BIOLOGY OF FIVE SPECIES OF SEAROBINS (PISCES, TRIGLIDAE) FROM THE NORTHEASTERN GULF OF MEXICO

THOMAS C. LEWIS AND RALPH W. YERGERl

ABSTRACT

Geographically, Gulf populations of Prionotus alatus appear to be restricted almost exclusively to the eastern portion of the Gulfof Mexico, while Bellator militaris, P. martis, P. roseus, and P. stearnsi occur over the entire Gulf. Bathymetrically, P. martis is a shallow shelf species; B. militaris and P. roseus, middle shelf species; P. alatus, middle to deep shelf species; P. stearnsi, deep shelf species. The size (standard length) of B. militaris, P. alatus, P. martis, and P. roseus showed a significant positive correlation with increasing depth ofcapture. Bellator militaris showed a significant "prefer­ ence" for fine sandy silt, clay, or mud bottoms. Prionotus stearnsi was captured in significantly greater numbers during daytime trawling and is postulated to swim actively in the water column at night. It appears to spawn from late summertofall or early winter, while the remaining species spawn from fall to spring or early summer. Adult P. stearnsi differed in food habits by consistently consum­ ing relatively large fishes, while juveniles of this species and all the age groups of the other four species fed consistently on crustaceans.

Searobins of the family Triglidae are commonly water. What little is known appears widely scat­ taken in shrimp trawls along the coast ofthe Gulf tered in the literature, usually in faunal lists. The of Mexico where they comprise an important ele­ only in-depth studies on the biology of western ment of the benthic shelf ichthyofauna (Miles North Atlantic triglids (Marshall 1946; McEach­ 1951; Hildebrand 1954; Springer and Bullis 1956; ran and Davis 1970) are on the two species (Pri­ Bullis and Thompson 1965; Roithmayr 1965; onotus carolinus and P. evolans) that do not occur Franks et al. 1972). They are not commercially in the Gulf. important in the GulfofMexico, but at least some Our study was undertaken to analyze the spe­ species are included among the bottomfishes cies composition of the northeastern Gulf triglid that are canned for pet food and reduced for fish fauna on the continental shelf between 20 and meal by commercial Gulf fisheries (Roithmayr 190 m, to determine the distribution and abun­ 1965). Triglids also present a rich source of food dance of this fauna, and to investigate aspects of for the larger, commercially important fishes from their biology. Thirteen species (Bellator brachy­ the Gulf. Prionotus ophryas, P. roseus, and P. chir, B. egretta, B. militaris, Prionotus alatus, P. stearnsi have been found in the stomachs of red martis, P. ophryas, P. paralatus, P. roseus, P. rubio, snapper, Lutjanus campechanus, taken off Pensa­ P. salmonicolor, P. scitulus, P. stearnsi, P. tribulus) cola, Fla. (Jordan and Swain 1885; Jordan and were collected, but only five species (B. militaris, Evermann 1887). Prionotus roseus was reported P. alatus, P. martis, P. roseus, P. stearnsiJ were from the stomachs of red grouper, Epinephelus taken in sufficient numbers to report on their morio, off Tampa, Fla. (Jordan and Evermann biology. 1887). Hildebrand (1954) regarded P. stearnsi as one of the most important forage fishes in the MATERIALS AND METHODS western Gulf where it was noted in the stomachs of rock sea bass, Centropristis philadelphica; red Specimens were collected from July 1969 to snapper; sand seatrout, Cynoscion arenarius; and October 1971 aboard the RV Tursiops and the inshore lizardfish, Synodus foetens. USNS Lynch. Most cruises were conducted Despite their importance as forage fishes, few or aboard the Tursiops from October 1970 to Octo­ no data are available on the biology of the Gulf ber 1971 as part of the "Gulf Shelf Project" species, particularly on those found in deeper conducted by the Edward Ball Marine Labora­ tory, Department of Oceanography, Florida State 'Department of Biological Science, Florida State University, University. Fishes were captured in a 16-foot Tallahassee, FL 32306. (4.9-m) try-net otter trawl with a %-inch (1.9-cm)

