Citation
Accuracy and Precision of Fin-Ray Aging for Gag (Mycteroperca microlepis)

Material Information

Title:
Accuracy and Precision of Fin-Ray Aging for Gag (Mycteroperca microlepis)
Creator:
DEBICELLA, JACLYN M.
Copyright Date:
2008

Subjects

Subjects / Keywords:
Age structure ( jstor )
Animal fins ( jstor )
Estimate reliability ( jstor )
Fish ( jstor )
Fish scales ( jstor )
Fisheries ( jstor )
Fisheries management ( jstor )
Grouper ( jstor )
Otolith organs ( jstor )
Symmetry ( jstor )
Gulf of Mexico ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Jaclyn M. Debicella. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2006
Resource Identifier:
436098789 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

ACCURACY AND PRECISION OF FIN-RAY AGI NG FOR GAG ( MYCTEROPERCA MICROLEPIS ) By JACLYN M. DEBICELLA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Jaclyn M. Debicella

PAGE 3

This thesis is dedicated to my parents, Cal and Maggie Debicella.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my advisors Dr. De bra Murie and Dr. Daryl Parkyn for their help, humor, and support. I have thoroughly en joyed working with them and am grateful to have had the opportunity to collaborate with them on this project. I also thank the other members of my committee including Dr. Mike Allen and Dr. Rich McBride, who provided valuable insight and a dvice throughout this process. Special thanks go to Northwest Seafood of Gainesville and the fishing guide boats of Steinhatchee for providing many of my grouper samples. Also, thanks go to Lew Bullock, Dr. Chris Koenig and Dr. Felicia Co leman for providing hard-to-get samples. I am very grateful to everyone who provide d field and lab assistance for this project including Liz Berens, Mark Butler, Julie Harris, Rick Kline, Steve Larsen, Eddie Leonard, Doug Marcinek, and Pat O’Day. I thank my family and friends for their continuing encouragement, patience, and humor, which has helped me th roughout the past two years. I am especially thankful to my parents, Maggie and Cal Debicella, for al ways supporting me and encouraging me in every endeavor. I also thank my brothers a nd sister, Dan, Joe and Di anne, for their love and support and Dana Shabica for her friendship. I would also like to express my thanks to Eddie Leonard for his friends hip, love, and sense of humor.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION........................................................................................................1 2 METHODS.................................................................................................................10 Fish Collection............................................................................................................10 Aging Structures.........................................................................................................11 Sagittal Otoliths...................................................................................................11 Fin Rays...............................................................................................................11 Aging Criteria.............................................................................................................13 Accuracy and Age-Class Assignment........................................................................14 Precision of Age Estimates.........................................................................................15 Comparative Age Estimates Using Fin Rays versus Otoliths ....................................16 3 RESULTS...................................................................................................................21 Fish Collection............................................................................................................21 Aging Structures.........................................................................................................21 Sagittal Otoliths...................................................................................................21 Fin Rays...............................................................................................................21 Aging Criteria.............................................................................................................23 Accuracy and Age-Class Assignment........................................................................24 Precision of Age Estimates.........................................................................................26 Comparative Age Estimates Using Fin Rays versus Otoliths ....................................28 4 DISCUSSION.............................................................................................................45 Aging Gag Using Dorsal Fin-Rays.............................................................................45 Challenges with the Fin-Ray Method.........................................................................48

PAGE 6

vi Annulus Formation in Otoliths and Fin Rays.............................................................51 Implications of a Nonlethal Aging Method for Gag...................................................56 LIST OF REFERENCES...................................................................................................59 BIOGRAPHICAL SKETCH.............................................................................................65

PAGE 7

vii LIST OF TABLES Table page 1. Measures of precision within-reader , between-readers, and between aging structures for otoliths and fin rays , where CV=coefficient of variation, APE=average percent error, rc=Lin’s concordance correlation, and n=sample size........................................................................................................................... .30 2. Percent agreement between ages estima ted from otoliths and dorsal fin rays relative to the otol ith-assigned age for Reader 1 (n=453)........................................31 3. Catch-at-age differences in gag aged using either otoliths or fin rays. Proportion-at-age is estimated from hookand-line caught gag in the Gulf of Mexico......................................................................................................................32

PAGE 8

viii LIST OF FIGURES Figure page 1. Mean annular diameter of the first and second annulus was taken by measuring perpendicular to the axis of the inner gr oove of the base of a dorsal fin ray of gag. Translucent zones for fin rays are enumerated, labeled here as 1, 2, and 3....18 2. Cross-section of a sagittal otolith from a: A) juvenile (3-yr old) gag showing dorsal and ventral areas of the otolith re lative to the sulcus; and B) older (7-yr old) gag showing annuli on the dorsal area to be clearer than on the ventral area. Marginal increment was therefore measur ed on the dorsal por tion (adjacent to the arrow) for all otoliths..........................................................................................19 3. Position of an annulus on the edge in : A) a gag otolith where the annulus is visible along the outer edge of the otolith and near the sulcus; and B) a dorsal fin ray of gag where the annulus is visi ble on the fast-growing, inner area of fin ray and part of the slow-growing, outer portion of the ray. Fish in A) and B) are not the same...................................................................................................20 4. Length frequency for all gag used in the present study. Samples where TL was unavailable were excluded (n=24)...........................................................................33 5. Annuli visible in basal versus distal sec tions of fin ray from an individual gag: A) basal section from just above pter ygiophore on a 5-yr old fish; B) section from a ray clipped level with the back of the gag; and C) section from ~ 2.4 mm distal on fin ray of the same fish, wh ere first annulus is present but more difficult to identify...................................................................................................34 6. Mean ( SE) diameter of the first and second annuli of fin rays of gag for each age class determined by resolved otolith age...........................................................35 7. Picture of a gag dorsal fin ray under A) br ight field (Khler il lumination) and B) the same fin ray under phase-contrast illumination. When viewed under phase contrast it was easier to determine if the annulus was on the edge or whether there was any opaque growth beyond the ultimate annulus.....................................36 8. Cross-section of dorsal fin rays of gag: A) showing how the first annulus can be partially occluded by blood vessels that r un through the center of the fin ray (3yr old fish); B) showing distinctly split rings for 2nd annuli. This 3-yr old fish could be mistakenly aged as a 5 or 6 yr old fish; and C) s howing annuli splitting into two distinct rings on ob lique inner portion of fin ray.......................................37

PAGE 9

ix 9. Cross-section of a dorsal fin ray of a gag showing: A) annuli compacting on the outer edge of the fin ray in the 5th and 6th years, whereas annuli are still evenly spaced near inner, oblique portions of the fin ray; and B) an 11-yr old gag with very clear, evenly spaced annuli on the i nner, oblique edge. Annuli are slightly more difficult to distinguish on the outer portion of the ray....................................38 10. Mean marginal translucent growth increment ( 1 SE) of age-4 gag otoliths over a 10-month period (April to January).......................................................................39 11. A) Qualitative marginal-increment anal ysis of the amount of translucent zone growth beyond the ultimate opa que zone for gag aged 2-5 yrs; and B) qualitative amount of opaque growth observed on the edge of dorsal fin rays of gag aged 25 yrs collected from January-December...................................................................40 12. Scatter plot comparison of the age es timate obtained using fin rays versus otoliths for gag. The diagonal line repr esents comparisons where otolith ages = fin ray ages. Circle size represents the relative sample size for that age combination..............................................................................................................41 13. Scatter plot comparison of the age estim ate obtained by Reader 1 versus Reader 2 using A) otoliths and B) fin rays for gag. The diagonal lines represent comparisons where age estimated by Reader 1 = age estimated by Reader 2. Circle size represents th e relative sample size for that age combination.................42 14. Scatter plot comparison of the age es timate obtained using otoliths versus fin rays for A) Reader 1 and B) Reader 2. The diagonal lines represent comparisons where age estimated for otoliths = age estimated for fin ray. Circle size represents th e relative sample size for that age combination.................43 15. Derived age structure for gag grouper fr om otolith-assigned ages and fin rayassigned ages. The number of samples in otolith age classes of 1 to 10+ were chosen according to the proportions of the age structure given in Fitzhugh et al. (2003)..............................................................................................................44

PAGE 10

x Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ACCURACY AND PRECISION OF FIN-RAY AGI NG FOR GAG ( MYCTEROPERCA MICROLEPIS ) By Jaclyn M. Debicella August 2005 Chair: Debra J. Murie Cochair: Daryl C. Parkyn Major Department: Fisher ies and Aquatic Sciences The ability to age fish accurately is an e ssential first step for assessment of a fish population and successful resource management. The most accepted method of aging long-lived marine fish uses ot oliths and is lethal. The s acrifice of large numbers of threatened fish or fish in protected areas is counter-intuitive to their conservation and may be avoidable through the use of alternat ive aging methods. The primary goal of this study was to assess the use of the nonlethal fin ray method for aging recreational and commercially important groupers, using gag ( Mycteroperca microlepis ) as a model species. Gag were collected from artificial re efs and live-bottom areas off the west coast of Florida, local fish markets and from fish ing guides. Paired samples of otoliths and dorsal fin rays were obtained from each fi sh and ages were determined using crosssections of each aging structure. Ages based on the conventional method of using otoliths ranged from 0 to 20 years old and fin ray ages ranged from 0 to18 years old.

PAGE 11

xi Within-reader agreement for Readers 1 and 2 was 92% and 95% for otoliths (Average Percent Error or APE=0.6 and 1.0, respectivel y) and 62 and 64% for fin rays (APE=6.8 and 9.3), with agreement of 1 yr of 97-99% and 94-95% for otoliths and fin rays respectively. Between-reader agr eement was 83% for otoliths (95% 1 yr) and 59% for fin rays (90% 1 yr). Percent agreement between otolith-based and fin ray-based ages was 56 to 62% (APE=5.9-6.2), with 87-95% agreement within 1 yr. Qualitative marginal-increment analysis indicated that ot oliths and fin rays completed deposition of the opaque zone (relatively fast growth) at approximately the same time each year (MayAugust). Four main aging inaccuracies were of particular concern when aging gag using fin rays, including an occluded first annul us, identifying false annuli or checks, compaction of annuli on the edge of aging struct ures and placing fish into the appropriate age classes. A test of symmetry for the two methods of aging gag showed an overall systematic bias for aging gag from 0-17 year s old with fin rays, which was statistically attributable to biases in agi ng 3 and 7 year old gag due to problematic false annuli. Age composition of gag (1 through 9, and 10 years of age) based on proportion-at-age in the hook-and-line landings was not sign ificantly different between otolith-based ages versus fin-ray-based ages. A test of symmetry indica ted a systematic bias in the aging that was significant only for 4 year old gag. Again, this bias was most likely due to problems with identifying checks. Further examination of bot h otoliths and fin rays is needed for older gag (>8 years) because otoliths from olde r gag were problematic and aged with less reliability than fin rays. The success in using the fin-ray aging method for gag, as a model grouper species, suggests that there is utility in fu rther developing the method for aging other serranids, such as the th reatened goliath and Nassau groupers.

