Abstract

The New England Aquarium is currently engaged in a captive breeding program for bluefin tuna (Thunnus thynnus). We have maintained wild caught tuna in captivity for the past two years. These fish are housed in a semi-closed 63,000 gallon recirculating sea water system and are fed a diet of previously frozen, vitamin-supplemented mackerel, squid, herring and capelin. Although these animals appear to adapt quickly to a captive environment and gain weight very rapidly, all animals develop a head deformity commonly termed "puffy snout." In addition some animals develop ocular lesions consisting of lenticular opacities with or without gas in and around the posterior aspect of the globe.

Gross postmortem, standard histology and three-dimensional Computer Aided Tomography (CAT) scan examinations of affected individuals have been compared to wild-caught normal animals. "Puffy snout" is characterized by an exaggerated connective tissue growth over most of the head region. In addition, CAT scans have revealed changes in skull bone morphology, density and cortical thickness in the captive animals as compared with wild-caught individuals. Theories of potential etiologies will be discussed.

Introduction

The fishes that comprise the family scombridae include the various species of mackerel and tuna. In general, this is a rather homogeneous group consisting of swift, predaceous, open water animals. Of the fishes in this group, the mighty bluefin tuna is perhaps the best known and most important commercially. Atlantic bluefin tuna (Thunnus thynnus) are known to travel thousands of miles during their yearly migrations, and to reach well over 500 kilograms as adults. The eastern Atlantic populations migrate up and down the east coast of the U.S. pursuing a great diversity of prey items.

Captive rearing attempts have taken two foci; public display and commercial rearing operations. Several aquariums in Japan have been moderately successful in exhibiting small bluefin tuna. Because of their high commercial value and increasingly limited supplies, commercial farming is being attempted in several countries. The New England Aquarium is currently undertaking a program to develop captive rearing techniques for Atlantic bluefin tuna. In addition, we are examining the feasibility of captive propagation for maintaining U.S. bluefin fisheries in a competitive position in the face of tuna rearing programs being developed in the Pacific and Mediterranean.

A common pathology, often referred to as puffy snout, was reported from nearly every institution housing bluefin tuna and related species of scombrids for research or display. We have begun to characterize puffy snout using four investigative techniques: gross observation and video recording of affected animals, gross necropsy and tissue sample collection for histological investigation, Computer Aided Tomography scans (CAT scans) of the skulls of affected animals, and wild (normal) specimens, and finally, blood samples were collected from both wild and captive individuals. The results of the blood value comparisons will be presented in an accompanying paper.

Materials and Methods

Capture, Transport and Temporary Holding

During the summer months of 1992 and 1993, small populations of juvenile bluefin tuna were collected off the coast of Cape Cod, Massachusetts (1992) and Wachapreague, Virginia (1992,1993) by hook and line. The tuna collected varied in size and weight, straight fork lengths ranged from 60 to 75 cm. Initial weights were from a weight of S to 13 kg at capture. Fish collected in Massachusetts were caught offshore and transported back to Woods Hole, Massachusetts via an onboard 700 gallon open water transport tank. When they arrived in Woods hole they were transferred and held in a 20 by 40 foot floating pen moored just offshore. The animals were later transported back to the New England Aquarium in Boston, Massachusetts using a cargo truck outfitted with a recirculating 700 gallon water system. Fish collected in Virginia were handled in 2 ways. In 1992 the majority of fish were held for varying times in the boat holding tank and released. Two fish were transferred to the transport tank and driven directly back to Boston. In 1993 the fish were caught offshore and transferred to a custom designed holding system located in Wachapreague, Virginia. The custom holding system featured an outdoor 12,000 gallon, twenty foot diameter, round fiberglass tank outfitted with two high pressure rapid sand filters and a large capacity titanium chiller unit. Water for the system was drawn from the surrounding estuary, filtered and recirculated. Despite our best attempts, the punishing heat and great volume of food added to the tank daily made it very difficult to maintain good water quality. Temperature fluctuations, coupled with high nitrogen levels, were a constant problem and were dealt with by performing frequent water changes and the use of a commercially available ammonia binder. In late summer, the tuna held in Virginia were transported back to the Aquarium in Boston in pairs via a refrigerated cargo truck with our 700 gallon recirculating transport system on board. Many of these specimens suffered capture and transport related injuries consisting of abrasions and minor lacerations, many specimens exhibited symptoms consistent with poor water quality.

Permanent Holding

The tuna were housed in an outdoor, 63,000 gallon, 20 by 40 foot rectangular painted concrete holding tank using a semi closed water circulation system. Water was drawn in from Boston harbor, filtered for particulates through two sets of high pressure sand filters and recirculated in the pool. An ozone system replete with a mazzei injector and a degassing chamber complete the system. During the summer months, the pool temperature was maintained at or near 19 C by periodically exchanging system water with newly filtered or sometimes raw harbor water. During the winter months, the temperature was maintained by the use of raw steam injected into the influent water lines. The tank was illuminated at night by four 150 watt floodlights to provide background lighting of the walls. Additionally, there were four high intensity 400 watt mercury halide floodlights that were used to simulate a 12 hour photo period during the winter months. Despite our attempt to regulate the animals photo period, it was likely that the tuna experienced an unnaturally short day length during the winter months.