Manuscript accepted August 1975. 93 FISHERY BULLETIN: VOL. 74, NO.1, 1976. FISHERY BULLETIN: VOL. 74, NO.1 square mesh body and a lis-inch (0.3-cm) square (catch per unit effort), and transformed [Y = log mesh cod end liner. (X + 1)1for analysis of the variance. Data for all The study area extended along the northeast­ stations (when available) were used for analysis ern Gulf of Mexico from east of the Mississippi regardless of whether or not the particular spe­ River Passes, La., to the western edge of Apala­ cies was present. chee Bay, Fla., over a depth range of 20 to 190 m Bottom temperature was recorded for most (Figure 1). The easternmost stations (between stations by bathythermograph and on a few long. 84°37'W and 85°30'W) were visited, with occasions either by expendable bathythermo­ few exceptions, in October and December 1970, graph or reversing thermometers. Bottom type and January, April, May, July, August, and was determined by examination ofsamples taken September 1971. The remaining stations were in a bucket dredge dragged over the trawl area. visited only once during cruises conducted in one Bottom type was divided into two major classes; of the following months: July, October, and De­ coarse sand overlain with shell hash (type I), and cember 1969; October and November 1970; Janu­ fine sandy silt, clay, or mud (type II). Data for ary, February, April, July, and October 1971. bottom type were not collected at some stations Station locations were determined through loran. and consequently fishes taken at these stations Station depth was recorded from fathometer were not used in analysis of bottom type. Night readings. Depths for a few stations were extrapo­ was considered to be that interval of time be­ lated from soundings recorded for that location tween 1 h after sunset and 1 h before sunrise at on "1100 Series" U.s. Coast and Geodetic Survey that time ofthe year, while day was considered to maps. (For complete station data and specimens be between 1 h after sunrise and 1 h before examined see Lewis 1973.) The principal investi­ sunset. Fish collected at dawn or dusk were not gators of the Gulf Shelf Project determined the used in the analysis of time of capture data. sampling regime for each station. One trawl The standard length (SL) of each fish was sample was taken at each station. Trawling time measured to the nearest millimeter. Identifica­ on the bottom ranged from 10 to 60 min. The time tions were made following Ginsburg (1950), duration for the majority of trawls at shallow Miller (1965), and Miller and Kent (1971). Speci­ stations (i.e. less than 90 m) was 10 min; for the mens were preserved in 10% Formalin2 origi- deeper stations, 20 min. In order to standardize these trawling efforts, catches were recorded as 2Reference to trade names does not imply endorsement by number of fish collected per 10 min trawling the National Marine Fisheries Service, NOAA.

FIGURE I.-Map of the study area sampled by the RV Tursiops and USNS Lynch between July 1969 and October 197J. 94 LEWIS and YERGER: BIOLOGY OF FIVE SEAROBINS nally, transferred to 40% isopropyl alcohol and were removed and the contents analyzed for deposited in the Florida State University identifiable remains. Food items were identified collection. at least to class, and where possible to order and Gonads were examined from specimens taken suborder. The importance of food taxa was judged in October, November, and December 1970, and by their numerical abundance. January, February, April, May, July, August, September, and October 1971. Size at sexual RESULTS maturity was determined by the first appearance ofripe or developing ova in females and enlarged . Bellator militaris (Goode and Bean) testes in males. Females with numerous ripe ova Horned Searobin were judged to be ready to spawn at or very near the date ofcapture. A ripe egg was determined to Bellator militaris was collected widely at be one that was transparent and filled with depths of approximately 20 to 100 m (Figure 2a) numerous oil globules. Its size was measured to and temperatures of 15° to 28°C. Specimens the nearest 0.1 mm with an ocular micrometer. ranged in size from 24 to 111 mm SL. This species Stomachs (including the posterior esophagus) showed the greatest density of all the species

21,,-°~+-----+------_._---t----"-=:----"---1 ..

