FISHERY BULLETIN: VOL. 82, NO. 2 



prior to death. Only teeth and intestinal contents 

 were sampled. For comparative bacteriological 

 studies, several osteichthyan fish were caught by 

 rod and reel or retrieved from the gill net men- 

 tioned above. Some elasmobranchs were occasion- 

 ally maintained in large concrete or fiberglass 

 tanks containing seawater piped from Sarasota 

 Bay. All fish examined from tanks were either 

 alive or had been dead for less than 1 h. Fish 

 caught in the Gulf of Mexico were either iced if 

 dead or kept in a wet hold until examined, which 

 was routinely less than 1 h. In one case (see below) 

 sharks were dead for 3 h before sampling. Water 

 samples were collected from either Sarasota Bay 

 or the tanks containing fish. 



Quantitative Analysis 



Swabbing was compared initially with two 

 other methods involving the use of membrane fil- 

 ters in a quantitative sampling of elasmobranch 

 skin. In the swabbing technique, a sterile alumi- 

 num foil template containing a 3.1 x 3.1 cm square 

 opening (9.6 cm 2 ) was placed on the skin, on the 

 side of each fish just posterior to the gills. A sterile 

 polyester-tip swab (Falcon No. 2069 3 ) was used 

 over the exposed area in all directions, and the tip 

 was broken off in a screw-capped tube containing 

 10 ml of sterile seawater. Decimal dilutions of this 

 were prepared in 9 ml of sterile seawater and 0.1 

 ml volumes spread (Buck and Cleverdon 1960) on 

 Bacto-Marine Agar (Difco Laboratories, Detroit, 

 Mich.). One procedure with membrane filters in- 

 volved placing sterile 0.45 /xm gridded membranes 

 (Millipore No. HAWG047SO) on shark skin and 

 pressing them down by rolling a sterile glass rod 

 across the membrane. The membrane was then 

 placed grid uppermost on the surface of an agar 

 plate. The second membrane filter technique was 

 similar to the first except that the membrane, 

 after exposure to the skin, was placed in a sterile 

 plastic screw-cap centrifuge tube with sterile sea- 

 water and agitated on a vortex mixer for 30 s. 

 Decimal dilutions and platings were then made as 

 indicated above. All plates were incubated at room 

 temperature (24°-26°C) for 3-5 d. Counts/cm 2 of 

 skin were calculated. 



Qualitative Analysis 

 In addition to skin, other body areas including 



3 Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



teeth, gills, and intestinal contents were sampled 

 by use of a swab. The upper third of plates of 

 eosin-methylene blue, tryptic-soy, brain heart in- 

 fusion, and marine agar (all Difco) was swabbed, 

 and the remaining two portions of the plate were 

 streaked sequentially with a sterile wire loop to 

 isolate colonies. Plates were incubated at room 

 temperature (24°-26°C) and 37°C for 1-5 d and 

 colonies were selected based on differences in 

 morphology. 



Qualitative changes in the bacterial flora on 

 shark flesh were assessed as a function of time and 

 temperature. Pieces of nurse shark, Ginglymos- 

 toma cirratum , flesh (about 2 cm 2 ) were cut asepti- 

 cally from one area of one side of the fish after the 

 skin had been removed. These pieces were placed 

 in sterile petri dishes and incubated at room tem- 

 perature (24°-26°C) and 5°C. Initially and after 3, 

 4, 5, and 7 d incubation, the surface of the flesh was 

 sampled using a sterile loop, which was used to 

 directly inoculate agar plates to obtain well- 

 isolated colonies. 



Tank and bay samples were collected at a depth 

 of about 30 cm in sterile bottles. One ml volumes 

 were diluted in 9 ml of sterile seawater and addi- 

 tional decimal dilutions prepared in a similar 

 manner. Spread plates (see above) were made on 

 marine agar. Representatives of various colonial 

 types were selected and identified after incubation 

 for 3-5 d at room temperature. 



All isolates were maintained on slants of either 

 marine agar or tryptic-soy agar. Gram reactions 

 were recorded by both conventional staining and 

 the KOH technique (Buck 1982). Gram negative 

 enteric bacteria were identified using either the 

 Enterotube II (Roche Diagnostics, Nutley, N.J.) or 

 API 20E ( Analytab Products, Plainview, N.Y.) sys- 

 tems. Hemolysis was detected on tryptic-soy agar 

 containing 5% horse blood. Other bacteria were 

 characterized using the methods of Shewan et al. 

 (1960) and Oliver (1982). 



RESULTS AND DISCUSSION 



Quantitative Analysis 



Counts by the swab technique averaged 115% 

 higher than those obtained by membrane filters 

 applied directly to agar plates (two experiments) 

 and 910% higher than counts by agitating the 

 membrane in seawater followed by dilution and 

 plating (three experiments). All subsequent 

 counts of skin bacteria on both elasmobranch and 

 osteichthyan fish were made using the swab 



376 



