An Estuarine Low Temperature Fish Kill in Mississippi, with Remarks on Restricted Necropsies

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Faculty Publications from the Harold W. Manter Laboratory of Parasitology Parasitology, Harold W. Manter Laboratory of

12-1974

Robin M. Overstreet Gulf Coast Research Laboratory, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/parasitologyfacpubs

Part of the Parasitology Commons

Overstreet, Robin M., "An Estuarine Low Temperature Fish Kill in Mississippi, with Remarks on Restricted Necropsies" (1974). Faculty Publications from the Harold W. Manter Laboratory of Parasitology. 484. https://digitalcommons.unl.edu/parasitologyfacpubs/484

This Article is brought to you for free and open access by the Parasitology, Harold W. Manter Laboratory of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications from the Harold W. Manter Laboratory of Parasitology by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

AN ESTUARINE LOW-TEMPERATURE FISH-KILL IN MISSISSIPPI, WITH REMARKS ON RESTRICTED NECROPSIES1

bY RO BIN M. OVE RSTRE E T G ulf C oast Res earch Laboratory O cean Springs , M iss iss ippi

ABSTRACT

In January 1973, large numbers of Mugil cephalus (s triped mullet), w eighing approxi- mately 250 gm each, died in tw o fres hw ater localities in tidew ater bayous of Jackson C ounty, M iss iss ippi. Fish identified as Mugil curema, M. cephalus, Megalops atlantica, Dormitator maculatus, and Fundulus grandis w ere found dead in other low s aline es tuarine areas . Fish-kills during cold periods are less commonly encountered in M iss iss ippi than in T exas or Florida. T his particular incident is attributed to conditions of s tress for fishes incompletely acclimated to the encountered low temperatures . T he most deleterious s tress w as the low s aline w ater w hich probably allowed a breakdown in the fishesÕ ion-osmoregulatory mechanisms. Striped mullet and other euryhaline fishes in s alinities greater than 6 ppt s urvived, as did fres hw ater centrarchids and ictalurids in areas w ith dying mullet. O ther s tress es thought to contribute to the w eak- ening of s triped mullet in Paige Bayou during the period of rapidly decreasing temperatures include s tarvation and high levels of pes ticide residues. I n examined fish, the alimentary tracts w ere devoid of food, the gall bladders w ere distended and leaking bile, the livers contained excess lipid material and w ere often s tained throughout w ith bile pigments, and the levels of DDT metabolites and endrin residues in the liver w ere higher than in control fish. Stres s caus ed by low levels of diss olved oxygen, toxic s ubs tances in the w ater, or disease w as dis- counted as a caus e of death.

INTRODUCTION

O n 16 January 1 973, following a s hort period of freezing temperatures , s everal hundred thous and dead and dying s triped mullet, Mugil cephalus Linnaeus , w ere obs erved floating in and lining the banks of Paige and C ooper Bayous near V ancleave, M iss iss ippi. Many times that number of fish, according to local res idents, had s unk or been s w ept out of the area earlier that day or the previous day by the tide. Such mass mortality prompted an immediate s earch of other areas for dead fishes and poss ible caus es for the deaths . Large fish-kills occurring during cold w eather, s uch as the one mentioned, and apparently n ot as s ociated w ith fishing mishaps , toxic s ubs tances in the w ater, or low concentrations of diss olved oxygen have been documented regularly from

1 T his s tudy w as conducted in cooperation w ith the U.S. Department of C ommerce, NOAA, National M arine Fisheries Service, under PL 88-309, Project No. 2-174-R and NOAA, O ffice of Sea G rant, under G rant No. 04-3-158-53. T he U.S. G overnment is authorized to pro- duce and distribute reprints for governmental purposes notw iths tanding any copyright notation that may appear hereon.

FISH C O LD-KILLIN M ISSISSIPPI 329

Florida (Willcox 1 887; Bangs 1 895; Finch 1 91 7; Storey and G udger 1 936; Storey 1 937; M iller 1 940 ; G alloway 1 941 ; T abb and M anning 1 961 ; and T abb et al. 1 962) and T exas (G unter 1 941 , 1 945, and 1 952; and G unter and Hildebrand 1 951 ), but n ot from M iss iss ippi and A labama. Near C or,pus C hristi, T exas , many people take advantage of the first cold s pell by collecting s tunned foodfish before the fishes move offshore. C hristmas (1 973:39) noted that extens ive cold-kills have not been reported from the M iss iss ippi es tuaries , bu t that less extens ive cold-kills do occur. Such deaths happen in G eorgia (Dahlberg and Smith 1 970 ) and North C arolina (Wells et al. 1 961 ), but are rare. I ass ume that the mortalities in M iss iss ippi w ere caus ed by the combination of low temperature and other, poss ibly s ynergistic, s tress -caus ing factors. Some of thes e factors w ill be discus s ed.

MATERIALS AND METHODS

Witness ing the extens ive fish-kill in Paige and C ooper Bayous , tributaries of Bluff C reek in Jackson C ounty, at T owns hip 6 South, Range 7 West, Sections 25, 26, 27, and 35, led the w riter to the ins pection of other areas in the county. T hes e included another tributary of Bluff C reek farther north in Section 1 6; the entrance of a canal s ituated parallel to Bellefontaine BeachÕ at T 8W, R8S, Section 1 3 and its eas tern end at T 8W, R 7S, Section 1 7; G raveline Lake; a canal at T 8S, R8W, Section 1 1 , as w ell as Simmons Bayou in Section 2, both in G ulf Park E s tates ; Davis and Heron Bayous below and near the bridges of old U .S. Highw ay 90 ; Davis Bayou at the G ulf C oas t Res earch Laboratory; the O cean Springs Small C raft Harbor at both Shearw ater and Pine Drive bridges ; and Bayou Porteaux and a nearby bayou, both off O ld Fort Bayou immediately north of O cean Springs . When dead fishes w ere found, as they w ere in s ome of the Bluff C reek tributaries ; Simmons Bayou, a tributary of Davis Bayou; upper Davis Bayou; and a tributary of the O cean Springs Small C raft Harbor at Pine Drive, the s tandard length w as meas ured and the gills , alimentary tract, liver, and gall bladder carefully examined. Repres entative s amples of liver and gill tiss ues from fish w hich w ere comatose and in a terminal s tate w ere fixed in phosphate-buffered 1 0 % formalin for s ectioning. For comparative purposes , additional material w as collected alive from Davis Bayou on January 1 9 and 22 and A ugust 3, 7,24 , and 29, 1 973 and immediately fixed. Liver tiss ue w as obtained both from near the base and tip of the left lobe w ith s ections cut at 4 y or 8 y and s tained us ing hematoxylin and eosin, M ass onÕs trichrome method, HallÕs method for bilirubin, and the periodic acid-Schiff technique w ith and w ithout dias tas e-hydrolys is. A Sudan black B s olution w as us ed to tes t for lipids in frozen tiss ue s ectioned at 1 0 -14 1.1 w ith a cryostat. T he techniques fol- lowed those edited by Luna (1 960 ). Livers from a s ample of s ix s triped mullet from the Paige Bayou kill w ere immediately frozen for pes ticide analys is. Samples for comparison w ere obtained from s triped mullet caught for human cons umption from Davis Bayou on 1 4 M ay 1 973. Six w ere frozen immediately upon death and s ix others left at 25¡C for 6 Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

330 OVERSTREET

hours to reflect any differences attributable to rapid deterioration of residues of pes ticide after death. E ach of the three s amples cons isted of livers from s ix pooled mullet. A liquots of 6.0 to 1 6.0 gm liver w ere added to double the amount of anhydrous s odium s ulfate, 1 0 0 ml nanograde isopropanol, and 1 0 0 ml nanograde hexane, and w ere completely macerated in a s tainles s s teel blender. T he blended tiss ue w as quantitatively trans ferred to a Buchner funnel to remove s olid material. A fter the liquid filtrate w as trans ferred to a s eparatory funnel and w as hed three times w ith distilled w ater to remove isopropanol, it w as filtered throug h anhydrous s odium s ulfate to remove final traces of w ater, and then concentrated to 5 ml in a K uderna- Danish concentrating apparatus . Pes ticide residues w ere then determined by gas- liquid chromatography, us ing a Ni63 electron-capture detector and calibrating w ith s tandard s olutions . A nalys es w ere performed us ing tw o different columns as a check on retention times . Samples of w ater from the s ites visited at the time of the fish-kill w ere collected and analyzed for content of s alt, chlorine, and calcium. Salinity w as determined w ith an A0 G oldberg temperature-compens ated refractometer; chlorosity by the low-precision, s ilver nitrate, titration method of Strickland and Parsons (1 968); and calcium w ith a Perkin-E lmer atomic abs orption s pectrophotometer M odel 305. C limatological data w ere obtained from monthly s ummaries of the U.S. Depart- ment of C ommerce, E nvironment-1 Data Service, for Biloxi, M iss iss ippi; C orpus C hristi, T exas ; and St. Petersburg, Florida.