A

FIGURE 2.-Distribution within the study area of: A, BelLalor militaris and B, Prionotus alatus.

B 1-- ---1 "--__--'''--_----'

95 FISHERY BULLETIN: VOL. 74. NO.1 caught, yielding 1.8 specimens per 10 min trawl­ ing within its depth range (Table 1). A This species was most abundant between 80 and 90 m (Figure 3a). There was a gradual increase in abundance to this depth range fol­ lowed by a sharp decrease. There was a sig­ nificant (P<0.001) positive relationship between 14 increasing size and increasing depth of capture.

A statistically greater (P<0.025) number ofB. 13 CD 51 Z militaris were taken over a fine sandy mud, silt, 14 II or clay bottom (Table 2). There was no statistical difference in catch for day versus night trawling II B (Table 3). z Bel/ator militaris appeared to reach sexual IE maturity at about 65 mm SL in both sexes. The 13 spawning season was protracted as indicated by •... the presence of females with numerous ripe ova ... 21 (0.7 to 0.9 rom in diameter) from November 1970 to July 1971. I 4 c Bel/ator militaris fed primarily on crustaceans ... c> 2 (90 to 95% of the total stomach contents). Juve­ 55 niles (Table 4) appeared to feed primarily on • ... 14 amphipods and natantian decapods; adults (Ta­ :IE -::> 4D ble 5), on natantian decapods, amphipods, and z 13 II z mysids. Adults also fed to a lesser extent on very c ... D small fishes (usually less than 15 mm 8L), :IE 34 polychaetes, bivalves, and gastropods. 55

14 TABLE I.-Number of specimens of five species of triglids 4D 13 collected and the mean number of fish per 10 min trawling. No. of Mean no. per E Species specimens 10 min traWling' Bellator mllitaris 277 1.8 Prionotus a/atus 162 1.0 P. martis 109 1.2 34 55 P. roseus 162 1.2 P. stearns; 113 0.7 25 65 105 145 115 'For trawls within the depth and geographic range of the species. DEPTH (mllll51

FIGURE 3.-Relationship of depth of capture versus catch per Prionatus alatus Goode and Bean unit effort. A. Bellator militaris, B. Prionotus alatus, C. Prio­ Spiny Searobin notus roseus, D. Prionotus martis, and E. Prionotus stearnsi. Number above each bar refers to the number of 10-min trawl­ With one exception, all specimens of P. alatus ing intervals at that particular depth. were collected east ofthe De Soto Canyon (Figure 2b). For this reason all analyses of this species this point followed by a gradual decline. As in were based only on data from stations east of the B. militaris, there was a significant (P

TABLE 2.-The relationship of bottom type to density for five species of triglids. Type I' Type II' Mean no. per Mean no. per Species 10 min trawling Variance 'N 10 min trawling Variance 'N 3F Bellator militaris 0.8 4.7 75 3.1 37.7 42 6.3' Prionotus alatus 0.8 2.2 53 0.7 2.3 85 0.6 P. martls 1.7 27.5 38 0.6 1.3 14 0.3 P. roseus 1.4 8.8 71 1.3 11.4 38 0.0 P. stearnsi 1.1 9.5 30 0.8 2.5 80 0.2 'Type I = coarse sand bottom overlain with shell hash. Type II = fine sandy mud, silt or clay bottom. 'N = the number of 10-min trawling intervals within the depth and geographic range of the species. 3For one factor analysis of the variance for data transformed to Y = 10g(X + 1). 'Significant at P<0.025.

