Vol. 9 No. 3 January 1996 ISSN: 0072-9027 REPORTS GULF RESEARCH Published by the UNIVERSITY OF SOUTHERN MISSISSIPPI GULF COAST RESEARCH LABORATORY Ocean Springs, Mississippi Gulf Research Reports Volume 9 | Issue 3 January 1996 Gonadal Maturation in the Cobia^ Rachycentron canadum, from the Northcentral Gulf of Mexico Jeffrey M. Lotz Gulf Coast Research Laboratory, jeff.lotz^usm.edu Robin M. Overstreet Gulf Coast Research Laboratory, robin.overstreet(^usm.edu James S. Franks Gulf Coast Research Laboratory, jim.franks(^usm.edu DOI: 10.18785/grr.0903.01 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation LotZ; J. M., R. M. Overstreet andj. S. Franks. 1996. Gonadal Maturation in the Cobia^ Rachycentron canadum, from the Northcentral Gulf of Mexico. Gulf Research Reports 9 (3): 147-159. Retrieved from http:// aquila.usm.edu/gcr/vol9/iss3/ 1 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9, No. 3. 147-159, 1996 Manuscript received August 17, 1995; accepted September 7, 1995 GONADAL MATURATION IN THE COBIA, RACHYCENTRON CANADUM, FROM THE NORTHCENTRAL GULF OF MEXICO Jeffrey M. Lotz, Robin M. Overstreet and James S. Franks Guif Coast Research Laboratory, P,0, Box 7000, Ocean Springs, Mississippi 39566-7000, USA ABSTRACT Gonadal maturation of cobia, Rachycentron canadtm, was evaluated by examining 508 specimens from its recreational fishery. Specimens were collected off southeast Louisiana to northwest Florida by hook-and-line during February through October 1987-1991. Fork lengths (FL) of these fish ranged from 580-1,530 mm, with corresponding weights of 2.0-43.5 kg. The femalermale ratio was 1 ;0,37. Using a combination of oocyte size- frequency and histological assessment of many of the fish, we determined that females wercripie from May through September, with atretic oocytes occuning in some fish from July through October. Degenerating hydra ted oocytes in July and October and the presence of resting ovaries in July suggest two major spawning pieriods; however, monthly gonosomatic indices peaking in May, followed by a steady decline, do not support that finding. Ovaries weieplaced. into undeveloped, early developing, mid-developing, or late developing categories based upon oocyte size-frequency distributions. Developing ovaries had two or three modes of oocytes larger than 30 pm. Batch fecundity was estimated to be 2.6x10^ to 1.91x10® oocytes, depending on the size of Fish/ovaries. The smallest female with oocytes exhibiting vitellogenesis was 834 mm FL. This fish was 2 years old based its otolith evaluation. The smallest male with an abundance of spermatozoa in its testes was 640 mm FL and 1 year old based on otolith evaluation; smaller males were not examined. Females larger than 840 mm FL had vitellogenic oocytes in March and April. A few fish still had vitellogenic oocytes in early October, but none did by late October. When Gilson's fluid was used to assess ovarian tissue, the fresh weight of the tissue was reduced by 20% after being stored for 3 months. The diameter of oocytes shrunk about 25% in Gilson's fluid which was 11% less than those fixed in formalin, embedded in paraffin, and sectioned. Tissue sections from specific individuals, each demonstrating a variety of different developmental stages, were similar regardless of whether they were obtained from the anterior, middle, or posterior portion of either ovary. INI’RODUCTION The cobiSL, Rachycentron canadurn (Goode 1884), is a pelagic fish that is found throughout most of the wann ocean waters of the world, except for the Pacific coast of North America (Migdalski and Fichter 1983). hi the western Atlantic Ocean, R. canadurn occurs from Massachusetts to Argentina and is common in the Gulf of Mexico (Shaffer and Nakamura 1989). In the Gulf of Mexico (Gulf), cobia migrate from their wintering grounds off die southern Florida coast into the waters of the northern Gulf in late March and April and return to their wintering grounds in late autumn and early winter (Biesiot et al. 1994; Franks 1991b). However, a relatively large number of fish appears to remain in the northcentral Gulf during winter months at depths of 1 00- 125 m (Howse et al. 1 992). The cobia is a highly prized recreational species in the Gulf and U.S. South Atlantic Ocean. Most of the U.S. cobia landings come from Gulf waters (Shaffer and Nakamura 1989). Although most cobia are caught by recreational anglers, some are caught incidentally in U.S. commercial fislieries (Shaffer and Nakamura 1989). Relatively little is known about the reproductive biology of cobia. Joseph etal. (1964) described eggs and juveniles collected from Chesapeake Bay and thenearby Atlantic Ocean andsuggested thaispawnmgoccumed during summer. Richards (1967) , also working in Chesapeake Bay, documented sexual dimorphism insize atmaturity, presented evidenceforspawning from late June through mid-August and postulated that multiple spawnings might occur. Dawson ( 197 1) .suggested dial spawning occurred during spring in the coastal waters of the northern Gulf of Mexico. P^ucane et al. ( 1978) reported evidence that cobia spawned off the coast of Texas m July and September, and Thompson et al. (1992) observed peak spawning in cobia from May through July inLouisiana coastal waters. Biesiot et al, (1994) described the biochemical changes in developing ovaries in cobia from the northern Gulf of Mexico and reported that spawning occurred during spring and summer. Our study was undertaken to answer the following questions forcobia in the northcentral Gulf of Mexico: 1 ) what is the minimum size (length) of fish at maiurity ; 2) what is the temporal period of reproductive activity; 3) does the cobia spawn more than once per spawning season, and if so, what is the estimated batch fecundity of a female? We attempted to answer these questions through an analysis of oocyte size-frequency distributions, gonadal histology and the gonosomatic index (GSI). 147 LoTZET AL. Materials and Methods Cobia examined in this study were caught by hook-and-line in the recreational fishery off southeast Louisiana, Mississippi, Alabama, and northwest Florida during February through October 1987-1991. Additional gonad samples from six cobia caught from the Gulf side of the Florida Keys in January 1991 were used ibr h islological evaluation only. In addition to those fish we caught, some were provided by recreational fishermen as well as state and federal fisheries agencies. Fish were stored on ice from the time of capture until examined dockside or when received at coastal fishing tournaments. Fork length (FL) and total length (TL) were measured 'm nun, and total body weight (TW) was recorded to the nearest 0.1 kg. The pair of gonads was removed, placed in a resealableplaslic bag, and stored in an ice slurry for up to 20 h. Total gonad weight was recorded to the nearest 0.1 g. A small subsample of each gonad was weighed to the nearest 0.1 g and fixed in 10% buffered formalin forhistological examination. A second subsample of each ovary was weighed to the nearest 0. 1 g and fixed in Gilson’s fluid to facilitate estimation of oocyte numbers and size- frequency distributions. A gonosomatic index (GSl) was calculated for both males and females: GSl = gonad weighl/total fish weight) x 100. Shrinkage of oocytes due to fixation was estimated by measuring the largest oocytes from fresh gravid ovaries, formalin-fixed-paraffin-embedded gravid ovaries, and Gilson’s-fixed gravid ovaries. An estimate of the weight loss due loGilson’s fixative was determined by weighing asample of fresh ovary at the time of collection , and then reweighing the same sample after 3 months in Gilson’s fixative. Ovarian tissue remained in Gilson’.s fluid for at least 3 months prior to estimating oocyte density and oocyte size- frequency distribution. ABioquant® image analysissystem was used to expedite oocyte counts and measurements. The number of oocytes in an aliquot was determined using a counting chamber. Oocyte density was determined from the number of oocytes in corrected- weight aliquots of Gilson’s fixed tissue and expressed as the number of oocytes per gram of fresh ovarian tissue. The total number of oocyles per female was obtained by multiplying the oocyte density by the total ovarian weight. The frequency distribution of oocytesizes was obtained by measuring the maximum distance across 100-200 randomly selected oocytes greater than 30 pm in diameter from an aliquot of Gilson’s-fixed tissue. Presumptive batch fecundity, the presumed number of eggs released during each spawning episode, was determined on the basis of the percentage of oocytes appearing as the most advanced standing stock of oocytes in late developing ovaries. Samples for histological analysis were embedded in parafiin from Hemo-De® xylene subsdtute.chilled, sectioned at 4 pm, stained with Gills hematoxylin, and counterstained with eosin-phloxine. Oocytes from the coverslipped slides were then staged according to sexual maturity. To determine whether the distribution of oocyte stages was homogeneous between ovaries and among anterior, middle, and posterior portions of each ovary; we examined histologically wedge- sluqred samples from the wall to the lumen at those sites. Size (length) at maturity was determined as the smallest fish which exhibited vitellogenesis or spennatogenesis. The agesof several fish examined during ihisstudy weredetermined as part of a concurrent study estimating age by otolith analysis (Franks etal. 1991a). Stanstical analyses were performed using Systat® software (Wilkinson 1990). Overall significance among group means was determined by the Kniskal-Wallis test (P<0.05); significance between pairs of means was detennined by the Mann Whimey U-teslusing Bonferroni’s correction (P<0.05). Batch fecundity data were transfonned to logarithms (log j^,) to normalize the data before correlationanaly ses were performed. Results A total of 508 cobia (374 females and 134 males) was sampled for reproductive analyses. The sex ratio of fish examined in Ihis study , 1:0.36 (female:male), wasrepresentative of the sex ratio of cobia entered in fishing toumamenis within our study area. Total weight and FL ranges among all specimens were 2.043.5 kg and 580- 1 ,530 mm, respectively. Seasonal pattern of maturation Alladultmales> 640mmFL (A= 1 34) and females >834 mm FL (Ar=361) were used for GSl calculations (Figure 1). Ovarian weights from the adult females ranged from 0.3% to 12.5 % of total body weight. Results of graphing GSl against month of collection revealed that the ovaries comprised an increasing proportion of body weight in spring, with a marked peak in GSl mean value of 5.0 in May, followed by a steady decline throughout summer and into autumn (Figure 1). Figure 1 also shows that the GSl for males is essentially the same as that for females. Figure 2 shows the seasonal dynamics of vitellogenesis. Three of four females caught off Mississippi during March 1991 were in early vitellogenesis. In April, when cobia first appeared in near shore waters of the northern Gulf, all females > than 834 mm FL were vitellogenic. The peak period of ovarian development occurred from April through Juneduring whichall lemales> 834 mmFL were vitellogenic. Ovarian developmeniinoursamples decreased inlate summer and autumn. Although a few fish were viieUog^c in early October, none were vitellogenic: in late October. 148 Gonadal Maturation in Gobi a from Northcentral Gulf of Mexico Apr May Jun Jul Aug Sep Oct Month Figure 1. Gonosomatic indices for adult male iN=lM) and female (Nr=361) cobia, Rachycentron canadum (means ± 1 standard error of the mean). Figure represents a composite of years 1987-1990. Weight loss due to Gilson’s fluid The mean weight loss of ovarian tissue fixed in Gilson’s fluid was 20% (S.E.=1,I%) of the fresh weight. This number was used as a correction factor for calculating the number of oocytes in fresh ovaries. Shrinkage of oocytes due to fixation The largest oocytes observed in fresh, well-developed ovaries were 950- 1000 min diameter. The largest oocytes in groupsof large oocytes observed in Gilson’s-fixed tissuefrom sitesadjacent to those where freshmaterial was obtained from the same ovaries were 700-750 pm. Therefore, we estimate thatdiameiershrinkageductoGiLson’streaim0iiwasaboui25%. Examination of ovarian tissue from individual females, both by histological techniques and by Gilson’s fixation, allowed for a more precise estimate of the relative shrinkage from the two treaunents. Examinationof 37 fish revealed that the diameter of oocytes fixed in Gilson’s fluid was 11% (S£.=3%) less than the diameter of oocytes treated by foimalin fixation, followed by paraffin embedding. Oocyte size-frequency distributions Oocyte size-frequency distributions were estimated for 131 cobia. Inspectionoftheoocyte size-frequency distributions coupled with examination of histological sections of ovaries allowed fish to be placed intoone of four groups representing various stages of ovarian development (Figure 3). Group I, Undeveloped, Twenty-nine of the 131 fish for which oocyte frequencies were determined wctc placed into the first group. Rgure 3 illustrates the oocyte-size distribution of a representative undeveloped fish. Undeveloped fish exhibited ovarieswhichcontainedthegreatestproportionofsmaUdiameter oocytes. Fish placed into this group possessed ovaries with 90- 100% of their Gilson-fixed oocytes with diameters less than 100 )im. The oocyte size-frequency distribution had a single mode ofoocyte diameters between 50 and 100 pm,and all eggs were less than 250 pm. Nineteen of the 29 fish were examined histologically and confirmed to beinactive, i.e. not vitellogenic. The mean GSl of fish with undeveloped ovaries was 0.84 (S.E.=:0.05). The mean number of oocytes per gram of ovarian tissue in these fish was 1.39x10^ (S.E.=1. 12x10^) with the mean number of oocytes per female fish of 1 . 1 8x 10* (S.E =1.78x10’). Group n, Early Developing. Forty-one of the 13 1 fish were placed in the early developing group. Figure 3 displays the distribution of oocyte sizes in the ovary of a fish in the early phases of vitellogenesis. Generally, fish in early development had ovaries with 50-90% of the oocytes smaller than 100 pm in diameter. The major mode was between 50 and 100 pm, but most of these fish had a small proportion of oocytes greater than 250 pm. Many fish showed signs of an additional minor mode ofoocyte sizes, particularly in the 250-400 pm nuige. The mean GSI of fish in this group was 2.11 (S.E.=0.18). The mean number of oocytes per gram of ovarian tissue was 4.26x10® (S.E.=1.8xl0*), with a mean number of oocytes per female of 1.10x10* (S,E.=L3xl0’), Month Figure 2. Monthly pattern of vitellogenic and non-vitellogenic female cobia, Rachycentron canadum, ^834 mm FL. Figure represents a composite of years 1989-1990. 149 Lotzet al. Not all fish could be correctly classified by their vitellogenic activity solelyonthebasisof oocyte size-finequency distributions. Seven fish which were ini tiallyclassified as havingundeveloped ovaries on the basis of oocyte frequencies were determined to be undergoing vitellogenesis based upon histological examinaiioa No fish with their largest oocytes less flian 150 pm were found to be developing. However, fish with their largest oocytes between 150 pm and 250 pm were either non- vitellogenic or undergoing vitellogenesis. All fish with at least one oocyte larger than 250 pm possessed oocytes with accumulated vitellin. Someof ihesefishcouldhaverepresented post-spawning fish with residual vitellogenic oocytes. Group I Group II Group III Group IV Oocyte diameter (iim) Figure 3. Oocyte size-frequency distributions of cobia, Rachycentron canadum. Group I: undeveloped ovary, Group II: early developing ovary, Group III: mid-developing ovary. Group IV: late developing ovary. Only oocytes 30 pm or greater were included in the groups. 150 Gonadal Maturation in Gobi a prom Northcentral Gulf of Mexico Group in, Mid-Developing. Figure 3 depicts the oocyte 450-500 pm. Some fish in this group displayed a third mode size-frequency distribution of a fish belonging to the mid- in the 250-400 pm range. The mean GSI of these females was developing group. Fish in this group exhibited oocyte- 4.20 (S.E.=0.29). The mean oocyte density was 2.21x10® diameter distributions with a major mode at 50-100 pm. A oocytes per gram (S£.=5,40xl0^), with a mean number of distinct second mode was recognized at 400-450 pm or oocytes per female of 1.47x10® (S.E.=2.47xlO''). Group Group Figure 4. Comparisons of GSI, oocyte diameter, oocyte density and total number of oocytes among the four stages (groups) of ovarian development in cobia, Rachycentron canadum (means ± 1 standard error of the mean). Groups I-IV are as in Fig. 3. Means which share a letter are not significantly different from one another. 151 Lorz Er al. Group IV, Late Developing. The late developing group of fish possessed the most well-developed ovaries and were considered to be close to spawning (Figure 3). Twenty- three fish were placed in this group. The frequency distribution was distinctly bi-modal or tri-modal, with the most advanced mode in the 500 to 650 pm range. No running-ripe females or fish with hydrated oocytes were observed during our study. The mean GSl of these fish was 5.94 (S.E.=0.38X with a mean oocyte density of 1.40x10^ (S.E.=:2.34xlO^) per gram of ovarian tissue. The mean number of oocytes perfanale was 1 .96x 10* (S.E.= 4.64x 1 0^. Acomparison of Uie four groups shows the relative s'lzeof the gonads, asa proportion of body weight (GSl)of female fish, increased as maturaiionproceeded(Figuie4). The differences among the four groups were statistically significant (p<0.05). Figure 4 further illustrates thiU the mean diameter of oocytes increased as maturation proceeded. The differences were significant betweenundeveloped(GroupI) and eariy developing (Group n)ovaries, as well asbetween early (GroupII)andmid- devcloping (Group IH) ovaries, but tlie differences between mid-developing (Group HI) and late developing (Group IV) ovaries were not statistically different. The density of oocytes per gram of ovarian tissue decreased as maturation proceeded and oocyte size increased (Figure 4). However, as Figure 4 indicates, there were no siatisticalLy significant differences among the groups in total number of oocytes per adult fish. Therefore, asinaturation proceeded, there was litilerecruitment of new oocytes, and the size of the gonads increased to accommodate the increasing size of vitellogenic oocytes. Histology From some initial histological material not used for other analyses, we compared developmental stages in the left and right gonad of nine females and one male and found no significant differences in the stages of an individual. From four of those female fish, representing individuals with different stages of oocyte development, we examined histological sections from the anterior, middle and posterior of both ovaries and found no significant differences in the presence of siagesamong those sites. Each ovary contained previiellogenic and one or more vitellogenic stages, occasionally with a somewhat patchy distribution of oocytes in specific stages. In the case of the mature male, the tubules wer^ especially filled with spermatozoa in the central portion. The walls of the efferent ducts contained more ripening germinal cysts with early developing stages in the periphery of the middle section of the testes than in those at either end. In summary, any section from a gonad provides a good indication of its developmental stage. Figures 5-18 illustrate the developmental stages and features in cobia ovaries, and Figures 19-24 show those in cobia testes. Of the gonads of 94 females and 49 males examined histologically, those of 14 females and 9 males were from fish 860 mm FL or less. Four of the females and two of the males were fish caught in the Gulf off the Florida Keys 18-19 January 1991; two of those females and one male fit into the <860 nun FL category. Some females were ripe May through September. Atretic oocytes occurred from J uly through October in nine fish from southeast Louisiana to northwest Florida, and degenerating hydrated oocytes occurred in three fish in Ju ly and October from the same location. However, a few fish in mid-June through /uly had ovaries in the resting state (containing bolli Group I and atretic oocytes), indicating they would produce eggs again because all females in August and September had high numbers of well-developed eggs. All four fish collected in January from the Florida Keys had atretic oocytes and degenerating hydrated oocyies- Size at Maturity The smallest female exhibiting developing oocytes measured 834 mm FL and was determined to be 2 years old on the basis of otolith evaluation. Histologically, this was the smallCvSt female fish with oocytes with a zona radiata and exhibiting vitellogenesis. Nine females (640 to 860 mm FL) examined histologically did not have developing oocytes. The smallest male exliibiting evidence of spermatogenesis was 640 mm FL. This particular male, 1 year old based on otolith evaluation, was not undergoing spermatogenesis although the tubules were filled with sperm. Actual onset of spermatogenesis may have occurred when this and other fish were smaller than 640 mm FL because no smaller male was examined histologically. Spawning The bi- and tri-modiil oocyte size-frequency distributions observed for Groups III and IV (Figure 3) suggest that oocytes continued to be matured throughout the spawning season; however, the exacinumberof spawns and spawning frequency could not be estimated from these data. The size of a batch spawn was estimated from the group of oocytes around the most advanced mode of oocytes in the 23 fish in the late developing group. The proportion of oocytes which were represen ted by the most advanced batch of oocytes ranged from 1 1 to 60% (mean = 28%; S. £.=3%). The estimated batch fecundity ranged from 2.6x10^ to 1.91x10* oocytes (Mean=4.8xl0^; S.E.=9.8xlO^. Among spawning fish, larger fish produced larger spawns. This is depicted by the significant positive linear relationship between the logarithm of batch fecundity and fork length and between the logarithm of batch fecundity and fish weight as measured for fish belonging to the late developing group (Figs. 25a, 25b). In addition, spawning fish with larger ovaries producelarger spawns. Figures 25c and 25d illustrate ihesignificant linear relation ship between the logarithm of batch fecundity and GSI, and between the logarithm of batch fecundity and total gonad weight of fish in late development. 152 Figures 5>10. Sectioned ovarian tissue from cobia, Rachycentron canadum. Numbers included in all figure legends that precede fish data are slide numbers. 5. Ovigerous lamella of immature fish lined by an ovary wall (tunica albuginea). Note mesovarium (m) and thin outer squamous epithelium (e) bordering collagen layer (c) and thick layer of smooth muscle (s) containing large blood vessel (v) (2703, July, 710 mm FL, 3.4 kg, 18 6 gm ovaries, 1 year old). 6. Lamellae of early ripening ovary showing medium-sized (m) oocyte starting vitellogenesis (small cortical alveoli and peripheral nucleoli extruding tVom nucleus into cytoplasm) and relatively large (1) later staged oocytes among various sized oocytes and small oogonia (2671, April, 1000 mm FL, 12.4 kg,92.0giu ovaries, 2 years old). 7. Close-up of same ovary.Nole lampbrush chromosome$(arrow)innucleusofmoredeveloped oocytes. 8. Ovary in similar stage as that shown in Fig. 7 but with more extensive cortical alveoli (a) (2672, April, 1230 mm FL, 465.0 gm ovaries). 9. Ripe ovary vrtth nuclei beginning to migrate marginally. Note postovulatory follicle (arrow) of released egg (2716, August, 1320 mm FL, 566.6 gm ovaries, 4 years old). 10. Ripe ovary similar to that in Fig. 9 but with few oocytes in early stages. Note ovigerous fold covered by squamous epithelium (arrow) (2839, August, 920 mm Fl^ 330.0 gm ovaries). 153 Figures 11-14. Sectioned ovarian tissue from cobia, Rachycentron canadum. Numbers preceding fish data arc slide numbers. 11. Ripe ovary at beginning of spawning period showing ripe non-hydrated oocytes with irregularly shaped nuclei with extruding nucleoli (arrow) (2670, April, 1225 mm FL,21.5 kg, 772.7 gm ovaries, 3 years old). 12. Ripe ovary in period between two apparent major periods of spawning. Note numerousempty follicles(f) (2700, July, 1231 mm FL,21.8 kg, 5^.6 gm ovaries, 3 years old). 13. Close-up of oocyte with the striated zona pelJucida (zona radiata) separating (arrowhead) from oocyte cytoplasm. Note follicular wall consisting of outer granules containing lipid droplets and inner zona pellucida consisting of darker thin outer layer (arrow) and thick pale Inner layer. The separated peripheral cytoplasm of the oocyte contains darkly staining yolk droplets (y) and clear cortical alveoli (a). External to the granulosa and divided by a conspicuous basement membrane (m) is the theca externa containing capillaries (c) 1 2700, July, 1231 mm FL, 21.8 kg, 584.6 gm ovaries, 3 years old). 