About Bioline  All Journals  Testimonials  Membership  News  Donations

Ichthyological Bulletin
J.L.B. Smith Institute of Ichthyology
ISSN: 0073-4381
Num. 63, 1995
J.L.B. Smith Institute of Ichthyology
Ichthyological Bulletin No.63, May 1995

T.G. Andrew^1, T. Hecht^1, P.C. Heemstra^2 and J.R.E. Lutjeharms^3

^1 Department of Ichthyology and Fisheries Science, Rhodes University, Grahamstown, South Africa.
^2 J.L.B. Smith Institute of Ichthyology.
^3 Department of Oceanography, University of Cape Town.

Code Number: FB95002
Size of Files:
    Text: 204.3K
    Graphics: line drawings (gif) - 183.8K
              photographs (jpg) - 529.4K


Recent collections of fishes from the South Atlantic islands of the Tristan da Cunha Group (Tristan, Inaccessible and Nightingale) and Gough Island have produced 25 new records. Fifty one species are known to occur in the near-shore waters of the islands (46 of these are documented by specimens and another 5 species are probable). A diagnosis, synonymy and in most cases, an illustration, are provided for each species. For certain species, brief notes on biology, relative abundance and seasonal distribution are included. The neritic ichthyofauna of the Subtropical Convergence (STC) region is characterized, and the importance of this frontal zone as a barrier to dispersal of species in the Southern Ocean is evaluated. Analysis of the zoogeographic affinities of the Tristan/Gough fish fauna resulted in a redefinition of the West Wind Drift Islands Province (WWDI), comprising the Tristan Group and Gough in the South Atlantic and St. Paul and Amsterdam Islands in the southern Indian Ocean. Vema Seamount (west of Cape Town) and Walters Shoal (south of Madagascar) are excluded from the WWDI Province, as the affinities of their poorly-known fish faunas appear to be more with South Africa and the tropical Indian Ocean respectively.


The south Atlantic islands known as the Tristan Group and Gough Island were formed by volcanic activity that resulted in massive cones thrusting up from the outer slopes of the Mid-Atlantic Ridge (Heydom and Lutjeharms, 1980). The Tristan Group comprises Tristan da Cunha, which is the largest island (about 13 km in diameter) and the two much smaller islands, Inaccessible and Nightingale. Tristan is about midway between South Africa and South America, at 37 degrees 05'S, 12 degrees 17'W, 1660 nautical miles from Cape Town (Fig. 1 ). The two smaller islands are about 20 miles south- west of Tristan. Gough Island lies 200 miles SSE of Tristan at 40 degrees 19'S, 9 degrees 56'W.

    Figure 1. The position of Tristan da Cunha and Gough island in the South Atlantic Ocean.

Little has been reported on the diversity and distribution of the neritic fishes of the Tristan da Cunha Group and Gough. Publications concerning the fishes of these islands have generally been descriptive in nature or have dealt with broad zoogeographical concepts. In the first published report on the fauna and flora of Tristan, Carmichael (1819) described four species of fish and mentioned another five that occur there. Barnard (1923; 1925) listed five more species. Norman (1935a) described an additional species, and Sivertsen (1945) added seven species that were collected by the Norwegian expedition to the islands. Rowan and Rowan (1955) added three oceanic species to the fish fauna of Tristan. Penrith (1967) reviewed the fishes of Tristan, Gough Island and the Vema Seamount northwest of Cape Town; he listed 23 species from Tristan, 8 from Gough and 13 from Vema, and compared the ichthyofauna of these localities with those from other areas of similar latitude. Miller (1982) collected three additional pelagic species from the waters around Gough Island, and Gon and Heemstra (1987) added another species.

Several species of midwater fishes are known from near the islands. Krefft (1974, 1978) reported 16 species from the Tristan/Gough region, and Hulley (1981 ) listed another 24 species collected by the FRV WALTER HERWIG during March 1971. Including midwater fishes, 68 species were previously known from these southern mid-Atlantic islands. We here report new records of 25 additional species.

    Figure 2. Percentage composition by mass of the eight most common species caught in the Tristan da Cunha fishery from October 1988 to March 1990.

The islanders depend on the fish resource to a large extent for food and for bait for the local rocklobster industry. Approximately 61.5 tons of linefish are harvested each year for these purposes. Figure 2 indicates the percentage composition by mass of the eight main species caught in the fishery from October 1989 to March 1990 (figures from present study).

The insular nature of the island ecosystem renders it vulnerable to over-exploitation from commercial fishing operations. Most of the fish species are bound to the islands for the completion of their life cycles; hence, the populations are more or less isolated and not supplemented by recruits from outside of the system. At present, only the palinurid lobster Jasus tristani is commercially exploited. The catch per unit effort for this species has decreased considerably since 1965 (Roscoe, 1979). If this industry were to collapse, there would be a strong incentive to exploit the fish stocks commercially. A great danger of over-exploitation of the fish resource exists in the future, if early management is not implemented. For this reason it is essential to obtain a thorough knowledge of the composition and dynamics of the fish fauna of the islands.