OBSERVATIONS AND RESULTS

A thin s heet of ice covered the s urface of Paige and C ooper Bayous during the evenings of January 1 3 and 14. T hes e are s ites roughly 1 to 5 m deep and completely fres h during that time of year. O n the following day, vast numbers of s triped mullet s urfaced and died. By the next day, January 1 6, the tide, about 0 .6 m in amplitude, removed the majority of fish, but a minimal es timation of a few hundred thous and carcass es s till remained. M any had s unk and a few fish continued to s urface and die. O ther than mullet, the only dead fish obs erved w ere four s pecimens of the emerald s hiner, Notropis atherimides Rafines que, 1 6 to 76 mm SL (averaging 71 .0 mm) w hich had apparently been killed by the propeller of an outboard motor. ÔÔShellcracker,ÕÕ Òbream,Ó Òbass ,Ó and ÒcatfishÓ actively fed w hen local res idents, w ho frequently fed them, placed food in the w ater. Becaus e of the pres ence of apparently healthy centrarchid and ictalurid fishes w hich w ould have als o been affected by toxic materials , analys is of the w ater for s uch s ubs tances w as not performed. A dditional areas nearer the coas t w ere then ins pected for dead fishes. T hese areas , w ith corres ponding obs erved dead and living fishes, and w ith values for s alinity, chlorosity, and calciu6 in the w ater, are listed in T able 1. O nly the low-saline areas of Davis Bayou, the Small C raft Harbor, and Simmons Bayou contained dead fish. Davis Bayou w as visited after most of the dead fish apparently had been s w ept aw ay: the tidal flow w as s trong, and the remaining fish, all s triped mullet, w ere lining the banks. Signs of raccoon and birds s urrounded many partially-eaten fish. T he Pine Drive area of the harbor, mostly less than 2 m in depth, had much less w ater movement, and the majority of dead fish had s unk. I n contras t to the presence of Overstreet in Gulf Coast

T able 1. Research Some C haracteristics of Water and Fauna on the Moming of 16 January 1973 at Several Localities in Jackson County, M iss iss ippi

Relative amoufit Living es tuarine Salinity C hlorosity C alcium Reports Locality of dead fish fishes obs erved PP t gm chloridelliter PPm

~ Paige Bayou V ery numerous None 0.30 6.5 0 .o (1974) U pper Davis Bayou Several ... 5.0 3.48 I7 3 rA Simmons Bayou M any ... LOW 3: 4(3). Small C raft Harbor Several None 6.0 4.45 94 0

Pine Drive Copyright b U nnamed tributary of None ... 7.5 5.48 175 c.sc O ld Fort Bayou M Bayou Porteaux None Striped mullet, 8.0 5.86 138 ' 1974, s potted s eatrout, 2 and others Heron Bayou None C yprinodontidae 8.5 5.16 Gulf

G raveline Lake None Striped mullet 9.0 5.45 Coast and others C anal entrance at None C yprinodontidae 10.0 6.68 145 Bellefontaine Beach and Poeciliidae Research G ulf C oas t Res earch None M any s pecies 10.5 6.47 141 Laboratory Pier

Small C raft Harbor None White mullet 11.0 6.99 126 Laboratory. Shearw ater Bridge and others C anal in Gulf Park None C yprinodontidae 14.0 8.44 165 E s tates and others

E ast end of Canal at None C yprinodon tidae 14.0 8.99 185 Used Bellefontaine Beach and others W L by permission. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

332 OVERSTREET

exclus ively dead M. cephalus in the mentioned areas, the harbor contained dead individuals of the w hite mullet, M. curema V alenciennes ; tarpon, Megalops atlantica V alenciennes ; and fat s leeper, Dormitator maculatus (Bloch); and n o obs erved living fishes. An es timated 50 0 fish had died, poss ibly all that had been pres ent, and other animals had been feeding on the carcass es. A few hundred meters s outhw es t of this area, w here the s alinity w as higher, w hite mullet and cyprinodontids s w am near the piers. M ullet from the latteL area appeared partially dazed the previous day. A few dead M. cephalus and OM G ulf killifish, Fundulus grandis Baird and G irard, w ere collected from Simmons Bayou in low-salinity w ater, bu t, according to fishermen, many dead fish w ere there earlier. Numbers and meas urements of the collected fish appear in T able 2. An average-sized M. cephalus w eighed about 255 gm (0 .56 lb).

Table 2. Fishes1 Dead in Conjunction with Low Temperature on 16 January 1973 in Jackson County, Mississippi

No. fish critically Standard length in mm Area Species examined Range Average

F'aige Bayou Mugil cephalus 25 170-278 231.0 Upper Davis Bayou Mugil cephalus 5 219-253 230.4 Simmons Bayou Mugil cephalus 13 203-261 232.5 Fundulus grandis 1 83 Ocean Springs Mugil curema 22 86-154 113.4 Craft Harbor Megalops atlanticu 6 226-289 251.5 Pine Drive Bridge Dormitator maculatus 21 40-102 67.0

~ -~ ~~~~ lThe largest and smallest observed specimens of all species were included in each sample.

T here w ere mullet and other fishes in s ome of the areas that w ere n ot ass ociated w ith mortalities . Fishermen caught s triped mullet during and after the cold w eekend of January 1 3-14 in G raveline Bay and Bayou Porteaux. T able 1 lists other obs erved fishes.

During the cold w eather, s triped mullet in the relatively high s alinity of M iss is- s ippi Sound congregated nocturnally about the mouths of bayous and s pread out on the mud-flats during the day. A ccording to a commercial fisherman, few , if any, of the mullet pres ent in Davis Bayou on the night of January 1 4 fed. He s aid he cu t more gall bladders than us ual w hile cleaning thes e fish, s uggesting enlarged bladders. By the following night, he reported, normal feeding behavior res umed. An examination of 1 3 of those fish refrigerated for 2 days confirmed an adequate intake of food: a full alimentary tract, normal-appearing gall bladder and liver, and s everal different paras ites , all features w hich contras t greatly w ith the condition of mullet from the more fres hw ater habitats on the following day. Numerous repres entatives of dead and dying fish from all areas w ere examined. T he gills of dying fish looked healthy; they w ere filled w ith w ell-oxygenated blood and w ere not coated w ith excess ive mucus . T he viscera had s everal as pects in Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

RSH C O LD-K ILLI N M ISSISSIPPI 33:

common: in s pecimem of each s pecies , all s tomachs and intes tines w ere empty, gall bladders w ere greatly distended, portions of livers and often other organs w ere s tained w ith leaked bile, liver parenchyma w as pale, and mes enteric and other visceral. depot fat w as extens ive. I t s hould be emphas ized that the liver's parenchyma w as often s tained throug hout in the comatose mullet. A ll of the tarpon, an infre- quent inhabitant of modern-day M iss iss ippi bayous , appeared thin. Figure 1 illus - trates viscera from a s triped mullet that had recently died. M ullet examined 3 O ctober 1 973 als o had empty alimentary tracts, enlarged gall bladders, and s mall pale areas in the livers (Fig. 2). M ullet, all ripe and nearly ready to s paw n, a condition als o apparently as s ociated w ith a lack of feeding, remained void of all but a minimal amount of food throughout O ctober and the first part of November. Figure 3 s hows the viscera of a fish from December 4 that had been feeding.

