent down to check if there is any fish there. If there is no fish then the search for the next payao begins. If fish are located then the sea anchors are set and the crews wait for the early morning to fish. The carrier vessel remains with the fishing vessel until its holds are filled with fish or the ice melts, then it returns to Calumpang while another carrier leaves Calumpang to go out to the fishing vessel.  The Catch  Total fish landings from the commercial ringnet fishery operating from Calumpang is estimated at 20,000 MT annually. The landings are composed of 3 to 30 species of pelagic fish of which tunas and tuna like species comprise fifty to one hundred percent (White, 1982). Approximately forty-five percent of the "tunas" consists of skipjack Katsuwonus pelamis, twenty-five percent yellowfin tuna Thunnus albacares,  twenty-three percent frigate and bullet tunas Auxis spp.,  and two percent each of bigeye Thunnus obesus and Eastern little tuna Euthynuus affinis.  Sixty  percent of the fish landed by the ringnet fishery supplies the two local canneries, thirty-five percent is exported out of General Santos for domestic fresh markets, and the remaining five percent supplies the local fresh fish market.  The Municipal Handline Fishery  Deep sea handline fishermen operate small bancas which are wooden outriggered canoe type vessels with inboard diesel engines, typically under 3 gross tonnes (Figure 9a). These handline fishermen target on the large tunas and billfish using weighted single hook handlines nearby payaoes in the Moro Gulf and Celebes Sea. These municipal fishermen often operate near payaoes under informal agreements with the owners of the payaoes. The municipal fishermen usually act as guardians while fishing the payao.  Figure 8  Photograph of a ringnet vessel brailling fish into a carrier vessel.  16 Figure 9a  Figure 9b  Photograph of a municipal handline motorized banca.  Photograph of a deep sea tuna handline.  17 There are an estimated 5,000 municipal bancas operating from General Santos City (Sevilla, 1987). Each boat has about six crew and can carry ice blocks in styrofoam insulated holds. Trip length is usually governed by how long the ice takes to melt and on average is about seven days. Handlines consist of various weight and hook types but the most common are a two kilogram lead weight, a #9 Chicago or tuna circle hook. The hook and weight are attached to 200 pound test line which is wound onto a wooden spool (Figure 9b).  Small tunas or scads  (Decapterus spp.)  are  jigged with light handlines and used as bait. A single hook is hidden in the bait and lowered to a depth of about 100 fathoms (at the thermocline) and then tied to the boat and left to fish. Hooked fish are played out, hauled in by hand and iced whole.  The Catch  Catches are ninety-five percent yellowfin tuna, three percent bigeye tuna, and two percent billfish  {Istiophoridae).  The prime market for all these fish is the Japanese sashimi auction. Prices  to the fishermen for sashimi grade tuna were about four U.S. dollars a kilogram during the first quarter of 1987. Lower quality tuna are sold to Dole Philippines Seafoods and are frozen for export canning markets. At Lion Beach fish are unloaded by the fishermen then weighed, labelled and the flesh graded by the fish buyers (exporters) (Figure 10). The quality of the flesh is assessed depending on whether the flesh is firm, pink, unbruised, and unburnt (Watson, 1988). If the quality is high enough, a price is set and the fish bought for the Japanese sashimi market. Lower grade fish are sold to the local fresh fish market at a lower price and the least desirable fish are sold for the canning market. Sashimi fish are dressed, fins and gills removed, packed in ice, trucked to Davao (4 hours trip), repacked in plastic bags and cardboard boxes, air freighted to Manila, then air freighted to Tokyo. The trip from Lion Beach to the Tokyo fish auction takes three to four days.  18  Figure 10  Photograph of yellowfin tuna at Lion Beach.  19  CHAPTER I AGE AND GROWTH OF JUVENILE YELLOWFIN TUNA  General Introduction  Age and growth assessments of fish is an integral part of stock assessment which forms the basis for most fisheries management recommendations. The ageing of fish has a long history dating back to the 1600's but the first verified use of scale annuli for carp was recorded in 1898 (Carlander, 1987). Since then annual marks, resulting from seasonal growth differences, on scales, fin rays, spines, opercular bones, vertebra, and otoliths have all been used to investigate age and growth of fish (Bagenal, 1974; Summerfelt and Hall, 1987). The age and growth of wild fish have also been estimated by length frequency analyses, back calculating size at a previous age, and tagrecapture studies. However some of these techniques, which are widely used on temperate fish, are of little utility in ageing fish less than a year of age, or in ageing tropical species where seasonal growth patterns do not result in clearly annual marks and where spawning is continuous throughout the year. The discovery of daily rings on otoliths by Pannella (1971) has been considered the most significant advancement in age determination in recent times. Ageing by means of daily structures allows for a level of accuracy and precision that previously was not possible using annual structures and provides a technique for ageing fish which could not be accurately aged by methods primarily applied in temperate fisheries. The greatest potential use of ageing by daily structures is in the investigation of early life history stages (sub-yearling) of all fish and the ageing of tropical fish. Direct ageing from daily structures on sagittal otoliths of juvenile yellowfin tuna from Philippine waters has not previously been attempted. However, this ageing technique has been used to age juvenile yellowfin tuna in the Central Pacific (Uchiyama and Struhsaker, 1981), adult yellowfin tuna in the Eastern Pacific (Wild, 1986), larval skipjack tuna (Radtke, 1983), larval black skipjack tuna (IATTC, 1989), juvenile bluefin tuna (Brothers et al, 1983), adult albacore tuna (Laurs  20 et al., 1985), and many other species of fish from very different environments and geographic regions (Gjosaeter et al., 1984). Daily increments are becoming generally accepted as accurate and precise ageing structures as the methods have been validated in a growing number of cases (Campana and Neilson, 1985).  This chapter investigates age and growth of juvenile yellowfin tuna, captured from payaos anchored in the Celebes Sea. The chapter is divided into two sections; section A focuses on estimating age and growth by increments on otoliths and section B compares these estimates with those derived from length frequency analysis and back calculations.  Section A. Estimates of age and growth as determined by increments in sagittal otoliths  Introduction  For age determination, accurate and precise interpretations of ageing structures must be made (Taubert and Coble, 1977; Brothers, 1978; Wild, 1986), the rate of formation of these structures validated for all age classes in the population (Beamish and McFarlane, 1983), and the time of first increment formation determined (Radtke, 1983). In this section, the morphology of sagittal otolith growth zones are identified by their microstructure under a scanning electron microscope and ageing structures or increments are clearly recognized so that an accurate interpretation of these structures can be made using light microscopy; the increment formation rate is validated as daily in juvenile Central Pacific yellowfin tuna; and 207 juvenile yellowfin tuna samples from the Philippines are aged using daily increments.  21  Methods and Materials  1.  1.1  The collection of samples  Sampling for juvenile yellowfin tuna Yellowfin tuna were collected for otolith samples from two fishing gear types at six landing  sites in General Santos City. Small yellowfin fish captured in the ringnet fishery were collected from carrier vessels upon landing at San Andres, RDFI, Amadeo, and Delapina sites in Calumpang, and the Lion Beach site in General Santos City. These fish were kept, on average, for three days in insulated fibreglass tanks filled with seawater and ice. A smaller portion of samples were collected from the municipal handline fishery. These handline caught yellowfin were collected upon landing at the Dole Philippines Seafood site at Calumpang and were usually kept on ice for up to six days prior to landing. Yellowfin tuna were identified by external features (Yesaki, 1982) such as body shape and lateral dorsal to ventral banding patterns (Figure 1.1a and 1.1b). Throughout this thesis, "tuna" has dropped from the common names of tuna species. Fresh (iced) samples obtained from carrier boats were transported back to the laboratory at the Notre Dame Dadiangus College where fork lengths (linear distance between the most anterior projection of the snout when the jaws are closed and the most posterior portion of the flesh on the lobes located at the fork of the caudal fin) were recorded to a millimetre using a measuring board. Whole wet weights (blotted fish) were recorded to a milligram using a triple beam balance.  Samples obtained from municipal handline fishermen  were measured and weighed at the landing site using the same methods as in the laboratory. Larger fish at the Dole site were weighed using a 10 kilogram dial type market scale. Fish collected from the ringnet fishery were gutted and livers examined to insure that no bigeye tuna {Thynnus obesus) were misidentified as yellowfin. Yellowfin liver lobes are asymmetric in length with the right lobe being somewhat narrower and longer than the middle and left lobes Figure 1.2). Bigeye liver lobes are all similar in length with the middle lobe being somewhat wider  22  Figure 1.1b  Photograph of juvenile yellowfin tunas showing vertical banding.  than the right and left lobes (Gibbs and Collette, 1967). Bigeye livers were found to be much darker in colour than those of yellowfin. Fish sampled from the handline fishery were not gutted and therefore fish not identified positively as yellowfin and verified by two fishermen were not sampled.  1.1.1  Conversion from thawed to fresh measurements Due to the unavailability of fresh samples, a few fresh frozen yellowfin were sampled from  the storage freezer at Dole. These frozen fish were completely thawed prior to sampling. Conversion factors from thawed to fresh fork lengths and weights were obtained from a sample of fish which were measured before freezing and after thawing. Methods used for the conversion test were standardized to those actually used at Dole. Fresh fish measurements were plotted against thawed measurements and the predictive regression equation was used to adjust thawed fork lengths and weights to their fresh equivalents.  1.1.2  Relationship between weight and fork length The relationship between weight and fork length was determined by least squares,  regression methods employed by the curve fitting program FISHPARM (Prager, Saila, and Recksiek, 1988) on a microcomputer. The weight length relationships for yellowfin from the present study, a previous Philippine study (Ronquillo, 1963 - females only) and a Hawaiian study (Tester and Nakamura, 1957) are plotted for comparison.  1.1.3  Change in body shape Small yellowfin have a laterally compressed body form in contrast with the robust form of the  adults. To investigate the change in body shape of yellowfin a condition factor was calculated for eight fork length intervals using the following equation, CF = W / L * 1000 3  where W = weight in grams  L = fork length in centimetres  25 Condition factors were plotted against fork length interval and data trends were taken as an indicator of body form changes.  1.2 Removal of sagittal otoliths To remove the otoliths, fresh or thawed yellowfin were laid flat and decapitated by a transverse cut just posterior to the preopercular edge. The cut edge of the head portion was then placed flat on the cutting surface and a frontal cut, slightly dorsal to the orbits was made to remove the dorsal portion of the head. When cut in this manner the cranial cavity and brain were exposed (Figure 1.3a). Once the brain tissue was removed the semi-circular canals could be seen on either side of the midline, in the lateral posterior region of the cranial cavity. Three pairs of otoliths (sagitta, asteriscus, and lapillus) are contained within the membranous labyrinth (inner ear) (Lowenstein, 1978). Part of this labyrinth, the sacculus, encloses the sagitta and is located in the prootic bone cavities ventral and slightly posterior to the medulla. With fine forceps both the membranous sacculi were carefully dislodged from their cavities and the sagitta removed. Sagittal otoliths were cleaned of membrane, blood, and tissue by rinsing in water, and soaking in bleach (sodium hypochlorite) until all membranes were digested (30 seconds) (Wild and Foreman, 1980). Clean otoliths were rinsed in water and acetone then allowed to air dry before being stored in plastic capsules. Samples were numbered sequentially from the first otolith pair taken on the first sampling day (Figure 1.3b).  2. Scanning electron microscope examination of sagittal otolith microstructure  The scanning electron microscope (SEM) examination of sagittal otoliths serves three functions; verifying, at the microstructural level, the existence of growth zones on otoliths, allows identification and interpretation of growth zones as bipartite increments, and justifies the use of a light microscope for increment examination. The SEM allows the examination of otolith structure at magnifications high enough to easily identify the crystalline "incremental" zone and the  Figure 1.3a  Figure  1.3b  Photograph of juvenile yellowfin head with brain exposed.  Photograph of otolith sampling set up.  protein matrix "discontinuous" zone which together make up one bipartite ageing structure or increment. The recognition of otolith growth zones, the conceptualization of these growth zones as distinct ageing structures or increments and gaining the confidence required to interpret these increments are critical to the study. The SEM  can also be used to measure the narrowest  increments on the otoliths. Increment widths less than 0.2 microns are less than the resolution powers of the LM.  Many authors have warned against the use of light microscopy for examining  otolith increments because of the inability to detect very narrow increments resulting from starvation experiments or other situations where growth is not appreciable (Campana and Neilson, 1985; Jones and Brothers, 1986; Morales-Nin, 1988). At the microstructure level, teleost fish otoliths consist of alternating growth patterns of incremental and discontinuous zones (Degens et al., 1969: Panella, 1974; Dunkelberger et al., 1980). The incremental or accretion zone (Mugiya, 1987) is composed mainly of calcium carbonate in the form of aragonite crystals and is recognized as an area with elongated crystal structure arranged along the axis between the primordium and the otolith edge. The discontinuous zone is composed mainly of an organic matrix in which amino acids dominate (Degens et al., 1969). When the etching procedure is done with care, these zones are recognized as raised ridges. When the etching has been disruptive, the matrix is dislodged and these zones appear as narrow grooves (Watabe et al., 1982, Morales-Nin, 1987). These discontinuous zones (ridges or grooves) intersect at right angles the crystal structure of the incremental zones. Recent studies have shown that the formation of these growth zones result from the antiphasic deposition of calcium and organic matrix in rainbow trout, Salmo gairdneri (Mugiya, 1988). One complete increment unit is identified as a bipartite structure comprised of one incremental and one discontinuous zone and in many fish is formed over a 24 hour period (Gjostaeter et al., 1984; Campana and Neilson, 1984). Sagittal otoliths extracted from yellowfin collected at the ringnet landing sites were prepared and examined under a JEOL SEM  at the Department of Mining and Metallurgical Engineering,  University of the Philippines, College of Engineering in Quezon City, Philippines. Clean dry otoliths were decalcified by submersion in a seven percent Ethylenedinitrilo tetraacetic acid (EDTA) solution  (adjusted to pH 7.4 with NaOH) for one minute (Radtke and Waiwood, 1980), rinsed in water, air dried then mounted sulcus side down (Figure 1.4) with double sided tape to a brass SEM  stub.  Mounted otoliths were gold coated in a vacuum for four minutes. Plated otoliths were examined at magnifications of 10 to 2000 times, under the SEM.  Increment microstructure was identified under  1500 to 2000 times magnification and overall increment morphology was examined closely at 400 times magnification for later identification and interpretation under light microscopy. Measurements were taken from electron microphotographs of both the narrowest and the widest increments seen on the otoliths.  3. Oxytetracycline (OTC) validation experiment  The formation rate of marks on calcified structures must be validated (Chilton and Beamish, 1982) for all ageing studies. The daily rate of increment formation has been validated for many larval and adult fish (Brothers et al., 1976). For tropical tunas, validation has been carried out by OTC injection of tagged and released yellowfin and skipjack (Wild and Foreman, 1980) and albacore (Laurs et al., 1985) in the Eastern Pacific, by OTC submersion of larval black skipjack (IATTC, 1989) in Panama, and by known age laboratory reared skipjack in Honolulu (Radtke, 1983). To validate the increment formation rate for yellowfin less than 40 centimetres in fork length, an oxytetracycline marking experiment was arranged at the Kewalo basin laboratory facility of the Southwest Fisheries Center/NOAA. The hypothesis tested was that the number of increments formed on the sagittal otoliths of juvenile yellowfin from a OTC mark to the otolith edge would equal the number of days elapsed from OTC injection. Yellowfin tuna were caught by a commercial pole and line vessel off Oahu the morning of November 9, 1987.  These fish were held in the fishing vessel's live baft well, and delivered to the  Kewalo Basin laboratory, at 1430 hours the same day they were captured. Fish were netted in the bait well, placed into a small tank, and transferred to the outdoor tanks where they were maintained  Figure 1.4  Electronmicrophotograph of a yellowfin otolith.  30 throughout the experiment. These fish were fed at least twice daily on a diet of assorted fish. Twelve healthy fish were injected intermuscularily with 17 ml of 100 mg/ml of oxytetracycline (McFarlane and Beamish, 1987) on November 19, 1987. Two fish died, one after 3 days and the other after 12 days from injection. Two fish were killed 21 days after injection, six fish after 28 days and the final two after 39 days. Fork lengths and wet weights were measured from freshly killed fish and sagittal otoliths were removed, cleaned, and stored in darkness. Otoliths were mounted sulcus side down onto glass slides with FLOTEX and examined for the location of the oxytetracycline mark using a ZEISS microscope with an ultra-violet light (UV) beam attachment. Both otoliths from each fish were examined and the otolith displaying the clearest OTC ring and increment configuration was selected for counting. The OTC mark was identified under the UV light and associated with a single increment by faint illumination with white transmitted light. Increments were counted from the OTC marked increment to the otolith edge along the primordium to post-rostrum axis (Figure 1.5) using the white transmitted light at 400 1000 times magnification. Counts were made until three consecutive readings were identical. This number was then taken as the increment count for that otolith. The number of increments counted between the mark and the otolith edge were plotted against the number of days lapsed between the time at marking and the time at sacrifice to determine the rate of increment formation. The otolith distance from the OTC mark to the post-rostral tip was measured with an optical micrometer and average increment widths were calculated.  4.  Ageing  4.1 Sample sizes: a priori estimate An estimate of the sample sizes required in each fish size interval was calculated using published data from Wild (1986) and a manipulated t-test formula from Larkin (1979). n > 2 *s *t 2  2 ( 2 t a i l K i jp =  . , . 05  df  / (difference between means)  2  2n2)  Figure 1.5  Diagram of a right sagittal otolith.  dorsal  32  The data on increment counts were grouped by four centimetre fish size intervals and the mean increment count, standard deviation, and the difference between the means were calculated.  The  larger the difference between sample means, the smaller the sample size required to reject the hypothesis that there is no significant difference between sample means. The number of samples required to detect a true difference between mean increment counts of each four centimetre group, 95% of the time, was estimated by iterations of the formula.  4.2  Preparation of otoliths One otolith from each sample was embedded in a fibreglass resin block and sectioned (J.  