PAGE 12

1 CHAPTER 1 INTRODUCTION Fisheries managers examine population para meters to determine the status of protected and exploited species. The ability to age fish accurately is an essential first step for age-based assessment of a fish population and successful resource management (Crumpton et al. 1984; Brennan and Cailliet 1989). The ability to construct an accurate age structure of a fish population is cri tical because many current stock assessment models in use for marine fish populations ar e age-structured and de pend on a catch-at-age matrix (e.g., Turner et al. 2001). These models allow for mortality estimates, agespecific growth rates, and age-specific fecund ity and maturity estimates. Reliability of these estimates ultimately influences the out come of stock assessments and thereafter management approaches for a fishery. If fish are not aged accurately, the stock assessment and resultant management decisions may be biased. Fish are generally aged by examining annual va riations in the growth of hard parts, such as scales, spines, fin rays, otoliths and/ or bones, including opercles, clavicles, and cleithra (Chilton and Beamish 1982; Camp ana 2001). The most accepted method of aging long-lived marine fish involves the us e of otoliths. Oto liths are composed of calcium carbonate (calcite and aragonite) embedded in an organic matrix and assist bony fishes in balance and sound perception (Ch ilton and Beamish 1982; Victor and Brothers 1982). Otoliths generally contain the best pe rmanent pattern of fish growth because calcium resorption is not known to occur from the otoliths and some deposition occurs each year (Chilton and Beamish 1982). However, the fish must be sacrificed to obtain

PAGE 13

2 the otoliths from the pars inferior located in the skull. The sacrifice of large numbers of threatened fish species or fishes in pr otected areas is counter-intuitive to their conservation and may possibly be avoided through th e use of alternative aging methods. There are several nonlethal alternative methods for aging fish, including the removal of external hard parts from fish such as scales, soft fin rays, and fin spines. Traditionally, scales have been used to age freshwater fish, as we ll as relatively shortlived (<5-6 years) marine fish (Beamish 1981). Although scales are an effective, nonlethal way to age rapidly growing fish w ith few year classes (Chilton and Beamish 1982), they often underestimate the age of fish with greater life spans (>10 years). This is a result of the annuli becoming crowded n ear the edge of the scale, which makes it difficult to enumerate the number of annuli in older fish (Beamish and Harvey 1969; Chilton and Beamish 1982; Dutka-Gianelli 1999 ). Another nonlethal method for aging fish is to use their soft fin rays or stiff sp ines, which can be removed and the fish released unharmed (Beamish and Harvey 1969; Mills and Beamish 1980). Fin ray sections have been used to age fishes for more than 60 years (Cass and Beamish 1983), including a variety of freshwater and saltwater species, such as white sucker ( Catostomus commersoni) (Beamish and Harvey 1969), white sturgeon ( Acipenser transmontanus) (Brennan and Cailliet 19 89), lake whitefish ( Coregonus clupeaformis) (Mills and Beamish 1980), lingcod (Ophiodon elongatus) (Beamish and Chilton 1977; Cass and Beamish 1983), walleye pollock ( Theragra chalcogrammma), Pacific cod ( Gadus macrocephalus) and albacore tuna ( Thunnus alalunga) (Beamish 1981), sockeye salmon ( Oncorhynchus nerka) and chum salmon (O. keta) (Bilton and Jenkinson 1969), and

PAGE 14

3 brown trout ( Salmo trutta) (Shirvell 1981). These fish are primarily from cold, temperate regions where annual ring depositions are relatively wide and distinct. The use of fin rays or spines, in partic ular, as nonlethal aging methods has not been thoroughly assessed using warm-temperate fish species, and is especially appealing for species that are threatened or in mari ne protected areas, such as goliath grouper ( Epinephelus itajara ) and Nassau grouper ( Epinephelus striatus ). White grunt ( Haemulon plumieri ), a warm-temperate species in the eastern Gulf of Mexico, has been successfully aged up to 10 years old using fi n rays (Murie and Parkyn 1999). For fish less than 10 years old there was 90% agreemen t between ages from sectioned otoliths and ages from fin rays, while fish 10 years old or greater had their age underestimated using the fin-ray method due to accumulation of a nnuli on the edge of the fin rays, without periods of growth between consecutive annu li (Murie and Parkyn 1999). Gray triggerfish ( Balistes capriscus) , another warm-temperate species from the northeastern Gulf of Mexico, have been aged up to 13 years using th eir single dorsal spine. The spine sections showed distinct, dark bands with agr eement between readers of 98% (Johnson and Saloman 1984). These two examples suggest that other warm-temperate fish species could be aged using fin rays or spines. Th e general applicability of the method therefore warrants further examination for warm water species, especially for species in which there is a conflict betw een researchers using lethal met hods to age fish and user groups concerned with conservation of these same fish populations. Chilton and Beamish (1982) have found that the advantage of using fin rays for age determination is that it is often possible to age older fish more accurately than if scales were used, and the fish still does not have to be sacrificed. The problem of the

PAGE 15

4 annuli becoming crowded and indi scernible in older, slow-gro wing fish when aging using scales (Beamish and Harvey 1969) can also be a problem when aging fi sh using fin rays. However, the age at which annuli begin to accumulate on the edge of the fin ray is species-specific and tends to occur at an olde r age than with scales (Chilton and Beamish 1982). Annuli accumulation on the edge of fin rays may therefore occur at an age close to or beyond that necessary for determining agespecific population parameters or a stock assessment for a particular species. In addi tion, the interpretation of the first annulus and identifying false annuli (“checks” ) can be difficult in fin-ray aging. Fins can also be more difficult and time consuming to handle, st ore, and prepare than scales or otoliths (Beamish 1981; Chilton and Beamish 1982). It is also important to examine different fins when aging a species for the first time to determine which fin provides the clear est pattern of growt h, as this can differ from species to species. For example, walle ye pollock are aged using their pectoral fin rays, whereas dorsal fin rays are used fo r lingcod (Beamish and Chilton 1977; Beamish 1981). In addition, it is practical to also de termine at what location and level on the body of the fish to remove the fin ray so that the first annulus is retained and minimal harm is imposed on the fish. This is important because Mills and Beamish (1980) found that small whitefish regenerated a deformed fin ra y after being clipped at the base of the pelvic fin, while fish larger than 22 cm did not regenerate rays at all. It may not be necessary to excise fin rays at their extrem e base, which is located within the epiaxial musculature, as long as the first annulus is retained and remains recognizable at the level that the fin ray is cut. Beamish (1981) found th at the best sections of fin rays for walleye pollock, Pacific cod, and albacore tuna cam e from 1.0-1.5 cm above the point where the

PAGE 16

5 ray base starts to flare. However, Bilton and Jenkinson (1969) noted when aging sockeye salmon with fin rays that the innermost ring on the section of the fin ray nearest the base was not present on sections from further out on the ray. Whenever a fish is aged, it is vital that the method used be validated across all age classes for a particular species. Validation pr ovides evidence that a technique is accurate and that the estimated ages ar e therefore as close to the tr ue value (age) as possible. Beamish (1981) and Campana (2001) recogniz e several techniques commonly used to validate an aging method. Several of these t echniques have been used to validate annual deposition in otoliths, including: marginal-i ncrement analysis, where changes in the width of the growth increment at the edge of the aging structure is measured over the course of a year; chemical marking of fish that are later recapt ured; and by comparing ages determined from different structures, su ch as otoliths, fin ra ys, and scales, to an aging structure that has been prev iously validated (Beamish 1981). Besides validating or determining the accu racy of using fin rays to age warmtemperate fishes, it is also essential to evaluate the precision or reliability of the method. There are two potential aging inaccuracies that are of particular c oncern when using fin rays. First, identifying the first annulus can be problematic and can result in systematic under-aging by one year (MacFarlane and King 2001) . This may not be critical for longlived species, such as many rockfishes ( Sebastes spp.) that live beyond 50 yrs (Leaman and Beamish 1984), but can have serious im pacts on population parameters and stock assessments for shortor moderately-lived sp ecies where there might be only 4 or 5 year classes that dominate the fishery (MacF arlane and King 2001). Second, and most importantly, ages based on fin rays may undere stimate the true age of older fish. The

PAGE 17

6 threshold age at which compaction of annuli and therefore under-aging occurs is speciesspecific and must therefore be estimated fo r each species of interest. Under-aging the older component of a population can result in a skewed age distribut ion, which could lead to overestimation of mortality (i.e., using a ca tch curve), as well as biased estimates of age-specific population parameters, such as growth, fecundity, and maturity. There are many warm-temperate fish gr oups in which a nonlethal aging method would be an asset, including the Serran idae or grouper family. Various groupers comprise important recreational and commer cial fisheries in Florida. Red grouper ( Epinephelus morio) make up approximately two thirds of the overall commercial and recreational grouper landings (Goodyear and Schirripa 1993), while the remainder of grouper landings is equally divided between gag ( Mycteroperca microlepis ) and all other grouper species (Shirripa and Goodyear 1994). Total landings of gag in Florida were 3.5 metric tons (7 million pounds) in 2001. Of the gag harvested in 2001, 89% were landed on the Gulf coast and 56% of those landi ngs were made by recreational anglers ( www.floridamarine.org ). Due to previous overexploita tion (Turner et al. 2001), gag are regulated by annual quotas, minimum size limits, and daily bag limits (Gulf of Mexico Fishery Management Council 2001). Members of the genera Epinephelus and Mycteroperca are particularly vulnerable to the threat of overfishing (M orris et al. 2000). The impact s of overfishing are amplified in these genera due to the tendency of many species to form predictable spawning aggregations. In particular , spawning aggregations of goliath grouper, Nassau grouper and gag have been heavily exploited. Due to low abundance, goliath grouper and Nassau grouper have been prohibited to harves t in Florida waters since 1985 and 1990,

PAGE 18

7 respectively (NMFS 1990a,b; Gulf of Mexico Fishery Management Council 2001). Although gag remain open to recreational and commercial fisheries in Florida, their harvest is heavily regulated as part of an annual quota, as well as by size and bag limits (Gulf of Mexico Fishery Management Counc il 2001). Beginning in the 1980’s, the average size of gag in the Gu lf of Mexico shifted toward smaller fish and the proportion of males decreased (from 17% to 1% in the last 20 years) in the population (Coleman et al. 1996). Since gag are protogynous, a severe reduction in large ma les in the population has the potential to limit re production. In addition, as fe males age they spawn more frequently and their fecundity increases with weight and leng th (Collins et al. 1996). Because of concerns that the removal of larger, older females and males has limited reproduction and subsequent juvenile recruitmen t, several regulations have been recently implemented for gag including harvest prohibi tion during the main months of their spawning season (March-April). In addition, marine protected areas have been set aside in the main gag spawning area to both protec t spawning aggregations of gag as well as other reef fishes (Gulf of Mexi co Fishery Management Council 2001). Thus, gag are a suitable grouper species in which to attempt to apply a nonlethal aging method not only from a management persp ective, but they are also suitable as a model species for other protected, moderately -lived, warm-temperate reef species, such as Nassau and goliath grouper, that occur in fa r lower abundance. In particular, gag are a moderately-aged grouper species, with a maximum age variously estimated as 22 years by Collins et al. (1987) and Harris and Co llins (2000) and 28 years by Fitzhugh et al. (2003). Maximum ages of Nassau grouper a nd goliath grouper are either younger (16 years for Nassau grouper, Olsen and La Place 1979) or in the same approximate age

PAGE 19

8 range (37 years for goliath grouper, although 90% of goliath grouper were <24 years old, Bullock et al. 1992). Gag are also advant ageous as a model species because ages estimated using otoliths have been validat ed using a variety of methods, including marginal-increment analysis (Hood and Shlieder 1992), chemical marking with oxytetracycline (Collins et al. 1996), and mark-recapture growth analysis (S chirripa and Burns 1997). This makes it possible to use otol ith ages as an indirect validation of finray ages if necessary. In addition, to asse ss the potential under-aging bias of fin-ray aging, it is necessary to sample relatively older groupers. Whereas this may still be problematic for gag, it is at le ast potentially feasible give n that fisheries are ongoing for the species, in contrast to Nassau and goliath groupers wher e the fisheries are closed. Developing nonlethal methods of fin-ray agi ng using gag as a model grouper species may therefore be more efficient and, ultimately, of direct benefit to ongoing recovery efforts of Nassau and goliath groupers if the methods and interpretation are tr ansferable to these protected species. Therefore, the primary goal of the present study was to assess the use of the fin-ray method for aging a warm-temperate moderate ly-lived grouper, using gag as a model species. The specific objectives were to: 1) ex amine different fins to determine which has the clearest pattern of altern ating opaque and translucent z ones necessary for aging gag and estimate precision for both otoliths and fin rays; 2) test the accuracy of the fin-ray method for gag by comparing the pattern and tim ing of the annual deposition in fin rays and otoliths, of which the latter has been prev iously validated; 3) compare fin-ray ages to ages obtained using otoliths, of which the latter is considered to be the best or least biased structure in aging fishes; and 4) determine if any resulting biases or error associated with

PAGE 20

9 the fin-ray method of aging would result in a significantly different age composition of gag compared to an otolith-based age composition. This final objective was necessary to determine if fin-ray aging is an appropriate method that could potentially be used to age threatened groupers or groupers residing in mari ne protected areas, eith er for direct stock assessment purposes or to assess the recovery in a truncated age stru cture that is typical of heavily or overexploited fish stocks.