Nutrition

Food intakes varied per individual, but a group of 6-20 kg fish would typically eat 5-6 kg of food per day. The fish were fed a variety of previously frozen food items consisting primarily of capelin (Mallotus villosus), silversides (Menidia menidia), smelt (Osmerus mordox), squid (Illex illecebrosus and Loligo pealed), mackerel (Scomber scombrus) and herring (Cluepea harengus). Food items were supplemented with a combination of multivitamins and minerals. Animals were typically fed until satiated 3 times per day.

Gross Observations

Video documentation of the development of puffy snout was performed in the captive specimens over time. The tuna were filmed on a semi-regular basis (approximately every week, weather permitting) using a Sony Hi8 compact video camera (model TR8 1) with a Sony underwater housing model 75M super marine pack. The Hi8 tapes were later recorded onto an SVHS VCR system for frame by frame analysis. Still photos were also taken at irregular intervals using a variety of high speed films and a Nikons 4 underwater camera. Ambient lighting was used in all cases; flashes and bright lights appeared to disturb the fish.

Post Mortem and Histological Examination

Bone and soft tissue samples were taken from both captive and wild specimens from variety of sites around the skull. A variety of bone sample sites were selected for comparison from around the head and axial skeleton. All samples were placed in 10% neutral buffered formalin immediately after collection. Samples of bone and connective tissue of both wild and captive type samples were sent to The University of Guelph Department of Pathology for histological examination. The samples containing bone were then transferred to a decalcifying fixative for approximately 30 days. All samples were then sectioned (paraffin embedded, 6 µm) and stained with hematoxylin and eosin in preparation for examination.

Computer Aided Tomography Scans

CAT scans offer a non invasive method of quantifying differences in both hard and soft tissues. One pair of tuna (I captive and I wild specimen) were taken to Massachusetts Eye and Ear Infirmary and another pair (also I captive and I wild) were taken to Tufts University Veterinary School Radiology Department for CAT scan examinations. Scans were taken at using a Siemens Somatom+ CAT scan machine at intervals ranging from 1-8 mm sections. Tufts University scans were done using a Technicare 2010 CAT scan machine. Hard copies of the scans were produced from both institutions and later digitized using a Hewlett Packard II cx 24 bit color scanner at 600 dpi. (gray scale). The CAT scan outputs of each pair of animals were produced using the same window, which ensures a consistent scan output. Because different techniques were used to produce the hard copy images at the two institutions, images produced at Tufts could not be directly compared to those produced at Massachusetts Eye and Ear Infirmary.

CAT scans were used to compare two features of selected bone in wild and captive bluefin tuna. The first was area measurements of selected bones. This was done using a commercially available imaging software (Adobe Photoshop version 2.5. 1). Areas were selected using the magic wand option using a tolerance of 25. The second feature examined was bone density, also with Photoshop. Units of area were measured in pixel counts. Bone density was compared using a 256 gray scale, with 0 representing black (least dense) and 255 representing white (most dense). Histograms were produced illustrating the distribution of density over a selected area of the frontal bones (Figures 6a-7b), The histograms were produced by summing all the pixels at a particular shade of gray. Bone densities of selected sites were measured on the Tufts animals using the standard Houndsfield units of measure.

Results

Gross Observations

Within six months in captivity the animals had more than doubled in size and within 3 months after capture changes in head morphology were readily apparent in some individuals. Changes in morphology varied by individual, with some specimens showing gross changes (see Figures 1-4) while others were relatively normal even after extended periods. Figure I was taken of an outwardly normal individual in late February, 1994. Figures 2-4 were taken of one individual at approximately 4-6 week intervals beginning in October, 1993. Figure 5 was taken in February just prior to the animals necropsy. Both animals were caught within 2 weeks of each other and were held and transported similarly. In the sampled animals the development of puffy snout was not consistent; in some cases it progressed from a focus along the mandible, frequently near the hook site, while in other cases it seemed to develop evenly along the head. In many cases, however, the end result was a head that appeared shortened and compressed from anterior to posterior. Many times eye shields were documented partially or wholly obscuring the eyes. Post cranial deformation was unremarkable with the exception of some curling of the pectoral and pelvic fins.

Figure 1	Figure 2
Figure 3	Figure 4
Figure 5

Histology

The condition was characterized grossly by what appeared to be an uncontrolled growth of connective tissues focused mainly around the head and mandible. Histology revealed that the predominant lesion of the mandible was a diffuse increase in the porosity of the compact bone due to an enlargement of osteocytic lacunae. There was also noted an increased vascularization of the compact bone. These lesions were not found in other sampled bones. Osteoclasts were present in moderate numbers on the medullary surface of compact bone and boney spicules.