14. Atretic oocyte in ovary offish before initial spawning period. Note the degenerated marginal nucleus (2671, April, lOOOnim FL, 12.4 kg, 92.0 gm ovaries, 2 years old). 154 LorrZET AL. Scomberomorusmaculatus, the Spanish mackerel (Finucane and Collins 1986) and Scomberomorus cavalla^ the king mackerel (Finucane el al. 1986). Ricliards (1967) reported the smallest mature female he examined from Chesapeake Bay was 696 mm FL, which is 138 mm shorter than the smallestmature female we observed. This discrepancy may reflect a slower growth rate for cobia in the cooler waters of thfe Chesapeake Bay area, rather than a regional difference in the age at maturation. Based upon scale Discussion Our study suggests, as did those of Thompson el al. (1992) and Biesiot et al. (1994), that the reproductive season for cobia in the northcentral Gulf of Mexico is protracted and extends from April through early October with greatest activity occurring in the spring and early summer. This also parallels the reproductive activity of other Gulf of Mexico coastal pelagic fishes such as Gonadal Maturation in Cobia from Northcentral Gulf of Mexico annuli, Richards ( 1967) sunnised (hat females of 700 nun FL were 2 years old (in their third year of life). In the Gulf of Mexico, it is unlikely thata700mm FL female would be2years old, since 2-year-old females examined in this study averaged 850 mm FL (Franks et al. 1991a). Thus, 700 mm FL mature females collected in Giesapeake Bay by Richards ( 1967) may have been the same age as fish measuring about 850 mm FL inourstudy. Ihesmallestmaiuremalewe found was 640mmFL (age 1). Richards(1967)repartedearliestmaturity inmalesat 518mmFLandage 1. Apparently males canmature when they are 1 yearold, whereas females are notmature until 2 years of age. The results of our study support Richards’ (1967) suggestion that cobia spawn more than once during the spawning season. Richards (1967) reached his conclusion on the basis of finding fish with partially spent ovaries. We reached our conclusion because we observed group- synchronous oocyte maturation in fish collected during the spawning season, characterized by the presence of at least two distinct size groups of oocytes that had undergone vitellogenesis in the same ovary as well as postovulatory follicles (empty follicles) and atretic hydrated eggs in a few ovaries from J uly through October (and in January from the Figures 15-18. Sectioned ovarian (bsue from cobia, Rachycentran canadum. Numbers preceding fish data are slide numbers. 15. Ovary of different fish than in Fig. 12 but during same interspawning period, showing a resting ovary with an abundance of clusters of atretic oocytes (ao) (2695, July, 944 mm FL, 8.7 kg, 171.1 gm ovaries, 2 years old). 16. Early phases of atresia (a) of some oocytes in ovary of post-spawned fish after end of spawning (2723, September, 940 mm FL, 9.1 kg, 60.9 gm ovaries, 2 years old). 17. Degenerating hydrated egg in post-spawned resting ovary, showing fibrous capsule (c) of atretic follicle containing the hydrated oocyte (o) with its membrane (m) and containing an abundance of inflammatory macrophages and fibrocytes (arrow) (2731, October, 1110 mm FL, 141.2 gm ovaries, 3 years old). 18. Resting ovary of fish in winter with atretic follicle contaioiog hydrated egg. Note homogeneous egg membrane (m) (2826, January, 991 mm FL, 10.9 kg, 120.0 gm ovaries, 2 years old). 155 * sijtf* jMGyy : [ -- Vj W ^ ■ ■-■ ■ i€j Gonadal Maturation in Cobia from Northcentral Gulf of Mexico Florida Keys). However, the GSI values we obtained, along with those of Biesiot et al. (1994) md Thompson et al. (1992) during most years, did not support the late summer spawning activity. Our estimates of batch fecundity are considerably larger than Richards’ (1967) estimates of total fecundity. We estimated the size of a batch spawn to be between r = 0.56 p < 0.05 Q _| ^ ^ , 1 800 1000 1200 1400 1600 Fork length (mm) 2.6x10® and 1.91x10* eggs, with anaverageof4.8xl0^ eggs per batch. Richards (1967) estimated total fecundity to be from 2x10® to 5x10® eggs per female, based on the total number of oocytes greater than 500 pm in diameter in the ovaries of six cobia. In our study, we used Gilson’s fluid rather than formalin as in Richards’ study (1967). Since we deteniiined that oocytes shrink 1 1 % more in Gilson’s than p < 0.05 6 ^ ^ ^ ^ 1 0 10 20 30 40 Total weight (kg) Gonoaomatic index Ovary weight (gm) Figure 25. Relationship between batch fecundity and fork length (a), total weight (b), GSI (c) and ovary weight (d) for female cobia, Rachycentron canadum* 157 Lotzet al. in formalin, we would probably have counted the same oocytes as Richards if we only consider eggs greater than 550 pm across. In late developing fish, oocytes greater than 550 pm generally constituted those used for our batch fecundity estimates. Those estimates are based on the advanced model group of developing oocytes and could overestimate fecundity if all eggs in the group are not released. It is not straightforward to estimate the total seasonal fecundity of fish that spawn more than once per season without knowledge of the number of spawns per season. Although nonsynchronous formation of oocytes in the ovaries was observed, presenting strong evidence for multiple spawning (Hunter et al. 1992), we were unable to calculate spawning frequency (Hunter and Goldberg 1980; Hunter and Macewicz 1985; Hunter etal. 1985) because of the lack of recently hydrated oocytes and the relatively small sample sizes of fish from a single location over aone- year period. There appear to be some yearly variations in spawning, probably controlled in part by year-to-year temperature and locality fluctuations. Even though we never observedmore than twoadvancedmodesof developing oocytes over 30 pm in an ovary at one time by measurement, we are unable to cottchide that cobia only spawn twice in a season. This study showed that the total number of oocytes remained nearly constant as fish matured through the four gonadal developmental stages and spawned. In addition, the histological data and the oocyte size-frequency distributions indicated that some fish were close to spawning throughout the protracted spawning season. Therefore, it is assumed that recruitment of primary oocytes throughout the reproductive season and possibly continual transfonnation of primary oocytes into vitellogenic oocytes ocemred throughout most of the reproductive season. Whether cobia spawn during the day or at night is not well understood. Ditty and Shaw (1992) postulated that cobia spawn during the day because all larvae {N-IA) examined from the Gulf, with one exception, were in similar late stages of development when collected during mid-morning. Behavior believed to be daytime spawning of cobia was observed off Panama City, Florida (Shaffer and Nakamura 1989). In the present study , we observed no fish with recently hydrated eggs, even though all examined fish were captured (by hook-and-line) during daylight hours. One explanation for the lack of fish with hydrated eggs in our study is that spawning cobia do not feed, as suggested by Richards (1967), and thus are not subject to capture by baited hook.. Another explanation is that at least some cobia spawn at night. Some cobia may spawn far offshore as suggested by the abundance of eggs found in the Gulf Stream offshore from North Carolina (Hassler and Rainville 1975) and by reported observations of cobia spawning approximately 80 km off the South Carolina coast (Shaffer and Nakamura 1989). Cobia examined in our study were caught near mainland and barrier island beaches, in ship channels, over shallow water wrecks, and at petroleum structures located no further than 40 km offshore. Acknowledgments We thank Leslie Christmas for her technical expertise with the image analysis system, Marie Wright and Lynn Montgomery forhistological prq)arations and examinations, and Jean Jovonovich Alvillar for printing photographs. For other valuable contributions, we thank Nate Jordan, Mike Zuber, David Lee, Mike Buchanan, Mike Alien, Niki Garber, Capl. J. C. Wells, Capt Daryl Simeon, Gulf Coast cobia fishermen, and directors of Gulf Coast fishing tournaments and rodeos. We thank marine enforcement personnel of the Mississippi Deparunent of Wildlife, Fisheries and Parks and the National Marine Fisheries Service for providing us with some specimens. The coastructive comments of two anonymous reviewers were greatly appreciated. Fiinding for this study was provided by the U.S. Dept of the Interior, U.S. Fish and Wildlife Service, Federal Aid in Sportfish Restoration Project F-91, through the Mississippi Dept of Marine Resources. Literature CfTED Biesiot P. M., R. M. Caylor, and J, S. Franks. 1994. Biochemical and histological changes during ovarian development of cobia, RachycerUron canadum^ from the northern Gulf of Mexico. Fish Bull 92:686-696. Dawson, C.E. 1971. Occurrence and description of prejuvenile and early juvenile Gulf of Mexico cobia, Rachycentron canadum. Copeia 1971(1)165-71. Ditty.J.G.andR.F.Shaw. 1992. Larval development, distribution, and ecology of cobia. Rachycentron canadum, (Family: Rachycentridae) in the northern Gulf of Mexico. Rsh Bull 90:668-677. Finucanc, J, H. and L. A. Collins. 1986. Reproduction of Spanish mackerel, Scomberomonis maculatus, from the southeastern United Slates. Northeast Gulf Sci 8:97-106. Finucane, J. H., L. A. Collins, and L. E. Barger. 1978. Ichthyoplankton/mack^l eggs and larvae. Environmental studies of the south Texas outer continental shelf, 1 977 . Final Rep to Bur Land Manage by NMFS (NOAA), Galveston, TX 77550, var. pag. Finucane, J. H..L A. Collins, H. A Brasher, and C. H. Salomon. 1986. Reproductive biology of king mackerel, ScomberomoruscavaUa, from the southeastern United States. Fish Bull 84:841-849. Franks.J.S,, J.T. McBee, and M.T. Allen. 1991a. Estimations of age in cobia, Rachycentron canadiun, from the northern Gulf of Mexico. J Miss Acad Sci 36(1):55. Franks, J. S., M. H. Zuber, and T. D. Mcllwain. 1991b. Trends in seasonal movements of (X)bia, /?ac^>r:£7zr/Tyi carzadum, tagged and ideasedinthenorthemGulfofMexico. JNBss Acad Sci 36(1):55. 158 Gonadal Maturation in Cobia from Northcentral Gulf of Mexico Hassler.W.W.andR.P.Rainville. 1975. Techniques for hatching and rearing cobia, Rachycentron canadtim, though larval and juvenile stages. l^bL UNC-SG-75-30, Univ, N.C. Sea Grant CoU Prog, Raleigh, NC 27650-8605, 26 p. Howsc, H. D.. R. M. Overstreet, W. E. Hawkins, and J. S. Franks. 1992. Ubiquitous perivenous smooth muscle cxn^ds in viscera of the teleost Rachycefitron canadum, with special emphasis on liver. J Morphol 212:175-189. Hunter, J. R. and S. R. Goldberg. 1980. Spawning incidence and batch fecundity in northern anchovy, Engraulis mordax. Fish Bull 77:641 -652. Hunter, J.R., N.C. H.Lo.andR.J.H.Lcong. 1985. Batch fecundity in multiple spawning fishes. In Lasker, R. L. (ed.). An egg production method for estimating spawning biomass of pelagic fish: Applicatiofi to the norlhcni anchovy, Engraulis mordax^ p. 79-94. NOAATcchRcp,NMFS36. Hunter, LR. and B.J.Macewicz. 1985. Measurement of spawning frequency in multiple spawning fishes. In Lasker, R. L. (ed.), An egg production method for estimating spawning biomass of pelagic fish: Application to the northern anchovy, Engraulis mordax, p. 79-94. NOAA Tech Rep, NMFS 36. Hunter, J, R., B. J. Maccwicz, N. C. Lo, and C. A. KimbrcU. 1992. Fecundity, spawning, and maturity of female Dover sole, Microstomus pac^icus , with an evaluation of assumptions and precision. Fish Bull 90:101-128. Joseph, E. B., J. J. Norcross, and W. H. Massman. 1964. Spawning of the cobia, Rachycentron canadwn, in the Chesapeake Bay area, with observations of juvenile specimens. Chesapeake Sci 5(l-2):67-71. Migd^ski, E- C, and G. S. Fichtcr. 1983. The fresh and salt water fishes of the world. Crown Publishers, Inc., New York, p 225-226. Richards, C. E. 1967. Age, growth and fecundity of the cobia, Rachycentron canadnm, from the Chesapeake Bay and adjacent Mid- Atlantic waters. Trans Am Fish Soc 96:343-350. Shaffer. R.V. and E.L. Nakamura. 1989. Synopsis of biological data on the cobia, Rachycentron canadum, (Pisces: Rachycentridae). NOAA Tech Rep, NMFS 82 [FAO Fish Synop 153], 21 p. Thompson, B. A.. C. A. Wilson, J. H, Render, M. Beasley, andC. Cauthron. 1992- Age, growth, and reproductive biology of greater amberjack and cobia from Louisiana waters. Final Rep, MARFIN Coop. Agreement NA90AA-H-MF722 to NMFS (NOAA). Coastal Fisheries Institute, LSU Center for Coastal. Energy, and Environmental Rcsoures, Baton Rouge, LA, 77 p. Wilkinson. L. 1990 Systat: The system for statistics. Systat, Inc., Evanston, IL. 159 Gulf Research Reports Volume 9 | Issue 3 January 1996 Food of Cobia^ Rachycentron canadum, from the Northcentral Gulf of Mexico Gabriele H. Meyer University of Southern Mississippi James S. Franks Gulf Coast Research Laboratory, jim.franks(^usm.edu DOI: 10.18785/grr.0903.02 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Meyer, G. H. and J. S. Franks. 1996. Food of Cobia, Rachycentron canadum, from the Northcentral Gulf of Mexico. Gulf Research Reports 9 (3): 161-167. Retrieved from http://aquila.usm.edu/gcr /vol9/iss3/2 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Rescaich Reports, Vol. 9. No. 3, 161-167, 1996 Manuscript received August 23. 1995; accepted September 18, 1S>95 FOOD OF COBIA, RACHYCENTRON CANADUM, FROM THE NORTHCENTR AL GULF OF MEXICO Gabriele H. Meyer* and James S. Franks^ ^University of Southern Mississippi. Department of Biological Sciences, Box 5018, Hattiesburg, MS 39406-5018, USA ^Guif Coa^ Research Laboratory, P.O. Box 7000, Ocean brings, Mississippi 39566-7000, USA AB^RACT The stomach contents of 403 cobLa, Rachycentron canadum, caught in the northcentral Gulf of Mexico recreational fishery from April through October of 1987-1990 were examined. Cobia ranged from 373-1,530 mm in fork length. Of the 403 stomachs, 287 (71.2%) contained at least one identifiable prey taxon. Crustaceans, consisting {uimarily of poTtunid crabs, were the predominant food. Crustaceans occurred in 79.1% of the stomachs and comprised 77.6% of the total number of identifiable prey. The second most important prey category was fish which was dominated by hardhead catfish, Arias felis, and eels. Fish occurred in 58.5% of the stomachs but only accou nted for 20.3% of the total number of prey. The importance of fish as prey increased with increasing size (length) of cobia, with the largest size class of cobia ( 1 ,150- 1 ,530 mm FL) showing the highest percent frequency occurrence of fish prey (84.4%). There were no significant differences between the diets of male and female cobia. Species composition of the diet indicated that cobia examined in this study were generalist carnivores in their feeding habits and fed primarily on benthic/epibenthic crustaceans and fishes. However, the occurrence of pelagic prey provided evidence of diversity in the foraging behavior of cobia. Feeding in cobia indicated their dependence upon prey availability rather than upon a few specific food organisms. Introduction Rachycentron canadum, commonly known as cobia or ling, is a widely distributed, pelagic fish which occurs worldwide in tropical, subtropical, and warm temperate seas, except in the central and eastern Pacific Ocean (Shaffer and Nakamura 1 989), In the western Atlantic , the cobia occurs from Massachusetts to Argentina (Briggs 1958), but is mostcommonin !he Gulf of Mexico (Migdalski and Fichter 1983), where it supports an important recreational fishery. In the Gulf of Mexico (Gulf) cobia range from Key West,Florida along thecoastto Campeche, Mexico (Dawson 1971). Cobia typically migrate during spring and summer from their wintering grounds off southern Florida to spawning/feeding grounds in the northern Gulf and return to their wintering grounds in late fall and early winter (Biesiot et aL 1994, Franks et al, 1991). The diet of R. canadum from the Gulf of Mexico, particularly the northern Gulf, is poorly known. Most of the previous research on the feeding habits of cobia was limited to simple descriptions of prey items found in a few stomachs. Miles (1949) reported the stomach contents of 11 cobia from Aransas Bay, Texas, and Knapp (1949, 195 1) noted the prey found in 24 cobia taken from the same area. Reid (1954), Boschung (1957), and Christmas etal, (1974) commented on feeding in a small number of cobia from Cedar Key, Florida (one fish) , coastal Alabama (four fish) and offshore Mississippi (eleven fish), respectively. These researchers found that crustaceans and fish made up the diet of R. canadum, although their conclusions varied on the relative importance of each prey type. Knowledge of the food habits of cobia is necessary for understanding the role of diet in their growth and survival and for comprehending the dynamics of the fishery. The purpose of this study was to describe the diet of cobia from the northcentral Gulf of Mexico. Materials and Methods Cobiaexamined in this study were caughtby hook-and-line in thenorthcentralGulf recreational fishery from April ihrou^ Octoberof 1987-1990. CobiaweretakenoffsoulheastLouisiana, Mississippi, Alabama, and northwest Florida between lat. 30^.0'-29’^.(yN and long. 86^-89^0.0^, The majorityofspecimensweretakenoffcoastalMississii^i. Some fish were provided by state and federal fisheries agencies. Fish were well-iced from the time of capture until Stomachs were removed at fishing docks or coastal fishing tournaments. Fork length (FL) was measured in mm and the sex was recorded. Most stomachs were placed in sealable plastic bags and stored in an ice slurry for short- term storage, usually 4-6 h. Stomachs were then either frozen or placed in 10% buffered formalin for later examination. Occasioiudly , when time permitted, stomachs were removed from fish, opened, and processed in the field. 161 Meyer AND Franks Stomachs were thawed or removed &om fornialin, opened, and scored as either containing food or empty. Stomach contents were gently rinsed with fresh water into a 0.5 mm mesh sieve. I^y items were separated, identified to the lowest possible taxon, and counted. Accurate identification and counts could be made in most cases since foods were generally swallowed whole. Some prey items were in advanced stages of digestion and could not be identified to species; however, those prey were often identifiable to the family or order level. Analyses All analyses were based on stomachs containing at least one identifiable taxon. Prey loo far digested for identification were not used in any computations. Additionally, some items found in stomachs were excluded because they wereprobaWy ingestedincidentaUy. Examples of these were tubes of Chaetopterus wonns, fragments of bivalve and gastropod shells, Sarga&sum weed, and pieces of coral, wood, and leather. Parasitic nematodes and acanthocephalons which occurred in some of the stomachs were also not considered in the diet analyses. Numeric abundance, frequency of occurrence and percent frequency of occurrence (%F) were tabulated forall identifiable prey. In addition, major prey categories (crustaceans, fish, and cephalopods) were analyzed for percent numeric abundance (%N) and percent frequency of occurrence. Three different fork length size classes of cobia, small (373-945 mm), medium (950-1,145 mm), and large (1,150-1,530 mm), were selected based on natural breaks within the size frequency distribution, and the percent firequency of occurrence of major prey within each was CfU3taco«na Fish Cephatopods Figure 1, Percent numeric abundance (%N) and percent frequency of occurrence (%F) of major prey categories of Rachyceniran canadumtrom the north central Gulf of Mexico. compared. A contingency table analysis and post-hoc test (Freeman-Tukey transformation) forproportionaldata were used to determine significant differences («=fi.05) between classes for each major prey category (Zar 1984). Major prey of male and female cobia were also compared. Since males tended to be smaller than females, only cobia within the size range 590-1,045 mm FL were selected. TTiis range contained most of the males sampled and reduced the confounding effect of size. Tests for significant differences («=0.05) were made using a Fisher exact test corrected for continuity. Results The stomach contents of 403 R. canadum, ranging from 373- 1 ,530mm FL, were examined. Of these stomachs, 287 (7 1 .2%) contained at least one identifiable prey taxon. Prey consisted of crustaceans, fishes, and cephalopods (Table 1). Another 35 (8.7%) stomachs contained only badly decomposed,unidentifiableremains. The remaining 81 stomachs (20.1%) were empty. Invertebrates Crustaceans were the iMiraary food of cobia and, essentially, dominated the diet. Crustaceans occurred in 79.1% of the stomachs and ranked first (77.6%) in numeric importance among prey (Figure 1). Crustaceans were represented by eight families of decapods and two families of stomatopods (Table 1). Portunid crabs were not only the predominant taxa among invertebrates consumed (Table 1) but also represented 60.7%N of total food items in the diet and occurred in 72.8% of the stomachs. The lesser blue crab, Callinectes similii, was the most abundant prey species found in the diet, comprising 36.5 %N and occurring in 48.8% of the stomachs. The iridescent swimming crab, Portunus gibbesih (12.5%N, 26.5%F) and the ladycrab, Ovalipes floridanus, (9.0%N, 23.3%F) were the next most important foods in the diet. Following the portunids in importance were the sicyoniids and penaeids (combmed=9.6%N). Other decqxxls, ).e., calllanassids. calappids, majids, pagurids andxanthids,occunedinfiequeittly (Table 1). Stomatopods, predominantly SquUlidae, comprised 6.9%N of the diet. Cephalopods comprised the other primary invertebrate prey group and were represented by two families, LoUginidae, the predominant group, and Octopodidae. Cephalopods were found in 13.2% of the stomachs but only made up 2.2%N of prey consumed (Figure 1). 