    Figure 3. The Tristan da Cunha Group with Gough Island inset showing the 183 m isobath around the islands (after Roscoe, 1979).

Being volcanic and situated on the Mid-Atlantic ridge, all of the islands have relatively narrow shelf areas (mostly about 1 nautical mile wide) which drop off into abyssal depths. Depths of up to 3000 m have been recorded between the northern islands. Roscoe (1979) estimated the planar areas around the islands down to the 183 m isobath (Fig. 3), which is about the maximum depth sampled for neritic fishes in the present study. The waters around the islands can be considered oceanic, because they are virtually unaffected by processes derived from the islands, such as silt deposition and fresh water run-off.

The inshore areas of the four islands differ considerably, although all are characterized by a rugged sea floor topography. Tristan has many boulder and sandy beaches that are not as common at the other islands, and there are many shallow reef areas with deeply incised rockpools. The slope of the littoral bottom at Tristan is more gentle than at the other islands. At Gough and Nightingale the drop-off is almost vertical from the shore down to about 10 m. Gough has few large rockpools. Gough and especially Nightingale have deeply incised coastlines, with many bays and offshore pinnacles. All of the islands are surrounded by a kelp (Macrocystis pyrifera) fringe occurring in depths of 10-40 m. The coastline of Inaccessible is more similar to that at Tristan than to the other islands, and it has few deeply incised features.

Knox (1960) stated that the Subtropical Convergence (STC) lies just north of Gough in summer and north of Tristan in winter. Relatively few physical and chemical measurements of this front have been undertaken in the central South Atlantic. In order to obtain a better understanding of the marine environment at the islands various oceanographic measurements in the inshore areas around the islands were undertaken during an 18-month period in the field.

Daily sea surface temperatures and salinities were measured from October 1988 to March 1990. Surface temperatures were measured using a Crawford bucket (Crawford, 1972) on the seaward side of the harbour breakwater at Tristan Island. Surface water samples were collected at the same locality. These samples were sent to Cape Town every three months where salinities were determined in parts per thousand (ppt) to an accuracy of 0.003 ppt, using a Beckman Model SR 9 salinometer.

Offshore temperature profiles were measured, on average, once a month off Tristan Island during the 18 month study period. Sampling periodicity was entirely dependent on sea conditions. A portable expendable bathythermograph (XBT) recorder and launcher where operated from an 8 m diesel powerboat or from the 35 m rock lobster vessel MFV TRISTANIA II. Care was taken to ensure that launching occurred in depths greater than 300 m as the XBT was capable of measuring a profile to this depth. The position, surface water temperature, salinity and weather conditions at each launch were recorded. These depths were usually obtained at a distance greater than 1 km from the shore.

Quality controls of the temperature profile recordings were carried out in the manner described in Lutjeharms (1985). In some cases traces were discarded due to spikes in the record, and in others only the trace above the spikes was used.

The STC forms the northern border of the Southern Ocean and the geographic limit to the influence of Subantarctic Surface Water. At and near the sea surface it consists of a strong horizontal gradient in temperature, salinity and a number of nutrients such as nitrate, nitrite and phosphate. These gradients are particularly well developed along the east coast of South America (Guretskiy, 1986) and south of the Brazil and the Agulhas Currents respectively (Peterson and Stramma, 1991).

According to Deacon (1933) the STC in the South Atlantic is marked by a surface temperature change of at least 4 C and a salinity change of at least 0.50 psu. Lutjeharms and Valentine (1984) have carried out a detailed analysis of the surface expression of the STC based on 70 crossings of the front, mostly south of Africa. They have established that the average width of the front here exceeds 200 km, that its temperatures span 11 C to 18 C on average at any one time and that it has a mean temperature, in the core of the front, of 14.2 C.

    Figure 4. The location of the STC in the South Atlantic Ocean in 1988, according to satellite thermal infrared imagery. Panel (a) shows a compilation of all frontal expressions from January to June 1988. Panel (b) gives the locations of the 15 C surface isotherm from June to August. Panel (c) gives the same for the 18 C surface isotherm for the summer months December, January and February. The Tristan da Cunha Group is indicated by the open circle.

Evidence that the islands fall within the meridional range of the STC can be seen in Figure 4. These data are derived from weekly products based on thermal infrared measurements made by satellite (Meeuwis, 1991). The locations of all distinct fronts for the period January to June 1988 are shown in Figure 4a. Figure 4b shows the locations of the 15 C surface isotherm which, according to Deacon (1966), represents the northern border of the STC in winter. Figure 4c shows the location of the 18 C isotherm in summer. All three diagrams indicate that Tristan da Cunha lies in the centre of the general meridional band in which the STC occurs.

    Figure 5. Mean monthly water temperatures at Tristan and Gough Island.

