SEDAR24-RD33

Transactionsof the American Fisheries Society' 111:465475, 1982 ¸ Copyrightby the AmericanFisheries Society 1982

Growth and Mortality of Red Snappersin the West-Central Atlantic Ocean and Northern Gulf of Mexico 1'2

RUSSELL S. NELSON AND CHARLES S. MANOOCH III

National Marine Fisheries Service BeaufortLaboratory, Beaufort, North Carolina 28516

Abstract Age, growthrates, mortality, and length-weightrelationships were determined for red snap- persLutjanm campechanua from twoareas in the westcentral Atlantic Ocean and two in thenorthern Gulf of Mexico.Growth ratesvaried only slightlyamong areas.The von Bertalanffygrowth equationwas Lt = 975(1 - e-ø'•6•t-ø'ø•) for fish from the Atlanticand Lt = 941(1 - e-ø'•7"+ø'") for fishfrom the Gulf of Mexico,where Lt is lengthat aget in years.The length-weightequation differed amongareas, but no differencewas evident between sexes. This equationfor fishfrom all areaspooled was W = 2.04 x 10-5 TLTM, where W is weightin gramsand TL is totallength in mm. Significantdifferences were found in totalinstantaneous mortality rates. The rateranged from 0.39 for fish off the coast of North Carolina and South Carolina to 0.78 for fish off the coastof Louisiana.Age of full recruitmentto the fisheryranged from 2 yearsoff Louisianato 6 yearsoff the Carolinas.Fishing mortality appears to be the majorcause of differencesin total mortality amongareas.

In this paper we examine annual growth, ly little information is available concerningits mortality, and age structurein populationsof ecology.Length-weight relationships, move- red snapper Lutjanuscampechanus from four ments,and growthhave been studied, but most widely separatedgeographic areas. Red snap- work hasbeen limited to singlelocations in the persare large, predatoryfish commonly occur- Gulf of Mexico, samplesizes have been small, ring in associationwith live-bottomand deep- and data virtuallyare absentfor growthof fish reef areasfrom Cape Hatteras, North Carolina older than 5 years and for mortality (Camber to the Campechebanks of Mexico.South of the 1955; Dawson 1963; Moseley 1966; Beaumar- Cuban Straitsthey apparentlyare replacedby iage 1969;Moe 1969;Bradley and Bryan 1975; the nominal congenericCaribbean red snapper Futchand Bruger 1976; Wade 1978; Wakeman L. purpureus.The deep-reefhabitats, essentially et al. 1979). Recentlife-history studies of other tropicalto sub-tropicalin bioticcomposition, are fish in the deep-reef taxocenehave been based characterizedby extensivealgal, soft coral, and on collectionsfrom narrow geographicranges sponge growth in conjunctionwith bedrock and do not describethe amount of geographic outcroppings,drowned coral reefs, and other variation in demographicparameters•partic- bottom irregularities(Brooks 1962; Struhsaker ularly growth and mortalityrates--often incor- 1969;Smith et al. 1975;Bright and Rezak1976). porated into managementmodels. The red snapper contributesheavily to both commercialand recreationallandings through- Methods out its range (Allen and Tashiro 1976; Hunts- Red snappers were sampled between 1974 man 1976; Nakamura 1976). Despiteits value and 1979 from recreational headboat hook-and- to both domesticand foreign fisheries,relative- line fisheries operating off North Carolina, South Carolina, and Florida, and from the commercial hook-and-line fishery off Louisi- ana. Samples were separated by geographic • Basedin part on a Master of Sciencethesis sub- areas: the Gulf of Mexico off Louisiana (Loui- mitted by the seniorauthor to the GraduateSchool siana);the Gulf off PanamaCity, Florida(West Faculty,Department of ,North CarolinaState Florida); the Atlantic Ocean off Florida from University,Raleigh, North Carolina. = Contribution 82-21-B 1, Southeast Fisheries Cen- Daytonasouth to Fort Pierce(East Florida); and ter, National Marine FisheriesService, Beaufort, North the Atlantic Ocean off North Carolina and Carolina 28516. South Carolina (Carolinas). All fish were col-