PAGE 21

10 CHAPTER 2 METHODS Fish Collection Gag samples were collected from a variety of sources, including: 1) fish traps, spear fishing, and hook and line from artificial reefs and live bottom areas at ~15 m depth off the west coast of Florida; 2) traps and hook and line from other, unknown offshore areas of the west coast of Florida (pers. comm.; Mr. Lew Bullock, Florida Marine Research Institute; Dr. Gary Fitzhugh, Nationa l Marine Fisheries Service, Panama City, FL; and Dr. Chris Koenig, Florida State Univer sity); 3) local fish markets, which land fish from various depths and areas in the northeastern Gulf of Me xico (e.g., Northwest Seafood, Gainesville, FL); and 4) fishing guide boats landing catch in Steinhatchee, FL. All whole fish were measured for maximum to tal length (TL) and weighed, with filleted gag measured only for TL. Sagittal otoliths were removed from each fish, rinsed in distilled water, and stored dr y until processed. Dorsal fin rays were collected from all fish, while dorsal fin spines and anal and pe lvic fin rays collected from a subsample of fish. By sampling only the mid-central rays of the fin it was easier to avoid frayed or damaged rays or rays that had grown and fused together (Chilton and Beamish 1982). Dorsal fin rays 3-7 and dorsal spines II-V were clipped at the base of the pterygiophores. Excess tissue was trimmed away with scissors and then fins were stored in a coin envelope with fin rays lying fl at and parallel to one anothe r with the cut surface exposed. Fins were air dried for approximate ly 2-5 days or stored frozen.

PAGE 22

11 Aging Structures Sagittal Otoliths In preparation for sectioning, the left otolit h (the right otolith wa s used if the left was unavailable) was heat annealed to a froste d slide with hot glue. The mounted otolith was then thin-sectioned using a Buehler Is omet 1000 digital sectioning saw (Buehler, IL) with a diamond wafering blade (7.6 cm diameter X 0.15 mm blade width). Sections 0.5 mm thick were taken through the nucleus of the otolith along a transverse plane. Otolith sections were rinsed in ethyl al cohol and then permanently mounted on glass slides with Histomount (National Diagnos tics, GA). Sections were viewed under transmitted light (25-45x power) using a ster eomicroscope. Under transmitted light, a combination of one dark opaque zone (or optic ally more dense band) with one translucent (clear light transmitting) zone was interprete d as a complete annulus, with the narrow opaque zone counted to determine the age (Col lins et al. 1987; Hood and Schlieder 1992; Harris and Collins 2000). Fin Rays For final processing, dried fins were boile d in distilled water for approximately 10 minutes and the skin removed from the base of the fin rays using a soft-bristled brush. Fins that were stored frozen were thaw ed and then submerged in boiling water for approximately 1 minute. Skin was then peeled away from the base of the fin rays using forceps. Cleaned fin rays were then air-dried after which they were coated with Hysol epoxy adhesive (Loctite Corporation, Rocky Hill, CT) and allowed to harden for at least 48 hours. Epoxied fin rays were cross-sec tioned using a Buehler Isomet 1000 saw with a diamond wafering blade (15.2 cm diameter X 0.5 mm blade width). Sections were cut to ~ 0.8 mm thick and were permanently mounted on glass slides using Flotexx. Fin-

PAGE 23

12 ray sections were viewed with transmitted light (4-10x power) using a compound microscope (Chilton and Beamish 1982). A narrowband green filter (wavelength 550 nm) was used to enhance contrast. In fin-ray sections, each annulus al so appears as a pair of alternating translucent a nd opaque zones. However, in fin rays it was the narrow translucent zone that was enumerated as per Chilton and Beamish (1982). A subsample of dorsal fin rays from 9 gag ranging from age 0 to 5 was used to assess changes in the appearance and presence of annuli from proximal to distal sections. One of the 4-5 rays in each of these sample s was cut where the fin ray stands free of the back of the fish to serve as a marker for the external portion of the rest of the fin rays. When the sample was aged it was possible to compare the ages of the rays below and at the level of the cut ray to examine whether any annuli (especially the first, which was laid down when the fish was smallest) were missing in the sections taken at the level where the fin ray stands free of the back of the fish. Rays from different fins (dorsal, anal, a nd pelvic) and dorsal fin spines were also assessed for clarity of annuli. Sections from the various struct ures were taken at variable widths (0.6-1.1 mm) to determine the best section width for each structure. The structures were comparatively aged to dete rmine which provided the most reliable, clear pattern of annuli for aging. In addition, a subsample of 138 fin rays fr om age classes 1-7 were examined for the mean diameter of the first and second annul us. Diameters of the first two annuli were measured perpendicular to the axis of the inner groove of the fin ray according to the method of Chilton and Beamish (1982) (Figur e 1) using a Motic M62 capture board and Motic image processing system (Versi on 1.3; Motic, Inc., Richmond B.C.).

PAGE 24

13 Measurements were always taken from the sec tion closest to the base of the fin ray for consistency. These measurements were take n to assess the possibility of a missed first annulus in older age classes. A single factor analysis of variance (ANOVA) was used to test for a significant difference between th e first and second annulus among year classes by using the difference between the two annular measurements (Penha et al. 2004). Aging Criteria Otoliths were aged according to criteria set forth by Collins et al. (1987) and Hood and Schlieder (1992). Aging criteria for gag fin rays was developed from general fin-ray aging criteria outlined in Chilton and Beamish (1982), with modifications to distinguish what an annulus looks like in a fin ray comp ared to an otolith. Aging criteria were initially developed based on an examination of a subsample of 50 fish previously aged by otoliths; these fish represented the complete size range of fish collected. A second attempt to age fin rays blind (without a prio ri knowledge of the otolith-based age) was then made. Aging criteria were subsequently adjusted to consiste ntly distinguish where the first annulus was located in fin rays, as well as which marks were true annuli versus checks or false annuli. A photographic referenc e was made for paired otolith and fin-ray sections. This reference guide included photographs of some co mmon difficulties in aging fin rays. The reference guide and a pr actice set of otoliths and fin rays were reviewed by both readers prior to aging. Fin rays were al ways aged before otoliths, since the otolith age was considered the best age estimate based on validation of annual ring deposition by marginal-increment analysis and mark-recapture studies (Collins et al. 1987; Collins et al. 1996; Hood and Shlieder 1992). Precision and accuracy were determined for a primary reader (JMD), as well as a secondary reader (DJM).

PAGE 25

14 Accuracy and Age-Class Assignment Quantitative marginal increment analysis was performed on otoliths to validate annual deposition for the “best” aging structur e. The 4 year old age class was used for this analysis because it was the only age cla ss with an adequate sample size for most months. The marginal increment in each ot olith was measured on the proximal medial surface on an axis dorsal to the sulcus, since th is was the clearest r eadable area of the sectioned otolith for gag. These measurements were made using the image capture system (Figure 2A). In particularly olde r gag, annuli on the dorsa l area of the otolith were more easily distinguishable than on the ventral area (Figure 2B). In this study, otoliths were considered by convention to be the “best” aging structure and were assumed to be 100% correct in estima ting the age of gag in compar ative analyses with fin-ray ages. To validate the accuracy of the fin ray and otolith aging methods, qualitative marginal-increment analysis was performed on both otoliths and fin rays independent of one another. Qualitative marginal-increment analysis on otoliths and fin rays was conducted by assessing if a sample had an opa que or translucent zone on the edge. An opaque zone in otoliths was considered “on the edge” if it appeared near the sulcus (Figure 3A). A translucent zone in fin rays was considered to be “on the edge” if it appeared in both the faster gr owing, oblique area of the fin ray, as well as some part of the slower growing, outer portion of the fin ra y (Figure 3B). Both the timing and number of minima present in the qualitative margin al increment for otoliths were graphically compared to the qualitative marginal increm ent for fin rays to determine whether the deposition of material on the edge of the aging structures occurred in a synchronous or differential manner.

PAGE 26

15 Age classes were assigned for each aging structure using the c onventional birthdate of January 1 (Chilton and Beamish 1982). With establishment of th e aging criteria, all fin rays and otoliths were r ead at least twice by the primary reader (JMD). As the alternative aging method, fin rays were always read first and independently of the paired otolith. All readings were independent of the age assigned in previous readings, and there was a minimum of 2 weeks between r eadings. In addition, no knowledge of the size or collection date of the fish was used when reading the aging structures. When age estimates for the primary reader were in agr eement, then that age was considered to be the resolved age. If the two age estimates disagreed, then the structure was aged again independently of the first two readings. A resolved age was then accepted if two out of three independent readings agr eed. If two out of three ages did not agree, a fourth read was conducted and either the median age or an age that was agreed upon in two out of four readings was used as the resolved age fo r the structure. If a resolved age could not be determined for a structure using thes e criteria then the structure was deemed unreadable. Precision of Age Estimates To determine between-reader precision, a s econdary reader (DJM) estimated ages for a randomized age-stratified subset of 111 fin rays and otoliths. Resolved ages were obtained in the same manner as for the primar y reader and all age estimates were based on the same aging criteria for defining an annulus as provided to the primary reader. Precision (reproducibility) was estimate d using the first two readings by determining percent agreemen t (Sikstrom 1983), coefficient of variation (CV) (Kimura and Lyons 1991), and average percent error (APE ) (Beamish and Fournier 1981). All measures of precision were ca lculated within reader, betw een readers and between aging

