xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002063883
007 cr mnu|||uuuuu
008 100318s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003015
Tyler-Jedlund, Amanda J.
Age, growth, and reproduction of Calamus proridens, the littlehead porgy, from the northeast gulf of Mexico
h [electronic resource] /
by Amanda J. Tyler-Jedlund.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 59 pages.
Thesis (M.S.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: A total of 1814 Calamus proridens ranging from 76 mm 361mm fork length (FL) were collected and processed along the central and northwest coasts of Florida between 2000 and 2007 to determine size, sex, age, and reproductive condition. Females ranged from 76-297 mm FL (mean FL=156 mm, n=1420), males ranged from 141-361 mm FL (mean FL=244 mm, n=297), and transitionals ranged from 131-307 mm FL (mean FL=207 mm, n=42). Sex ratios sorted by length class, age, and overall were significantly different from the 1:1 ratio for gonochoristic species (P is less than 0.0001). Sagittal otoliths (sectioned and whole) from 1438 C. proridens were used to determine age. Marginal increment analysis suggested that a single annulus is formed each year in the spring. Ages ranged from 0 to 10 years with 88% of the fish being between 0 and 4 years. Females ranged in age between 0 and 6 years, while males ranged between 1 and 10 years. Growth was rapid in the first two years and then began to slow down. The von Bertalanffy growth model fitted to all observed data was L[subscript (t)]=306[1-e[superscript -0.254(t+1.69)]]. The data were further broken down into central and southern strata and the von Bertalanffy growth model showed that fish in the central region grow larger than those in the southern region. Histological analysis confirmed that C. proridens are protogynous hermaphrodites with delimited type gonads. Estimates indicated that 50% of the females in the sample had transitioned into males by age 4 and a FL of 231 mm. Calamus proridens mature at a small size; 50% of the samples were mature by 132 mm and within the first year. The samples obtained indicated that the peak spawning season is in the spring. The presence of hydrated oocytes and post ovulatory follicles in the same ovary suggests that they are multiple spawners.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Joseph J. Torres, Ph.D.
x Marine Science
t USF Electronic Theses and Dissertations.
Age, Growth, and Reproduction of Calamus proridens the Littlehead Porgy, from the Northeast Gulf of Mexico by Amanda J. Tyler-Jedlund A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science College of Marine Science University of South Florida Major Professor: Joseph J. Torres, Ph.D. David A Mann, Ph.D. James A. Colvocoresses Ph. D. Date of Approval: May 29, 2009 Keywords: life history, protogynous hermaphrodite Florida, otol iths, fishery Copyright 2009, Amanda J. Tyler-Jedlund
Acknowledgements I would like to thank my committee, Joseph Torres, David Mann, and Jim Colvocoresses, for their support during this study and their critic al review of this manuscript. I want to thank Rich Lehnert for having the insight to l ook at this interesting and complex species. I would like to tha nk the Fisheries-Independent Monitoring program and the crews of the baitfish survey s for providing samples, especially Keith Fisher and Jenna Tortorelli. The National Ma rine Fisheries, Panama City Lab, and the Fisheries-Dependent Monitoring program also provided samples. Hist ological sections were prepared by the Louisiana State Univ ersity and the Florida Wildlife Research Institute (FWRI). Otoliths were processe d by the Marine Fisheries Biology group at FWRI. Janet Tunell provided guidance in th e age and growth process. Sue LowerriBarbieri, Doug DeVries, Melissa Cook, and Harry Grier were instrumental in guiding me through the different classes of reproductive biology. I w ould like to thank Michael Murphy, Rich McBride, and Robert Muller who helped strengthen my statistical knowledge and provided valuable insight in th e overall results for the age and growth section. I want to thank my husband Tobias who helped process samples and for his encouragement and emotional and moral suppo rt. This studied was funded by the Federal Aid for Sportfish Restorati on Project Number F-43 and th e Florida saltwater fishing license sales.
i Table of Contents List of Tables i List of Figures ii Abstract iii Introduction 1 Material and Methods 5 Geographic coverage 5 Sample Collection 5 Age and Growth 8 Age Validation 12 Back-Calculation 12 Reproduction 13 Results 17 Sample Collection 17 Age and Growth 21 Reproduction 34 Discussion 45 Age Validation 45 Age and Growth 46 Reproduction 48 Conclusion 54 References 56
ii List of Tables Table 1 Relationship between standard, fork, and total length, and fork length and total weight of Calamus proridens from the northeast GOM. 7 Table 2 Histological criteria to asse ss reproductive classes for female Calamus proridens. 15 Table 3 Mean back-calculated FL at annulus formation of Calamus proridens all sexes combined. 26 Table 4 Sample sizes, observed mean fo rk length (mm) at age, and range for Calamus proridens. 27 Table 5 Growth parameter estimates of Calamus proridens from the von Bertalanffy growth model with st andard error and 95% confidence intervals. 29 Table 6 Sample sizes, observed mean fork length (mm) at age, and predicted length at age from the Von Bertalanffy growth model for Calamus proridens for all samples, and th e central and southern strata 30 Table 7 Age length key for Calamus proridens from the northeast GOM. 33 Table 8 Number and percen tage of females, males, and transitionals of Calamus proridens from the northeast Gulf of Mexico grouped into10 mm length classes with sex ratio. 36 Table 9 Number and percen tage of females, males, and transitionals of Calamus proridens from the northeast Gulf of Mexico by age with sex ratio. 37
iii List of Figures Figure 1. Sampling locations of Calamus proridens from the FisheriesIndependent Monitoring (FIM) progr amÂ’s baitfish cruise surveys, FIM gear testing cruises, and samples collected from the National Marine Fisheries (NMFS) in Panama City from 2000 Â– 2007. 6 Figure 2. Whole otolith from an age one Calamus proridens (164 mm FL) (top) and a sectioned otolith from a six year old C. proridens (274 mm FL) (bottom); both samples were captured in November 2005. 9 Figure 3. Relationship between depth and fork length for Calamus proridens collected in the northeast Gulf of Mexico. 18 Figure 4. Length A) and age B) frequency distribution of Calamus proridens collected from the northeast Gulf of Mexico from 2000-2007; includes sexed and unsexed samples. 19 Figure 5. Length-weight relationship for Calamus proridens from the northeast Gulf of Mexico. 20 Figure 6. Mean marginal increment for Calamus proridens ages 2-10. 22 Figure 7. Relationship between otolith radius (mm) and fork length (mm). 23 Figure 8. Mean back-calculated fork le ngth between ages 1-5 as a function of age at capture. 25 Figure 9. Von Bertalanffy growth model fitted to all observed length at age data for Calamus proridens 28 Figure 10. Von Bertalanffy growth curves fitted to observed length at age for the central and southern regions of the west coas t of Florida. 31 Figure 11. Length A) and age B) frequency distribution of Calamus proridens collected from the northeast Gulf of Mexico from 2000-2007. 35 Figure 12. Macroscopic views of A) ovaria n, B) transitional, a nd C) testicular gonads from Calamus proridens collected in the Gulf of Mexico. 39
iv Figure 13. Histological sections of Calamus proridens gonads. 40 Figure 14. Percentage of female littlehead porgy, Calamus proridens by age and length class (10 mm size interv als) with a logistic function (continuous line) fitted to the observed data. 41 Figure 15. Percentage of mature littlehead porgy, Calamus proridens by length class (10 mm size interval s) with a logi stic function (continuous line) fitted to the observed data. 43
v Age, Growth, and Reproduction of Calamus proridens the Littlehead Porgy, in the Northeast Gulf of Mexico Amanda J. Tyler-Jedlund ABSTRACT A total of 1814 Calamus proridens ranging from 76 mm Â– 3 61mm fork length (FL) were collected and processed along the central and northwest coasts of Florida between 2000 and 2007 to determine size, sex, age, and reproductive condition. Females ranged from 76-297 mm FL (mean FL=156 mm, n= 1420), males ranged from 141-361 mm FL (mean FL=244 mm, n=297), and transitiona ls ranged from 131-307 mm FL (mean FL=207 mm, n=42). Sex ratios sorted by length class, age, and overall were significantly different from the 1:1 ratio for gonochoristi c species (P<0.0001). Sagittal otoliths (sectioned and whole) from 1438 C. proridens were used to determine age. Marginal increment analysis suggested that a single annulus is formed each year in the spring. Ages ranged from 0 to 10 years with 88% of the fish being between 0 and 4 years. Females ranged in age between 0 and 6 years, while males ranged between 1 and 10 years. Growth was rapid in the first two years and then began to slow down. The von Bertalanffy growth model fitted to all observed data was L(t)=306[1-e-0.254(t+1.69)]. The data were further broken down into central a nd southern strata and the von Bertalanffy growth model showed that fish in the centr al region grow larger than those in the southern region. Histological analysis confirmed that C. proridens are protogynous
vi hermaphrodites with delimited type gonads. Es timates indicated that 50% of the females in the sample had transitioned into males by age 4 and a FL of 231 mm. Calamus proridens mature at a small size; 50% of the sa mples were mature by 132 mm and within the first year. The samples obtained indicated that the peak spaw ning season is in the spring. The presence of hydrated oocytes a nd post ovulatory follicles in the same ovary suggests that they are multiple spawners.