465 SEDAR24-RD33

466 NELSON AND MANOOCH lected between June 1978 and May 1979 in magnified image from the focus to each ob- Louisiana and West Florida. In East Florida, servedring. sampleswere taken from April 1976to October Otoliths (sagittae)were exposedby a trans- 1978. Samplingoccurred from April 1974 to versecut with a hacksawthrough the cranium November 1978 in the Carolinas. midwaybetween the posterioredge of the orbit Differencesbetween the commericalfishery and the preopercle.They were removed with (Louisiana) and the recreational fisheries (all forcepsand storedin paperenvelopes. All oto- other areas)should be noted.Fish taken by the lithswere sectioned laterally through the focus recreational fisheriesprobably were caught with a BuchlerIsomet 11-11803 low-speed saw within 90 km of the landingport, whereassam- by proceduresof Berry et al. (1977). Sections plesfrom the commercialfishery may have been of 0.18 to 0.28 mm thickness were mounted on taken 150 or more kilometersfrom port. Ad- glassslides with Permount3, and projectedand ditionally, the recreationalfisheries concentrate viewed in the same manner as scales. All mea- fishing pressureduring the daylight hours, surements were taken from the focus to each whereasthe commercialfishery operates around observedring alongthe medioventralridge. the clock.Because larger red snappersmay be Scales and otoliths fi'om all fish less than 235 more activeduring the day than smallerones mm and greater than 800 mm were used, be- (Bradleyand Bryan 1975),the recreationalcatch causeboth sizeextremes were infrequently en- mightbe slightlybiased toward larger fish. countered.The commonlysampled sizes were Fishwere measuredand weighedat the port dividedinto 50-mmlength classes and subsam- of landing, but no weightswere taken in Lou- pled. All scalesand otolithswere read at least isiana becausefish were gutted before exami- twice at an interval of 2 or 3 months. Those from nation.Fish sampled in the Gulf of Mexicohad fishgreater than 800 mm were read three times, fork lengths(FL) recordedbut thosefrom the and either the mode or the mean was used as Atlantic had total lengths(TL) recorded.All estimatedage. lengthsreported henceforth are TL. The re- Ratesof instantaneoustotal mortality (Z) were lationship of FL to TL is TL = 1.0712 estimatedby the regressionmethod (Ricker FL + 1.700; r = 0.99; N = 180. Scales were re- 1975)and the methodof Robsonand Chapman moved from all fish. Otoliths were removed ( 1961).Instantaneous rates of natural mortality from, and sex determined for, some fish. (M) were determinedwith the procedurede- The 4/0 to 9/0 hook sizescommonly used in scribedby Pauly (1980). An estimateof the in- all fisheries tended to exclude fish smaller than stantaneousrate of fishing mortality (F) was 180 mm from the catch. However, 263 small obtained by subtraction:F = Z - M (Ricker red snapperswere obtained from research 1975). trawlingoff the Carolinasand in SaintAndrews Bayin WestFlorida through the courtesyof the Results and Discussion South Carolina Marine Resources Research In- stitute, Charleston, and the National Marine Validityof AgeDetermination FisheriesService, Panama City, Florida. Scalesamples fi-om 2,151 fish representing Scaleswere taken from beneaththe tip of the all study areas were examinedand 76% were posteriorlyextended pectoral fin, storedin pa- usable.On legible scales,recurring patternsof per envelopes,cleaned, rinsed, and dried be- compressedcirculi in conjunctionwith marked fore being read and measured.They were ex- crossingover of circuliin the lateral fieldswere aminedby eye and with a hand lensfor signs consistentlyevident. We hypothesizedthat these of regenerationand potential legibility. If avail- compressedcirculi were annuli and took mea- able, four scalesfrom each fish were mounted surementsfi'om the focus to the midpoint of on glassslides and viewed on an Eberbach3pro- each.Circuli were quite closetogether in recent jector at 27x magnification.All measurementsannuli of the largestfish and exactdetermina- were taken along the anterior radius of the tion of the number of annuli became increas- inglydifficult as fish lengths exceeded 850 mm. Otolithsfrom 142 red snapperscollected off a Reference to trade names does not imply en- the Carolinas were examined and 90% were dorsementby the NationalMarine FisheriesService. legible.Left and right otolithsfi'om eight ran- SEDAR24-RD33

GROWTH AND MOR2'ALYi'Y OF RED SNAPPERS 467

Gulf of Mexico 2.C

1.1• /'

1.1• o__------o• .• •o,..,..---.•// x\ \ \ ,.•-•'• o 1.• 4 - -'"*. // x •X., /' o/ •o

1.(

0.8

0.6

0.4

0.2 i i i i i I i i i I i I

2.2 Atlantic o 2.C

1,6

1.zl 4 \. / \ / o.,,...... "'•--.,,/'

1.; 1.œ \ x"/ __,- 0.• ... 6 •<...... • V / ,--- • • .... 50' •¾.. --... /%/ •.•._...•-.'-*•...... •=:-•i•7:;'-"---• ...... 0.1• \:.,. - ,-/..

I JAN FEBi MAR i APR I MAY i JUN I JUL i AUG I SEP I OCT I N •) V DEC i

MONTH F•OOR]•1.•Mean monthlyscale increment for red snappersaged 2 to 6 yearscollected in theAtlantic Ocean (1974- 1978) and theGulf of Mexico(1978-1979). TheApril incrementfor 4-year-oldred snappers is a singleobservation. domly selectedfish of 53 to 523 mm were read (N = 10; SD = 9.41), 68.2 mm in August (N = independentlyand agreed; thereafter we sec- 22; SD = 15.4),82.7 mm in September(N = 34; tionedonly the left otolithfrom the remaining SD = 16.7), and 103.5 mm in October (N = 11; fish.Otolith sectionsviewed in transmittedlight SD = 28.5). Samplesof otoliths were not avail- containednarrow opaque zones and wider hya- able for all months,but 18 1-year-oldfish col- line zones. The opaque bands were distinct lectedduring June in the Carolinasshowed the around the entire surface of the medioventral beginningof new growthjust beyond an an- arm of the section and were counted as annuli. nulus at the sectionedge, indicatingthat scale Otolith annuli alsowere mostdifficult to sepa- and otolith annulus formation occurs simulta- rate at older ages. neously. To test the hypothesisthat observedrings Given simultaneous annulus formation, scales were formed only onceannually, we calculated and otoliths obtained from the same fish should mean monthly marginal growth on scalesby show the same number of annuli, and mean month of collectionfor all aged fish. Minimum lengthsat each age as determined from scales values were unimodal for all age classes;fish and otoliths obtained from fish within the same from the Atlantic areas showed annulus for- populationshould agree. For 43 fish,we deter- mation during April and May, whereasGulf of mined age from scalesand otoliths indepen- Mexicofish formed annuli duringJune andJuly dently. The agreementwas 77%. Of the ten (Fig. 1). Young-of-yearfish collectedin West disagreements,nine involveda differenceof 1 Florida had mean lengthsof 36.2 mm in July year and one a differenceof 2 years.Observed SEDAR24-RD33

468 NELSON AND MANOOCH

T^BLE1.--Mean range and standard error of themean (SE) of observedtotal length in mmfor redsnappers aged by otolith•and scales from theCarolinas 1974-1978.