Uchiyama pers. com.)  as most were too large and opaque, when whole, for proper examination with  a light microscope. A two millimetre frontal section was taken in a plane containing both the primordium and the post-rostral tip (Figure 1.5) using an ISOMET slow speed saw fitted with two diamond edge blades.  Sections were sanded by hand on 400 and 600 grit sandpaper until the  primordium and the post-rostral tip could just be seen under the surface. Careful polishing on 0.3 /jm lapping paper was done until the primordium and the post-rostral tip were revealed.  Sections  were then etched in a one percent hydrochloric acid (HCL) solution (one second), rinsed in water, air dried and mounted with FLOTEX onto a glass slide. Smaller otoliths were not embedded in fibreglass but were simply etched whole by submersion in five percent HCL  (three seconds), then rinsed in water, and air dried. These small  otoliths were mounted whole, sulcus side down, with FLOTEX onto a glass slide. For eight samples, one otolith of each pair was prepared as a whole mount and the other prepared as a section. Counts were completed on all eight pairs and a t-test for paired comparisons was used to determine significant differences in counts between methods of preparation and between left and right otoliths of a pair.  33 4.3 Counting increments Otolith sections and whole mounts were selected at random by sample number and examined under immersion oil with a WILD phase-contrast light microscope at 200 - 400 times magnification and white transmitted light. Discontinuous zones were used as a reference when counting increments. The narrow and well defined ridges of the discontinuous zones made them much easier to discern as opposed to the variable width of the incremental zone structures. The overall morphology of an increment was recognized from prior SEM examinations and increments counted as a single bipartite structure containing both an incremental and a discontinuous zone. After an initial scanning of the otolith, two preliminary counts were made along a straight line path between the primordium and the post-rostral tip (counting path). Then four counts were made, two in each direction along the counting path, and the arithmetic mean of these four counts taken as the increment count for that otolith. A random sub-sample of 20 otolith sections was re-examined under LM, three months after completing the initial counts. A t-test for paired comparisons was used to determine significant differences between initial and re-examined counts. Ages for yellowfin over similar fork lengths for the Eastern Pacific (Wild, 1986) were plotted along with the data collected from the Philippines. Fork lengths for the Eastern Pacific yellowfin were measured on frozen fish, therefore the fork lengths were adjusted to their fresh fish equivalents using the predictive regression equation calculated earlier.  4.4 Sample sizes: posteriori adjustment The data collected from 150 otoliths were used to adjust the a priori sample size estimate. The smaller standard deviation of increment counts for the small size classes allowed for the narrowing of the size intervals to three and two centimetres. The same t-test formula used in the a priori estimate was used iteratively to obtain adequate sample sizes to detect differences between these new sample means. Standard deviations and 95% confidence limits for the sample means were calculated.  34  5.  5.1  Growth  Growth in fork length and weight The von Bertalanffy and Gompertz growth functions were fit by non-linear least squares  regression methods using the curve fitting program FISHPARM (Prager, Saila, and Recksiek, 1988) on a microcomputer. To determine the linear growth stanzas for fork length and weight, the data were sorted by age and size was regressed against age. Beginning at the smallest point, successive points were added to the regression and r values were computed for the resulting two 2  stanza line. The cut-off or transition point between regression lines was identified when r values 2  were maximized over the entire data set. Decisions as to the best fit equation for a set of data were made on the basis of highest correlation (r values), and least bias, as determined by the most randomly distributed pattern of residuals. Functional regression lines were determined for all linear relationships, unless otherwise stated, as all the variates are subject to error in measurement or inherent variability or both (Ricker, 1973). The yellowfin growth rate from zero to the lowest data point was extrapolated by assuming a linear rate from 0 to 15.2 cm.  5.2 Comparison of growth Ages for yellowfin over similar fork lengths obtained by Wild (1986) were plotted along with the data collected from the Philippines. Thawed fork lengths for the Eastern Pacific were adjusted to fresh equivalents using the predictive regression equation calculated earlier. Linear growth equations for yellowfin from Uchiyama and Struhsaker (1981) and Wild (1986) for yellowfin of similar size are plotted along with the linear growth equations obtained in this study for comparison.  35 Results  1.  Otolith samples collected  Between March 4 and August 17, 1987, 545 yellowfin sagittal otolith pairs were collected from six landing sites along the shoreline in and around General Santos. Of these 545 otolith pairs sampled, 357 were taken from fish caught by ringnet gear and 188 were taken from fish caught by handline. The yellowfin sampled, ranged from 15.2 to 79.0 centimetres in fork length and 50.5 to 9200.0 grams in whole wet weight. One hundred and eight samples were taken from previously frozen fish.  1.1 Conversion between thawed and fresh measurements The conversion between thawed and fresh fork lengths and weights were determined by predictive regression equations (Figure 1.6a and 1.6b) All measurements taken from thawed samples were converted to their fresh equivalents using these equations, L, = 1.028 * L, - 0.117  n = 24  1^ = 0.997  W  n = 24  r = 0.995  f  = 1.025 * W  where  t  + 0.626  2  L, = fresh fork length in centimetres L, = thawed fork length in centimetres W, = fresh whole weight in grams W, = thawed whole weight in grams  Fish typically shrink when frozen and in this study as much as 1 centimetre in a 40 cm fish was lost after freezing.  1.2 Relationship between weight and fork length The allometric equation describing the relationship between fresh weight and fork length is shown in Figure 1.7 and is,  Figure 1.6a  Relationship of fresh against thawed fork lengths.  t h a w e d w e i g h t in g r a m s  Figure 1.7  Relationship between weight and fork length.  fork length in centimetres  W = 0.0146 * F L  3064  where W = whole wet weight in grams L = fork length in centimetres A comparison of the weight - length relationship between the Philippine and Hawaiian yellowfin is shown in Figure 1.8. The data from Ronquillo show a slightly heavier fish below 70 cm and a slightly lighter fish above 70 cm than the present study. The Hawaiian yellowfin follow a similar trend yet the departure from the present study is less. Overall the relationships are very similar. The intersection of all three curves just below 70 cm may indicate the point of departure in growth between males and females. The samples from both this study and the Hawaiian study do not distinguish between males and females whereas the Ronquillo study shows the results for females only.  1.3 Change in body form Condition factors calculated for fork length intervals show an increasing trend from 19 to about 35 cm followed by a levelling off between 35 and 56 cm (Figure 1.9). Although the means are not significantly different, the data suggest that yellowfin less than about 35 cm in fork length increase in length at a more rapid rate than weight. After the 35 cm point the weight increases in proportion to the length cubed.  2. Sagittal otolith microstructure  Otolith samples viewed under the SEM show the typical growth pattern of concentric layers or increments. When examined under 2000X magnification these growth increments displayed the distinctive bipartite ultrastructure of incremental and discontinuous zones (Figure 1.10a). Then by reducing the magnification, these very distinct growth zones coalesce to form bipartite increment patterns at lower magnifications. Electron microphotographs show the concentric pattern of incremental and discontinuous zones which form from the primordium by successive layers along the outer edges of the otolith (Figure 1.10b).  Figure 1.8  Comparison of the weight - fork length relationship between the Philippines and Hawaii.  f o r k l e n g t h in c e n t i m e t r e s  40  Figure 1.9  Condition factor at fork length for juvenile yellowfin. means +  sd  21 -,  -a  §16-u  1  5 -I 15  1  1  1  1  1  1  1  1  20  25  30  35  40  45  50  55  m e a n f o r k l e n g t h i n t e r v a l in  centimetres  1 60  41  Bipartite increment structure was easily recognized and ageing increments could be identified at magnifications less than 400X. The narrowest increments were typically the first few surrounding the primordium and the outer most increments at the edge of the otolith. These increments were no less than 1.5 microns in width (Figure 1.11a) which is greater than the 0.2 micron limit to the resolution capabilities of light microscopy (Campana and Neilson, 1985). Measurements of the widest daily increments were in excess of 40 microns (Figure 1.11b). The SEM examination and measurement of these widest increments were instrumental in verifying daily increment patterns under the light microscope. The presence of subdaily marks have been reported by many authors (Wild and Foreman, 1980; Taubert and Cole, 1982; Campana, 1983; Neilson and Geen, 1982, 1984). These subdaily structures were distinguished by their generally faint and incomplete increment morphology. Subdaily marks or structures were common within wide increments (>25 urn) when examined under the light microscope but when examined under the scanning electron microscope, there was no evidence of repeated subdaily marks. These very slight disturbances in surface topography may well be an artifact of the etching process. Ridges and grooves resulting from differential etching along some crystalline formations, particularly in the "rapid growth" or wide incremental zones may appear to be subdaily marks under light microscopy. Transmitted light is reflected differently through areas of differing densities and surface configurations, whereas the images of the SEM show only variations in the surface topography.  3. The rate of increment formation  Sagittal otoliths of all twelve fish injected with oxytetracycline displayed phosphorescent marks when exposed to ultraviolet light (Figures 1.12a and 1.12b). The green phosphorescent band of OTC was concentrated within a single increment and on some otoliths seemed to extend partially into the following increment. The number of increments from the OTC mark to the otolith edge was plotted against the number of days from injection (Figure 1.13). The slope of the predictive  43  Figure 1.12b  Otolith viewed under phorescent light, showing the OTC mark.  days lapsed from  marking  46  regression line was not significantly different from 1, indicating that the number of increments was directly proportional to the number of days lapsed. The daily formation of increments on sagittal otoliths of Central Pacific yellowfin from approximately 25.0 to 40.0 cm in fork length was validated. The overall range of validated fork lengths for yellowfin has now been extended from Wild's 40 to 110 cm to 25 to 110 cm. All increment counts are taken as a direct estimate of fish age in days. Average increment widths in the region between the OTC mark and the edge of the otolith varied from 1.1 to 27.7 microns. Two fish that did not initiate feeding and died prior to sampling incorporated the narrowest increments measured, averaging widths of 1.1 and 3.3 microns. Increments continued to form at a daily rate on otoliths despite the severe stress which ultimately resulted in death.  4.  Ageing  Of the 545 otolith pairs collected, 280 were selected on the basis of fish size and were prepared for ageing. A total of 207 yellowfin were aged by means of counting increments on sagittal otoliths (Appendix I). Sixty-eight whole otoliths were aged representing fish between 15.2 and 28.2 cm in fork length, and 139 sectioned otoliths were aged representing fish between 16.2 and 79.0 cm in fork length. Sixteen of these yellowfin were frozen previous to sampling. Age estimates ranged from 36.5 to 594.0 days, standard errors ranged from 0 to 18.2 and averaged at 4.38. Precision in counting generally decreased with increasing age. Mean increment counts are plotted against mean fork length interval (Figure 1.14). Counts made on left and right otoliths, whole mounted or sectioned from the same sample were not significantly different (Table 1.1). A random subsample of 20 otoliths reexamined three months after the initial counts were completed showed no significant difference between counts (Table 1.2).  Figure 1.14  Plot of increment counts by fork length interval showing means, and +,- one standard deviation.  360 +  90 +  0  10  20  30  40  fork length interval m e a n in c e n t i m e t r e s  50  48 Differences between whole mounted and sectioned otolith increment counts done on left and right otoliths of a pair.  whole  section  difference  57 40 81 80 84 40 59 95  55 54 89 85 80 45 63 93  -2 + 14 +8 +5 -4 +5 +4 -2  t - test  n = 8 t, = 3.96  p < 0.01  Table 1.2  Differences between first and second increment counts on 20 randomly selected otolith sections.  Item  first  second  difference  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20  446 352 88 216 594 99 57 135 447 341 215 472 89 370 446 339 211 94 64 47  425 343 90 211 568 98 55 150 452 345 217 473 85 387 460 350 216 83 66 48  -19 -9 +2 -5 -26 -1 -2 + 15 +5 +4 +2 +1 -4 + 17 + 14 + 11 +5 -11 +2 +1  t-test  n = 20  t, = 4.859  p < 0.001  50  5.  Growth  The von Bertalanffy growth function was fit to the fork length at age data obtained from the otolith increment counts. Data points plotted with the fitted von Bertalanffy growth curve and the plot of residuals are shown in Figures 1.15a and 1.15b. The form of the von Bertalanffy function is, L, = Lo,, * {1 - exp (  k[t  }  r = 0.975  tol)  2  where L, = fork length at age t  L^, = 88.49 cm  k = .0026 and  t = -49.10 days, for these Philippine yellowfin. n  The Loo value computed of 88.49 centimetres is not realistic given that adult yellowfin reach fork lengths far in excess of this. Regardless of this, the von Bertalanffy growth function does not describe the data very well due to the absence of a decrease in growth rate over the ages examined here. Two linear stanzas were also fit to the fork length at age data. The transition point between the two stanzas occurred at 122 days and 35 cm in fork length. The first stanza occurs between 15.2 and 35.0 cm, is described by the regression equation L  = 0.25 * age + 5.87  and represents fork length growth of 2.5 mm  n = 115 per day. The second stanza occurs between 35.0 and  79.0 cm, is described by the regression equation L = 0.096 * age + 22.42 and represents fork length growth of 0.96 mm  n = 92 per day. Data points plotted with both the regression  lines (r = 0.979) and the plot of residuals are shown in Figures 1.16a and 1.16b. Comparing the r 2  values of the fitted equations, the linear stanzas give a marginally higher value than the von Bertalanffy function. However, comparing the plot of residuals, the von Bertalanffy function shows clumped distributions about the mean, over time, indicating consistent biases resulting from the overestimation of fork lengths at when residuals are clumped below the 0 line and an  2  Figure 1.15a  Fork length at age data and fitted von Bertalanffy growth function.  100i  : 100 = 8 8 . 4 9  CO  C  o  80  n = 207  60  r  2  , k = 0.0026  t  0  =  -49.10  = 0.975  40c 0  20-  0 0  100  200  300  400  a g e in d a y s  Figure 1.15b  Scatter plot of fork length residuals.  a g e in d a y s  500  +  600  Figure 1.16a  Fork length at age data and fitted linear growth stanzas.  100 fl = 0 . 0 9 6 * a g e + 2 2 . 4 2 CO CD  n = 92  80-  '  CD  E  '-+-'  en  60-  r  z  = 0.979  o c -C  40-  CP  c  _0)  1_  20-  o  fl = 0 . 2 5 * a g e + 5 . 8 7  4—  n = 115 0 0  100  200  300  400  a g e in d a y s  Figure 1.16b  Scatter plot of fork length residuals.  10  co CD  D >  -10 a g e in d a y s  500  600  underestimation when clumped above the 0 line. The residuals from the linear stanzas show a more even distribution about the mean, over time, indicating that estimated fork lengths are less biased. Growth in fork length from 15.2 to 79.0 cm in yellowfin from the southern Philippines is best described by the two linear stanzas. The Gompertz growth function was fit to the weight at age data and the residuals plotted in Figures 1.17a and 1.17b. The form of the Gompertz function is, W  t  = W  * exp » " - « " <G  0  [  r = 0.964  t ] ) >  2  where W, = weight at age t in grams W  0  = 142.0  G = and  gm  5.42  g = .0025 for these yellowfin. Two linear stanzas were also fit to the weight  at age data. The transition point between the stanzas occurred at 2350 grams and 300 days of age. The first stanza occurs between 50 and 2350 grams, is described by the regression equation W = 9.15 * age - 394.30  n =  175  and represents growth of 9.15 gms per day. The second stanza occurs between 2350 and 9200 grams, is described by the regression equation W = 23.68 * age - 4916.66  n = 32  and represents growth of 23.68 grams per day. Data points plotted with both regression lines (r = 2  0.953) and a plot of the residuals are shown in Figures 1.18a and 1.18b. Comparing the r values 2  of the fitted equations, the Gompertz function gives a marginally higher value than the linear stanzas. Residuals resulting from fitting both the Gompertz and the linear stanzas show clumped distributions about the mean, over time, indicating consistent biases. At young ages the linear stanzas result in residuals that are slightly less bias than those from the Gompertz, as indicated by the less extreme departures from the zero line. The linear stanzas best describe growth in weight for these yellowfin between 50 and 9200 gm.  Calculating from the regression equations, Philippine  yellowfin attain a fork length of 57.5 centimetres and a weight of 3726 grams in a year.  Figure 1.17a  W  8000  n = 207  7000  E a  6000  C  JZ  = 142.40  9000  CO  CD  Weight at age data and fitted Gompertz growth function.  r  0  G = 5.42 g = 0.0025  2 = 0.964  5000 4000  cn 3 0 0 0 '<$u 2000 1000 0 1000  0  +  100  —  i  200  1  1  300  400  1—  500  600  500  600  a g e in d a y s Figure 1.17b  Scatter plot of weight residuals.  2000 1500-1000OT  a >  500-0 -500-1000--1500-2000 0  100  200  300  400  age in days  54  Figure 1.18a  90008000-  Weight at age data and linear growth stanzas.  wt = 2 3 . 6 8 * a g e n -  4916.66  32  700060005000-  = 0.953  400030002000n =  1000wt = 9 . 1 5 * a g e -  0 -1000 — 0  1  100  200  300  400  1  175  394.30 r—  500  600  500  600  a g e in d a y s Figure 1.18b  100  Scatter plot of weight residuals.  200  300  400  a g e in d a y s  56 Comparison of age at fork length between the Philippine yellowfin and the adjusted Eastern Pacific yellowfin show the Philippine fish slightly larger at fork length values over similar sizes (Figure 1.19a). Linear growth models fit fork length at age data for otolith aged yellowfin from the present study in the Celebes Sea, the Eastern Pacific, (Wild, 1986), and the Central Pacific, (Uchiyama and Struhsaker, 1981) (Figure 1.19b). The rate of growth in fork lengths and fork lengths at a year are shown in Table 1.3. The rates of growth calculated for the second stanza of the Hawaiian, the Philippine, and the entire data set for the Eastern Pacific fish are virtually identical. Over the first stanza however, growth rate is greater for the Philippine fish. The size at a year is greatest for the Philippine yellowfin and least for the adjusted Eastern Pacific yellowfin. The data are difficult to compare because of differences in methods between studies such as the measurement of fork length, the use of frozen samples in the Eastern Pacific, the small sample size (n=14) taken in the Central Pacific, and possibly the differences between otolith readers in the interpretation of increments. The interpretation of increments, although consistent within one study may  not have  been between studies.  Discussion  1.  Ageing by sagittal otolith increments  Philippine yellowfin otoliths examined under the SEM  at magnifications of up to 2000X  confirmed the existence of bipartite growth increments analogous to those described as daily in yellowfin of the Eastern Pacific (Wild, 1986), skipjack in the Central Pacific (Radtke, 1983), and bluefin in the Western Atlantic (Brothers et al., 1983). These increments are clearly visible from the etched surface of whole mounted otoliths or of frontal sections taken in a plane which contains both the primordium and the post-rostral tip. Increment widths were large enough to resolve under light microscopy.  57  Figure 1.19a  Comparison of fork lengths at age for the Western Pacific and the Eastern Pacific.  