162 Food of Cobia from Northcentral Gulf of Mexico TABLE 1 Prey items occurring in stomachs of cobia, Rachycentron canadum^ ffom the northcentral Gulf of Mexico, 1987-90. Percent frequency of occurrence based on N=2S7, Prey Total numb^ of individual prey items Frequency of occurrence Percent frequency of occurrence INVERTEBRATES Crustaceans Decapoda Pcnaeidae Penaeus aztecus 3 1 0.3 Penaeus setiferus 1 1 0.3 Penaeus sp- 34 8 2.8 Trachypenaeus sp. 37 9 3.1 Sicyoniidae Sicyonia brevirostris 62 15 5.2 Sicyonia sp. 102 18 6.3 Callianassidae Callichirus islagrande 1 1 0.3 Paguridae sp. 2 2 0.7 Calappidae Calappa flammea 2 1 0.3 Hepatus epheliticus 2 2 0.7 Majidae Libinia emarginata 1 1 0.3 Portunidac Arenaeus cribrarius 16 8 2.8 Callinectes sapidus 5 5 1.7 Callinectes similis 909 140 48.8 Ovalipes floridanus 224 67 23.3 Portunus gibbesii 312 76 26.5 Portunus sayi 1 1 0.3 Portunus spinicarpus 16 3 1.0 Portunus spinimanus 30 17 5.9 Xanthidae Menippe adina 1 1 0.3 Stomatopoda Lysiosquillidae Lysiosquilla scabricauda 2 2 0.7 Squillidae Squilla chydaea 2 2 0.7 Squilla empusa 78 21 7.3 Squitla neglecta 1 1 0.3 Squilla sp. 88 40 13.9 Cephalopods Loliginidae Lotigo pealei 1 1 0.3 Unid. loUginids 47 33 11.5 Octopodidae Octopus sp. 6 4 1.4 FISH Squatinidae Squatina dumeril 1 1 0.3 Dasyaddae Dasyatis sp. 7 7 2.4 Torpedinidae t^arcine brasiliensis 4 3 1.0 Anguillifonnes 133 52 18.1 163 Meyer AND Franks Prey Total number of mdividual prey items Fi^equency of occurrence Percent frequency of occurrence Clupeidae Brevoortia patronus 19 3 1.0 Brevoortia sp. 2 2 0.7 Unid. clupeids EngrauUdae 4 4 1.4 Anchoa sp. 2 1 0.3 Unid. engraulid 1 1 0.3 Axiidae Arias felis 138 70 24.4 Ophidiidae 5 4 1.4 Ogcocephalidae Halieuiichthys aculeatus 1 1 0.3 Syngnatliidae 2 2 0.7 Triglidae Prionotus sp. 48 7 2.4 Serranidae Diplectrum bivitatum 33 2 0.7 Unid. serranids 2 1 0.3 Carangidae Decapterus punctatus 26 18 6.3 Seriola dumerili 1 1 0.3 Unid. carangid 1 1 0.3 Lutjanidae Latjanus campechanus 3 3 1.0 Sparidae Lagodon rhomboides 10 10 3.5 Unid. sparid 1 1 0.3 Sciaenidae Menticirrhus sp. 3 3 1.0 Micropogonias undulatus 9 3 1.0 Cynoscion sp. 1 1 0.3 Lehstomus xamhurus 1 1 0.3 Mugilidae Mugil sp. 5 3 1.0 Uranoscopidae Astroscopus y-graecum 5 5 1.7 Trichiuridae Trichiurus lepturus 3 1 0.3 Stromateidae Peprilus burti 1 1 0.3 Peprilus sp. 3 1 0.3 Bothidae Citharichthys sp. 12 3 1.0 Etropus crossotus 1 1 0.3 Etropus sp. 2 1 0.3 Soleidae Symphurus plagiusa 1 1 0.3 Symphurus sp. 1 1 0.3 Balistidae Batistes capriscus 1 1 0.3 Unid. balistids 4 3 1.0 Tetiaodontidae Chilomycterus schoepfi 2 2 0.7 Unid. tctraodontids Number of stomachs examined 6 Total 2,491 3 1.0 403 Number (and %) of stomachs containing identifiable prey 287 (71.2) Number (and %) of stomachs containing Number (and %) of empty stomachs only decomposed, unidentifiable remains 35 (8.7) 81 (20.1) 164 Food of Cobia from Northcentral Gulf of Mexico Fish Although contributing substantiaUy to the diversity of the diet, fish were not as important as crustaceans. Fish occurred in 58.5% of the stomachs and accounted for 20.3%N of all prey consumed (Figure 1). A wide variety of fishes was consumed, including twenty families of bony fishes and three families of cartilaginous fishes (Table 1). The hardhead catfish. Arius felis, and eels (Order Anguilliformes) were by far the predominant fishes in the diet. Arius fetis^ found in 24.4% of stomachs, exhibited the highest numeric percentage (273%) among fish and contributed 5.5%N to the total diet Eels occurred in 18,1% of stomachs, comprised 26.3 %N of fish in the diet, and accounted for 5.3%N of total items in the diet. Fish less frequently encountered in the diet included round scad, £>ecupre/‘U5puncraruj(Carangidae)andpmfish, Lagodon rhomboides (Sparidae). Other identified fish occurred only rarely (Table 1). Comparison of diet among size classes of cobia Crustaceans dominated the diet of the small (77.2%F) and medium (84.8%F) size classes of cobia, and made up a primary portion (65.6%F) of the large size class (Figure 2). Despite these high frequencies, contingency table analysis (x^ 10.25, df=2, p<0.05 ) and the corresponding post-hoc tests indicated all three size classes were significantly different from each other. Portunid crabs, particularly CalUnectes siwi/is, were the mostimportantprey consumed in all size classes of cobia (Table 2). TABLE! Percent frequency of occurrence of major taxa in the stomachs of three size classes of Rachycentron canadum from the northcentral Gulf of Mexico. Fork length (mm) 373-945 950-1145 1150-1530 N==57 N=164 N=64 Crustaceans (Portunid crabs) (63.2) (80.5) (64.1) CalUnectes similis 35.1 53.0 51.6 Portunus gibbesii 17.5 31.1 23.4 Ovalipes floridanus 19,3 28.0 15.6 Stomatopods 24.6 19.5 25.0 Fish Anguilliformes 14.0 19.5 18.8 Ariusfelis 7.0 22.0 43.8 Cephalopods Loliginidae 17.5 9.1 14.1 In contrast, the importance of fish as prey increased with increasing size of cobia. the largest size class showing the highest percent frequency of occurrence (84.4%) (Figure 2). The increase infishoccuirence was attributable to the hard-head catfish, Arius felis, which increased from 7.0%F in the small size class to 43.8%F in the large cobia (Table 2). Again, con^gency table analysis (x^27.77, df=2, /kO. 001) and post-hoc tests indicated that all size classes were significantly different from each other. The percentage of cephalopods (predominantly squid) remained consistently low across the three size classes (Figure 2, Table 2). No significant differences were found. Comparison of the diets of male and female cobia Ihe diet of male and female cobia within the size range of 59()-l,045inmFLappearedtobesimilar(Table3). Crustaceans were the dominant prey in both sexes. Although females showed a higher percent frequency of occurrence (86.8%) of crustaceansihandidmales(79.2%),thesedifferenceswerenot significant, Portunid crabs were the major component of crustaceans ingested by both sexes. Fish occurred with greater frequency in the diet of males (60.4%F) than in the diet of females (46.2%F), partially due to a greater occurrence of eels in the male diet (Table 3). Males, however, fed less frequently on catfish. As with thecrustaceanprey , no significant differences were found between the diets of male and female cobia with respect to fish or cephalopod prey. TABLES IN^rcent ft^uency of occurrence of m^jor taxa from the stomachs of male and frmaleRorhycenl^on canadum from the northcentral Gulf of Mexico. Size range from 590-1045 mm FL. Male Female iV=48 N=106 Crustaceans (All Crustaceans) (79.2) (86.8) (Portunid crabs) (70.8) (80.2) CalUnectes similis 33.3 52.8 Portunus gibbesii 20.8 32.1 Ovalipes floridanus 16.7 30.2 Stomatopods 14.6 19.8 Fish (All Fish) (60.4) (46.2) Anguilliformes 27.1 17.9 Ariusfelis 6.3 17.0 Cephalopods Loliginidae 10.4 14.2 165 Meyer AND Franks Discussion We found crustaceans, primarily portunid crabs, to be the dominant foods of cobia both in terms of numeric abundance and percent frequency of occurrence. Fishes were second in order of importance. These results vary somewhat from the findings of other researchers. Miles (1949) reported crabs, shrimps, and fishes in near equal numbers in the stomachs of cobia taken from Aransas Bay , Texas, and, similarly, diristmas et al, (1974) found the numbers of fishes and crustaceans to be approximately the same in their samples from northern Gulf waters off Mississippi. In sharp contrast, Knapp (1951) observed a predominanceof fishes (83.3 %F), followed by stomatopods (58%F), penaeid shrimps (46%F) and crabs (42%F) in the diet of cobia caught near Aransas Bay. Texas. The conclusions reached in previous studies were based on examinations of a limited number (24 or less) of stomachs. Although cobia examined in our study were collected by hook-and-line and, therefore, did not represent a random sample, we believe our findings represent amorc definitive description of the diet of cobia in the northern Gulf of Mexico, due, in part, to our high sample number (7V=287) and extensive geographical range. Although crustaceans were the dominant food, our results also indicated that larger cobia, males and females alike, consumed fish with significantly greater frequency than did smaller cobia. This may reflect an ontogenetic shift toward fish as prey in larger cobia. Our results, however, showed no significant differences in the diet of male and female cobia within a range of comparable sizes which may be attributable to the relatively low sample size of males. Although not statistically different, we did encounter fish more frequently in the stomachs of males than females which also may be an indication of an ontogenetic shift toward fish prey since most of the large males, and not the large females, were included in the male>female comparative analysis. The species composition of the diet revealed that cobia fed primarily on or near the seafloor. Theportunids, sicyoniids, penaeids, and stomatopods, though capable of swimming, are primarily benthic or cpibenthic inhabitants. Octopi, as well as many of the fish prey (e.g., bothids, uranoscopids, airids, triglids, dasyatids, eels), also reside on or near the bottom. However, other prey such as carangids, clupeids, and squid are pelagic organisms, and their presence in the diet indicated flexibility in the foraging behavior of cobia. In summary, we found that the primary foods of cobia from the northcentral Gulf of Mexico were benthic or Fork ler\gth (mm) Figure 2. Percent frequency of occurrence of major prey categories for three size classes of Rachycentron canadum from the northcentral Gulf of Mexico. epibenthic crustaceans and fishes, although some feeding did occur in the water column and nearsurface. Additionally, our results indicate that the cobia is an opportunistic carnivore and that feeding appears to depend more on prey availability rather than upon a few specific food organisms. Acknowledgments We thank Tom Mcllwain for his advocacy of this work. We would also like to thank coastal sportfishermen and the directors of coastal fishing tournaments in Mississippi and Florida for their spirit of cooperation in helping us obtain samples. We are grateful to Richard HeardandBruceComynsfor their helpinidentifying the invertebrates and the fishes, respectively. The field assistance provided by Carol Cleveland, Mike Buchanan and T. J. Becker was greatly appreciated. We give our thanks to Terry McBee for his help in the field and for his valuable assistance with data analysis. Our gratitude is extended to Barbara VLskup and David Abom for reviewing an earlier version of the manuscript. JefleryLotz also provided helpful suggestions which improved the manuscript. The constructive comments of two anonymous reviewers were also appreciated. We thank marine enforcement personnel of the Mississippi Dept, of Wihllife, Fisheries and Parks for providing us with some specimens of cobia. Funding for this study was provided by the U.S, Dept, of the Interior, U.S. Fish and Wildlife Service, Federal Aid in Sportfish Restoration Project F-91, through the Mississippi Dept, of Marine Resources. 166 Food of Cobia from Northcentral Gulf of Mexico Literature Cited Biesiot,P.M.,R.M.Caylor, and J.S. Franks. 1994. Biochemical and histological changes during ovarian development of cobia, Rachycentron canadum, from the northern Gulf of Mexico. Fish BuU 92:686*696. Boschung, H. T., Jr. 1957. The fishes ofMobUe Bay and the Gulf coast of Alabama. Ph.D. Dissertation, Univ. Alabama, Tuscaloosa, AL, 626 p. Briggs, J.C. 1958. A listofFloridafrshes and their distribution. Bull Fla State Mus Biol Sci 2:221-318. Christmas, J. Y., A. Perry, and R. S. Waller. 1974. Investigations of coastal pelagic fishes. Comp Rep Bur Comm Fish ProjNo.2-128-R. 105 p. Dawson, C. E. 197 1 . Occurrence and description of prejuvenile and early juvenile Gulf of Mexico cobia, Rachycentron canadurn, Copeia 1971:65-71. Franks, J.S.,M.H.Zubear,andT.D.McIIwain. 1991. Ttetids in seasonal movements of cobia, tagged and released in the northern Gulf of Mexico. J Miss Acad Sci 36(1);55. Knapp, F. T. 1949. Menhaden utilization in relation to the conservation of food and game fishes of the Texas Gulf Coast. Trans Am Fish Soc 79:137-144. . 1951. Food habits of the sergeantfrsh, /{ocAyc^nTro/i canadus. Copeia 1951:101-102. Migdalski,EC.,andG.S.Hchter. 1983. Thefreshandsahwaterfishes oflheworid. Crown Publisheis, Inc., New Ycdc,pp. 225-226. Miles, D.W. 1949. A study of the food habits of the fishes of the Aransas Bay area. M.S. Thesis, Univ, Houston, TX, 70 p, Reid, G. K., Jr. 1954, An ecological study of the Gulf of Mexico fishes, in the vicinity of Cedar Key, Florida. Bull Mar Sci Gulf Caribb 4:1-94. Shaffer, R. V., and E. L. Nakamura. 1989. Synopsis of biological data on cobia Rachycentron canadurn (Pisces: Rachycentridae). NOAA Tech Rep, NMFS 82 (FAO Fish Synop 153], 21 p. Zar,J.H. 1984. Biostadstical analysis. 2nded. Prentice-Hall, Inc., Englewood Cliffs, N.J., 718 p. 167 Gulf Research Reports Volume 9 | Issue 3 January 1996 Seasonal Migration in the Southern Hogchoker^ Trinecks maculatusfasciatus (Achiridae) Tanya L. Peterson University of Southern Mississippi DOI: 10.18785/grr.0903.03 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Peterson, T. L. 1996. Seasonal Migration in the Southern Hogchoker, Trinectes maculatusfasciatus (Achiridae). Gulf Research Reports 9 (3): 169-176. Retrieved from http;//aquila.usm.edu/gcr/vol9/iss3/3 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9 No. 3, 169-176, 1996 Manuscript received January 9, 1995; accepted March 20, 1995 SEASONAL MIGRATION IN THE SOUTHERN HOGCHOKER, TRINECTES MACULATUS FASCIATUS f ACHIRIDAE) Tanya L. Peterson* UniversityofSouthemMississippi,Depar&nentofBiologicalSciences, Hattie^urg,Mississippi39406, USA ABSTRACT Life history patterns often respond to local environmental conditions- The seasonal migration pattern of the northern hogchoker has been described, but the southern subspecies rarely has been studied. To document the migratory movements and the habitat characteristics of the southern hogchoker, long-term survey data and specimens collected during 1993 were examined. Moderate depth (5.8-6.4 m), low water clarity (0-1.2 m), moderate oxygen concentration (4-9 ppm), and sand-mud substrata generally defined hogchoker habitats. Hogchoker habitats only showed seasonal shifts in temperature and salinity characteristics. Hogchokers were only collected in low salinity (0-2 ppt) waters during the winter, but exhibited three abundance peaks in relation to bottom salinity during the summer samples at 0, 5, and 18 ppt. The survey data and the data from the 1993 specimens support the hypothesis that southern hogchokers are following a migration pattern similar to that described for the northern subspecies. Introduction The distribution of maculatus fasciatus, the southern hogchoker, extends south from approximately South Carolina to the Yucatdn peninsula. The range of the northern subspecies (T. m. maculatus) extends from the South Carolina coast north to Massachusetts (Hildebrand and Cable 1938, Gilbert and Kelso 1971). Hogchokers are small estuarine fish with a complicated migration pattern. Newly-hatched individuals begin moving into freshwater areas following summer estuarine spawning and begin migrating into low salinity areas the next spring. This downstream distance is extended progressively each year until maturity, when spawning occurs in the outer areas of the estuary. A return migration into freshwater occurs each fall for the winter period (Dovel et al. 1969). Life history patterns may vary in response to local ecological conditions and the timing of environmental factors can often dictate differences in the evolution of these traits (Steams 1976, Boyce 1979). The migratory movements and many life history factors of the northern hogchoker have been widely studied. However, the southern subspecies has only been the subject of a few studies along the Atlantic coast (Castagna 1955, Smith 1986). The purpose of this study was to determine if the movement pattern documented in the northern subspecies is also present in a Gulf population of southern hogchokers. While individual collections have documented hogchokers in both freshwater streams and estuaries along the Mississippi coast, 1 also examined continuous survey data to clarify seasonal movement patterns. These survey data were also used to describe habitat characteristics of * Current address : University of South Florida, Department of Biology, Tampa, Florida 33620, USA T. m.fasciatus. Finer details of the migration pattern were investigated by examining the reproductive condition and age of hogchokers along the salinity gradient during 1993. Materials and Methods Hogchoker distribution and habitat data were obtained from a fishery survey conducted since 1980 along the salinity gradient from the Back Bay of Biloxi offshore to Horn Island by personnel of the Gulf Coast Research Laboratory (GCRL), CX:ean Springs, Mississippi. The survey samples consisted of standardized 10 minute tows with a 4.9 m flat otter trawl. Deeper trawls, at stations 83 and 84, required 30 minute tows wilha 12.2 m flat otter trawl. Both trawls caisisted of a 19.1 mm stretch mesh body with a 6.4 mm mesh cod end liner. Monthly collections were made atsix sites along the sahnity gradientand hogchokers were commonly caughtatfour of these sites: Bayou Bernard (36), Keesler Marina (34), Biloxi East Channel (37), and Bellefouniain Buoy 8 (32) (Figure 1). Figure 1. Distribution of sampling localities for the GCRL survey in the Back Bay of Biloxi, Mississippi. Peterson Hogchoker abundance, bottom temperature, bottom salinity, bottom oxygen concentration, walerclarity (Secchi disk) and several habitat classifications (water body, bottom morphology, and substratum) were measured at each site. A factor analysis (SPSS^-V2.1) of all environmental variables was performed to specifically describe hogchoker habitats. Although the assumptions were validated, this procedure was abandoned because even if all 1 1 variables were included, only 40% of the variance in hogchoker abundance could be explained. This variability is probably attributable to the complicated and interactive nature of the factors associated with thehogchoker’sseasonalmovements along the salinity gradient. Instead, frequency distributions of the number of hogchokers collected during the survey were examined for each variable. In examining the trends, the presence of 45 individuals was considered to be biologically meaningful. The data were examined in three sets: the entire sample, a summer sample (May through August), and a winter sample (October through March). When no differences occurred between the seasonal data sets, only the entire data set is presented. To address reproductive condition, the specimens collected during the 1993 monthly survey trawls were examined. Two additional stations, 80 and 81, were sampled in the Mississippi Sound during June and July 1993 using 12,2 m trawls. A low salinity site (station 30), near the mouth of Old Fort Bayou, was also sampled in May and September 1993 with 3.2 mm mesh seines. All T. maculatus were fixed in 10% seawater formalin. Each specimen was weighed (0.1 mg), measured (SL and TL, 0.0 1 mm) and then dissected to remove the gonads and otoliths. Gender was determined andreproductivecondition of females was classified following Smith (1986) (Table 1). Gonadosomatic indices (GSI) (Nielson and Johnson 1983) were calculated for each specimen as the wet gonad weight divided by whole wet body weight multiplied by 100. The assumptions of a linear relationship and a 0 y-intercept for a gonad and body weight regression were validated before using the GSI. The average GSI values were compared between stations. The otoliths were embedded in Ciba Geigy® media and sectioned with a Beuhler Isomet® Low Speed Saw into 1 .0-2.0 mm increments. The sections were hand ground to a0.20-0.50nim thickness using 600 and 1500 gritsandpaper and then polished (Beuhler® Alpha Micropolish 11). The otoliths were aged by three independent readers and only usedifatleasttwoofthe three readingsagreed. Symmetrical growth between otoliths and the annular formation of rings were validated in Peterson (in preparation) . Age distributions were compared between stations along the salinity gradient. TABLE 1 Female reproductive classifications based on external morphology of the ovaries, following Smith. Category Color Shape Tissue Immature pale small, equilateral triangle; no posterior elongation undifferentiated; no vascularization Resting light yellow more robust; some posterior elongation (length = 2X height) folicular development; no vascularization Developing deep yellow more elongated posteriorly; becoming turgid and distended appears granular from folicular development; slight vascularization Ripe dark yellow to orange extended to the distal end of the coelomic cavity; very distended appears granular with distinct eggs visible; highly vascularized Spent pale to yellow; often with a reddish hue elongated but extremely flacid; appearing deflated folicular material loose, with atretic eggs present; vascularization disrupted 170 Seasonal Migration in the Southern Hogchoker Results A total of 936 collections were made from 1980-1993. The salinity gradient from Biloxi Bay out to Horn Island was stable, but there was substantial yearly variation in the absolute salinities (Figure 2). Hogchoker abundance was not seasonally associated with changes in water depth, water clarity or oxygen concentrations. Hogchokers occupied depths from 2.4-6.7 m, with most captured in 5. 8-6.4 m of water (Figure 3a). The available habitat range extended from 0.6 to 1 1.6 m. The water clarity of the sample area usually ranged from 0.0-2.4 m, with a few samples reported from 3.1-7 .6 m. Hogchokers were mostly associated with habitats of 0.0- 1 .2 m of visibility (Figure 3b). The Back Bay of Biloxi had oxygen concentrations ranging from 4-12 ppm over the study area. The highest abundance of hogchokers were collected in habitats with 4-9 ppm oxygen (Figure 3c). There were definite seasonal trends, as expected, in the temperature of habitats utilized by hogchokers. Throughout the sample period, temperatures ranged from 1-36“C. During summer months, hogchokers were most abundant in temperatures of 24-32®C; while during the winter peak abundance occurred between 12 and 19®C (Figure 4). Hogchokers also did not exhibit seasonal shifts in the use of habitat types. They were usually in the Bay and Bayou, sometimes collected in the Sound, and only rarely in the Gulf (Figure 5a). They were also commonly collected in natural and dredged channels, but less so in open water without submerged vegetation (Figure 5b). Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Figure 2. Monthly average salinity levels of four Sampling stations in the Back Bay of Biloxi from a 14 -year survey. 95% confidence intervals were lai^e and^ for clarity, are presented in Peterson (1994). Depth (m) 0 2 4 6 8 10 12 14 16 Oxygen Concentration (ppm) Figure 3. Hogchoker abundance distribution in relation to habitat depth, water clarity and bottom oxygen concentration. 171 Peterson In addition to these categories, only three others (sand beaches with submerged vegetation and marsh areas both with and without submerged vegetation) were sampled, a total of 1 1 times, throughout the sample period. Hogchokers were documented in each of these habitat types, but they were most abundant in habitats with mud bottoms (Figure 5c). However, during the study there were no sites classified as sand substrata, probably indicating an actual sand-mud combination along the salinity gradient. Hogchokers were collected during all months , but they were most abundant from April to September (Figure 6a). Abundance also varied among years, with a peak from 1988-1989 (Figure 6b). This peak is followed by a sharp decline in collection abundance , with a subsequent increase in 1993. Hogchokers did exhibit a seasonal shift in terms of habitat salinity. During the winter, hogchokers were most abundant in salinities of 0-2 ppt (Figure 7), with only rare occurrences in higher salinity areas. During the summer, three abundance peaks occurred , atapproximately 0, 5 , and 18 ppt. Hogchokers were documented throughout most of this salinity range during summer collections. Figure 4. Hogchoker abundance distribution in relation to bottom temperature during summer and winter sample periods. ■o (D O 0) o o to L_ (D o .c o D) O X Bay Bayou Sound Gulf Figure 5. Distribution of hogchokersamong categories of water body types, bottom morphology (Coding: 1-open water without submerged vegetation, 2-dredged channel, 3-natura] channel, 4-sand beach without submerged vegetation, 5-marsh with submerged v^etation, 6-marsh without submerged vegetation), and substrata characteristics. Sample n’s are the number of collections with hogchokers in each category. 172 Seasonal Migration in the Southern Hogchoker In 1993 , salinity generally increased along the gradient (Figure 8). Hogchokers were collected during April only at lower salinity stations. During June, most hogchokers were collected at the outer and higher salinity stations. Hogchokers were again only collected in low salinity areas in September. The salinity ranges during January- July 1993 were similar to the survey data means. Immature individuals, males, and resting and developing females were collected in April. In June, only resting females were collected in low salinities, while ripe and developing females werecollectedat the Sound stations. Ripe females were only collected during June at the high salinity stations. Hogchokers, including a spent female, were only collected in July at a low salinity station in the bay. Only immature specimens were collected at the mouth of Old Fort Bayou during September (Figure 9). The gonadosomatic indices were low (< 0.5) for all specimens collected in April and male GSI values never exceeded 0.5 in the 1993 samples. However, female GSI values increased to 5 .0-6. 1 at the outer stations during June, and then decreased in the next month (Figure 10). Figure 6. Total collection abundance of hogchokers during each month and year of the survey. During the summer months in whichhogchokers were collected, a trend of increasing age along the salinity gradient was exhibited (Figure 1 1). Young of the year fish were only collected at the low salinity sites and the oldest specimen, in its fifth summer, was collected at the outermost station. On the spawning grounds, considered to be stations 80 and 8 1 based on the reproductive data, the mean age of hogchokers was approximately 3 to 4 years. Discussion The annual pattern of hogchoker abundance could be indicating population trends. If so, the Biloxi population of southern hogchokers increased in size from 1984 to 1989, followedby asharp population decline in the following years and only in 1993 did an upward trend return. This could be the result of variation in annual recruitment and survival. To my knowledge, no major environmental phenomena occurred which could explain this pattern and annual salinity flucmations do not correspond with the population trends. Further investigations of hogchoker population dynamics are needed to explain this annual variation. Hogchoker habitats can be described as areas of moderate depth, low water clarity, moderate oxygen concentration, and mud-sand substrata. Temperature and Salinity (ppt) Figure 7. Distribution of hogchokers in relation to bottom salinity during summer and winter sample periods. 