Agedby scales Agedby otoliths Mean Length Mean Length Age N length range SE N length range SE 1 27 224 137-336 13.2 18 188 137-335 12.2 2 14 379 288•t27 9.8 16 337 218•t01 11.2 3 15 453 391-497 8.7 7 457 395•t90 14.3 4 49 536 435-610 5.5 13 536 460-570 8.2 5 75 577 515-718 4.0 10 579 510-630 13.1 6 65 629 520-732 5.5 8 657 612-685 8.4 7 35 684 605-795 7.4 11 721 645-755 8.6 8 12 753 692-815 11.2 7 769 731•805 10.2 9 7 786 720-806 12.0 7 814 800-840 5.3 10 4 845 820-875 13.2 4 844 812-875 15.3 11 14 855 820-910 6.7 4 841 824--850 6.1 12 8 893 860-930 9.4 1 830 13 2 867 856-878 11.0 3 873 856-885 6.0 14 2 943 930-955 12.5 1 888 15 1 1,025 2 910 890-930 20.0 16 1 978

Age 5 50 50 n=1480 n=516 40 40 30 30 20 20 10 10 6ø½I Age1 0 60 40 5O Age6 n=1364 Z 30 4(• 20 30 O 10 20

[i. 0 10 I- I Age2 Z 60 I11 356 5O Age 7 I11 40 n=994 50 Q. 3O 4• 2O 30 10 Age3206• n=233 Age8 n=120

FOCUS TO ANNULUS (ram X 27) FICURE2.---Frequency distributions ofmeasurements fromscale focus to each annulus for redsnappers aged 1 to8 years. SEDAR24-RD33

GROWTH AND MORTALITY OF RED SNAPPERS 469

T^BL• 2.--Weightedback-calculated mean total length (TL) in mmand mean annual increment (MAl) for redsnappers byarea and age.

Age Area I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Louisiana TL 137 267 379 469 546 613 665 707 751 783 794 891 906 (scales) MAI 137 130 112 90 77 67 52 42 44 32 11 N 402 382 247 170 131 111 86 53 22 11 4 1 1

West Florida TL 177 298 390 470 538 597 642 675 723 762 784 (scales) MAI 177 121 92 80 68 59 45 33 48 39 N 443 424 301 198 121 92 67 33 15 7 1

East Florida TL 146 266 354 448 527 601 652 690 714 733 780 759 (scales) MAI 146 120 88 94 79 74 41 38 24 19 47 N 357 288 190 110 72 36 29 18 11 7 2 1

Carolinas TL 153 275 357 467 542 607 667 723 748 780 803 830 864 913 969 971 (scales) MAI 153 122 82 110 75 65 60 56 25 32 23 27 34 49 56 N 331 304 290 275 226 151 86 51 39 32 28 14 6 4 2 1 (otoliths) TL 150 266 356 449 525 585 653 715 770 792 821 855 882 904 939 MAI 150 116 90 93 76 59 69 62 55 22 29 34 27 49 35 N 112 94 78 71 58 48 40 29 22 15 11 7 6 4 2

All areas TL 154 278 371 465 540 606 658 703 740 773 800 829 870 913 969 pooled MAI 154 124 93 94 85 66 52 45 37 33 27 29 41 43 56 (scales) N 1,533 1,398 1,028 753 550 390 268 155 87 57 35 16 7 4 2

lengthsof fishaged by scalescompared well with ment betweencaptures was 6 mm TL. Our data observedlengths of fishaged by otolithsfor the from East Florida indicate mean monthly Carolinas (Table 1). Analysisof variance re- growth incrementsof 5.2 mm TL for fish be- vealedno significant(P • 0.65) main or inter- tween3 and 9 yearsof age. active effects betweenlength at age and the We feel the data supportour hypothesisthat structureused in determiningage. observedrings are annual marks.Both scales Observedage as determinedfrom scalean- and otolithsare usefulfor estimatingage, but null was associateddirectly with increasing scalesare easierto collectand process.The use length of red snapperfrom all areas.Frequency of both scale and otoliths is advised when sam- distributions of the measurements from scale ple sizesare small,and for fishgreater than 800 focusto each annulusexhibited single modes mm TL. for eachage, a consistentincrease in modalfish Growth length with increasingannulus number, in- We determinedgrowth rates from observed creasingoverlap of distributionsas annulus lengths and back-calculatedlengths. Because number increased,and decreasinglength in- fish from the Gulf and fish from the Atlantic crements between modes as annulus number were not sampledequally over the sametime increased(Fig. 2). span,observed length is not a reliableindicator Further support for our age determinations of growth differences. Therefore, we com- can be taken from two taggingstudies. Beau- paredonly back-calculatedlengths. mariage(1969) reportsa mean annual growth In order to back-calculatelengths, we first increment equivalentto 95 mm TL for red determinedthe relationbetween fish length and snappers291490 mm TL tagged and recap- scaleradius. This was best describedby the tured on the west coast of Florida. Our data regressionof log•0scale radius on log•0fish TL. (Table 2) indicate mean annual increments of Analysisof covariancerevealed no significant 98 mm TL for fish in the samesize range. Moe effectassociated with sex,but a highlysignifi- (1969) mentionsthe return of a red snapper5.8 cant (P • 0.001) effect associatedwith area of yearsafter it wastagged at 370 mm TL near St. collection.The equationsdescribing the fish- Augustine,Florida. The monthlygrowth incre- length: scale-radiusrelation for eacharea were SEDAR24-RD33