TABLE 3.-Comparison of day versus night trawling for five species of triglids. Night Day Mean no. per Mean no. per Species 10 min trawling Variance 'N 10 min trawling Variance 'N 'F Bellator mifitaris 1.9 30.5 67 2.1 24.3 61 0.1 Prionotus alatus 1.4 14.9 76 0.8 3.4 65 0.3 P. martis 2.0 27.0 33 1.1 4.7 29 1.8 P. roseus 1.0 5.6 64 0.9 9.4 56 0.4 P. stearns; 0.2 1.7 55 1.6 7.9 59 18.5' 'N = the number of 10-min trawling intervals within the species' depth and geographic range. 'For one factor analysis of the variance for data transformed to Y = 10g(X + 1). 'Significant at P

TABLE 4.-Percent of total stomach contents for the juveniles TABLE 5.-Percent of total stomach contents for the adults of of five species of searobins (n = the number of stomachs that five species of searobins (n = the number of stomachs exam· contained identifiable remains). ined that contained identifiable remains).

Bellator Prionotus P. P. P. Bellator Prionotus P. P. P. militarls alatus martis roseus stsarnsl mllltaris alatus martis rossus stsarnsi Taxa n = 15 n = 24 n = 5 n = 14 n ~ 10 Taxa n ~ 59 n ~ 30 n ~ 25 n ~ 54 n ~ t4 Crustacea: Crustacea: Ostracoda 1.1 3.0 Ostracoda 1.2 0.3 0.1 Copepoda 7.9 2.2 Copepoda 3.9 Stomatopoda 10.4 Stomatopoda 2.4 4.3 1.5 Amphipoda 41.6 16.4 38.4 10.2 86.8 Amphipoda 30.1 2.5 21.t 3.0 4.5 Isopoda 3.0 7.7 2.2 Isopoda 1.7 1.8 0.5 Mysidacea 4.5 13.4 13.3 Mysidacea 18.6 5.0 6.4 4.7 Dscapoda: Decapoda: Natantia 25.8 31.3 7.7 71.6 2.2 Natantia 30.9 71.3 39.8 82.4 9.1 Reptantia 7.9 9.0 23.1 1.8 Reptantia 5.2 11.0 10.5 4.5 9.1 Megalops 1.1 4.5 4.4 Megalops 0.9 1.0 1.2 9.1 loea loea 0.5 Annelida: Annelida: Polychaeta 6.7 1.5 23.1 1.3 Polychaeta 0.8 4.6 2.0 Mollusca: Mollusca: Bivalvia 3.4 1.5 1.8 Bivalvia 1.5 0.3 0.6 0.2 Chordata: Gastropoda 0.3 0.1 Vertebrata Cephalopoda 4.5 Osteichthyes 6.0 2.2 Echinodermata: Ophiuroidea 1.2 0.2 Chordata: Cephalochordata 11.7 eter) were collected from November 1970 to April Vertebrata: 1971. No females were collected in Mayor June Osteichthyes 2.0 2.5 2.9 0.8 63.7 and those collected in July 1971 were not ripe, indicating that spawning ceased somewhere dur- Prionotus roseus Jordan and Evennann ing this interval. Bluespotted Searobin Prionotus alatus fed primarily on crustaceans (91 to 97% of total stomach contents). Juveniles Specimens of P. roseus ranging in size from 240 (Table 4) fed on decapods, amphipods, mysids, and to 170 mm SL were collected throughout thestudy stomatopods; adults (Table 5), chiefly on decapods. area atdepths of20 to 90 m (Figure 4a) andbottom Small fishes (usually less than 15 mm SL) made temperatures of 16° to 28°C. It ranked, with P. up the only substantive non-crustacean food item martis, second in density within its depth range in both adults and juveniles. (Table 1). 97 FISHERY BULLETIN: VOL. 74, NO.1

.0

"

flo •

A

.0 .0 11° .0 .0

FIGURE 4.-Distribution within the study area of: A, Prionotus roseus and B, Prionotus martis eel and Prionotus stearnsi (.).