1 Introduction The littlehead porgy ( Calamus proridens ) occurs in the Atlantic Ocean and Gulf of Mexico (GOM) from Florida to Loui siana, the Campeche Bank, and through the Greater Antilles (Darcy, 1986, Dubovitsky, 1977 a, Pierce and Mahmoudi, 2001, Randall and Caldwell, 1966). Their distribution is in fluenced by salinity, temperature, water depth, and habitat structure, as well as; sp ecies and life history ch aracteristics (Darcy, 1986). They are typically found associated w ith natural or artific ial reefs, offshore platforms, and live bottoms consisting of sponges and corals (Darcy, 1986, Pierce and Mahmoudi, 2001). Calamus proridens have not been found in estuaries or in waters where salinity is less than 30 ppt (Darcy, 1986) Adults in the GOM are distributed from shallow near shore waters to depths as great as 60 meters (Darcy, 1986). Calamus proridens is an important component of the nearshore/offshore fish assemblage of the west coast of Florida, a nd subtropical and tropical American waters in general (Darcy, 1986, Pierce and Mahmoudi 2001, Randall and Caldwell, 1966). For example, Calamus proridens has been reported to be the most important and abundant member of the family Sparidae in the commercial fishery in the Campeche Bank, Yucatan. In the 1960s and 1970s research vessels collected over 500 kg per haul (Dubovitsky, 1977a). They are often incide ntally harvested by the commercial reef fishery in Florida (Darcy, 1986, Dubovitsky, 1977a, Randall and Caldwell, 1966). The red snapper ( Lutjanus campechanus ) commercial fishery harvested over 1.72 million pounds between 2002 and 2006 along the Florid a west coast (personal communication
2 from the Fish and Wildlife Research Instit ute) indicating that numerous porgies were probably also harvested even though no sp ecific numbers are re ported. Recreational harvest of porgies between 2002 and 2006 (not including pinfish, red porgy, or sheepshead) totaled over 67,000 kg from the Atlantic and Gulf coasts (personal communication from the National Ma rine Fisheries Service). Even though C. proridens have a commercial and recreational importance, specific life history informa tion is not well known, especially in the GOM. Randall and Caldwell (1966) did a diagnostic description on four new species of Sparids, including C. proridens. In the 1970s Dubovitsky reported life history characteristics for the Campeche Bank region, however his studies we re based on length frequency analysis and the data do not specify which length was used (Standard Length-SL, Fork Length-FL, or Total Length-TL) in the analysis. Da rcy (1986) did a biological synopsis on C. proridens and C. arctifrons with most of the findings para phrased from DubovitskyÂ’s studies. Many species in the family Sparidae have been reported to be protandrous (male to female) (Besseau and Brusle-S icard, 1995, Buxton and Garratt, 1990, Lee et al. 2008) or protogynous (female to male) (B uxton and Garratt, 1990, Garratt, 1986, Haung et al. 1974, Kokokiris et al. 1999) hermaphrodites. Most research on sex-changing fish involves species within the families Serra nidae (groupers and seabasses), (Bullock et al. 1996, Cochran and Grier, 1991, Coleman, 1981, Fischer and Petersen, 1987, Shapiro, 1987, Thurman, 2004), Labridae (wrasses), (S hapiro and Rasotto, 1993, Warner and Swearer, 1991) and Scaridae (pa rrotfish) (Munoz and Warner, 2004). The sex reversal in those families is undelimited, where the male and female tissues are intermixed or separated within the gonad duri ng the course of sex reversal. Little research has been
3 conducted on the sex-changing processes of sparids, which are delimited (male and female gonads are separated by connective tissu e) and most work has been done on South African species (Besseau and Brusle-Si card, 1995, Buxton and Garratt, 1990, Garratt, 1986, Haung et al. 1974, Lee et al. 2008) or the red porgy ( Pagrus pagrus ) (DeVries, 2006, Hood and Johnson, 2000, Kokokiris et al. 1999, Kokokiris et al. 2006, Pajuelo and Lorenzo, 1996). Disturbingly, fish populations are bei ng depleted from overfishing (Alonzo, 2003, DeVries, 2006). The Magnuson-Stevens Act wa s reauthorized in 2006 with the goal of continuing to protect commercial and recreational fish communities as global fishery resources provide tremendous economical and ecological value. As the more highly valued fish become depleted and more hi ghly regulated, commercial and recreational fishermen typically switch to more abundant species that are not as highly monitored (DeVries 2006). Past stock assessment models were lim ited by the data that were available, however, they are now beginni ng to include life-history pa rameters such as mating behaviors, reproductive patterns, survival, and recruitment (Alonzo, 2003). Sex-changing fish add another level of complexity to st ock assessment models th at can have a large impact on their results and subsequent ma nagement decisions (Alonzo, 2003, Armsworth, 2001, Davis and Berkson, 2005). A better unders tanding of the protogynous life history strategy can benefit modeling efforts and our understanding of the sensitivity of protogynous species to overfishing (DeVries, 2006). Unlike dioeciou s (separate sexes) species, a size selective fisher y on sex-changing fish can negatively impact reproductive rates and reduce population size to very low levels (Alonzo, 2003, Armsworth, 2001,
4 Davis and Berkson, 2005). Including life history information in stock assessments will further our understanding of those speciesÂ’ vulnerability to exploitation, and may help eliminate future fish populations from fi shing moratoria or population crashes. Calamus proridens are suggested to be monandric (arising only from sex change) protogynous hermaphrodites (female to male) w hose growth rates may be both ageand sex-dependent. However, there is little information on age and size of the transition period and reports have come only from the Campeche Bank. Hermaphroditic fishes are especially vulnerable to exploitation, as fisher ies typically target the largest fishes first. This practice tends to remove the largest fishes from the system and disrupts the reproductive ability of the population (A lonzo, 2003, Armsworth, 2001, Crabtree and Bullock, 1998, Davis and Berkson, 2005). Precise timely information describing size at age and sex and age at sex-reversal will allow an accurate assessment of C. proridens stocks and provide detailed information fo r the GOM. Including new information from a single-species study into a multi-species model will allow further development for an ecosystem-based approach to managing fisheries resources in the GOM. The main objectives of the present resear ch were to describe the age and growth, and reproductive behavior of Calamus proridens from the northeast Gulf of Mexico concentrating on age and size at sexual transition.
5 Materials and Methods Geographic coverage The majority of the samples were obtained in the spring (April and May) and fall (October and November) from 2003-2007 during the Fisheries-Indepe ndent Monitoring (FIM) programsÂ’ baitfish surveys aboard the R/V Tommy Monro Samples were collected using a 65-ft balloon trawl that was towed for 30 minutes per sample. Spring surveys were conducted between 2003 and 2007 and th e fall surveys between 2004 and 2006. Samples were collected at randomly selected sites off the central west coast of Florida between the 28o N and 26o N (Fig. 1). Samples were al so obtained from the FIM geartesting cruises in October 2006 and July 2007 aboard the R/V Suncoaster and R/V Bellows using baited Chevron traps with 60 minute soak times. The October 2006 cruise collected samples between 27o 48Â’ N and 27o 26Â’ N; the July 2007 cruise collections were from the Florida middle grounds between 28o 50Â’ N and 28o 20Â’ N. Additional samples were collected from the Fisheries-Depende nt Monitoring (FDM) program, the National Marine Fisheries (NMFS) archive databa se from Panama City, and recreational fishermen; all additional fishes were caught using hook and line. Sample Collection Standard, fork, and total lengths (mm) were recorded from samples taken on the baitfish survey. General linear regression was used to determine length to length relationships (over 1500 fish were used) and had R2 values of 0.99 for all length combinations (Table 1). Those models were then used to obt ain missing lengths from other samples. All
6 Figure 1. Sampling locations of Calamus proridens from the Fisheries-Independent Monitoring (FIM) programÂ’s baitfish cruise surveys ( ), FIM gear testing cruises ( ), and samples collected from the National Marine Fisheries (NMFS) in Panama City from 2000-2007. Samples collected by FDM and recrea tional fisherman did not have latitude and longitude coordinates and are not represen ted in the map, but were collected in the Tampa Bay area.