173 Peterson salinity were the only variables in which hogchokers showed a seasonal shift in resource use. These results correspond with Smith’s (1986) suggestion that temperature initiates the migration while salinity is the directing factor. During winter months, hogchokers were most common in low salinity waters of 12“19®C. During the summer, there were three main hogchoker locations, all of 24-32*^!. One area was at the freshwater interface, another at 5 ppt, and the final locality at 18 ppt. The 1993 collections exhibited abundance trends similar to the survey data. Hogchokers were first collected in April only in low salinity areas, their GSI values were low and none of the specimens were mature. As summer progressed, the older maturing and mature fish began moving out into the estuary. The abundance peak in the survey data at the fireshwater interface were most likely immature individuals, and 2-3 year oldmaturing specimens probably represent the 5 ppt peak (Peterson in prep.). Peters and Boyd (1972) reported a similar summer distribution of juvenile hogchokers in a North Carolina population, with the highest abundance occurring at 25^*0 and 5-10 ppt. The final peak at 18 ppt represents the hogchoker spawning grounds. The only specimens seen in spawning condition, as well as the oldest individuals, were collected in this area. The distribution range then shortens as the fish move back to low salinity and fireshwater habitats. This would explain why the winter survey samples produced only a single abundance peak in low salinity (0-2 ppt) areas. 30 25 20 15 10 5 0 30 25 20 15 10 5 0 30 36 34 37 80 81 30 36 34 37 80 81 Stations Stations Figure 8. Habitat salinity and hogchoker abundance at each station along the salinity gradient during 1993. 174 Seasonal Migration in the Southern Hogchoker Dovel et al. (1969) reported that hogchoker spawning areas are in full seawater (30ppt). The data presented here suggest that southern hogchokers are spawning in lower salinity waters, approximately 15-18 ppt. This may not represent a true difference, but instead may be an artifact of environmental characteristics. The northern Gulf of Mexico does not have the higher salinity inshore areas as 30 36 34 37 80 81 30 36 34 37 80 81 Stations Wm Male I I Developing Female Immature [TTTITTm Ripe Female Resting Female Spent Female Figure 9. Percent occurrence of gender and female reproductive conditions of hogchokers collected at each station along the salinity gradient during 1993. Figure 11. Mean age of hogchokers collected at each station along the salinity gradient during 1993. those reported for the northern Atlantic coast. Moreover, the spawning area reported by Dovel et al. (1969) was determined by egg presence in plankton samples, not the presence of ripe hogchokers. Further studies of the northern subspecies, or hogchoker egg collections in the Gulf of Mexico, are required to verify this possible subspecies difference. Stations Figure 10. Mean gonadosomatic indices of hogchokers collected at each station along the salinity gradient during 1993. 175 Peterson OPPT INCREASING S AUNTY 18 PPT I 1 SPAWNING AREA (SUMMER) WINTER RANGE EXTENSION WITH SALINITY YEARS ALL INDIVIDUALS CAPABLE OF SPAWNING Figure 12. Proposed seasonal migration pattern of southern hogchokers^ modified from Dovel et al. (1969). The survey data and the 1993 movement patterns support the hypothesis that the Back Bay of Biloxi population of southern hogchokers uses a migration pattern similar to that described for the northern subspecies. A migration pattern, modified from Dovel et al. (1969), is presented as that followed by the southern subspecies (Figure 12). The conservation of this character between subspecies and over time agrees with McDowell’s (1993) suggestion that diadromy is a stable process and not a transitional mechanism of evolving to a freshwater lifestyle. Acknowledgments I gratefully acknowledgemy advisor. Dr. S. T. Ross of the University ofSouthmiMississippl,forhissupportandguidance. James Warren and theGQlL fisheries staff provided technical assistance, long-term survey dataand laboratory equipment. I thank W.T. SlackandM.T. O’Connell for reading the otoliths. M. Peterson, C. Peterson, J. Petersai, and E Curtis provided unending support throughout this study. Two anonymous reviews provided many helpful comments on this manuscript. References Cited Boyce, M. S. 1979. Seasonality and patterns of natural selection for life histories. Am Nat 114(4):569-583. Castagna,M. 1955. Astudy ofthehogchoker,rrwiecrej/m3(c«/or«j (Bloch and Schneider) in the Wakulla River, Florida. Master’s Thesis. Fla State Univ. Dovel, W. L., J. A. Mihursky and A. J. McErlean. 1969. Life history aspects of the hogchoker, Trinectes maculatus, in the Patuxent River Estuary, Maryland. Chesapeake Sci 110(2): 104-1 19. Gilbert, C. R. and D. P. Kelso. 197 1 . Fishes of the Tortuguero Area, Caribbean Costa Rica. Bull Fla State Mus Biol Sci 1 6(1 ): 1 -54. Hildebrand, S. F. and L. E. Cable. 1938. Further notes on the developmentandlife histories of someteleosts atBeaufoaLNorth Carolina. US Fish Bull 48(24): 505-642. McDowall,R.M. 1993. A recent marine ancestry for anadromous fishes? Sometimes yes, but mostly no. Environ Biol Fishes 37(4):329-335. Nielson, L. A. and D. L. Johnson. 1983. Fisheries Techniques. Blacksburg, VA. Southern Printing Company. 468 pp. Peters, D. S. and M. T. Boyd. 1972. The effect of temperature, salinity, and availability of food on the feeding and growth of the hogchoker, Trinectes maculatus. J Exp Mar Biol Ecol 7:201-207 Peterson, T. L. 1994. The effects of salinity on the survival, growth, behavior, and metabolism of juvenile hogchokers, Trinectes maculatus fasciatus (Achiridae), a possible explanation of seasonal migration. Master’s Thesis. Univ Southern Miss. Peterson, T. L. In prep. Partial life history of southern hogchokers, Trinectes maculatus fasciatus (Achiridae), in the Back Bay of Biloxi, Mississippi. Smith, S. 1986. Reproductive ecology, population dynamics and seasonal movement of the hogchoker in the Elizabeth River, Virginia. Master’s Thesis. CoU William and Mary, Va Instil Mar Sci, Stearns, S. C. 1976. Life history tactics; a review of the ideas. Q Rev Biol 51(l):3-47. 176 Gulf Research Reports Volume 9 | Issue 3 January 1996 Reproductive Strategies in a Population of Gobiosoma bosci (Osteichthyes: Gobiidae) with Slow and Fast Maturing Individuals Candace H. Conn Lamar University David L. Bechler Lamar University DOI: 10.18785/grr.0903.04 Follow this and additional works at; http://aquila.usm.edu/gcr Part of the Marine Biology Commons Recommended Citation Conn, C. H. and D. L. Bechler. 1996. Reproductive Strategies in a Population of Gobiosoma bosci (Osteichthyes: Gohiidae) with Slow and Fast Maturing Individuals. Gulf Research Reports 9 (3): 177-182. Retrieved from http://aquila.usm.edu/gcr /vol9/iss3/4 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9, No. 3, 177-182, 1996 Manuscript received June 2, 1995; accepted August 14, 1995 REPRODUCTIVE STRATEGIES IN A POPULATION OF GOBIOSOMA BOSCI (OSTEICHTHYES: GOBIIDAE) WITH SLOW AND FAST MATURING INDIVIDUALS Candace H. Conn and David L. Bechler* Lamar University, Department of Biology, P, O. Box 10037, Beaumont, Texas, 77710, USA ABSTRACT The reproductive biology of Gobhsoma bosci collected from November 1986 to October 1987 in the McFaddin Wildlife Refuge in southeast Texas was studied by using morphometric data. Males achieved greater weights per unit length than females, and longevity was about 12 to 13 months. GSI values and mean monthly ovum diameters indicated that the breeding season ran from April to September, with a major activity peak in May and a minor peak in September. Significant differences in male and female standard lengths (SL), ovum diameter, and egg number existed for sexually mature specimens between the first and second peaks of reproductive activity. An egg versus length analysis produced a positive linear relationship. An accessory gonadal structure index (ASGI) was developed and revealed that maximal AGS development corresponded with the male GSI, but did not produce discemable peaks. Two reproductive strategies were followed and depended upon time of hatching and growth rate. Some individuals that hatched early in the breeding season grew rapidly and were capable of egg laying by August or September. Individuals hatched late in the breeding season delayed breeding until the following season. Introduction The distribution of the naked goby, Gobiosoma bosci, extends from Long Island, New York, to the state of Campeche, Mexico (Hoese and Moore 1977). In view of the markedenvironmental differences throughout its range, G. bosci may be expected to vary its reproductive strategies. In a review of life history phenomena. Cole (1954) stated that the age at which reproduction begins is one of the most significant characteristics, and as such will influence the reproductive success of an individual. Therefore, southerly populations of G. bosci have at least two potential reproductive strategies available to them (Steams and Crandall 1984). First, an individual may grow throughout the non-reproduedve season after hatching and then engage in breeding during the subsequent breeding season. Alternatively, young-of-the-year may grow rapidly enough to reproduce before the end of the breeding season, and thereby gain a breeding season not available to fi^ with delayed breeding. As such, individuals within the same populationmay employ one strategy or theother depending upon their time of hatching and environmental conditions. * Corresponding author; current address is Dept, of Biology, Valdosta State University. Valdosta. GA 31698 Gobiosoma bosci, a cavity nester preferring hard substrates (Bechler etal. 1990), has been studiedextensively along the Atlantic Coast. Dahlberg and Conyers (1973), who reviewed much of the literature, postulated that spawning seasons for different populations of G. bosci were variable, and depended on location. They related initiation and termination of spawning towatertemperature and indicated peak spawning activity was in the warmest months. May through August. However,they didnotdiscuss any variations inbreeding strategies. Other reports on G.ho.^ci by Neio(1976), Crabtree andMiddaugh(1982),andFitzsimons and Seok (1989) revealed many facts about the life histoiy and ecdogy of G. bosci. However, none of the above studies examined reproductive strategies of the naked goby in detail along the Gulf Coast. In this study, we concentrated specifically on life history data related to the reproductive biology of G. bosci in a southeast Texas salt marsh. Description of the Study Area The McFaddin National Wildlife Refuge is located in southern Jefferson County, 20 km southwest of Sabine Pass, Texas (GiiffithandBechler 1995), asubtropical to temperate region of the United States. The refuge is a brackish marsh dissected by meandering creeks, man-made cuts, and shallow lakes; thelargest,CIamLake, was theprimaiy siteof this smdy. 177 Conn and Bechler Clam Lake is relatively shallow, 1 .0 to 1 .3 m in depth, with the deepest area 2.8 m. The substrate ^d banks along the edges of the consist of a firm clay. A fine silt substrate lies 1 to5 m from shore. The clay hanks are riddled with holes and tunnels inhabited by Rhithropanopeus harrisi, the mud crab, two species of Vca, the fiddler crabs, and CaUinectes sapidus, the blue crab. The predominant flora surrounding the lake is Phragmites australis, a reed grass, and Spartina altemiflora, a salt marsh grass. Tidal influences in Clam Lake are minimal. Wind-generated wave action undercuts the banks of the lake and results in the gradual erosion of the shoreline and a subsequent exposure of roots of Phragmites and Spartina. These roots are washed free of soil and extend into the water except during drought and low tide. Large clumps of Phragmites roots break off and fall into the water. These remained c lose to the banks of the lake below the water and provided habitat for G. bosci. Methods and Materials Beginning in November 1986, monthly collections of G, bosci were made in Clam Lake fw* a period of 12 months, except forJuly.whenno collection was made. Sixty to 100 fish were collected each month by using a 3.23 mm mesh seine. Specimens were liardened in 10% fornialin for 24 hours, washedin water for 24 hours, and preserved in 55% isopropyl alcohol. A minimum of 28 specimens (14 males, 14 females) were examined from eachmonthly collection. Standard length (SL) was measured to the nearest 0.01 mm with aMitutoyodial caliper. Total body weight was measured to the neaiest0.(XX)l g with a Metller AE 100 analytical balance. Gonads were removed and wet weighed togetherasaunit foruseinobtaining Gonadal Somatic Indices (GSI). Ten accessory gonadal structures (AGS), which are related to sexual maturation of male gobies (Miller 1984, Cole and Robertson, 1988, Lahnsteineretal. 1992), were removed and wet weighed each month. The weight of the AGS was divided by the total wet body weight and multiplied by 100 to produce an Accessory Gonadal Structure Index (AGSI) similar to that of the GSI. hi females, lOovaccntrally locatedonthesuxfaceoftheovaryand randomly selected were measured in 10 animals each month with an ocular micrometer. During the months of April and May, the beginning of the breeding seascai, the diameters of 175 to 200 ova on the surface of the ovaries were measured in ^ fish to determine if more than one size class of ova were present Eggs large enough to be teased from the ovaries woe counted by hand with an Olympus zoom stereo microscope system. These ova were classified as to their develo{Hnental stageby following HeinsandRabito(1986). Salinity and water tOTip^ature were taken each month with a Yellow Springs TSC meter. All statistical analyses employed MINITAB (1986) software. Results General Ecology Salinity during the year in which the fish were collected ranged from 0 to 12 ppt. Water temperature ranged from 7 to 32®C. During the months of highest reproductive activity, April to June, salinity varied from 2 to 6 ppt and mean water temperature was 28’’C. Day length increased from 12.48hourson 1 April 1987 to 14.05 hours on 30 June 1987 (United States Naval Observatory 1965). In September, when there was a smaller peak in reproductive activity, the salinity was 12 ppt and water temperature was 28®C. Day length decreased from 12.81 hours on 1 September 1987 to 1 1.90 hours on 30 September 1987. Specimens for preservation were collected in clam shells, algal mats, grass clumps, and Phragmites australis roots along the edge of the lake. Seining over soft sediments produced few specimens compared to the number collected close to the banks. Life History Length-frequency relationships examined by month indicated that G. bosci life expectancy was not more than 12 to 13 months (Figure 1). This conclusion was based on the fact that during the months of August and September specimens greater than 25 and 29 mm respectively were not collected. However, from November to June individuals exceeding 29 mm (range=30.()-44.4 mm) were common. A broad range of size classes from December to April indicated that G. bosci most likely reproduced throughout the spring and summer. The smallest specimens (minimum=ll.l mm) were collected from August to October and represented offering newly recruited into the population during late spring and summer. The maximum size of specimens progressively increased from October to April when a maximum size of 44.4 nun was reached. From April to June, the beginning of the reproductive season as discussed below, the size of G. bosci decreased from the maximum to 40.0 mm (Figure 1). The maximum size of G. bosci in August was 24.8 mm, in September, 29.7 mm, and in October, 36.6 mm. The increase in maximum size between each pair of consecutive months was4.9 mm and 6.9 mm with an averageof 5.9 mm/month. Thus if a growth rate of approximately 6 mm/month through the summer is assumed, the largest individuals in August would be no more than about four months old. These fish indicate that all gobies collected in this month represented a new generation. 178 Reproductive Strategies in G. bosci Length-weight relationships comparing males and females showed that males not only reached a greater maximum standard length (44.4 mm) than females (34.7 mm), but were heavier per unit body length (Figure 2). A curvilinear relationship between standard length and weight existed for males and females. The regression analysis produced the following equations: Reproductive adults 4-5 months old Months Reproductive adults 12-13 months old No large reproductive adults Figure 1. Length-frequency histogram for the months involved in the breeding season of G. bosci in the McFaddin National Wildlife Refhge, Sabine Pass, TX. Size class increments are based on 2 mm intervals. Data for the month of July are missing. The month of October is given for size comparisons, but is not part of the breeding season. SL(mm) Weighty = 0.281 + 0.0364SL + 0.0018SL2 Weight,^ = 0.291 + 0.0309SL + 0.0008SL2 These two regression models were significanlly different (N^367; R^O.958: DF=3, 363; F=6.1891; P<0.005) and showed that males were heavier than females per unit length. The results of l-tests indicate that the quadratic (t=4.03, P<0.01) and linear (t;=;3.27, P<0,01) components were responsible for the differences between the models. Ova easily teased from the ovaries were classified as mature to ripe (Heins and Rabito 1986). Mean number of mature to ripe ova per female was 466 (N=40, SD= 178.2) witharangeof 116to 1030. These ova made up two distinct size classes in each of six females examined from April and May (Figure 3). Mean diameter for all stages of developing ova in May was 0,45 mm (N=12, SD=0.1153). The September mean ovum diameter was much lower. 0.19 nun, but the variance greater (N=8 , SI>=0.1538). Mean ovum diameter of mature and ripening eggs in April and May was 0.49 mm (range: 0.35 to 0.56 mm) and, in September, 0.42 mm (range: 0.34 to 0.50 mm). A two sample t-test showed a significant difference in the diameter of mature and ripening ova for the two periods (N=23; 1^2.75; DF=21; P=0.012). Not only were ovum diameters smaller during the second breedingpeak, but ovum number per female was also significantly lower (N=29: DF=3, 25: RM).73; F=22.42; P=0.0001). A regression analysis comparing egg number versus SL for females from April and May against females from August and September produced the following models: Bggs^,^-577 + 45ASL EggSp^i = -382 + 34.0SL Figure 2. Length-weight relationships for male and female Figure 3. Ovum size class frequencies for six females collected G. bosci are given. The 367 data points used to compute the in April and May. Size class increments are based on regression lines are not shown. 0.05 mm intervals. 179 Conn and Bechler Figure 4. Linear relationship between SL and egg number for all females collected during the breeding season in the McFaddin National Wildlife Refuge, Sabine Pass, TX. The difference between the two models resulted from significant differences in the coefficients of the slopes (t=2,22. P=0.036), but not in the intercepts (t=0.51, P=0.617). Themeaneggnumberduringthefitstpeakwas505 (SD=137.1) and 252 (SI>=95,2) during the second peaL The total number of maturing to ripe ova in the ovaries of all females produced the following regression model (Figure 4): Fecundity = -416 + 39.5SL The positive correlation between SL and ova number was significant (N=37; DF=1, 35; F=45.836; P<0.001; R^O.567). Mean male GSI values were low from November to January, began to rise in February and March, and peaked in April (Figure 5). GSI values then declined slightly during June, produced a second peak in August, and then declined in September. The April peak had a mean value of 1.266 and the August peak a mean of 1.284. Male GSI values peaked one month prior to the female peak GSI values for both reproductive periods. Female GSI values showed abimodal distribution, but with a more dramatic change than males (Figure 5). Female GSI values remained low November through December, began a gradual increase in January, increased rapidly February through April, and peaked in May at a mean value of 14.26. GSI values declined in August, peaked again in September at a mean value of 4.10 and declined in October. In male G. bosci the AGS was mitten-shaped with a small lobe extending lateral to the main lobe, attached to the body wall near the vent, and extended dorsally between the inner body wall and the intestinal tract. Each AGS was attached to the caudal end of the testis by a small duct. During December the AGS were small and difficult to locate and remove. Inmonths of high reproductive activity, the AGS were large and completely filled the ventral portion of the abdominal cavity. AGSI values (Figure 5) followed a pattern similar to GSI values for males. AGSI values were low November through December and rose slowly from January to March. Unlike GSI values however. AGSI values did not peak in April but continued to climb and peaked in May, remained high through August, and then declined until October. Ovarian maturation in female G. bosci was determined by ovum diameter. Mean ovum diameters (Figure 6) remained low October through February, and then began to increase in March. The increase in ovum diameter in March signaled the onset of ovum maturation. Therefore, the mean ovum diameter for March was used as a criterion against which ovum diameters for fish examined in other months were compared. All females whose mean ovum diameter was larger than 0.1924 mm, the mean ovum diameter for March, were considered to possess maturing ova. Of 32 females examined in August and September, 13 had a mean ovum diameter greater than 0.1924 mm, the minimum mean size of maturing ova. The mean SL of these 13 females was 20.8 mm (range=16.9 to 24.0 mm). Of the 40 fish examined in April and May , 3 1 had ovum diameters exceeding 0. 1924 mm. The mean standard length of these 31 females was 24,1 nun (range=18.9 mm to 29.6 mm). A comparison of SL indicated that sexually mamre females from August and September were significantly shorterthan females from April and May (N=22; DF=20; t;=4,92; P=0.0001). In addition, significantly fewer females possessed maturing ovaries during the second breeding peak (N=32, DF=1, X^24.953, P<0.001). 15 o CO o 0 Month Figure 5. Monthly mean GSI and AGS values for males and females. Data for the month of July are not given. 180 Reproductive Strategies in G. bosci Male sexual maturity was determined on the basis of GSI values (Figure 5). GSI values* which began increasing in March, were used as an indicator of maturing testes (N=16, ^=0,5624). Of23malesexaminedinApril,20had a GSI exceeding 0.5624. The mean SL of these 20 males was 30.6 mm (range: 19.5 to 44.4 mm). GSI also peaked in August Mean SL of males with a GSI value above 0.5624 was 20.5 mm (range: 17.3 to 24,6 mm), A two sample t-test comparing maturing males from each time period indicated that males matured at a shorter SL in the late summer (N=61; DF=56.9; 1=6.10; P=0.0001). Discussion According to conclusions by Dahlberg and Conyers (1973), populations of G. bosci in the northern parts of its range should reproduce late in life after having grown and matured over the winter. Populations at the southern end of the range, such as in south Florida and Mexico, would be expected to mature earlier and possibly at a smaller size during the protracted breeding season. The primary factor controlling the age at first reproduction, and thus the reproductive tactic employed by an individual, would be the length of the breeding season in combination with the time of hatching. The onset of the reproductive season in the Clam Lake population of G. bosci, determined by changes in GSI and AGSl values, and ovum diameters, indicated the following: (1) the reproductive season ex tended from at least late April to September, a period of five to six months; (2) the reproductive season possessed bimodal peaks in spawning activity; (3) the presence of mature and ripening ova from April to September indicated that breeding occurred throughout the summer; and (4) individual females possessed multiple size classes of ova suggesting that they might breed more than once during the season, a fact supported by field experiments (Conn 1989). Individuals involved in the first reproductive peak in May were significantly larger than those in the September peak. The absence of large individualsinthelasthalf of the summer indicated that the oldest individuals were not Living more than 12 to 13 months and were dying out by August. Therefore, at least some individuals involved in the second breeding peak had hatched out during the early part of the breeding season, matured rapidly, and were breeding by the end of the breeding season. Among these individuals, only 40.6% produced eggs capable of being Figure 6, Monthly mean ovum diameters. Data for the monthof July are notgiven. Open portions of barsrepresent standard deviations. ovulated and laid, and indicating that the majority of individuals did not reproduce until the following spring. From this analysis we conclude thatindi vidua! members of the G. bosci population in Clam Lake followed two distinct reproductive strategies based upon their time of hatching. Some individuals that hatched early in the breeding season matured rapidly, reproduced at the end of the breeding season, and laid alow number of eggs. These same individuals then had the opportunity for further growth over the winter with a second opportunity to produce a greater number of eggs in the early part of the next breeding season. Individuals hatched later in the reproductive season were too small to reproduce by the end of the season and delayed breeding until the following year. These conclusions support Steams and Crandall (1984) and Steams and Koella (1986), who found that plasticity in reproductive strategies allows individuals to vary their age and size at first reproduction. Our data also indicate that time of hatching is a critical factor in determining which reproductive strategy individual G. bosci employ in protracted breeding seasons. Acknowledgments Terry Ling and Roy King are thanked for assistance in the field. Jim Kiakowoski and the staff of the McFaddin National Wildlife Refuge are thanked for providing access to the refuge. John Sullivan is thanked for reading an earlier draft of this manuscript. 