00 • Western P a c i f i c , p r e s e n t s t u d y CO <D  v  80  E a s t e r n P a c i f i c , Wild ( 1 9 8 6 )  adjusted  ~<D E c O  60  _c -i—• cn c JD  40  o u_  20  *****  0 0  100  200  300  400  500  600  700  age in d a y s Figure 1.19b  Comparison of linear growth equations for the Western, Eastern, and Central Pacific.  100 •Western P a c i f i c , present study - E a s t e r n P a c i f i c , Wild ( 1 9 8 6 ) adjusted  co  CD ^_  -4—'  80  Central P a c i f i c , U c h i y a m a & S t r u h s a k e r  CD  a> o H—'  CD  (1981) 60  4 0 -  2 0 -  0 0  100  200  300  400  a g e in d a y s  -t500  +  600  700  58  Table 1.3  Estimates of growth for yellowfin aged by daily sagittal otolith increments.  Region  growth rate (mm/day)  size range fork length  size in a year (cm)  average temp °C  Philippines  4.07 2.50 0.96  0 - 15 15 - 35 35 - 79  57.46  29.5  1.40 0.90  < 64.2 64.2 - 93.0  53.46  40 - 135  51.08  Hawaii  Eastern Pacific 1.05  1  2  source  1  this study Uchiyama and Struhsaker, (1981)  2  26.5  linear projection from the lower end of the data set back to zero adjusted to fresh fork lengths  Wild, (1986)  With a highly mobile pelagic species such as yellowfin, the validation requirements for using increments as ageing structures are difficult to fulfil. Therefore in this study much has been inferred from previous work. An oxytetracycline validation of the increment formation rate for yellowfin in the Philippines was not attempted. But given the daily nature of increment formation in Eastern Pacific yellowfin 40 - 110 cm (Wild and Foreman, 1980) and the present validation of Central Pacific yellowfin 25 - 40 cm, it is unlikely that the occurrence of a rate other than daily would exist in Philippine yellowfin. Daily formation of increments on sagittal otoliths seems to be ubiquitous in yellowfin. However, daily rates of increment formation have not been found in skipjack and albacore tunas. For adult skipjack, Wild and Foreman (1980) reported that the rate was less than daily and considered the interruption of increment deposition due to the energy demands of maturation and reproductive activities. Laurs et al. (1985) found no evidence of this in albacore but could not dismiss the possibility of occasional interruptions or systematic under counting. In both immature (present study) and adult (Wild and Foreman, 1980) yellowfin increments were formed at a daily rate, therefore maturation and reproductive activities do not seem to have an effect on increment rate in yellowfin. It is not known when increments begin to form on yellowfin otoliths. Otoliths have been shown to be present at hatching for many fish, yet the onset of increment formation varies between species. Species of fish having relatively large eggs and long incubation periods, such as the grunion, Leuresthes Atlantic cod,  tenuis,  Gadus morhua  may form 2 to 4 increments prior to hatching (Brothers et al., 1976). (Ftadtke and Waiwood, 1980) and skipjack tuna (Ftadtke, 1983) have  relatively small eggs and were found to initiate increment formation at hatching. Yellowfin eggs are small, about 1.0 mm  in diameter: incubation times are 24 - 38 hours at 26°C and the yolk sack is  absorbed and exogenous feeding begins 3 days from hatching (Mori et al., 1971). Yellowfin, like skipjack, probably begin to form increments at hatching so that one increment is formed 1 day after hatching and continue to form throughout life.  60  Sagittal otolith increments on Philippine yellowfin fulfil all the basic requirements of ageing structures. Increments were identified by microstructure examination of growth zones and interpreted as bipartite structures. These bipartite increments are validated as daily for fork lengths between 25 and 110 cm and are assumed to form from hatching. This study is the first to validate daily increment formation on the sagittal otoliths of yellowfin between 25 - 40 centimetres in fork length and use daily otolith increments in ageing yellowfin from the Celebes Sea in the Philippines.  2. Growth  Philippine yellowfin were estimated to grow at slightly lower rates to those of the Eastern Pacific, over the 40 to 79 centimetre size range of overlap between the data but the size at one year estimates were 6.38 centimetres greater for the Philippine fish. This difference in size is equivalent to 60 or 67 days calculating from Eastern Pacific and Philippine growth rates. Even though the Hawaiian study consists of only 14 individual fish, the growth rate and size at age are very similar to those of the Philippine study over the fork lengths 62 to 79 cm. Below 62 cm, the Philippine growth rates are slightly faster overall and the size at age one is 4 cm greater than the Hawaiian fish. This difference is equivalent to 42 or 44 days calculating from Hawaiian and Philippine growth rates. Extrapolating from zero to the lowest data point for the Philippines, the growth rate is about 4.07 mm/day. This linearized growth rate is slightly higher than the 3.91 mm/day observed for  Thynnus sp.  observed for black skipjack  (1.0 - 19.4 cm in 47 days) and less than the 5.5 mm/day  Euthynnus lineatus  (2.8 - 17.0 cm in 26 days) tank reared at the  Anchotines laboratory of the Inter-American Tropical Tuna Commission (lATTC, 1988a). The discrepancy in size at one year may be due to either a true difference in growth during the early life history of these fish or a misinterpretation of increments in one or two of these studies. Assuming that increments were interpreted correctly for all the studies, the differences in size at age can be separated into either endogenous or exogenous factors. True differences in the inherent pattern of growth between these populations of fish cannot be discounted. Semi-independent sub-  61  populations of yellowfin in the Eastern, Central, and Western Pacific have been separated by morphometric studies (Royce, 1964; Cole, 1980; IATTC, 1988b). Environmental factors can be divided into biotic factors such as the quantity and quality of food and abiotic factors such as the hydrodynamics, salinity, and temperature of ambient water bodies. Many of these environmental factors are beyond the scope of this study but sea surface temperatures ranges were highest in the Philippines at 27.75 - 30.0°C (ODC, 1988), intermediate for the Eastern Pacific at 25.9 - 27.8°C (Schaefer, 1987), and lowest for Hawaii at 22.8 - 25.6°C (Uchida and Uchiyama, 1986) and may account in part for the differences in size (Brett, 1979). Assuming that endogenous and exogenous factors are similar between the Eastern Pacific, Central Pacific, and Philippine fish, the size differences in fish may be attributed to misinterpretation of increments. During early life growth is rapid and increments are very wide. At this stage the possibility of misinterpreting increments is greatest, after this initial phase increments become more consistent and thus easier to interpret. Wild (1986) noted that the area between the 20th and 170th increments were the most difficult to interpret and it is possible that in this area either increments were underestimated by about 2 months in this study or conversely, overestimated by Wild. The similarity of the growth rates but the difference in the size at age seems to point at some initial discrepancy in counting increments. The large departure between the Hawaiian and Philippine size at age at sizes less than 62 cm may be due in part to the small sample of the Hawaiian study but may also suggest problems between counting paths used. Uchiyama and Struhsaker (1981) counted increments between the primorium and the post-rostral tip but along a path over the sulcus rather than a straight line path counted in both Wild's and this study. These discrepancies however can not be resolved without the OTC validation of increment formation in the region of the otolith where the increments are very wide. This region corresponds to fish between zero and 25 cm in fork length. Points of change in growth occur at 35 cm in fork length and 2350 gm in weight over the size range examined. These transition points possibly reflect ontogenetic changes in the early life history of yellowfin. Ontogenetic shifts are often associated with fish size because of the importance  62 size has in determining vulnerability to predators and feeding ability (Nikolsky, 1963; Balon, 1985). Bioenergetics and growth are closely related thus ontogenetic shifts are often reflected in growth rate changes (Larkin et al., 1957). The first fork length stanza (15 - 35 cm) is characterized by a relatively rapid increase in fork length as opposed to weight.  At these small sizes there is an advantage in rapidly increasing  length to avoid predation (Nikolsky, 1963). In the payao environment, 68% of the large yellowfin diet consists of small tunas (Yesaki, 1983). The artificial payao environment may enhance growth by adding selection pressure, because of the high rate of cannibalism, for fast growth in small fish to avoid being preyed upon. The second fork length stanza (35 - 79) shows a change to a slower fork length growth but a continued weight growth. During this stanza it is possible that the developments in body form, swim bladder, and physiology allow the transition from the juvenile mode to the adult mode of lifestyle. The change in fork length growth to a slower rate at the 35 cm transition point and the continued weight growth results in a  1  filling out" of the body shape and  indicates a transformation from the compressed body form of the juvenile fish to the robust scombroid form of the adults. Weight growth increases after the 2350 gm (51 cm) transition point and corresponds quite well to the weight at which the development of the gas bladder increases (Magnuson, 1973). The gas bladder in yellowfin is absent in fish less than 1000 gm, increases rapidly after 2000 gm, and is fully developed in fish of 8000 gm. The gas bladder reduces fish density and less lift is required to maintain hydrostatic equilibrium. Body form and density relates to swimming activity and in tunas is significant for the transition to continuous and high speed swimming (Webb and Weihs, 1986; Magnuson, 1978). Small yellowfin are also characterized by the lack of swim bladders (Magnuson, 1973), subcutaneous heat exchangers, deep red muscle (Dr. K. A. Dickson pers. com.) and the brachialostical arrangements for ram jet ventilation (Roberts, 1987). As the juvenile body shape aligns to that of the adult, the transition to continuous swimming may be accomplished.  Coupled  with the transition to continuous swimming is the development of ram jet ventilation, deep red  63 with the transition to continuous swimming is the development of ram jet ventilation, deep red muscle masses, and subcutaneous metabolic heat exchangers (Roberts, 1978; Collette, 1978; Dr. K. A. Dickson, pers. com.). Recent work by Dickson (IATTC, 1988b) on black skipjack suggests that deep red muscle and heat exchangers develop and begin to function at sizes between 11 and 28 cm in length. Black skipjack is a smaller scombroid and does not possess a swim bladder but like yellowfin is a predatory tuna with similar physiology and high speed swimming capacity. In yellowfin all these tuna adaptations for a high speed lifestyle are developed early in their life histories and are reflected in the growth stanzas. Attainment of the adult form and lifestyle may be achieved at fork lengths between 35 and 70 cm. Once the small yellowfin successfully complete the development to that of the adults they most likely leave the surface waters and migrate deeper in the water column and thus are not caught in the surface ringnet fisheries. Adult yellowfin exist in the waters just above the thermocline, making short trips vertically through the water column both into the deep and surface layers at various times of the day (Yonimori, pers. com., 1987; Holland, 1988). It is speculated by White (1982) and Yesaki (1986) that these yellowfin >50 cm disperse out into the Moro Gulf and Celebes Sea where they were spawned, and enter the Western Pacific. Catch data from the surface pole and line fisheries in the Western Pacific indicate an abundance of yellowfin <100  cm  (Kamimura and Honma, 1962) but no direct evidence supports this Celebes Sea yellowfin dispersion. Tagging studies are underway with the small tunas presently caught in the ringnet fishery in the Celebes Sea (FAO/BFAR, 1988). With information about the dispersion or residency of the yellowfin in the Celebes Sea, the conspicuous absence of landing for the intermediate size range, between about 70 and 115 cm, may be accounted for, and the migration or movements of yellowfin determined for the region.  Section B.  Length frequency analysis and back calculations  Introduction  Age and growth of wild fish may also be determined by the inspection of length frequency modes and by back calculating the size of an individual fish at previous ages. To determine the accuracy of the daily increment ageing technique used in Section A, estimates of age and growth derived from different techniques are used to compare with the estimates derived in Section A and possibly verify the technique. In this section growth rates estimated in Section A are compared with growth estimated by the progression of prominent length frequency modes. Age at fork length and growth rates are compared with back calculated fork lengths at previous ages and back calculated growth rates.  Methods and Materials  1. Length frequency analysis  Length frequencies have been used to separate age classes of fish as early as 1892 by C. G. J. Petersen. Later techniques were developed to separate age classes by statistical, mechanical, graphic, and maximum likelihood means (Harding, 1949; Cassie, 1954; Tanaka, 1956; MacDonald and Pitcher, 1979). The form of the length frequency distributions result from recruitment, growth, mortality, and sampling biases, and are obscured by variability in recruitment dates and individual > growth rates (MacDonald, 1987). Age classes are most easily determined from length frequencies when recruitment is a discrete event and when growth is rapid so that little overlap occurs between successive age groups. This technique has proved reliable during rapid growth of young fish. Using length frequency data collected over time, strong cohorts may  be monitored and growth  extrapolated from the progression of these prominent modes over time.  65  1.1. Data collection from the ringnet fishery Yellowfin were sampled from carrier vessels servicing the commercial ringnet/payao fishery operating in the Celebes Sea and Moro Gulf areas. At two landing sites in Calumpung, San Andres and RDFI, fish unloaded from various carrier vessels were sampled between 0700 and 1200 hours, ten days per month, from February 1987 to March 1988. To unload the catch from the carrier vessels, rattan baskets or wooden boxes are used to scoop fish out of the holds and transport them to the auction floor. Baskets or boxes were intercepted at random before reaching the auction floor and sampled. When sampling, yellowfin were selected from the mixed catch and fork lengths measured to a half centimetre using a measuring board.  1.2 Data analyses The fork lengths collected were grouped by two centimetre intervals and percent frequencies plotted per month from February 1987 to March 1988. Prominent length frequency modes were identified by eye and their progression plotted over time.  2.  Back calculated fork lengths at previous ages  Back calculation is the process of determining the size of an individual fish at some previous age. Utilizing time marks laid down on the hard parts of fish, previous sizes of fish are estimated from established relationships between the size of the fish, the size of the hard part and the time interval between marks. For back calculating, two relationships must be established; the relationship between the size of the size of the body part and the size of the fish over the sizes examined and the relationship between growth zones on hard parts and known time intervals. The relationship between the distance from the post-rostral tip to the primordium (counting path length) on sagittal otoliths and fork length was determined for yellowfin. Then using the relationship between  66 increments and time has been established earlier, fork lengths at previous ages are back calculated.  2.1  Relationship between otolith size and fish size Measurements were made on whole otolith mounts and prepared otolith sections used in  Section A for ageing. The linear otolith distance between the primordium and the post-rostral tip (counting path) was measured using an optical micrometer and a light microscope at 100X to 200X magnification. The allometric relationship between the fork length of the fish and the counting path was estimated by least squares methods using the FISHPARM package on a microcomputer.  2.2 Measurements of increment widths Otolith microincrements were used for back calculations as they have been shown to be directly related to time in days as shown by the marking experiment in Section A. Five otolith sections were selected, by size and clarity of increments, from the prepared otolith sections used for ageing earlier. Sections were viewed at 400X magnification, under a light microscope with a 35 mm camera attached. Photographic slides were taken of all increments along the path from the primordium to the post-rostral tip. A photographic slide of a slide micrometer was also taken for use as a scale for measurements. The slides were projected and the image size adjusted with the scale so that a 2 cm distance on the image was equal to 5 jum. The otolith increments were sequentially traced onto paper, taking care to match up increments at each end of the slide. The traces were then lined up and the distance between increments were measured using a metric ruler. Every effort was made to measure increment widths along a straight line path between the primorium and the post-rostral tip. Distances from the primordium to each successive increment were determined.  2.3  Back calculations Using the allometric equation relating the counting path length to the fork length of the fish,  fork lengths at previous ages were back calculated. Distances between successive sagittal otolith  67 increments reflect fork length growth on a daily time scale.  3. Growth patterns  The detailed history of growth of an individual fish can be studied by examining daily increment widths on fish otoliths. Information on recruitment (Victor, 1982), individual growth and size-selective mortality (Rosenberg and Haugen, 1982), within season growth differences (Jones, 1985), modification of increments by environmental factors (Eckmann and Rey, 1987) and correlations between increments, growth, and life history transitions (Brothers and McFarland, 1981; Tanaka et al., 1987) have all been extracted by otolith increment investigations. The pattern of increment widths over time was investigated to identify transitions in growth rates by changes in otolith increment width trends. Increment width patterns reflect early life growth histories and transitions may pinpoint significant events in yellowfin ontogenies. Successive increment widths were plotted over time from the initial increment surrounding the primorium to the last increment at the post-rostral tip of the otolith.  68 Results  1. Length frequency a n a l y s e s  Monthly length frequencies are shown in Figures 1.20 and 1.21.  Yellowfin less than 24 cm  were present in the landings in every month of the year, suggesting that spawning is possibly a continuous event. Peaks in spawning appear to be twice yearly as evidenced by the three strong cohorts which passed through the fishery between February 1987 and March 1988.  The greatest  numbers of fish were landed between August and October and the least between April and May. The smallest fish were recruited to the fishery at 16 cm and were present during July and December in 1987.  Yellowfin between 16 and 70 cm in fork length were taken by the ring net gear  but the bulk of the landings are made up of fish between 24 and 44 cm.  The 70 cm maximum size  in the landings may represent the upper limit to ring net gear catchability. This may be due to the size of the largest fish attracted by lights to the surface from under payaoes, the smallest fish able to escape the net, or the maximum size of juvenile yellowfin in the area. The progression of prominent modes are shown in Figure 1.22.  There appears to be a  change in growth rate at about 40 cm in fork length. The average growth rate is estimated to be 2.00 mm/day (sd=.31, n=7) over fork lengths between 20 and 40 cm, and 0.6 mm/day (sd=0.0, n=4) between 40 and 54 cm.  69 Figure 1.20  Fork length frequency distributions for yellowfin caught in the ringnet fishery from February to August 1987.  fork length in  centimetres  70 Figure 1.21  Fork length frequency distributions for yellowfin caught in the ringnet fishery from September 1987 to March 1988.  Sep'87  20  0 20-  Oct  0 o  20-  c  CD =3  Nov  0  O" CD  ^  20  §  0  Dec  O CD  Cl.  20  Jan'88  0 20  Feb  0 20  0  Mar 0  10  20 30 40 50 60 fork length in c e n t i m e t r e s  70  71 •  Figure 1.22  Modal progression of prominent length frequencies.  60  50 + TO  | c <p o  J  -o  40  o-  30  •  20  O  average growth rate < 40 cm = 2.0 mm/day  10 +  > 40 cm = 0.6 mm/day  0 0  1  2  3  4  5  6  7  8  9  1 0 1 1 1 2 1 3 1 4 1 5  t i m e in m o n t h s f r o m Feb 1 9 8 7  72  2. Back calculations  2.1  Relationship between otolith size and fish size Two  hundred and sixteen sectioned otoliths were measured for counting path lengths.  