Age and Growth Ofthe Leopard Grouper, Mycteroperca Rosacea, in the Southern Gulf Ofcalifornia, Mexico1 J

Age and Growth ofthe Leopard Grouper, Mycteroperca rosacea, in the Southern Gulf ofCalifornia, Mexico1 J. Gabriel Diaz-Uribe,2 Juan F. Elorduy-Garay,3 and Ma. Teresa Gonztilez-Valdovinos+

Abstract: Growth of the leopard grouper, Myeteroperca rosacea (Streets, 1877), was analyzed in its natural habitat. Age determination was based on the reading of otoliths, and the method was validated under three main criteria: (1) pro­ portionality, (2) seasonality, and (3) concordance with another method. Otolith growth is proportional to organism growth, with a slight degree of allometry, and the otolith registers the growth of the individual, even at advanced ages. The opaque growth zone in the otolith is deposited once a year, between July and October. Thus, taken together, one opaque and one hyaline mark represent an annual cycle. Back-calculated lengths-at-age agreed reasonably well with ob­ served lengths-at-age at the time of capture, considering that back-calculated lengths represent an exact age (birthday), and observed lengths are taken at an intermediate age between birthdays. Fish length and otolith age data were fitted to the von Bertalanffy growth function by two methods: (1) linear regression (Ford-Walford and Beverton), using transformed data, and (2) nonlinear re­ gression, by iteration. Although the nonlinear regression gave a fit with unbiased error, parameters resulting from linear regressions had a better biological meaning for the species. The resulting parameters were compared with those reported for other species of the family Serranidae.

THE LEOPARD GROUPER, Mycteroperca rosacea Sur) to Banderas Bay (Jalisco), including the (Streets, 1877), is found in subtropical shal­ Gulf of California (Thomson et al. 1987, low coastal waters, in a restricted geographi­ Heemstra 1995). The family Serranidae cal area from Magdalena Bay (Baja California forms an important part of the catches from the artisanal fishery of Baja California Sur, and the leopard grouper surpasses other spe­ 1 This study was partially funded by Direcci6n Gen­ cies both in volume and frequency of catches eral de Investigaci6n Cientifica y Superaci6n Academica­ (Rodriguez-Medrano 1990). The biology of Secretaria de Educaci6n Publica (DIGICSA-SEP) and the leopard grouper is poorly known. There Direcci6n de Estudios de Posgrado e Investigaci6n (DEPI)-966509 research projects. ].F.E.-G. received fel­ are works referring to its description and tax­ lowships from Comisi6n para el Fomento de Actividades onomic status (Rosenblatt and Zahuranec Academicas (COFAA) and Estimulo al Desempeiio 1987), distribution and abundance in La Paz Docente (EDD) at the Instituto Politecnico Nacional. Bay, ReS. (Villavicencio 1983), and feeding Manuscript accepted 9 August 2000. habits (Bermudez-Almada and Garcia-Laguna 2 Instituto de Recursos, Universidad del Mar, AP. 47, Puerto Angel, Oaxaca, 70902, Mexico. 1985), but previous studies on age and growth 3 Departamento de Pesquerias y Biologia Marina, for this species are lacking. Centro Interdisciplinario de Ciencias Marinas (CICI­ Individual growth rate, age structure, and MAR), AP. 592, La Paz, B.C.S., 23000, Mexico (tele­ age offirst maturity and recruitment are use­ phone: 52 (1) 122 53 44; fax: 52 (1) 122 53 22; E-mail: [email protected]). ful parameters for the evaluation of pop­ 4 Departamento de Biologia Marina, Universidad ulations, and the confidence level at which Aut6noma de Baja California Sur, AP. 19-B, La Paz, they are estimated depends, in large part, B.C.S., 23080, Mexico. on the method used for age determination (Ricker 1979, Sparre and Venema 1992). The Pacific Science (2001), vol. 55, no. 2:171-182 reading of scales and otoliths has been the © 2001 by University of Hawai'i Press. most frequently used method for age deter­ All rights reserved mination of fish. In slow-growing, long-lived