7 Table 1. Relationship between standard, fork, and total length, and fork length and total weight of Calamus proridens from the northeast GOM. SL = standard length (mm), FL = fork length (mm), TL = total length (mm), WT = total weight (g). The fork length range for all length-length regressions was 76 Â– 361 mm and the total length range was 90 Â– 421 mm. Total weight range for leng th-weight regression was 12.5 Â– 1115 g. Y=ax+b Y X N a b R2 SL FL 1545 0.88 -0.54 0.997 FL SL 1545 1.13 1.21 0.997 SL TL 1532 0.77 -4.31 0.992 TL SL 1532 1.29 7.05 0.992 FL TL 1558 0.87 -4.25 0.995 TL FL 1558 1.14 5.93 0.995 WTlo g 10 FLlo g 10 1298 2.82 -4.18 0.993 Fork lengthweight power equation WT=7e-05FL2.82
8 results are reported based on fork length (FL) measurements unless otherwise noted. Total weight in grams was recorded when possi ble. Frozen weights were taken in the lab when fresh weights could not be obtained in the field. Fresh and frozen weights were combined for length-weight analysis. Lengt h and weight were log transformed and an Analysis of Covariance (ANCOVA) was used to test for differences between sexes (Froese, 2006, Ricker, 1975, Sokal and Rohlf, 1981). Location and depth (meters) were recorded from the baitfish survey, gear te sting cruises, and NMFS samples. Surface salinity and temperature were recorded from the baitfish surveys but only salinity was recorded from the NMFS samples. Age and Growth Fish age has been determined using calcifi ed structures such as fin rays, scales and otoliths (Ihde and Chittenden, 2003, Pa nnella, 1974). Otoliths are usually the clearest and most accurate age markers when compared to other calcified structures (Chambers and Miller, 1995). Depending on the species, age assessment can be done with whole or sectioned otoliths (Gettel et al. 1997, Thurman, 2004). In this study, fish were aged using sectioned and whole sagittal ot oliths (Fig. 2). Readab ility of large whole
9 Figure 2. Whole otolith from an age one Calamus proridens (164 mm FL) (top) and a sectioned otolith from a six year old C. proridens (274 mm FL) (bottom), both samples were captured in November 2005. OR= otolith ra dius. Note the wide marginal increment subsequent to the sixth annulus. 1 2 4 3 6 5 OR 1
10 otoliths can be affected due to clouding a nd crowding of rings, therefore, 202 otoliths were aged whole and then sectioned to identi fy the limits of aging using whole otoliths. Since there was a 96% agreement on whole vers us sectioned otoliths for ages zero and one, all otoliths were aged whole first and then any whole otoliths with more than one annulus were sectioned and aged. Any questio nable whole otoliths were also sectioned. The left otolith was used for age estimati on unless it had been broken, lost, or damaged, in which case the right one was substituted. Otoliths were embedded in Araldite resin and cured at 60C for three hours. The otolith was then hot-glued onto cutting paper, and three transverse sections were cut to approximately 0.5 mm thick using a multi-bladed low-speed saw. The sections were rinsed, dried, and mounted on a slide with Flo-Texx mounting medium. Annuli were counted using a dissecting microscope with either transmitted or reflective light. All otoliths were enumerated twice, independent of each other. In cases where the two independent reads did not agree, a third read was done. If an agreement could not be reached, the otolith was not used in analysis. To assign an age to each fish a birth date of April 1st was assumed based upon C. proridens peak spawning period. A margin code was also recorded with a range of 0 Â– 3, with 0 having no margin (i.e. the last annulus is on the edge) and 3 a large margin. Samples collected from April 1st to July 31st with a margin code of 3 were assigned an age equa l to the annulus count plus one. Samples with a margin code of 0 or 1 and collected during January 1st through March 31st, were assigned an age equal to the a nnulus count minus one. Otherw ise, fish age was equal to annulus count.
11 To evaluate growth, observed lengths at age were fitted to the von Bertalanffy growth model (Ricker, 1975) by using nonlin ear regression techniques with a Marquardt algorithm (PROC NLIN; SAS 1996). The orig inal von Bertalanffy equation used to calculate growth was: L(t) = L *(1-e-k(t-to)) where L(t) = the fork length of the i th individual at age t ; L = the asymptotic maximum fork length; k = Brody growth coefficient; t = age; and t0 = the hypothetical age at which fork length is zero. Due to large variability in the model when a ll length and age data was combined from all areas; I separated the data into three regions based on latitude (Fig 1). There were not enough samples in the northern region to constr uct a growth curve. The central region was broken into two strata: a central (27 latitude) and a southern (26 latitude) to determine if growth differed by latitude. A randomization analysis of 1,000 iterations was used to determine if there was a significant difference between the central and southern strata (Manly, 1991). The AkaikeÂ’s Information Criteria (AIC) was used to measure the goodness of fit of the von Bertalanffy model to the data base d on location. The smaller the AIC number the better the data fits to the model. The formula for the AIC was: AIC=n*log(RSS/n)+2*P Where n=sample size RSS= residual sum of squares from the growth model P = number of parameters in the model Observed ages at length were used to construct an estimated size at age key (Ricker, 1975). Aged fish were assigned to 20 mm FL intervals and calculated age
12 distribution (as a percenta ge) for each size interval. Age validation Since age estimates had not been previously done on C. proridens, annual deposition of opaque rings was validated us ing marginal increment analysis. The marginal increment is the distance from the pr oximal edge of the la st visible opaque ring to the otoliths edge. The timing of op aque ring formation can be determined by examining the monthly mean marginal increm ent. If the mean marginal increment displays a yearly cycle, it can be inferred th at opaque ring formation is a yearly event (Barbieri et al. 1994, Crabtree and Bullock, 1998, Hood and Johnson, 2000). Measurements were taken along the ventral ri dge of the sulcal groove from the core to the distal edge of each annulus and to the otoliths edge (Fig. 2) using Image Pro image analysis software. Otolith radius and opaque bands were only marked on core sections and on samples where the two independent re ads agreed on incremen t count. Since the distance from the core to the first annulus is usually the largest, the marginal increment was calculated using a proportion equation: MI= OR Â– I L IL Â– IL-1 where OR = otolith radius IL = distance from otolith core to last increment IL-1 = distance from otolith core and the next to last increment Back-calculation Back-calculation is a technique used to in fer a length from an earlier time based on a known length and age (Francis, 1990). Annu li measurements with fish length are used to estimate fish length at time of annul us formation. Back-calculation is often used
13 to help develop an age-length model to determin e lengths of fish at ages that are rarely caught. The biological intercept method fr om Campana (1990), which assumes a linear relationship between otolith length and fish le ngth and uses a biological intercept rather than a statistical one, was used to back-c alculate fork length at annulus formation: La=Lc+(Oa-Oc)(Lc-Li)(Oc-Oi)-1 where Li and Oi = fish and otolith length at biological intercept Lc and Oc = fish and otolith length at capture La and Oa = fish and otolith length at age The biological intercept was calculated as the mean fish length and otolith radius of all age 0 fish. To detect LeeÂ’s phenomenon that back-calculated lengths decrease with increasing age a linear regression analysis was used to detect any trends (Francis, 1990, Gotelli and Ellison, 2004, Schirripa, 2002, Sokal and Rohlf, 1981). Reproduction Gonad samples from various size rang es were collected for histological evaluations. In the field, gonads were re moved and placed in 10% buffered formalin. Later, the gonads were rinsed and soaked in water for two one-hour periods, followed by a final 12-hour soaking prior to being transferre d to 70% ethanol. To evaluate variability in the gonad tissue a subsample (n = 44) of all the gonads had an anterior, middle, and posterior section taken from both lobes. Preliminary results indicated that there was no difference in the reproductive class between lobe s. Therefore, the lobe that was in the best condition was used. Gross observati ons of gonads undergoing sexual succession showed the configuration of ge rminal tissues to be of the delimited type, based on the Sadovy and Shapiro (1987) criteria of herm aphrodites. Connectiv e tissue delimits the
14 male tissue from female tissue and the male tissue begins growing posteriorly above the female tissue. This result was confirmed m acroand microscopically. Therefore, at a minimum, the posterior section was used for histology. An additional section was occasionally taken for further evaluation of the transitional process. In 100 samples, only a middle section was obtained; those were staged and analyzed as far as possible. Samples were stained with Hematoxylin-Eosin (H&E). A few samples were also stained with Periodic Acid-SchiffÂ’s reagent Matalin Yellow (PAS/MY). Females were assigned a reproductive class based on oocyte devel opment criteria from Brown-Peterson et al. (2007) (Table 2). Transitionals were identifie d based on criteria from Sadovy and Shapiro (1987) (Table 2). When exact reproductive class could not be determined, samples were categorized as immature, spawning, mature or unknown. Males were classified as immature (developing) or mature (spawning capable). The overall sex ratio for the entire sample, by 10 mm size intervals, and by age were tested for significant differences from a 1:1 ratio with chi-square expectancy analysis I estimated length and age at which 50% of the population underwent sex reversal from female to male by fitting the observed data with a logistic regression (proc logistic; SAS). The same method was used to estimate the size and age at which 50% of the population reached maturity. To prevent misclassifying of regenerating (primary gr owth mature) specimens as immature, only samples collected in the spawning season we re used for analysis. Females were considered sexually mature if the gonad was assigned to the developing class or the more advanced gonad classes.
15 Table 2. Histological crite ria to assess reproductive classes for female Calamus proridens Female Class Description Immature Only oogonia and primary growth present Â– no atresia, no signs of previous spawning Developing (Cortical avelio) Primary growth, cortical avelio and/or early vitellogenesis. No POFÂ’s Spawning Capable (Vitellogenic) Vitellogenic (VI) Â– may have primary growth and cortical avelio. Partially and fully yolked oocytes. Early VI with evidence of POFÂ’s Actively Spawning (Final oocyte maturation) Ovulating (within 12 hour s), germinal vesical migration (GVM), hydrated oocytes, POFÂ’s, advanced yolk coalescence and breakdown of nuclear membrane Regressing Atresia Â– VI undergoing alpha and beta atresia Â– less developed oocytes often Present, possible POF Regenerating (Primary growth mature) Only primary growth present, muscle Bundles, enlarged blood vessels May not be able to determine from immature. Transitionals Proliferating male tissue, spermatocytes Spermatid, spermatozoa. Some atresia in females. Immature females, mainly only primary growth present.
16 Male and female length and age frequenc y distributions were compared using the Kolmogorov-Smirnov (KS) two-sample test (Sokal and Rohlf, 1981). Analysis of variance (ANOVA) and t-test analyses were us ed to compare differences in mean lengths and mean length at age by sex using the stat istical software package STATISTICA.