Counting path lengths ranged from 1.43 mm to 5.08 mm and associated fork lengths ranged from 16.4 cm to 79.0 cm.  The relationship between the sagittal otolith counting path length and the fork  length shown in Figure 1.23. and is best described by the equation,  L = 10.54 * C P L  1185  r = 0.963 2  where CPL = counting path length in millimetres L = fork length in centimetres  2.2  Measurements of increment widths All the increments along the counting path for five otoliths sections were photographed on a  series of 5 (>8 increments per slide) to 18 (>19 inc/slide) consecutive slides (Figures 1.24a, 1.24b and 1.25a, 1.25b). The number of increments traced from photographic images matched those from previous ageing.  Each increment was interpreted to be equivalent to a time interval of one  day (Section A; Wild and Foreman, 1980).  2.3  Back calculations Individual back calculated growth histories are plotted along with the growth stanzas derived  from otolith ageing (Figures 1.26a and 1.26b). Back calculated growth histories are a record of individual fish growth and thus reflect slightly different information than that shown in the population growth estimate of Section A. Back calculated fork lengths and estimated growth rates at monthly intervals are shown in Table 1.4. (sd=0.88, n=10)  Growth rates over fork lengths 9.1 - 32.5 cm are 2.66 mm/day  and over fork lengths 32.5 - 54.8 is 0.67 mm/day (sd=0.64, n=8).  Figure 1.24a  Figure 1.24b  Photograph of a sectioned otolith showing the increments surrounding the primordium, 400X magnification.  Photograph of a sectioned otolith showing the narrow increments near to the post-rostral tip, 400X magnification.  75  Figure 1.25a  Photograph of a sectioned otolith showing the increment patterns, 400 X magnification.  76  Figure 1.26a  0  50  Comparison of back calculated fork lengths at age and the linear growth stanzas derived from otolith ageing.  100  150  200  250  300  350  400  a g e in d a y s Figure 1.26b  Comparison of the mean back calculated fork lengths at age and the linear growth stanzas derived from otolith ageing.  o  -10-  °0  50  100  150  200 a g e in d a y s  250  300  350  400  77 Table 1.4 Back calculated fork lengths by monthly intervals.  1  2  sample # 3  4  5  days  30 60 90 120 150 180 210 240 270 300 330  spawned fork length age in days  9.7  7.0 17.9  8.3 17.1 25.5 30.4  9.6 20.7 25.7 30.9 35.6  10.7 22.4 31.5 36.3 40.5 44.3 47.1 49.2 51.9 53.2 54.8  May 16.4 41  Jan 26.0 76  Dec 33.2 122  Dec 39.1 178  Jul 55.0 341  backc mean fl  n  mean growth sd  9.1 19.5 27.6 32.5 38.1 44.3 47.1 49.2 51.9 53.2 54.8  5 4 3 3 2 1 1 1 1 1 1  3.02 3.54 2.50 1.65 1.48 1.27 0.93 0.07 0.09 0.04 0.05  0.43 0.37 0.59 0.06 0.08  78 Lee's Phenomenon is an apparent change in growth rate over time when sizes at previous ages are back calculated from marks on hard parts. Lee's Phenomenon may result from a higher mortality of the larger fish in an age class. A reversed Lee's Phenomenon (Lee, 1912) is apparent within the largest four of the five samples. The back calculated fork lengths are increasingly greater for fish captured at increasingly larger sizes. There may be seasonal variation in growth as those samples spawned in January and December are smaller at all ages than those spawned in May and July.  3. Pattern of growth  A consistent pattern of early growth is shown by plotting increment widths against increment number or days from birth (Figure 1.27). The general trend in increment widths is that of uniform width of ~3.5 /jm after hatching to about 10 days, rapid increase in widths to —55 pm at about 30 days, slow decrease in widths to 5 urn at about 120 days, then a more or less steady increment width fluctuating between 2 and 8 pm beyond 120 days. This generalized growth pattern is present on all otoliths examined.  80 Discussion  1.  Length frequency analysis  Despite the possibility of uniform growth, prolonged spawning, and sampling bias of a single gear type, prominent length frequency modes were easily identified and their progression easily traced. The success of the length frequency analysis seems to be in the large sample sizes taken each sampling day, the large number of sampling days each month, the selection of the two busiest landing sites which received landings from nine fishing companies and over 25 catcher boats, the young ages of yellowfin, and their rapid growth. Large and varied samples provided a reasonable representation of the population sampled and most prominent modes were normally distributed. The young yellowfin sampled were growing rapidly therefore there was little overlap between modes. Because of the relatively limited landings of small yellowfin by other gear types, the length frequency data were collected solely from the ring net landings. Therefore the length frequency analysis is subject and sensitive to gear selectivity. When large numbers of recruits enter the fishery over prolonged periods, estimates of growth tends to be underestimated. Also at the other extreme, when the fish exit the fishery, growth tends to be underestimated because the faster growing individuals leave earlier and are not represented in the sample. Therefore the growth rates derived from length frequencies are less than the actual growth rates and are most pronounced at the upper extreme of the data. This seems to be the case when the growth rates are compared with the rates derived from otolith analysis. The change to a slower growth rate at 35 cm in fork length observed in the otolith analysis is also seen in the length frequency analysis at about 40 cm in fork length. Estimates of growth rates derived from the progression of the prominent length frequency modes correspond well with those derived from otolith analysis and for this study serves as a verification of the accuracy of the otolith technique.  81  2.  Back calculations  Both the methods and the applications of back calculations have been scrutinized since it was developed in the early 1920's (Hile, 1970; Carlander, 1981; Smith, 1982). For this study the purpose of performing back calculations was simply to verify estimates of fork length at age and estimates of growth derived from ageing by otolith increments in Section A. The sample size used for back calculating was very small due to the difficulty of measuring increments and may not be representative. Nonetheless the back calculated estimates were very near those derived by otolith analysis. The similarity of fork lengths at age and growth rates does not necessarily validate the otolith ageing technique and possibly all that can be said is that this verification does not point to any gross error. But the similarity of age and growth estimates derived from the three methods; otoliths, length frequencies, and back calculations, supports the verification of accurate age and growth estimates. A reversed Lee's phenomenon is indicated in the back calculated lengths at age. This suggests that the chances of survival are increased with enhanced growth. The sample size is possibly not large, enough to draw any conclusions from but if this trend is apparent, it would suggest that there is size selective mortality which effects adversely on the smaller fish of an age group. This phenomenon has been observed in larvae of bluefin tuna (Brothers et al., 1983), turbot (Rosenberg and Haugen, 1982), yellow perch (Post and Prankevicius, 1987), and Pacific herring (Robinson, 1988). Seasonal variations in growth may also be indicated in the back calculated lengths at age. The samples spawned in May and July are larger at age than those spawned in March and January, and December. Again the sample size is too small to draw any conclusions but the data suggest that fish spawned during May and June may encounter more favourable conditions for growth than those fish spawned during January and December.  82 3. Growth comparisons  Table 1.5 shows growth estimates for yellowfin, over similar sizes to this study, derived from various methods. The fastest growth rates are 4.07 mm/day for Philippine 0.0 - 15.2 cm and 3.91 mm/day for Eastern Pacific 1.0 - 19.4 cm and were estimated by otolith examination. The slowest growth rates found are 0.60 mm/day for Philippine 40.0 - 54.0 cm by length frequencies and 0.68 mm/day for Philippine 38.1 - 54.8 cm by back calculations. The fastest growth rates are expected to occur in the smallest fish and for yellowfin this is the case. The slowed growth between about 38 and 55 cm possibly reflects the body shape changes discussed earlier. Growth rates derived from otoliths for Philippine, Hawaiian, and Eastern Pacific yellowfin (last stanzas) are all within 0.06 mm/day. The differences in rates between these regions may in fact be population or environment specific and therefore reflect true differences in growth between these fish. Possible technical problems for comparisons may be due to the difference in size ranges and sample sizes between these studies. The Eastern Pacific sample contains a greater proportion of large fish than the Philippine sample thus growth rates may be biased towards the larger fish with slower growth rates and conversely the Philippine sample may be biased towards the smaller and faster growing fish. All the yellowfin otolith studies conclude that linear models for growth in fork length describe the data better than the standard von Bertalanffy growth function (von Bertalanffy, 1938) which has, in the past been the most popular growth model used. The growth rates derived from length frequency analyses for the Philippines, Indian Ocean, Papua New Guinea, and the Eastern Pacific are within 0.24 mm/day. Linearized growth for Philippine fish by length frequency analysis are 0.11 mm/day less than that derived from otoliths. In the Eastern Pacific, the growth rate estimated by length frequency analysis is 0.21 mm/day greater than that derived from otoliths. Eastern Pacific studies are difficult to compare due to the possibility of growth differences between years and areas (Davidoff, 1963; Wild, 1986;. Bayliff, 1988). In general all growth rates over similar size ranges are comparable regardless of the technique used.  Table 1.5  Comparison of growth rates for juvenile yellowfin.  area  size cm  growth rate mm/day  method  source  Philippines  20.0 - 40.0 40.0 - 54.0 20.0 - 54.0  2.00 0.60 1.08  If If  this study  0.0 9.1 19.5 27.6 32.5 38.1  - 9.1 - 19.5 - 27.6 - 32.5 - 38.1 - 54.8  3.02 3.54 2.50 1.65 1.48 0.68  be be be be be be  0.0 15.2 35.0 15.2  - 15.2 - 35.0 - 79.0 - 79.0  4.07 2.50 0.96 1.19  otoliths otoliths otoliths  Indian Ocean  30.0 - 70.0  0.95  If  Anderson, 1988  Papua New Guinea  70.0  1.13  If  Wankowski, 1981  Hawaii  < 64.2 64.2 - 93.0  1.40 0.90  otoliths otoliths  Uchiyama and Struhsaker, 1981  Eastern Pacific  84.0  1.17  If  Davidoff, 1963  52.5 - 62.4 62.5 - 72.4 72.5 - 82.4  0.42 - 1.16 0.13 - 1.36 0.43 - 1.01  tagging  Bayliff, 1988  40.0 - 135.0  0.96  otoliths  Wild, 1986  1.0 - 19.4  3.91  observed  IATTC, 1988a  'linearized growth rate  1  1  84 4. Growth patterns  The pattern of increment widths is interesting as it gives a daily record of individual fish growth and allows for the study of growth during early life history stages. All five otoliths show similar patterns of growth which reflect similar life histories and possibly exemplify the yellowfin under payaos in the Celebes Sea. The first eight to ten increments surrounding the primordium are consistently narrow and evenly spaced, followed by increments that increase rapidly in width to a maximum at about one month of age, then slowly over the next one to two months increments decrease in width then remain relatively steady, fluctuating slightly and very slowly diminishing in width over time. These changes in increment widths may indicate life history changes in individual fish (Brothers and McFarland, 1981; Radtke, 1982). The first ten increments may reflect the first life phase from yolk sac absorption to exogenous feeding (Mori et al., 1971). For larval bluefin tuna in the Atlantic, Brothers et al. (1982) indicate that a more rapid growth stanza is reached by 7 mm standard length, about eight or nine increments from the primordium. Black skipjack otoliths also display nine consistent increments from the primordium which are followed by increments difficult to interpret (IATTC, 1989), also indicating a possible change in life history. The second increasing growth phase may reflect a very rapidly growth in fork length and possibly a change in habits at fork lengths about 16 cm. At the fourty day point fish are about 16 cm and are first caught in the surface ringnet fishery. The 16 cm size may point to morphological or behaviourial changes which result in the affinity to payaoes. The slow decline in growth over the next two months and the levelling off at about 120 days correspond to a fork length of about 35 cm and marks the beginning of the second growth stanza detailed in Section A. Four otoliths back dated to spawning indicate that spawning activities occur during the new moon. Yellowfin eggs are about 1 mm, clear with a yellow oil globule, and neutrally buoyant. This spawning strategy would ensure the least disturbance from tidal flow and also provide maximum cover of darkness (no moon) for the newly hatched larvae.  85 Little is known about the early life history phases of yellowfin but the general form of growth discussed here is common for fish larvae (Zweifel and Lasker, 1976) and is very similar to early growth in bluefin (Brothers et al., 1982) and black skipjack (IATTC, 1989). The examination of otolith increments and their widths is a very powerful tool in assessing the early life history of yellowfin.  86  CHAPTER II MONSOONS, SPAWNING, RECRUITMENT, AND GROWTH  Introduction  Reproductive effort or fecundity estimates and recruitment forecasts are central to fisheries management. Population stability and year class fluctuations are determined largely by the annual number of eggs developing for spawning and the survival of the early stages after hatching. Spawning seasons of tropical fishes are generally more prolonged than those of temperate fish (Nikolsky, 1963; Cushing, 1970). Temperate fish tend to have one sharply defined recruitment pulse which is consistent with the short spawning period. Tropical fish on the other hand, having prolonged spawning seasons, would be expected to have a more or less constant recruitment. However, it has been reported that well defined pulses of recruitment occur twice a year for Philippine neritic fishes (Pauly and Navaluna, 1983) and tropical near shore fishes with pelagic larvae (Johannes, 1978) . Yellowfin, as well as other tunas and tropical fishes, have an extended reproductive season within which they spawn successive batches of ova (Schaefer and Orange, 1956; Bunag, 1956; Nikolsky, 1963; Johannes, 1978; Hunter et al., 1987; Schaefer, 1987). Although it seems well established that multiple spawning and extended reproductive seasons for yellowfin are apparent in the Philippines, little is known about the relationship between spawning activity, recruitment, and growth of recruits with environmental factors such as monsoon events for the region. Studies elsewhere have shown that in tropical waters environmental factors leading to variations in food supply, in water temperature, salinity, turbulence, and in competition or predation may impose seasonality in spawning, recruitment, and growth (Longhurst, 1971; Cushing, 1975; Lasker, 1978; Scott, 1979; Bye, 1984). In this chapter monsoon seasons are indicated; peaks in spawning activity, recruitment, and growth of juvenile yellowfin were investigated; then the timing of the monsoons were compared with  these peaks in spawning, recruitment and juvenile growth. Monsoon seasons were determined by published information on sea surface temperature, prevailing wind, and prevailing current trends for the Celebes Sea region. Spawning activity was derived by adult yellowfin condition, hepatosomatic, and gonadosomatic indices collected and plotted over time. Recruitment was estimated by converting the monthly length frequencies to age frequencies then back dating to their birth month. Condition indices and growth rates for juvenile yellowfin were derived and plotted over time.  Methods and Materials  1. Monsoon seasons The strongest environmental influence in the Southeast Asian region are the monsoons. Monsoonal winds and rains effect both water circulation and the seasonal distribution of its physical, chemical, and biological properties. Wind stirring, vertical current shear, and internal waves tend to enhance eddy diffusion and increase the nutrient supply from deeper waters to the surface layer (Seckel, 1972). Increased nutrients increase the productivity of the upper layer which could enhance larval fish growth. Hjort (1914) suggested the strength of a year class fluctuated with the availability of food to the very earliest larval stages. Cushing (1973) elaborates on this hypothesis and suggests that the recruitment to a fish stock is determined in part by the match between larval production and appropriate food and in part by density dependant processes in juvenile life. Another extension of Hjort's hypothesis by Lasker (1981) takes into account the degree of physical disturbance in the environment, explaining that a stable environment is critical for larvae so sufficient concentrations of food for the first feeding larvae are not disturbed. Because tunas spawn pelagic eggs in the upper mixed layer of the open oceans, an account of the seasonal environmental factors affecting the waters of the southern Philippines is an essential component to the investigation of yellowfin spawning, recruitment, and growth. The northeast monsoon affects the region north of the equator from December to February and the southwest monsoon from June to August with transition periods in between (Soegiarto,  88 1981). For the Philippine region, the monsoon winds occur at a 5 and 7 month periodicity (Pauly and Navaluna, 1984). The southwest monsoon begins in May and continues to September and the northeast monsoon begins in October and continues until April peaking in January. The April to May period is characterized by high temperatures, no rain, and weak trade winds. This is the weak transition period or tradewind season between the NE and SE monsoons and is referred to in General Santos City as the "summer". The transition period between the SW  and NE monsoons in  October is not such a calm period and marks the beginning of a cooling trend. The environmental factors of sea surface temperature and prevailing wind and current associated with the monsoons were extracted from published literature for the Celebes Sea region.  1.1  Sea surface temperatures In the tropics, there is little seasonal change in sea surface temperature. Although  temperature may not play a key role in the timing of spawning in general, there is some evidence of spawning associated with water temperatures in the tropics (Johannes, 1978). From August 1986 to August 1987, sea surface temperatures for the Celebes Sea were obtained from weekly OPC KM MCSST charts of the Oceanographic Data Center, Washington D.C.  100  Monthly mean  temperatures were calculated from the weekly means and plotted over the year. Data from September to December 1987 were repeated from the same months in 1986 as only one year of data were obtained.  1.2  Prevailing wind and current The transport and disturbance of pelagic eggs and post hatch larvae are dependant on  wind induced water movements and the prevailing current of the surface waters. There are many reports of spawning peaks of tropical coastal species during periods of calm between monsoonal winds (Wourms and Bayne, 1973). It is proposed that spawning during these periods would reduce the loss of larvae to the open ocean. Similarly, Watson and Leis (1974) reported that for Hawaiian marine fishes, spawning peaks coincide with changes in the local prevailing current. They proposed  89 that during the period of current shifts, the flow rates are reduced and would allow development of the larvae before they were taken out to sea. Johannes (1978) has compiled spawning data for 18 locations in Micronesia within the boundaries of either the equatorial counter current or the north equatorial current. He found that for 13 of the 18 locations, spawning peaks occurred at times of the year when the prevailing wind and current were at their weakest. These trends all support the idea that inshore fishes have evolved reproductive strategies to maximize the movement of offshore larvae to nearby shore habitats. Data for the prevailing wind strength (Beaufort Scale) and direction were obtained from the Atlas of pilot charts, South Pacific and Indian Oceans (U.S.N.O.O., 1966). The most frequent prevailing wind were plotted for each month. Prevailing surface current data were obtained from the Atlas of surface current, Indian Ocean (U.S.N.H.O., 1944). The two most frequent prevailing currents were plotted each month.  