171 172 PACIFIC SCIENCE· April 2001 species, the growth marks are unreadable (GW). Whenever possible, total weight near the edge of scales, causing an underesti­ (TW) and sex were also determined. After mation of age. Otoliths have proven more measuring, both sagittae otoliths were ex­ useful for age determination in this kind of tracted by an oblique cut at the cranial basis. organism (Boehlert 1985, Devries and Frie At the laboratory, otoliths were washed with 1996). Considering the long life cycle and soap and tap water and stored dry. slow growth rate of species composing the Before age determination, a small sub­ family Serranidae (Manooch 1987), we con­ sample of otoliths was subjected to different sidered that otoliths would be the most reli­ combinations of high temperatures (300­ able structure for age and growth analysis of 400°C) and times (5-30 min) to improve the the leopard grouper. Even the reading of contrast between marks (Elorduy-Garay and otoliths does not ensure a priori that this Diaz-Uribe 1994). The temperature-time method of age determination is dependable. combination with best contrast was chosen It must be demonstrated that the marks used and applied to all otoliths considered in the to determine age are related to a specific time study. interval, and that they can be observed for Otolith readings and measurements were most of the life cycle of the species (Beamish made using a dissecting microscope, an ocular and McFarlane 1983, Casselman 1983). Only micrometer, and reflected illumination, at in this way can the bias in estimation of 10 x total magnification. Reading consisted growth parameters be reduced, or at least of the counting of growth marks and the known, and the consequences in resource determination of edge type (i.e., opaque or management appreciated (Beamish and hyaline). Each otolith was read twice in­ McFarlane 1983). dependently. When both readings were The von Bertalanffy growth function has identical the otolith was declared readable, proven to be a good model in studies of and the number of marks (rings) and edge growth because of the applicability of its pa­ type were assigned to the corresponding fish. rameters in more complex models describing Whenever there were differences, otoliths population dynamics (Sparre and Venema were submitted to a third reading. Ifthe third 1992). Because of the commercial importance reading agreed with either of the previous of the leopard grouper, the estimation of two readings it was assigned as definitive; if growth parameters is especially valuable be­ not, the otolith was considered unreadable cause of its relevance to fishery management. and discarded. In this study we analyze the validity of otolith Three types of morphometric relation­ reading for age determination ofMycteroperca ships were analyzed: TL versus SL, GW ver­ rosacea caught in the Bay of La Paz, B.C.S., sus TW, and TL versus GW. The first and vicinities, and the fitting of data to the two were evaluated using linear regression by Von Bertalanffy growth function is also least squares and Student's t-test under the presented. null hypothesis a = 0, b = 0, and R2 = 0 (ex = 0.05). The third relationship was eval­ uated by the potential model Y = axb using MATERIALS AND METHODS nonlinear regression by least squares, with the From January 1991 to June 1992 individuals program FISHPARM 3.1 (Prager et al. 1989); of M. rosacea were selected monthly from in this case, Student's t-test was ran under the both commercial and experimental catches null hypothesis b = 3 and R2 = 0 (ex = 0.05). made around Espiritu Santo and Cerralvo The use of otoliths for age determination Islands, located off the western coast of the of M. rosacea was validated with three basic Gulf of California (Figure 1). A maximum criteria (Beamish and McFarlane 1983, Cas­ of 15 fish per length interval (50 mm), per selman 1983): month, was randomly selected from the (1) Proportionality between otolith growth catches. Each fish was measured in total (TL) and fish growth was tested by selecting a and standard (SL) lengths and gutted weight stratified subsample, as a function of fish Age and Growth ofMycteroperca rosacea . Diaz-Uribe et al. 173

San Jose I.

., = ~ San Fancisco I. t -i~ Espiritu Santo I.

Bay of La Paz

.' .:".-,_.r. -. "'.: ". "::.::.;.;;:.;:r.....,..~ ...... :-:. ..:.

1100 00'

FIGURE 1. Fishing areas from which the samples ofM. rosacea were obtained. length and independent of the month of cap­ and fish growth were adjusted to a potential ture. A maximum of 20 pairs of otoliths per model. length interval was selected. The otoliths (2) Seasonality in the formation of growth were measured in long diameter (OLD) and marks was analyzed by the monthly frequency long radius (OLR). Data on otolith growth of otolith edge type, using the total sample. 174 PACIFIC SCIENCE· April 2001