PAGE 27

16 structures. Measures of percent agreement, coefficient of variati on, and average percent error are commonly used to assess precision in aging methods and therefore facilitated comparison with other studies. Lower va lues for CV and APE represent greater precision. Between-structure precision wa s also determined by calculating the concordance correlation coefficient (L in 1989, 2000; Murie and Parkyn 2002). The concordance correlation ranges from .0 to 1.0, with 0 representi ng no reproducibility. The concordance correlation has been determin ed to be more robust than the coefficient of variation, paired t-test s, Pearson’s correlation coeffi cient and the least squares approach (Lin 1989; 2000). Comparative Age Estimates Using Fin Rays versus Otoliths The comparison between resolved otolith and fin ray ages was visualized by plotting the fin-ray age as a function of the sectioned otolith age for each fish. A line denoting the ideal 1:1 relationship was used for reference (Beamish and Harvey 1969). Age estimates from otoliths and fin rays were also compared quantitatively using a test of symmetry to determine systematic aging di fferences between stru ctures (Bowker 1948; Hoenig et al. 1995). If a significant differe nce was detected, then the test of symmetry was repeated assuming symmetry for the most extreme cases until no bias could be detected (Hoenig et al. 1995). In addition to direct tests of any diffe rence between the two aging methods, an applied test of the fin-ray method was conduc ted by comparing the pr oportions at age of a gag population using otoliths versus fin rays. Expected proportions were based on the Fitzhugh et al. (2003) hook-and-li ne data from 1997-2000, as th is most closely resembled the time period and gear from which the curr ent studies’ samples were collected (original data obtained from Dr. Gary Fitzhugh, NMFS, Panama City, FL). Gag older than 10

PAGE 28

17 years were pooled together in to a 10+ age group, based on the catch-at-age matrix used in current gag stock assessment (Turner et al. 2001). A random agestratified subsample from all available otoliths was then drawn to match the expected proportions. Fin rays that were paired with the drawn otolith subs ample then formed the basis of the proportion at age based on the fin-ray method of aging. The age compositions (number in each age class) for each aging method were compar ed using a Wilcoxon Signed Rank test to determine if there was any significant differe nce in the estimation of gag age structure between fin rays and otoliths. To detect any systematic bias in fin-ray aging within the estimated age structure, a test of symmetry was also performed on the subsample of fin rays and otoliths. If a significant difference was detected, then the test of symmetry was repeated assuming symmetry for extreme cases until no bias could be detected (Hoenig et al. 1995).

PAGE 29

18 Figure 1. Mean annular diameter of the first and second annulus was taken by measuring perpendicular to the axis of the inner groove of the base of a dorsal fin ray of gag. Translucent zones for fi n rays are enumerated, labeled here as 1, 2, and 3.

PAGE 30

19 ventral dorsal sulcal groove A B Figure 2. Cross-section of a sagittal otolith from a: A) juvenile (3-yr old) gag showing dorsal and ventral areas of the otolith re lative to the sulcus; and B) older (7-yr old) gag showing annuli on the dorsal area to be clearer than on the ventral area. Marginal increment was theref ore measured on th e dorsal portion (adjacent to the arrow) for all otoliths.

PAGE 31

20 B A 1 2 3 4 Figure 3. Position of an annulus on the e dge in: A) a gag otolith where the annulus is visible along the outer edge of the otolith and near the sulcus; and B) a dorsal fin ray of gag where the annulus is visi ble on the fast-growi ng, inner area of fin ray and part of the slow-growing, outer portion of the ray. Fish in A) and B) are not the same.

PAGE 32

21 CHAPTER 3 RESULTS Fish Collection A total of 461 paired fin ray and otolit h samples were collected from gag grouper from the northeastern Gulf of Mexico (365 from fish traps, spear fishing, and hook and line; 96 from local fish markets sampling the commercial fishery). Aging structures were collected from gag ranging in size be tween 257 mm and 1265 mm TL (n= 437 fish) (Figure 4). Paired aging stru ctures from 24 gag were used without any associated size data. Aging Structures Sagittal Otoliths Sectioned sagittal otoliths of gag were clear to read along the medial area dorsal and ventral to the sulcus in younger fish, but on ly along the dorsal portion of the sulcus in relatively older fish (Figure 2A and 2B). Under transmitted light, the nucleus of the otoliths appeared opaque dark brown and op tically dense. Each subsequent annulus consisted of a pair of one narrow dark opaque zone and one broad relatively translucent zone. Narrow opaque zones were enumerated to determine the number of annuli present in the otolith for consistency with other st udies (Collins et al. 1987; Hood and Schlieder 1992; Harris and Collins 2000). Fin Rays Initial observations of the dorsal fin rays and spines, pe lvic fin rays, and anal fin rays indicated that dorsal fin rays and dorsal fin spines had the cleare st sections and were

PAGE 33

22 relatively easier to read, both due to their clarity and size. Pe lvic fin rays were readable but very small in cross-secti on with crowded annuli, and anal fin rays were not readable beyond age 4. Preliminarily, examination of a sm all (n=22) subsample of otoliths, dorsal fin rays and dorsal fin spines indicated that percent agreement between dorsal fin rays and otoliths was slightly great er (71%) than agreement between dorsal fin spines and otoliths (65%) for fish ranging in age from 3-7 years. Dorsal fin rays were used for all further analysis in this study because they were easier or as easy to r ead as fin spines and more fin ray samples were available than fin spine samples. Dorsal fin ray sections removed from wher e the fin ray stands free of the dorsum of the fish provided clear, readable sections, while still retaining all annuli (n=9, ages ranging from 0-5). No annuli were found to be missing in rays removed and sectioned at the base of the dorsal fin versus rays secti oned just above the pter ygiophores (Figure 5A and B). However, as sections proceed distally up the fin ray, above th e level of the dorsal surface, the first year becomes increasingly di fficult to identify (Figure 5C). This is because it becomes less distinct and the sinus of the blood vessel begins to acquire an irregular shape. For a representative subsample of gag aged 2-7 years, mean diameter of the first fin ray annulus averaged 0.56 mm ( 0.01 mm SE, range of 0.35-0.82 mm, n=118 ) and 0.81 mm ( 0.01 mm, range of 0.52-1.12 mm, n=118) for the second fin ray annulus. A single-factor ANOVA detected no significant difference among gag age classes in the difference between the first and second a nnular diameter measurements (F=1.63, n=118, p=0.157) (Figure 6). This indicated that th e first and second annuli did not converge in older fish.

PAGE 34

23 Aging Criteria In general, fin rays were more difficult and time consuming to prepare and interpret than otoliths. In fin rays, annuli that were enumerated appeared as translucent bands on an opaque background (Figure 1). Viewing the e dge of the fin ray was facilitated by use of phase contrast on a compound microscope, which also helped to eliminate checks (Figure 7). Annuli were count ed beginning from the central core of the fin ray, outwards toward the edge. The oblique, inner portions of the fin ray were the fastest growing areas where annuli appeared earlier than the outer edge of the fin ray, which was a relatively slower growing area. Three impediments to aging were of part icular concern: 1) an occluded first annulus; 2) identification of checks or spli t rings; and 3) enumeration of compacted annuli in older fish. In some dorsal fin rays , the first annulus was partially obscured by the core of the ray. If the fin ray core a ppeared irregular (not round), then the first annulus was usually indistinct. In these cases the annulus wa s visible typically at least partially outside the core and appeared as a light, translucent area (Figure 8A). Checks, or false annuli, were present on mo st of the fin ray sections from gag. Often annuli were composed of more than a si ngle ring, with all rings associated with one annulus extending around the entire ray. It wa s therefore important to view all fin ray sections for a fish (~9-15) to determine if certain rings always extended around the entire ray. If the rings merged together or were not complete in the majority of sections the ring was considered a check. The 2nd, 3rd, and 4th annuli in gag fin rays were often split into two very distinct rings for each annulus (Figure 8B). This was determined by comparing fin rays that had these types of checks to the otoliths from the same fish. It was also

PAGE 35

24 difficult to follow an annulus onto the oblique , inner portions of th e ray, as the annulus often appeared to split in this area (Figure 8C). Annuli were relatively widely spaced for th e first few years and then became more closely spaced in fish aged greater than 5, and became compacted ( tightly aggregated) on the outer edge of the ray in fish aged ol der than 6 (Figure 9A). Once annuli began compacting on the edge they appeared as thin but discrete translucent lines rather than broader bands consisting of multiple checks. Annuli appeared to compact more along the slower growing outer edge of the fin ray and they were sometimes easier to differentiate in older fish along the internal, oblique edge. Annuli on the internal edge do not appear to begin compacting before age 10 (Figure 9B). Accuracy and Age-Class Assignment For quantitative marginal-increment analysis of otoliths by age class, sample size throughout most months was only adequate fo r 4-year old gag. The observation of only one minimum in the marginal increment ove r a 12-month period validated the deposition of one annulus (the opaque zone) per year in this age class (Figures 10). Opaque zone completion in otoliths occurred from June-August, at which time the average marginal increment was at a minimum. Qualitative marginal-increment analysis for gag otoliths from 2-5 year old fish showed the same trend as quantitative margin al-increment analysis from the 4 year old gag (Figure 11A). Including multiple age classes in the qualitative marginal-increment analysis was feasible because these otoliths di d not need to be measured, as is the case with quantitative marginal-incre ment analysis, and this allowed an increase in the sample size to n=306. Translucent z ones for otoliths were at a minimum for both quantitative and qualitative analyses from June through August.

PAGE 36

25 Similarly, qualitative marginal-increment an alysis for gag fin rays was done for 2-5 year old gag (n=247), with the difference in sample sizes between qualitative marginalincrement analysis for otoliths and fin rays due to lower numbers of fin rays having a resolved edge. As with otoliths, the translu cent zone in fin rays was at a minimum from June through August, thereby indicating that the opaque zone was being deposited during this time (Figure 11B). Therefore, based on a birthdate of approximately February/March (Hood and Schlieder 1992) and completion of the de position of the opaque zone beginning approximately June 1st, age classes for otoliths were assigned as follows: 1. Fish collected from January through May were advanced one age class if the otolith had a significant amount of tran slucent zone (almost complete based on the previous interannular distance), since the fish would be one age class older within that same year. If th ere was an opaque annulus on the edge during this period, the age class would not be advanced because it would have to be assumed that the annulus had st arted to form early relative to the majority of fish. 2. Fish collected from June through December were kept in the same year class as the number of opaque annuli, unless th e fish were collected between June and August and had a large amount of translucent zone growth on the edge. In this latter case, the age of the fish was advanced by one age class because it would have to be assumed that the fish was either in the process of laying down an opaque annulus or was to depos it an opaque annulus later than an average fish.