2. S p a w n i n g  Bunag (1956) investigated ova diameters in ovaries of yellowfin in the Visayan region of the Philippines and concluded that yellowfin are multiple spawners, releasing more than one batch of eggs during the spawning season, and that there was no well defined spawning season. Yesaki (1983) using gonad indices, determined for the southern Philippines, that the north Celebes Sea is the principal spawning ground and the spawning period for yellowfin extends from September through May, with a minor seasonal peak from September to December, and a major peak from February to May. For 1987, the spawning activity of yellowfin in the Celebes Sea was determined by collecting various weights and associated fish fork lengths in order to derive a condition, hepatosomatic, and gonadosomatic indices for the adult yellowfin landed at Lion Beach. Condition indices allow relative comparisons of the well being or condition of the yellowfin over time. Changes in condition factor indicate physiological changes and when correlated with changes in gonadosomatic and  hepatosomatic indices may provide the pinpointing of significant life history events such as spawning.  2.1  Data collection from the handline fishery Large yellowfin tuna were sampled from the municipal deep sea handline fishery at Lion Beach.  Ten days per month, from 0600 to 1000 hours, fish unloaded from bancas and offered for sale to the exporters were measured for fork lengths and weights and sexed. Yellowfin were identified by distinctive external features such as elongated yellow second dorsal and anal fins and vertical banding patterns on the flanks. Fork lengths were measured, to a centimetre using a two meter wooden calliper and whole wet weights recorded to a kilogram from a hanging 100 kilogram fish scale. Sex was easily determined when fish were dressed due to the mature state of the gonads. Ovaries were large, pink to orange in colour and appeared grainy in texture, whereas testes were large but creamy white in colour and appeared smooth. Often males were "running ripe" and milt was extruded from undressed fish by running a hand in a posterior direction along the flesh of the belly.  2.2. Condition index Adult yellowfin were sampled 10 days per month for condition index at Lion Beach between February 1987 and January 1988. While workers dressed fish, sexes were determined by a visual assessment. A condition index was calculated for each sex using the following equation,  C.I. = W  / L * 10,000 3  b  where W  b  = whole wet weight in kilograms  L = fork length in centimetres  Only adult fish over 35 kilograms are included in the analysis. Condition indices were calculated and plotted by sex from February 1987 to January 1988.  91 2.3 Gonadosomatic and hepatosomatic indices Gonadosomatic indices reflect relative gonad size thus changes over time provide information on the spawning period. All the yellowfin sampled at Lion Beach were over 100 cm in fork length and were sexually mature (Wade, 1949; Yesaki, 1983) therefore have all been included in the analysis. The liver functions as a dynamic storehouse of glycogen. Glycogen stores reflect the feeding rate over short periods of time thus hepatosomatic indices provide a relative measure of the nutritional state of the fish (Tyler and Dunn, 1976). At the Dole Philippines site in Calumpang, yellowfin destined for the sashimi market were sampled once a week, depending on whether samples were available. Yellowfin were measured and weighed in the same manner as above but when the fish were dressed, the entire viscera of each tuna was obtained from the workers. Both whole gonads and whole livers were separated from the remaining viscera, blotted to remove excess moisture, then weighed using a 5 kilogram kitchen scale. Indices were calculated for each sex using the following formulae, G.I. = W  / L * 10,000 3  g  LI. = W, / L * 10,000 3  where W  g  = weight of both gonads in grams  W, = weight of the liver in grams L = fork length in centimetres Gonad and liver indices were plotted by sex monthly from February 1987 to January 1988.  3. Recruitment patterns  Pelagic fish recruitment has important effects on fisheries. Since Hjort's (1914) suggestion that the strength of the year class is determined by the availability of food to the earliest larval stages, evidence has been gathered and the basic hypothesis elaborated on. Starvation, predation on larval fish (Blaxter and Hunter, 1982) and environmental factors have all been shown to affect  recruitment. Cushing (1975) proposes that "the magnitude of recruitment is linked to the match or mismatch of the production of larvae to that of their food". Successful fish recruitment is dependant on conditions which provide the proper food for fish larvae during "critical periods" (May, 1974). Food availability for larval fish has been related to biological dynamics (primary production) until a level is reached whereby physical processes (wind generated turbulance) destruct the biological processes (Lasker, 1981). In the Philippines where seasonal variation is marked by wind direction and force, optimal environment windows for successful pelagic fish recruitment may periods. The continuous spawning of tropical fish may  exist over short  be effected by these windows, resulting in  sharply peaked recruitment during favourable times (Bakun, et al., 1982). In the Philippines, previous studies on yellowfin biology have used the presence of larvae as an indicator of recent spawning. Wade (1950) described many juvenile tunas in Philippine waters and concluded that during all months of the year larval forms were present. For yellowfin Wade determined that the months of greatest spawning intensity were March through December, and the least spawning during January and February.  A similar study by Baguilat (1987) in the Sulu Sea,  reported yellowfin larvae were concentrated near the island of Mindanao during both the October to November 1982 and the February to March 1983 cruises comprising her study in 1982. This technique of determining spawning relies on the assumption that the survival of the pelagic eggs and larvae are not influenced by biotic and abiotic factors prior to the sampling. all these fish larvae sampled were past the vulnerable early stages, ft may  As  be argued that these  samples represented recruitment success rather than actual spawning activity. In this study larval samples were not available to estimate recruitment so relative recruitment frequencies were derived from monthly length frequencies taken from the ringnet fishery landings. Monthly fork length frequencies were first converted to age frequencies using the following equations from Chapter I, Section B, age = (L - 5.87) / 0.250 age = (L - 22.42) / 0.096  for fish between 15 and 35 cm for fish between 36 and 79 cm  Then for each month, age frequencies were back dated to age 0, the month in which they were spawned. Fork length frequencies from February 1987 to March 1988 were cumulated and plotted to display relative recruitment intensity from February 1986 to January 1988.  4.  Growth of juvenile yellowfin  The timing of spawning periods to coicide with the most favourable conditions for the development of the eggs and larvae are determined by selection (Nikolsky, 1963). Thus fish spawned during the peak times should encounter the best conditions and be "better off than those fish spawned during the off season. A condition factor and fork length growth rate were used as simple indicators of well being and were determined for juvenile yellowfin caught in the ringnet fishery.  4.1  Condition index Fork lengths and whole wet weights were collected at ringnet landing sites in Calumpang  between February 1987 and January 1988.  A condition factor was calculated for juvenile yellowfin  between 26.0 and 28.5 cm in fork length using the following formula, CF. = W  / L * 10,000 3  b  where W  b  = whole wet weigh in grams  L = fork length in centimetres As shown earlier in Chapter I, Section A, condition factors increased with increasing size for yellowfin between 18 and 35 centimetres in fork length. Therefore the size range of small yellowfin used to determine condition indices over time were limited to 2.5  4.2  cm.  Growth over time Juvenile yellowfin between 19 and 21 cm in fork length were sampled from ringnet catches  at Calumpang six times between March and August 1987.  Otoliths were extracted and aged by  94  examination of sagittal otolith increments (Chapter I Section A). An average growth rate was calculated for each of the six samples. Significant differences in age at fork length between monsoon and tradewind seasons were tested by an analysis of covariance.  Results  1. Monsoon seasons  Surface water circulation during the Southwest monsoon is dominated by a northerly current and inflow of water from the Pacific into the Celebes Sea is strong. During the Northeast monsoon the southerly water flow causes a cyclonic surface water pattern (Soegiarto, 1981). The North Equatorial current flows westward and splits into northern and southern portions upon reaching the Philippine islands. The southern current flows south along the Mindanao coast and enters the Celebes Sea. Sea surface temperatures, prevailing wind force and direction and prevailing current direction by month for the Celebes Sea are shown in Figure 2.1. The tradewind season beginning at the end of March is marked by a decrease in winds and the rapid increase in sea surface temperatures (SST) from a minimum in February to a maximum in April. The onset of the southwest monsoon at the end of April is indicated by the slow decrease in SST, a change in winds to those from the Southwest, and an increase in wind strength. The southwest monsoon reaches wind strengths of up to force 4 and continues until September. Between September and November the tradewind season is indicated by a change in prevailing wind direction and a change from a slowly decreasing to an increasing SST. Unlike the tradewind season in April and May, the winds do not abate, only change in direction. The Northeast monsoon begins in November, SST begin to decrease and strong winds prevail, reaching their greatest strength (force 4) in January. Winds slowly decrease through to the end of April. The Northeast monsoon is greater in length and wind strength than the Southwest monsoon. Throughout the Celebes Sea the prevailing current direction  31  <u  b ° o> 21-  "0986 data)  •29  £ <1) -K  27"  force +  prevailing wind strength and direction  / / I  /  50SK frequency  prevailing current direction and drift (1 >1.5  1.5  knots  A ~y 7—7* -7  T* / /  1.5  >1.5  knot) >1.5  1.5  -7 7*  ^7*  c >, ID  u  1986  O  d Figure 2.1  J  1987  f  m  a  m  j  j  a  s  Sea surface temperatures, prevailing wind and current by month for the Celebes Sea plotted along with relative recruitment frequency.  co  96  is to the West and Southwest. SST range from 27.75 to 30 °C, wind strength ranges from 0 to force 4, and currents range from 0.5 to 2 knots.  2. Spawning  Adult yellowfin indices are shown in Figure 2.2. Mean monthly condition indices are significantly higher ( F-ratio = 8.51, p < 0.01) for males than females indicating that for the same fork length, males are heavier (Table 2.1). Condition indices show a minimum in October and a maximum in November for both sexes. If the assumption is made that these fish sampled over the year are from the same stock, the drop in condition indices in October show a general loss of weight and may indicate a period of intense spawning. Mean monthly hepatosomatic indices are significantly higher (F-ratio = 37.55, p < 0.01) for females than males, indicating that females have larger liver masses and store more glycogen for short term energy (Table 2.2a). Female liver indices show a maximum in October and a minimum in January. Male liver indices are at a minimum in June and September and a maximum in July and November. Peaks in the hepatosomatic index may indicate periods of good feeding and conversely lows may indicate poor feeding. Gonadosomatic indices again show females with higher values than males, showing that females have larger gonad masses than males of the same fork length (Table 2.2b). This difference however is not significant (F-ratio = 1.94, p = 0.18). Peaks in indices are shown in April and October for both females and males, indicating spawning is occurring or will occur shortly thereafter.  3. Recruitment  Fishing effort, in terms of days out fishing, remains constant throughout the year as there is no regulations on fishing time in the Philippines. Sampling effort also remained constant thus the recruitment frequencies derived from the sampling (Chapter I, Section B) are a true reflection of the  Figure 2.2  Adult yellowfin condition, hepatosomatic, and gonadosomatic indices by sex plotted over time along with the relative recruitment frequency.  98  Table 2.1 Condition indices for adult yellowfin.  Females  Males  month  mean  Sd  n  mean  Sd  n  February 1987 March April May June July August September October November December January 1988  18.51 19.29 19.08 19.31 19.29 18.87 18.54 18.63 17.67 19.48 18.72 18.45  1.79 2.06 1.79 1.42 1.64 2.08 1.65 1.80 1.34 1.75 1.77 1.85  40 22 30 46 94 145 97 89 26 •82 25 55  19.15 19.31 19.82 19.88 19.78 19.62 19.31 19.39 18.35 20.05 19.23 18.96  1.65 2.25 2.12 1.67 1.94 1.80 1.73 1.87 1.81 1.96 1.82 1.95  191 77 161 195 134 371 366 216 162 132 116 128  99  Table 2.2a Hepatosomatic indices for adult yellowfin.  Females  Males  month  mean  sd  n  mean  sd  n  February 1987 March April May June July August September October November December January 1988  1.43 1.45 1.58 1.54 1.44 1.40 1.72 1.49 1.31  0.49 0.56 0.60 0.25 0.21 0.06 0.11 0.44 0.21  18 '6 11 8 6 2 2 5 3  1.13 1.04 0.96 0.85 1.18 1.08 0.96 1.26 1.32 1.23 -  0.36 0.20 0.28 0.14 0.41 0.21 0.25 0.21 0.27 0.46 -  32 7 12 19 9 14 23 5 4 6 -  Table 2.2b Gonadosomatic indices for adult yellowfin.  Females  Males  month  mean  sd  n  mean  sd  n  February 1987 March April May June July August September October November December January 1988  2.64 2.84 2.46 2.35 2.38 2.23 2.75 2.56 2.25  0.85 0.44 0.66 0.66 0.68 0.63 0.47 0.75 0.24  23 7 17 8 6 4 2 5 4  2.30 2.94 2.43 2.05 2.04 2.07 1.98 2.75 2.25 2.38 -  0.79 0.74 0.42 0.63 0.74 0.65 0.51 0.32 0.22 0.47 -  58 7 12 31 10 20 25 6 5 6 -  100 yellowfin population. Relative recruitment frequencies show that recruitment occurs in all months of the year with peaks in November in 1986 and May-June and October-November in 1987 (Figure 2.2 and 2.3). Data at either end of the time scale may be incomplete but the presence of three peaks over the September 1986 to January 1988 period are clearly identified. Recruitment appears to be low in the January-February period and also in August.  4. Growth  Condition factors for small yellowfin between 26.0 and 28.5 cm in fork length reach a minimum in May and peak in August (Table 2.3a). A second low occurred in November and high in February (Figure 2.3). Growth rates for fish between 19 and 21 cm in fork length over the period from March to August reach a minimum on April 20 and a maximum on July 25 (Table 2.3b). Growth rates are significantly slower during April.  Discussion  The Northeast monsoon is the most disruptive seasonal event in the southern Philippines and the tradewind season between the NE and SW monsoon is the most calm season. Two monsoon events occur yearly but in the southern regions the full force of the monsoons is not felt as the major typhoons sweep over the central Visayas region of the Philippines. Wind driven current, mixing in the upper layers, and the development of upwelling gyres may occur during both monsoons but the data is not available to make an assessment here. The associated biological processes are again not available and only the end product is represented as fish spawning, recruitment, and growth.  101 Table 2.3a Condition factors for juvenile yellowfin between 26.0 and 28.5 cm in fork length.  month  mean  February 1987 March April May June July August September October November December January 1988  20.2 20.0 19.6 13.7 17.5 22.9 23.9 22.4 16.6 15.0 16.1 15.7  sd  1.5 1.0 0.8 1.1 2.0 . 3.0 1.9 3.4 1.2 1.4 1.2 1.6  n  82 37 7 2 13 117 425 464 52 93 74 78  Table 2.3b Growth rates for juvenile yellowfin between 19 and 21 cm in fork length.  date  mean  sd  n  March 10 April 5 April 20 June 24 July 25 August 13  3.56 3.18 3.08 3.56 3.58 3.49  0.09 0.07 0.11 0.15 0.18 0.15  '7 7 4 9 6 5  force 4  1987 f  prevailing wind strength and  I  |  m  I I |  a  I  Figure 2.3  I  m  |  I I  j  direction  1  j  I I  J  •I I |  I  a  I  |  I  s  I  |  I  o  I  |  I  n  I  II  d  Prevailing wind, juvenile yellowfin growth rate, condition factors, and relative recruitment frequency.  I [  j  103 Spawning occurs in all months of the year as the gonadosomatic indices indicate that mature females (G.I. >16) (Kikawa, 1959) are present throughout the year and the recruitment of young yellowfin to the fishery is continuous. The October to November peak in spawning activity is evident by a dip in condition and peaks in hepatosomatic and gonadosomatic indices, and coincide well with the recruitment peak. However the May to June recruitment peak does not coincide with the increase in spawning activity. From the gonadosomatic and hepatosomatic indices, spawning seems evident from April to June. Successful recruitment only overlaps the spawning peak over the latter two months. It seems that "survival windows" exist over short periods of time (Bakun et al., 1982) effectively restricting the recruitment pattern. The conditions for larval survival do not seem to be met in the month of April thus recruitment failed over this month. These recruitment "windows" coincide with the five and seven month periodicity of the monsoons and occur for two months immediately at the turn around and onset of both monsoons. Peaks in recruitment occur when SST's are 29.38 °C (sd = 0.375, n = 3), winds are increasing and changing directions, and current is flowing to the west, southwest between one to two knots. Physical and biological processes in the surface layers appear to lead to increased survival of the eggs and larvae. The change and onset of monsoon winds appear to enhance recruitment. Lows in recruitment occur during low temperatures (27.75 °C) and high winds (Beaufort force 4). Strong winds may force Ekman-type upwelling which past wind speeds of about 5-6 m/s generate enough turbulance in the upper layers to disrupt biological processes and adversely effect recruitment (Cury and Roy, 1989). Juvenile yellowfin condition factors and growth rates both reflect the "well being" of the fish and appear to be positively correlated. In general, peaks and lows are matched where data overlap. Condition factors reach a maximum in August and growth rates reach a maximum on July 25. The minimum condition factor was found in May and a minimum growth rate on April 20. The condition factor shows a trend which follows the strength of the monsoon winds. Growth data suggest that the most favourable time for these yellowfin is at the height of the southwest monsoon. During this period it is possible that an upwelling gyre is created in the Celebes Sea by wind driven  104 processes. Thus productivity is increased through the system. Condition factors over the northeast monsoon however are not as high, possibly showing that the winds are two forceful and destruct the stability of the environment causing a decrease in the concentration of food items (Lasker, 1981). Recruitment frequencies appear to be negatively correlated with the condition factors and suggest that months which are favourable to juvenile yellowfin between 26.0 and 28.5 cm in fork length are not favourable for yellowfin larvae. In general, the monsoonal events and tradewind seasons appear to have strong influences on the spawning, recruitment, and early growth of yellowfin in the Celebes Sea. This is not surprising as tunas spawn pelagic eggs in the upper mixed layer thus both eggs and larvae are subject to the ambient ocean processes.  Further investigation of the spawning, recruitment, and  growth of these yellowfin must include some detailed oceanography. Research effort in this area, concentrating on the oceanographic processes, could yield valuable information for recruitment studies and possibly lead to stock size predictions for fisheries.  105  CHAPTER 111 ESTIMATES OF AGE AND  GROWTH FOR ADULT YELLOWFIN  Introduction  Age and growth of adult yellowfin have been estimated by length frequency analysis (Yabuta and Yukinawa, 1957, 1959; Hennemuth, 1961; Davidoff, 1963), weight frequencies (Moore, 1951; Shomura, 1966; Suzuki, 1971; Le Guen and Sakagawa, 1973), scales and vertebrae (Yabuta, et al., 1960; Shomura, 1966; Yang et al., 1969; Huang and Yang, 1974; Suzuki, 1974), tagging (Wild and Foreman, 1980; Bayliff, 1988), and microincrements on otoliths (Uchiyama and Struhsaker, 1981; Wild, 1986). In all these studies, except for Wild (1986) and Shomura (1966), the separation of data by sex was not considered and except for Uchiyama and Struhsaker (1981) and Wild (1986), data were fitted to a von Bertallanfy growth function. In this chapter general biological population aspects will be investigated for the adult yellowfin landed from the handline fishery in the Moro Gulf and Celebes Sea. Equations describing the relationship between weight and fork length by sex are derived from data collected from Lion Beach between February 1987 and January 1988.  