B'------~------'-----"--->..---J

This species was most abundant between 60 mysids, and amphipods; adults (Table 5) even and 70 m. As with the previous two species, P. more exclusively on decapods. roseus showed a significant (P

Prionotus stearnsi Jordan and Swain Depth Distribution Shortwing Searobin The triglids collected in this study fit into four Prionotus stearnsi was collected widely at bathymetric categories: 1) shallow shelf and in­ depths of approximately 60 to 185 m (Figure 4b) shore species, 2) shallow shelfto midshelfspecies, and temperatures from 14° to 21°C. Specimens 3) shallow to deep shelfspecies, and 4) midshelfto ranged in size from 11 to 117 mm SL. It ranked deep shelf species. fifth in density within its depth range. Prionotus martis is a shallow shelfand inshore Prionotus stearnsi was fairly evenly distributed species. Springer and Bullis (1956) reported it within its depth range, but was slightly more from 200 fathoms (366 m) but we feel that this abundant at shallower depths (Figure 3e). Unlike record is based on either a misidentification or the previous four species, there was no significant incorrect station data. All other specimens in relationship between increasing size and increas­ their paper came from 25 fathoms (46 m) or less. ing depth of capture. The maximum depth for our study, 44 m (24 There was no significant difference in catch be­ fathoms), is probably the maximum depth tween bottom types (Table 2). There was, however, reached by this species. It also enters shallow a significantly (P

The occurrence of P. stearnsi in such greater a function of the size attained by the particu­ numbers during the day is difficult to explain. lar species. Two opposing hypotheses can be postulated. First, P. stearnsi may be a diurnal species, active Spawning Season over the bottom during the day, and perhaps burrowing during the night and thus eluding Spawning seasons for the triglids collected in capture. Or second, P. stearnsi may be nocturnal; this study can be separated into two ill-defined during the day it may rest on the bottom exposed categories: 1) Late summer to fall or early win­ to daytime trawls, while at night it may ascend ter and 2) late fall to spring or summer (see into the water column to feed beyond the reach of Figure 5). the trawl. We favor the second possibility be­ Prionotus stearnsi appears to fit into the first cause ofthe general physiognomy ofthis species. category. In our study ripe females were collected Food habits, as will be discussed, also suggest a only in October and December. Longley and more actively swimming existence compared Hildebrand (1941) reported collecting a ripe fe­ with other triglids. male in August off the Tortugas. These limited data and a large number ofvery small specimens Reproduction in collections from October, December, and Janu­ ary indicate that P. stearnsi probably spawns Sexual Maturity from late summer to late fall or early winter. The paucity of ripe females suggests that this species Bellator militaris and P. stearnsi.are rather may spawn at greater depths than those sampled small triglids maturing at 65 and 60 mm SL and in this study. reaching a maximum size around 120 and 135 Three of the remaining four species (J3. mili­ mm, respectively (Ginsburg 1950). Prionotus taris, P. alatus, P. roseus) had obviously pro­ alatus, P. martis, and P. roseus mature at about tracted spawning seasons from fall to late spring 100 mm SL and attain at least 189 mm (Ginsburg or summer. The presence of a number of small 1950), 166 mm (Reid 1954), and 225 mm (Gins­ individuals collected throughout the year further burg 1950), respectively. Marshall (1946) found corroborated the length of the reproductive that P. carolinus and P. evolans mature at about period. 140 and 180 mm SL, respectively, and attain a Prionotus martis was in spawning condition in much larger size than any Gulf species. It ap­ October, December, and April. The presence of pears that the size at sexual maturity is largely only a few juveniles in this study leads us to

1111110' mllllllil

P,la.atus IlltUI _I----?---

p,la.atul 'a.. uI ~_.?..