PAGE 37

26 Qualitative marginal-increment analysis for fin rays revealed the translucent zone (i.e., the enumerated annulus) to be on the e dge over an extended period of time (Figure 11B). Translucent zones begin to be deposit ed in July and were not complete until the following spring, during April and May. An opaque zone was completed primarily between June and August. The timing of the appearance and duration of the translucent and opaque zones in fin rays indicated that they were being laid down at approximately the same time of the year as the translucent and opaque zones in ot oliths. The difference in interpreting annuli between the two structures was therefore ba sed on enumerating opaque zones for otoliths and translucent zones for fin rays. Therefore, resolved ages for fin rays were assigned as follows: 1. Fish collected from January through May w ith a translucent zone on the edge were kept in the same age class as the number of annuli (number of translucent zones), unless there was significant zone of opaque growth. In this latter case, the age class of the fish was advance one year because it would have to be assumed that these fish were taking longer than averag e to form the translucent zone. 2. Fish collected from June through December with an apparent opaque zone past the last translucent annulus we re kept in the same age class as they were aged. 3. Fish collected from August through Decem ber had one year subtracted from their ages if they had a translucent annulus on the edge, since this would represent the start of the slower growth period for that year continuing into the next year (the relatively faster opaque growth period for the present year had already occurred). Precision of Age Estimates Overall, annuli on otoliths (opaque zones) we re easier to identify and interpret than annuli on fin rays (translucent zones), excep t for older gag. Resolved ages of gag determined using otoliths ranged from 0 to 17 years with 0% of the otoliths illegible for Reader 1 and 1% requiring a four th read. Reader 2 had resolved ages ranging from 0 to 20 yrs with 0% illegible and 1.8% of otoliths requiring a four th read to obtain resolved

PAGE 38

27 ages. Resolved ages of gag determined usi ng dorsal fin rays ranged from 0 to 16 years with 5% illegible by Reader 1 and 3% requiring a fourth read. Reader 2 had resolved fin ray ages ranging from 0 to 18 years with 0.9% deemed unreadable and 9.0% requiring a fourth reading for resolution. Within-reader measures of precision for otol iths and fin rays showed that for both readers, precision was higher with otoliths than with fin rays (Table 1). For Reader 1, 92% of otoliths (APE=0.96) and 62% of fin ra ys (APE=4.85) were consistently read with no difference. However, 99% of otolith readings were within 1 year and 94% of fin ray readings were within 1 year of each other. Within-reader precision was consistent for both Reader 1 and Reader 2 fo r both otoliths (92% and 95%, respectively) and fin rays (62% and 64%, respectively). Scatter plot s comparing the first two otolith readings indicated that within -reader precision for both the primary and secondary reader was much more variable for gag grea ter than 8 years of age than for gag <8 years old (Figures 12A and 12B). Although it was difficult to qua ntify this difference due to the small number of gag sampled that were >8 year s of age, both reader s found it difficult to enumerate opaque annuli along the normal reading axis on the sulcus in otoliths of older gag. Opaque annuli in these older fish were extremely thin, as if drawn by a light pencil line, and had to be counted by moving from one area to another on the sectioned otolith. Qualitatively, Reader 1 observed a large amount of translucent zone material at the edge of the otoliths in these older gag, without any apparent opaque zones, and therefore concluded that she was underaging older gag us ing otoliths. In contrast, scatter plots comparing the first and second fin ray readings for either reader (Figures 12C and 12D) indicated more overall variability in the precis ion of the readings but did not indicate that

PAGE 39

28 this variability increased with gag age. Both readers were more consistent in their aging of older gag using fin rays than using otoliths. Between-reader measures of precision base d on resolved ages were also higher for otoliths than for fin rays (Table 2). Scatter plots of resolved otolit h ages for Reader 1 and Reader 2 indicated that Reader 1 aged gag >8 years old, in general, to be younger than the same fish aged by Reader 2 (Figure 13A). Fin ray ages differed between readers by as much as 4 years, but were within 1 year of each other in 90% (APE=6.57) of all fin rays aged by both readers (Table 2). In gene ral, resolved fin ray ages by Reader 1 were younger than for resolved ages for the same fish by Reader 2 (Figure 13B). Comparative Age Estimates Using Fin Rays versus Otoliths Fin-ray ages agreed exactly with otolith ages 62% (Reader 1) and 58% (Reader 2) of the time, with agreement 1 year much higher (87-95%) (Table 1). Lin’s concordance correlation coefficien t between otoliths and fin rays indicated a near perfect concordance of age estimates from the two techniques ( rc>0.93 for both readers). Other measures of precision such as CV and APE were, in general, comparable to those calculated for within-reader agreement of fin ray ages. The majority of fish age estimations by Reader 1 fell on the 1:1 line for fish < 4 years old (Table 2, Figure 13A). For fish between 4 and 9 years old, the fin ray method began to consistently underestimate the age of the fish when compared to the otolith method. After age 9, fin ray ages showed no systematic bias towards underaging (Figure 14A). A test of symmetry (Hoenig et al. 1995) for all age classes (a ges 1-17) rejected the null hypothesis of symmetry about a main dia gonal that indicates a 1:1 relationship of otolith ages to fin ray ages ( 2 = 52.77, df = 26, p<0.001). However, inspection of the

PAGE 40

29 individual 2 components of the symmetry table revealed that the bias was attributed to 3 and 7 year old gag being underage d by 1 year (i.e., very high 2 components). If it is assumed that aging inaccuracies of 1 year fo r 3 and 7 year old gag are symmetric and the test of symmetry is rerun on this basis, then the bias associated with aging gag using fin rays becomes nonsignificant over 1-17 years ( 2 =32.77, df=26, p>0.10). Age composition based on otolith ages a nd paired fin ray ages using proportions from hook-and-line catch at age given in F itzhugh et al. (2003) showed no significant difference in the number of fish in each ag e class between an otolith-estimated age composition and a fin ray-estimated age composition (Figure 15) (Wilcoxon Signed Ranks test, two-tailed, n=7, T+=17, p>0.50) (Table 3). A test of symmetry revealed bias associated with aging gag using fin rays ( 2 = 26.3, df=12, 0.0050.25).

PAGE 41

30 Table 1. Measures of precision within-re ader, between-readers, and between aging structures for otoliths and fin rays , where CV=coefficient of variation, APE=average percent error, rc=Lin’s concordance correlation, and n=sample size. % Agreement ( 1 yr) CV APE rc n Otoliths Within Reader Reader 1 92 (99) 1.36 0.96 0.97 460 Reader 2 95 (97) 0.84 0.59 0.99 110 Between Readers 83 (95) 3.31 2.34 0.96 110 Fin rays Within Reader Reader 1 62 (94) 6.86 4.85 0.93 453 Reader 2 64 (95) 9.28 6.38 0.97 110 Between Readers 59 (90) 9.30 6.57 0.96 110 Between Structures Reader 1 (all data) 62 (93) 8.41 5.94 0.93 453 Reader 2 58 (95) 8.92 6.12 0.98 110 Reader 1 56 (87) 8.79 6.21 0.95 110

PAGE 42

31 Table 2. Percent agreement between ages estimated from otoliths and dorsal fin rays relative to the otol ith-assigned age for Reader 1 (n=453). Age Class Number of Fish (n) % Agreement Between Structures % Agreement Between Structures 1 year 0 7 86 100 1 51 88 100 2 34 82 94 3 119 75 98 4 87 62 94 5 60 50 98 6 32 41 84 7 41 34 83 8 12 16 67 9 3 0 33 10 1 0 100 11 1 0 0 12 1 0 0 13 1 0 0 14 1 0 100 15 1 0 0 17 1 0 100

PAGE 43

32 Table 3. Catch-at-age differences in gag aged using either otoliths or fin rays. Proportion-at-age is estimated from hookand-line caught gag in the Gulf of Mexico based on Fitzhugh et al. (2001). Number of gag in each age class (1 through 10+) using the otolith aging method is based on a random agestratified subsample from the present study that re presents the proportions from Fitzhugh et al. (2001). Number of gag in each age class using the fin ray aging method corresponds to placement of gag into an age class based on the assumption that the otolith-based ag ing method provides the true age. Age Class 1 2 3 4 5 6 7 8 9 10+ Proportion-at-age (Fitzhugh et al. 2001) 0.01 0.040.15 0.36 0.18 0.12 0.06 0.02 0.010.04 Aging Otolith 1 7 30 70 36 24 12 5 2 7 Method Fin Ray 1 10 36 50 43 25 11 2 2 7

PAGE 44

33 n=4370 5 10 15 20 25 30 35 4030 0 350 400 45 0 500 5 50 60 0 650 7 00 7 50 800 8 50 9 00 950 10 0 0 1 0 5 0 1 1 00 11 5 0 1 2 0 0 1 2 50Maximum Total Length (mm)Number of Fish Figure 4. Length frequency for all gag us ed in the present study. Samples where TL was unavailable were excluded (n=24).

PAGE 45

34 1 2 3 4 5 1 B C A 1 2 3 4 5 2 3 4 5 Figure 5. Annuli visible in basal versus distal sections of fin ray from an individual gag: A) basal section from just above pter ygiophore on a 5-yr old fish; B) section from a ray clipped level with the back of the gag; and C) section from ~ 2.4 mm distal on fin ray of the same fish, where first annulus is present but more difficult to identify.

PAGE 46

35 n=1380 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1234567 Resolved Otolith AgeMean Annular Diameter (mm) 1st annulus 2nd annulus Figure 6. Mean ( SE) diameter of the first and second annuli of fin rays of gag for each age class determined by resolved otolith age.

PAGE 47

36 A B 1 2 3 4 1 2 3 4 Figure 7. Picture of a gag dorsal fin ray unde r A) bright field (Khl er illumination) and B) the same fin ray under phase-contra st illumination. When viewed under phase contrast it was easie r to determine if the annulus was on the edge or whether there was any opaque growth beyond the ultimate annulus.

PAGE 48

37 A B C 1 2 3 1 3 2 2 1 3 Figure 8. Cross-section of dorsal fin rays of gag: A) showing how the first annulus can be partially occluded by blood vessels th at run through the center of the fin ray (3-yr old fish); B) showi ng distinctly split rings for 2nd annuli. This 3-yr old fish could be mistakenly aged as a 5 or 6 yr old fish; and C) showing annuli splitting into two distinct ring s on oblique inner portion of fin ray. 2

PAGE 49

38 A B compacted Evenly spaced 1 2 3 4 5 6. . . . . Figure 9. Cross-section of a dorsal fin ray of a gag showing: A) annuli compacting on the outer edge of the fin ray in the 5th and 6th years, whereas annuli are still evenly spaced near inner, oblique portions of the fin ray; a nd B) an 11-yr old gag with very clear, evenly spaced an nuli on the inner, oblique edge. Annuli are slightly more difficult to disti nguish on the outer portion of the ray.

PAGE 50

39 n=82 0.00 0.05 0.10 0.15 0.20 JFMAMJJASOND MonthMean Growth Increment (mm) Figure 10. Mean marginal translucent growth increment ( 1 SE) of age-4 gag otoliths over a 10-month period (April to January).

PAGE 51

40 n=306 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 123456789101112 MonthProportion of gag otoliths opaque little translucent significant translucent n=2470 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 123456789101112 MonthProportion of gag fin rays little opaque significant opaque translucent Figure 11. A) Qualitative marginal-increment analysis of the amount of translucent zone growth beyond the ultimat e opaque zone for gag aged 2-5 yrs; and B) qualitative amount of opaque growth obs erved on the edge of dorsal fin rays of gag aged 2-5 yrs collected from January-December. A B

PAGE 52

41 Reader 1 Reader 2 02468101214161820 Otolith Reading 1 (years) 0 2 4 6 8 10 12 14 16 18 20Otolith Reading 2 (years) n=460 02468101214161820 Otolith Reading 1 (years) 0 2 4 6 8 10 12 14 16 18 20 n=110 02468101214161820 Finray Reading 1 (years) 0 2 4 6 8 10 12 14 16 18 20Finray Reading 2 (years) n=454 02468101214161820 Finray Reading 1 (years) 0 2 4 6 8 10 12 14 16 18 20 n=110 Figure 12. Scatter plot comparison of the ag e estimate obtained using fin rays versus otoliths for gag. The diagonal line repr esents comparisons where otolith ages = fin ray ages. Circle size represents the relative sample size for that age combination.

PAGE 53

42 02468101214161820Reader 1 Otolith Reading (yrs) 0 2 4 6 8 10 12 14 16 18 20Reader 2 Otolith Reading (yrs) n=110 02468101214161820Reader 1 Fin ray Reading (yrs) 0 2 4 6 8 10 12 14 16 18 20Reader 2 Fin ray Reading (yrs) n=110 Figure 13. Scatter plot comparison of the age estimate obtained by Reader 1 versus Reader 2 using A) otoliths and B) fi n rays for gag. The diagonal lines represent comparisons where age estimated by Reader 1 = age estimated by Reader 2. Circle size represents th e relative sample size for that age combination.