Data collected over 1982, 1983, and 1984, were  obtained from BFAR and differences in the relationship between weight and fork length between sexes and within sexes between years were determined.  Because of the difficulty obtaining otolith  samples from adult yellowfin, estimates of age and growth could not be derived from microstructural increments as in Chapter I. Given the spawning periods derived in Chapter II, fork length frequencies for the months of May and October were produced for yellowfin caught in both the ringnet and handline fisheries. Using the Petersen method of length frequency analysis, age and growth estimates were made for the adult yellowfin. Sex ratios by fork length and age were derived and the question of differential growth and mortality discussed.  106  Methods and Materials  1. Data collection from the handline fishery  Fork lengths and whole wet weights by sex for yellowfin collected from Lion Beach and Bula during 1982, 1983, and 1984 for the BFAPAFAO Tuna Project were obtained from the BFAR office in Manila. Sampling procedures were identical to those used at Lion Beach in 1987 (Chapter II).  1.1  Relationship between weight and fork length  Weight and length relationships by sex were determined for 1982, 1983, 1984, and 1987. Functional regressions were fit to the logarithms of whole wet weight and fork length to determine the relationships. Relationships between weight and fork length for yellowfin from the Eastern Pacific (Chatwin, 1959), Central Pacific (Nakamura and Uchiyama, 1966), and the Philippines were compared. Sexually dimorphic growth has been found for yellowfin in the Eastern Pacific. Wild (1986) determined that females were larger than males at about one year of age then thereafter female growth diminishes and males are consistently larger than females for the same age. Sexually dimorphic growth has been suggested for yellowfin from the Central Pacific (Murphy and Shomura, 1972), bigeye (Shomura and Keala, 1963), and albacore (Otsu and Sumida, 1968) but significant differences were not established. At Lion Beach, sexually dimorphic differences were suspected as males seemed to be consistently heavier than females for the same fork length. To test this hypothesis and determine if there is a significant difference in weights and fork lengths between males and females, first a test of homogeniety of slopes were performed on data for 1982, 1983, 1984, and 1987,  and if the assumption of homogeniety of slopes is plausible, covariance analyses  were performed. The statistical package SYSTAT (Wilkinson, 1988) was used to perform tests on a  107 microcomputer. Growth has been found to be significantly different between years for yellowfin in the Eastern Pacific (Hennemuth, 1961; Davidoff, 1963; Wild, 1986). To determine if growth is significantly different between years, logarithms of weights and fork lengths for females and males, were tested for homogeniety of slopes and covariance analyses were performed using SYSTAT.  2. Age and growth of adult yellowfin  Otoliths were not available for microstructure ageing analysis because the primary market for the adult yellowfin required the heads to remain intact. Therefore length frequencies were used to estimate age and derive growth for adult yellowfin from the Celebes Sea.  Fork lengths collected by  sex at Lion Beach were grouped by two centimetre intervals and percent frequencies plotted per month from February 1987 to January 1988.  The analysis of length frequencies has many  difficulties. The most significant being the determination of the "starting point". The choice of birth month or size at age with which to "zero" the growth curve is critical for length frequency analyses. Therefore in this study the uncertainty was removed by using the peak spawning months of May and October as the starting points. Fork length frequencies for the month of May from the ringnet landings (juvenile yellowfin) were combined with the handline landings (adult yellowfin) by sex and plotted. Fork length frequencies for the month of November were treated in the same manner. Size at age  was  determined by the Petersen method of identifying age classes by prominent length modes. Growth was estimated by calculating the difference in fork lengths between prominent modes and dividing by the time lapse between the modes. Growth curves for yellowfin from the Eastern Pacific (Hennemuth, 1961; Davidoff, 1963; Wild, 1986), and the Philippines are compared. Growth curves for yellowfin from the Western Pacific (Yabuta and Yukinawa, 1959; Yabuta, et al., 1960), Central Pacific (Moore, 1951), and the Philippines are compared.  108 3.  Sex ratio  Investigations of sex ratios have shown a preponderance of males over the largest size classes (Marr, 1948; Murphy and Shomura, 1955; Kume and Joseph, 1966; Otsu and Sumida, 1968). The male dominated catches have been attributted to a combination of differential growth, mortality, and availability between the sexes. Differential growth rates are examined and their relative contributions to the overall sex ratios displayed in the landings evaluated. First the phenomenon of the preponderance of males is investigated for fork length intervals. Second, the effects of differencial growth rates are removed by converting all fish lengths to ages, given the growth rates calculated earlier, and sex ratios compared at age. If differential growth is largely responsible for the sex ratios seen earlier than the age converted sex ratios should be near the expected 50:50. Third, to identify possible changes in availability to the handline gear between sexes over time, monthly sex ratios were determined.  3.1 Sex ratio by fork length To determine if significant differences exist between the proportion of males and females at fork length intervals, chi-squared tests were performed for each 5 centimetre fork length interval between 110 and 170 cm. Overall fork length frequencies were determined for both females and males from February 1987 to January 1988.  3.2 Sex ratio by age Using the estimates of growth derived from the length frequency analysis, fork lengths at each two month interval between 26 and 50 months were estimated for each sex. Then age frequencies were derived by grouping fork length frequencies by their associated ages. To determine if significant differences exist between the proportion of males and females at age intervals, chi-squared tests were performed for each two month age interval between 26 and 50 months. Overall age frequencies were determined for both sexes.  109 3.3 Sex ratio by month Sex ratios were determined by month between February 1987 and March 1988 regardless of size or age.  Significant differences between proportions of males and females were determined by  chi-squared tests.  Results and Discussion  1. Relationship between weight and fork length  The regression equation describing the relationship between wet whole weight and fork length for female (Figure 3.1a) yellowfin is, logWT = 2.523 * logFL - 3.714  n = 721  r = 0.705  n = 2099  r = 0.747  2  and for male (Figure 3.1b) yellowfin is, logWT = 2.548 * logFL - 3.742  2  where logWT = logarithm of whole wet weight in kilograms logFL = logarithm of fork length in centimetres  The relationship between weight and fork length by sex for 1982, 1983, 1984, and 1987 are shown in Figure 3.2. 1982 females  logWT = 2.53 * logFL - 3.74  n = 702  r = 0.70  1982 males  logWT = 2.71 * logFL - 4.08  n = 1342  r = 0.73  1983 females  logWT = 2.73 * logFL - 4.14  n = 907  r = 0.70  1983 males  logWT = 2.93 * logFL - 4.54  n = 1440  ^ = 0.80  1984 females  logWT = 3.12 * logFL - 4.95  n = 391  r = 0.89  1984 males  logWT = 3.20 * logFL - 5.11  n = 527  r = 0.89  2  2  2  2  2  110  Figure 3.1a 2.D00T  Relationship between log weight and log fork length for female adult yellowfin.  iogWT - 2.523 + log FL - 3.714 n - 721 r - 0.705  .900-£  2  1.800 +  V"  O  t  1.700 +  1.600--  cj> 1.500 + o 1.400  4  ~T S\ S\  2.1  2.000  Q  1  K n  2,200  2.250  !og fork Isngth in centimetres  Figure 3.1b  Relationship between log weight and log fork length for male adult yellowfin.  .000 IogWT - 2.5+3 * logFL - 3.742  1.900-  n - 20S9 r - 0.747 2  01  1.800-  3  1.700 +  -  1.600  S> 1.500 o  + +  1.400.300 2.000  1 2.050  , 2.100  1 2.150  log fork length in centimetres  1  2.200  2.250  111  Figure 3.2  Relationships between log weight and log fork length for males and females for 1982, 1983, 1984, and 1987.  1982  1983  2.0 males males  1.8 --  logWt = 2.71 » logFL - 4.08 n = 1842  r  2  '--  = 0.73  logWt = 2.93 * logFL - 4.54 n = 1440 r  = 0.80  ' /  ''•'  females  1.6 --  logWt = 2.53 * logFL - 3.74 n = 702  2  2  r  • /'  = 0.70  2  females logWt = 2.73 * logFL - 4.14 n= 907 r  = 0.70  2  1-4+.  Cn O  1984  1987 77~  CD / /  o 1.9 +  males logWt = 3.20 * logFL - 5.11 n = 527  r  = 0.89  2  males  "  logWT = 2.55 * logFL - 4.74  " //  n = 2099  //  r  2  = 0.75  1.7-"  1.5--  " "  females females  logWt = 3.12 * logFL - 4.95 n = 391  r  2  logWT = 2.52 * logFL - 3.71  = 0.89  n = 720  1.32.00  2.10  2.20  2.05  log fork length in c e n t i m e t r e s  r  2  = 0.71  2.15  2.25  112 The assumption of homogeniety of slopes for IogWT against logFL between sexes was  not  met for 1982, 1983, and 1984 therefore slopes were not parallel, and significant differences (p < 0.01) in the relationships between weight and fork length exist. For 1987, a test of homogeniety of slopes was plausible (p = 0.729) so an analysis of covariance was performed and significant differences (p < 0.01) were found for logWt and logFL between males and females. Males are heavier than females of the same fork length. Because sexually dimorphic relationships between weight and fork length exist, to test for differences between years, females and males were treated separately. The assumption of homogeniety of slopes for IogWT against logFL between years was not met for either females and males. Therefore significant differences exist for relationships between weight and fork length between years for females and males. Many previous studies on yellowfin have not differentiated females from males. Clearly from this sample taken from the Celebes Sea sexes should be separated to reduce the variability in parameters dealing with descriptive population statistics. Relationships between weight and fork length for yellowfin from the Eastern Pacific, Central Pacific, and the Philippines are shown in Figure 3.2.  Male Philippine yellowfin are heavier at fork lengths between 100 and 125 cm and thereafter  are lighter than both the Eastern and Central Pacific yellowfin. Female yellowfin show a similar trend and are heavier at fork lengths between 100 and 120 cm but again are lighter at fork lengths above 120.  Some difficulties in comparing absolute weights and fork lengths include the use of  fresh or frozen samples and the measurement of fork lengths. Recent studies on the morphometric and meristic variation of the yellowfin in the Pacific basin indicate that the Eastern Pacific (Mexico . and Equadorj, Central Pacific (Hawaii), and the Western Pacific (Japan and Australia) suggest separate biological groups (IATTC, 1988c). These separate groupings may in the weight-length relationships.  also reflect differences  113  Figure 3.3  Comparison between the relationship between weight and fork length for yellowfin from the Philippines, Eastern, and Central Pacific.  150  120-  o  o Eastern tropical Pacific  •  • Central Pacific  A  A Philippines — males  A  A Philippines — f e m a l e s  I, 9 0  o ".5 c  60-  . A'  30  0 80  100  120  140  +  160  f o r k l e n g t h in c e n t i m e t r e s  180  200  114 2. Age and growth  Monthly fork length frequencies sex are plotted in Figures 3.4 and 3.5 for males and 3.6 and 3.7 for females. The length frequencies for the peak spawning months of May and October were selected and combined with the juvenile length frequencies obtained from the ringnet fishery (Figures 3.8a for females and 3.8b for males). Females were estimated to reach a fork length of 127 centimetres at age 3 and males a fork length of 136 centimetres. Assuming that growth between the sexes does not begin to differ until after one year of age, the growth rate estimated for females between years 1 and 3 is 2.90 cm/mo and for males is 3.27 cm/mo. Thereafter the growth rate decreases from the age 3 rate, at 0.67 cm/mo for females and 0.10 cm/mo for males, every month after age 3. An overall growth curve for yellowfin is shown in Figure 3.9. A comparison of growth curves of yellowfin from the Philippines and Eastern Pacific is shown in Figure 3.10. All the growth curves were derived from length frequency analysis with the exception of Wild's Eastern Pacific curves for males and females. Wild derived his curves from otolith ageing. In general, the Philippine yellowfin are larger at age than the Eastern Pacific yellowfin. The curves by Wild are similar to the curve for Philippine females, although the size at age one is somewhat smaller. A comparison of growth curves of yellowfin from the Philippines, Western, and Central Pacific is shown in Figure 3.11. The curves for the Western Pacific show the greatest size at age 2 but for ages 1 and 4 show smaller sizes than the Philippine yellowfin. In general, the Philippine yellowfin are the largest at age one and obtain the largest sizes at age four. Both the Central Pacific (Moore, 1951) and the Eastern Pacific sample of Wild's for males have sizes at age four between that of Philippine males and females. It is difficult to compare all these growth curves and predicted lengths at age due to the use of frozen or fresh samples, of various models to describe growth, and the sensitivity of the assumed ages to the birth month chosen (Moore, 1951), size at age chosen (Hennemuth, 1961; Davidoff, 1963), and assumed age at recruitment.  Figure 3.4  .  Fork length frequencies for male yellowfin caught in the handline fishery between February and July 1987.  Feb n = 205  r—i  — | 1  ii  ii  -  n  n  i—i  i i  Mar n = 70  —  r—1 i  n  r-i  i  ii  | 1 r—1  Apr n=167  i—1  May n=189  r-i  n  rp  i—i  1I r—i  i  i —  i  i  Jun n = 210 •  —  i—|  i  11—i  I  i  i  1  i—i i—i  _ i _ r-p  r—i  i  r-1  r—|  i ...  nn  i  1  n  r-i  .  i  [—|  Jul n = 383  r—1  r—  • 90  n  i—i  i  70  _|_  1  i  .  n  I  i—]  i i  .  i—|  — nfl  110  Fork  length  -L-  130 in  .a. r~i  I—I  150 centimetres  i 170  190  116  Figure 3.5  15 ..  Fork length frequencies for male yellowfin caught in the handline fishery between August 1987 and January 1988.  Aug n = 374 nr-irp  0  Sep n= 225  15  na  0 u c CD D  Oct n=173  cr CD  >-• rp n  0 c CD  J=L  Nov n = 1 31  O i  CD CL  JZLLTl  0  Dec n=11 6  15 0  Jan n= 383  15  n  0  70  90  110  130  nn 150  Fork length in c e n t i m e t r e s  170  190  117  Figure 3.6  Fork length frequencies for female yellowfin caught in the handline fishery between February and July 1987.  Feb n = 47  20  nn a  0  Mar n = 31  20  ODD o c CD  3 cr  a.  Apr n = 34  20  CD  o c CD U L_ CD  i—i n~i i—11  r~i rp  J H  May n = 52  20  oa  0  Jun n = 101  20  0  20  ..  Jul n = 166  a  H  0  70  90  110  130  Fork l e n g t h in  150  centimetres  170  118  Figure 3.7  Fork length frequencies for female yellowfin caught in the handline fishery between August 1987 and January 1988.  Aug n = 1 13  20  n _. T  Sep n = 108  20  nn  CJ c OJ  D  cr CD  Oct n = 35  20  r~i  0 c CD  O  i_  4i  na-  Nov n = 103  20  CD  CL  HL  i l  Dec n = 64  20  0 20  JZL  _L  Jan n = 69  nnH 70  a  90  110  130  Fork length in  rpn 150  centimetres  170  119 Figure 3.8a  Combined fork length frequencies from the ringnet and handline fisheries for the months of May and October 1987, for females.  25 7 20  mo  e  May females  --  ro  O  E rO  >, ro  15--  o c Q>  10 - -  cr  I 2.  il  .1 5  mo  ••|  H i l l  o E  Oct females  lO  2 0  15 10 - -  5 --  H 10  r  L  ill  20  30  Figure 3.8b  40  50  60  70  80  90  100  110  120  130  140  150  160  170  Combined fork length frequencies from the ringnet and handline fisheries for the months of May and October 1987, for males.  257  mo  May  20-  males  OT aj  >15  •°  +  o c §•  O  £  ro  E ro 5  L ii  c  5 mo  CD O  il  Hi  l l l i l  r"  1  1  h  Oct males rO  15 - 10 - -  5 --  ill 10  20  30  40  50  + +  60  70  80  +  90  H 100  II  h 110  120  130  140  150  liL 160  170  120  Figure 3.9  Growth curves for yellowfin derived from fork length frequencies,  180 —  0  6  males females  12  18  24  30  a g e in m o n t h s  36  42  48  54  Figure 3.10  Comparison of growth curves for yellowfin from the Philippines and the Eastern Pacific.  Figure 3.11  0  Comparison of growth curves for yellowfin from the Philippines, Western, and Central Pacific.  o —  o Philippines - females  • --  • Philippines - males  A  A Western Pacific - Yabuta and Yukinawa  1  2  age in years  3  4  123 Linearized growth rates over sections of the Eastern Pacific growth curves were calculated by Wild (1986) and linearized growth rates for the Western Pacific calculated from lengths at age by Shomura (1966) are shown along with growth rates of the Philippine samples (Table 3.1). Wild found that the growth rates between sexes was not significantly different and therefore pooled the data and presents only one growth rate. The slowest growth rate was estimated for yellowfin from the Western Pacific. The fastest rate was determined by length frequency modes for the Central Pacific. The estimated rate for the Central Pacific as determined by otoliths is close to that for females in the Philippines. The growth rate for Philippine males is near to that calculated as an overall rate for Eastern Pacific yellowfin determined by otoliths. The differences in the growth rates between females and males may be accounted for in part by the differences in liver and gonad weights between the sexes and associated energy differential in producing eggs versus sperm. Female Philippine yellowfin have larger gonad and liver masses than males and must have larger basal energy requirements to maintain them. Yesaki (1983) determined that the feeding rate was not significantly different between females and males from the Moro Gulf and Celebes Sea areas. Thus given the same food intake, females would not have as much energy as males to channel into body growth. The differences in growth rates between sexes in the Philippine sample are larger than those in the Eastern Pacific (Wild) sample. This may be due to the spawning activities in the Celebes Sea area and the mature state of the gonads found in all of the samples. Females expend more energy producing eggs than males do in producing sperm. The resulting growth differential would tend to increase with the length of the spawning period. The Eastern Pacific sample was taken over a large area, north of the equator and east of 137° W and may or may not have contained individuals in spawning condition. Improvements could be made to the length frequency analysis if data were available for the intermediate sized yellowfin. The assumption of linear growth between ages one and three was accepted due to the character of growth up to one year and the absence of length frequency data.  Table 3.1  Growth rates for adult yellowfin.  Area  method  Growth rate cm/mo  FL range cm  Eastern Pacific  If  3.54  70 - 135  Hennemuth, 1961  If  3.34  69 - 134  Davidoff, 1963  otoliths  3.21  40 - 135  Wild, 1986  If  3.92  47 - 130  Moore, 1951  otoliths  2.98  52 - 93  Uchiyama and Strusaker, 1981  If  2.39  51 - 137  Yabuta and Yukinawa, 1959  scales  2.38  54 - 140  Yabuta, et al., 1960  If If  2.90 female 3.27 male  57 - 127 57 - 136  present study  Central Pacific  Western Pacific  Philippines  source  125 3.  3.1.  Sex ratios  Sex ratios by fork length The percentage of males in each 5 centimetre fork length interval were plotted from 110 to  170 cm (Figure 3.12a). Significant differences to the expected 50:50 female:male proportions were evident in all fork length intervals from 135 to 170 cm. Males dominate the landings and make up 100% of the landings over 165 cm. The largest female and male in the sample was 161 and 170 cm respectively. Fork length frequencies for the total sample in 1987 are plotted for females (Figure 3.12b) and for males (Figure 3.12c). The greatest number of females were in the 135 cm interval and largest numbers of males were in the 140 cm interval.  3.2.  Sex ratios by age  The percentage of males in each two month interval were plotted from 28 to 48 months (Figure 3.13a) Significant differences to the expected 50:50 female:male proportions are evident from age 30 to 44 months, thereafter females appear to dominate. The oldest female and male in the sample were 52 and 50 months respectively. Age converted fork length frequencies, using the differential growth rates derived earlier, are plotted for females (Figure 3.13b) and for males (Figure 3.13c). For females the most frequent age interval was 38 months of age, and for males was 36 months of age.  3.3.  Sex ratios by month The percentage males in each month from February 1987 to March 1988 are plotted in  Figure 3.14. Significant differences to the expected 50:50 female:male proportions were apparent in all months except November 1987. Aside from the November sample males dominated all the monthly landings at Lion Beach. The two peak spawning months of June and November show the highest proportions of females in the landings.  126 Figure 3.12a  Percent males by fork length for adult yellowfin from the handline fishery.  