181 Conn and Bechler Literature Cited Bechler, D.L., C.H. Conn, and T.S. Ling. 1990. A novel nest- trap for ecological and behavioral studies of cavity-nesting fishes. J Fish Biol 36:277-278. Cole, K.S. and D.R. Robertson. 1988. Protogyny in the Caribbean reef goby, Coryphopterus personatus: gonad ontogeny and social influences on sex-change. Bull Mar Sci 42:317-333. Cole, L.C. 1954. The population consequences of life history phenomena. Quart Rev Biol 29:103-137. Conn, C.H. 1989. The reproductive behavior of Gobiosoma hosci^ the naked goby. [Unpublished masters thesis] Elcpartment of Biology, Lamar University, TX. 50 p. Crabtree, R.E. and Middaugh. 1982. Oyster shell size and the selection of spawning sites of Chasmodes bosquianus^ Hypleurochilusgemmatus, Hypsoblenn uis ionthas (Pisces , Blenniidae) and Gobiosoma bosci (Pisces, Gobiidae) in two South Carolina estuaries. Estuaries 5:150-155. Dahlberg, M.D. and J.C. Conyers. 1 973. An ecological study of Gobiosoma bosci and C. ginsburgi (Pisces, Gobiidae) on the Georgia coast. Fish Bull l\:n9~'2Xn, Fitsimons, JJML and K. Seok. 1989. The effect of size, sex, prior tcsidoncy, and past e^xrioioe on temtoial ddtnsc by the naked g!±>y,Gobiosormboxi. Ptoc Louisiana Acad Sd 52:3542. Griffith, S.A and D.L. Bechler. 1995. The distribution and abundance of the bay anchovy, AnchoamUchiUi, in aSoutheast Texas Marsh Lake System. Gulf Res Rep 9; 117-122. Hiens.D.C. andF.G. Rabito, Jr. 1986. Spawning performance in North American minnows: direct evidence of the occuirence of multiple clutches in the genus Notropis. J Fish Biol 28:343-357. Hoesc, H.D, and R.H. Moore. 1977. Fishes of the Gulf of Mexico, Texas A&M University Piess,College Statiem. 327 p. Lahnsteiner, F., M, Sciwald, and R.A. Patzner. 1992. The seminal vesicles of the male grass goby Zosterisessor ophiocephalus (Tclcostei, Gobiidae). Fine structure and histochemistry. Zoomorphology 111:239-248. Miller, P.J- 1984. The Tokology of Gobioid Fishes. Pages 119- 153 in G.W. Potts and R.J. Wootton, eds. Fish reproduction, strategics and tactics. Academic Press Inc., London. MINTTAB, Inc. 1986. Minitab Reference Manual, 5th release. State College, Pa. 236 p. Nero, L.L. 1976. The natural history of the naked goby, Gobiosoma bosci (Pcrciformes; Gobiidae). [Unpublished masters thesis] Department of Biology, Old Dominion University, VA. 86 p. Stearns, S.C. and R.E. cirandall. 1984. Plasticity for age and size at sexual maturity: A life-history response to unavoidable stress. In: Potts, G.W. andR. J. Wootton (eds), Fish Reproduction, Strategies and Tactics, p 13-33. Academic Press, New York. Steams, S.C. and J.C. Koella. 1986. The evolution of phenotypic plasticity m life-history traits: prediction of reaction norms for age and size at maturity. Evolution 40:893-913. 182 Gulf Research Reports Volume 9 | Issue 3 January 1996 Cryopreservation of Sperm of Spotted Seatrout (Cynoscion nebulosus) William R. Wayman Louisiana State University R. Glenn Thomas Louisiana Department of Wildlife and Fisheries Terrence R. Tiersch Louisiana State University DOI: 10.18785/grr.0903.05 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Wayman, W. R., R. Thomas and T. R. Tiersch. 1996. Cryopreservation of Sperm of Spotted Seatrout (Cynoscion nebulosus). Gulf Research Reports 9 (3): 183-188. Retrieved from http://aquila.usm.edu/gcr /vol9/iss3/5 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9, No. 3. 183-188, 1996 Manuscript received April 20, 1995; accepted May 29, 1995 CRYOPRESERVATION OF SPERM OF SPOTTED SEATROUT (CYNOSCION NEBULOSUS) William R. Wayman^ R. Glenn Tbomas^ and Terrence R. Tiersch^* ^ School of Forestry, Wildlfe, and Fisheries, Louisiana Agricultural Experiment Station, Louisiana State University Agricultural Center, Baton Rouge, Louisiana 70803. USA * Lomsuma Department ofWiUUife and Fisheries, LyleS. St.AirumtMarineBiologicdlLdboratory, P. O. Box 37, Grand Isle, Louisiana 70358, USA ABSTRACT Cryoprcscrvation of fish sperm has applications in preserving genetic resources from slocks of endangered fishes, replenishing fisheries, reducing the number of males needed in hatchery situations, and allowing repeated spawning of specific males. As part of a larger study on artificial breeding of sciaenid fishes, we develops procedures for collection, handling, refrigerated storage, and cryopreservalion of spotted seatrout sperm. Hanks' balanced salt solution (HBSS) was used as an extender for collection and storage of sperm . Sperm motility in relation to graded concentrations of HBSS was used to determine the osmolality at which sperm were activated. Based on these findings, HBSS was prepared at 201 mOsm/kg as an extender for sperm storage. To determine if ions present in HBSS were involved in sperm activation, separate activating solutions were prepared by the addition of NaCl, CaCl,, KCl, Na^HPO^, or MgSO^ to aliquots of a stock glucose solution (1 85 mOsm/kg). The chemicals were added at the concentration of each found in 1-x HBSS. Only the glucose solution containing 8 gA NaCl (424 mOsm/kg) produced activation of sperm. We also evaluated four chemicals as cyroprotectantsi methanol, glycerol, dimethyl sulfoxide (DMSO), and n,n-dimethyl acetamide. Two freezing rates were evaluated by placing samples at either of two heights within a nitrogen vapor shipping dewar. The highest post-thaw motilities were in 10% DMSO with an average retention of 60% of initial motility at the lower position in the dewar, and 37% at the upper position. A third freezing rate was produced using a computer-controlled freezer programmed for a rate of -45"C/min, yielding a retention of initial motility of 31%. Our freezing and transport of cryopreserved sperm in shipping dewars demonstrate the utility of this procedure for field applications. Introduction The faniilySciaenldlaecontaiits several species iinpoitant to recreational and commercial fisheries. The red drum {Sciaenops ocellatus\ black drum (Pogonias cromis), and spotted seatrout {Cynosdon nebulosus) all have large commercial fisheries that \vere closed or restticted to prevent overfishing. TTie decline of these fisheries has stimulated interest in developmentofmethods such asartificialspawning and the use of cryopreserved spenn to aid in restoration efforts. Cryc^jreservation of sperm can be used to ixeserve genetic resources from stocks of fisltes that are endangered andtoaidinreplenishingfisheries. Cryopreservedspennean be used to reduce the number of males maintained in the hatchery and allows repeated spawning of specific males when females are in spawning condition. Cryopreservation can be used to preserve gametes of improved stocks, to study hybridization and crossbreeding, and to accelerate genetic research. The first studies of cryopreservation of fish sperm were performed in marine fishes to aid in hybridization of herring stocks that spawned at different times of the year (Blaxter 1953). Most subsequent studies, however, have been of freshwater species, especially salmonids (see reviews by Scott and Baynes 1980; Stoss 1983). Previous work in reproductive biology of sciaenids has addressed natural spawning(SaucierandBaltz 1993), induced spawning (Colura 1974;ThoinasandBoyd 1988), and hybridi 2 aiion(Henderson- Arzapalo and Colura 1984; Henderson-Arzapalo etal. 1994). Cryopreservation of sperm from the Atlantic croaker (Micropogonias undulatus) was studied by Gwo et al. (1991) and represents the only report on cryopreservation of sp^m from a sciaenid species. As partofalarger study on artificial breeding of sciaenids, we developedprocedures forcoliection,handling,refiigerated storage, and cryopreservation of spotted seatrout sperm. Our objectives were to: 1) determine the relationship of osmotic pressure and sperm activation to allow safe storage; 2) evaluate the effect on sperm activation of specific ions used in the extender solution; 3) evaluate the effectiveness of different cryoprotectants; and 4) evaluate the success of different freezing rates. Motility estimates were used as a measure of sperm viability. To our knowledge, this is the first report of the cryopreservation of sperm of spotted seatrout ♦Corresponding author 183 Wayman et al. Materials and Methods Blood Plasma Osmolality and Extender Preparation Blood samples were collected from 34 spotted seatroui caught from April-August, 1994 in Barataria Bay, LA (29‘'19’ N, 89^56’ W). Water in the Bay ranged in osmolality from 450-750 mOsm/kg during the collection period. The blood samples were allowed to clot, and 10 pi of plasma were used to determine osmolality with a vapor pressure osmometer (model 5500, Wescor Inc., Logan, UT). The osmolality of blood was 356.0 ± 18.4 mOsm/kg (mean ± SD). This value is similar to plasma values (350 mOsm/L) obtained for red drum (Oocker et al. 1983), another member of the family Sciaenidae. Sperm of marine species typically become motile in solutions of higher osmotic pressure than the blood plasma. Therefore. Hanks’ balanced salt solution (HBSS) was prepared (Tiersch et aL 1994) using reagent grade chemicals (Sigma Chemical Coip., St Louis, MO) at an osmotic pressure (3(X) mOsm/kg) below that of the blood plasma to ensure that sperm remained inactive when placed in the extender for storage. Estimation of Sperm Motility Thepercentmodlity of each sperm sample was estimated using darkfield microscopy at lOOx immediately after addition of the activating solution. Percent motility was defined as the percentage of progressively motile sperm within each activated sample. The osmolality of the activated sperm mixture was determined by removing 1 0 pi of diluted sample directly from the microscope slide for analysis by a vapor pressure osmometer. The threshold activation point was defined as the osmotic pressure at which 10% of the sperm became motile. The complete activation point was defined as the lowest osmotic pressure that elicited the highest percentage of motile sperm. Osmotic Analysis of Sperm Activation Sperm samples were collected by manual stripping of three males caughiin April 1994 inBarataria Bay, Louisiana. For this, fish were dried with a towel, and gentle pressure was applied to the abdomen. Spenn were collected in 75-pl hematocrit tubes, transferred to 1 . 8 -ml centrifuge lubes, and diluted with I ml of HBSS (300 mOsm/kg). Sperm activation was evaluated according to Bales et al. (1996) by dilution of 2 -pl aliquots of sperm with 20 pi of solutions ranging in osmotic pressure from deionized water (8 mOsiii/kg) to double-strength HBSS prepared at 600 mOsm/kg. Because HBSS is highly ionic, solutions of mannitol (Sigma) were prepared at three concentrations (200, 300, and 350 mOsm/kg) and used to test activation of sperm in solutions deficient in ions. Ionic Analysis of Sperm Activation Because there was a persistent low level of spontaneous motility (1%) of sperm placed in the HBSS extender solutions, we tested storage of sperm in HBSS al osmotic pressures as low as 152 mOsm/kg. To determine if ions present in HBSS were involved in sperm activation, separate activating solutions were prepared by addition of NaCl, CaClj, KCl, Na^HPO^, or MgSO^ to aliquots of a stock glucose solution (185 mOsm/kg). The chemicals were added at the concentration of each found in Lx HBSS. Glucose and .sucrose .solutions prepared at higher osmolalities were used as control treatments to test effects on sperm activation (Table 1), and motility estimates were performed as described above. Artificial seawater (Forty Fathoms Bio-crystals Marinemix, Marine Enterprises International, Inc., Baltimore, MD) prepared at an osmolality of 628 mOsm/kg was used to establish the level of complete motility. Evaluation of Cryoprotectant Toxicity Initial cryopreservation studies were performed in the field at the Louisiana Department of Wildlife and Fisheries Lyle S. St. Amant Marine Biological Laboratory on Grand Terre Island . Sperm from two males caught on April 9 were collected by surgical removal and smashing of testis. The sperm were stored in HBSS (186 mOsm/kg) at 4°C. We evaluated four reagent-grade chemicals (Sigma) as cyroproteciants: methanol, glycerol, dimethyl sulfoxide (DMSO), and n 41 -dimethyl acetamide (DMA), Each cryoprotectant was mixed at 50%:50% (v:v) with HBSS before addition to sperm mixtures. All cryoprotectants were used at concentrations of 5% and 10% except DMA, which wasonly usedata 5% concentration because of acute toxicity at higher concentrations (data not shown). Tlie time between addition of cryoprotectant to the sperm and initiation of the freezing procedure was 15 min. Motility was estimated at the initiation of the freezing procedure to determine the acute toxicity of each cryoprotectant to spotted seairout sperm. The sperm used in this study were subsamples of the samples used in the cryoprotectant toxicity and freezing rale study described below. Evaluation of Cryoprotection and Freezing Rates Sperm were cryopreserved in 0.5-inl straws (IMV International Corp., Minneapolis, MN) with two replicates per fish (n = 2) for each treatment. Straws were sealed using glass balls (Minitube of America, Madison, Wl), and were placed into an RPE embryo freezer (Peter Elsdcn and Associates, Collierville, TN) designed for the cyropreservation of mammalian embryos. The RPE embryo freezer consisted of a metal cylinder, with holes drilled for sixteen0.5-nil straws, designed tocreateauniform freezing 184 Cryopreservation of Sperm of Spotted Seatrout rate when lowered into nitrogen vapor in a vapor shipping dewar (Model CP'35, Taylor-Whmon» Theodore, AL). Two fteezing rates were accomplished by placing the freezer at either of two heights within the dewar. Placing the center of the RPE embryo freezer 220 mm from the bottom of the dewar yielded the fastest freezing rate; placement 320 mm from the bottom of the dewar yielded a slower freezing rale. To document the freezing rates, a straw filled with HBS S was inserted into the freezer and the temperature was recorded using a type-T thermocouple and a strip chart recorder. The recording was initiated at the time the straws were placed into the RPE embryo freezer and ended when the straws were removed (30 min after reaching -80“C). Straws were transferred immediately to a larger shipping dewar (Model CP-65, Taylor-Wharton) for storage. After 72 hours of storage, the straws were thawed for 7 sec in a water bath at 40'C. The straws were dried and the ends cut to release the sperm into 1.8-ml tubes. For estimation of sperm motility, a 2-pl aliquot of each sample was activated with 20 pi of HESS (600mOsm/kg) to obtain maximal motility. Inan experimentperformed in the laboratory atLouisiana State UnivCTsity, a computer-controlled iSneezer (Kryo-10, Planer Products Ltd., England) was used to produce a third freezing rate (45*CAnin). Spenn from the two males used for the dewar studies were transported ai4®C and stored for 24 hr before analysis. Dimethyl sulfoxide at 10% waschosenas the cryoprotectant for this experiment, based on results of the experiments performed in the shipping dewar. Straws were frozen using atwo-stepprocedure. Thestrawswerefirslcooled to a temperature of for 5 min, and then frozen atarate of 45‘CymmurUil reaching -80®C. The straws were maintained at -80C! for 20 min, removed from the freezer, and plunged immediately into liquid nitrogen for storage. After 48 hr of storage, g)erm san^les were thawed as described above and motility estimated. TABLE 1 Activation of spotted seatrout sperm. Sperm were stored at 4”C in Hanks’ balanced salt solution (HBSS) at either 152 mOsm/kg or 201 mOsm/kg for 2 days prior to analysis. Two-pl aliquots of sperm were activated with 20 pi of activating solution. Motility was estimated under lOOx dark-field microscopy. Sugar solutions and artificial seawater were used to establish control values. Solutions containing ionic components of HBSS were prepared in a glucose solution (185 mOsm/kg) to supplement osmotic pressure. Storage solution n Activating solution ingredients(osmolality) Final osmolality^ Percent motility 152 2 Artificial sea water (628 mOsm/kg) 585 18 201 2 597 35 152 2 Sucrose (757 mOsm/kg) 724 18 201 2 713 35 152 1 Glucose (321 mOsm/kg) 318 5 152 1 Glucose (258 mOsm/kg) 258 1 152 2 Glucose (207 mOsm/kg) 211 0 201 2 214 0 152 1 Glucose + 8g/L NaCl 424 10 201 1 Glucose + 0.16g/L CaCl 189 0 201 1 Glucose + 0.4 g/L KCl 199 0 201 1 Glucose + 0.06 g/L Na2HP04 183 0 ^ mOsm/ Kg 185 Wayman et al. Statistical Analysis All percent motility values were arcsine-squareroot transfoimed prior to statistical analysis. Motility data derived jfrom the osmotic analysis of sperm activation were compared using a paired Student’s l-test (Microsoft Excel 5.0, Microsoft Corp,). In the cryoprotectant toxicity study, differences in pre-freezing motility were determined using a one-factor analysis of variance (SAS 6.08, SAS Institute Inc ., Cary, NC). In the study evaluating post-thaw motility as a function of cyioprotectant and freezing rale, differences were determined using a two-factor analysis of variance (SAS 6.08). Means were separated using Duncan’s multiple range test, and were considered significant when P < 0.05. Results Osmotic Analysis of Sperm Activation Hie motility of spotted seauout sperm increased with increased osmolality of the HBSS or mannitol activating solutions, with maximum motility (90%) observed at ~375 mOsm/kg and above (Figure 1). Motility of sperm activated in mannitol solutions was not signifrcantly different (F > 0.27) from motility of sperm activated in HBSS, In general, ^-5% of the sperm (subihreshold) became motile at 242 mOsm/kg. The threshold activation point (10% motility) was 262 mOsm/kg, and the complete activation point (90% motility) was 370 mOsm/kg. Ionic Analysis of Sperm Activation To evaluate the activating effect of the various ions contained in HBSS, individual chemical compements of HBSS (at the concentration used in 1-x HBSS) were dissolved in glucose solutions (prepared at 1 85 mOsm/kg to supplanent the osmotic pressure) and used as activating solutions for spotted seaiiout sperm. Only' the glucose solution containing 8 g/1 NaCI proved sperm activation. The osmolality of this solution was 424 mOsm/kg (Table 1), above the complete activation point identified in Figure 1. Speim activated by sucrose solutions at osmolalities of 724 mOsm/kg produced motility equal to diat of sperm activated by aitifrcial seawater at 628 mOsm/kg (Table 1). Spoiled seatrout spemi stored at 152 mOsnVkg demonstrated reduced motility of as much as 50% compared to sp^m stored at 201 mOsm/kg. Evaluation of Cryoprotectant Toxicity The average initial motility of sperm samples was75% at the time of addition of ciy oprotectants. Speim motility at the time of breezing was reduced significantly (P = 0,(X)01) by exposure to glycerol and DMA (Table 2). This loss of motility was likely due to acute toxic effects of the chemicals on sperm. Of the four cyroprotectants, exposure to glycerol reduced pre- fireezcmotility lhemost(to~ 1%). Pre-fieezemolility of sperm exposed to methanol or DMSO was not significantly different (P > 0.05) from motility of control sperm not exposed to ciyopiotectants. Figurel. PercentofmotllityofspottedseatroutspermactivatedwithsolutioiisofHaiiks'balancedsaltsolution (squares) <»* mannitol (triangles) spaniung a range of osmotic pressures. Each point represents the mean of sperm from three fish. Motility of sperm activated with mannitol was not significantly different (P >0.05) from motility of sperm activated with HBSS at corresponding osmotic pressures. All standard errors were less than 10%. 186 Cryopreservation of Sperm of Spoited Seatrout TABLE 2 Mean motility^ (+ SD) before freezing and after thawing of sperm of spotted seatrout (n=2). Sperm was frozen at one of two positions in a nitrogen vapor shipping dewar: lower (-3^ "C/min) and upper (-2.5 *’C/nim); or in a computer-controlled freezer (-45.0 "C/min). Sperm frozen in 10% DMSO had significantly higher (F=0.001) post- thaw motility than sperm frozen in other cryoprotectants. Position within the dewar had no significant effect on post-thaw motility (P=0.15). Pre-freeze modlity values sharing a letter were not significantly different Cryoprotectant Concentration Pre-freeze motility^(%) Post-thaw motilitv (%) Lower Upper Freezer ControP -- 70±0‘ 0±0 0 + 0 " Methanol 5% 75±6‘ 1±1 0 + 0 10% 75 + 6* 0±0 0±0 ” DMSO 5% 65 ±6- 13 ±4 13 + 11 ” 10% 63 ±3* 45 ±21 28+11 - 10% 72 ±4 “ ” 22+ 18 DMA 5% 49125’’ 3±3 1+0 -- Glycerol 5% 1 ± r 1 ±0 1+0 - 10% 1 ± p 1 + 0 1 + 0 “ ^ Initial motility at time of collection was >75%. ^ Motility estimated 15 min after the addition of cryoprotectant. ^ Hanks’ balanced salt solution without cryoprotectant Evaluation of Cryoprotection and Freezing Rate Ten percent DMSO produced the highest post-thaw motility of the four cryoprotectants studied {P = 0.001). Sperm frozen in other cryoprotectants yielded motilities of between 0% and 3%. Samples frozen without any cryoprotectant contained no motile sperni after thawing. The average rates of freezing for the shipping dewar were -3.5artmait of Wildlife and Fisheries for administrative support We also thank J. Buchanan, J. Lighter, M. Mayeaux , B. McNamara, D. Suggs, and R. Talbot for assistance in collection of samples. LiTERATOitE Cited Bates, M.C., W.R. Wayman, and T.R. Ticrsch. 1996. Effect of osmotic pressure on the activation and storage of channel catfish sperm. Trans Amer Fish Soc 125: in press, Bi]]anl,R. 1984.LaocHiservation des gambles ell’insSminationartificidle chez le bar et la daurade. In: G. £amab6 and R. Billard, (eds.). L’ Aquaculture du Bar et des Sparides, p 227-231. INRA Publications, Paris. Blaxtcr, J.H.S. 1953. Spring storage and cross-fertilization of spring and autumn spawning herring. Nature 172:1 189-1 190. Chambeyron, F. and Y. Zonar. 1990. A diluent for sperm cryopreservation of gilthead seabream, Sparus aurata. Aquaculture 90:345-352. Colura, R. L. 1974. Induced spawning of the spotted seatrout, Cynoscion nebulosus (Cuvier). Proc Fifth Ann Workshop World Mar Soc, p 319-330. Crocker, P.A., C.R. Arnold, J.A. DeBoer, and G.J. Holt. 1983. Blood osmolality shift in juvenile red drum, Sciaenops ocellatus L, exposed to fresh water. J Fish Biol 23:315-319. Gwo, J., K. Strawn, M.T. Longnecker, and C.R. Arnold. 1991. Cryopreservation of Atlantic croaker spermatozoa. Aquaculture 94:355-375. Henderson- Arziqjalo, A. and R.L. Colura. 1984. Black drum x red drum hybridization and growth. J World Mar Soc 15: 412-420. Henderson-Arzapalo, A.,R.L. Colura, and A.F.Maciorowski. 1994. A comparison of black drum, red drum, and their hybrid in saltwater pond culture. J World Aquacult Soc 25; 289-296. Palmer, P.J., A.E. Hogan, and C.G. Barlow. 1994. Cluiled storage of pikey bream {Acanthepagrw berda) sperm and activation in different salinities. Asian Fish Sci 7:35-40. Saucier, M. C. and D. M. Baltz. 