470 NELSON AND MANOOCH

TABLE3.--Gr0wth parameters and asymptotic standard errors (ASE) calculated from the von Bertalanffy growth equations (Lt = L©(1- ½-Ket-t0))for redsnappers, 1974-1979, by area. L• is theaverage length at theasymptotic age predictedby the equation, K isthe growth coefficient, and to is the hypothetical ageat whichfish would have zero length - if theybad always grown in themanner described bythe equation. Gulf of Param- Mexico, Atlantic Ocean, All areas, eter Louisiana WestFlorida EastFlorida Carolinas pooled pooled pooled L• 950 941 970 970 941 975 970 ASE 13.5 16.1 19.9 10.5 11.0 9.3 8.0

K 0.175 0.170 0.155 0.165 0.170 0.160 0.162 ASE 0.005 0.005 0.006 0.004 0.004 0.003 0.002

to 0.10 -0.10 -0.01 -0.01 -0.10 0.00 0.010 ASE 0.021 0.025 0.025 0.024 0.018 0.017 0.012

Louisiana ceivedmuch attentionregarding the appropri- logmTL = 1.109logmSCL + 0.014; atenessof its application(Knight 1968; Yama- N=403; r 2=0.98; guchi 1975;Bayley 1977). In thisstudy, the plot Panama City of observedlength on age aswell as the length- 1og•0TL = 1.044log•oSCL + 1.787; frequency distributionsof annuli suggested N= 513; r 2= 0.98; asymptoticgrowth. No pronouncedinflection waspresent in the curveover the youngerage Daytona groups,and samplesincluded representatives log•0TL = 1.074log•oSCL + 0.092; of age groups 1-16 collectedthroughout the N= 356; r •= 0.96; year. Therefore, we chosethe von Bertalanffy Carolinas model as an empiricallybased description of logmTL = 1.067log•oSCL + 0.117; growth.The equationis N = 332; r • =0.96; Lt = L•(1 - e-•(t to))+ Et; SCL = magnifiedscale radius (mm x 27). For whereLt is lengthat aget (years),K is the growth each fish, all annuli measurements were substi- coefficient, L• is the average length at the tuted for SCL in the appropriateequation. The asymptoticage predictedby the equation,to is weightedmean TL at the time of annulusfor- the hypotheticalage at which fish would have mation and weightedmean annual growth in- zero length if the fish had alwaysgrown in the crementfor eachage group then were calcu- mannerdescribed by the equation,and Et is the lated (Table 3). error associatedwith age determination. The mean back-calculatedlengths are close ProcedureNLIN, whichuses the Marquardt to thosepresented by Moseley(1966) for Texas method as containedin the StatisticalAnalysis and Futch and Bruger (1976) for Florida. Systempackage (Helwig and Council1979), was Moseleyshowed annual growth incrementsof usedto fit length at time of eachannulus for- approximately90 mm standardlength from the marionto the growthequation. The procedure end of the first year through the end of the utilizesweighted least-squares estimates of the fourth year. Futch and Bruger reported parametersof nonlinearmodels, calculates re- weightedmean lengthsof 151, 316, 401,473, siduals,and regressesresiduals on the partial and 554 mm TL (convertedfrom FL to TL) for derivatives of the model until the iterations ages1 through 5 caughtoff Clearwater,Flori- converge.All estimatesfor L • are consistentwith da. In our study,first-year growth washighest maximumobserved lengths in eacharea (Table for WestFlorida ( 177 mm), followedby that for 3). the Carolinas (153 mm), East Florida (146 mm), The effect of area of collection on the fit of and Louisiana(137 mm). There did not appear the growth equations was evaluated with a to be any predictablepattern of increaseor de- modification of the maximum-likelihood-ratio creaseof mean length at the early ages(Lea's test (Neeter and Wasserman 1974; Nelson phenomenon)(Tesch 1968; Nelson 1980). 1980). The von Bertalanffyequation was fit to The von Bertalanffygrowth equation has re- the data by area, and the residual sums of SEDAR24-RD33