Histological s ections of gills and livers from s triped mullet involved in the kill w ere compared w ith tiss ues collected 6 days and then again 7 months later. T he gill tiss ue ofthe former w as n ot diseas ed, had little mucus , and had far few er ciliate pro- tozoans and no monogeneans and copepods , all of w hich w ere common paras ites in material from Davis Bayou w here there w ere n o mortalities . Sections of liver s tained w ith hematoxylin and eosin and w ith Hall's method for bilirubin revealed no con- s picuous bile pigments nor pathological changes in the hepatic ducts that differed greatly among dying mullet and others collected in January and A ugus t, even though bile pigments discolored the parenchyma throug hout much of the livers of the dying fish. E xtens ive cytoplas mic vacuolation, however, did occur in the hepatocytes of the dying fish (Fig. 4). V acuolation, to a less er extent, appeared in all livers w ith the leas t amount found in a fish caught 24 A ugus t (Fig. 5). U s ing routine s tain- ing procedures , tha extraction of either lipids or glycogen can caus e vacuolation. Frozen s ections retain lipids and s ome glycogen us ually remains in s tandard s ections .

Lipid material, but n ot glycogen, w as exhibited in hepatocytes of the fish from the mass mortality. Preparations positive for Sudan black als o occurred for tw o of three livers of fish collected A ugus t 3 and 7 that w ere originally s us pected as being excess ively fatty by gross examination. Negative s tains for fat res ulted for livers of the remaining fish collected January 1 9 and 22 and A ugust 3, 7, 24, and 29. T he Sudan black s tained material throug hout the entire liver, but more abun- dantly in the s ubcaps ular parenchyma, in the dying fish (Fig. 6) and appeared only on distinct portions of the other tw o. Figure 7 s hows vacuolated tiss ue w ithout lipids and Fig. 8 s hows less extens ive lipid accumulation in a mullet during the s ummer. Preparations positive for PA S, hydrolyzed by diastase and thus indicative of the pres ence of glycogen, w ere demons trated for all of the livers not collected from the fish-kill except tw o in A ugus t. O ne of those had the least amount of vacuolation of all livers. Figure 9 s hows the abs ence of glycogen in liver from a dying fish and Figs . 1 0 and 1 1 s how moderate and cons iderable deposits of glycogen, res pectively.

In addition to vacuolation, the livers from the dying fish occas ionally revealed foci of centralobular hepatic necrosis (Fig. 1 2). Such a finding w as n ot unus ual, s ince all livers poss ess ed this poss ibly typical characteristic. Figure 1 3 reveals s ome inflammatory respons e as s ociated w ith an hepatic duct even from fish exhibiting no Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISH C O LD-K ILLIN M ISSISSIPPI 335

Figures 4-5. 4. E xtens ive vacuolation in hepatocytes of Mugil cephalus dying on 1 6 January 1 973, high power view , M ass onÕs trichrome method. 5. M inimal vacuolation in hepatocytes of Mugil cephalus on 24 A ugus t 1 973, high power view , M ass onÕs trichrome method.

other pathological lesions . M ore pronounced w as the inflammation of s pecific ves- s els in fish caught in A ugus t and containing excess ive amounts of fat (Figs . 1 4-16). In a few s ections of livers, but not those from the dying fish, a pigment, apparently a fixation artifact, s urrounded s ome portal veins and occas ionally other vess els (Fig. 1 7). Hepatic parenchyma of s triped mullet, not in distinct lobules , w ere arranged in dual-plated muralia, es pecially notable in Fig. 8, s eparated by s inus oids . T hese s inus oids typically appeared more cons picuous peripherally. Residue levels of s ome chlorinated hydrocarbon ins ecticides appear in T able 3. Livers from M. cephalus dying during the kill poss ess ed higher concentrations of endrin, p,pÕ-DDE , and p,pÕ-DDD than livers collected later from apparently healthy individuals . A llowing for individual variation, little difference exiked betw een pooled livers of fres hly caught and decomposing mullet.

T able 3. Pes ticides in Parts per M illion (milligrams per kilogram) Recovered from Livers of Striped M ullet C ollected 1 6 January and 1 4 M ay 1 973 in Jackson C ounty, M iss iss ippi

Source of mullet Fish-kill Fres h A ged 6 hours Ins ecticide Paige Bayou Davis Bayou Davis Bayou C hlordane,&is0 mer 0 .44 0 .73 0 .38 E ndrin 1 .0 2 0 .52 0 .39 p ,p’- D DT 0 .38 0 .29 0 .33 o,pÕ-DDT 0 .5 1 0 .60 0 .35 p,pÕ-DDE 0 .64 0 .33 0 .32

p,pÕ-DDD ’ 1 .58 0 .53 0 .73 E xamination for internal and exiernal paras ites revealed few organisms. T he fishes from the Small C raft Harbor, s omew hat decomposed, had no obvious paras ites . Dying s triped mullet from Paige Bayou could be examined much more Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

336 OVERSTREET

Figurm 6-8. Frozen Iiver tissue from Margil Figures 9- 1 1. Peripheral areas of liver tissue cephulus stained in Sudan black B solution to from Mugit cephulus stained using periodic test for lipids. 6. Extensive lipid deposition acid-Schiff technique. 9. Liver of fish from in fish from 16 January 1973 kill, high power kill on 16 January 1973 revealing no glycogen view, 7. Vacuolated tissue not containing deposits, medium-high power view. 10. Liver lipids in fish from brackish water on 22 Jan- from 7 August 1973 revealing moderate glyw uary 1973, high power view. 8. Fatty tissue gen deposits, medium-high power view. Dark- from fish on 7 August 1973, medium power ly stained ams representing glycogen wm view. Note dual-plated muralia radiating from hydrolyzed with diastase. 1 1. Fish of 22 Jan- central vein. uary 1973 revealing extensive glycogen deposits, mediumhigh power view. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

Figures 12-17. Signs suggestive of pathological changes in her time of Mugif cephrrlus. 12. Necrosis around central vein of fish From 16 January 1973 kill, medium-high power view, hcmatoxylin and eosin. 13. Minimal white blood cell involvement around hepatic duct of least pathotogically altered liver examined, Crom fish on 24 August 1973, mediumhigh power view, hematoxylin and eosin. 14. More were degeneration and inflammation within portal area of liver from 3 August 1973, low power view, Masson's trichome method. 15. A Iess in- flamed area of same tissue as Fig. E4 with moderate lymphocytic infiltration, mediumhigh power view. 16. High power view of still another mea in the liver illustrating extensive involvement by eosinophils, the large cells surrounding the vein which contains smaller lymphocytes and erythrocytes. hematoxylin and eosin. 17. Suhstancc, apparently a fixation artifact, surrounding vessel of fish from 29 August 1973, medium-high power view, hematoxylin and eosin. Similar pigment surrounded occasional veins%arteries, and hepatic ducts of a few ether mullet. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