PAGE 54

43 02468101214161820Reader 1 Otolith Age (years) 0 2 4 6 8 10 12 14 16 18 20Reader 1 Finray Age (years) n=453 02468101214161820Reader 2 Otolith Age (years) 0 2 4 6 8 10 12 14 16 18 20Reader 2 Finray Age (years) n=110 Figure 14. Scatter plot comparison of the ag e estimate obtained using otoliths versus fin rays for A) Reader 1 and B) Read er 2. The diagonal lines represent comparisons where age estimated for otoliths = age estimated for fin ray. Circle size represents th e relative sample size fo r that age combination.

PAGE 55

44 n=1940 10 20 30 40 50 60 70 80 12345678910+ Age (years)Number of gag Otolith Fin Ray Figure 15. Derived age structur e for gag grouper from otolith-assigned ages and fin rayassigned ages. The number of samples in otolith age classes of 1 to 10+ were chosen according to the proportions of th e age structure given in Fitzhugh et al. (2003).

PAGE 56

45 CHAPTER 4 DISCUSSION Aging Gag Using Dorsal Fin-Rays Dorsal fin rays proved to be a viable, a lternative nonlethal ag ing method for gag. Although measures of precision (i.e., aging re liability) were lower when aging gag using fin rays compared to sectioned otoliths for both within-reader and between-reader aging (Table 1), this precision wa s influenced primarily by eith er underaging or overaging gag fin rays by one year. In part icular, tests of symmetry demonstrated that specific age classes of gag are prone to aging bias, which wa s attributed to the presence of strong false annuli or “checks” in these ages. Ricker (1975) suggests that 80-90% agreemen t between sets of ag es (within-reader precision) is good. Within-reader precision was consistent for both Reader 1 and Reader 2 for both otoliths (94% and 95%, respectively) and fin rays (69% and 66%, respectively), indicating that each reader ha d approximately the same consistency aging the structures and similar difficulty interpreti ng fin rays. Another moderately long-lived marine species, bluefish ( Pomatomus saltatrix ) from the Chesapeake Bay region , had similar levels of within-reader agreement (86% and 63% for sectioned otoliths and dorsal fin spines, respectively, with ages rangi ng from 1-14) (Sipe and Chittenden 2002). Although between-reader agreement for gag grouper otolith ages was good (83% ranging from 0-17 years old), the relatively high variability observed both within and between Reader 1 and Reader 2 in age estima tes of gag older than 10 years using otoliths (Figure 12 and 13) was unexpected in comp arison to the lower variability in age

PAGE 57

46 estimates of these older gag using fin rays. Otoliths of older gag were difficult to age consistently due to the presence of a very wide translucent zone on the edge of the structure, with obviously either missing or i ndistinct opaque zones within the translucent zone. Fin rays for these same fish still ha d distinct opaque and translucent zones, making it easier for both readers to enumerate fin rays of older gag when compared to otoliths of older gag. Variation in otolit h age estimates increasing in older gag because of the lack of clarity of the opaque zones in a wide tr anslucent zone has also been recognized by other experienced gag otolith readers (pers. comm., Dr. Gary Fitzhugh, NMFS, Panama City). The difficulty in observing the opaque zones in older gag may occur because as the fish ages and the rate of somatic grow th decreases, the propor tion of opaque growth visible in the otolith is mini scule in relation to the larg e amount of translucent zone deposition that occurs over a longer period of reduced somatic growth. In the present study, between-reader agreem ent for gag grouper otolith ages (83%) was slightly greater than between-reader agreement documented in Hood and Schlieder (1992) (76% from samples ranging from 1-17 y ears old). However, it is important to keep in mind that Hood and Schlieder (1992) ag ed more fish in older age classes (31% older than 7 years) than in the present study (5% older than 7 years). Some reef fish species are typically easier to age than ga g, including white grunt with 98% agreement between readers, with a maximum age of 18 years in the Gulf of Mexico (Murie and Parkyn 2005). Johnson and Saloman (1984) al so found much higher percent agreement for between-reader dorsal spine precision for gray triggerfish, at 98% (maximum age of 13 years, with most fish 8 years) with an APE = 0.0072.

PAGE 58

47 Studies on comparative aging of fin rays or spines in relation to otoliths have met with variable results. Sikstrom (1983) not ed only 49% agreement between fin rays and otoliths for arctic grayling ranging in age from 1 to 12 y ears, which increased to 94% 1 year. In Sikstrom (1983), otolith ages were generally higher than fin ray ages when they did not agree. This was attri buted to difficulty in identifyi ng the first annulus in arctic grayling and difficulty obtaining sections su fficiently close to the base of the ray (Sikstrom 1983). Agreement between struct ures for gag was slightly higher than agreement between structures for arctic gr ayling (56-62% for gag compared to 49% for arctic grayling). Although the first annulus in gag was some times difficult to identify, it did not cause consistent disagreement between otolith and fin ray ages. Sectioned otoliths and dorsal fin spines of bluefish also had a re latively low agreement of 52%, which increased to 68% for bluefish younger than age 5 (Sipe and Chittenden 2002). In contrast, agreement between otoliths and dor sal fin rays was 90% for white grunt less than age 10 (Murie and Parkyn 1999). Agreem ent over all ages (0-20 years) of gag grouper based on otoliths and dorsal fin rays appeared to be moderate (62%) in comparison to these other studies. The ages obtained from reading otoliths in th is study were assumed to be the “best” estimate of the age of the gag (i.e., 100% accurate) and were therefore useful in comparing ages obtained by other methods. Th is assumption is obviously a problem if it proves to be false, as proved to be the cas e in aging gag >10 year s using otoliths. The extent to which ages of older gag based on otoliths have confounded the comparison of ages based on fin rays is presently unknow n. To rectify this assumption, it will be necessary to improve the methods used curren tly to age older gag using otoliths, perhaps

PAGE 59

48 by either taking thinner sections or polishing or staining the otolith sections to increase the contrast of the opaque zones (Vanderkooy and Tisdel 2003). The acceptability of using otolith ages for these older gag as the “best” ages will depend on being able to age the fish with either equal or greater precision than the alternative agi ng structure (fin rays in this case). Challenges with the Fin-Ray Method Potential aging inaccuracies using fin rays can be largely attributed to four main problems: 1) identification of the first annul us; 2) differentiating false annuli or checks from true annuli; 3) compaction of the annuli on the edge; and 4) interpretation of the edge of the aging structure. The first potential problem in assigning ages to fin rays of gag was identification of the first annulus. The first annulus in fi n rays was sometimes partially occluded by the basal sinus that runs through the center of th e ray (Figure 8A), or the region where the blood vessel sinus should be located was almo st completely missing. This particular problem is not uncommon with fin ray and spine aging (Beamish and Chilton 1977). Partial occlusion and a general blurriness or lightness near the core made distinguishing the first year difficult on some rays. This pr oblem was less common in fish aged 0-3 than in fish 4 years of age and older. Beam ish and Chilton (1977) pr oposed measuring the mean radius of the first and second annulus as a method for determining if the first annulus has been reabsorbed or occluded by th e sinus in the center of the fin ray. Penha et al. (2004) further proposed measuring the mean radius of the first annulus and comparing between age groups to determine if the first annulus has been missed in any age groups. These methods were applicable fo r determining if the fi rst annulus had been missed when reading fin ray sections of gag since the diameters of the first and second

PAGE 60

49 annuli were significantly distinct from one another (Figure 6). In addition, there was no trend of increased first annular radius with ag e, which would be expected if resorption or occlusion of the first annulus was occurring in older fish. An occl uded first annulus in fin rays of some species is a serious problem due to vasculari zation of fin rays. However, in comparison to lingcod, gag do not have heavily vascularized fin rays (pers. comm., Debra Murie). Being able to identify the approximate location of the first and second annuli in gag fin rays by using a measuremen t based on their diameter is advantageous for both readers lacking experi ence with the method and experienced readers that need to revisit the reference aging collection pr ior to reading additional fin rays. Secondly, problems in differentiating check s from true annuli led to systematic aging bias in 3, 4, and/or 7 year old gag, based on tests of symmetry. The systematic bias in underaging 3 year old gag, or aging 4 year old gag as either 3 or 5 years, was most likely due to misinterpretation of checks ve rsus annuli. Between the ages of 3 and 6, checks were particularly difficult to distingu ish from annuli and gag in this age range had a tendency to have each annulus split into tw o or more distinct checks. Knowing this, readers may have had a preconception of the split-annulus pattern and may have overcompensated by putting 4 year old gag in age class 3 when in fact each of the observed “checks” was actually an annulus. This may also be the case for 7 year old gag that were systematically underaged by 1 year . When aging, the reader may have assumed that since gag of ages 3-6 had problematic checks, this may extend to gag with 7 apparent annuli. The compaction of annuli at the edge of fi n rays in older gag, as a potential source of underaging, was not as signi ficant a problem as initia lly suspected based on other

PAGE 61

50 studies using fin rays (Beamish and Harvey 1969; Beamish 1981; Sipe and Chittenden 2002). If underaging of fin rays was occurring then a scatter plot of otol ith versus fin-ray ages would reveal data consistently below the 1:1 line that indicat es perfect agreement between structures. For fi n rays, this typically occu rs when the annuli become compacted on the edge of the structure because readers cannot differentiate the number of annuli at the edge, and therefore there is an increasing difference between the otolith ages and the fin ray ages with increasing fish age (Chilton and Beamish 1982). For gag, age estimates for fin rays by Reader 1 began to show evidence of underaging between age 5 and age 9, beyond which there appeared to be no systematic underaging as evidenced by data falling to either side of the 1:1 line (Figure 14A). Again, this aging error can be attributed to problematic checks in these age classes, as evidenced by systematic underaging by 1 year in age classes 3 and 7. This was different than the systematic underaging of fin rays by 1 or 2 years that wa s evident in gag >10 years aged by Reader 2 (Figure 14B). However, obtaining resolved ag es for these older ga g using otoliths was problematic for both readers (Figure 13), with Reader 1 concluding that these older gag were being underaged, which was suppported by Reader 2 consiste ntly assigning these older gag into older age classe s than Reader 1 (Figure 13). Without increa sed precision in aging older gag using otoliths, which is s upposed to provide the “best” age for the fish, it was not possible to state that the differe nce from the 1:1 line was indicative of a systematic or unacceptable underaging error. The results of this study indicate that older gag can be aged more precisely with fin rays th an with otoliths. This needs to be further examined with larger sample sizes.