GROWTH AND MORTALITY OF RED SNAPPERS 471 squaresand degreesof freedomwere totalled MortalityEstimates to give a full-modelresidual sum of squares, Age-lengthkeys derived from agedfish were RSS(F)and degreesof freedom,df(F). The yon applied to all fish collected,and catchcurves Bertalanffyequation was then fit to all data then were constructed(Fig. 3). Fishused in age combined to obtain a reduced-model residual determinationswere selected randomly from the sum of squares,RSS(R) with corresponding total catch, and keys were constructedsepa- degreesof freedom,df(R). An F-statisticwas rately for eacharea, sothis methodshould not calculatedby the equation: bias the resultingestimates of total mortality (Westrheim and Ricker 1978). F = RSS(R)- RSS(F). RSS(F) Analysisof covarianceshowed no effect due dr(R) - dr(F) dr(F) ' to yearon the slopeof theregressions from East The effect of area wassignificant (P (0.01) Florida and the Carolinas,and the regression for all study areas, for pooled Atlantic data of Z on year of collectionshowed no trend of comparedto pooledGulf data,for areaswithin increaseor decreasein mortalityover the years the Gulf, and for areas within the Atlantic. Al- sampled,so yearswere pooled to yield the best thoughsignificant statistically, the actualdiffer- estimate(Table 4). Area did have an effect on encesare small.Data for populationswithin the the mortalityregressions (P < 0.01). Louisiana Gulf and within the Atlanticare certainlysim- showedthe highestrate of instantaneoustotal ilar enoughto be pooled.Thus, red snappers mortality(0.78 by regression;0.94 by Robson of ages1 through5 from the Atlanticwould and Chapman 1961), East Florida an interme- have predictedlengths of 144, 264, 369, 458, diate rate (0.50; 0.50), and West Florida (0.42; and 533 mm at the time of annulus formation, 0.44) and the Carolinas(0.39; 0.42) the lowest and thosefrom the Gulf 160,283,385,472, and rates.The meanage in the catchwas 2.4 years 546 mm. By the timea fishfrom the Gulf was for Louisiana,4.1 yearsfor West Florida, 3.0 10 yearsof ageit wouldreach 772 mm and a years for East Florida, and 6.1 years for the similarlyaged fish from the Atlanticwould be Carolinas.There wasa strongnegative corre- 778 min. lation (r = -0.83) betweeninstantaneous mor- Analysisof covarianceindicated that the re- talityand meanage. lationshipbetween length and weightdiffered Studiesof other fish speciesin the deep-reef betweenareas (P (0.01), but not betweensexes taxocene(red porgy Pagruspag'rus; vermilion (P 3 0.50). Althoughstatistically significant, the snapperRhomboplites aurorubens) have demon- biologicalimplications of area differencesseem strated depth-specificsize segregation (Ma- minimal.On the average,a 500-ramfish would nooch and Huntsman 1977; Grimes 1978). The weigh1,842 g if takenoff WestFlorida, 1,889 fishinggrounds off the Carolinascan be divid- g if takenoff EastFlorida, and 1,951g if caught ed into inshore ((35 m depth) and offshore off the Carolinas.The equationsare ()35 m depth) habitats.To evaluatethe size distributionof red snapperwe comparedlength West Florida W = 1.82 x 10 distributions, catch curves, and estimated mor- N= 143; r e= 0.99; tality(Z) for fishcollected inshore and offshore. East Florida W = 1.36 x 10 5 TL:•.in7; Length distributionswere similar. The modal N= 142; r == 0.99; length classeswere 575-600 mm inshoreand 600-625 mm offshore. Both groups of data Carolinas W = 3.15 x 10 5 TL'2'887; showeda modalage of 6 years,and analysisof N = 177; r == 0.99; covariancerevealed no differencein mortality between inshore (Z = 0.42) and offshore (Z = All areas,pooled W = 2.04 x 10 • TLZ'":•; N=462; r z=0.99; 0.38) data. Few red snapperswere taken in- shore,and mostof thesewere from the deepest W is weight in grams. inshore reefs. No size segregationby depth is The data do not allow an assessment of vari- evident. ation in growth betweenindividual reef sys- The Pauly (1980) method of determiningM tems.Overall geographic patterns are probably dependson the mean annual temperature(T), the resultof differing ratesof growthon many and parametersof thevon Bertalanffy equation distinctreef systems. as follows: log•0M= 0.654 logn•K- 0.280 SEDAR24-RD33

472 NELSON AND MANOOCH

East Florida 81 Louisiana y=-O.51X+8.9 /]\\., y=-0.78X+8.8 r=-0.97

4 \• .

z u.i o , I u.i u.

West Florida Carolinas o y=-O.42X+8.5 y=-O.38X+6.9 r=-0.98 r=-0.97

/// •'•'•

, I • I , [ t I i i I • I • I , I , I , I • I , I , I • I 2 4 8 8 10 12 14 2 4 8 8 10 12 14 16

AGE FIGURE3.--Catch curves for redsnappers from all areas.The solid line described bythe equation (y = bx + a; y = loge frequency;x = agein years)indicates the age range used in regressionestimates of total annual instantaneous mortality (Z). Z is equalto theabsolute value of theslope (b) of theregression line. logj, L• + 0.463 logj0T. In calculatingM, we method is that there is no way to include infor- used50-year mean temperatures with 75 m as mationabout the age-sizestructure of the pop- a standarddepth. The mean annual tempera- ulation in the calculation of M. ture was 22 C for Louisiana, 22 C for West Both methodsused to calculateZ are subject Florida, 24 C for East Florida, and 21 C for the to the assumptionsthat: (1) recruitment and Carolinas (Churgin and Halminski 1974a, survivalare constantin the populationunder 1974b).All estimatesof M are virtuallythe same. examination;(2) fishingeffort has been con- Inasmuch as there is little or no variation in the stantover the time spanrepresented by the old- abioticregime and predator and prey estfish in the ageseries examined; (3) the sam- assemblages among deep-reef habitats ple is random; and (4) for the populationin throughoutthe range of our study, we would questionall sizeclasses should be presentwithin expectto find little differencein natural mor- the area of collection. To address the first as- tality. A possibleshortcoming of the Pauly sumptionwe examineda seriesof length-fre-

TABLE4.--Modal andmean age of thecatch in years,and instantaneous rates of total(Z), natural(M), andfishing (F) mortalityfor all areas.M wasdetermined bythe method of Pauly (1980), Z bycatch-curve regression and by the method ofRobson and Chapman (1961), and F bydifference.