338 OVERSTREET

critically, but harbored few paras ites . E xtemally, the mobiline peritrich Trichodina s p. lightly infes ted the gills of most, and the piscicolid leech MyzobdeZZu Zugubris Leidy, 1 85 1 occurred feeding on about 5% of s everal hundred s uperficially examined fish. O ne s uch fish had a light infection of Trypanosoma s p., a protomonad flagel- late infecting the blood, bu t blood s amples of the other mullet w ere n ot collected. T he neoechinorhynchid acanthocephalan, Flroridosentis elongatus Ward, 1 953, infected one of ten fish examined for intes tinal paras ites . O n January 25, res idents along C ooper and Paige Bayous reported that s everal dead mullet w ere again pres ent on the s urface. By the time the area w as visited, a few thous and mullet floated on the s urface, and others could be s een dead on the bottom. E xamination revealed them to be bloated and w ell decomposed. T he adipose eyelids of the fish had become trans lucent w ith hemorrhaging at the bas e, and filamentous algae, up to 4 cm long, covered the external s urface of many fish. I be- lieve all the floating fish had been s unken fish on January 1 5 and had filled w ith gas from decomposition and then s urfaced. Healthy mullet had by this time partially repopulated the bayou. T hat s ame day, as s istants examined s ites of other kills , as w ell as additional s ites , w ithout finding any trace of dead fish. E arlier, on January 22, a res ident living on a canal off M ary Walker Bayou in G autier, M iss iss ippi, reported that about 300 0 dead mullet had floated into his canal, but not adjacent ones , during the previous three days. T hese mullet w ere als o bloated and contained cons iderable filamentous algal growth. T hey probably died earlier in the low-saline w es tern region of M ary Walker Bayou. Personnel at nearby fishing camps northw es t of the U.S. Highw ay 90 bridge over the West Pascagoula River s tated on January 22 that they had obs erved n o unus ual number of dead fish recently, except on the tw o days following the freeze on January 1 3 w hen they s aw large numbers of dead mullet in the upper West Pascagoula River. Since Bluff C reek joins that river, those mullet probably cons tituted the s ame ones under s tudy. In order for the reader to understand relative temperature values, tw o figures are pres ented. Figure 18 reveals that the air temperature reached its lowes t level, -4.4" C , on January 1 3, after declining s teadily for 8 days . For a comparative examination of temperatures in areas producing extens ive mortalities , monthly values for average and lowes t air temperature in C orpus C hristi, T exas ; St. Petersburg, Florida; and Biloxi, M iss iss ippi w ere obtained. During w inter months , temperature values from T exas w ere us ually lower than those from Florida, and therefore closer to those in M iss iss ippi and better to compare for critical differences w ith those from M iss iss ippi w here fish are not as prone to mortality during cold s pells . T his comparison, indicated by Fig. 1 9, s hows that both the average and the lowes t temperatures during cold months w ere lower in M iss iss ippi than in T exas . A ir temperatures replace those for w ater temperatures becaus e of unavailability of the latter data. A lthough s lightly higher in value, air temperatures reflect the general condition of the w ater rather w ell. T he w inter temperature of s hallow w ater and s mall bodies of w ater s urrounded by land is s lightly lower than that in deep and open w aters. O n 1 7 December 1 973, the temperature once again fell below freezing, and a thin s heet of ice covered s ome of the coas tal bayous . A ll the areas w here dead fish w ere found earlier in January w ere examined. T he s alinity of the w ater in those areas along the coas t ranged from 9.0 to 1 9.0 ppt, and numerous different fishes Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISH COLD-KILLIN MISSISSIPPI 339

?!=I IO ce 0) E

I 15 I 15 31 December January

Figure 18. Values for the lowest daily air temperature during December 1972 and January 1973 in Biloxi, Mississippi.

Average Lowest Mississippi - Mississippi ...... Texas ...... Texas

c a E al I- O

I 1 1968 1969 1970 1971 1972

Figure 19. Monthly values for average and lowest air temperatures in Corpus Christi, Texas, and Biloxi, Mississippi, between January 1968 and February 1973. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

340 OVERSTREET

w ere pres ent. Paige and C ooper Bayous w ere completely fres h, bu t, according to operators of fishing camps , all the mullet had moved out about 2 w eeks earlier during s ome less s evere cold w eather. In any event, n o dead fish w ere obs erved on or following December 1 7.

DISCUSSION

T he death of fishes reported here most certainly res ulted from the low temperatures , but the caus e appears to be more complicated than temperature alone. Why did the es tuarine s triped mullet die in Paige Bayou, w hile, in the s ame area, centrarchids and ictalurids fed eagerly, and w hile mullet s urvived in M iss iss ippi Sound and bayous of relatively high s alinity? Stres s on the fish in different amounts and in addition to that caus ed by low temperature probably can be attributed to factors involved in osmoregulation, temperature-rate-change, acclimation, nutrition, pes ticides , as w ell as other poss ibilities. M any w ill be discus s ed.

Salinity- Osmoregulation

A marine fish, s uch as any of the dead fishes , w hich is capable of entering brackish w ater normally counteracts osmotic and ionic changes in low s aline w ater by regulating blood and other body fluids . I ts ability to regulate osmolarity als o depends on temperature. T he pres ent res ults s uggest that at the encountered temperature or its rate of change, mullet and the other obs erved euryhaline fishes could not tolerate concentrations of or less than about 6.0 ppt s alinity w ith 4.5 gm chloride per liter and 94 ppm calcium. E xact values of concentrations are probably meaningless , and pres umably, if the level of calcium ions or, perhaps , other ions w ere higher, then s alinity per se w ould have even less meaning. C alcium ions affect permeability of membranes , bu t their exact function remains poorly understood. C ameron (1 970 ) s tudied blood of s triped mullet from the natural environment in T exas and found lowered hematocrit values , hemoglobin concentrations , and red blood cell concentrations during w inter. In opposition to his finding for the pinfish, Lagodon rhomboides (Linnaeus ), C ameron did not find a decrease in hemoglobin and hematocrit value in mullet w hen s alinity w as decreas ed. Striped mullet can be trans ferred from s eaw ater to fres hw ater, if dilution is gradual (e.g. C ummings 1 955). M cFarland (1 965) s howed that the fish regulated s erum ion concentration, mus cle ion concentration, and osmolarity about as w ell w hen living in fres hw ater as w hen in a hypersaline s olution. A lthough aw are of a poss ible influence by temperature, M cFarland did not s tudy that relations hip. Also not s tudied in relation to temperature w as the prolactin concentration of the mul- letÕspituitary gland (Sage 1 973). Sage found that this hormone, one that apparently acts by reducing s urface permeability [see review by Johns on (1 973)], w as involved both w ith initial and long-term adaptation of M. cephalus from s eaw ater to fres h- w ater. Suppres s ion and alteration of ion-osmoregulatory mechanisms in fishes at low temperatures has been discus s ed (M orris 1 960 ; Hous ton 1 973). A n alteration in acid-base balance can als o accompany cold-shock (Boucher-Firley 1 935). Poss ibly, in the kill reported here, at leas t a breakdown of the ion-osmoregulatory mechanism took place. I do not know how long individual mullet inhabited the fres hw ater bay- Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISH C O LD-KILLI N M ISSISSIPPI 34 1

ous off Bluff C reek before the fish-kill, bu t the fish w ere devoid of many ectopara- s ites common in brackish areas, and a large group resided there continuous ly for s everal w eeks. Immigration takes place, s ince a few mullet reinhabited the area by January 24. T here w as n o reason to s us pect low concentrations of diss olved oxygen during the cold period, es pecially w ith the rich s upply of oxygen in the gills of dying fish and the pres ence of live fres hw ater fishes. Res ults from s triped mullet in T exas s uggest that even if the oxygen concentration decreas ed, the fish could readily make appropriate phys iological and behavioral adjus tments to the hypoxic environments (C echand Wohls chlag 1 973). O n the other hand, diss ociation of any linked s ys tems, s uch as ventilation and circulation, and the s hut-down of any one of them, w ill lead to death (Roberts 1 973).