100 485 »  s e x r a t i o s by f o r k l e n g t h  7 5 -  * p<.oai * * p<.ao5 50-  25 110  120  130  1 40  Fork length  30  150  160  170  frequencies  Figure 3.12b  females  1987  n = 720 20  10 --  30  -+males  Figure 3.12c  1987  n = 2099 20  -  I0--  0 110  120  130 140 150 f o r k l e n g t h in c e n t i m e t r e s  160  170  Figure 3.13a  Percent males by age for adult yellowfin from the handline fishery.  100  120 *  sex ratios by age 736 *  17 *  483 *  310  75 +  78 * 747 *  50  270 *  * p< .001  25 26  28  30  32  34  36 Age  38  40  42  44  46  48  50  frequencies  40 Figure 3.13b  females 1 9 8 7 n = 720  30  20 +  10 +  40 males 1 9 8 7  Figure 3.13c  n = 2099  30 +  20  10  0 26  28  30  32  34  +  36 38 40 age in months  42  44  46  48  50  128  Figure 3.14 100n  Percent males by month from the handline fishery. :  129 In all fork lengths above 130 cm males dominate the landings and at lengths above 160 cm females become extinct. Removing the difference in growth rates, sex ratios by age show an opposite trend with females absent from the landings until 30 months, then increasing in proportion to dominate the landings at ages 46 and 48 months and at 50 months making up 50 % of the landings. The absence of females below 30 months of age is due to the size at recruitment into the handline fishery. Generally fish less than about 110 .cm are not taken in the deep sea handline fishery.  The males grow faster than females and as a consequence are taken in the handline  fishery at earlier ages than the females. Differential growth rates account in part for the male dominated sex ratios yet do not fully explain the discrepency as age converted sex ratios still show a preponderance of males. Fishing mortality is greater for males than females due to the longer period of exposure to fishing. Total mortality, if anything, is less for females than males, over 3 years of age, as evidenced by the greater proportions of older females in the sample. This leaves the question of differential availability between the sexes to account for the sex ratios. Suda and Schaefer (1965) have shown that there is a significant difference in sex ratio between surface and deep water gear. Deep water longlines catch a higher proportion of males than the surface purse seine and troll gear. Surface gears such as trolling (Tester and Nakamura, 1957) and gillnet (Maldeniya and Joseph, 1986) indicate slightly higher proportions of females overall. Given the differential growth rates higher proportions of females should be present over some fork lengths at some point below where the males begin to dominate.  For the surface gears  (Tester and Nakamura, 1957; Maldeniya and Joseph, 1986) in size classes below 110 cm, a greater proportion of females were found in the landings. Females may exist at shallower depths than the males in the Philippines and as a result are caught less frequently in the deep sea handline fishery. Yellowfin spawn in the upper layers and it is possible that the females spend a greater proportion of their time in shallower water. Towards the end of the spawning peaks in June and November, the proportions of females in the landings are at their highest and may indicate that after the peak spawning females move deeper in the water column. It is also possible that after the spawning  130 peaks, females are more aggressive feeders thus are more likely to be caught. Sex ratios for yellowfin in the handline fishery are biased heavily toward males above the 130 cm fork length interval, below the 46 month interval, and during all months of the year aside from November. Differential growth rates account for the absence of females above 160 cm in fork length and part of the disproportionate sex ratios but can not be the sole cause. Age converted sex ratios show that females are older than the males in the landings but males still predominate. Total mortalities for females appears to be less than males at ages over 3 years. Mortalities of females below 3 years of age may be greater than males. Males are landed at almost three times the rate that females and are nowhere more abundant than males in the handline fishery. Females may exist in shallower water than the males and may be subject to heavier fishing mortalities in the surface ringnet gear. To determine if differential mortality exists in the younger age classes, the proportions of females and males should be examined in the smaller fork lengths and ages. The absence of yellowfin between about 70 and 110 cm in fork length in landings from all the tuna fisheries is conspicuous. The reason for this absence unknown. It seems unlikely that these intermediate sized fish are simply not vulnerable to any of the gear types and raises the question of whether these fish actually exist in the area.  131 General  Summary  Microstuctural growth zones on the etched surface of juvenile yellowfin sagittal otoliths were identified and recognized as bipartite structures analagous to those described as daily in yellowfin from the Eastern and Central Pacific. Daily increment formation on sagittal otoliths was validated for yellowfin between 25 to 40 centimetres in fork length. Sagittal otoliths were extracted from juvenile yellowfin taken in the ringnet fishery and prepared for ageing. Increments were counted on prepared otoliths along a path between the primordium and the post-rostral tip. This ageing technique was shown to be both accurate and precise. The analysis of fork length frequencies taken from the ringnet fishery were obscured somewhat by the constant recruitment of young fish into the fishery. But the progression of prominent modes was consistant with the growth rates determined by otolith ageing. Estimates of size at age and growth from back calculations were also consistant with those from the otolith ageing. Back calculations showed a reversed Lee's Phenomenon and possibly indicates higher survival for those individuals in the age class with higher growth rates. Bi-annual monsoons occur at five and seven month periodicity in the southern Philippine region. Spawning, recruitment, and growth of yellowfin appear to be affected by seasonal changes in environmental factors related to the monsoons. Spawning is continuous throughout the year with peaks of activity in the April through June and October through November periods. Relative recruitment frequencies show peaks in recruitment in May through June and October through November. Juvenile yellowfin sampled between March and August show that growth is significantly lower during the month of April. Spawning, recruitment, and growth are at their highest during the onset of the monsoons. Growth of juvenile yellowfin seems to be inversely related to the relative recruitment frequency. The weight-length relationship for adult yellowfin is significantly different between sexes and between years for each sex. Sex ratios showed a preponderence of males over all fork lengths over 135 cm, ages between 30 and 44 months, and in every month except November. A  132 combination of differential growth, mortality, and availability between the sexes probably account for the preponderance of males in the landings. The growth history of yellowfin has been estimated for fork lengths between 15 and 79 centimetres between 110 and 161 cm for females 110 and 170 cm for males. The growth of yellowfin from 0 to 15 cm and from 79 to 110 cm is extrapolated from information in otoliths and from length frequency analysis. Otolith data indicate, an initial slow growth from hatching to eight or ten days, then growth increases rapidly to a maximum at about 30 days. Growth rates are estimated to be between 3 and 4 mm/day over the first month. At 30 days yellowfin are about 15 cm and begin to enter the surface ringnet fishery. Growth begins to decrease after 30 days and seems to level off around the 120 day point. Between 30 and 120 days, growth is best described by a linear stanza and corresponds to fork lengths between 15 and 35 cm.  Growth rates in this  stanza are 2.5 mm/day. After 120 days growth appears to remain steady. At the 120 day or 35 cm point, another linear growth stanza begins and continues to 79 cm.  Growth in this stanza is 0.96  mm/day or 2.9 cm/mo. The transition points marking a change in growth rate may indicate morphological and physiological transitions in the early life history of yellowfin. A change in body shape from the laterally compressed juvenile to the robust adult occurs along with the development of physiological machinery for high speed sustained swimming. Growth is assumed to be linear between age one and three. Differential growth between sexes is assumed to occur after age one. Estimated growth rates between age one and three are .37 cm/mo greater for males. Females appear to grow at a linear rate of 2.9 cm/mo from 35 to 127 cm in fork length. Males however begin to grow faster than females and depart from the females at fork lengths around 60 to 70 cm.  Males grow at a rate of 3.27 cm/mo from about 70 to 136 cm in  fork length. After age three, female growth rates decrease 0.07 cm/mo and males 0.10 cm/mo every month. Females grow slower than males, reach smaller fork lengths but appear to have lower mortality rates at ages over three years.  133 The growth curve for yellowfin in this region is still incomplete for very small fish that are not caught in the ringnet fishery (<15 cm) and for fish between about 70 and 110 cm in fork length. The absence of "intermediate" sized fish is a mystery. The intermediate yellowfin are possibly not attracted to the payaoes, not available to the fishing gear, or move out of the area. Further studies of age and growth should focus on the small and intermediate sized yellowfin to give an overall growth curve for the fish in this area.  /  The adult fish caught in the handline fishery are constantly supplying the ringnet fishery with new recruits. But it is not known whether individual fish remain in the area and spawn continuously over the year or whether various stocks of adult fish move into and out of the area over short periods just to spawn. The movements of yellowfin within the Celebes Sea and the exchange of fish between the Celebes Sea and the Pacific Ocean or adjacent seas is of interest for further studies. An overall picture of the population dynamics of the tuna stock in the Celebes Sea should be investigated. Stock assessments and yield analyses may  be determined using size at age and  mortality information. If the Celebes yellowfin move out into the Western Pacific they become vulnerable to the tuna fisheries operating in that region. The impact of each fishery in the question of over-exploitation should be considered. Fishing effort for the tuna fisheries in the Moro Gulf and Celebes Sea areas are revolve solely around payaos. Estimates of fishery production per payao or per payao month may well be the best indicator of stock abundances. Little is known about the numbers or concentrations of payaoes for the area, yet the entire fishery is based on these attraction devices. Fishing companies are aware of these statistics on payaoes but are reluctant to divulge any information. Satellite imagery may  answer some of these questions in regard to the number and location of payoes and  will allow a quick assessment of effort. Research on the influence of payaoes on the tuna stock should also be stressed. Little is known about the behavior and interaction of small and large tunas under payaoes. From a practical point of view, the industry would benefit greatly from design improvements of the payao.  134  The importance of oceanographic processes for recruitment studies have been shown and are emphasized here. Pelagic production processes governed by seasonal environmental influences should be investigated and their relation to yellowfin recruitment studied.  Larval otolith studies may  give further insight into the recruitment of yellowfin. Validating a daily increment formation rate for increments in otoliths of larval yellowfin would allow for the investigation of early history events on a daily basis. Significant events in the development of. yellowfin may also be investigated on a fine time scale. Development of physiological processes related to high speed sustained swimming may be pinpointed and related to possible changes in growth. The Celebes Sea tuna fisheries are unique in that they target on both a juvenile and adult form of yellowfin. Both fisheries exist year round because the area is south of the "typhoon belt", markets exist year round, and the fishery is unregulated by the government. Increased demand for fishery products has resulted in increased fishing pressure. Payaoes that are owned by fishing companies are deployed further and further from General Santos City, effectively "staking" fishing claims further into the Celebes Sea.  A recent infiltration into the Celebes Sea by longline vessels  has caused some gear conflicts as payao lines are severed by these steel longlines. The point at which the area is saturated with payaoes and space for all gear types becomes limiting lies in the near future. Directives for reducing gear conflicts and possible means of reducing effort by some sort of payao control should be given some consideration.  Management of the tuna resources may be a  academic exercise as the government of the Philippines has very little in the way of monetary resources for even the most basic monitoring of compliance to regulations. The responsibility of resource management, for the conservation of the tuna stocks, should be placed on the industry. Managers need to inform the industry on possible management practices that would reduce the possiblity of over-exploitation of the tuna stocks. These may be as simple as reducing fishing effort during the peak spawning months, or as complex as a harvest strategy, in time and space, for the entire fishing area.  135 Literature Cited Anderson, Ft. C, 1988. Growth and migration of juvenile yellowfin tuna (Thunnus albacares) in the central Indian Ocean. Manuscript Marine Research Section, Ministry of Fisheries, Male', Maldives. 28-39. Aprieto, V. L, 1980. Philippine tuna fisheries: Status, trend, and management. Jour, of the Natural Resources Management Forum, ll(5):1-43. B.F.A.R., 1987. Status of the Philippine fisheries for tuna. In Report of the Second Meeting of the tuna Research Groups in the Southeast Asian Region. Manila, Philippines, August 25-28, 1987. B.R.C., 1988. Integrated fisheries development project: A total approach to develop the local fishing industry. Business Resource Center, Notre Dame Dadiangus College, General Santos City, Philippines. Newsletter July-August 4(5):4-5. B.R.C., 1989.  Facts and Figures. B.R.C. Newsletter January-February 5(2):8.  Baguilat, R. B., 1987. Occurance and description of tuna and tuna-like larvae in the Sulu Sea. Paper presented at the Second Meeting of the Tuna Research Groups in the Southest Asian Region. Manila, Philippines, August 25-28, 1987. Bagenal, T. B. (ed.), 1974. The Ageing of Fish. Brothers: London, England.  Proceedings of an International Symposium.  Bakun, A., J. Bayer, D. Pauly, J. G. Pope, and G. D. Sharp, 1982. resources. Can. J. Fish. Aquat. Sci. 39(7): 1059-1070. Balon, E. K., (ed.), 1985.  Unwin  Ocean science in relation to living  Early Life Histories of Fishes. Dr W Junk Publishers: Netherlands.  Barut, N., 1987. Tuna Stock Assessment Research Project. Status Report. Bayliff, W. H., 1988. Growth of skipjack, Katsuwonus pelamis, and yellowfin, Thunnus albacares, tunas in the eastern Pacific Ocean, as estimated from tagging data. Inter-Am. Trop. Tuna Comm. Bull. 19(4) :310-358. Beamish, R. J. and G. A. McFarlane, 1983. The forgotten requirement for age validation in fisheries biology. Trans. Amer. Fish. Soc. 112:735-743. BFAR, 1984. Fisheries Statistics of the Philippines. 34:84 Bureau of Fisheries and Aquatic Resources. Manila, Philippines. Brett, J. R., 1978. Environmental factors and growth. In W. S. Hoar and D. J. Randall (eds) Fish Physiology. Volume Vlll. Growth. Academic Press: New York, NY. Brothers, E. B., E. D. Prince, and D. W. Lee, 1983. Age and growth of young-of-the-year bluefin tuna, Thunnus thynnus, from otolith microstructure. In L. D. Prince and L. M. Pulos (eds), Proceedings of the International Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, Billfishes, and Sharks. U.S. Nat. Mar. Fish. Ser., N.O.A.A. Tech. Rep. N.M.F.S. 8:49-59. Brothers, E. B., C. P. Mathews, and R. Lasker, 1976. adult fishes. Fish. Bull. U.S. 74:1-8.  Daily growth increments in otoliths from larval and  Brothers, E. B., and W. N. McFarland, 1981. Correlations between otolith microstructure, growth, and life history transistions in newly recruited French grunts [Haemulon flavolineatum (Desmarest), Haemulidae]. Rapp. P.-v. Reun. Cons. int. Explor. Mer. 178:369-374.  136 Bunag, D. M., 1956. Spawning habits of some Philippine tuna based on diameter measurements of the ovarian ova. Philipp. J. Fish., 4(2):145-177. Bye, V. J., 1984. The role of environmental factors in the timing of reproductive cycles. In G. W. Potts and Ft. J. Wootton (eds), Fish Reproduction: Strategies and tactics. 187-205. Campana, S., 1983. Feeding periodicity and the production of daily growth increments in the otoliths of steelhead trout (Salmo gairdneri) and starry flounder {Platichthys stellatus). Can. J. Zool. 61:15911597. Campana, S. E. and J. D. Neilson, 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci. 42:10141032. Carlander, K. D., 1981. Caution on the use of the regression method of backcalculating lengths from scale measurements. Amer. Fish. Soc. Bull., Fisheries 6(1):2-4. Cassie, R. M., 1954. Some uses of probability paper in the analysis of size frequency distributions. Aust. J. Mar. Freshwater Res. 5:513-522. Chatwin, B. M., 1959. The relationships between length and weight of yellowfin tuna (Neothunnus macropterus) and skipjack tuna (Katsuwonus pelamis) from the eastern tropical Pacific Ocean. InterAmer. Trop. Tuna Comm., Bull. 3(7):305-352. Chilton, D. E., and R. J. Beamish, 1982. Age determination methods for fishes studied by the groundfish program at the Pacific Biological Station. Can. Spec. Publ. Fish. Aquat. Sci. 60:102 p. Cole, J. S., 1980. Synopsis of biological data on yellowfin tuna, Thunnus albacares (Bonnaterre, 1788) in the Pacific Ocean. Inter-Am. Trop. Tuna Comm., Spec. Rep., 2:71-150. Cury, P., and C. Roy, 1989. Optimal environmental window and pelagic fish recruitment success in upwelling areas. Can J. Fish. Aquat. Sci., 46:670-680. Cushing, D. H., 1970. 33:340-362.  The regularity of the spawning season in some fishes. J. Cons. int. Explor. Mer.  Cushing, D. H., 1975.  Marine Ecology and Fisheries. Cambridge University Press.  Davidoff, E. B., 1963. Size and year class composition of catch, age, and growth of yellowfin tuna in the eastern tropical Pacific Ocean, 1951-1961. Inter-Am. Trop. Tuna Comm., Bull., 8(4):199-251. Degens E. T., W. G. Deuser, and R. L Haedrich, 1969. Molecular structure and composition of fish otoliths. Mar. Biol. 2:105-113. Dickson, K. A., 1988. Comparisons of aerobic and anaerobic metabolic capacities at in vivo temperatures in myotomal muscle of tunas and ectothermic Scombrids. (Abstr) 39th Annual Tuna Conference, May, 1987, Lake Arrowhead, CA. Dunkelberger, D. G., j. M. Dean, and N. Watabe, 1980. The ultrastructure of the otolithic membrane and otolith in the juvenile mummichog, Fundulus hereroclitus. J. Morphol. 163:367-377. Eckmann, R., and P. Rey, 1987. Daily increments on the otoliths of larval and juvenile Coregonus spp., and their modification by environmental factors. Hydrobiol. 148:137-143. Fonteneau, A., 1980. La croissance de I'albacore de I'Atlantique est. Int. Comm. Conserv. Atl. Tunas, Coll. Vol. Sci. Pap., 9:152-168.  137 Gibbs, Ft. H. and B. B. Collette, 1967. Comparative anatomy and systematics of the tunas, genus Thunnus. Fish. Bull. U.S. 66(1):65-130. Gjosaeter, J., P. Dayaratne, O. A. Bergstad, H. Gjosaeter, M. I. Sousa, and I. M. Beck, 1984. Ageing tropical fish by growth rings in the otoliths. F.A.O. Fish. Circ. 776:54. Harding, J. P.,1949. The use of probability paper for graphical analysis of polymodal frequency distributions. J. Mar. Biol. Assoc. U.K. 28:141-153. Hennemuth, R. C, 1961. Size and year class composition of catch, age, and growth of yellowfin tuna in the eastern tropical Pacific Ocean. Inter-Am. Trop. Tuna Comm., Bull. 5(1):1-112. Hile, R., 1970. 474.  Body-scale relation and calculation of growth in fishes. Trans. Amer. Fish. Soc. 99:468-  Hjort, J. 1914. Fluctuations in the great fisheries of northern Europe viewed in the light of biological research. Rapp. P.-v. Reun. Cons. Perm. int. Explor. Mer. 20:1-228. Hunter, J. R., B. J. Macewicz, and J. R. Sibert, 1987. The spawning frquency of skipjack tuna, Katsuwonus pelamis, from the South Pacific. Fish. Bull. U.S. 84(4):895-903. lATTC, 1988a. Inter-Am. Trop. Tuna Comm., Quart. Rep. First quarter, January - March 1988. IATTC, 1988b. Inter-Am. Trop. Tuna Comm., Quart. Rep. Third quarter, July - September 1988. IATTC, 1988c. Inter-Am. Trop. Tuna Comm., Quart. Rep. Fourth quarter, October - December 1988. IATTC, 1989.  Inter-Am. Trop. Tuna Comm., Quart. Rep. First quarter, January - March, 1989.  IPTP, 1987. Report of the Second Meeting of the Tuna Research Groups in the Southeast Asian Region. Manila, Philippines. 