This information was compared with the sea to survey at sea. Individuals in the sample surface temperature in southern La Paz Bay. ranged from 283 to 975 mm TL, but 90% of The Secretaria de Desarrollo Social, Dele­ the fish were between 350 and 700 mm TL gaci6n RC.S., kindly provided temperature (Figure 2). Sex was determined for 16% of data. the sample, comprising fish 286-808 mm in (3) Concordance with another method was length (TL). None of the fish had virginal analyzed by the back-calculation of lengths at gonads. past ages, as a reference method. The sub­ Because of the small dispersion of data, the sample used was the same as that used in the linear model fitted to the TL-SL relationship proportionality analysis. At this stage, the explains nearly 98% of the variability radii at each growth mark (annuli) were (P < 0.01) (Table 1). The GW-TW rela­ measured in each otolith. Average length­ tionship presented the same behavior, with at-age was estimated using the equation: 99% (P < 0.01) of explained variance (Table 1). Ordinates at the origin were not sig­ nificantly different from zero (P > 0.05) in both regressions. Therefore, the standard where L j is the length of the fish when the length of the leopard grouper represents an otolith radius was Ri; a and b are the parame­ average of 86-88% of the total length, and ters of the regression offish length on otolith the gutted weight represents an average of radius from the proportionality analysis. 87-90% of the total weight. Back-calculated lengths were compared with The potential model fitted to the TL-GW observed lengths at the time of capture. relationship explains 97% of the observed Individual growth was calculated fitting variability (P < 0.01), so it has an important the age-length data to the Von Bertalanffy predictive value. The coefficient b = 2.97 is growth function: not significantly different from 3 (P> 0.05), L, = L",[1 - e-k('-to)j so the leopard grouper exhibits isometric growth (Table 1). where L, is the average length of the fish at Heating the otoliths at 325°C for 25 min age t; L", is the average maximum length of was the combination that rendered best re­ the analyzed stock; k is the growth coefficient; sults. Under these conditions, opaque rings and to is the parameter of origin of the turned yellow, and hyaline ones turned dark growth curve. brown. The immersion of these otoliths in Two separate methods were used to derive glycerin several minutes before the readings the growth function. In the first method notably improved the contrast between rings. (designated LRC) we followed the Ford­ Nearly 95% of the otoliths were used in age Walford procedure to calculate L", and the determinations. The remaining were consid­ Beverton procedure to calculate k and to ered unreadable, because oflack of agreement (Ricker 1975, Sparre and Venema 1992). The in the number of rings (2 %) or lack of second method utilized a nonlinear regres­ agreement both in the number of rings and sion analysis (NLR) with the program edge type (3 %). FISHPARM 3.1 (Prager et al. 1989). Confi­ Otolith growth is strongly correlated to dence intervals (a = 0.05) were calculated for overall fish growth. All the correlation co­ each parameter and compared with those efficients (r) were significant (P < 0.01), in calculated by the LRC method. spite of the fact that data present higher dis­ persion than the previous regressions (Table 2). In every correlation where TL is involved RESULTS the exponent is significantly greater than 1 In total 769 fish were sampled during the (P < 0.01), whereas in every correlation study period, except for the months ofJune, where GW is involved the exponent is sig­ November, and December 1991 when nificantly different from 3 (P < 0.01). This catches were low, and it was impossible for us means that there is a certain degree of allom- Age and Growth of Myeteroperea rosacea . Diaz- Uribe et at. 175

20

15

>. u c 10 Q) ::J rr ~ u.. 5

o 275 375 475 575 675 775 875 975 Length class (mm)

FIGURE 2. Frequency distribution ofsize classes (TL) for the whole sample ofM. rosacea. etry, in which the fish grows proportionally in that year, the tendency in the other months faster than the otolith. indicates that individuals with opaque otolith In 1991 the frequency of individuals with edges predominated (>50%) from June to opaque otolith edges was low in January and November. The frequency ofindividuals with February and highest between July and Oc­ opaque otolith edges was again low at the tober (Figure 3). Although there is no infor­ beginning of 1992. This pattern follows the mation for June, November, and December temperatures registered in the Bay of La Paz.