Acclimation Lack of thermal acclimation w ould s eem to be more res pons ible for the reported deaths than low s alinity. Both, however, are interrelated as has been s uggested earlier and w ill be explained later. A cclimation, or the conditioning of an aquatic poikilotherm to a particular s tate by its previous thermal history, allows the animal to s urvive almost any low temperature. Brett (1 960 ) s tated ÒAll s pecies to date, endemic to the U nited States and C anada, given adequate acclimation, have s hown the ability to tolerate temperatures in the region of the freezing point of fres h w ater. Rapidly ons etting low temperature, however, can cons titute one of the great- es t threats to s urvival w hen fish are accus tomed to relatively w arm w ater becaus e of the inherent s low rate of acclimation to low temperature.Ó T he lowes t temperature encountered in this s tudy apparently preceded proper acclimation. In addition to gradual reduction in temperature allowing for proper acclimation, it probably als o allows es tuarine fishes to leave unfavorable areas s uch as low- s aline habitats. T he rate of temperature decrease illus trated in Fig. 1 8 for the period preceding the kill had been s urpas s ed during the previous month, bu t the lowes t temperature had not. Fish, however, remained in near-freezing w ater longer in December. Lack of proper acclimation probably determines w hy mass mortalities occur more frequently in T exas and Florida than in M iss iss ippi. Figure 1 9 illus trates the difference in temperatures betw een Biloxi, M iss iss ippi, and C orpus C hristi, T exas . Fishes in M iss iss ippi living in w ater normally cooler than in T exas are necess arily acclimated to lower temperatures . C ons equently, a s udden drop to near-freezing levels w ould affect those fishes less . T he inability of tropical marine animals n ot accus tomed to low temperatures to s urvive or function during s uch periods w as discus s ed as long ago as 1 91 4 (M ayer 1 9 14). Doudoroff (1 942) s tudied the effects of acclimation and low temperatures on marine fishes and defined tw o s tages referred to as primary and s econdary chill- comas. A t low temperatures , fishes almost immediately ceased all movements and died from primary chill-coma, bu t could recover from the initial s hock at less extreme temperatures . A fter s ome hours or days , these s hocked fish again s howed dis- tres s and died from s econdary chill-coma. Doudoroff (1 945) could delay the lethal Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

342 OVERSTREET

effect at the s econdary coma point by placing the fish in an isosmotic s olution, apparently relieving a malfunctioning ion-osmoregulatory mechanism. Brett (1 956) s uggested a combination of three caus es of death for young s almon in cold s aline s olutions s lightly hypertonic to their blood: disturbance of the central nervous sys- tem, an ups et osmotic balance, and a delayed unknown caus e. U ps et osmotic balance, as mentioned earlier, w ould s eem to be involved in the die-off, or Òsecondary chill-coma,ÕÕ of the fishes in this s tudy w hich died 1 to 3 days following the ons et of ice formation on the w aterÕs s urface. M uch of the mys tery of acclimation and the effect of temperature on it is ans w ered by an understanding of biochemical adaptations . Hochachka and Somero (1 971 ) review this topic. In brief, for any given enzyme, the s tabilization of its activity res ulting from increases in the enzyme-subs trate affinity w ith dropping temperature occurs only over a certain range of temperatures . T w o or more variants of a given enzyme, if acting together, can then promote thermally independent enzymic functions over a w ider range of temperatures . A pparently, the crucial pro- cess in cold-acclimation is the biosynthes is of enzyme variants, either poly- morphisms or isozymes , better adapted for catalys is at low temperatures than the primary enzyme in the s ame individual. T he ability for the organismÕs complex adaptive respons es to function may als o depend on the rate and extent of fluctua- tions of environmental parameters in addition to temperature, s uch as diss olved gases and s alinity. Side11 et al. (1 973) meas ured the time course of acclimation from 1 5¡C to 5¡C and to 25°C for tw o compens ating enzymes , and found those enzymes in s hort goldfish reached a s teady s tate quicker than in large ones . In a s tudy on metabolic rates , M orris (1 965) found body-size had a s ignificant effect on res piration rate of acclimated yellow bullhead, bu t that there w ere no differences by s ize in the ability to acclimate. T o my knowledge, all members of every euryhaline s pecies in the pres ent s tudy died. With marginal temperature-stres s , however, mortalities differentiate by length. G unter (1 947) obs erved large individual Anchoa mitchi2li (V alenciennes) and Menidia belyllinn (C ope) having more difficulty s urviving a natural cold- s hock in T exas than s mall fish. M orris (1 962) clarified this problem by meas uring oxygen cons umption of Aequidens portalegrensis, a native Brazilian cichlid. Stimu- lation w ith a cold-shock following cold-acclimation caus ed depress ed res piration in fish larger than 6.0 gm only.

Nutrition A ll fish examined during the period of the kill, both living and dead, poss ess ed empty s tomachs and intes tines . Becaus e of this and s ince their gall bladders w ere greatly distended w ith bile w hich leaked throug hout much of the liver tiss ue and often on adjacent visceral tiss ues, I as s ume the fish w ere under s ome s tres s . Whether s uch s tres s affects the lethal thermal thres hold is unknown. U nder average temperature conditions , however, fish beyond postlarval s tages can w iths tand long periods of s tarvation. As illus trated by Fig. 2, s exually ripe mullet occas ionally s top feeding during non-w inter months long enough to caus e distended gall bladders and partially discolored livers. E ven though I have often obs erved carnivorous fishes w ith empty digestive tracts, adult mullet, w hich feed mostly on plant material and s ediments, typically have material in the intes tine. T his obs ervation is based on examination of Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISHC O LD-KILLIN M ISSISSIPPI 343

s everal hundred s pecimens for the pres ence of paras ites and on comments by commercial fishermen. Frozen s ections of livers from the dying mullet revealed cons iderable amounts of lipid, a finding not limited to fish encountering low temperatures . T he amount of permanent damage, if any, to the hepatocytes is unknown. O ther than poss ess ing relatively large vacuoles, they s eemed normal. Lipid accumulation in hepatocytes often follows periods of s tarvation, poison- ing, or extreme s tres s (Hinton et al. 1 973). E xperimentally, it can be caus ed by feeding animals diets high in fat, low in protein, and deficient in lipotropic s ub- s tances s uch as choline and lecithin. Deficiencies in vitamins A, B,, B6, C , E , and nicotinic acid have all been implicated in the condition. Dixon (1 973) review ed the literature concerning s olubilization and cellular fatty change, including a brief discus s ion on the relations hip betw een the change and temperature. Since mullet killed in the w inter w ere s tarved, w ere osmotically s tres s ed, and contained high quantities of ins ecticides , excess ive fatty globulation in hepatocytes could be expected. U s ing lipid s tains , one can be reas onably s ure vacuolation in histological s ections s ignifies lipids rather than glycogen. Fish that had lipid-positive livers w ere either dying or s us pected upon capture of containing copious fat from the abnor- mally greasy texture of their flesh and viscera. V acuolation in livers of other fish w as probably caus ed by glycogen that had been extracted during preparation of the s ections as reported by Leske and M ayersbach (1 969). Hinton et al. (1 972), w ho positively identified glycogen as the primary caus e of vacuoles in livers of largemouth bass , and others pointed out the need to critically differentiate caus es of vacuolation. New el1 (1 970 :437) s ummarized the biochemical changes in liver of cold- adapted fish as extremely complex w ith increases in glycolys is, glyconeogenesis, glycogen s ynthes is, lipogenesis, and hexose monophosphate s hunt participation. Both lipid and glycogen s torage products may vary according to temperature, s ea- s on, composition and amount of food cons umed, reproductive cycle, s ex, and age.

T he liver and s keletal mus cle s erve as the main s ites of lipid s torage in fish, in contras t to adipose tiss ues in mammals (Bilins ki 1 ,969). Bilins ki generalized that the liver acts as the s torage s ite for s uch reserves in s luggish, bottom-dw elling fishes, w hereas s keletal mus cle plays this role in more active s pecies. T he s triped mullet, however, from this s tudy and additional obs ervations appears to utilize adipose tis- s ue as w ell as liver and mus cle for s torage. Such s torage appears s eas onal and s easonal variation in the amount of oil pres ent in many fishes from the northern G ulf of M exico is marked (T homps on 1 966).

T he concentration of glycogen in liver als o varies s eas onally and w ith res pect to nutritional s tate. Sw ift (1 955) inves tigated s eas onal variation in food reserves of brown trout. I n livers of rats, at leas t, the concentration of glycogen fluctuates cyclically on a daily basis and this rhythm varies throug hout the year (Leske and M ayersbach 1 969). Black et al. (1 966) reported a 38% reduction in liver-glycogen for rainbow trout s tarved 84 hours w hen compared to controls . In hatchery-trout deprived of food for 2 w eeks, glycogen and vacuolation disappeared (Simon et al. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

344 OVERSTREET

1 967). T he apparent abs ence of glycogen in the dying mullet could be becaus e of a s tarved condition. G rowth of adult s triped mullet practically ceases during midw inter w here s tudied in northw es tem Florida (Broadhead 1 953:23) and along the s outhern North A merican A tlantic coas t (A nderson 1 958). G rowth may als o decreas e during the mulletÕs s paw ning period. Pres umably during thes e periods , mullet utilize their re- s erves, s uggesting that the pres ence or abs ence of vacuolation in hepatocytes repres ents a normal s eas onal phenomenon.