1993. Spawning site selection by spotted seatrout, Cynoscion nebulosus, and black drum, Pogonias cromis, in Louisiana. Environ Biol Fish 36:251 -272. Scott, A.P. and S.M. Baynes. 1980, Areviewof thcbxology, handling and sUxage of salmonid spennatozoa. J Biol 17:707-739. Stoss, J. 1983. Fish gametepresejvationandspennaiozoan physiology. In: W,S. Hoar, D J. Randall and E.M. Donaldson, (eds.). Fish physiology, Vol. 9B, p 305-351. Academic Press, New York. Thomas, P. and N. Boyd, 1988. Induced spawning of spotted seatrout, red drum, and orangemouth corvina (Family: Sciaenidae)withluteinizing hormone-releasing hormone anabg injection. Contrib Mar Sci 30:43-48. Tiersch, T.R., C.A Goudie, and GJ. Carmichael. 1994. Cryopreservation of channel catfish sperm: storage in cryoprotectanls, fertilization trials, and production of channel catfish with cryoprcserved sperm. Trans Amer Fish Soc 123:580-586. 188 Gulf Research Reports Volume 9 | Issue 3 January 1996 Aricidea (Allia) bryani, 2iNew Species ofPolychaete (Polychaeta: Paraonidae) from the Northern Gulf of Mexico Gary R. Gaston University of Mississippi Jerry A. McLelland Gulf Coast Research Laboratory , Jerry.McLelland^usm.edu DOI: 10.18785/grr.0903.06 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Gaston, G. R. and J. A. McLelland. 1996. Aricidea (Allia) bryani, a New Species ofPolychaete (Polychaeta: Paraonidae) from the Northern GulfofMexico. Gulf Research Reports 9 (3): 189-195. Retrieved from http:// aquila.usm.edu/gcr/vol9/iss3/6 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9, No. 3, 189- 195, 1996 Manuscript received September 5, 1995; accepted October 9, 1995 ARICIDEA (ALLIA) BRYANI, A NEW SPECIES OF POLYCHAETE (POLYCHAETA: PARAONIDAE) FROM THE NORTHERN GULF OF MEXICO Gary R. Gaston* and Jerry A. McLelland^ ‘Biology Department, University of Mississippi, University, Mississippi 38677 , USA ^Invertebrate Zoology Section, Gulf Coast Research Laboratory, P.O. Box 7000, Ocean Springs, Mississippi 39566, USA ABSTRACT Aricidea bryani, a new species of polychaele (Polychaeta: Paraonidae) belonging to the subgenus A//iaStrclzov 1973, is described from shallow subtidal sediments along the northern shore of Mississippi Sound, an estuary of the northern Gulf of Mexico. The new species is distinguished from other members of the subgenus by the presence of tubcrculate ncuropodial lobes in the anterior 15-20 setigers, a ciniform median antenna that extends posteriorly to setiger three, and by modified ncuropodial setae that are abruptly tapered at mid-length, but lack terminal aristae. Introduction Specimens of an undescribed species of Aricidea Webster l879(subgenusA//inStielzov 1973)werccoIlected near the historic Biloxi Lighthouse (Harrison County, Mississippi) during January 1991 to April 1992. The collection site (type locality) was a shallow, subtidal, mesohal ine habitat in water less titan 2 m deep. Collections were made adjacent to a fishing pier that extended seaward from an artificial, public beach on the northern shore of Mississippi Sound, an estuary of the northern Gulf of Mexico. Aricidea Webster, 1879, is one of six genera in the family Paraonidae Cerruti, 1909; the others are Paraonis, Paraonides, Cirrophorous^ Levinsenia^ and Sabidius. Based on the presence and characteristics of modified neuroselae, Strelzov (1973) proposed subdividing the genus Aricidea into four subgenera as follows: Aricidea s. str; Aedicira Hartman, 1957; and two new subgenera, AUia and Acesia. Fauchald (1977) elevated Strelzov’s four subgenera to generic rank without giving specific reasons, and Hartley (1981) corrected the subgenus name of Acesta to Acmira, since the former was preoccupied in the phylum Mollusca. The taxonomic scheme used here follows Strelzov (1973) and Hartley (1981). Holoiypes, paraiypes, and additional material from the type locality are deposited in the U.S. National Museum of Natural History, Smithsonian Institution (USNM), Washington, DC. Other specimens are deposited in the museum of the Gulf Coast Research Laboratory (GCRL), Ocean Springs, Mississippi, and in the personal collections of the authors. Taxonomy Family Paraonidae Cerruti, 1909 Genus Aricidea Webster, 1879 Subgenus AUia Strelzov, 1973 Aricidea (AUia) bryani, new species Figures 1-4 Synonym: Aricidea cf. alisdairi (Gaston 1984, in part). Material Examined Type material. Northern Gulf of Mexico, Biloxi, Mississippi (30®10’N, 88°53'50'’W). Holotype 22 mm length, 1.0mm width (USNM 172124), 4 paratypes(USNM 172125), June 30, 199 1 ;26 paraiypes from two collections made on March 28, 1991 (USNM 172126) and April 2, 1992 (USNM 172127), shallow subUdal (0.5-1.0 m), sediment type: well-sorted fine sand; all type material collected by K. Matulewski. Additional material, Apalachee Bay, FL (30®1 .5 ’ N, 84®16.2'W): 5 specimens collected during July 1991,4.6m depth, muddy sand (83% sand); Oyster Bay, FL (30^.0’N, 84®17.2'W): 4 specimens collected during July 1991,2.1 m depth, sandy mud (57% silt-clay); Mississippi Sound, MS (30®15.3’N, 88°26.rW): 1 specimen collected during July 1991 . 4.9 m depth, sandy mud (90% silt-clay); Chandeleur Sound, LA (29®58.2'N, 88°5 l.rW): 5 specimens collected during July 1991, 2.4 m depth, muddy sediments (96% sill-clay); Terrebone Bay, LA (29“45.0’N, 90°15.0*W): 5 specimens collected during July 1991,2.3 m depth, muddy sediments (95% silt-clay); Caracahua Bay, TX 189 Gaston and McLeixand (28"37.6’N, 96®22.5'W): 1 specimen collected during July 199 1» 1.2 m depth, sandy mud (73% silt-clay); Off St. Peteisburg, FL (27^57 ’(X).4”N, 8356’00.5”N, 83®27’29.6’’W) (USNM 090216): 1 specimen collected during August 1977, 30 m depth, clayey, sandy silt; Off St. Petersburg, FL (27’^52’30.5’’N, 83®33’59.0”W) (USNM 090217); 1 specimen collected during August 1977, 34 m depth, clayey, sandy silt; Off St. Petersburg, FL (27°57’28.8^’N, 83®42’29.2”W) (USNM 090218): 1 specimen collected during July 1976, 37 m depth, silty- very fine sand; Off Apalachicola, FL (29°30’N, 84®27'W) (USNM 090219): 1 specimen collected during July 1976, 24 m depth, medium-fine sand. DESCRimON Based on holotype and paratype material. Body dorsoventrally flattened in branchialregion,moiecylmdrical in postbranchial region. Prostomium triangular, bluntly pointed anteriorly (Figure 1 A). Nuchal slits very slender, inconspicuous, directed anterolaterally. Median antenna cirrifonn, widest near base, extending to anterior edge of setiger 3 (Figure IB). Two small eyes present, faint in preserved specimens. Branchiae beginning on setiger 4, numbering 25-33 pairs on large specimens, 14-18 on very smjill specimens. Branchiae broad basaUy, tapering to slender, bluntly pointed terminus; overlapping across dorsum and perpendicular to body axis (Figures 1, 2). Branchiae subequal, except those of last pair, which are similar in shape and smaller, about 2/3 length of others (Figures 2D, 4A). Ciliated bands occurring across dorsum between pairs of branchiae and on branchiae, but less developed in the lastpair (Figure 4 A)* Notopodial postsetal lobes digitate on seligers 1 and 2 (slightly longer on setiger 2), twice as long on setiger 3 and thereafter, thinner in postbranchial region to pygidium (Figures 2, 3). Neuropodial postsetal lobes tuberculate from setiger 1 to about setiger 18, absent on seiiger20 and thereafter (Figures 1 , 2, 3). Notosetae all thin and capillary, numbering about 22-25 per fascicle in prebranchial and branchial region and about 8- 1 3 per fascicle in postbranchial region. N eurosetae number about 22-28 per fascicle in prebranchial and branchial region and about 35 per fascicle in postbranchial region, stouter and somewhat shorter than notosetae in prebranchial region and anterior two-thirds of branchial region (Figures 1, 2); becoming thin, similar to notosetae in posterior third of branchial region and posteriorly to about setiger 60 at which point fascicles include setae abruptly tapering at about midlength with pubescent fringe on seial shaft where abrupt tapering occurs (Figure 3C); pubescent fringe extending from midlength to terminus of shaft, forming sharp dp (Figure 4B); abruptly tapering neuroseiae becoming more numerous in far-posterior setigers (Figure 3B), comprising most of setae in fascicles near the pygidium. Body narrowing at pygidium to half its postbranchial width. Pygidium with two ventrolateral and one ventromedial anaj cirri: ventromedial cirrus about half the length of two lateral cirri (Figure 1C). Variation. The median aniemia generally is cirriform, but antennae of some paraiypes have a greater swelling near the base than that of the holotype. Additionally, some smxill (young) specimens have short, tuberculate antennae thatappearto be inadevelopmental stage. Specimens from offshore Florida have similarnumbersof branchiae, similar notopodial and neuropodial lobes, similar neurosetae and notosetae, but shorter antennae than on other specimens of A. hryani (to setiger 1 in USNM 090216,090217, 090218, 090223; to middle of setiger 2 in USNM 090219). Remarks. Aricidea hryani^ n. sp., belongs to the fsubgenusAllia, which currently has 17 species characterized by the presence of neurosetae that are markedly thicker than corresponding notosetae. These characteristic neurosetae have abruptly tapering shafts, w ith more abrupt tapering ‘m far posterior segments. Aricidea (Aricidea) fragilis, which is the type species for the genus, has similar neurosetae to species of Aricidea (AUia), bulk distinguished by the presence of pseudocompoimd (pseudoarticulate) neurosetae in some jXDsterwrsegmenls (see Hartman 1957: 317). Although we were not able to examine type specimens, our examination of A. fragilis specimens from offshore North Carolina (USNM 51181), from Fivers Island near Beaufort, North Carolina, and from Perdido Key, Horida, revealed that all had the characteristic pseudocompound neurosetae in posteriorsegments. We arc inclined, therefore, toraaintain the subgenus Allia based on its modiftod seta! type as set foitli by Strelzov (1973). Aricidea (Aricidea) fragilis is otherwise distinguished from A. hryani by the presence of short (not tuberculate) neuropodial lobes to sctiger40 (not to 20), and by its short, subulate median antenna, which extends at most to setiger 2 (Webster 1879, Pettibone 1965:129). Comparative characteristics of the species of Aricidea (Allia) are summarized in Table 1, Four species of Allia, like A. hryani, have neuropodial lobes and a ciiriform antenna that extends posteriorly to at least setiger 3, but differ from the new species in the following respects: A. ahranchiata lacks branchiae; A. quadrilobata has a long, filamentous antenna extending to setiger 9, well-formed, digitate neuropodial lobes anteriorly, and biramous notopodial cirri in the branchialregionof largerspecimens; A. suecica (=A. nolani; Strelzov 1973) has modified setae with terminal aristae on posterior segments; A.pseudanne 190 New Species of Paraonid Polychaete TABLE 1 Species of Aricidea, subgenus Allia with some comparative morphological characteristics for specimens of maximum size. Includes data from Hasan (1960), Pettibone (1963), Hartman (1965), Day (1967), Imajima (1973), Strelzov (1973), Katzmann and Laubier (1975), Hartley (1984), and Gaston (1984). Neuropodial lobes Antenna Species . end at setiger type end at setiger Pairs of branchiae A. marianne Katzman Laubier, 1975 tuberculate 13 short, ovoid, digitate tip 0 12-19 A. bulbosa Hartley, 1984 tuberculate 15 short, fusiform 0 21 A. albatrossae Pettibone, 1957 tuberculate 12-25 short, subulate 0 26-30 A. hartmani (Strelzov, 1968) tuberculate 12 short, conical-clavate w/cerratophore 0-2 15-19 A. ramosa Annenkova, 1934 tuberculate 17 short, branched 1* 13-17 A. roberti Hartley, 1984 tuberculate 17 short, cylindrical 1 22-26 A. Claudia e Laubier, 1967 tuberculate 9 basally enlarged, sharply attenuated tip 1 16 A. curviseta Day, 1963 tuberculate 20 cirriform 1 44 A. bryani n. sp. tuberculate 18-20 cirri form 3 25-33 A. abranchiata Hartman, 1965 tuberculate 2-4 cirriform 4-7 0 A. suecica Eliason, 1920 tuberculate 25 cirriform 5 30 A. pseudanne Katzmann & Laubier, 1975 tuberculate 5-6 cirriform 5-6* 13 A.quadrilobata Webster & Benedict, 1887 conical, fusiform 25 cirriform, slender 9 27 A. monieae Laubier, 1967 absent cirriform 0 9 A . facilis Strelzov, 1973 absent - short, club-shaped 0 9-15 A.pulchra Strelzov, 1973 absent - cirriform 1 18 A. alisdairi Hasan, 1960 absent - cirriform 2 43 A. Imajima, 1973 absent - long, clavate 2 20 * Based on illustration is smaller, has an antenna extending to setigers 5-6, only 6-13 pairs of branchiae, and has neuropodial lobes only on the first six setigers. Two other species are similar to A. bryanU but are distinguished as follows: A, roberti has a short, cylindrical antenna, fusiform notopodial postsetal lobes on first two setigers (longer than those of A . bryani)^ and neuropodial lobes to scliger \1\A. bulbosa has bulbous posterior branchiae. The new species was originally described as Aricidea (Allia) cf . alisdari by Gaston (1984), but not all of Gaston's specimens weredelennined lobe A . btyani. Some specimens had similar nolosetac and neurosetae to A. bryani, but longer notopodial lobes in the first two setigers and either shorterantennae (USNM 090220, 090222) or longeranienna (USN M 090221) than A. bryani Another specimen (USNM 090215) had a short, clavate antenna (to setiger 1) and pseudoariiculate neurosetae in the posterior region. Ecology. Aricidea (Allia) bryani was abundant in the littoral zone of Mississippi Sound. Specimens at the type locality occurred in densities of 28-555/m’ during January 1991 to April 1992 (Malulewski 1995). Sediments of the area consisted primarily of well-sorted fine sand that periodically contained considerable amounts of silt and detritus. Most preserved specimens of the new species had a distinct corkscrew shape, and juveniles almost always were green, possibly due to ingested algae. Adults were whitish with an orange-brown stripe dorsally in some specimens. Presence of small specimens in the monthly collections at the type locality led the authors to conclude that larval settlement occurred from January to March during 1991. Gnly large specimens (adults) were collected during the summer. Specimens from the type locality were usually filled with ingested sand grains that were packed in the far-posterior gut. A combination of the filled gut and the 191 Gaston and McLelland Figure 1. Aricidea (AlUa) bryani, neyf species. A. Anterior end, dorsal view. B. Anterior end, lateral view. C. Pygidium (holotype), dorsal view. D. Seligerl. £. Setiger3. Parapodium in posterior view, number of setae reduced for clarity. Scales: A, B = 0.5mm; C, D, E =: 0.1mm. 192 New Species of Paraontd Polychaete Figure 2. Aricidea (Allia) bryani, new species. Representative branchial parapodia, posterior views, number of setae reduced for clarity. A. Setiger 6. B. Setiger 9. C. Setiger 16. D. Setter 29. Scale = 0.1mm. 193 Gaston and McLelland thin body wall resulted in loss of the far-posterior region uptake, and may help keep the body surface free of debris, andpygidiumofmostspecimeasduringsample preparation. Similar bands were evident in scanning electron Aricidea (Allia) bryani was one of over 40 species of micrographs of Aricidea roberti and Aricidea suecica polychaetes that inhabited the type locality. (Hartley 1984). Ciliated bands across the dorsum and on the branchiae Etymology. The species is named in honor of Bryan (Figure 4A) probably function in motility, enhance oxygen Deaver Gaston, son of the senior author. Figure 3. Aricidea (Allia) bryani^ new species. Representative postbranchial parapodia, anterior views, number of setae reduced for clarity. A. Setiger37. B. SetigerllS. C. Detail ofneuroseta from setiger 118. Scales: A, B=:0.1mm; C =0.05mm. 194 New Species of Paraonid Polychaete Acknowledgments We gratefully acknowledge Kenneth Matulewski, who collected the type material and hydrographic data as part of his M.S. degree research and whocontributed to an earlier version of tliis manuscript We also thank Robert Allen (GCRL) for conducting the scanning electron miciuscope work and taking the photographs and Carol M Cleveland (UM) for reviewing the manuscript We are grateful to Dr. Kristian Fauchald (USNM, Smithsonian Institution) for lending specimens for review. Specimens collected during July to August 1991 were provided by the U.S. Environmental Protection Agency Environmental Monitoring and AssessmentProgram (EMAP). Specimens collected during 1976 and 1977 were provided by the U.S . Department of the InteriorBLMMoniloringProgram (Gaston 1984). Literature Cited Day, J. H. 1967. A Monograph of the Polychaeta of Southern Africo. 2. Sedentaria.,p559-878. London: BritMus (NatHist). Fauchald, K. 1977. The Polychaete Worms. Definitions and Keys to Orders, Families, and Genera. Los Angel Co Mus Nat Hist Sci Ser 28:1-190. Gaston, G. R. 1984. Chapter 2, Family Paraonidae. In: J. M. Uebclacker & P. G. Johnson (eds). Taxonomic Guide to the Polychaetes of the Northern Gulf of Mexico, p 2-1 - 2-53. Barry Viltor and Associates, Inc., Mobile, Alabama. 7 vols. Hartley, J. P. 1981. The family Paraonidae (Polychaeta) in British waters: A new species and new records with a key to species. J Mar Biol Assoc UK 61:133-149. Hartley, J.P. 1984. Cosmopolilanpolychaetespccies; the status of Aricidea belgicae (Pauvcl, 1936) and ivotes on the identity of A. suecka Eliason, 1920 (Polychaeta; Paraonidae). In: P.A. Hutchings ,cd. . Proceedings of the Pist Internationa] Polychaete Conference, Sydney, p 7-20. Linncan Soc NSW. Hartman, O, 1957. Orbiniidae, Apistobranchidae, Paraonidae andLongosomidae. Allan Hancock PacExped 10(3): 31 1-388. Hartman, O. 1965, Deep-water benthic polychaetous annelids off New England to Bermuda and other North Atlantic areas. Occas Pap Allan Hancock Found 28:1-378. Hasan, S. A, 1960. Some polychaetes from the Karachi coast. Ann Mag Nat Hist (13). Imajima, M. 1973. Paraonidae (Polychaeta) from Japan. Bull Natl Sci Mus (Tokyo) 16:253-292. Kalzmann, W. & L. Laubier. 1975. Paraonidae (Polych^tes sfidenlaires) de TAdriatique. Ann Naturhis Mus Wien 79:567-588. Matulewski, K. 1995. Temporal variation in the macrobenthic community structure of a subtidal artificial beach in Biloxi, Mississippi. M.S. Thesis. Uni v South Miss, Hattiesburg. A Figure 4. Aricidea (Allia) bryani, new species. Scanning electron micrographs. A. Posterior branchial region, dorsal view. B. Neuropodial fascicle from posterior region, anterior view (setiger 130). Scales: A = 0.1mm; B = 0.05mm. Pettibone, M. H. 1963. Marine polychaete worms of the New England Region. 1 . Aphroditidae through Trochochaetidae. Bull US Nat Mus 227:1-356. Pettibone, M.H. 1965. Two new species of Aricidea (Polychaeta, Paraonidae) from Virginia and Florida, and redescription of Aricidea fragilis Webster. FVoc Biol Soc Wash 78:127-140. Strelzov.V.E. 1973, Polychaetewormsofthe family Paraonidae Cerruti, 1909— Polychaeta Sedentaria. Academy of Sciences of the USSR, Order of Lenin, S. M. Kirov Kola affiliate, Murmansk Marine Biology Institute (English translation, Amerind Publishing Company Private, Limited, New Delhi, 1979). 212 pp. Webster, H.E. 1879. Annelida Chaetopoda of the Virginian coast. Trans Albany Inst. 9:202-269. 195 Gulf Research Reports Volume 9 | Issue 3 January 1996 Observations on an Isolated Population of Sagitta hispida Conant (Chaetognatha) in a Tropical Lagoon System of Northeast Yucatan (Mexico) Jose N. Alvarez-Cadena Universidad Nacional Autonoma de Mexico Eduardo Suarez-Morales Colegio de la Frontera Sur, Mexico Jerry A. McLelland Gulf Coast Research Laboratory , Jerry.McLelland^usm.edu DOI: 10.18785/grr.0903.07 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Alvarez-Cadena; J. N., E. Suarez-Morales and J. A. McLelland. 1996. Observations on an Isolated Population of Sagitta hispida Conant (Chaetognatha) in a Tropical Lagoon System of Northeast Yucatan (Mexico). Gulf Research Reports 9 (3) : 197-204. Retrieved from http://aquila.usm.edu/gcr /vol9/iss3/7 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports, Vol. 9, No. 3. 197-204, 1996 Manuscript received July 17, 1995; accepted September 27, 1995 OBSERVATIONS ON AN ISOLATED POPULATION OF SAGITTA HISPIDA CONANT (CHAETOGNATHA) IN A TROPICAL LAGOON SYSTEM OF NORTHEAST YUCATAN (MEXICO) Jos4 N. Alvarez-CadenuS Eduardo Su^rez-MoralesS and Jerry A. McLelland^ ‘ Universidad Nacional Aut6noma de Mixico, Instituto de Ciencias del Mar y Limnologia. Estacidn "Puerto Morelos", P,0. Box 1152, 77500, Canc&n, Q,Roo, Mixico ^ Colegiode la fronteraSur, UnidadChetumal, P.O.Box424, 77000, Chetumal. Q.Roo.Mi^dco ^ Invertebrate Zoology Section, Gulf Coast Research Laboratory, P.0, Box 7000, Ocean Springs. MS 39566-7000, USA ABSTRACT Monthly zooplankton collections were carried out from January to December 1991 at two sampling sites, Cuenca Norte and Bojdrquez lagoon, in the Nichupt6 lagoon system, a partially enclosed network of interconnected waterways located in the northeastern region of the Yucatan Peninsula (Mexico) adjacent to the Caribbean Sea. Only one species of Chactognatha, Sagitta hispida Conant, was present and was more abundant at Cuenca Norte (total density 450.6 organisms/m^) than at Boj6rquez (138*6 organisms/m’). The latter site is smaller, more physically isolated, and more environmentally stressed than the former. From monthly gonadal and length-frequency analyses of 1390 specimens, it was found that (1) total length significantly differed among four successive maturity stages, (2) juvenile and immature specimens occurred in greater numbers at Bojdrquez, while more mature specimens comprised a greater percentage of individuals found at Cuenca Norte, and (3) individuals collected atBojdrquez, where slightly higher temperatures were recorded, were significantly smaller than those from Cuenca Norte. The latter two findings indicate that Sagitta hispida spawns at a higher frequency at Boj6rquez, possibly due to the cumulative effect of higher temperature. iNTRODUCnON Chaetognaths are among the most abundant holoplanktonic animals in oceanic, neritic and coastal environments (King 1979; 0resland 1990). Like most other zooplanktcrs, chaetognaths produce more generationsat lower latitudes where temperatures are higher (Dunbar 1941, 1952, 1962; McLaren 1963; Alvarifio 1965; Sameoto 1971). This higher breeding frequency has an impact on the composition of otha zooplankiers such as copepods which are repwted to be their main ixey item (Reeve 1970; Szyper 1978; Pearre 1980; Canino and Gram 1985; Alvarez-Cadena 1993). Sagitta hispida Conant, 1895, is a conspicuous chaetognath commonly occurring in neritic waters on both sides of the tropical-equatorial Atlantic Ocean (Alvariflo 1965). It has often been reported as abundant in coastal waters of the Gulf of Mexico (Pierce 1951; McLelland 1989), along the east coast of the United States (Pierce 1953; Pierce and Wass 1962; Owre 1960; Grant 1962, 1963) and from some areas of the Caribbean Sea (Suilrez- Caabro 1955; Michel 1984; Su^z-Morales et al. 1990; McLelland and Heard 1991; McLelland et al. 1992). In tropical regions, especially in areas with dense submerged vegetation, Sagitta hispida has been observed with a near bottom distribution (Owre 1972; McLelland and Heard 1991) leading Bieri (1991) to consider the species as “quasi-planktonic**. Otlier field studies demonstrate a marked diel migration for the species; mature specimens are rarely collected in surface waters during the day (McLelland and Heard 1991). Sweatt and Forward (1985) determined in laboratory studies that 5. jpidifldemonstiates all-or-none upward vertical movement when ambient light intensity is below approximately lO'® "^ photons m'^ s'*. There are no pre viou $ reports on the chaetognath fauna in the NichupuJ lagoon system (NLS), located on the Mexican coast of the Caribbean Sea. In this p^r, we document the unique occurrence of Sag/rm hispida in this lagoon system and compare the populations of the species at two sites with (Uffming hydrological conditions, while considering data on body length and four preadult maturity stages. Study Area The NLS is located adjacent to the Caribbean Sea (21" 07’N, 86" 46'W) in the northeastern region of the Yucatan Peninsula (Figure 1). The climate in the region is subhmnid and warm (lowest temperatures are higher than 1 8" C) with the main rainy season in summer and moderate rainfall in winter (type AWl (X*)(i’) g of Garcia 1964), Although the NLS (type IV-B of Lankford 1976) at present is largely surrounded by tourist facilities, it was originally bordered by mangrove vegetation. In tropical oligotrophic 197 Alvarez-Cadena et al. Figure 1. Study area and sampling sites in the Mexican Caribbean. systems such as NLS, the typical submerged vegetation is characterized by Thalassia testudinum, Halodule sp. and rhyzophytic or calcareous algae. The soil in the area is highly porous and permeable (BuUerlin 1958; Ldpez-Ramos 1974). Freshwater runoff from rivers that typify other lagoon systems is lacking. Subterranean springs and “cenotes’ ’ (karstic waterdeposits) provide variable amounts of freshwater input into the system. Three climatic regimes are reported annually in this area: “nortes”, dry. and rainy seasons (Merino and Otero 1991). During the present study period (1991), ‘hortes”, identified by the strong northern winds which blow in the area, extended from December to March. The dry season, with dominant southeastern winds and low precipitation, extended from April to July. The rainy season, with the same wind pattern but higher precipitation, extended from August to November. For this study, two intrinsically different sampling areas at adjacent locations within the NLS (Figure 1) were compared. Station 1 consisted of two sites in Bojdrquez lagoon, a nearly enclosed, shallow (1.5 m average depth) lagoon with two narrow openings to other parts of the NLS. Bojdrquez lagoon is further characterized by high salinity, nutrient enriched water from organic pollution, and patches of Thalassia testudinum and Halodule. Anthropogenic stress at this site is considerable and was detailed by Alvarez-Cadena and Segura-Puertas (in preparation). Station 2 was located in Cuenca Norte, a larger, less stressed body of water averaging 2.5 m in depth and with a small, but distantopeningtotheCaribbeanSeatothenorth. Of the two, station 2 is, on the whole, more representative of the NLS. Materials and Methods Zooplankton samples and surface hydrographic data (salinity and temperature) were collected at the two stations monthly from January to December 1991, All collecting was performed between 0900 and 1100 hrs. A conical plankton net (0.42 m diameter, 330 urn mesh) was equipped with a General Oceanics flowmeter and lowed in a circular path near the surface for 5 minutes at 1 .5 knots. Samples were preserved in thefield with buffered (lithium carbonate) formalin at a concentration of 5% formalin-seawater. In the laboratory, all chaetognaths collected were counted and examined for species identification, whereupon at least 50 animals were randomly removed from each sample, measured to the nearest 0. 