Regression Robsonand Chapman Area Modal age Mean age M Z (SE) F Z (SE) F Louisiana 2 2.4 0.20 0.78 (0.! 1) 0.58 0.94 (0.04) 0.74 West Florida 2-3 4. ! 0.19 0.42 (0.05) 0.23 0.44 (0.09) 0.25 East Florida 2-3 3.0 0.19 0.50 (0.07) 0.31 0.50 (0.04) 0.31 Carolinas 5 6.1 0.18 0.39 (0.05) 0.21 0.42 (0.04) 0.24 SEDAR24-RD33

GROWTH AND MORTALITY OF RED SNAPPERS 473 quency histogramsand catch curves for the 1979). Manooch and Huntsman (1977), dis- Carolinas(1974-1978) and East Florida (1976- cussinga similarsituation regarding red porgy 1978). In all cases,modal age wasconstant by from off the coasts of North Carolina and South area over the yearsexamined, the mean age in Carolina, concludedthat mortality was largely the catch did not differ over the years within the result of natural causesand that fishing area, and there were no unusually weak or playedonly a minor role. strongyear classesapparent during the period Conclusions examined.As previouslymentioned, Z did not vary betweenyears in theseareas. We feel that The red snapper, although resident in a the first assumptionprobably is satisfied,and functionally tropical habitat (Huntsman and that any variation in survival or recruitment MacIntyre 1971;Smith et al. 1975),can be aged within the ageseries examined is random. Mor- by annular markingson both scalesand oto- tality estimatesderived by the regressionmeth- liths. We found that annulus formation occurs od are least sensitive to minor violations of this in April and May in the Atlanticand June and assumption(Ricker 1975;Robson_and Spangler July in the Gulf of Mexico.The formation of 1978),and all statisticaltests are performedon annuli does not seem to be associated with win- theseestimates. Fishing effort, as indicatedby ter temperature minima. Camber (1955) the number of activevessels, did not change thoughtthat red snappersceased feeding dur- appreciablyduring the courseof this study in ing the spawningseason, and Moseley(1966) the Carolinas, East Florida, or West Florida. We believed a feeding cessationduring spawning do not think effort in the Gulf commercial fish- periods accountedfor annulus formation. Al- ery has changedsignificantly. Whether hook- though our data support this hypothesis,the and-linefishing is a random method of collec- presenceof an annuluson age-1(immature) fish tion is not clear. It certainlyexcludes fish not could indicate that someinnate physiological activelyfeeding. Although hook sizesused do rhythmn, correlated with external factors that excludesmaller fish, they do not seemto affect controlspawning, might be the stimulusfor an- sizefrequencies in the rangeused for mortality nulusformation. Hickling (1935) suggestedthis estimates.We think our sampleis a true ran- explanation in connectionwith annulus for- dom sampleof the catch,and that the catchis mation in juvenile hake Merlucciusmerluccius. representative of red snapper populations Red snappersbecome sexually mature after 2- within the age seriesexamined in each area. 3 years(Bradley and Bryan 1975). Fishing mortality appearsto be the major In general, growth of the red snapper and causeof differencesamong total mortality rates. other speciescommon to the outer reef habitats Catchesfor the intensivecommercial fishery off is similar.The K valuesof the growthequation Louisianaand for the recreationalfishery off for red snapperranged from 0.155 to 0.175. EastFlorida, wherethe reefsare fairly closeto Other workers have shown similar K values for shore and easilyaccessible, are similar in age the vermilionsnapper (0.198: Grimes 1978),the structureand reflectthe highestmortality rates. yellowtailsnapper Ocyurus chrysurus (0.160: Pie- The recreational fisheries off West Florida and dra 1969), the red grouper Epinephelusmorio the Carolinas are located farther offshore and (0.179: Moe 1969), the red hind Epinephelus require larger boats and more sophisticated guttatus(0.180: Burnett-Herkes 1975), and the electronicgear to be exploitedsuccessfully. Both white grunt Haemulonplumieri (0.1084: Ma- of theseareas exhibit lower mortalityrates than nooch 1978). Growth for these speciesis ini- the other areasand catchesare composedof tially fast and then slowssteadily, as larger sizes older,larger fish. Fisheries in the Gulf and East associatedwith long life expectancyare at- Floridahave muchlonger histories than the one tained.Although most predatory pelagic fishes off the Carolinas ((;amber 1955; Huntsman from these same general areas do not live as 1976).The agestructure of red snapperoff the long asspecies in the reef-fishcomplex, growth Carolinas reflects the relatively light fishing is fast throughout their lives, with K valuesof pressureand shorthistory of the fishery.A large from 0.23 to 0.34 for bluefish Pomatomus salta- fraction of the total biomassconsists of large, trix and 0.48 for Spanish mackerel Scombero- older individuals,a conditionindicative of sys- moru•maculatus (Huntsman and Manooch 1978). tems subjectto little exploitation(Regier et al. In Brazil, the Caribbeanred snapperappears SEDAR24-RD33