17 Results Sample Collection A total of 1814 C. proridens were collected and processed from the Gulf of Mexico between 2000 and 2007. Samples were collected from all months, with the majority collected in the spring and fall. During the spring and fall baitfish surveys, samples were collected in waters with salinities ranging from 33.7 to 37.8 ppt and at temperatures between 19.4C and 28.9C. Calamus proridens from the northeast GOM were collected in temperatures ranging from 16.0C to 24.8C. Fish were collected at depths between nine and 60 meters. Fish le ss than 160 mm were only collected in depths less than 40 m, while larger fish were di stributed among the whol e range (Fig. 3). The C. proridens used in the present study ranged in length from 76 to 361 mm (mean=174 mm, SD=52, n=1814) (Fig. 4). The total weight range for individuals was 12.5-1115 g (mean=146.8, SD 136.6, n=1298). The relationships between SL, FL, and TL, and FL to total weight are presented in Table 1. Analysis of covariance showed no significant difference between sexes for the length-weight relationship so all samples were pooled, including samples that c ould not be sexed (ANCOVA; n=1298, F=0.05, P>0.05, Fig. 5).
18 Fork Length (mm) 50100150200250300350400 Depth (m) 0 10 20 30 40 50 60 70 Figure 3. Relationship between depth and fork length for Calamus proridens collected in the northeast Gulf of Mexico.
19 A Fork Length (mm) 50100150200250300350 Frequency 0 20 40 60 80 100 120 140 160 B Age 0246810 Frequency 0 100 200 300 400 500 Figure 4. Length A) and age B) frequency distribution of Calamus proridens collected from the northeast Gulf of Mexico from 2000-2007; includes sexed and unsexed samples. N = 1415 N=1814
20 Figure 5. Length-weight relationship for Calamus proridens from the northeast Gulf of Mexico. There was no significant difference be tween sex as a result all samples were pooled for analysis, P<0.001. 0 200 400 600 800 1000 1200 050100150200250300350400Fork Length (mm)Total Weight (g) N=1298 W=7e-05FL2.82
21 Age and Growth Otoliths were collected from 1438 specimens; 1047 were sectioned, 371 remained whole, and 22 could not accurately be aged a nd were not used for analysis. There was 94% agreement between the tw o independent reads on whole otoliths; 36 did not agree and were read a third time to produce an ag reed age. Sectioned otoliths had a 95% agreement between the two independent reads, 10 samples were cut again using the right otolith and two samples did not produce an agreement on age. An agreed age was determined on all remaining samples. Marginal increment analysis for all samp les combined showed annulus deposition occurring once a year during th e spring and early summer (Fig. 6). The majority of the samples were taken in April and October, which also had the largest differences in marginal increment. The mean marginal increment in April (mean=0.30, SD=0.25) was significantly smaller than the mean marginal increment for October (mean=0.69, SD=0.22, P=0.04). In April, marginal incremen ts were either larg e (52%), indicating individuals were about to lay dow n an annulus, or were on the e dge or small (38%), indicating the annulus had recently been laid down. Even though annulus deposition for individual ages could not be validated due to small sample sizes, I assumed for age and growth analysis that each opaque ba nd represented an annual mark. The biological intercept method for back -calculating previous length at age assumes a linear relationship between otolit h radius and fish le ngth. The relationship between otolith radius and fork length was described by the following general linear regression (Fig. 7): FL = 211.01 OR + 49.25 (n = 990, r2 = 0.70). LeeÂ’s phenomenon was not observed in back -calculated lengths as there was no
22 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 JanFebMarAprMayJunJulAugSepOctNovDecMonthMean Marginal Increment Figure 6. Mean marginal increment for Calamus proridens ages 2-10, n=663, error bars = standard deviation. Numbers above error bars indicate sample size. 18 12 14 293 19 9 9 2 1 162 92 3
23 0 50 100 150 200 250 300 350 400 00.20.40.60.811.21.4Otolith Radius (mm)Fork Length (mm) Figure 7. Relationship between otolith radius (mm) and fork length (mm). The linear regression equation is FL = 211.01x + 49.25, wh ere FL=fork length and OR= otolith radius (r2 = 0.7044, n=990).
24 indication that back-calculate d length decreased as a functio n of age (linear regression, P>0.5, Fig. 8). Back-calculated lengths of C. proridens were 124 Â– 136 mm at the time of first annulus formation (Table 3) a nd 104 Â– 212 mm at age one (Table 4). The observed means for each annulus group were larger than the back-calculated means indicating that there is probably some growth after annulus formation. Calamus proridens ranged in age from 0 to 10 and most fish (88%) were between ages 0 and 4 (Fig. 4B and Table 4). Only 14 fish (1%) were older than age seven. The largest specimen was 361 mm and was six ye ars old. The smallest specimen collected was 76 mm and age 0. There were two 10 year olds that were 308 and 355 mm. Size increased rapidly until age two (mean size = 193 mm) where it began to level off; and over 50% percent of the total growth was co mpleted by age two. Growth began to slow down after age three and then began to sligh tly increase after age six. Females were not observed after the age of six and males were observed to age 10. The von Bertalanffy growth equation, fitted to observed length at ages (all sexes combined) was L(t) = 306*(1-e-0.254(t+1.69)) (Fig. 9 and Table 5). The estimated L of 306 mm appears low based on observed maximum le ngth of 361 mm. The predicted sizes at age were similar to the observed length at age up to age 9; this coul d be due to the small sample size in the larger age groups (Table 6) The growth parameters estimated for the southern strata were L=268.5 mm, K=0.32, to=-1.54, and the central strata estimated parameters were L=292.9 mm, K=0.32, to= -1.36 (Fig. 10). The two growth curves were significantly different (P<0.01) based on th e randomization analysis. It appears that
25 0 50 100 150 200 250 300 12345678910AgeMean back-calculated fork length1 2 3 4 5 Figure 8. Mean back-calculated fork length be tween ages 1-5 as a function of age at capture. The lack of a significant trend (P>0.05) indicates the absence of LeeÂ’s phenomenon.
26 Table 3. Mean back-calculated FL at annulus formation of Calamus proridens all sexes combined. Back-calculated lengths were es timated using the biol ogical intercept method (Campana and Jones, 1992). Annulus group N 1 2 3 4 5 6 7 8 9 10 Observed mean SD 1 221 136 176 22 2 165 133 179 197 21 3 189 132 176205 218 26 4 130 132 176207232 243 22 5 64 131 168194221241 250 24 6 43 127 162187208227244 254 29 7 25 126 162191217239259275 283 31 8 12 126 163187207228245262277 285 21 9 3 125 162189211232250273290304 311 18 10 1 124 166201228252274293309328 347 355 Weighted means 133 174201223236249271281310 347
27 Table 4. Sample sizes, observed mean fo rk length (mm) at age, and range for Calamus proridens. Female Male Transitional All Age N Mean (SE) Range N Mean (SE) Range N Mean (SE) Range N Mean (SE) Range 0 233 128 (1.00) 76-165 1 136 136 242 127 (0.97) 76-165 1 372 157 (0.83) 104-212 34 173 (2.86) 141-203 11 168 (5.13) 131-201 422 158 (0.83) 104-212 2 217 188 (1.07) 141-243 19 220 (4.65) 189-262 9 207 (7.59) 173-253 259 193 (1.26) 141-262 3 116 203 (1.90) 170-294 50 236 (2.82) 200-285 14 227 (6.36) 179-266 190 215 (1.86) 170-294 4 56 229 (2.88) 190-283 68 248 (1.86) 205-289 4 239 (15.57) 212-284 139 240 (1.78) 190-289 5 10 240 (8.21) 207-294 54 259 (2.93) 211-322 1 208 208 75 253 (2.79) 207-322 6 8 265 (7.63) 235-297 27 270 (6.22) 220-361 49 258 (4.56) 212-361 7 20 284 (5.23) 260-352 1 307 307 25 279 (5.47) 228-352 8 9 290 (7.11) 265-322 9 290 (7.11) 265-322 9 3 301 (16.13) 276-331 3 301 (16.13) 276-331 10 2 332 (23.50) 308-355 2 332 (23.50) 308-355
28 Age 024681012 Fork Length (mm) 50 100 150 200 250 300 350 400 Figure 9. Von Bertalanffy growth model fitt ed to all observed length at age data for Calamus proridens Solid line = predicted growth curve, dotted lines = upper and lower 95% confidence intervals.
29 Table 5. Growth parameter estimates of Calamus proridens from the von Bertalanffy growth model with standard error and 95% confidence intervals. The AIC= AkaikeÂ’s Information Criteria quantifying th e fit of the growth model to the length at age data. Standard error for each parameter is in parentheses and the 95% CI is the lower and upper ranges. Parameters L K to AIC Strata n Estimate ( SE) 95% CI Estimate ( SE) 95% CI Estimate ( SE) 95% CI Southern 754 268.5 (5.96) 256.8 280.2 0.320 (0.026) 0.27 0.37 1.54 (0.12) -1.78 -1.29 1892 Central 558 292.9 (6.86) 279.4 306.4 0.317 (0.025) 0.27 0.37 1.36 (0.11) -1.56 -1.15 1395 All 1415 306.0 (5.72) 294.8 317.2 0.254 (0.015) 0.22 0.28 1.67 (0.09) -1.89 -1.51 3688
30 Table 6. Sample sizes, observed mean fork le ngth (mm) at age, and predicted length at age from the Von Bertalanffy growth model for Calamus proridens for all samples, and the central and southern strata All Central Southern Age N Mean (obs) PredictedN Mean (obs) PredictedN Mean ( obs) Predicted 0 242 127 130 111129 131 131126 129 1 422 158 157 153162 160 259155 154 2 259 193 192 110200 198 136186 188 3 190 215 218 81 219 224 100208 211 4 139 240 236 66 241 241 58 233 226 5 75 253 252 17 255 256 43 245 236 6 49 258 267 10 260 267 17 231 247 7 25 279 274 5 276 272 8 256 252 8 9 290 281 4 295 279 1 285 255 9 3 301 286 1 295 281 1 276 259 10 2 332 291
31 Age 0246810 Fork Length (mm) 50 100 150 200 250 300 350 Central region Theoretical growth curve for cenrtal region Southern region Theoretical growth curve for southern region Figure 10. Von Bertalanffy growth curves fitted to observed length at age for the central and southern regions of the west coast of Fl orida. Parameter estimates are given in Table 5.