1 mm, and examined for gonadal condition. Measurements were made from the tip of the head to the tail excluding the tail fin. Maturation stages were assigned as follows and were based on various classifications reviewed by Alvariflo (1965); Stage I. Ovaries not visible, or if present are very small, not reaching anterior part of posterior fin; few small ovocites piesent Seminal vesicles not present and no sperm visible in tail segment Stage 11. Ovaries visible, usually reaching mid space between lateral fins or posterior part of anterior fm; small ovocites present Some thickening occurring near end of posterior fin indicating primordium of seminal vesicles; sperm pre.sent but not occupying entire tail segment Stage in. Ovaries reaching midpoint of anterior fm; ovocites rounded. Seminal vesicles visible but somewhat flattened; spenn present throughout tail segment Stage IV. Ovaries reaching anterior part of anterior fin;ovocitesrounded,arrangedusuallyintworows. Seminal vesicles oval to rectangular, testes full of sperm. 198 SaGHTA HISPIDA FROM A TROPICAL LaGOON SYSTEM Figure 2. Annual variation of salinity at Boj6rquez and Cuenca Norte. Boj6rquez data are mean values recorded from two sampling sites. Figure 3. Annual variation of temperature at BoJ6rquez and Cuenca Norte. BoJ6rquez data are mean values recorded from two sampling sites. Stage V. Ovaries reaching posterior part of ventral ganglion, ovocites fully matured- Seminal vesicles completely full or sometimes “spent” (empty). Chaetognath data firom the two sites in Bojdrquez were pooled as the mean lengths of the animals did not show any significant differences. Likewise, pooled values are given for salinity and temperature recorded from Bojbrquez (Figures 2,3). Data from the two stations were compared statistically using mean length measurements (total N=1390) for maturity stages I-IV. A two-way analysis of variance (ANOVA) was used to test for variation among stages at each station and for differences between stations. A t-test was employed to further compare the differences among the mean lengths of each maturity stage between the two stations. Testing was perfonned at «=0.05withconfidence intervals of 99%. 8 r 7 6 A 5 O) « A 9 4 k ^ 5 A Cuenca Norte □ B0J6RQUE2 2 I > ' ' I II III IV Stage Figure 4. Annual mean length variations of Sagitta hispida maturity stages I-IV at Cuenca Norte and Bojdrquez. Results Salinity at Bojdrquez was usually higher than at Cuenca Norte (Figure 2). Lower salinity was observed at both stations during January (“nortes**) with 29.63 and 19.72®/oo respectively. Highest values were recorded in August for Bojdrquez (35,44°AX))andmJulyforCuencaNorle(34.68®/oo)- T^nperature was usually slightly higher at Bojdrquez than at Cuenca Norte (Figure 3). At both stations, the lowest temperature was recorded in January and the highest in July-August The taxonomic analysis of chaetognalhs collected at both sites during the survey revealed the presence of only one species, Sagitta hispida. It was about three times as abundant at Cuenca Norte (450.6 organisms/m^) than at Bojdrquez with a total density of 138.6 organisms/m^ (Table 1). I^portionally, stage I was the most common of all maturity stages, and comprised a higher percentage of the population at Bojdrquez (89.2%) than at Cuenca Norte (7 1.2%). Stages n to IV were consistently higher in percentage at Cuenca Norte than at Bojdrquez (14.1% stage n, 6.4% stage III, and 8.3% stage IV from Cuenca Norte and 7,6% stage II , 1.8% stage m, and 1.4% stage IV for Bojdrquez). Only seven adult specimens of stage V (8.5-9.0 mm total length) were collected during the survey, six at Cuenca Norte and one at Bojdrquez. Because of this extreme scarcity, comparative analyses were omitied for this stage. Mean lengths were variable during the annual cycle (Table 2, Figures 4, 5a and 5b), especially those of stage I at Bojdrquez which ranged from 1 .59 mm in August to> 4 mm in March, April and November (variation range = 64%). Stage I animals at Cuenca Norte did iK)t show this range of size variability (3.18 mm in May to 4.8 mm in August; variation range 33%). The mean lengths of the remaining stages showed less variability at both sampling areas. 199 Alvarez-Cadena et al. TABLE 1 Numbers analyzed montbly for maturity stages I-IV and total density (organisms/m^ of SagUta hispida at Bojdrquez and Cuenca Ncule. Data for Bojdrquez represent pooled values from two sites. Bojdrquez Stage Jan Feb Mar Apr May Jun Jul Aug Oct Nov Dec Total (org./m’) % I 49 82 41 16 70 80 82 78 98 72 40 18 726 89.2 n 0 8 8 9 1 2 1 13 2 13 5 0 62 7.6 m 0 0 0 2 1 0 1 5 0 3 3 0 15 1.8 IV 1 0 1 1 0 0 0 4 0 2 2 0 11 1.4 OrgymS 11 5.9 3.1 2.1 4.1 15.3 5.1 39.5 19 13.5 13 6.3 138.6 - Cuenca Norte Stage Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total (org./ffl^) % I 29 30 38 39 21 37 42 18 35 44 35 42 410 111 n 11 12 8 8 0 7 1 9 7 3 7 8 81 14.1 m 5 6 2 2 0 2 4 6 4 1 5 0 37 6.4 IV 5 2 2 1 0 4 3 17 9 2 3 0 48 8.3 Org./m3 5 9.6 12.5 15.8 44.3 82.5 49.4 133.3 17.9 0.3 31 49 450.6 - Length data for each stage at the two sites (1390 measurements) were tested using a two-way analysis of variance ( ANO V A) for variation in mean length in sampling sites vs. the four maturity stages. Residual data showed a normal distribution with the Kolmogorov-Smimov test. It was found that significant differences ( « = 0.05) existed in length vs. sampling site (Fs=9.55,p=0.002), maturity stage (Fs=227.97, p<0.001), and interaction between the two factors (Fs= 13.12, p<0.00 1). It was further found that total length significantly differed among successive maturity stages and that individuals collected at Cuenca Norte were significantly larger than those caught at Bojdrquez for each of the four maturity stages (Figure 4). Both parameters were constant at all levels and the terms of significant interaction were not evident, but could be explained by the larger difference between stage I lengths at both sampling sites (Figure 5a, b). In order to compare mean length differences of each maturity stage between the two sampling sites, t-tests were used with confidence intervals of 99% and testing HO at <^=0.05. For stage I, with 1 136 individual length measurements (410 from Cuenca Norte and 726 from Bojdrquez), differences were highly significant (t=- 17.58), exhibiting the widest range of difference between the two sampling areas. For stage II, with less available data (Ns 143), differences were also significant (t=-3 .46). However, the same analyses with stage III and IV data revealed no differences al 99% CJ. between the lengths of Cuenca Norte and Bojdrquez specimens. Finally, when this analysis was performed pooling the total number of observations (1390) from all stages at each station group, the differences were found to be significant (ts 18.58) between the two sampling areas. 200 SaGITTA HISPIDA FROM A TROPICAL LaGOON SYSTEM TABLE 2 Monthly mean length (mm) of Sagitta hispida for maturity stages I-IV at Bojdrquez and Cuenca Norte. Data for Bojdrquez represent pooled values from two sites. Bojorquez Stage Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. OcL Nov. Dec. Grand Mean I 2.61 2.77 4,17 4.72 3.32 2.62 2.17 1.59 2,98 2 J 3 4.36 3.34 2.956 n — 5.66 5.87 6.44 5.70 — ... ... 5.56 ... 5.761 m — — — 7.00 5.50 “ — ... — 6.76 6.660 IV 7.45 — 6.90 ... ... — ... ... ... 8.00 ... 7.070 Cuenca Norte Stage Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct Nov. Dec Grand Mean I 4.27 4.72 4.28 4.14 3.18 3.21 3.68 4.77 4.46 4.23 4.46 3.34 4.181 n 6.10 6.84 6.21 6.01 — 5.13 6.00 6.61 6.02 6.40 6.01 — 6.251 m 7 . 4 S 7.80 6.70 6.90 -- 6.00 6.65 6.98 7.02 6.70 6.40 6.985 IV 8.10 7.25 7.76 7.70 — 6.54 7.13 7,34 7.57 7.70 7.76 7.481 Discussion Identification of immatme chaetognaths is often difficult and confusing because of their small size and undeveloped, taxonomically important sexual features. This is particularly complicated inlocalideswheieseveralspedesco-occur. Sagitta /i/rp/d47,however,canbeieadilyidentifiedbyitsconspicuously large, wide head, square to bean-shaped eye pigment and the presence of gut diverticulae in larger specimens (McLelland 1989). Immature specimens can bedistinguished from similar species by examining the anterior teeth which, under high magnification, appear to lie flat against the head, their tips formingalinenearlypeipetKliculartothebody axis; fiirthennore, the teeth have longitudinal ridges which give them a quadrangular cross-sectional £q^)earance. Theseridgescanbe seen in SEM photographs by Cbsper and Reeve (1970). In mature specimens, the ovaries often extend past the midpoint ofthe anterior fins andcontamioundovaananged in two rows. The lateral fins are completely rayed, with the posterior fins leachingthcseminalve^cles. Thelatteraieovaltoiectangular when ripe, and sq^aiBted from the caudal fin by half their length. Sagitta hhpida has been reported to tolerate wide ranges of salinity and temperature. Experiments by Reeve and Walter (1972) on growth rates in laboratory populations found that the species will not survive more than 15 days at temperatures above 33 They also found that the species can grow to maturity in salinity ranges from 25-4()®A)0, but failed to mature at 20°/oo and below. At both stations intheNLS, temperatures were never below 25®C or above 33‘C. On the other hand, salinities were not below 28®/oo or above 36®/oo. Thus, the ranges were well within the known ranges of growth and maturation and may explain the local success of 5. hispida. The fact that the species was represented in our samples mainly by immature specimens likely indicates a combinarion of continuous spawning of the population, a greater mortality of older individuals, and the known differential vertical distribution of juvenile and adult chaetognaths. Sagitta hispida isknown to breed more or less continuously throughout the year in tropical and subtropical waters (Pierce 1951; Owre 1960), resulting in a large proportion of immature specimens representing multiple generations coexisting at different stages of development (Pearre 1991). A concentration of immature chaetognaths in the upper water column is thought to be an indication of shallow-water spawning (Stone 1969) or better survival conditions for younger individuals (Raymont 201 (mm) (rnrn) Alvarez-Cadgna et al. Figure 5. Monthly mean length of Sagitta hispida maturity stages 1-IV at Cuenca Norte and Bojdrquez. 202 Sagitta hispida from a Tropical Lagoon System 1963; McLelland 1984). Adults of S. hispida have been reported to dwell near the bottom or associate with submerged seagrass beds during the day (Owre 1972; Sweatt and Forward 1985; McLelland and Heard 1991)and migrate upward at nighL Although our sampling efforts were near the surface* our data confirms that young individuals can be found abundantly in the water column even during the day. Inadditional samples (notieportedhem) collected overa24-hour period in April and October in Bojdrquez lagoon* larger numbers of adult animals were collected at ni^t* further supporting the assumption of diel migration for this species. The statistical analyses of length variations in the surveyed populations of Sagitta hispida clearly showed that (1) successive maturity stages are well defined in regards to corresponding body length* (2) differences between sampling stations were more apparent in stages 1 and II* and (3) Cuenca Norte individuals were consistently larger at all stages as compared to those from Bojdrquez. Food availability did not seem to be the reason for differences in size between the two stations. In a concurrent study. Alvaiez-Cadena and Segura- Puertas (in preparation) found that cc^)q)ods in Bojdrquez. where chaetognaths were smaller, were nearly three times as abundant as those of Cuenca Norte (6177.6 and 2236/m^ lespecdvely. aimual abundance). On the other hand, it is possible that when food is overly abundant, the metabolic energy of the chaetognath is shifted to repixxluctive output instead of growth, accounting for a smaller ^ at maturity (fast^ rate of maturation) at Bojdrquez. The two dominant genera of the NLS copepod population. Acartia and Paracalanus, were iqwrted by Reeve (1966) to be the main food items for 5. hispida in Biscayne Bay. Florida, Temperaturc differences between the two stations might account for the observed dissimilarities in body sizes and relative pn^xationsofinatmity stages. Forsomezooplaiiktonic orgaiusms, temperature seems to be related not only to the attainment of larger size at maturity, but also with the number of generations produced by the species. Dunbar (1941) mentioned that *'it is generally true that zooplanktos of high latitudes (colder waters) devdopmoreslowly .reach largersize and live longer than related forms in warmer areas.** McLaren (1963, 1966) reported that Sagitta elegans, a ciicumbOTeal species, required more lime to reach maturity at lower temperatures. It may be argued that temperature differratces between Bojdrquez and Cuenca Norte are very small (on the Older of l^C or less) and not significant enough for these differences. However, the cramilalive effect of temperature, rather than just the slightly higher values, may be responsible for the smaller mean size yet tq>parently higher spawning frequency of^. /u’spiV/aatBoJdrquez. Sameoto(1971)repQrted that once S. elegans had accumulated 738®C degree-days, the species would reach maturity. Jakobsen (1971) remarked that small differences in temperaturecouldpromote distinct gonadal development due to the ex tent of the period they exerted their influence. Dunbar (1962) also found that although hydrographic differences (e.g., temperature) were not large, they also had a cumulatiA^ effect on the biology of 5’. elegans. Reeve and Walter (1972) reported that Sagitta hispida completes its entire life cycle in 18 to 50 day.s. Although thenumberof generations forS. Aijpfdacannotbeaccurately determined in this work because samples were collected at monthly intervals, apparently a higher number of generations is produced at Bojdrquez than at Cuenca Norte as evidenced by the higher fiequency of juveniles (stage I) recorded throughout the year. A more frequent sampling effort might clarify this. The population of /li^p/dflintheNLS.especiaUy Bojdrquez, is largely isolated from continual genetic input from coastal populations by the combined effects of low- energy tidal flow and poor circulation which hinder exchange with the adjacemmarine environm^t. Residence of water in the NLS, as a whole, has been estimated at about two years (Merino et al. 1990), except for occasional catastrophic events such as Hurricane Gilbert in 1988 which overwashed the narrow land barrier separating the lagoon system from the Caribbean Sea. In Bojdrquez, wind-aided circulation has been further diminished by the dense line of hotels and other tourist facilities along the Caribbean seacoast which obstruct the dominant southeast trade winds from reaching the lagoon. In a personal communicaiion to the second author, S. Van der Spoel suggested that tangible variations begin to appear in a typical planktonic species population that has been isolated for 200 generations. Given a more or less continual rate of growth and a generation turnover of 20-30 days for Sagitta hispida, such population divergences should occur within 10- 15 year intervals. Thus, it is possible that because of its isolation, a population of Sagitta hispida characterized by smaller adults is evolving in Bojdrquez lagoon. Acknowledgments We wish to thank F. Escobar de la Llata for helping with the map and some of the figures. F. Ruiz-Renteria and Ma. E. IslasrLanderos collaborated during the field sampling and salinity determinations. This paper was substantially improved by the comments of anonymous reviewers. 203 Alvarez-Cadena et al. Literature Cited Alvarez-Cadena, J.N. 1993. Feeding of the chaetognath Sagitta elegans Verrill. Estuarine Coast Shelf Sci 36:195-206. Alvarino, A. 1965. Chaetognaths. Ann Rev Oceanogr Mar Biol 3:115-194. Bieri, R. 1991. Six new genera in the chaetognath family Sagiltidae. Gulf Res Rep 8(3):221-225. Buttcrlin. J. 1958. Rcconocimiento gcol6gico del Tw^ritorio de Quintana Roo. Bol Asoc Mex Gcol Pel 10(9-10):531-570. Canino» M.F. and G.C. Grant. 1985. The feeding and diet of Sagitta tenuis (Chactognatha) in the lower Chesapeake Bay. J. Plankton Res 7(2):175-188. Cosper, T.C. and M.R. Reeve. 1970, Structural details of the mouthpartsof achaetognath as revealed by scanning electron microscopy. Bull Mar Sci 20:441-445. Dunbar, M.J. 1941, The breeding cycle in Sagitta elegans arctica Aurovillius. Can JRes SeetD Zool Sci 19:258-266, , 1952. Marine macroplankton from the Canadian eastern Arc tic. 2 . Medusae, S ipho nophorae, Ctenophora, Pteropoda and Chaetognatha. Can J Res Sect D Zool Sci 20(3);71 -77. . 1 962. The life cycle of Sagitta elegans in Arctic and subarctic seas, and the modifying effects of hydrographic differences in the environment. J Mar Res 20:76-91. Garcia, E. 1964. Modificacionesalsistemadeclasilicaci6nclimdtica de Koppen, para adaptarlo a las condidones particulares de la Repfiblica Mexicana, Offset Larios, Mexico. 246 p. Grant, G.C. 1962. The Chaetognatha of the inner continental shelf waters offVirginia, their taxonomy, abundance.anddependence on physical factors of the environment, MS Thesis. College of William and Mary, Williamsburg, VA. 60 p. 1963. Investigations of inner continental shelf waters off lower Chesapeake Bay, Part IV. Descriptions of the Chaetognatha and a key to their identification . Chesapeake Sci 4<3); 107-1 19. Jakobsen,T, 1971. <>itiicHobgyofSflgittociiegani'VetrillaiKlSagf^a soosa I Muller in Inner Oslo^rd. Nerw J Zool 19: 201-255. King, K. R. 1979. The life history and vertical distribution of the chaetognath, Sagitta elegans^ in Dabob Bay, Washington. J Plankton Res 1: 153-167. Lankford, R.R, 1976. Lagoons in Mexico, In: K. Wilcj' (ed.), Estuarine Processes. Academic Press, New York, p 182-215. L6pez-Ramos, E. 1974. Estudio geoldgico de la Penhisula de Yucaldn. Bol Asoc Mex Gcol Pet 25:25-76, McLaren, LA, 1963- Effects of temperature on growth of zooplankton and the adaptive value of vertical migration. J Fish Res Board Can 20(3):685-727. 1966. Adaptive significance of large size and long lifeof the chaetognath Sog/tw elegans in the arctic. Ecology 47:852-855. McLelland.J.A. 19W. Observadonson chaetognath distributions in the northeastern Gulf of Mexico during the summer of 1974. Northeast Gulf Sci 7(l);49-59. . 1989. An illustrated key to the Chaetognatha of the northern Gulf of Mexico with notes on their distribution. Gulf Res Rep 8(2): 145-172. and R.W. Heard. 1991. Notes on some chaetognaths from Pine Cay, Turks and Caicos Islands (British West Indies). Gulf Res Rep 8(3):227-235. McLelland,J.A., R.W. Heard and J.D. Hardy, Jr. 1992. Arrow worms (Chaetognatha) from the near-shore waters of Tobago. Caribb Mar Stud 3:33-40. Merino, I.M. and L. Otero. 1991 . Atlas ambiental costero de Puerto Morelos, Quintana Roo. Centro de Investigaciones de Quintana Roo (CIQRO). 80 pp. Merino, I.M., S. Czitrom, E. Jordan, E. Martin, P. Thom6 and O. Moreno. 1990. Hydrology and rain flushing of the Nichupte lagoon system , Canedn, Mexico. Estuarine Coast Shelf Sci 30:223-237. Michel, H.B. 1984. Chaetognatha of the Caribbean Sea and adjacent areas. NOAA Tech Rep NMFS 15, US Dept of Commerce. 33 pp. 0resland, V. 1990. Feeding and predation impact of the chaetognath Eukrohnia hamata in Gerlache Strait, Antarctic Peninsula. Mar Ecol Prog Ser 63:201 -209. Owre,H.B. 1960. Plankton of the Florida Current, Part VI. The Chaetognatha. Bull Mar Sci 10(3):255-322. . 1972. The genus Spadella and other Chaetognatha. Sarsia 49:49-58. Peatre,S.Jr. 1980.Fee(fingbyChac(ognstha: Thcidationofpmysizcto predabT size in several species. Mar Ecol PiogrScr 3:125-134. 1991 . Growdiandiqxoduction. ImTheBiologyofChaelognaths. Q. Bone, H. Kspp, and A-C. Pknot-BuUs, eds. Oxford Science F^Jblicaticns, Oxford University ftess, p 61-75. Pierce, E.L. 1951. The Chactognatha of the west coast of Florida. Biol Bull (Woods Hole) 100:206-228. . 1953. The Chaetognatha over the continental shelf of North Carolina with attention to their relation to the hydrography of the area. J Mar Res l2(l):75-92. and M.L. Wass. 1962. Chactognatha from the Florida Current and coastal water of the southeastern Atlantic states. Bull Mar Sci 12(3):215-234. Raymont, J.E-G. 1963. Plankton and Productivity in the Oceans , Int Scr Monogr Pure Appl Biol 1 8 . Pergamon Press Ltd., Oxford, UK. 660 p. Rceve,M.R, 1966. Observations on the biology of a chaetognath. In; Harold Barnes (cd.), Some Contemporary Studies in MflrineScience. George AlknBodUnwinLld,Londm,p613-630. .1970. The biology of chaetognaths. L (Quantitative aspects of growth andegg production in In: StecleJ.H. (e(L),MarineFoodChains. 