25 - 28 August 1987. Colombo Indo-Pacific Tuna Development and Management Programme. Johannes, R. E., 1978. Reproductive strategies of coastal marine fishes in the tropics. Env. Biol. Fish 3:6584. Jones, C, 1985. Within-season differences in growth of larval Atlantic herring, Clupea harengus Fish. Bull. U.S. 83(3):289-298.  harengus.  Jones, C. and E. B. Brothers, 1987. Validation of the otolith increment ageing technique for striped bass, Morone saxatilis, larvae reared under suboptimal feeding conditions. Fish. Bull. U.S. 85(2): 171-178. Kikawa, S., 1959. Notes on the regional difference of spawning season of Pacific yellowfin tuna. Rep. Nankai Reg. Fish. Res. Lab. 11:59-76. Kume, S. and J. Joseph, 1966. Size composition, growth and sexual maturity of bigeye tuna, Thunnus obesus (Lowe), from the Japanese long-line fishery in the eastern Pacific Ocean. Bull. Inter-Am. Trop. Tuna Comm. 11(2):45-99. Larkin, P. A., 1978. A handbook of elementary statistical tests. University of British Columbia, Biology 300 (Biometrics) manual. Larkin, P. A., J. G. Terpenning, and R. R. Parker, 1957. Size as a determinant of growth rate in rainbow trout, Salmo gairdneri. Trans. Am. Fish. Soc. 86:84-96.  138 Lasker, R., 1981. The role of a stable ocean in larval fish survival and subsequent recruitment. In Lasker, R. (ed), Marine Fish Larvae. Washington Sea Grant Program. University of Washington Press: Seattle, Wash. Lasker, R., 1978. The relation between oceanographic conditions and larval cnchovy food in the California Current: identification of factors contributing to recruitment failure. Rapp. P-v. Reun. Cons. int. Explor, Mer. 173:212-230. Laurs, R. M., R. Nishimoto, and J. A. Wetherall, 1985. Frequency of increment formation on sagittae of north Pacific albacore (Thunnus alalunga). Can. J. Fish. Aquat. Sci. 42:1552-1555. Lee, R. M., 1912. An investigation into the methods of growth determination in fishes. Publ. Circ. Cons. Explor. Mer. (63):3-35. Longhurst, A. R., 1971. The clupeoid resources of tropical seas. Oceanog. Mar. Biol. Rev. 9:349-385. Lowenstein O., 1971. The labyrinth. In W. S. Hoar and D. J. Randall (eds), Fish Physiology. Volume V. Sensory systems and Electric Organs. 207-240. MacDonald, P. D. M., and T. J. Pitcher, 1978. Age-groups from size-frequency data: A versatile and efficient method of analyzing distribution mixtures. J. Fish. Res. Board Can. 36:987-1001. Magnuson, J. J., 1973. Comparative study of adaptations for continuous swimming and hydrostatic equilibrium of scombriod and ziphoid fishes. Fish. Bull. U.S. 71:337-356. Magnuson, J. J., 1966.  Continuous locomotion in scombroid fishes. (Abstr.) Am. Zool. 6:503-504.  Magnuson, J. J., 1978. Locomotion by scombrid fishes: hydrodynamics, morphology, and behavior. In W. S. Hoar and D. J. Randall (eds) Fish Physiology. Vol. VII. Locomotion. Academic Press, New York, NY. Maldeniya R. and L. Joseph, 1986. On the distribution and biology of yellowfin tuna CL albacares) from the western and southern coastal waters of Sri Lanka. IPTP Coll. Vol. Work. Doc. 1:51-61. Marr, J. C. and M. B. Schaefer, 1948. Definitions of body dimensions used in describing tunas. U.S. Fish. Bull. 47(51 ):241-244. McFarlane, G. A. and R. J. Beamish, 1987. Selection of dosages of oxytetracycline for age validation studies. Can. J. Fish. Aquat. Sci., 44:905-909. Moore, H. L, 1951. Estimation of age and growth of yellowfin tuna (Neothunnus macropterus) in Hawaiian waters by size frequencies. U.S. Fish Wildl. Serv., Fish. Bull. 52:133-149. Morales-Nin, B., 1988. Caution in the use of daily increments for ageing tropical fishes. Newsletter of the Asian Fisheries Society, Fishbyte 6(2): 5-6. Mori, K., S. Ueyanagi, and Y. Nishikawa, 1971. The development of artificially fertilized and reared larvae of the yellowfin tuna, Thunnus albacares. Far Seas Fish. Res. Lab. Bull. 5:219-232. Mugiya, Y., 1987. Phase difference between calcification and organic matrix formation in the diurnal growth of otoliths in the rainbow trout, Salmo gairdneri. Fish. Bull. U.S. 85(3):395-401. Mugiya, Y., 1984. Diurnal rhythm in otolith formation in the rainbow trout, Salmo gairdneri: seasonal reversal of the rhythm in relation to plasma calcium concentrations. Comp. Biochem. Physiol. 78(A) :289-293.  139 Mugiya, Y., N. Watabe, J. Yamada, J. M. Dean, D. G. Dunkelberger, and M. Shimizu, 1981. Diurnal rhythm in otolith formation in the goldfish, Carassius auratus. Comp. Biochem. Pysiol., 68A(4):659-662. Murphy, G. I. and R. S. Shomura, 1972. Pre-exploitation abundance of tunas in the equatorial central Pacific. U.S. Nat. Mar. Fish. Serv., Fish. Bull., 70(3):875-910. Nakamura, E. L. and J. H. Uchiyama, 1966. Length-weight relations of Pacific tunas. In T. A. Manar (ed), Proceedings, Governor's Conference on Central Pacific Fishery Resources, State of Hawaii, Honolulu, 197-201. Neilson, J. D., and G. H. Geen, 1982. Otoliths of Chinook salmon (Oncorhynchus tshawytscha): Daily growth increments and factors influencing their production. Can. J. Fish. Aquat. Sci. 39:1340-1347. Neilson, J. D., and G. H. Geen, 1984. Effects of feeding regimes and diel temperature cycles on otolith increment formation in juvenile Chinook salmon (Oncorhynchus tshawytscha). Fish. Bull. U.S. 83(1):91-101. Nikolsky, G. V., 1963.  The ecology of fishes. Academic Press: London, England.  Orange, C. J., 1961. Spawning of yellowfin tuna and skipjack in the eastern tropical Pacific, as inferred from studies of gonad development. Inter-Am. Trop. Tuna Comm., Bull., 5(6):459-546. Otsu, T. and R. F. Sumida, 1968. Distribution, apparent abundance, and size composition of albacore (Thunnus alalunga) taken in the longline fishery based in American Samoa, 1954-65. U.S. Bur. Comm. Fish., Fish. Bull., 67(1):47-69: Pannella, G., 1971.  Fish otoliths: daily growth layers and periodical patterns. Sci., 173(4002):1124-1127.  Pannella, G., 1974. Otolith growth patterns: an aid in age determination in temperate and tropical fishes. In T. B. Bagenal (ed.). The ageing of fish. Proc. Int. Symp. on the Ageing of Fish. Unwin Bros., Ltd., Surrey, England:28-39. Parsons, T. R., Takahashi, M., and B. Hargrave, 1977. Biological Oceanographic Processes. Edition. Pergamon of Canada Ltd.: Ontario, Canada.  Second  Pauly D. and N. A. Navaluna, 1983. Monsoon-induced seasonality in the recruitment of Philippine fishes. F.A.O. Fish. Rep. 291(3):823-833. Peters, R. H., 1983. MA.  The ecological implications of body size. Cambridge University Press: Cambridge,  Post, J. R., and A. B. Prankevicius, 1987. Size selective mortality in young-of-the-year yellow perch (Perca flavescens) evidence from otolith microstructure. Can. J. Fish. Aquat. Sci. 44:1840-1847. Prager, M. H., S. B. Saila, and C. W. Recksiek, 1987. FISHPARM: A Microcomputer Program for Parameter Estimation of Nonlinear Models in Fishery Science. Department of Oceanography, College of Sciences, Old Dominion University, Norfolk, Virginia. 23529-0276. Tech. Rep. 87-10. Radtke, R. L, and K. G. Waiwood, 1980. Otolith formation and body shrinkage due to fixation in larval cod (Gadus morhua). Can. Tech. Rep. Fish. Aquat. Sci. 929:iii + 10 p. Radtke, R. L, 1982. Otolith formation and increment deposition in laboratory-reared skipjack tuna, Euthynnus pelamic, larvae. In L. D. Prince and L. M. Pulos (eds), Proceedings of the International Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, Billfishes, and Sharks. U.S. Nat. Mar. Fish. Ser., N.O.A.A. Tech. Rep. N.M.F.S. 8:99-103.  140 Ricker, W. E., 1975. Computation and Interpretation of Biological Statistics of Fish Populations. Bull. Fish. Res. Board Can. 191:382. Ricker, W. E., 1979. Growth rates and models. In W. S. Hoar, D. J. Randall, and J. R. Brett (eds.) Fish Physiology. Volume Vlll. Growth. Academic Press: New York, NY. Ricker, W. E., 1973.  Linear regressions in fishery research. J. Fish. Res. Board Can. 30:409-434.  Roberts, J. L, 1978. Ram gill ventilation in fish. In G. D. Sharp and A. D. Dizon (eds) The physiological ecology of tunas. Academic Press: New York, NY. Robinson, S. M., 1988. Early life history characteristics of Pacific herring, Clupea harengus pallasi Valenciennes 1847, in the Strait of Georgia, British Columbia: Hydrodynamics, dispersal, and analysis of growth rates. PhD. Thesis, Zoology Department, University of British Columbia, Vancouver. Ronquillo, I. A., 1963. A contribution to the biology of Philippine tunas. Proceedings of the World Scientific Meetings on the Biology of Tunas and Related Species, La Jolla, California, 2-14 July 1962, F.A.O. Fish. Rep. (6), 3:1683-1752. Rosenberg, A. A., and A. S. Haugen, 1982. Individual growth and size-selective mortality of larval turbot (Scophthalmus maximus) reared in enclosures. Mar. Biol. 72:73-77. Royce, W. F., 1964. A morphometric study of yellowfin tuna, Thunnus albacares Wildl. Serv. Fish. Bull. 63:395-443.  (Bonnaterre). U.S. Fish.  Schaefer, K. M., 1987. Reproductive biology of black skipjack, Euthynnus lineatus, an eastern pacific tuna. Int-Am. Trop. Tuna Comm., Bull. 19(2): 169-260. Schaefer, M. B., and C. J. Orange, 1956. Studies of the sexual development and spawning of yellowfin tuna (Neothunnus macropterus) and skipjack (Katsuwonus pelamis) in three areas of the easter Pacific Ocean, by examination of gonads. Inter-Am. Trop. Tuna Comm., Bull. 1(6):281-349. Scott, D. B. C, 1979. Environmental timing and the control of reproduction in teleost fish. Symposium Zool. Soc. Lond. no. 44:105-132. Seckel, G. R., 1972. Hawaiian caught skipjack tuna and their physical environment. 72(3):763-787.  Fish. Bull. U.S.  Shomura, R. S., 1966. Age and growth studies of four species of tunas in the Pacific Ocean. In T. A. Manar (ed), Proceedings, Governor's Conference on Central Pacific Fishery Resources, State of Hawaii, Honolulu, 203-219. Shomura, R. S. and B. A. Keala, 1963. Growth and sexual dimorphism in growth of bigeye tuna (Thunnus obesus): a preliminary report. FAO Fish. Rep., 6(3): 1409-1417. Smith, C. L, 1982. Summary of round table discussions on back calculation. In L. D. Prince and L. M. Pulos (eds), Proceedings of the International Workshop on Age Determination of Oceanic Pelagic Fishes: Tunas, Billfishes, and Sharks. U.S. Nat. Mar. Fish. Ser., N.O.A.A. Tech. Rep. N.M.F.S. 8:4547. Soegiarto, A., 1981. The oceanographic features of the southeast Asian waters. In Sien, C. L, and C. MacAndrews (eds), Southeast Asian Seas, frontiers for development. McGraw-Hill: Singapore. Sokal, R. R., and F. J. Rohlf, 1981.  Biometry. Second Edition. W. H. Freeman and Co.: Ney York, NY.  141 Suda, A. and M. B. Schaefer, 1965. General review of the Japanese long-line fishery in the eastern tropical Pacific Ocean 1956-1962. Inter-Am. Trop. Tuna Comm. 9(6):305-462. Summerfelt, R. C, and G. E. Hall, (eds.), 1987. Iowa.  Age and Growth of Fish.  Iowa State University Press:  Tanaka, S, 1953. Precision of age-composition of fish estimated by double sampling method using the length for stratification. Bull. Jap. Soc. Sci. Fish. 19:657-670. Tanaka, K., Y. Mugiya, and J. Yamada, 1981. Effects of photoperiod and feeding on daily growth patterns in otoliths of juvenile Tilapia nilotica. Fish. Bull. U.S. 79(3):459-466. Tanaka, K., O. Tabeta, N. Mochioka, J. Yamada, and S. Kakuda, 1987. Otolith microstructure and ecology of the conger eel (Conger myriaster) larvae collected in the Seto Inland Sea, Japan. Nippon Suisan Gakkaishi 53(4):543-549. Taubert, B. D. and D. W. Coble, 1977. Daily rings in otoliths of three species of Lepomis mossambica. J. Fish. Res. Board Can., 34(3):332-340.  and Tilapia  Tester, A. L. and E. L. Nakamura, 1957. Catch rate, size, sex, and food of tunas and other pelagic fishes taken by trolling off Oahu, Hawaii, 1951-55. U.S. Fish Wildl. Serv., Spec. Sci. Rep. Fish. 250:25. Tyler and Dunn, 1979. Ration, growth, and measures of somatic and organ condition in relation to meal frequency in winter flounder Pseudopleuronectes americanus, with hypotheses regarding population homeostasis. J. Fish. Res. Board Can. 33:63-75. U.S.N.H.O., 1944.  Atlas of suface current, Indian Ocean. U.S. Navy Hydrographic Office.  U.S.N.O.O., 1966. Atlas of pilot charts, South Pacific and Indian Oceans. U.S. Navy Oceanographic Office. Uchida, R. N., and J. H. Uchiyama (eds), 1986. Oceanography of northwestern Hawaiin Islands. Fishery atlas of the northwestern Hawaiian Islands. U.S. Nat. Mar. Fish. Ser., N.O.A.A. Tech. Rep. N.M.F.S. 38:21-27. Uchiyama, J. H., and P. Strusaker, 1981. Age and growth of skipjack tuna, Katsuonus pelamis, and yellowfin tuna Thynnus albacares, as indicated by daily growth increments of sagittae. Fish. Bull. U.S. 79(1):151-162. Victor, B. C, 1982.  Recruitment and population dynamics of a coral reef fish. Science 219:419-420.  von Bertalanffy, L, 1938.  A quantitative theory of organic growth. Hum. Biol., 10(2): 181-213.  Wade, C. B., 1951. Larvae of tuna and tuna-like species from Philippine waters. U.S. Fish. Wildl. Serv. Fish Bull. 51:445-485. Wankowski, J. W. J., 1981. Estimated growth of surface-schooling skipjack tuna, Katsuonus pelamis, and yellowfin tuna, Thunnus albacares, from the Papua New Guinea region. U.S. Nat. Mar. Fish. Serv., Fish. Bull., 79(3):517-545. Watabe, N., K. Tanaka, J. Yamada, and J. M. Dean, 1982. Scanning electron microscope observations of the organic matrix in the otolith of the teleost fish Fundulus heteroclitus (Linnaeus) and 77/ap/a nilotica (Linnaeus). J. Exp. Mar. Biol. Ecol., 58(1): 127-134. Watson W., and J. M. Leis, 1974. Ichthyoplankton in Kaneohe Bay, Hawaii. A one-year study of fish eggs and larvae. Univ. Hawaii Sea Grant Tech. Rept. TR-75-01.  142 Watson, C, R. Bourke, and R. Brill, 1987. A comprehensive theory on the etiology of burnt tuna, (abstr.) 39th Tuna Conference, May 1987, Lake Arrowhead, CA. Webb, P. W. and D. Weihs, 1986. Functional locomotor morphology of early life history stages of fishes. Trans. Amer. Fish. Soc. 115:115-127. White, T. F., 1982. Indo-Pacific tuna development and management programme. F.A.O./U.N.D.P. South China Sea Fisheries Development and Coordinating Programme. Working Paper: 114. Wild, A. and T. J. Foreman, 1980. The relationship between otolith increments and time for yellowfin and skipjack tunas marked with tetracycline. Inter-Am. Trop. Tuna Comm., Bull. 17(7):507-560. Wild, A., 1986. Growth of yellowfin tuna, Thunnus albacares, in the eastern Pacific Ocean based on otolith increments. Inter-Am. Trop. Tuna Comm., Bull. 18(6):421-482. Wilkinson, L, 1988.  SYSTAT: The System for Statistics. SYSTAT, Inc.: Evanston, IL.  Wourms, U. P., and O. Bayne, 1973. Development of the viviparous brotulid fish, ilucoeteoides. Copeia 1973:32-40.  Dinematichthys  Yabuta, Y., M. Yukinawa, and Y. Warashina, 1960. Growth and age of yellowfin tuna II. Age determination (scale method). Rep. Nankai Reg. Fish. Res. Lab., 12:63-74. Yabuta, Y. and M. Yukinawa, 1959. Growth and age of the yellowfin tuna (Neothunnus macropterus) in the equatorial Pacific. Study of length frequency distributions I. Rep. Nankai Reg. Fish. Res. Lab. 17:111-120. Yesaki, M., 1983. Observations on the biology of yellowfin (Thunnus albacares) and skipjack (Katsuwonus pelamis) tunas in Philippine waters. F.A.O./U.N.D.P. Indo-Pacific tuna development and management programme. Working paper:7. Yesaki, 1982. Illustrated key to small and/or immature species of tuna and bonftos of the southeast Asian region. F.A.O./U.N.D.P. South China Sea Fisheries Development and Coordinating Programme. Working Paper:4. Zweifel, J. R. and R. Lasker, 1983. Prehatch and posthatch growth of fishes - a general model. Fish. Bull., U.S. 74:609-621.  Appendix 1  Yellowfin measurement data, mean increment counts, and standard deviation.  forklength centimeters  weight grams  sample number  mean increment count  standard deviation  15.2 15.4 16.4 16.4 16.5 17.6 18.0 18.1 18.2 18.4 19.1 19.1 19.1 19.2 19.3 19.4 19.4 19.4 19.6 19.7 19.7 19.9 20.2 20.4 20.4 20.4 20.4 20.5 20.6 20.6 20.6 20.7 20.8 20.8 20.8 20.8 20.8 20.9 21.0 21.0 21.0 21.1 21.2 21.3 21.4  50.5 59.8 70.5 72.4 72.2 83.1 96.8 100.7 107.4 109.2 116.7 119.4 117.5 118.8 123.3 125.0 125.2 118.0 134.8 130.4 125.1 131.3 141.7 138.8 141.0 149.9 150.2 150.8 142.5 148.8 148.4 148.2 157.2 149.1 174.4 153.0 139.0 159.6 161.7 158.0 149.6 163.2 145.3 163.4 174.6  474 467 464 457 471 463 462 114 102 454 469 51 529 112 110 456 12 511 103 59 16 461 517 204 536 53 104 106 542 166 56 518 447 455 241 173 23 514 115 449 451 539 203 13 251  36.5 39.0 39.5 39.0 39.2 46.3 45.5 51.0 50.0 48.8 56.0 52.0 52.0 59.3 58.7 57.3 53.5 49.0 63.0 56.5 52.0 51.5 57.5 60.5 60.2 60.2 66.5 63.2 56.7 64.5 56.5 60.0 57.5 61.0 70.0 66.0 58.0 61.3 60.6 60.3 58.3 59.5 70.5 60.0 67.3  0.6 0.0 0.5 1.4 0.5 0.5 1.0 1.4 2.3 1.2 0.8 0.8 1.9 2.6 0.9 0.9 1.5 0.8 1.8 1.2 1.2 1.9 2.3 1.2 0.9 0.5 1.9 2.2 0.9 2.3 1.2 1.1 1.0 1.1 1.8 1.1 1.4 1.5 1.2 1.5 0.5 2.4 1.5 0.8 0.9  143 Appendix 1  continued  forklength centimeters  weight grams  sample number  mean increment count  standard deviation  21.4 21.5 21.5 21.6 21.7 21.8 21.8 21.9 21.9 21.9 22.2 22.3 22.3 22.3 22.6 23.3 23.4 23.8 23.8 23.9 24.1 24.4 24.5 24.5 24.5 25.1 25.3 25.4 25.4 25.6 25.8 25.8 25.9 26.0 26.1 26.5 27.0 27.2 27.2 27.4 27.4 27.5 27.7 28.2 28.3 28.5 28.7  165.5 167.8 166.6 176.4 184.5 171.7 167.4 174.9 174.9 182.0 197.0 188.8 200.3 187.8 202.8 238.8 236.0 227.8 240.0 236.8 260.2 266.8 283.5 278.1 263.5 280.0 289.6 288.2 308.9 331.2 294.7 325.7 310.2 303.0 312.5 360.3 338.3 341.8 377.3 360.4 374.0 331.4 386.3 362.4 373.8 448.3 406.6  52 541 480 100 538 522 254 194 194 197 105 243 244 107 247 490 489 116 113 123 448 483 101 484 119 160 158 174 488 226 206 481 186 162 111 485 475 216 446 129 250 169 255 210 18 31 171  58.3 65.7 60.3 61.5 59.2 62.2 72.8 65.3 70.7 66.5 71.5 71.5 72.2 68.7 77.0 74.0 64.8 65.8 69.5 77.0 69.0 69.3 75.2 73.5 74.0 79.3 75.3 78.0 74.8 78.8 77.7 81.5 83.0 76.0 79.0 80.5 83.5 80.7 82.5 79.3 82.0 78.8 78.3 91.8 83.5 81.5 86.7  0.9 0.9 1.5 0.9 0.9 0.9 0.9 1.2 1.8 1.9 2.5 2.3 1.7 3.5 0.8 1.1 0.9 0.9 1.2 1.6 2.9 2.6 2.2 1.2 2.2 1.2 1.8 1.4 0.9 0.9 1.5 1.2 1.8 1.6 0.8 1.7 1.2 0.9 1.7 1.9 0.8 2.7 1.8 1.7 2.3 1.2 2.7  144 Appendix 1  continued  forklength centimeters  weight grams  sample number  mean increment count  standard deviation  29.7 29.8 29.8 29.8 30.0 30.0 30.0 30.2 30.3 30.6 31.0 31.0 31.3 31.3 32.5 33.0 33.1 33.2 33.2 33.4 33.5 33.8 34.0 34.7 34.7 35.1 35.7 35.8 36.0 36.2 36.4 36.5 36.6 36.6 36.8 37.2 37.5 37.7 37.8 37.9 38.0 38.1 38.1 38.2 38.7 39.0 39.1  453.5 465.6 550.4 550.1 460.4 521.5 478.7 536.2 493.4 593.8 567.0 517.8 528.8 574.4 655.8 669.5 668.3 716.3 669.7 665.9 684.8 738.5 705.5 737.5 761.3 761.7 805.4 920.0 851.9 838.4 850.0 860.0 888.2 957.5 922.5 988.7 949.0 969.3 1010.0 1054.7 975.2 998.2 1050.0 1074.0 1103.7 1180.0 1128.3  164 5 35 , 32 20 3 28 65 63 33 118 126 128 4 2 7 62 120 6 136 124 181 81 91 150 151 89 440 77 78 443 138 79 45 95 143 156 97 398 172 76 83 154 86 134 145 139  91.2 88.7 98.5 94.7 95.2 89.2 95.5 97.2 96.0 98.5 104.2 98.5 103.5 99.0 99.0 113.3 108.2 122.0 112.5 110.2 110.2 108.7 117.0 117.2. 140.2 146.0 152.7 135.2 128.7 122.2 163.2 152.2 163.0 141.0 141.7 160.0 145.2 151.7 179.0 182.2 146.2 153.7 168.2 179.0 143.0 201.2 175.0  1.7 1.7 2.3 1.8 0.9 1.2 3.3 3.3 2.9 2.3 0.5 2.3 1.7 1.4 3.1 3.2 2.5 3.5 1.9 3.5 2.8 4.8 3.1 3.5 2.9 7.7 3.3 0.9 3.3 3.5 8.2 9.2 6.6 5.5 7.5 4.2 6.3 2.3 5.3 13.0 5.0 5.3 3.0 9.8 2.5 7.9 8.1  145  Appendix 1  continued  forklength centimeters  weight grams  sample number  mean increment count  standard deviation  39.5 39.6 39.7 39.8 40.0 40.3 40.9 40.9 41.1 41.5 42.0 42.0 42.4 42.8 43.0 43.3 43.8 44.6 45.2 45.2 45.5 45.6 46.1 46.5 47.0 47.7 48.0 48.8 49.9 50.2 50.3 51.0 51.1 52.0 52.1 52.2 52.2 52.6 53.1 53.2 53.5 53.6 54.2 54.3 54.3 54.7 54.8  1158.0 1185.3 1086.5 1235.0 1265.5 1246.4 1350.0 1300.0 1322.6 1288.5 1458.3 1560.0 1449.4 1423.5 1490.3 1560.0 1564.4 1764.9 1929.4 1739.7 1700.0 1682.5 1780.0 1977.0 1924.9 1990.0 2000.6 2023.5 2436.7 2450.0 2581.5 2535.8 2293.3 2550.0 2747.0 2620.0 2960.0 2630.1 2590.8 2780.0 2850.0 2759.9 3000.0 3100.0 2750.0 3070.0 3200.0  96 209 211 239 140 311 397 367 148 152 316 391 43 137 302 381 144 153 279 142 378 318 269 263 235 411 281 280 272 437 308 335 295 416 298 435 394 277 300 392 360 348 429 419 383 373 404  162.5 188.5 187.5 177.5 203.2 168.5 212.2 215.2 199.0 173.5 194.0 207.5 198.2 216.2 185.0 24102 236.0 252.2 203.0 222.2 273.0 271.5 233.5 247.2 276.0 290.5 245.5 233.2 271.2 299.7 290.5 325.0 285.0 292.5 314.5 290.5 350.7 285.5 277.0 326.0 320.0 314.7 346.5 304.7 303.5 350.7 339.2  3.4 8.6 7.7 1.7 5.3 7.5 13.7 4.0 13.2 6.8 5.0 9.5 8.4 5.8 5.7 8.5 7.7 13.7 5.6 4.5 4.8 8.3 5.0 6.5 16.1 6.5 4.0 7.8 8.5 11.4 7.9 8.6 4.2 5.8 11.9 9.9 7.0 10.1 5.7 11.1 12.8 8.0 10.2 11.8 1.0 2.2 13.0  146 Appendix 1  continued  forklength centimeters  weight grams  sample number  mean increment count  standard deviation  55.0 55.0 55.3 55.4 55.7 55.7 56.0 56.5 56.5 57.0 57.2 57.9 58.1 58.2 59.8 65.1 66.5 70.0 70.3 75.0 79.0  3170.0 2800.0 3305.5 3290.0 3040.0 3590.0 3260.0 3400.0 3380.0 3650.0 3730.0 3580.0 3460.3 3560.0 3789.9 4428.4 6500.0 6370.0 6560.0 8850.0 9200.0  355 385 133 436 345 431 432 441 380 396 405 403 498 342 504 350 421 388 341 390 386  329.5 341.3 329.5 358.7 344.5 376.5 354.7 348.5 308.7 381.2 317.2 352.0 387.2 393.2 405.2 162.2 446.5 446.7 472.5 520.0 594.0  9.7 4.8 7.5 12.3 12.7 9.9 10.5 7.5 7.4 13.7 13.0 6.1 18.2 8.1 8.0 13.3 8.5 8.0 12.3 16.7 12.2  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0097612/manifest

Comment

Related Items