32 while the BrodyÂ’s growth coefficient is the same for both strata the fish in the southern stratum do not grow as large as the fish in the central stratum. The model predicted lengths at age that were similar to the observe d lengths at age for the central and southern strata; again sample sizes were small for th e older age group (Table 6). While there were not enough samples to construct a growth curve from the northern region, the oldest fish (age 10) were collected in that re gion. The AIC numbers were lowest for the southern and central data indica ting a better fit to the model than when all the data were combined (Table 5). Estimates and statistics for all growth models are presented in Table 5. An age-length key was also created for C. proridens based on the observed age and length data collected (Table 7).
33 Table 7. Age length key for Calamus proridens from the northeast GOM. Length class = fork length at 20 mm intervals, N= number sa mples, values are proportions within each age class. Age Length class (mm) N 0 1 2 3 4 5 6 7 8 9 10 60 1 1.00 80 10 1.00 100 61 0.90 0.10 120 155 0.77 0.23 140 256 0.21 0.77 0.02 160 208 0.01 0.64 0.300.04 180 200 0.21 0.520.250.02 200 163 0.05 0.400.370.120.040.02 220 114 0.100.300.330.150.100.03 240 120 0.080.210.480.160.08 260 86 0.020.120.190.280.190.16 0.03 0.01 280 24 0.080.170.250.250.08 0.13 0.04 300 9 0.220.110.44 0.11 0.11 320 5 0.200.20 0.40 0.20 340 3 0.330.33 0.33 360 1 1.00
34 Reproduction Sex was assigned in the field to 1710 speci mens; sex was unable to be determined on 104 samples. Samples were comprised of 1371 (80.2%) females, 297 (17.4%) males, and 42 (2.5%) transitionals (T able 4). Females ranged from 76 to 297 mm FL and were 0 to 6 years old (Fig. 11). The mean length fo r females was 156 mm, with a mean age of 1.4 years; 82% of females aged were between 0 and 2 years. Male s ranged in length from 141 to 361 mm with a mean size of 244 mm. They ranged in age from 1 to 10 years old, with a mean age of 4.2 years. Over ha lf (58%) of the males aged were between 3 and 5 years; while only a few were older than 7 years. There was a significant difference in the age and length distribut ion for males and females: ma les were significantly older and larger than females (Age: KS Dmax=0.627 P<0.01, t-test F=2.41, t= -28.3 P<0.001; Length: KS Dmax=0.757, P<0.01, t-test F=1.06, t= -35.2 P<0.001). Females dominated the smaller size classes up to 190 mm, and male s dominated the larger size classes (Table 8). Length at age was also significantly di fferent for males and females for all ages except age six (ANOVA P<0.01). The overall sex ratio (1:4.6 males to female s) was significantly different from the expected 1:1 for a gonochoristic fish ( X2 = 691.5, P<0.001). Sex ratio by length class for males and females also had a significant devi ation from the expectant 1:1 in all size classes except length groups 220, 230, and 290 mm (Table 8). Sex ratio was further broken down by age and again chi square anal ysis showed a significant deviation from 1:1 for all ages except age 4 (Table 9). A total of 388 gonads were processed fo r histology and were either ovarian, ovotestes (hermaphroditic females), transitional, testicular, or hermaphroditic males.
35 A Fork Length (mm) 50100150200250300350 Frequency 0 20 40 60 80 100 120 140 B Age 012345678910 Frequency 0 100 200 300 400 Figure 11. Length A) and age B) frequency distribution of Calamus proridens collected from the northeast Gulf of Mexico from 2000-2007. Females = white bars, males = dark grey bars, transitionals = black bars. N = 1339 N=1710
36 Table 8. Number and percentage of females, males, and transitionals of C. proridens from the northeast Gulf of Mexico grouped in to 10 mm length classe s with sex ratio. Chi square test was conducted to test for significa nt difference from a 1:1 sex ratio, indicates P<0.01. Length Class Females Males Transitionals Male:Female N % N%N% 70 2 100.00 80 23 100.00 90 43 100.00 100 72 100.00 110 118 100.00 120 132 100.00 130 127 98.45 20.88 140 133 97.79 31.281:44.33 150 128 95.52 52.1810.431:25.60 160 106 89.83 73.3752.371:15.14 170 97 90.65 63.0441.991:16.17 180 86 89.58 94.8510.531:9.56 190 108 92.31 52.3941.891:21.60 200 80 81.63 126.6863.221:6.67 210 36 66.67 1411.6043.021:2.57 220 20 42.55 2426.8032.581:0.83 230 23 45.10 2424.9743.301:0.96 240 10 17.86 4358.2232.271:0.23 250 12 21.82 4153.3721.541:0.29 260 2 4.65 4083.9410.761:0.05 270 7 18.92 3053.651:0.23 280 2 12.50 1345.6111.351:0.15 290 4 57.14 34.681:1.33 300 787.5011.05 310 1100.00 320 4100.00 330 1100.00 340 1100.00 350 3100.00 360 1100.00 Total 1371 80.18 29717.37422.461:4.62
37 Table 9. Number and percentage of females, males, and transitionals of C. proridens from the northeast Gulf of Mexico by age cl ass with sex ratio. Chi square test was conducted to test for significant differen ce from a 1:1 sex ratio, indicates P<0.01. Age Females Males Transitionals Male:Female N % N%N% 0 233 99.57 0.0010.43 1 372 89.21 348.15112.641:10.94* 2 217 88.57 197.7693.671:11.42* 3 116 64.44 5027.78147.781:2.32* 4 56 43.75 6853.1343.131:0.82 5 10 15.38 5483.0811.541:0.19* 6 8 22.86 2777.14 0.001:0.30* 7 0.00 2095.2414.76 8 0.00 9100.000.00 9 0.00 3100.000.00 10 0.00 2100.00 0.00 Total 1012 75.58 28621.36413.061:3.54*
38 Macroscopic pictures of ovarian, transitiona l, and testicular are shown in Fig. 12. Ovotestes Â– ovarian tissu e with small quantity of testicular tissue (male cyst) attached to the ovarian wall (Fig. 13) were observed in 39% of the samples that had a complete ovarian wall. The testicular tissue of the ovotestes ranged in size and development; small and large cysts contained primarily spermata gonia, although some larg er cysts exhibited ongoing spermatogenesis. The majority of the ovotestes were observed in October, November, and April; however they were also observed in January, July, and December. The specimens with the testicular tissue undergoing spermatogenesis were larger and older (mean FL = 218 mm, mean age=2.3 years) than the ones only containing spermatogonia (mean FL = 165 mm, mean age= 1.1 years). Ovotestes only containing spermatogonia were observed as early as ag e 0, while ovotestes w ith spermatogenesis were not observed until age 1; both were observed up to age 4. Histological evidence confirms Calamus proridens is a monandric protogynous hermaphrodite. Transitionals were determin ed by the presence of proliferating male tissue and degenerating female tissue (Fig. 13). A total of 42 transitionals were observed, 38 of them confirmed through histology. Most specimens undergoing sex reversal ranged in size from 131-307 mm and were 1 to 5 years old (Fig. 11). One transitional was also observed at age 0 and another one at age 7. The mean size at sex reversal was 207 mm with a mean age of 2.4 years; and the majority of fish undergoing sex reversal (83%) were ages 1 to 3. The estimated ag e and length at which 50% of the population had transitioned from female to male wa s 4 years and 231 mm (Fig. 14). Transitionals were observed in January, April, July, Oc tober, and November, however, no samples were collected in September and less than 10 samples were collected in June, August, and
39 Figure 12. Macroscopic views of A) ovarian, B) transitional, and C) testicular gonads from Calamus proridens collected in the Gulf of Mexico. O=ovarian tissue, T=testicular tissue. B A C O T
40 Figure 13. Histological sections of Calamus proridens gonads. Transitional go nads with proliferating male tissue and degenerating female tissue from A) a 131 mm FL specimen collected in April 2007, B) a 167 mm FL specimen collected in April 2003, and C) a 225 mm FL specimen collected in April 2006. D) A spawning female gonad from a 205 mm FL specimen collected in April 2003. E) A spawning male gonad from a 216 mm FL specimen collected in April 2004. F) A regenerating female gonad with a small male cyst from a 163 mm FL speci men collected in November 2005. T=testicular tissue, O=ovarian tissue, S=Sperm, A=atretic oocyte, SP=spermatogenesis, VI= vitellogenic oocytes, GVM=germinal vesicle migration, POF=post ovulatory follicle, HO=hydrated oocyte, VD=vas deferens, MC=male cyst. Black scale bars are 0.1mm POF HO T O VD S T O D B E SP A S A F MC O M C T O POF GVM VI
41 Age 024681012 0 50 100 Length Class (mm) 50100150200250300350400Percent Females (%) 0 50 100 Figure 14. Percentage of female littlehead porgy, Calamus proridens by age and length class (10 mm size intervals) wi th a logistic function (con tinuous line) fitted to the observed data. Arrows indicat e predicted length and age at which 50% percent of the females transition into males. N = 1298 L50 = 4 years N = 1668 L50 = 231 mm