474 NELSON AND MANOOCH to reach the samesize as the red snapper,but the Texas continentalshelf. Pages 248-288 in growsmore slowly,with a coefficient(K) of 0.090 Bullisand Jones(1976). (de Menezesand Gesteira 1974). BROOKS,H. K. 1962. Observations on the Florida The total annual mortalityand age structure middle-grounds.Geological Society of America SpecialPaper (RegionalStudies) 68:65-68. of the catch seems to reflect the amount of fish- BULLIS,H. R., ANDA. C. JONES,editors. 1976. Pro- ing pressurein each study area. The fishing ceedings:colloquium on snapper-grouperfish- componentof mortalityappears to accountfor ery resourcesof the western central Atlantic the variation in Z between all areas. Ocean. Mississippi-AlabamaSea Grant Consor- Significantdifferences in mortality and age tium, Report 17, publishedin cooperationwith structureof the populationsof red snapperin Texas A&M University Sea Grant College, Bi- loxi, Mississippi,USA. the four areas would have more impact on BURNETT-HERKES,J. 1975. Contributionto the biol- managementmodels than the slight,but signif- ogyof the red hind, Epinephelusguttatus, a com- icantdifferences in growth.Similar differences merciallyimportant serranidfish from the trop- mayalso exist for otherspecies in the deep-reef ical western Atlantic. Doctoral dissertation. taxocene. Universityof Miami, CoralGables, Florida, USA. C^MBER,C. I. 1955. A surveyof the red snapper Acknowledgments fishery of the Gulf of Mexico, with specialref- erenceto the CampecheBanks. Florida Board of This researchwas made possible through the Conservation Marine Research Laboratory effortsof personnelfrom the NationalMarine Technical Series 12. Fisheries Service, Southeast Fisheries Center CHURGIN,J., ANDS. J. HALMINSKI.1974a. Temper- laboratories in Beaufort, North Carolina and ature salinity,oxygen, and phosphatein waters off United States. Volume 1. Western North At- Panama City, Florida. Specialappreciation is lantic. United States National Oceanic and At- extended to Carl Saloman who directed field mosphericAdministration, Environmental Data samplingin the Gulf of Mexico.We alsowould and InformationService Key to Oceanographic like to thank David Dickey,Department of Sta- Records Documentation 2. tistics,North Carolina StateUniversity, for aid CHURGIN,J., ANDS. J. HALMINSKI.1974b. Temper- in the analysis,and WilliamW. Hasslerand John ature, salinity,oxygen, and phosphatein waters off United States. Volume 2. Gulf of Mexico. M. Miller, Department of Zoology,North Car- United StatesNational Oceanicand Atmospheric olina State University, for valuable comments Administration, Environmental Data and Infor- and criticism. mation ServiceKey to OceanographicRecords Documentation 2. References DAWSON,C. E. 1963. Length and weightrelation- ALLEN,D. M., ANDJ.E. TASHIRO.1976. Statusof the shipsof youngred snappersfrom the northern U.S. commercialsnapper-grouper fishery. Pages Gulf of Mexico. Transactions of the American 41-76 in Bullisand Jones(1976). FisheriesSociety 92:310-311. BatkEY,P. 1977. A methodfor findingthe limitsof DE MENEZES, M. F., AND T. C. V. GESTEIRA. 1974. applicationof the yon Bertalanffygrowth model Idade e crescimentodo pargo, purpureus and statisticalestimates of the parameters.Jour- Poey,do notre e nordestedo Brasil.Arquivos de nal of the Fisheries Research Board of Canada Ci6ncias do Mar 14:81-85. 34: 1979-1984. FUTCH,R. B., ^NoJ. E. BRUGER.1976. Age, growth, BEAUMARIAGE,D. S. 1969. Returns from the 1965 and reproduction of red snapper in Florida Schlitztagging program including a cumulative waters.Pages 165-184 in Bullisand Jones (1976). analysisof previousresults. Florida Department GRIMES,C. B. 1978. Age, growthand length-weight of Natural Resources Marine Research Labora- relationshipof ,Rhomboplites tory TechnicalSeries 59. aurorubens,from North Carolina and South Car- BERRY,F. H., D. W. LEE,ANDA. R. BERTALINO.1977. olina waters. Transactions of the American Fish- Progressin bluefin tuna ageingattempts. Inter- eriesSociety 107:454-456. national Commission for the Conservation of HELWIG,J. T., ^NOK. A. COUNCIL,editors. 1979. SAS Atlantic Tunas ScientificPapers 6:305-317. usersguide, 1979 edition. SAS Institute, Raleigh, BRADLEY,E., ANDC. F. BRYANIII. 1975. Life history North Carolina, USA. and fisheryof the red snapper(Lutjanus campe- HICKLING,C. G. 1935.Seasonal changes in the ovary chanus)in the northwesternGulf of Mexico:1970- of immature hake, Merlucciusmerluccius L. Jour- 1974. Proceedingsof the Gulf and Caribbean nal of the Marine BiologicalAssociation of the Fisheries Institute 27:77-106. United Kingdom21:311-318. BRIGHT,T. j., ANDR. REZAK.1976. Fishing banks of HUNTSM^N,G. R. 1976. Offshore bottom fisheriesof SEDAR24-RD33