M ullet als o us e their reserves during other periods . C ummings (1 955) main- tained s triped mullet in a fres hw ater pool for 3 to 4 months after trans fer from s eaw ater. He then divided them into groups of ÒhealthyÓ and ÒfeebleÓ s pecimens and for each individual meas ured its blood for concentration of chloride. T hose ÒfeebleÓ fish had cons iderably less chloride and als o had s maller livers w ith corres - pondingly distended gall bladders. T hey appeared nutritionally deprived.

A need to s tudy s eas onally large numbers of livers from male and female s triped mullet, in addition to understanding the s torage of lipid and glycogen, is apparent to determine the reas on for the pres ence of the inflammation as s ociated w ith many hepatic vess els . A re the various w hite blood cells as s ociated w ith a pathological condition, caus ed by a natural s tate preceding or immediately following death, or related to haematopoies is? E osinophils indicate the pres ence of foreign proteins and lymphocytes indicate a chronic condition. Necrosis of liver cells adjacent to vess els are known to be caus ed by aquatic pollutants (Hinton et al. 1 973).

Pesticides Levels of pes ticide res idues , es pecially that of endrin, appear rather high in the mullet tes ted; these ins ecticides probably decreas ed the mulletÕs res istance to the effects of low temperatures . T he values for both endrin and combined metabolites of DDT in livers of fish from the kill did s urpas s those of apparently healthy fish in M ay. T hose for aged and fres h tiss ue in M ay agreed fairly w ell. Little meaningful data exist for levels of pes ticides in es tuarine fishes including mullet. Hans en and Wils on (1 970 ) reported DDT res idues for homogenized w hole es tuarine fishes , other than mullet, near Pens acola, Florida, as rarely exceed- ing 0 .1 ppm, except in fishes from the lower es tuary in s ummer and fall w hen the amount of DDT and its metabolites reached 1 .3 ppm. A t leas t part of that residue w as acquired during a s praying program. O ther pes ticides , including endrin, did n ot exceed 0 .0 2 ppm. A lfred J. Wils on, Jr. (personal communication) analyzed s everal apparently healthy s triped mullet from Jolly Bay near Panama C ity, Florida, for DDT and its metabolites . He obtained the following values recorded in ppm from pooled liver, the organ tes ted in this s tudy: DDE = 0 .30 , DDD=0 .34, and DDT= 0 .28. A ll registered lower than for mullet in M iss iss ippi. C orres ponding values for residues in mus cle w ere less , 0 .1 4, 0 .0 7, and 0 .0 3 ppm. T iss ues of s torage s uch as liver parenchyma and fat typically incorporate higher levels of pes ticides than other tiss ues, excepting poss ibly nervous tiss ues, g ut, gills, and blood (e.g. Lane and Livings ton 1 970 ). Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FrSH C O LD-KILL IN MISSISSIPPI 345

C roker and Wilson (1 965) s tudied the effects of 0 .2 pounds of DDT applied per acre of a tidal marsh near Pens acola. Small mullet, 25 to 50 mm long died first. Some pooled homogenated mullet in this length-group n ot killed by the pes ticide accumulated 1 2.33 ppm DDT, and others 100 to 1 75 mm long poss ess ed u p to 39.24 ppm. O ther mullet died after accumulating far less pes ticide. When 1 pound per acre of dieldrin w as applied to a s alt marsh in St. Lucie C ounty, Florida, all individuals of s everal s pecies died. T he s triped mullet, by far the most common fish pres ent w ith the poss ible exception of the s maller G ulf killifish, res ponded first to the toxic s timulus (Harrington and Bidlingmayer 1 958). E ndrin, found in high concentration in my s tudy, w as found to be the most toxic of 1 0 chlorinated hydrocarbon ins ecticides s tudied by Henderson et al. (1 959). Butler (1 963) listed the concentration of numerous pes ticides in s eaw ater necess ary to kill 50 % of tes ted w hite mullet in 24- and 48-hour periods .

T he amount of pes ticide that a fish can tolerate depends on numerous variables. T hese include genetic tolerance and environmental history. G enetically res istant s tocks of s everal fishes are known. Some populations of mosquitofish, Gambusia affinis (Baird and G irard), resist over 1 0 0 times the level of endrin killing s us ceptible individuals (Fergus on et al. 1 966). E nvironmental and biological parameters affect the rate, amount, and means w hich pes ticides are pres ented to, accumulated by, and released from the individual. Pes ticides can be acquired from food or w ater either quickly or s lowly. Influenced by individual and s pecies-specific biochemistry, phys iology, and behavior, as w ell as s tage of maturity and age, accumulation in the fish varies. Different organs res pond differently to pes ticides , and the pres ence of additional pes ticides or chemicals may caus e s ynergistic effects. T emperature, s alinity, and, perhaps , diss olved gas may influence uptake and deposition of pesticides. A fish w ith a high level of pes ticide in its fat and a low level in its mus cle may be a healthy fish until it is s tarved, at w hich time large quantities of pes ticide are released into the blood s tream, killing or s everely s tres s ing the fish. T he levels reported here from mullet cons equently have little s ignificance other than that cons iderable endrin and DDT persisted in the liver. If the deaths had been caus ed entirely from pes ticides , residue levels probably w ould have been higher, behavior of dying fish w ould have differed, and other fishes , either in the s ame or adjacent areas, w ould have als o been affected.

G overnmental agencies need to perfect values for maximal permiss ible concentrations of residues in fish cons idered fit for human cons umption. M ount (1 967) points out that levels , unacceptable by health s tandards , in edible portions of fishes may res ult from exposure by fishes to concentrations in w ater that do not directly harm them. Butler (1 969) pres ents a good review of the s ignificance of DDT residues in es tuarine faunas.

Parasites Paras ites of many kinds infect internally or infes t externally the s triped mullet in M iss iss ippi. Both mullet and other s pecies , however, examined during the kill w ere depauperate of paras ites . O n many of the mullet from Paige Bayou, a leech occurred and, although it removes blood from the fish, too few had fed to caus e much s tress . I n addition, noninfes ted fish died as readily as infes ted ones. A Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

346 OVERSTREET

trypanosome, poss ibly trans mitted by the leech, occurred in low numbers in the blood. E ven heavy infections of trypanosomes in fishes s eldom s everely harm their hosts. Trichodina s p., the ciliate, occurred on the gills in s mall numbers. T his and related s pecies often infes t local mullet in enormous numbers, both on the gills and s kin. Related s pecies caus e mortalities in fishes , bu t us ually confined larval or young fishes s uch as cultured channel and blue catfishes are most s erious ly diseased. Post- metamorphosed flatfish in hatcheries have died from ciliate infes tations ( h rd om and Howard 1 971 ); however, w hen mortalities occur there is us ually an as s ociation of the ciliate w ith a monogenetic trematode (Pearse 1 972). T he s piny-headed w orm Roridosentis elongatus infected too few mullet to be involved w ith mortalities . T he impoverished paras itic fauna s uggests that it in no w ay affected the hostsÕ health. A mlacher (1 961 :257), however, pointed out that fishes s uffering from fatty degeneration are more s ens itive than are healthy ones to infectious diseases. T he lack of paras ites is pres umably becaus e the fish had been in fres hw ater long enough to caus e osmotic or reproductive s tres s in the ectoparas ites w ith res ulting s loughing. A ls o, there w ould be n o introduction of additional ecto- and endoparas ites acquired from brackish intermediate hosts. C ommon, bu t cons picuous ly abs ent paras ites of local mullet, include hemiurid, monorchiid, haploporid, haplosplanchnid, hetero- phyid, and monogenetic trematodes , an as caridoid nematode, ergasilid copepods , an argulid, myxosporideans , and ciliates. O ther paras ites infect this host, bu t are less commonly obs erved in M iss iss ippi.