42 December. Sex reversal was most often observed (64%) during the non-spawning season in October and November when the fe males were usually regenerating, although transitionals were also observed during th e spawning season (26%). One transitional collected during the spawning season appeared to have equal amounts of male and female tissue in the same reproductive class. The female tissue ha d hydrated oocytes and post ovulatory follicle (POFs), and the male tissue was full of sperm. However, simultaneous hermaphroditism was unlikely (or at least ineffective) because the male did not have the means to disperse sperm: the vas deferens was not developed. Calamus proridens mature at a small size. Estimated size for females at 50% maturity was 132 mm (SE=2.25 mm) and maturity increased rapidly as length increased (Fig. 15). The largest immature observed was 156 mm. The logistic regression indicated that 50% of the individuals mature by the firs t year and all probably mature after the first year. All reproductive classes for female gonads were observed in the histological samples. Due to difficulty with artifact s in the histology and possible postmortem degeneration of the gonad, gonads classified as spawning-capable and actively spawning were grouped together as spawning. It was also difficult to determine the difference between immature and regenerating in some gonads. In these cases, the gonads were classified as unknown when it could not be de termined (n= 41). Female gonads showed signs of spawning from February through May; no samples were collected in June and by July the few samples observed were regenera ting. Atretic oocytes were observed in regressed gonads during April. Spawni ng females ranged in size from 138-297 mm (mean = 200 mm SE=3.74). Calamus proridens are batch spawners and may spawn daily
43 Length Class (Fork Length) 100120140160180200220240260280300Percent Maturity 0 50 100 Figure 15. Percentage of mature littlehead porgy, Calamus proridens by length class (10 mm size intervals) with a logistic function (cont inuous line) fitted to the observed data. Arrows indicate predicted size at wh ich 50% of the population was mature. N = 95 L50 = 132 mm
44 as the presence of POFÂ’s were observed with hy drated oocytes (Fig. 13). In the fall only immature or regenerating classes were observed. A total of 136 males were collected fo r histological evaluation. Hermaphroditic males (n=38) had testes that had functional ma le tissue but had traces of ovarian tissue. Timing of male reproduction was difficult to determine as the majority (89%) of the males examined had testes with sperm. Six samples had visible sperm ducts filled with sperm (Fig. 13) and I could clearly classify th em as spawning, while nine testes only had the presence of spermatogonia and spermato cytes and were most likely immatures. There was no distinction or separation in length between the possible groups.
45 Discussion Otoliths are a valid method for ageing C. proridens Annuli on whole otoliths can be accurately counted in fish up to one year of age. However, visibility of annuli decreases in whole otoliths as fish increase in age. Scales have been used to age other Calamus spp. but have not had the same accuracy as otoliths (Horvath et al. 1990, Waltz et al. 1982). In the present st udy, annuli in sectioned otolith s were easily identifiable with distinct hyaline and opaque bands. The f act that only two sect ioned otoliths were discarded from over 1000 samples due to disagr eement emphasizes the ease and accuracy of reading sectioned otoliths for C. proridens Therefore, sectioned otoliths should be the preferred method of aging C. proridens Clarity of rings can decrease as age increases, a fact that should be taken into consideration when using otoliths to age older specimens of Calamus or other sparids. Age validation The use of opaque bands in otoliths to age C. proridens has not been previously validated. Age validation studies for C. leucosteus, (whitebone porgy) and C. nodosus, (knobbed porgy) using scales and otoliths ha d difficulty verifying annulus deposition (Horvath et al., 1990; Waltz et al., 1982). The marginal increment (MI) analysis in this study indicated that C. proridens deposited a new annulus during the spring and summer. While the marginal increment data may appear somewhat variable, the variability could be due to small sample sizes in some mont hs. Seven months had less than 15 samples and five of those had less than 10 samples. It may be formation of the opaque band occurs
46 over a long period of time due to a protract ed spawning period. However, a significant difference in marginal increment was obser ved between April (smaller MI) and October (larger MI). Annulus deposition for the red porgy, Pagrus pagrus, has been verified and occurs during the same time period as C. proridens (Hood and Johnson, 2000, Pajuelo and Lorenzo, 1996). Marginal increment an alysis was able to show that annulus deposition is a yearly occurrence; however, validation for individual ages is still a good idea. Age and growth Results from the present study suggest that Calamus proridens in the GOM are moderately long-lived with an observed ma ximum age of 10 years and a maximum size of 361 mm. Similar maximum sizes were observed in the Campeche Bank with a maximum size of ~340 mm (type of length meas urement unspecified), but ages were not determined (Dubovitsky, 1977a, Dubovitsky, 1977b) The present study is the first to determine age of C. proridens There are limited age and gr owth studies on porgies, and of those Calamus proridens is most similar to the whitebone porgy, Calamus leucosteus in age and size. Calamus leucosteus from the South Atlantic Bight had a maximum age of 12 and a maximum size of 407 mm FL (Waltz et al. 1982). The red porgy, Pagrus pagrus and the knobbed porgy, Calamus nodosus, live longer (age 18 and 17 respectively) and obtain a larger size (470 mm TL and 544 mm TL respectively) than C. proridens (Hood and Johnson, 2000, Horvath et al. 1990) Back-calculated estimates of length at the time of annulus deposition did not exhibit LeeÂ’s phenomenon, and were in the sa me range as the observed mean length at age. However, it seemed most appropriate to estimate growth curves using observed data
47 since the observed data had age zero fish, where back-calculated estimates did not. The asymptotic maximum length (L) of 306 mm estimated from the von Bertalanffy growth model for all data co mbined appeared low based on an observed maximum length of 361 mm. While separating th e data based on latit ude created a better fit to the model, the small number of larger a nd older fish could have also contributed to the low estimates of L. Similar findings of low L were found with the whitebone porgy (Waltz et al. 1982). The rapid growth of C. proridens from the GOM in the first two years was also observed in C. proridens collected from the Campeche bank by Dubivinsky (1979b), even though specific growth parameters were not reported in his studies. Based on growth parameter comparisons C. proridens grows at a similar rate to the knobbed porgy (Hood and Johnson, 2000, Horvath et al. 1990), red porgy (DeVries, 2006) and the white grunt, Haemulon plumieri (Murie and Parkyn, 2005). Calamus proridens also grew initially faster than most groupers and snappe rs; this could be attr ibuted to grouper and snappers much longer life span or that mo st grouper and snapper studies do not have young-of-the-year data. Calamus proridens collected in the southern strata (26 latitude Â– off of Charlotte Harbor) grew at a similar rate, but attained a smaller maximum size than in the central strata (27 latitude Â– off of Tampa Bay). Diffe rences could be attributed to water quality, habitat, food availability, predators, comp etition and/or fishing pressure (Murie and Parkyn, 2005, Thurman, 2004). Local differences in size composition could be due to differences in mortality rates rather than growth rates (Murie a nd Parkyn, 2005). Further research on environmental parameters and addi tional biological studie s could be collected
48 to further explain the differences that were observed. While there is not a commercial fishery for C. proridens, they are a significant bycatch in other commercial fisheries such as the red snapper and shrimp fishery. Ev en though they are not as heavily fished recreationally as groupers, snappers, and red porgy; C. proridens are recreationally fished. With increased restrictions on other reef species, fishermen will begin to fish for species with little to no restrictions. Having data available will help obtain accurate stock assessments and to help establish and mainta in an overall ecosystem based approach to monitoring reef species in the GOM. An age-length key has not b een previously estimated for C. proridens The age length key is another tool to us e in stock assessments that can be used to estimate age for fish caught in the northeast GOM. Based on this key fish cau ght between 140 Â– 179 mm FL, in the northeast of the GOM, would be ag e one. Due to low sample sizes in the large length classes the age length key may not be an accurate estimate for age of fish greater than 300 mm FL. This also indicates that the largest fish are not necessarily the oldest fish. Reproduction Female C. proridens in the northeast GOM reached maturity at a small size (132 mm) and within the first year. Research on the whitebone porgy, whic h in most respects is quite similar to the littlehead porgy in life history patterns, indicated that they probably mature at age 1 (Waltz et al. 1982). They also found hydrated eggs in specimens as small as 179 mm FL. In comparison, the red po rgy matures at a grea ter age (ages 2 4) and a larger size (226-250 mm TL) (Hood and Johnson, 2000, Kokokiris et al. 1999, Pajuelo and Lorenzo, 1996).