TABLE 1 Morphometric Relationships Involving Body Length and Weight in M. rosacea

Parameters

Relationship n a SE (a) b SE (b) R2

TL-SL 762 -10.583 12.998 0.872 4.36 x 10-3 0.981 GW-TW 111 5.997 120.312 1.128 8.51 x 10-3 0.994 TL-GW 762 1.43 x 10-5 1.58 X 10-6 2.970 1.70 x 10-2 0.973

Note: Total length (TL) - standard length (SL) (function, y = a + bx); gutted weight (GW) - total weight (TW) (function, y = a + bx); total length (TL) - gutted weight (GW) (function, y = ax'). R', coefficient of determination; n, sample size. 176 PACIFIC SCIENCE· April 2001

TABLE 2 Morphometric Relationships between Otolith Features and Body Length and Weight in M. rosacea

Parameters

Relationship a SE (a) b SE (b) R'

OLDvs. TL 10.640 1.416 1.656 0.055 0.8360 OLRvs. TL 34.640 3.276 1.639 0.055 0.8357 OLDvs.GW 0.088 0.035 4.241 0.159 0.7928 OLRvs. GW 1.538 0.452 4.288 0.161 0.7977

Note: All relationships were fitted to the potential model y = ax'. Linear dimensions are expressed in mm and weight in g. OLD, otolith long diameter; OLR, otolith long radius; TL, fish total length; GW, fish gutted weight, n = 187 for all cases.

In general, the temperature increase corre­ one hyaline ring represents a year in the life lates well with the increase of opaque otolith of these fish. edge development and vice versa. There is no The back-calculation of lengths at past evidence ofanother season with a high degree ages showed that the readings on otoliths of opaque otolith edge development in the were consistent, except for the age groups population. Thus, a set of one opaque,and of 12 or more years. In these, the back-

100 30 -. ~ 0 '-' 28 G80 -. 0 z 0 w '-' ::J \ 26 W a 60 ~ W q \ ::J ~ , u.. ",,\ 24 ~ '~ ~ W w C) 40 , 0 " 0- \", ' 22 :2: w W W I- ::Ja 20 4: 20 0- 0 0 18 JAN MAR MAY JUL SEP NOV JAN MAR MAY

~ Opaque edges -+- Temperature

FIGURE 3. Monthly frequency of individuals of M. rosacea exhibiting opaque otolith edges in the total sample of 769 fish, plotred with mean monthly sea surface temperature in the Bay of La Paz. Age and Growth ofMyeteroperca rosacea . Diaz- Uribe et at. 177

1000 o

800

-. E • ...... E 600 J .....~ II I Cl C Q) -.oJ tU 400 ~

200 - Linear Regression Non Linear Regression <> Observed Lengths ± S.D. • Back-calculated Lengths ± S. D. o o 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Age (years)

FIGURE 4. Growth curves (in total length, TL) of the leopard grouper fitted by two methods: NLR (nonlinear re­ gression), using FISHPARM 3.1 (Prager et al. 1989) and LRC (linear regression). Back-calculated lengths-at-age (closed circles), and observed lengths-at-age (open circles) are shown.