P,lonolul mlllil ~----?-----_.~

FIGURE 5.-Spawning seasons for five species of searobins. P,lanolul Itu,nll ----?---

101 FISHERY BULLETIN: VOL. 74, NO.1 believe that the young develop in shallower tools for searching along the bottom as in other water. The bulk of spawning appears to take triglids. place from late fall to late winter or early spring since all specimens less than 45 mm SL that we have examined came from March and April ACKNOWLEDGMENTS collections from water less than 20 m deep. Also, We thank Patrick M. McCaffrey for the sugges­ Hastings (1972) collected small specimens of tion that initiated this study and for supplying us P. martis during February to April only (great­ with and helping to collect most ofour specimens. est abundance in April) during his seasonal stud­ We express our gratitude to George C. Miller, ies of the jetty fauna at Destin and Panama who confirmed some of our identifications early City, Fla. in this study, Robert W. Hastings, Christopher C. Koenig, and Robert L. Shipp, who provided us Food Habits with their assistance and encouragement. Ship­ time aboard the RV Tursiops for the Gulf Shelf Rapid retrieval of the trawl from the bottom often resulted in eversion ofstomachs, especially Project cruises was funded through National Science Foundation Contract No. GD-28174, Pat­ in the deeper water species. Hence, analysis of food habits was impeded by small sample sizes. rick M. McCaffrey, principal investigator. The senior author received support through a Na­ Also, the use of numerical abundance of taxa to determine dietary preferences presents an obvi­ tional Science Foundation Trainee Fellowship ous bias. Large numbers of small individuals during the study. would appear dominant when, in fact, they might make only a small percentage of the volume of LITERATURE CITED food consumed. This was the case in the domi­ nance ofamphipods in the stomachs ofjuvenileP BULLIS, H. R, JR, AND P. J. STRUHSAKER. stearnsi. In general, however, individuals of the 1970. Fish fauna of the western Caribbean upper slope. numerically dominant taxa tended to be domi­ Q. J. Fla. Acad. Sci. 33:43-76. nant in size also. BULLIS, H. R., JR., AND J. R. THOMPSON. 1965. Collections by the exploratory fishing vessels On the basis of these limited data, four of the Oregon, Silver Bay, Combat, and Pelican made during five species (13. militaris, P alatus, P martis, P 1956 to 1960 in the southwestern North Atlantic. U.S. roseus) and the juveniles of the fifth (P stearnsi) Fish Wild\. Serv., Spec. Sci. Rep. Fish. 510, 130 p. appear to feed primarily on benthic crustaceans BURNS, C. 1970. Fishes rarely caught in shrimp trawl. Gulf Res. and other benthic organisms. Reid (1954) and Rep. 3:110-130. Springer and Woodburn (1960) examined P sci­ FRANKS, J. S., J. Y. CHRISTMAS, W. L. SILER, R. COMBS, R tulus latifrons and P tribulus crassiceps from the WALLER, AND C. BURNS. northeastern Gulf and also found that both spe­ 1972. A study of nektonic and benthic fauna of the cies fed primarily on crustaceans. Likewise, Mar­ shallow Gulf of Mexico off the State of Mississippi as related to some physical, chemical, and geological shall (1946) found the same to be true for P factor. Gulf Res. Rep. 4, 148 p. carolinus and P evolans from the Atlantic coast. GINSBURG, 1. In contrast, the adults of P stearnsi appear to 1950. Review of the western Atlantic Triglidae (fishes). consume primarily other fishes. The food habits Tex. J. Sci. 2:489-527. GUNTER, G. of the adults ofthis species are different from all 1945. Studies on marine fishes of Texas. Pub\. 1nst. other western North Atlantic triglids examined. Mar. Sci., Univ. Tex. 1, 190 p. Its piscivorous habit lends support to our earlier HASTINGS, R. W. contention that this species is more mobile than 1972. The origin and seasonality of the fish fauna on a its congeners. This type of diet would imply an new jetty in the northeastern Gulf of Mexico. Ph.D. Thesis, Florida State Univ., Tallahassee, 555 p. active pursuit of their prey. - HILDEBRAND, H. H. The fusiform shape of this species also implies 1954. A study ofthe fauna of the brown shrimp (Penaeus an active mode of existence. The head of P. aztecus Ives) grounds in the western Gulf of Mexico. stearnsi with its terminal mouth does not appear Publ. Inst. Mar. Sci., Univ. Tex. 3:233-366. HOESE, H. D., B. J. COPELAND, F. N. MOSELEY, ANDE. D. LANE. to be adapted for bottom feeding. The free rays of 1968. Fauna of the Aransas Pass Inlet, Texas. III. Diel the pectoral fins are more slender and less de­ and seasonal variations in trawlable organisms of the veloped; they likely are not used extensively as adjacent area. Tex. J. Sci. 20:33-60. 102 LEWIS and YERGER: BIOLOGY OF FIVE SEAROBINS