49 The population of C. proridens in the northeast GOM was dominated by females as indicated by a highly skewed sex ratio to wards females, a resu lt consistent with protogyny, other sparids, and specifically other Calamus spp. The Campeche Bank population from DubovitskyÂ’s (1977 b) studies also showed a female skewed sex ratio, but had a smaller ratio than the GOM with a 1:2.7 in favor of females. The knobbed porgy did not show a significant bias toward females in the overall sex ratio, but this could be due to the lack of small fish in th eir collections. Sex ratio statistics were not reported for the whitebone porgy, but females accounted for about 80% of the smaller size classes and males accounted for 70% of the larger size classes (Waltz et al. 1982). Reproductive biology in the family Sparidae is complex. In the sparid family there are species exhibiting protandry, prot ogyny, and rudimentary hermaphroditism, in addition to separate sexes (Alekseev 1982, 1983, Buxton and Garratt, 1990, Garratt, 1986). A rudimentary hermaphrodite consists of an immature intersexual gonad that later matures as a male or female with evidence of sex reversal (Buxt on and Garratt, 1990). Although reproduction of many sparids has been described, mainly in South African species, there are only two studies on reproductive biology within the genus Calamus. The histological analysis of the present study has shown C. proridens to be a monandric sequential protogynous hermaphrodite, which is consistent with the criteria from Sadovy and Shapiro (1987) on hermaphroditism. The nature of sex reversal in the sparid family is of the delimited type where testicular and ovarian tissues are separated by connective tissue. In protogynous species th e testicular tissue proliferates and envelops the ovarian tissue as it begins to regress; this is th e opposite of other protogynous families, such as Serranidae (groupers an d seabasses) (Bullock et al. 1996, Cochran and Grier, 1991,
50 Coleman, 1981, Fischer and Petersen, 1987, Shapiro, 1987, Thurman, 2004), Labridae (wrasse) (Shapiro and Rasotto, 1993, Wa rner and Swearer, 1991), and Scaridae (parrotfish) (Munoz and Warner 2004), where one sex infiltra tes within the ovarian or testicular cell wall and ca nnot be seen without histological evaluation. During my research, at times, I was able to see both testicular and ovarian tissue macroscopically (Fig. 11) and could define them as transitionals, however, since most transitionals were observed when the female was in the rege nerating class and could not be seen macroscopically, histological evaluation was needed to fully demonstrate protogynous hermaphroditism in the genus Calamus The highly biased sex ratio toward female s and the significantly different length and age distributions were further eviden ce for protogyny. Moreover, the fact that no males were found in the smaller size classes a nd that they dominated the larger sizes and older ages strongly suggested that males aris e only from sex change (Fig. 10 and Table 4). The presence of an ovotestis is comm only observed in delimited gonads; however, the presence of a small testicular cyst does not mean that a female will undergo sex reversal (Kokokiris et al. 1999). The presence of large females demonstrated that not all females undergo sex reversal. The mean length of transitionals was 207 mm which was slightly smaller than the Campeche Bank population (240 mm length measurement unspecified) (Dubovitsky, 1977b). However, a 207 mm FL converts to a 241 mm TL from the length to length conversion from Table 1 and would coincide almost exactly with DubovitskyÂ’s findings. While the majority transitioned around the mean I did find specimens that transitioned at smaller and larger sizes (Table 4). The mean length of transitionals was always larger
51 than the females at the same age, which could indicate that the larger females in each age group were the first to change sex which was also observed in the protogynous common pandora, Pagellus erythrinus (Alekseev, 1983). The increase in growth rate after sex reversal could also be an explanation for w hy the mean length at age of males was larger than females. Ovotestes are common in the sparid family and occur in a wide range of ages and sizes, indicating that sex revers al may not be limited to a sp ecific size or age (Kokokiris et al. 1999). Kokokiris et. al ( 1999) suggested that there we re three possible pathways for sexual pattern and reproduction in the red porgy. The suggested pathways were 1) females reproduce and then change sex, 2) fe males change sex before reproducing as a female, or 3) the female may not undergo sex change at all and rema in female (Kokokiris et al. 1999). Due to the differences in si ze and age of observed fish undergoing sex reversal, the sexual pattern in C. proridens may be similar to that of the red porgy. While many studies have tried to explain the biological reasons for hermaphroditism, one theory has not become clearly dominant. The overall goal for species survival is to maximize abundance by a high rate of reproduction coupled with a high rate of survival of offspring to repr oductive age. Hermaphroditism has been one adaptation to help species reach this goa l. The implications for sex reversal are particularly unclear for the sparid family. The Â“size advantageÂ” model suggests that a species will change sex when it is advantageous to reproduce first as one sex and then as the other sex (Ghiselin, 1969, Munoz and Warner, 2004, Warner, 1975). From an a priori standpoint, it would seem to be mo st advantageous to be female when the body cavity is large and can hold more eggs, so why are the males larger in C. proridens and
52 why are females transitioning at such a small size? In ot her protogynous species, such as groupers and wrasses, studies have demonstrated a complex social structure that controls sex reversal (Fischer and Petersen, 1987, Warner and Swearer, 1991). Unfortunately, mating systems have only received moderate atte ntion in the sparids, and appear to be at least as complex as their reproductive biology. One study by Buxton and Garratt (1990) described three different types of mating from three different species in the sparid family: 1) demersal spawners, 2) pair spawners, and 3) dense spawning aggregates with similarly sized individuals. Dubovits ky (1977a) suggested that C. proridens are sequential spawners, have different spawning grounds for different groups, and indicated that they do not form dense spawning aggregations. Mating systems within the sparids are apparently complex, knowing the mating struct ure for individual species would be most helpful in determining cues for sex reversal. The sex ratio/size ratio threshold hypothe sis combined with th e density dependent hypothesis could explain induc tion of sex reversal in C. proridens The sex/size ratio models suggest that females may be induced to change sex (in the presence of a male) when a threshold number of females is reached or a threshold level of small to large individuals is reached for a given female (Lutnesky, 1994, Ross et al. 1983, Shapiro, 1987). For example, in the sizeratio model, Ross et al. (1983), showed that in the saddleback wrasse ( Thalassoma duperrey ), sex change was stimulated by the presence of at least one other smaller fish and did not requ ire removal of the male or largest fish. The theory is that if there is a biased ratio toward larger fish (mainly males) and few smaller fish (mainly females) there are too few fema les for a new male to mate with and too many large males to compete with (Ross et al. 1983). However, if the ratio of smaller to
53 larger fish is reversed and there are few larger fish, it may be reproductively advantageous to change sex. The sexratio model is similar in that there are a certain number of interactions a female is used to having with a male a nd once that interaction rate is changed (decreased with increase in females) the female will change sex. In addition, different densities can induce sex change depending on the encounters assumed to be important in the sex change process, this model is similar to the sex/size-ratio model except it incorporates fish density and its effects on behaviors and simple proximity (Lutnesky, 1994). Therefore, sex change in C. proridens may not be as much about between-sex interactions as it is about in teractions within sexes. Many variables contribute to sex reversal: environmental, bi ological, social/behavio ral, sex allocation, density-dependent, mortality, fishing pressure and/or population changes. More than likely it is not just one factor that infl uences sex change but a combination of environmental and biological factors. More detailed field and/or experimental studies on mating behaviors are needed to better unde rstand the reasons for sex reversal in C. proridens
54 Conclusion Even though Calamus spp. provides an important commercial and recreational fishery in the Campeche Bank and in the Gulf of Mexico, the status of fishing stocks has not been determined. Life history inform ation for many porgies is unknown or limited. The present research include d life history information on age, growth, reproduction, and sex change that can be in cluded in stock assessment models and can help manage fisheries resources in the GOM as management advances fr om a single-species to an ecosystem based model. Hermaphroditic species respond differe ntly than gonochor istic species to overfishing (Alonzo, 2003, Crabtree and Bu llock, 1998). Models for estimating dioecious stocks are not appropr iate for estimating stocks of sex-changing species. Sizeselective fishing pressure on a hermaphroditic species could affect the overall dynamics of the population and influence sex reversal. For example, if all the males are being removed this will bias the sex ratio towards females, which may affect overall spawning success resulting, in smaller a nd smaller females undergoing sex reversal (Alonzo, 2003, Crabtree and Bullock, 1998, Hood and J ohnson, 2000). Managing stocks of sexchanging fish will require considering the sex reversal pattern, however, it must be incorporated within the cont ext of the mating system (Alonzo, 2003). Stock assessment models should not be based on one life history parameter, but encompass as many life history parameters as are available. Models th at are limited in information must be used with caution and evaluated frequently as new information is discovered.
55 While this research is a start in understa nding the life history characteristics of C. proridens, additional parameters still need to be estimated. Additional research on fecundity, mortality, behavioral and mating characteristics ar e still needed to better understand the biology of the species and to improve stock assessment models.