calculated lengths were, on average, shorter methods. The 95% confidence interval for k than for the other age groups (Figure 4). Be­ estimated by the NLR method is 0.0567 to cause of the small number of fish older than 0.0750 yr-l; the value of k (0.0925 yr-l) esti­ 11 yr, it is not possible to evaluate if the bias mated by the LRC method is outside this in the averages was due to a random factor or range. Although the nonlinear regression to an intrinsic factor of the otolith (Table 3). analysis (NLR) explains a larger amount of The growth model was fitted to a set of the variance in the data set, the goodness offit data containing the back-calculated lengths­ should be evaluated in view of the biological at-age, up to 11 yr, and the observed lengths­ information of the species (Figure 4). at-age (Table 4). Estimated parameters in the LRC method were different from those DISCUSSION in the NLR method (Table 5). According to the standard error estimated by the Although sex could be determined only for a NLR method, L oo could vary from 1142 to small number of fish, we can state that juve­ 1310 mm (df = 28; P = 0.05); the Loo = 1083 niles were not present in the sample. Even the estimated by the LRC method is outside that smaller fish in our samples were adults, which range. The growth coefficient k also showed agrees with the reports of Heemstra (1995), differences between the two estimation who referred to juveniles of M. rosacea as TABLE 4 smaller than 300 mm TL. The same author Observed Mean Lengths of Individuals ofM. rosacea at reported a maximum length of 700 mm for Various Ages as Determined by Otolith Annual Marks the species, but Thomson et al. (1987) and Allen and Robertson (1994) mentioned that Age (yr) TL (mrn) SD n this species can reach up to 1000 mm and 12.2 kg of weight, although they did not 2+ 291 7.616 4 specify the kind of length and weight. This 3+ 336 13.730 29 4+ 385 25.232 142 5+ 434 31.899 141 6+ 477 30.924 98 7+ 535 37.180 119 TABLE 5 8+ 582 39.179 84 Growth Parameters for M. rosacea Calculated by Two 9+ 631 26.179 67 Regression Methods 10+ 681 32.472 24 11+ 722 20.124 8 12+ 747 23.629 3 Nonlinear Linear 13+ 768 3.536 2 Regression Regression 14+ 786 15.556 2 15+ 808 1 Parameter Average SE Average SE 16+ 860 1 17+ 866 1 L", 1226 41.16 1083 18+ 875 1 k 0.06586 0.00447 0.09245 21+ 975 1 to -1.393 0.1798 -0.2446 r 2 0.99603 0.98034 Note: TL, total length; SD, standard deviation of mean length; n, sample size. Note: L., in nun, k in year-I, and to in years. Age and Growth of Myeteroperca rosacea . Diaz-Uribe et al. 179 divergence is shown in the length distribution lengths by age at the time of capture repre­ found in our study sample. Although most sent the length at an intermediate age be­ individuals caught are less than 700 mm TL, tween two birthdays, because at the time of larger fish are encountered; our study regis­ age assignment most otoliths show a fraction tered lengths to 915 mm TL, close to the of edge, indicating that the birthday has maximum reported by Thomson et al. (1987) already passed. However, back-calculated and Allen and Robertson (1994). lengths at previous ages represent the exact Heating otoliths is a useful technique to birthday lengths. Therefore, the consistency improve contrast and ease the reading of M. of the two methods was analyzed by the fol­ rosacea otoliths; each mark acquires a differ­ lowing criteria: (1) each back-calculated ential coloration. length belongs to an exact age (1, 2, 3, ... , The allometry phenomenon in the rela­ etc.); (2) each length at the time of capture tionship of otolith size to fish size has been belongs to an intermediate age (1.5, 2.5, ... , reported for several species, and discussed etc.). In this way, both data sets intercalate from the physiological point of view (Simkiss reasonably well between 3 and 11 yr of age, 1973). The extreme case in which otoliths within what could be the individual growth grow to a maximum size and then grow only curve. Because juveniles younger than 3 yr in thickness has been documented for several and adults older than 11 yr were scarcely species of soles (Williams and Bedford 1973). represented in our sample, we consider only In M. rosacea the evidence indicates that the ages within this range to be accurately re­ fish grows at a higher rate than the otolith. flected by otolith characteristics. The value of b = 1.656 indicates that when Individual growth is a process that implies the fish increases its length 100%, the otolith a change in body mass in a determined time grows approximately 53%. Other species, period. To express this phenomenon in terms such as Ocyurus chrysurus (Manooch and of length, von Bertalanffy developed his Drennon 1987) and Lutjanus peru (Rocha­ model under the hypothesis that organisms Olivares and Gomez-Munoz 1993), have b grow isometrically. That is, that the organism values near 1.6. However, otolith growth keeps the same body form during the growth does not completely stop in these cases. Al­ period. Whenever this happens, the change though there are no precise criteria specifying in body mass or respiratory surface due to the critical level at which allometry would growth is proportional to the change in adversely affect age determination by exami­ length cubed or squared, respectively. Pauly nation of otoliths, the values obtained for M. (1979) found that many fish species do not rosacea do not seem to cause problems, be­ comply with this criterion, so the Von Berta­ cause even at the oldest ages, rings could be lanffy growth model (VBGM) represents only counted and measured. a special case. The value of b = 2.97 in the Whenever growth marks are deposited leopard grouper is not significantly different periodically, it is possible to assess this pro­ from 3, therefore the use of the VBGM can cess using the ratio of edge types (opaque or be considered adequate in this case. hyaline) in otoliths (Williams and Bedford Fitting the growth curve by NLR is a sta­ 1973). The prevalence of opaque edges in tistically powerful method because no trans­ summer-autumn and hyaline edges in win­ formation of the data is required and the, ter-spring represents a defined pattern in M. error of the parameters is estimated in an rosacea otoliths. unbiased way (Prager et al. 1989). But this is a We consider that reading of otoliths is a condition a priori, and the estimations must consistent method for determining age at the be evaluated as a function of the biological time of capture, because observed lengths implications of the adjusted parameters. Sev­ were generally larger than back-calculated eral authors have discussed the close rela­ lengths at each age. Comparison was made tionship of w;., and k with the biological cycle taking into account that each method gives of species (Pauly 1979, Manooch 1987, complementary information. That is, average Sparre and Venema 1992). Munro and Pauly 180 PACIFIC SCIENCE· April 2001