Hypotheses by Local Residents A ll interview ed long-time res idents of the Bluff C reek area s eparated mortalities happening during cold periods from those caus ed by depletion of diss olved oxygen during w arm periods . Several, however, s till cons idered limited oxygen as the reas on for the reported deaths . Res idents remembered mas s -mortalities in the pas t, but cons idered the pres ent kill as one of the w orst. A ccording to a few individuals , the las t tw o occurred in 1 966 and 1 969. Since mullet died exclus ively w hile other fishes lived, most res idents attributed deaths to s pecific behavioral traits of the mullet. Some as cribed the caus e as ruptured or frozen gill filaments from their characteristic behavior of jumping out of the w ater. O thers s uggested that the fish fed on s omething disagreeable s uch as an algae or other matter available only during cold periods or that paras ites killed them. I discount all the hypothes es . Fish had enough oxygen and their gills appeared healthy. Becaus e the mullet had n ot fed, they probably did n ot feed on toxic s ub- s tances . T oxins in the w ater s hould als o kill or affect other fishes. Paras ites infected few er hosts in s maller numbers than expected and n o unus ual s pecies w ere pres ent. I obs erved n o les ions or s igns indicative of bacterial or viral diseases.

ACKNOWLEDGMENTS I w ould like to acknowledge the help of many people w ho gave us eful infor- mation for this s tudy: res idents near the mas s -mortality, James M dlette of the M iss iss ippi G ame and Fish C ommiss ion, s ports fishermen, and commercial fishermen, especially David Boothe, w ho all s howed an interes t in the problem. T echnical Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISH C O LD-K ILLIN M ISSISSIPPI 347

ass istance w as provided by E dw ard C . Whatley, Jr., Ronnie G . Palmer, and T heresa A nn St. A ndrie of the G ulf C oast Res earch Laboratory. In addition, others at the Laboratory helped: Harold D. Hows e, William E. Haw kins , and Rosemary C heek s tained histological s ections ; William W. Walker analyzed material for chlorinated hydrocarbon ins ecticides ; M ary M arshall determined values for chlorosity and calcium; and Richard Waller identified the s hiner. R oy T . Saw yer of the C ollege of C harles ton verified my identification of the leech and A lfred J. Wils on, Jr. of the E nvironmental Protection A gency, G ulf Breeze, Florida, provided values for pes ticides in mullet from Florida.

LITE RA TURE C IT E D

A M LA C HER, E RWIN 1 961 . T extbook of Fish Diseases, T. F. H. Publications , Jersey C ity. T rans lated from the G erman by D. A . C onroy and R. L. Herman. 30 2 p. A NDE RSO N, WILLIA M W. 1 958. Larval development, growth, and s paw ning of s triped mullet (Mugil cephalus) along the s outh A tlantic coas t of the U nited States . Fishery Bull. Fish Wildl. Serv. U . S. 144:50 1-519. BANG S, OUTRA M 1 895. T he pres ent s tanding of the Florida manatee, Trichecus lafirosfris (Harlan), in Indian River w aters. A m. Nat. 29:783-787. BILINSK I, E . 1 969. Lipid catabolism in fish mus cle. I n Fish in Res earch, (editors, 0 . W. Heuhaus and J. E . Halver), A cademic Press , New York. pp. 1 35-152. BLA C K, E DG A R C ., N. J. BOSOM WORTH and G E O RG INA E . DOC HERTY 1 966. C ombined effect of s tarvation and s evere exercise on glycogen metabolism of rainbow trout, Salmo gairdneri. J. Fish. Res. Bd C an. 23(9):1 461 -1463. BOU C HE R-FIRLE Y, S. 1 935. Recherches biochimiques s ur les te)le)oste)ensapodes (anguilla, congre, murkne). A nnls Ins t. O cganogr., M onaco 1 5:21 7-327. BRE T T , J. R. 1 956. Some principles in the thermal requirements of fishes. Q . Rev. Biol. 31(2):75-87. 1 960 . T hermal requirements of fish-three decades of s tudy, 1 940 -1970 . I n Biological Problems in Water Pollution, T rans actions of the 1 959 Seminar (editor, C . M. T arzw ell), Public Health Service, Robert A . T aft Sanitary E ngineering C enter Tech. Rept. W60 -3. pp. 1 1 0 -1 1 7. BRO A DHE A D, G O RDON C . 1 953. Investigations of the black mullet, Mugil-. cephalus L., in northw es t Florida. Florida Bd. C ons . T ech. Ser. 7:l-34. BUT LE R, PHILIP A . 1 963. C ommercial fisheries investigations . In Pesticide-Wildlife Studies : A Review of Fish and Wildlife Service Inves tigations During 1 961 -62. Fish Wildl. Serv. U . S., C ir. 1 67:ll-25. 1 969. T he s ignificance of DDT res idues in es tuarine fauna. I n C hemical Fallout, (editors, M . W. M iller and G . G . Berg), C harles C . T homas , Springfield. pp. 20 5-220 . C A M E RON, JAM E S N. 1 970 . T he influence of environmental variables on the hematology of pinfish (Lagodon rhomboides) and s triped mullet (Mugil cephalus). C omp. Biochem. Phys iol. 32: 1 75-1 92. C E C H, JOSE PH J., JR. and DONA LD E . WOHLSC HLA G 1 973. Res piratory res pons es of the s triped mullet, Mugil cephalus (L.) to hypoxic conditions . J. Fish Biol. 5(4):421 -428. C HRIST M A S, J. Y. 1 973. C ooperative G ulf of M exico E s tuarine Inventory and Study, M iss iss ippi. Phase I. A rea Description. G ulf C oas t Res earch Laboratory, O cean Springs , M iss . pp. 1 -71. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