56 Literature Cited Alekseev, F. E. 1982. Hermaphroditism in sparid fishes (Perciformes, Sparidae). 1. Protogyny in porgies, Pagrus pagrus, P. orphus, P. ehrenbergi and P. auriga from West Africa. Journal of Ichthyology 22 :85-94. Alekseev, F. E. 1983. Hermaphroditism in Porgies (Perciformes, Sparidae). 1. Sexual structure of the population, mechanism of its formation and evolution in Scups, Pagrus pagrus, P. orphus, P. ehrenbergi and P. auriga Journal of Ichthyology :61-72. Alonzo, S. H. 2003. The effects of size-selective fisher ies on the stock dynamics of and sperm limitation in sex-changing fish. Fishery Bulletin 102 :1-13. Armsworth, P. R. 2001. Effects of foshing on a protogynous hermaphrodite. Canadian Journal of Fisheries and Aquatic Sciences 58 :568-578. Barbieri, L. R., M. E. C. Jr., and C. M. Jones. 1994. Age, growth, and mortality of Atlantic croaker, Micropogonias undulatus, w ith a discussion of apparent geographic changes in population dynamics. Fishery Bulletin 92 :1-12. Besseau, L., and S. Brusle-Sicard. 1995. Plasticity of gonad develpment in hermaphroditic sparids: ovotestis ontogeny in a protandric species, Lithognathus mormyrus Environmental Biology of Fishes 43 :255-267. Brown-Peterson, N.J., S.K. Lowerre-Barbie ri, B.J. Macewicz, F. Saborido-Rey, J. Tomkiewicz and D.M. Wyanski. 2007 An improved and simplified terminology for reproductive classification in fi shes. Pre-print doc. doi: http://hdl.handle.net/10261/11844 Bullock, L. H., M. F. Godcharles, and R. E. Crabtree. 1996. Reproduction of the yelloedge grouper, Epinephelus flavolimbatus from the eastern Gulf of Mexico. Bulletin of Marine Science 59 :216-224. Buxton, C. D., and P. A. Garratt. 1990. Alternative reproductiv e styles in seabreams (Pisces: Sparidae). Environmental Biology of Fishes 28 :113-124. Campana, S. E., and C. M. Jones. 1992. Analysis of otolith microstructure data. P.^Pp. 73-100 in Otolith microstructure examination and analysis Chambers, R. C., and T. J. Miller. 1995. Evaluating fish growth by means of otolith increment analysis: Special properties of individual-level long itudinal data. P.^Pp. Cochran, R. C., and H. J. Grier. 1991. Regulation of sexual succession in the Protogynous black sea bass, Centropristis striatus (Osteichthyes: Serranidae). General and Comparative Endocrinology 82 :69-77. Coleman, F. 1981. Protogynous Hermaphroditism in the Anthiine Serranid Fish Holanthias martinicensis Copeia 4 :893-895. Crabtree, R. E., and L. H. Bullock. 1998. Age, growth, and reproduction of black grouper Mycteroperca bonaci in Florida waters. Fishery Bulletin 96 :735-753. Darcy, G. H. 1986. Synopsis of Biological data on the Porgies, Calamus arctifrons and C. proridens (Pisces: Sparidae). NOAA Technical Report NMFS 44 148 :i-19.
57 Davis, M. L., and J. Berkson. 2005. Effects of a simulated fishing moratorium on the stock assessment of red porgy ( Pagrus pagrus ). Fishery Bulletin 104 :585-592. DeVries, D. 2006. The life history, reproductive ec ology, and demography of the red porgy, Pagrus pagrus in the northeastern Gulf of Me xico. PhD Dissertation, Florida State University, Tallahassee. Dubovitsky, A. A. 1977a. Distribution, migrations and some biological features of littlehead porgy ( Calamus proridens Jordan and Gilbert, 1884) family Sparidae, of the Gulf of Mexico. Dubovitsky, A. A. 1977b. Sex ratio and length-sex composition of the Campeche-Bank littlehead porgy ( Calamus proridens Jordan and Gilbert 1884) population. P.^Pp. 115122 in Symposium on progress in marine resear ch in the caribbean and adjacent regions FAO, Rome (Italy). Fisheries Resources and Environmental Div. Fischer, E. A., and C. W. Petersen. 1987. The evolution of se xual patterns in the seabasses. BioScience 37 :482-489. Francis, R. I. C. C. 1990. Back-calculation of fish length: a critical review. Journal of Fish Biology 36 :883-902. Froese, R. 2006. Cube law, condition factor and weight-length relationships: history, meta-analysis and recommendations. Journal of Applied Ichthyology 22 :241-253. Garratt, P. A. 1986. Protogynous hermaphroditism in the slinger, Chrysoblephus puniceus (Gilchrist & Thompson, 1908) (Teleostei:Sparidae). Journal of Fish Biology 28 :297-306. Gettel, G. M., L. A. Deegan, and C. J. harvey. 1997. A comparison of whole and thinsectioned otolith aging techniques and validation for Arctic grayling. Northwest Science 71 :224-232. Ghiselin, M. T. 1969. The Evolution of Hermaphroditism among Animals. The Quarterly Review of Biology 44 :189-208. Gotelli, N. J., and A. M. Ellison. 2004. A Primer of Ecological Statistics Sinauer Associates, Inc., Sunderland. Haung, C.-C., C.-F. Lo, and K.-H. Chang. 1974. Sex reversal in one sparid fish, Chrysophrys major (Perciformes, Sparidae). Bulletin of the Institute of Zoology, Academia Sinica 13 :55-60. Hood, P. B., and A. K. Johnson. 2000. Age, growth, mortality and reproduction of red porgy, Pagrus pagrus from the eastern Gulf of Mexico. Fishery Bulletin 98 :723-235. Horvath, M. L., C. B. Grimes, and G. R. Huntsman. 1990. Growth, mortality, reproduction and feeding of the knobbed porgy, Calamus nodosus along the southeastern United States coasts. Bulletin of Marine Science 46 :677-687. Ihde, T. F., and M. E. Chittenden. 2003. Validation of presumed annual marks on sectioned otoliths of spotted seatrout, Cynoscion nebulosus in the Chesapeake Bay region. Bulletin of Marine Science 72 :77-87. Kokokiris, L., S. Brusle, M. Kentouri, and A. Fostier. 1999. Sexual maturity and hermaphroditism of the red porgy Pargus pargus (Teleostei: Sparidae). Marine Biology 134 :621-629. Kokokiris, L., A. Fostier, F. Athanassopo ulou, D. Petridis, and M. Kentouri. 2006. Gonadal changes and blood sex steroids le vels during natural sex inversion in the protogynous Mediterr anean red porgy, Pargus pargus (Teleostei: Sparidae). General and Comparative Endocrinology 149 :42-48.
58 Lee, M.-F., J.-D. Huang, and C.-F. Chang. 2008. Development of ovarian tissue and female germ cells in the protandrous Black Porgy, Acanthopagrus schlegeli (Perciformes, Sparidae). Zoological Studies 47 :302-316. Lutnesky, M. M. F. 1994. Density-dependent protogynous sex change in territorialharemic fishes: models and evidence. Behavioral Ecology 5 :375-383. Manly, B. F. J. 1991. Randomization and Monte Carlo Methods in Biology Chapman & Hal, 2-6 Boundary Row, London SE1 8HN, UK. Munoz, R. C., and R. R. Warner. 2004. TEsting a new version of the sixe-advantage hypothesis for sex change: sperm competition and size-skew effects in the bucktooth parrotfish, Sparisoma radians Bevhavioral Ecology 15 :129-136. Murie, D. J., and D. C. Parkyn. 2005. Age and growth of the white grunt ( Haemulon plumieri ): A comparison of two populations along the west coast of Florida. Bulletin of Marine Science 76 :73-93. Pajuelo, J. G., and J. M. Lorenzo. 1996. Life history of the red porgy Pagrus pagrus (Teleostei: Sparidae) off the canary Islands, central east Atlantic. Fisheries Research 28 :163-177. Pannella, G. 1974. Otolith growth patterns: an aid in age determination in temperate and tropical fishes. P.^Pp., D. o. Geology, ed. University of Puerto Rico. Pierce, D. J., and B. Mahmoudi. 2001. Nearshore Fish Assemblages Along the Central West Coast of Florida. Bulletin of Marine Science 68 :243-270. Randall, J. E., and D. K. Caldwell. 1966. A review of the sparid fish genus Calamus with descriptions of four new species. Bulletin of the Los Angeles County Museum of Natural History 2 :1-47. Ricker, W. E. 1975. Computation and Interpretation of Biological Statistics of Fish Populations The Journal of the Fisheries Re search Board of Canada, Ottawa. Ross, R. M., G. S. Losey, and M. Diamond. 1983. Sex change in a coral-reef fish: dependence of stimulation a nd inhibition on relative size. Science 221 Sadocy, Y., and D.Y. Shapiro. 1987. Criteria for the diagnosis of hermaphroditism in fishes. Copeia 1 :136-156. SAS Institute 1996 SAS/STAT guide for personal co mputers. Version 6 edition. SAS institute, INc., Cary North Carolina. Schirripa, M. J. 2002. An evaluation of back-calcu lation methodology using simulated otolith data. Fishery Bulletin 100 :789-799. Shapiro, D. Y. 1987. Differentiation and evolution of sex change in fishes. BioScience 37 :490-497. Shapiro, D. Y., and M. B. Rasotto. 1993. Sex differentiation and gonadal development in the diandric, protogynous wrasse, Thalassoma bifasciatum (Pisces, Labridae). Journal of Zoology 230 :231-245. StatSoft, Inc. 2005 STATISTICA (data analysis software system), version 7.1. www.statsoft.com. Sokal, R. R., and F. J. Rohlf. 1981. Biometry W. H. Freeman and Company, New York. Thurman, P. E. 2004. Life history and energy budget of roughtongue bass, Pronotogrammus martinicensis (Serranidae: Anthiinae), Univ ersity of South Florida, Saint Petersburg.
59 Waltz, C. W., W. A. Roumillat, and C. A. Wenner. 1982. Biology of the whitebone porgy, Calamus leucosteus in the South Atlantic Bight. Fishery Bulletin 80 :863-874. Warner, R. R. 1975. The adaptive significant of sequent ial hermaphroditism in animals. The American Naturalist 109 :61-82. Warner, R. R., and S. E. Swearer. 1991. Social control of sex change in the bluehead wrasse, Thalassoma bifasciatum (Pisces: Labridae). Biological Bulletin 181 :199-204.