01iverandBoyiM/S(CIRKIPEDIA; RHIZOCEPHALA) PARASITIZING CRABS OF THE GENUS CALLINECTES IN THE SOUTHWESTERN GULF OF MEXICO Fernando Alvarez and Jorge Calderdn Coleccidn de Crustdceos, Imtituto deBiologta, Universidad NacionalAutdnoma de Mexico, Apartado Postal 70-153, Mdxico 04510, D,F., Mexico. ABSTRACT A preliminary study on the interaction between the parasitic hamac\(iLoxothylacus texanus and two of its host species, the blue crab Callinectes sapidus and the dark blue crab C. rathbunae, in the Gulf of Mexico is presented. Data were obtained from 923 crabs, 162 C. sapidus and 761 C. rathbunae, dsiposit^in the Coleccidn dc CrusUccos, Instiuito de Biologfa, Universidad Nacional Aut6noma de Mexico (UNAM), that were collected in 14 coastal lagoons and sites along the Mexican coast of the Gulf of Mexico. The distribution of L. texanus parasitizing each one of the host species, mean host size variation, distribution of number of parasite externae per host, and morphological modifications of the abdomen of the hosts are analyzed. Introduction The crabs of the genus Callinectes, mainly the blue crab C. sapidus Rathbun, and the dark blue crab C. rathbunae Contreras, support one of the most important commercial fisheries, both in terms of volume and value, within the Gulf of Mexico (Anonymous 1994). Of the biotic factors that affect blue crab populations negatively in the Gulf of Mexico, the parasitism by ihizocephalan barnacles may be one of the most important, periodically reaching very high prevalences (Wardle and Tirpak 1991; LorSn et al. 1993). Rhizocephalanbarnaclesparasitizesusceptibleshrimps and crabs through a planktonic larval stage from which an endoparasitic phase originates, a phase that is not evident unless the host is examined histologically. During the intemalphaseof theparasite, the hosiextemal morphology changes; the male abdomen becomes broader through a process that has been called feminization (Reinhard 1950). The emergence of a reproductive body called **extema” follows the endoparasitic phase. The externa emerges through the internal surface of the abdomen of the host after molting, while the host exoskeleton is still soft. In host species parasitized by the rhizocephalan family Sacculinidae, hosts will not molt once the externa has appeared and the mean size of parasitized hosts is usually significantly less than that of unparasitized hosts (Reinhard 1956; O’Brien and Van Wyk 1984), However, the most important effect caused by this parasitism is that host gonads do not mature (parasitic castration sensu O’Brien and Van Wyk 1984). The effects of the parasitisra by rhizocephalans at the population level are: a) parasitized individuals are not removed by the commercial fishery from the population because they do not attain legal size, and b) the parasitized fraction, which does not reproduce, competes with unparasitized individuals for food and space. Studies have been carried out on the distribution of rhizocephalans (Hochberg et al. 1992), host size distribution (CTuistmas 1969; Adkins 1972; Ragan and Matheme 1974), changes in prevalence during outbreaks (Guistmas 1969; Park 1969; Wardleand Tirpak 199 l),therelationship between parasite size (externa size) and host size (Reinhard 1950; Wardleand Tirpak 1991),andthe morphological changes that parasitized crabs undergo (Reinhard 1950; Hochberg et al. 1992). In ^ite of the great economic impomnee of the blue crab Gshery in Mexico, only two studies have recorded data (Mi the prevalence of Loxothylacus texanus Boschma, its seasonal variation, and host size variation: Lordn et al. (1993), who analyzed thecrab populations of Alvarado lagcxin and L6zaro- (Mvez et al. (in press), who studied parasitized crabs in Tamiahua lagoon. The objective of this study is to present additicxial records of parasitized crabs of the genus Callinectes within the Gulf of Mexico in order to update the known distribution of L. texanus, to determine what host species are being parasitized, to establish the host size range, and to present figures of the most common type of moiphological variations of parasitized crabs. Materials and Methods Data on parasitized blue crabs from the southwestern Gulf of Mexico were obtained through the examination of all the C. sapidus and C. rathbunae from theGulf of Mexico deposited in the Coleccidn de Crustdeeos, Instituto de Biologia,UniversidadNacional AutdnoinadeM^xico (UNAM). 205 Alvarez AND Calder6n Fourteen localities were represented in these samples, including coastal areas and coastal lagoons (Altamira, Chairel, Pueblo Viejo, Tamiahua, Casitas, La Mancha, Mandinga, Alvarado, Soniecomapan, Coaizacoalcos, Machona, Atasta, T6rminos, and Champotdn) as depicted in Figure 1. Detailed descriptions of these coastal lagoons and sites can be found elsewhere in a number of papers (Contreras 1985; Y^ez-Arancibia and Day 1988; Rosas 1989). Each crab was identified and examined for the presence of rhizocephalan exiemae. All crabs were sexed and the shape of the abdomen recorded; sex was determined through the inspection of gonopods and genital pores. In this way feminized crabs (crabs thatareparasitized but which do not yet show the parasite externa) were also found. The inlemal surface of the abdomen of all crabs was examined in search of small extemae or scars of Loxothylacus texanus and the number of extemae per crab was recorded. The distribution of number of parasite exiemae per host was compared to a Poisson (random) distribution with achi-square test. The most common types of abdomens of parasitized male crabs were identified (triangular and rounded) and parasitized individuals were classified accordingly by species. A G-test of independence was used on a 2 x 2 contingency table, to test if triangular abdomens were equally frequent in males of both host species, and a Student’s t-test was used to compare their mean sizes. The two types of abdomens for parasitized males and the extra broad abdomen of parasitized females were figured to aid in their identification in the field. Crab sizes correspond to carapace width in millimeters (mm), and mean values are followed ± one standard error. 206 Distribution of Loxothylacus texanus Parasitizing Calunectes in Southwestern Gulf of Mexico TABLE 1 Samples of crabs examined by species and locality within the Gulf of Mexico. Values represent number of parasitized crabs/ number of unparasitized crabs, and mean host size/mean size of unparasitized crabs, followed by ± one standard error. Asterisks represent significant differences between means. All material is deposited in the Instituto de Biologia, UNAM. Callinectes sapidus Callinectes rathbunae Altamira, Tamaulipas 0/9 * Chairel, Veracruz 0/21 Pueblo Viejo, Veracruz 5/1 87.8 ± 10.4 0/90 Tamiahua, Veracruz 9/53 123.0 ±3.7/ 116.4 ±2.5 0/289 Casitas, Veracruz 1/4 101.4 La Mancha, Veracruz 0/13 0/5 Mandinga, Veracruz 0/5 6/123 120.2±7.2/ 109.9 ±1.5 Alvarado, Veracruz 0/4 0/112 Sontecomapan, Veracruz 24/32 91.6 ±1.9/ 111.2 ±3.0* 67/26 86.0 ±1.6/ 104 ±3.3* Coatzacoalcos, Veracruz 0/2 Machona, Tabasco 1/0 94.3 Atasta, Campeche 1/0 IS. 2 T^rminos, Campeche 0/7 V5 81.9 ±3.3/ 113.5 ±3.8* Champot6n, Campeche 0/1 * P< 0.001 Results A total of 923 crabs, 162 C. sapidus and 761 C. rathbunae, was examined; from this total, 38 C. sapidus (23.5%) and 83 C. rathbunae (10.9%) were parasitized (Table 1). Parasitized crabs were found in eight of the 14 coastal lagoons and sites that were examined (Pueblo Viejo, Tamiahua, C^sitas, Mandinga, Sontecomapan, Machona, Atasta, and T6nninos) as shown in Table 1. In our samples, C. sapidus was parasitizeil in Pueblo Viejo, Tamiahua, and Sontecomapan lagoons, Veracruz. The mean size of parasitized C. sapidus ranged from 87 .8 ± 10.4 mm in Pueblo Viejo lagoon to 123.0 ± 3.7 mm in Tamiahua lagoon; the range of sizes of parasitized C. sapidus was from 70.0 to 134.6 ram, both from Tamiahua lagoon, Veracruz. C. rathbunae was parasitized in Casitas and the coastal lagoons of Mandinga and Sontecomapan, Veracruz; Machona lagoon, Tabasco; and Atasta and T6rminos lagoons , Campeche. The mean size of parasitized C. rathbunae ranged from 81.9 ± 3,3 mm in T6rminos lagoon, Campeche, to 120.2 ±7.2 mm in Mandinga lagoon, Veracruz; the range of sizes of parasitized C. rathbunae went from 45.3 mm in Sontecomapan lagoon, Veracruz, to 144.1 mm in Mandinga lagoon, Veracruz. In Tdrminos lagoon, parasitized C. rathbunae and parasitized crabs of both species in Sontecomapan lagoon were significantly smaller than unparasitized crabs; while there were no significant differences in the rest of the localities. The number of parasite extemae in C. sapidus varied from one to three; 32 crabs had one externa (84.2%), five had two (13.2%), and one had three (2.6%). In turn, in C, rathbunae^ the number of parasite extemae ranged from one to four; 50 crabs had one parasite externa (60.2%). 20 had two (24.7%), 11 had three (13.6%), and two had four(2.5%). Assuming that the distribution of parasites is the result of similar processes in all the sites studied and since these distributions departconsiderablyfromnoimality, the distribution of number of extemae per host was analyzed only by species and not by locality. In C. sapidus, the distribution of extemae corresponds to arandom distribution (Table 2), while in C. rathbunae, the distribution departs considerably from random and approaches a contagious one, with a larger proportion of hosts having multiple extemae (Table 3). Out of 38 parasitized C. sapidus, 1 1 (28.9%) were males and 27 (71.1 %) were females; while in C. rathbunae, there were41 parasitized males and 42parasitized females. In C. sapidus, three of 1 1 males (27.3%) had a triangular abdomen, while the remaining eight had a broad abdomen. For C. rathbunae, nine out of 41 crabs (22%) had a triangular abdomen (Figure 2).The frequency of ^jpearance of triangular abdomens was independent of the host species 207 Alvarez AND Calder6n TABLE 2 Distribution of externae of Loxothylacus texanus on 162 CaUinecfes sapidus* Observed ft’equencies are compared (Chi-square test) to the expected frequencies of a Poisson (random) distribution. No. externae per host Observed frequencies Expected frequencies {O-EflE 0 124 122.72 0.013 1 32 34,11 0.130 2 5 4.74 0.014 3 1 0.44 0.712 Total 162 162.01 X^.869.P>0.05 (x^[ 1]= 1 .83, P> 0.05); mother words, triangularabdomens are equally frequent in both host species. Mean host sizes for males with triangular abdomen were 83.76 ± 8.29 mm (range 67.2 to 92.7 mm) for C. sapidus and 82.25 ± 6.53 Iran (range45.3 to 1 15.8 mm) for C. rathbunae\m significant differences were found between the two mean values (t-test, t[10]=0.121,P>0.05). Discussion Two species of the genus Callinectes, C. sapidus and C. rathbunae, are parasitized by Loxothylacus texanus in the southwestern (julf of Mexico. C. sapidus is parasitized, as indicated by the records presented here and complemented with the information provided by Lor^^n et al. (1993), throughout the coast of the State of Veracruz from Tamiahua lagoon south to Alvarado and Sontecomapan lagoons. No clear pattern of variation of host size can be discerned in C. sapidus along the Mexican coast, contrary to what Hochberg et al. (1992) found for the northern and eastern Gulf of Mexico. While in one previous report (Lor^ et al., 1993) C. rathbunae was found to be a second host for L. texanus, no information on the extent of the distribution of this association was available prior to this report. C. rathbunae is an endemic of the Gulf of Mexico, occurring south from the United States-Mexico border to probably T^rminos lagoon, Campeche (Williams 1974); however, it is parasitized only southwards from Casitas, Veracruz, to Campeche. It is relevant to note that although very large samples of C. rathbunae have been obtained from Tamiahua lagoon (Ldzaro-CMvez et al. in press), this species has never been found parasitized in that area, confirming that it is parasitized only in the southern portion of its range. The size range of parasitized C. rathbunae (45.3 to 144. 1 mm) TABLE 3 Distribution of externae of Loxothylacus texanus on 761 CaUinectes rathbunae. Observed frequencies are compared (Chi-square test) to the expected frequencies of a Poisson (random) distribution. No. externae per host Observed frequencies Expected frequencies (0-E)2/E 0 678 640.57 2.18 1 50 110.17 32.86 2 20 9.47 11.70 3 11 0.54 202.61 4 2 0.02 169.93 Total 761 760.77 X^19.30,P<0.0001 is greater than that for C. sapidus (70,0 to 134.6 mm), and nodefined pattern of host size variation along a geographic gradient is evident with the available data. The number of L. texanus exiemae appearing in the two host species differed statistically. In C. sapidus, the occurrence of externae was not significantly different from a random distribution, indicating that the chances of becoming parasitized are the same for all individuals. However, in C. rathbunae, the number of externae per host approached a contagious distribution, suggesting that this species may occur naturally in a more aggregated pattern that may favor multiple infections (Hoeg 1982). The recognition of parasitized crabs in the field is based on the presence of externae of the parasite and on the identification of aberrant forms of the abdomen. In this study, twotypesofabdomenswererecognizedforparasitized males: rounded, similar to a mature female abdomen, and triangular, such as those of immature females. The recognition of the two types of abdomens for parasitized males was first made by Reinhard (1950) with blue crabs from Galveston Bay, Texas. Parasitized males with triangular abdomens may not occur in all populations, as the existence of two morphologies for abdomens of parasitized males was not discussed in an investigation of parasitized blue crabs from the west coast of Florida (Hochberg et al. 1992). The frequency of appearance of both forms in males and mean size of crabs bearing a triangular abdomen did not differ significantly between host species. The origin of the triangular abdomen may be related to the number of times an infected host molts between the time of infection and the time of emergence of the externa (Alvarez 1993), so the extent of feminization would be related to the duration of the internal phase of the parasite. Consequently, the appearance of the two types of abdomens could vary seasonally and geographically. 208 Alvarez AND Calder6n Literature Cited Adkins, G. 1972. Notes on the occurrence and distribution of the rhizoccphaian parasite {Loxothylacus texanus Boschma) of blue crabs {Callinectes sapidus Rathbun) in Louisiana estuaries. LA Wildlife and Fisheries Comm , Tech Bull 2, 13 p. Alvarez, F. 1993. The interaction between a parasitic barnacle, Loxothylactis panopaei (Cirripedia, Rhizocephala), and three of its crab host species (Brachyura, Xanthidae) along the cast coast of North America, Ph.D. Dissertation, Univ. Maryland, College Park, MD, 188 p. Anonymous. 1994, Atlas Pesquero de Mexico. InstitutoNacional de Pesca, Secretaifa de Pcsca, 234 p. Christmas, J. Y. 1969. Parasitic barnacles in Mississippi estuaries with special reference to Loxothylacus texanus Boschma in the blue crab {Callinectes sapidus). Proc Ann Conference SE Assn Gome and Fish Comm 22:272-275. Contreras, F, 1985. Las lagunas costerasmexicanas. Secretaiia de Pesca, 263 p, Hochberg, R. J., T, M. Bert, P- Steele, and S. D. Brown. 1992. Parasitization of Loxothylacus texanus on Callinectes sapidus: aspects of population biology and effects on host morphology. Bull Mar Sci 50:117-132. Hoeg, J. T. 1982. The anatomy and development of the rhizocephalan barnacle Clistosaccus pagwri Lilljeborg and relation to its host Pagurusbemhardus (L.). JExp Mar Biol Ecol 58:87425. Uizaro-Chfivez, E., F. Alvarez, and C. Rosas. In press. Records of Loxothylacus texanus (Crustacea: Rhizocephala) parasitizing the blue crab, Callinectes sqpidus, in Tamiahua Lagoon, Mexico. J Crust Biol. Loran, R. M., A. J. Valdez and F, Escudero. 1993. Algunos aspectos poblacionales de las jaibas Callinectes spp. en lagunas de Alvarado, Veracruz. Cienc Pesquera 10:15-32. O’Brien, J. and P. Van Wyk. 1984. Effects of crustacean parasitic castrators (cpicaridean isopods and rhizocephalan barnacles) on growth of crustacean hosts. In: A. M. Wenner (ed.). Crustacean Issues 3. A. A Balkema, Rotterdam, p 191-218. Park, J. R. 1969. Preliminary study of Biscayne Bay. Q J Fla Acad Sci 32: 12-20, ' Ragan, J. G. andB. A. Matheme, 1974. Studies on Loxothylacus texanus. In: R. L. Amborski, M. A. Hood, and R. R. Miller (eds.), Proc Gulf Coast Reg Symp Diseased Aquat Anim. LSU-SG-74-05. Baton Rouge, LA. p 185-203. Reinhard, E. G. 1950. An analysis of the effects of a sacculinid parasite on the external morphology of Callinectes sapidus. Biol BuU 98:277-288. Reinhard. E. G. 1956, Parasitic castration of Crustacea. Exp Parasit 5:79-107. Rosas, C. 1989. Aspectos ecofisiol6gicos delaajaihas Callinectes sapiduSy Callinectes rathbunae y Callinectes similis de la zona sur de la Laguna de Tamiahua, Veracruz (Crustacea: Decapoda: Portunidae). Tdsis Doctoral, Facultad de Ciencias, UNAM, Mexico, 215 p. Wardle, W, J. and A. I- Tirpak. 1991. Occurrence and distribution of an outbreak of infection of Loxothylacus texanus (Rhizocephala) in blue crabs in Galveston Bay, Texas, with special reference to size and coloration of the parasite’s external reproductive structures. J Crust Biol 11:553-560, WiUiams, A. B. 1974. The swimming crabs of the genus Callinectes {Doc&podsi: Portunidae). Fish Bull 72:685-798. Yfihez-Arancibia, A, & J. W. Day. 1988. Ecologfa de los Ecosistemas Costeros en el Sur del Golfo de Mexico: La Regi6n de la Laguna deT6rmin(>.s. UNAM-Organizaci6nde Estados Americanos, 5l8 p. 210 Gulf Research Reports Volume 9 | Issue 3 January 1996 Cholangioma in a Wild-Caught Sheepshead Minnow (Cyprinodon variegatus) from the Northern Gulf of Mexico Lee A. Courtney U.S. Environmental Protection Agency DOI: 10.18785/grr.0903.09 Follow this and additional works at: http:/ / aquila.usm.edu/ gcr Part of the Marine Biology Commons Recommended Citation Courtney; L. A. 1996. Cholangioma in a Wild-Caught Sheepshead Minnow (Cyprinodon variegatus) from the Northern Gulf of Mexico. Gulf Research Reports 9 (3): 211-213. Retrieved from http:// aquila.usm.edu/gcr /vol9/iss3/9 This Article is brought to you for free and open access by The Aquila Digital Community. It has been accepted for inclusion in Gulf and Caribbean Research by an authorized editor ofThe Aquila Digital Community. For more information, please contact Joshua.Cromwell^usm.edu. Gulf Research Reports. Vol. 9, No. 3, 211-213, 1996 Manuscript received June 30. 1995; accepted August 14, 1995 CHOLANGIOMA IN A WILD-CAUGHT SHEEPSHE AD MINNOW (CYPRINODON VARIEGATUS) FROM THE NORTHERN GULF OF MEXICO Lee A. Courtney U.S. Environmental Protection Agency, National Health and Environmental Effects Research Laboratory, Gulf Ecology Division, Center for Marine and Estuarine Disease Research, J Sabine Island Drive, Gulf Breeze, Florida 32561-5299, USA ABSTRACT A single caseof acholangiomaoccurred in the liver of a wild-caught sheepshead minnow (Cyprinodon variegatus). This is the first biliary neoplasm and second case of a hepatic neoplasm reported from a wild-caught specimen! of this species. The fmdings further demonstrate the susceptibility of the sheep.shead minnow to neoplasm development and add support to its selection as a subject for field monitoring of carcinogenic exposure. Introduction The use of fish species as sentinels of carcinogen exposure in the aquatic environmeni is a significant tool for identifying compromised ecosystems. Many important attributes of fish sentinels make their use in experimental carcinogenicity testing and fieldmoni toting both efficacious and advantageous (Dawe and Couch 1984), One important consideration, particularly regarding field monitoring for environmental carcinogens, is the relatively low rate of spontaneous neoplasm developmentinfishes. Furthermore, a high correlation has been demonstrated between environmental contamination and most reported epizootics of hepatic neoplasms in fishes (Harshbarger and Clark 1990: Baumaim 1992). These factors emphasize the importance of hepatic neoplasms in non-treated fishes and, particularly, in wild populations. The sheepshead minnow, Cyprinodon variegatus Lac6p5de,has demonstrated its susceptibility tochemically- induced hepatic neoplasm development in various caicinogenstudies(e.g.,CouchandCouitncy 1987; Hawkins et al. 1991). It is a small estuarine teleost with a limited home range inhabiting coastal waters from New England to northern South America. To date, the only reports of spontaneous neoplasm development in sheepshead minnows involve thyroid adenomas in aquarium-held specimens (Nigrelli 1952; Lightner and Meineke 1979) and a single case of hepatocellular adenoma in a wild- caught specimen (Oliveira et al. 1994). This paper describes a cholangioma found in a wild-caughtshecpshead minnow from an ongoing study on P-gly coprotein expression in tissues of teleost fishes and the possible role of xenobiotics in the disruption of its function as a iransepithelial efflux pump (Hemmer and Courtney, personal conununication). Materials and Methods Approximately 30 sheepshead minnows (Cyprinodon variegatus) were collected in a lagoon off Santa Rosa Sound on the north shore of Santa Rosa Island, Florida, approximately 2 kilometers east of the Navarre Beach bridge. Specimens were collected by seine net, cul ventrally to open the visceral mass and immersed in Bouin’s solution in the field. They were fixed for 48 hours, washed in running water for six hours and stored in 70% ethanol at room temperature. Liver, intestine and kidney tissues were dissected firom 10 of the preserved specimens, dehydrated in a graded ethanol series, cleared in xylene and embedded in paraffin. ScctianswerecutonaTotaiymicrotomeatSpm, mounted on poIy-L-ly sine coated slides and air dried. Initial sections were processed for immunohistochemical labeling with four different P-gly coprotcin antibodies [monoclonals C219, C494 and JSB-1; polyclonal mdrfAb-l)] and counterstained with Mayer's hematoxylin (Hemmer et al. 1995), Additional sections were stained with Harris* hematoxylin and eosin. Results and Discussion During evaluation of samples for P-glycoprotein antibody reactions, a neoplastic lesion was found in the liver of one specimen, an adult female (4.5 gm wet weight; 50 mm SL). The lesion was a well-circumscribed cholangioma approximately 750 x 960 pm in greatest dimension. It occupied -2.5 % of the liver area in the plane of sectionexamined ajid was situated at the periphery of the liver (Figure 1). The lesion had regular, well-defined borders with no invasion into surrounding parenchyma. It consisted of numerous well-formed bile ducts possessing nonnal-appearing cuboidal epithelium within a minimal 211 Figure 1. Low power magnification showing well-circumscribed cholangioma located at periphery of liver of a sheepshead minnow {Cyprinodon variegatus). (H&C; 76x). Figure 2. High power magnincatiou of the cholangioma showing well-defined bile ducts consisting of normal-appearing cuboidal epithelium, ranging from circular to multilocular in profile and surrounded by a minimal connective tissue matrix. (H&E; 257x>. 212 Cholangioma in a Wild-Caught Sheepshead Minnow (Cyprinodon variegatus) matrix of connective tissue (Figure 2). Bile duct profiles ranged from small circular configunflions to relatively large muliilocular structures. No mitotic figures were noterl. No positive reaction with any of the P-glycoproiein antibodies tested was observed. The architecture and cellular profile of this lesion resembled thatofcholangiomas described from field collections (e.g„ Dawe et al. 1964; Myers etal. 1987;BuntonandBaksi 1988) and of chemically- induced cholangiomas reported from various small fish species(e.g..Hawkinsetal. 1988;GrizzleandThiyagarajah 1988). The present case is important in that it represents only the second case of a hepatic neoplasm from a wild-caught sheepshead minnow. Furthermore, this is the first report of a biliary neoplasm from this species that was not experimentally-induced or from any other wild-caught fish species in the Gulf of Mexico. Numerous specimens from wild and laboratory-reared populations of sheepshead minnows from the northern Gulf of Mexico have been used in toxicity and carcinogenicity studies with no neoplastic lesions reported from any untreated experimental specimens (see Couch and Courmey 1987) and only one neoplasm observed from a wild specimen (Oliveira et al. 1994). These observations support a very low rate of spontaneous neoplasm development in the sheepshead minnow, an important consideration in the evaluation of results of carcinogenicity tests and in field monitoring. Furthermore, Vogelbein el al. (1990) demonstrated the significance of utilizing small fish species that have a restricted home range in monitoring for environmeniaJ carcinogens. This study showed that a few hundred meters can make a significant difference in the histology of fishes located around a point-source contamination. Most other sheepshead minnows used at the Gulf Breeze EPA laboratory were collected at sites several kilometers west of the location where these specimens were sampled. The significance of finding one hepatic neoplasm at this site in a sample of only 10 fish histologically examined is unclear. As in the report of Oliveira et al. (1994), the presentcasedoesnoiconsiiiute an epizootic, and the collection site is considered uncontaminated, with no significant commercial or residential development in its general vicinity and noother apparent sources of xenobiotic contamination. Nevertheless, this report further demonstrates the susceptibility of the sheepshead minnow to neoplastic development and supports its selection as a subject for field monitoring of carcinogenic exposure. Acknowledgment I wish to thank Dr. John Fournie for examining the slides and confirming the diagnosis. This is Contribution 932 of the U .S. Environmental Protection Agency National Health and Enviromnenial Effects Research Laboratory, Gulf Ecology Division, Gulf Breeze, Florida (formerly the U.S. EPA Environmental Research Laboratory, Gulf Breeze). Literature Cited Baumann, P.C. 1992. The useoftumors in wild populations offish to assess ecosystem health. JAquatBcosysHeaUh 1:135-146. BuntDn,T.EandSM.Baksl 1988. Clv)langiumamwhitepesch(M