348 OVE RSTRE E T

C RO K E R, RO BERT A . and A LFRE D J. WILSON 1 965. K inetics and effects of DDT in a tidal marsh ditch. T rans . Am. Fish. SO C .94(2): 1 52-1 59. C U M M ING S. E . G . 1 955.'T he relation of the mullet (Mu@ cephalus) to the w ater and s alts of its environment: s tructural and phys iological as pects. Ph.D. diss ertation, North C arolina State C ollege, Raleigh. 1 0 3 p. DA HLBERG , M IC HA E L D. and FRE DE RIC K G . SM ITH 1 970 . M ortality of es tuarine animals due to cold on the G eorgia coas t. E cology 51 (5): 931 -933. DIXO N, K E NDA L C . 1 973. Solubilization and cellular fatty change. Histochem. J. 5(4):363-387. DOU DORO FF, PE T E R 1 942. T he res istance and acclimatization of marine fishes to temperature changes. I. E xperiments w ith CirelZa nigricans (A yres ). Biol. Bull. 83(2):21 9-244. 1 945. T he res istance and acclimatization of marine fishes to temperature changes. 11. E xperiments w ith Fundulus and Atherinops. Biol. Bull. 88(2):1 94-20 6. FERG U SO N, DE NZE L E ., J. LA RRY LU DKE and G E O RG E G . M URPHY 1 966. Dynamics of endrin uptake and release by res istant and s us ceptible s trains of mosquitofish. T rans . Am. Fish. SO C .95(4):335-344. FINC H, RU Y H. 1 91 7. Fish killed by the cold w ave of February 2-4,1 91 7, in Florida. M on. Weath. Rev. A pril, 1 91 7:1 71-173. G A LLO WA Y, J. C . 1 941 . Lethal effect of the cold w inter of 1 939-40 on marine fishes at K ey West, Florida. C opeia 1 941 (2):1 1 8-11 9. G UNTE R, G ORDON 1 941 . Death of fishes due to cold on the T exas coas t, January, 1 940 . E cology 22(2): 20 3 - 20 8. 1 945. Studies of the marine fishes of T exas . Publs Ins t. M ar. Sd. U niv. Tex. 1 (1 ):1 -190 . 1 947. Differential rate of death for large and s mall fishes caus ed by hard cold w aves. Science 1 0 6(2759):472. 1 952. T he import of catas trophic mortalities for marine fisheries along the T exas coas t. J. Wildl. M gmt 1 6(1 ):63-69. G U NT E R, G O RDON and H. H. HILDE BRA ND 1 951 . Des truction of fishes and other organisms on the s outh T exas coas t by the cold w ave of January 28-February 3,1 951 . E cology 32(4):731-736. HA NSE N, DA V ID J. and A LFRE D J. WILSO N, JR. 1 970 . Res idues in fish, w ildlife and es tuaries . Pes tic. M onit. J. 4(2):51 -56. HA RRING TON, ROBERT W., JR. and WILLIA M L. BIDLING M A YER 1 958. E ffects of dieldrin on fiihes and invertebrates of a s alt marsh. J. Wildl. M gmt 22(1 ):76-82. HE NDE RSO N, C RO SWE LL, Q . H. PIC KE RING and C . M . T A RZWE LL 1 959. Relative toxicity of ten chlorinated hydrocarbon ins ecticides to four s pecies of fish. T rans . Am. Fish. SO C .88(1 ):23-32. HINT O N, D. E ., M . W. K E NDA LL and B. B. SILV E R 1 973. U s e of histologic and histochemical as s es s ments in the prognosis of the effects of aquatic pollutants. Biological M ethods for the A s s ess ment of Water Q uality, A . S. T. M . ST P 528~ 1 94-20 8. HINT O N, D. E ., RO BE RT L. SNIPE S and M IC HA E L W. K E NDA LL 1 972. M orphology and enzyme histochemistry in the liver of largemouth bas s (Micropterus salmoides). J. Fish. Res , Bd C an. 29(5):531-534. HOC HA C HKA , P. W. and G . N. SO M E RO 1 971 . Biochemical adaptation to the environment. In Fish Phys iology, (editors, W. S. Hoar and D. J. Randall), A cademic Press , New York. pp. 99-156. HOU ST O N, A RT HUR H. 1 973. E nvironmental temperature and the body fluid s ys tem of the teleost. In Res pons es of Fish to E nvironmental C hanges (editor, W. C havin), C harles C . T homas , Spring- field. pp. 87-1 62. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

FISH C O LD-KILLIN M ISSISSIPPI 349

JOHNSO N, DONA LD W. 1 973. E ndocrine control of hydromineral balance in teleosts. A mer. Zool. 1 3(3):799- 81 8. LA NE , C HA RLE S E . and ROBERT J. LIVING STON 1 970 . Some acute and chronic effects of dieldrin on the s ailfin molly, Poecilk lutipinna. T rans . A m. Fish. SO C .99(3):489-495. LE SK E , R. and H. V. M A YERSBAC H 1 969. T he role of histochemical and biochemical preparation methods for the detection of glycogen. J. Histochem. C ytochem. 1 7(8):527-538. LU NA , LE E G . (editor) 1 960 . M anual of Histologic Staining M ethods of the A rmed Forces Ins titute of Pathol- ogy, 3rd E d. M cG raw -Hill Book C omp., New York. 258 p. M A YER, A LFRE D G . 1 91 4. T he effect of temperature upon tropical marine animals . h bl. C amegie Ins t. Wash., Pap. T ortugas Lab. 6:3-24. M C FARLA ND, W. N. 1 965. T he effect of hypersalinity on Serum and mus cle ion concentrations in the s triped mullet, Mugil cephulus L. Publs Ins t. M ar. Sci. U niv. T ex. 1 0 :1 79-186. M ILLE R, E . M O RTO N 1 940 . M ortality of fishes due to cold on the s outheas t Florida coas t. E cology 21 (3): 420 -421 . M ORRIS, R. 1 960 . G eneral problems of osmoregulation w ith s pecial reference to cyclostomes . Symp. Zool. SOC. Lond. 1 :l-16. M ORRIS, ROBERT W. 1 962. Body s ize and temperature s ens itivity in the cichlid fish, Aequidens portalegrensis (Hens el). A m. Nat. 96(886):35-50 . 1 965. T hermal acclimation of metabolism of the yellow bullhead, Zctuluris nutulis (LeSueur). Phys iol. Zool. 38(3):21 9-227. M O U NT, DONA LD I. 1 967. C ons iderations for acceptable concentrations of pes ticides for fish production. A m. Fish. SOC. Spec. Publ. No. 4:3-6. NE WE LL, R. C . 1 970 . Biology of Intertidal A nimals , A merican E lsevier Publishing C omp., Inc., New York. 555 p. PE A RSE , L. 1 972. A note on a marine trichodinid ciliate paras itic on the s kin of captive flatfish. A quaculture 1 :261 -266. PU RDOM , C . E . and A . E . HOWA RD 1 971. C iliate infes tations : a problem in marine fish farming. J. C ons . Int. E xplor. M er 33(3):51 1 -51 4. RO BE RT S, JOHN L. 1 973. E ffects of thermal s tres s on gill ventilation and heart rate in fishes .In Res pons es of Fish to E nvironmental C hanges (editor, W. C havin), C harles C . T homas , Spring- field. pp 64-86. SA G E , M A RTIN 1 973. T he relations hip betw een the pituitary content of prolactin and blood s odium levels in mullet (Mugil cephulus) trans ferred from s ea w ater to fres h w ater. C ontr. M ar. Sci. U niv. T ex. 1 7:1 63-167. SIDE LL, BRU C E D., F. RA Y WILSO N, JE FFRE Y HA ZE L and C . L. PRO SSE R 1 973. T ime course of thermal acclimation in goldfish. J. C omp. Phys iol. 84(2): 1 1 9-127. SIM ON, RA YMOND C ., A LE XA NDER M . DOLLA R and E DWA RD A . SM U C K LE R 1 967. Des criptive clas s ification on normal and altered histology of trout livers. I n T rout Hepatoma Res earch C onference Papers. U . S. Dep. Health E duc. Welfare; U . S. Dep. Inter. Res. Rep. 70. pp. 1 8-28. STORE Y, M A RG A RE T 1 937. T he relation betw een normal range and mortality of fishes due to cold at Sanibel Is land, Florida. E cology 1 8(1 ): 10 -26. Overstreet in Gulf Coast Research Reports (1974) 4(3). Copyright 1974, Gulf Coast Research Laboratory. Used by permission.

350 OVE RSTRE E T

STORE Y, VA RG A RE T and E . W. G U DG E R 1936. M ortality of fishes due to cold at Sanibel Is land, Florida, 1886-1 936. E cology 1 7(4): 640- 648. ST RIC K LA ND, J. D. H. and T . R. PA RSO NS 1968. A practical handbook of s eaw ater analys is. Bull. Fish. Res. Bd C an. 167:l-311. SWIFT, D. R. 1955. Seasonal variations in the growth rate, thyroid gland activity, and food reserves of brown trout (Salmo trutta Linn.). J. E xp. Biol. 32:751-764. T A BB, D. C ., D. L. BUBRO W and R. B. M A NNING 1962. T he ecology of northern Florida Bay and adjacent es tuaries . Florida Bd. C ons . T ech. Ser. 39: 1-81. TA BB, D. C . and R. B. M A NNING 1961. A checklist of the flora and fauna of northern Florida Bay and adjacent brackish w aters of the Florida mainland collected during the period July, 1957 through September, 1960. Bull. M ar. Sci. G ulf C aribb. 11(4):552-649. T HOM PSO N, M A RY H. 1966. Proximate composition of G ulf of M exico indus trial fish. Fishery Indus t. Res. 3(2):29-67. WE LLS, HA RRY W., M A RY JA NE WE LLS and I. E . G RA Y 1961. Winter fish mortality in Pamlico Sound, North C arolina. E cology 42(1):217-219. WILLC O X, JOSE PH 1887. Fish killed by cold along the G ulf of M exico and coas t of Florida. Bull. U S. Fish C ommn 6:123.