0.7

0.6 181 Epinephelus o Mycteroperca 0.5 • M. rosacea v Lutjanus 0.4 K 0.3

C8I 0.2 Y 'VV 0 181 V' 0.1 0 JIt8l NlR 0 0 200 400 600 800 1000 1200 1400 1600 1800 Loo

FIGURE 5. Growth performance (<1» for several species of the families Serranidae and Lutjanidae (cited in Manooch 1987) and for M. rosacea by NLR (nonlinear regression) and LRC (linear regression).

(1983) developed the index = log k + length in the southern Gulf of California, 2/3 log ~, called the "growth performance following the concept of Pauly (1979). In this index" because it identifies specific kinds of context, the value ofLoo = 1083 mm obtained growth in different taxonomic groups. An av­ by LRC seems more reasonable than the erage value of = 1.66 has been estimated one estimated by the NLR method (L oo = for the family Serranidae, varying from 1.3 5 1226 mm). to 2.26 (Manooch 1987). We estimated Manooch (1987) gave the values of k for = 1.735 (using LRC) and = 1.693 (using the family Serranidae, which vary from 0.074 NLR) (Figure 5). Both values are consistent to 0.63 per year. The leopard grouper has with the "growth performance index" of the k = 0.066 estimated by NLR, which is out­ family Serranidae. side the range reported by Manooch, but the The asymptotic size of the model (ex­ value estimated by LRC (0.092) is within the pressed as Loo or ~) has been generally de­ range. fined as the average size a species would reach It is important to stress that within the if growth were indefinite (Ricker 1975). An­ serranids the leopard grouper is one of the other interpretation was proposed by Pauly slowest growing species. This could be re­ (1979), who argued that the asymptotic size lated to the feeding regime of the species. should be equal to the average size of the Buesa (1987) analyzed the growth rates of oldest fish in a stock. The maximum length several species of demersal fish and found registered by us was 975 mm TL. It seems evidence that those with fish as a main item in reasonable that the maximum length of their diet had slower growth rates than the 1000 mm reported by Thomson et al. (1987) ones with more diverse diets or the opportu­ is a good reference point of the asymptotic nistic ones. The leopard grouper h:j.s been Age and Growth ofMycteroperca rosacea . Diaz-Uribe et at. 181 classified as fully piscivorous (Hobson 1965), 512 in B. R Murphy and D. W. Willis, but Bermudez-Almada and Garcia-Laguna eds. Fisheries techniques, 2nd ed. Ameri­ (1985) found that, besides fish, M. rosacea also can Fisheries Society, Bethesda, Mary­ feeds on crustaceans in a 2: 1 ratio (fish: land. crustaceans). In spite of utilizing foods with Elorduy-Garay, J. F., and J. G. Diaz-Uribe. high energetic content, the low growth rates 1994. Age validation of Caulolatilus affinis, offish with this feeding regime are due to the Gill 1865 (Pisces: Branchiostegidae) from large energetic cost of searching and chasing the Gulf of California using otoliths. Sci. prey (Buesa 1987). Mar. 58:277-287. Heemstra, P. C. 1995. Serranidae. Pages 1201-1813 in W. Fisher, F. Krupp, W. ACKNOWLEDGMENTS Schneider, C. Sommer, K. E. Carpenter, We thank the fishermen of El Sargento and and V. H. Niem, eds. Guia FAO para la La Ventana for allowing us to sample their identificacion de especies para los fines de catches. We also are grateful to Richard la pesca: Pacifico centro-oriental. Vol. 3. Radtke and an anonymous reviewer who made FAO, Rome [in Spanish]. important contributions on early drafts of the Hobson, E. S. 1965. Diurnal-nocturnal activ­ manuscript to improve its quality and clarity. ity of some inshore fishes in the Gulf of California. Copeia 3:291-302. Literature Cited Manooch, C. S., III. 1987. Age and growth of snappers and groupers. Pages 329-374 in Allen, G. R, and D. R Robertson. 1994. J. J. Polovina and S. Ralston, eds. Tropical Fishes of the tropical eastern Pacific. Uni­ snappers and groupers: Biology and fish­ versity of Hawai'i Press, Honolulu. eries management. Westview Press, Boul­ Beamish, R]., and G. A. McFarlane. 1983. der, Colorado. The forgotten requirement for age valida­ Manooch, C. S., and C. L. Drennon. 1987. tion in fisheries biology. Trans. Am. Fish. Age and growth ofyellow-tail snapper and Soc. 112:735-743. queen triggerfish from the U.S. Virgin Is­ Bermudez-Almada, B., and G. Garcia-La­ lands and Puerto Rico. Fish. Res. 6:53­ guna. 1985. H:ibitos alimenticios de los 68. peces de las zonas rocosas de la Bahia de Munro, J. L., and D. Pauly. 1983. A simple La Paz, Baja California Sur. B.S. thesis, method for comparing the growth offishes Universidad Nacional Autonoma de Mex­ and invertebrates. Fishbyte, ICLARM ico, Mexico, D.F. [in Spanish]. 1:5-6. BoeWert, G. W. 1985. Using objective crite­ Pauly, D. 1979. Gill size and temperature as ria and multiple regression models for age governing factors in fish growth: A gener­ determination in fishes. Fish. Bull. 83:103­ alization of von Bertalanffy's growth for­ 117. mula. Verh. Inst. Meereskunde, Univ. Kiel Buesa, R J. 1987. Growth rate of tropical 63:1-156. demersal fishes. Mar. Ecol. Prog. Ser. Prager, M. H., S. B. Saila, and C. W. Reck­ 36:191-199. siek. 1989. FISHPARM: A microcomputer Casselman, J. M. 1983. Age and growth as­ program for parameter estimation of non­ sessment of fishes from their calcified linear models in fishery science, 2nd ed. structures-techniques and tools. In E. D. Old Dominion Univ., Oceanogr. Tech. Prince and L. M. Pulos, eds. Proceedings Rep. 87-10: 1-18. of the International Workshop on Age Ricker, W. E. 1975. Computation and inter­ Determination of Oceanic Pelagic Fishes: pretation of biological statistics of fish Tunas, Billfishes and Sharks. NOAA populations. Fish. Res. Board Can. Bull. Tech. Rep. NMFS 8:1-17. 191:1-382. Devries, D. R, and R V. Frie. 1996. Deter­ ---. 1979. Growth rates and models. mination of age and growth. Pages 483- Pages 677-743 in W. S. Hoar, D. J. Ran- 182 PACIFIC SCIENCE· Apri12001