PAGE 62

51 A fourth type of aging inaccuracy can be at tributed to difficulty in interpreting the edge of the fin ray. The edge on gag fin ray sections often appeared cracked, making annuli difficult to distinguish. The cracked edge may be attributed to a thin layer of skin that was not removed through the short boili ng process. Although fi ns are boiled until the skin is soft and then lightly brushed, a th in layer of skin may remain. Even a small amount of skin folded over the edge of th e sectioned fin ray can obscure the edge. Another possibility is that the outer edge of the dried or froz en fin ray is easily fractured during the drying, mounting and/or sectioning process. This damage may result in tiny cracks that are present along the outer edge of the fin. Alternative methods of removing the skin and associated tissue around the fin ra ys should be examined if it is found that that observed cracking results fr om the processing methodology. Annulus Formation in Otoliths and Fin Rays Understanding the timing of annulus formati on is critical when comparing different aging methods, but is often overlooked or li mited to only one of the aging structures being examined. In particular, timing of the deposition of annuli in paired otolith and fin ray samples has never been compared to our knowledge. This comparison is necessary in order to determine if ages estimated using different methods can be interpreted in the same manner and is essential for placement of fish into proper age classes. Qualitative marginal-increment analysis for otoliths and fin rays indicated that the two structures deposited complete opaque and translucent z ones at approximately the same time each year. It is important to note that qualitative marginal-increme nt analysis for otoliths did not reveal a high proportion (>0.5) of opaque growth in any months. The highest proportion of gag otoliths with complete opa que zones identifiable on the edge occurred in May and June. This is most likely due to fast growth of the otolith occurring in a very

PAGE 63

52 short period of time, associated with a growth spurt in the fish. The same is true for opaque zones in fin rays, which represent fast er zones of growth. The highest proportion of completed opaque zones occurs from June-A ugust, but does not become greater than 0.5. This is also likely due to concentrated, short, fast growth periods. Larger sample sizes in the summer months may have produced more distin ctive patterns in qualitative marginal-increment analysis. It is also possi ble that the variability in the start of fast growth periods, combined w ith the short duration of fa st growth, may have made distinctive patterns (where th e proportion of fish with an opaque zone on the edge is >0.5) difficult to observe. The low proportions of the opaque zone on the edge of the structures may also be a resu lt of the applied aging criter ia. Annuli were designated as “on the edge” of the otolith if the opaque z one was visible near the sulcus. However, annuli can often be identified on the dorsal and ventral tips of the otolith sections earlier than near the sulcus. If the criteria for ot oliths was extended to include annuli visible on the edge of any part of the otolith secti on, the proportion of otolit hs with annuli on the edge would likely increase. Criteria for fin rays could also change the proportion of annuli judged “on the edge” of th e structure. The criteria us ed in this study designated annuli as “on the edge” if a translucent zone was visible on the edge of the oblique, inner part of the fin ray section, as well as on so me portion of the outer edge of the fin ray. Since fin rays lay down the translucent zone on the inner, oblique area of the fin ray first, if the criteria were changed to include all fi n rays with translucen t zones on the oblique sections only as “on the edge”, the proporti on with annuli on the edge would increase. Marginal-increment analysis provided evidence of annual deposition in the fin rays and otoliths of young (2-5 yr old) gag (Figure 11). Since the otolith aging method of

PAGE 64

53 gag has not been validated for fish >7 yr of age (Collins et al. 1987; Hood and Shlieder 1992), it must be assumed for now that one annulus is deposited each year in older fish as well. Whether this is correct or not is not of paramount im portance for the present study because the deposition of annuli is simila r in both of the aging structures under consideration when viewed relative to one anot her (rather than in absolute terms). So otoliths and fin rays appear to deposit annuli at the same ra te in older gag, regardless of the accuracy of that rate. Marginal-increment analysis was fundamental to understanding how to place gag into age classes. Because opaque zones were enumerated in otoliths and translucent zones were enumerated in fin rays, fish woul d have been placed into incorrect age classes if the timing of the deposition of these z ones was not properly understood. In most studies investigating alternate aging methods for a fish, two or more aging structures are enumerated independently and then compared. Often, fish are placed into age classes based solely on annual mark counts (Crumpt on et al. 1984; Sipe and Chittenden 2002; Ihde and Chittenden 2002). Having young gag (0 and 1 yr olds) in the samples allowed us to determine that although gag have an opa que core to their otolith, they do not lay down their first opaque zone (annulus) in the otolith until the second calendar year of life when they are ~15 months old. In contrast , the fin ray does lay dow n a translucent zone (annulus) in the first calendar year when the fish is ~8 months old. Due to the enumeration of opaque zones in otoliths and tr anslucent zones in fin rays, these structures would place a gag into different age classe s if the timing of deposition had not been evaluated (i.e., by otolith a gag would be placed in age 0 in the fall of their first year and by using the corresponding fin ray the same fish would be placed into age 1). Without

PAGE 65

54 understanding the timing of deposition (via ed ge or marginal-increment analysis) the aging structure under in vestigation may be interpreted inappropriately. Analyzing the timing of deposition and composition of a nnuli are therefore crucial steps in understanding the comparability of two aging structures. Opaque zones observed in gag fin rays vi sibly represented somatic growth of the fin (i.e., there was tissue deposit ion). If it is assumed that the growth in the dorsal fin rays is comparable in timing to the rest of the somatic body growth of the gag, then this suggests that the opaque zone of the fin rays represents relatively fast growth in gag, whereas the translucent zone represents rela tively slow or no so matic growth. Narrow translucent zones in fin rays occurred over an extended period of time in which water temperatures decreased and stayed generally cooler in coastal waters of the Gulf of Mexico where the majority of gag were co llected (Kline 2004). With decreased water temperature somatic growth is slowed, and henc e growth of the fin ray is minimal. Wide, opaque zones in fin rays that are believed to be associated with relatively faster somatic growth occurred as water temp erature increased in the spri ng and early summer months. Conventional aging methods for marine r eef fishes in the southeastern U.S. enumerate opaque zones, with the understand ing that these opaque zones represent a period of relatively slower growth (Hood and Schlieder 1992; Harris and Collins 2000; Vanderkooy and Tisdel 2003). However, base d on the near simultaneous deposition of complete opaque and translucent zones in otol iths and fin rays of gag (Figure 11A and 11B), it would appear that the opaque zone in gag otoliths more like ly represents a period of relatively faster growth. The translu cent zone in gag fin rays was narrow and was observed on the edge of the fin ray structures for a relatively long period of time, from

PAGE 66

55 fall through to spring (Figure 11B). If fin-ra y growth increments were indicative of body growth, then somatic growth of gag was at a minimum through the late fall and into the early spring. Although the translucent zone in the otoliths was also observed on the edge of the otolith over an extended period of time similar to the fin rays, the translucent zone nevertheless continued to be deposited over th at time period and was therefore relatively wide. The continued deposition of the transl ucent zone in the otoliths of gag over a period of little or no apparent somatic grow th (i.e., dorsal fin growth) indicated that otolith growth at that time was uncoupled from somatic growth in gag (Casselman 1990; Wright et al. 1997). This unc oupling of somatic growth from otolith growth has been observed previously in other fi shes. Most notably, experime ntal studies on arctic char have shown that even once somatic growth has ceased, otoliths continue to grow over the life of the fish (Mosegaard et al. 1988). In the southeastern U.S., Murie and Parkyn (2005) have also noted that linear regressi ons relating white grunt TL to otolith radius result in poor regression fits (r2=0.43-0.51), indicating that somatic growth is not directly proportional to otolith size. Thus older wh ite grunts (e.g., 14-18 years old) have been observed to have approximately the same TL as younger fish (e.g., 45 yrs old), but the latter group have larger otol iths (increased otolith radi i) (Murie and Parkyn 2005). Wright et al. (1997) also s uggest that there is evidence against a ny direct functional relationship between somatic growth rate and otolith accretion rate. Experimental studies that have either reduced or eliminated food in take of fish have noted that the fish stop growing in body size and weight, yet their otoliths continue to growth, even through periods of starvation (Casselman 1990; Wright et al. 1997). If deposition of calcium carbonate and otolin is physio logically linked to metabolic rate, including maintenance

PAGE 67

56 metabolism throughout the year, then deposition in the otoliths during periods of little or no somatic body growth is reasonable (Wright et al. 1997). In addition, otoliths are used as a sensory structure by the fish for heari ng and balance (Victor and Brothers 1982) and their increase in size may be at least partially decoupled from an increase in the size of the fish’s body due to their function, as ha s been observed for other sensory structures such as eyes (Pankhurst and Montgomery 1994). Implications of a Nonlethal Aging Method for Gag In summary, this study showed that fin rays have the potential to be a useful nonlethal aging structure for gag. There are cu rrently no critical standards for acceptable levels of APE, but Morison et al. (19 98) recommends a threshold APE of 5 for production aging. In this study, within-r eader APE for fin rays was 4.85-6.38 and between-structure (otoliths versus fin rays) APE was 5.94-6.21. The advantage of the fin-ray aging method is that the fish do not need to be sacrific ed for age determination. Removal of dorsal fin rays, in particular, shou ld provide minimal harm to the fish. Fin rays may be a better structure to remove from the fish than dorsal fin-spines, because dorsal fin-spines are used for defense against predators whereas fin rays largely serve as a supportive structure for the fin. Although this study focused on the use of dorsal fin rays, dorsal fin spines may also prove to be a usef ul structure and should be further examined. In addition, to further reduce possible harm to the fish, the recommendation of this study is that fin rays be rem oved from the dorsal fin where the rays meet the body of the fish. By removing the fin rays at this level, all annuli were retained and minimal risk of infection would be imposed on the fish. It is therefore not necessary to remove the fin rays at the base of the pterygiophores, whic h is advantageous for groupers since they

PAGE 68

57 have a fleshy base to their dor sal fin and fin excision at a lower level could create an open wound. The disadvantages of the fin-ray method ar e that fin rays are more time consuming to prepare and process (Murie and Parkyn 1999) and interpret than otoliths. For older fish (>8 years old), the fin ray method needs to be further evaluated to determine its reliability, but this comparison ultimately de pends on obtaining reliable age estimates for these same fish using otoliths, which proved to be a significant problem that needs further resolution. Significant aging e rrors may be caused by using only fin rays to age fish beyond the age at which the method breaks down. This could lead to error in estimations of gag grouper population parameters, such as mo rtality and growth rates. For the limited gag samples >8 years old that were assessed in this study (n=7), the fin-ray method was useful and underor over-estimated gag ages by only 1-2 years in most cases. The test of symmetry did not indicate any systematic bias for older fish and w ithin-reader precision was higher for fin rays (when aging older fish ) than for otoliths. Use of the fin ray method for aging older fish should be further investigated. It was also apparent that developing the methodology a nd interpretation of using fin rays to age gag in a nonlethal manne r could be of direct benefit to ongoing management efforts for gag, as well as prot ected Nassau and goliath groupers. For gag, the fin-ray method has important applicati ons for studies currently being conducted offshore in the Middle Grounds of the Gulf of Mexico, where specific areas are closed to fishing during the spawning season each year to protect aggregations of gag and other reef fishes (Gulf of Mexico Fishery Ma nagement Council 2001). As part of the examination of the effectiveness of thes e reserves, the age structure of the gag

PAGE 69

58 populations inside and outside of these reserv e areas are being asse ssed to determine if the age structure of gag in the reserve ar eas are showing signs of recovery (i.e., proportion of gag in older age classes is in creasing) (pers. comm ., Dr. Chris Koenig, Florida State University, Talla hassee, FL). Gag from the reserve areas are therefore currently being aged using th e nonlethal fin ray aging method so that age samples can be taken in a non-destructive manner (i.e., removi ng 1-2 fin rays versus the otoliths). The fin-ray method for gag has been shown to be a viable nonlethal alternative to otolith aging and is potentially more accurate and pr ecise than otoliths when aging fish older than 10 years old. Using gag as a model grouper species, th e fin-ray aging method has also been demonstrated to be an appropriate method to explore for use in aging other serranids, such as goliath and Nassau groupers. Like gag, these serranids are moderately longlived, warm-temperate fish with overlapping latitudinal ranges. This would be an especially appropriate method for aging thes e threatened serranid s since it would allow for monitoring a rebuilding of th e age structure of their stocks without sacrificing a large number of already scarce fish.