GROWTH AND MORTALITY OF RED SNAPPERS 475

the United Statessouth Atlantic coast. Pages 192- tific Translations,Jerusalem, Israel. TT69-59016, 221 in Bullisand Jones(1976). NationalTechnical Information Service, Spring- HUNTSMAN,G. R., ANDI. G. MACINTVRE.1971. Trop- field, Virginia, USA. ical coral patchesin OhslowBay. Underwater REGIER,H. A., J. E. PALOHEIMO,AND V. F. GALLUCCI. Naturalist 7(2):32-34. 1979. Factors that influence the abundance of HUNTSMAN, G. R., AND C. S. MANOOCH III. 1978. largepiscivorous fish. Pages 333-341 in H. Clep- Coastalpelagic and reef fishesin the SouthAt- per, editor. Predator-prey systemsin fisheries lanticBight. Pages 97-106 in H. Clepper,editor. management.Sport FishingInstitute, Washing- Marine recreationalfisheries 3. SportFishing In- ton, District of Columbia, USA. stitute,Washington, District of Columbia,USA. RICKER,W. E. 1975. Computationand interpretation KNIGHT,W. 1968. Asymptoticgrowth: an example of biologicalstatistics of fish populations.Fish- of nonsensedisguised as mathematics. Journal of eries Research Board of Canada Bulletin 191. the Fisheries Research Board of Canada 25:1303- ROBSON, D. S., AND D. G. CHAPMAN. 1961. Catch 1307. curvesand mortality rates. Transactionsof the MANOOCH,C. S., IIl. 1978. Age, growth, and mor- AmericanFisheries Society 90: 181-189. tality of the white grunt, Haemulonplumieri La- ROBSON,D. S., ANDG. R. SPANGLER.1978. Population cbpbde,from North Carolinaand South Caroli- abundanceand survival.Pages 26-51 in S. D. na. Proceedings of the Annual Conference Gerking,editor. Ecology of freshwaterfish pro- Southeastern Association of Fish and Wildlife duction. Blackwell ScientificPublications, Oxford, Agencies30:58-70. England. MANOOCH,C. S.,III, ANDG.R. HUNTSMAN.1977. Age, SMITH, G. B., H. M. AUSTIN, S. A. BORTONE,R. W. growth,and mortalityof the red porgy,Pagrus HASTINGS,AND L. H. OGREN. 1975. Fishes of the pagrus.Transactions of the American Fisheries Florida middle ground with commentson ecol- Society106:26-33. ogyand zoogeography.Florida Marine Research MOE, M. A., JR. 1969. Biologyof the red snapper Publication 9. from the eastern Gulf of Mexico. Florida De- STRUHSAKER,P. 1969. Demersal fish resources: com- partment of Natural ResourcesMarine Research position,distribution, and commericalpotential LaboratoryProfessional Papers Series 10. of the continental shelf stocks off the southeast- MOSELEV,F. N. 1966. Biologyof the red snapper, ern United States.Fishery Industries Research Lutjanusaya Bloch, of the northwesternGulf of 4:261-300. Mexico. Publications of the Institute of Marine TESCH,F. W. 1968. Age and growth. Pages93-123 ScienceUniversity of Texas 11:90-101. in W. E. Ricker, editor. Methods for assessment NAKAMUA, E. L. 1976. Recreational fisheries for of fish productionin freshwaters. Blackwell Sci- snappersand groupersin the Gulf of Mexico. entificPublications, Oxford, England. Pages77-85 in Bullisand Jones (1976). WADE,W. 1978. Studiesof migratorypatterns of NEETER,J.,ANDW. WASSERMAN. 1974. Applied linear commercially valuable marine finfish of Ala- statisticalmodels. Richard D. Irwin, Homewood, bama. FederalAid in FishRestoration Project Illinois, USA. 2-280-R, CompletionReport, United StatesNa- NELSON,R. S. 1980. Growth and mortalityaspects of tional Marine FisheriesService, St. Petersburg, natural populationsof red snapper, Luteanus Florida, USA. campechanus,in the westcentral Atlantic and Gulf WAKEMAN,J. M., C. R. ARNOLD,D. E. WOHLSCHLAG, of Mexico. Master's thesis. North Carolina State ANDS.C. RABALAIS.1979. Oxygenconsumption, University,Raleigh, North Carolina,USA. energyexpenditure and growthof the red snap- PAUY, D. 1980. On the interrelationshipsbetween per. Transactionsof the American FisheriesSo- naturalmortality, growth parameters, and mean ciety 108:288-292. environmentaltemperature in 175 fish stocks. WESTRHEIM,S. J., ANDW. E. RICKER.1978. Bias in Journal du Conseil,Conseil International pour usingan age-lengthkey to estimateage-frequen- L'Explorationde la Mer 39:175-192. cy distributions. Journal of the Fisheries Re- PIEDA,G. 1969. Materialson the biologyof the yel- search Board of Canada 35: 184-189. lowtailsnapper (Ocyurus chrysurus Bloch). Pages YAMAGUCHI,M. 1975. Estimatinggrowth parameters 251-269 in A. S. Bogdanov,editor. Soviet-Cu- from growthrate data. Oecologia (Berlin) 20:321- ban fisheryresearch. Israel Programfor Scien- 322.