dall, and J. R. Brett, eds. Fish physiology. fish in relation to ageing. Pages 1-12 in Vol. 8. Bioenergetics and growth. Aca­ T. B. Bagenal, ed. Ageing of fish. Unwin demic Press, Orlando, Florida. Brothers, Surrey, England. Rocha-Olivares, A., and V. Gomez-Munoz. Sparre, P., and S. C. Venema. 1992. Intro­ 1993. Validacion del uso de otolitos para duction to tropical fish stock assessment. determinar la edad del huachinango del FAO Fish. Tech. Pap. 306/1: 1-376. Pacifico Lutjanus peru (Perciformes: Thomson, D. A., T F. Lloyd, and A. N. Lutjanidae) en la bama de La Paz y aguas Kerstitch. 1987. Reef fishes of the Sea of adyacentes, B.C.S., Mexico. Cienc. Mar. Cortez: The rocky-shore fishes ofthe Gulf 19:321-331. of California. University of Arizona Press, Rodriguez-Medrano, M. C. 1990. Composi­ Tucson. cion especifica de la captura artesanal de Villavicencio, C. 1983. Distribucion y abun­ escama de isla Cerralvo, B.C.S., Mexico. dancias relativas de la familia Serranidae en B.S. thesis, Universidad Autonoma de Baja la Bahia de La Paz, B.C.S. Mem. VII California Sur, La Paz, B.C.S., Mexico [in Congr. Nac. Zool., Jalapa, Ver., Mexico Spanish]. [in Spanish]. Rosenblatt, R. H., and B. J. Zahuranec. 1987. Williams, T, and B. C. Bedford. 1973. The The eastern Pacific groupers of the genus use of otoliths for age determination. Mycteroperca; including a new species. Pages 114-123 in T B. Bagenal, ed. The Calif. Fish Game 53:228-245. ageing of fish. Unwin Brothers, Surrey, Simkiss, K. 1973. Calcium metabolism of England.