PAGE 70

59 LIST OF REFERENCES Beamish, R.J. 1981. Use of fin ray sections to age walleye pollock, Pacific cod, and Albacore, and the importance of this method. Transactions of the American Fisheries Society 110: 287-299. Beamish, R.J. and H.H. Harvey. 1969. Age de termination in the white sucker. Journal of the Fisheries Research Board of Canada. 26: 633-638. Beamish, R.J. and D. Chilton. 1977. Age determination of lingcod ( Ophiodon elongates ) using dorsal fin rays and scal es. Journal of the Fisherie s Research Board of Canada 34 (9): 1305-1313. Beamish, R.J. and D.A. Fournier. 1981. A met hod for comparing the precision of a set of age determinations. Canadian Journal of Fisheries and Aquatic Sciences 38: 982983. Bilton, H.T. and D.W. Jenkinson. 1969. Age determination of sockeye and chum salmon from examination of pectoral fin ra ys. Journal of the Fisheries Research Board of Canada 26 (5): 1199-1203. Bowker, A.H. 1948. A test for symmetry in contingency tables. Journal of the American Statistics A ssociation. 43: 572-574. Brennan, J.S. and G.M. Cailliet. 1989. Co mparative age-determination techniques for white sturgeon in California. Transacti ons of the American Fisheries Society118: 296-310. Bullock, L.H., M.D. Murphy, M.F. Godcharles and M.E. Mitchell. 1992. Age, growth, and reproduction of Goliath grouper, Epinephulus itajara , in the eastern Gulf of Mexico. Fisheries Bulletin 90 (2): 243-249. Campana, S.E. 2001. Accuracy, precision a nd quality control in age determination, including a review of the use and abuse of age validation methods. Journal of Fish Biology 59 (2): 197-242. Cass, A.J. and R.J. Beamish. 1983. First ev idence of validity of the fin-ray method of age determination for marine fishes. North American Journal of Fisheries Management 3: 182-188.

PAGE 71

60 Casselman, J.M. 1990. Growth and relative size of calcified stru ctures of fish. Transactions of the American Fisheries Society 119: 673-688. Chilton, D.E. and R.J. Beamish. 1982. Age de termination methods of fishes studied by the Pacific Biological Station. Canadian Special Publications on Fisheries and Aquatic Sciences 60. Coleman, F.C., C.C. Koening and L.A. Co llins. 1996. Reproductive strategies of shallow water groupers in the eastern Gulf of Mexico and the consequences of fishing spawning aggregations. Environm ental Biology of Fishes 47 (2): 129-141. Collins, C.W., C.W. Waltz, W.A. Roumillat an d D.L. Stubbs. 1987. Contribution to the life history and reproductive biology of gag in the South Atlantic Bight. U.S. National Marine Fisher ies Service Fisheries Bulletin 85: 648-653. Collins, C.W., S.B. Van Sant and G.R. Se dberry. 1996. Age validation, movements, and growth rates of tagged gag, black sea bass, and red porgy. In Biology, fisheries and culture of tropical groupers a nd snappers. F. Arreguin-Sanchez, J.L. Munro, M.C. Balgos and D. Pauly, editors. ICLA RM Conference Proceedings 48: 158-162. Crumpton, J.E., M.M. Hale and D.J. Renf ro. 1984. Aging of three species of Floridacatfish utilizing three pectoral spin e sites and otoliths. Proceedings of the Annual Conference SEAFWA: 335-341. Dutka-Gianelli, J. 1999. Comparat ive age and growth of sheepshead, Archosargus probatocephalus (Walbaum 1792) (Pisces: Sparidae), from the northwestern coast of Florida. M.S. Thesis, University of Florida, Gainesville. 68 pages. Fitzhugh, G.R., L.A. Lombardi-Carlson and N.M. Evou. 2003. Age structure of gag ( Mycteroperca microlepis ) in the eastern Gulf of Me xico by year, fishing mode, and region. Proceedings of the Gulf a nd Caribbean Fisheries Institue 54: 538-549. Florida Fish and Wildlife Commission, FMRI . 2003. Species account for gag grouper. www.floridamarine.org . Goodyear, C.P. and M.J. Schirripa. 1993. The red grouper fishery of the Gulf of Mexico. National Marine Fisheries Service, SE Fisheries Ce nter, Miami Lab, Miami CRD92/93-75. 122 pages. Gulf of Mexico Fishery Management Council. 2001. Commercial fishing regulations for Gulf of Mexico Federal Wa ters via the internet. ( http://www.gulfcouncil.org ). Harris, P.J. and M.R. Collins. 2000. Age, growth and age at maturity of gag, Mycteroperca microlepis , from the southeastern United States during 1994-1995. Bulletin of Marine Science 66(1): 105-117.

PAGE 72

61 Hoenig, J.M., M.J. Morgan, and C.A. Brow n. 1995. Analysing differences between two age determination methods by tests of symm etry. Canadian Journal of Fisheries and Aquatic Sciences 52: 364-368. Hood, P.B. and R.A. Schlieder. 1992. Age, growth and reproduction of gag, Mycteroperca microlepsis , in the eastern Gulf of Me xico. Bulletin of Marine Science 61 (3): 581-591. Ihde, T.F. and M.E. Chittenden. 2002. Comp arison of calcified structures in aging spotted seatrout. Transactions of the American Fisheries Society 131: 634-642. Johnson, A.G. and C.H. Saloman. 1984. Age, growth, and mortality of gray triggerfish from the Northeastern Gulf of Mexico. Fisheries Bulletin 82 (3): 485-492. Kimura, D.K. and J.J. Lyons. 1991. Between-Reader Bias and Variability in the AgeDetermination Process. Fisheries Bulletin 89: 53-60. Kline, R.J. 2004. Metabolic rate of the gag grouper Mycteroperca microlepsis in relation to swimming speed, body size, and seasonal temperature. M.S. Thesis University of Florida, Gainesville. Leaman, B. M., and R. J. Beamish. 1984. Ec ological and management implications of longevity in some Northeast Pacific gr oundfishes. Internatio nal North Pacific Fisheries Commission Bulletin 42:85. Lin, L.I. 1989. A concordance correlation coefficient to evaluate reproducibility. Biometrics 45: 255-268. Lin, L.I. 2000. A note on the concordance co rrelation coefficient. Biometrics 56: 324325. MacFarlane, G.A. and J.R. King. 2001. The validity of the fin-ray method of age determination for lingcod ( Ophiodon elongatus ). Fisheries Bulletin 99: 459-464. Mills, K.H. and R.J. Beamish. 1980. Comparison of fin ray and scale agedeterminations for lake whitefish and their implications for estimates of growth and annual survival. Canadian Journal of Fi sheries and Aquatic Sciences 37 (3): 534544. Morison, A.K., S.G. Robertson, and D.C. Smith. 1998. An integrated system for production fish aging: image analysis and quality assurance. North American Journal of Fisheries Management 18: 587-598. Morris, J.V., C.M. Roberts and J.P. Hawkins. 2000. The threatened status of groupers (Epinephelinae). Biodiversity and Conservation 9: 919-942.

PAGE 73

62 Mosegaard-Nin, B., E. Gutierrez, and S. Massuti. 1988. Uncoupling of somatic and otolith growth rates in Arctic char ( Salvelinus alpinus ) as an effect of differences in temperature response. Canadian Journal of Fisheries and Aquatic Sciences 45: 1514-1524. Murie, D.J. and D.C. Parkyn. 1999. Age, grow th, and sexual maturity of white grunt. Prepared for FMRI, St. Petersbur g. Final Report: Part II. 57p. Murie, D.J. and D.C. Parkyn. 2002. Compar ison of total mortality of white grunt from the headboat fishery on the gulf coast of Florida during spawning and postspawning seasons. North American J ournal of Fisheries Management 22: 806814. Murie, D.J. and D.C. Parkyn. 2005. Age and growth of white grunt ( Haemulon plumieri ): a comparison of two populations al ong the west coast of Florida. Bulletin of Marine Science 76 (1): 73-93. NMFS (National Marine Fisherie s Service) 1990a. Fishery management plan for the reef fish of the Gulf of Mexico. Amendment 2. Gulf of Mexico Fisheries Management Council. Tampa. 20p. NMFS (National Marine Fisheries Service) 1990b. Fishery management plan for the snapper-grouper fishery of th e S. Atlantic Bight region. Amendment 2. Gulf of Mexico Fisheries Management Council. Charleston. 28p. Olsen, D.A. and J.A. LaPlace. 1979. A study of a Virgin Islands grouper fishery based on a breeding aggregation. Proceedings of the Gulf and Caribbean Fisheries Institute 31: 130-144. Pankhurst, N.W. and J.C. Mont gomery. 1994. Uncoupling of vi sual and somatic growth in the rainbow trout Oncorhynchus mykiss . Brain Behavioral Evolution 44 (3): 149-155. Penha, J.M.F, L.A.F. Mateus, and M.P. Petrere. 2004. A procedure to improve confidence in identification of the first a nnulus in fin-spines of fishes. Fisheries Management and Ecology 11 (2): 135-137. Ricker, W.E. 1975. Computation and Interp retation of Biological Statistics of Fish Populations. Bulletin of the Fisheries Research Board of Canada 191: 382p. Schirripa, M.J. and K.M. Burns. 1997. Growth estimates for three species of reef fish in The eastern Gulf of Mexico. Bulleti n of Marine Science 61(3): 581-591. Schirripa, M.J. and C.P. Goodyear. 1994. Status of the gag stocks of the Gulf of Mexico: Assessment 1.0. CRD-93/94-61. NMFS, SE Fisheries Center, Miami Lab., Miami 156p.

PAGE 74

63 Shirvell, C.S. 1981. Validity of fin-ray agi ng for brown trout. Journal of Fish Biology 18: 377-383. Sikstrom, C.B. 1983. Otolith, pectoral fin ray, and scale ag e determination for arctic grayling. Progressive Fish -Culturist 45 (4): 220-223. Sipe, A.M. and M.E. Chittenden, Jr. 2002. A comparison of calcified structures for aging bluefish in the Chesapeake Bay re gion. Transactions of the American Fisheries Society 131: 783-790. Turner, S.C., C.E. Porch, D. Heinemann, G.P. Scott and M. Ortiz. 2001. Status of gag in the Gulf of Mexico, Assessment 3.0. Sustai nable Fisheries Division Contribution: SFD 01/ 02-134. Southeast Fisher ies Science Center, Miami. Vanderkooy, S. and K. Guindon-Tisdel (ed itors). 2003. A practical handbook for determining the ages of Gulf of Mexi co fishes. GSMFC. Ocean Springs, MS. Victor, B.C. and E.B. Brothers. 1982. Age and growth of the fallfish Semotilus corporalis with daily otolith in crements as a method of annulus verification. Canadian Journal of Zoology 60: 2453-2550. Wright, P.J., H. Mosegaard, B. Morales-Ni n, and A. Geffen. 1997. The regulation of otolith formation and its si gnificance to age determination, size back-calculation and elemental composition. In The pres ent status of otolith research and applications. Proceedings of a works hop, held at ORSTOM, Brest, France. May 27-29. P.J. Wright, editor. Pages 4-21.

PAGE 75

65 BIOGRAPHICAL SKETCH Jaclyn Debicella was born on October 14, 1977, in Bridgeport, CT, and grew up in Shelton, CT. She graduated from Bucknell Un iversity in Lewisburg, PA, in 1999 with a Bachelor of Science in biol ogy. Jaclyn worked as science instructor at Seacamp in the Florida Keys and as a project assistant w ith the National Fish and Wildlife Foundation before taking a position as a fisheries resear ch assistant with the University of Florida Department of Fisheries and Aquatic Sciences.