JORDAN, D. S., AND B. W. EVERMANN. MILLER, G. C., AND D. M. KENT. 1887. Description of six new species of fishes from the 1971. A redescription of Prionotus beani (Pisces, Trig­ Gulf of Mexico, with notes on other species. Proc. U.S. lidae). Q. J. Fla. Acad. Sci. 34:223-242. Nat!. Mus. 9:466-476. MOE, M. A., JR. AND G. T. MARTIN. JORDAN, D. S., AND J. SWAIN. 1965. Fishes taken in monthly trawl samples offshore of 1885. Description of three new species of fishes (Prio­ Pinellas County, Florida, with new additions to the fish notus stearnsi, Prionotus ophryas and Anthias vivans) fauna of the Tampa Bay area. Tulane Stud. Zoo!. collected at Pensacola, Florida, by Mr. Silas Stearns. 12:129-151. Proc. U.S. Nat!. Mus. 7:541-545. REID, G. K., JR. LEWIS,T. C. 1954. An ecological study of the Gulf of Mexico fishes, 1973. Biology of searobins (Pisces: Triglidae) from the in the vicinity of Cedar Key, Florida. Bul!. Mar. Sci. northern Gulf of Mexico. M.S. Thesis, Florida State Gulf Caribb. 4, 94 p. Univ., Tallahassee, 84 p. LONGLEY, W. H., AND S. F. HILDEBRAND. RICHMOND, E. A. 1941. Systematic catalogue of the fishes of Tortugas, 1968. A supplement to the fauna and flora ofHorn Island, with observations on color, habits, and local distribution. Mississippi. Gulf Res. Rep. 2:213-254. Carnegie Inst. Wash. Pub!. 535:1-331. (Tortugas Lab. ROITHMAYR, C. M. Pap. 34.) 1965. Industrial bottomfish fishery of the northern Gulf MARSHALL, N. of Mexico, 1959-63. U.s. Fish Wild!. Serv., Spec. Sci. 1946. Observations of the comparative ecology and life Rep. Fish. 518, 23 p. history of two sea robins, Prionotus carolinus and Prio­ SPRINGER, S., AND H. R. BULLIS, JR. notus evolans strigatus. Copeia 1946:118-144. 1956. Collections by the Oregon in the Gulf of Mexico. McEACHRAN, J. D., AND J. DAVIS. U.S. Fish Wild!. Serv., Spec. Sci. Rep. Fish. 196, 1970. Age and growth of the striped searobin. Trans 134 p. Am. Fish. Soc. 99:343-352. SPRINGER, V. G., AND K. D. WOODBURN. MILES, R. M. 1951. An analysis of the "trash fish" of shrimp trawlers 1960. An ecological study of the fishes of the Tampa operating in Apalachicola Bay and the adjacent Gulf Bay area. Fla. State Board Conserv., Mar. Lab., Prof. of Mexico. M.S. Thesis, Florida State Univ., Talla­ Pap. 1, 104 p. hassee, 46 p. Topp, R. w., AND F. H. HOFF, JR. MILLER, G. C. 1972. Flatfishes (Pleuronectiformes). Memoirs of the 1965. A new species of searobin (Triglidae). Q. J. Fla. Hourglass Cruises. Vo!. IV. Part II. Fla. Dep. Nat. Re­ Acad. Sci. 19:259-266. sour., Mar. Res. Lab., 135 p.

103