The epidermis/dermis samples revealed that in many sampled areas the epidermis was entirely absent, possibly due to erosion, while other samples appeared normal. The basil lamellae was mildly to moderately thickened, with individual variation. In some cases the distinction between spongy and compact dermal layers was obvious, while in other cases this boarder was difficult to differentiate. In areas where spongy dermis was discernible, it appeared looser than on other areas of the cranium.

Sample of the eye shields were submitted and appeared very uniform and were comprised primarily of dermis covered by an attenuated layer of epidermis. Samples taken from immediately around the eye had a similar description. In the second submission, the eye flap was comprised of an epidermal layer of normal thickness, while the dermis had a very loose appearance. There was also evidence of mild lymphocytic perivascular cuffing and a diffuse increase of lymphocytes near the basal layer of the epidermis. The diagnosis based on these samples included osteocytic perilacunar bone resorption in the mandible and cranial dermal hyperplasia.

Other noteworthy histological results include evidence of moderate parasitic gastritis, cataracts, and in I individual sampled, moderate nephrosis.

Computer Aided Tomography (CAT) Scans

Two features of selected bones were quantified. The first feature was a total area measurement of the selected bones. The wild tuna samples had up to 40% more total area of bone (measured in pixels) than the captive specimens in the frontal bones (see Figures 6a -7b). Furthermore, the distribution of the pixels in the wild specimen was skewed further to the right (more dense) on the grayscale than the captive specimens. The percent of dense bone (greater than 145 on the gray scale) in the wild specimens was up to 9 times greater than a comparable bone site (frontal bone) in the captive animals. Additionally, bone density measurements were taken of the frontal bones as well as the parasphenoid (in Houndsfield units); in every case the density of comparable bones in the wild animals exceeded that found in the captive specimens.

Figure 6a	Figure 6b
Figure 7a	Figure 7b

Discussion

Post mortem and histological examinations revealed little evidence for infectious disease, with the exception of parasitic gastritis. Additionally, no evidence exists of a trauma related injury that may account for this unusual condition.

Two main factors contribute to the gross changes that outwardly characterize puffy snout. Changes include an externe overgrowth of connective tissue, especially around the eyes, and in an overall morphological change in the head shape, often leading to an apparent shortening and thickening of the head. In many cases, the connective tissue overgrowth is first observed along the mandibular surface and just anterior to the eye, often producing shields in front of the eye. In extreme cases, the connective tissue will completely cover the eye. As the connective tissues enlarge around the lower jaw, normal mandibular function can be impaired. These two factors often lead to feeding difficulties which may exacerbate pathologies already in progress.

The results of connective tissue sampling were somewhat disappointing because of inconsistent sampling sites. In some cases the epidermis was eroded away, in other samples it was intact. In some cases the distinction between spongy and compact dermal layers was obvious, in others cases it was not. The variability of the results makes it difficult to reach any conclusions based on this small sample set, however, further investigations into soft tissue analysis will undoubtedly prove fertile.

Two main changes were described in the bone. The first was a decrease in the total area of bones at selected sites; captive reared animals had bones that had smaller cross sectional areas when compared with equivalently sized wild caught animals. The second feature we examined was the change in bone density between wild caught and captive tuna. The captive animals have bones that are up to 9 times less dense than wild caught specimens of a similar size. A third feature of this unusual condition which we examined, but did not quantify, was the changes in the geometry of the affected bones. This may be a major contributing factor in the gross morphological changes that are observed.

The CAT scans are particularly informative because they illustrate the changes in cephalic shape, the changes in cross sectional area of bones and changes in bone density and geometry. Although CAT scans can provide an unparalleled vantage from which to view changes in bones, this investigative technique is not without its shortcomings. An apparent artifact may have affected the results from one of the CAT scans. In 2 cases, the area of selected bones (parasphenoid and frontal) was greater in the wild specimens as opposed to the captive animals. An exception was found in the dentary samples. Although they were sampled from approximately the same size animals the gross changes in morphology resulted in a sampling of two incongruous sites. In this instance the site selected was the point where the premaxillary and dentary bones first appear in together in a single scan. Because of the apparent shortening of the premaxillary bone and the apparent enlargement of the dentary bone, the absolute position where these two bones are both visible in a scan is different on the two specimens. In the captive animal, the scan site was posterior to the site sampled in the wild specimen, and correspondingly of a greater thickness.

Although this is preliminary study of a small sample of animals, the results cannot be dismissed. The phenomenon of puffy snout is a quantifiable, reproducible pathology. Indeed, the condition persists in some animals currently housed at the New England Aquarium and research into the causative agents of this unusual syndrome continues.

Acknowledgments

I would like to extend my sincerest thanks to the following people for their invaluable contributions to this paper: Karen Johnson for all her help with the Tufts CAT scans, Dr. D. MacPHEE for the excellent histological work, Leah Faretra, and Joshua Singer for their help filming, gathering and sorting the data, Carl Howard and Renato Perelli for their help with video and photography, Bruce Wyman for his patience and technical assistance, and Karen Cowan for her understanding and proofing assistance.