Monthly average temperatures calculated for all the available sea surface temperature data from Tristan da Cunha and Gough Island also suggest that these islands lie within the STC (Fig. 5). Temperatures at Tristan in the warmer six months lie between 15 C and 19 C, and in the colder six months between 13 C and 15 C. Temperatures at Gough Island are on average 3 C lower than at Tristan during all months. If Deacon's (1966) definition of the STC is used, this would place the Tristan da Cunha group at the northern edge of the temperature range that the STC occupies and Gough Island at the southern border or entirely south of an STC influence. Monthly standard deviations are large, usually about 2 C (Fig. 5) from which it can be concluded that the seasonal temperature cycles at both Tristan da Cunha and Gough Island may be considerably different from year to year.

According to the range of salinities reported for the STC by Lutjeharms (1985) all the measurements taken during the present study indicate that the Tristan da Cunha group lies within the STC. Figure 6 shows salinity values from this study as compared to salinities at the STC given by Whitworth and Nowlin (1987). The mean salinity proposed by these authors at the northern edge of the STC is indicated by the broken line. Therefore, according to their definition all the samples taken at Tristan Island during this period were of Subtropical Surface Water, that is from north of the STC. The salinity measurements from this study do not clearly indicate any meridional movement of the STC on a seasonal basis. This is in contradiction of the view held by Knox (1960).

    Figure 6. Offshore sea surface salinities measured in the vicinity of the Tristan da Cunha group from 1988 to 1990. The broken line denotes the saline limit of Subtropical Surface Water according to Whitworth & Nowlin (1987).

    Figure 7. Profiles for ocean temperature to 400 m in the vicinity of the Tristan da Cunha group. ]he date when each profile was measured can be read off by comparing the numbered location above the top axis with the months on the bottom axis. Each numbered location corresponds to the 15 C isotherm on the temperature scale. The location of the 10 C isotherm at 300 m depth is shown as a step for each temperature trace that extended that deep.

Figure 7 shows the results of the measurement of offshore temperature profiles over the 18 month period. Two distinctive processes may be inferred from this figure. Firstly, a seasonal warming of the upper 50 m is evident. This was particularly clear for Station 8, taken on 13 February 1989 when the surface temperature was 21.0 C and the temperature at 50 m was 14.5 C. In winter the surface layer had cooled down to a uniform 13.3 C (Station 15). The second process is the event-controlled, vertical mixing of heat into the water column. This is evident from the step-like structure of a number of temperature traces (eg. Stations 5, 9, 13, 14). All the heat absorbed in the upper 50 m in February is mixed to 100 m by March (Station 9), to 200 m by August (Station 14) and to 300 m by September (Station 15). Whereas the temperature at 300 m had stayed more or less constant at 11 C, in August it suddenly increased to 12 C while the surface temperature had decreased to 13.2 C. By 17 October (Station 17) the top 80 m was again 0.5 C warmer with a distinct step at that depth showing the vertical penetration of the mixing process. By 23 November (Station 20) the top 40 m had again become warmer than 15 C.

Considered in conjunction with the findings of Lutjeharms and McQuaid (1986) who investigated changes in the thermal structure of Southern Ocean fronts with the onset of summer, the results presented here are quite significant. As summer progresses, surface water becomes increasingly warmer. This is due to insolation and not due to the overflow of warmer water from the north, in which case the salinity would also have risen. An increase in salinity was not evident from the present study (Fig. 6). Since the water masses below 100 m are not directly involved in this heating, they remain unchanged and a strong stratification between the upper 100 m and the water below is established. This has been observed previously (Lutjeharms and McQuaid, 1986). With the onset of winter storms this upper, warmer layer is mixed out. Another possible mixing mechanism would be convective overturning as the sea surface cools.

The effect at the sea surface of this seasonal process might be an apparent southward shift of the STC during the summer months. In fact, the strongest horizontal thermal gradient might be found further south during the summer. The saline front will most probably not follow this thermal front, nor will the nutrient expression, or the major subsurface thermal expression of the front. For biological activity in the frontal zone this relative immobility of the major part of the STC itself is probably not as important as an increase in stratification at the front during the summer. Fournier et al. (1979) found that changes in the inclination of the Shelf-Break Front off Nova Scotia associated with varying degrees of mixing and stratification of the water column resulted in different phytoplankton standing stocks. They suggested that the steepness of the front influenced the depth to which phytoplankton were mixed which in turn controlled the amount of illumination received by a given cell. Therefore at times when the water column was well stratified (decreased frontal inclination) illumination available to phytoplankton would be at a maximum resulting in the highest production. Similarly, Marra et al. (1990) proposed that enhanced phytoplankton growth at the thermohaline Shelf-Break Front in the Middle Atlantic Bight could be explained by nutrient enrichment from turbid underlying water to organisms in clear well-lighted surface water. Allanson et al. (1981 ) hypothesised that increased primary productivity at the STC during the summer was mainly a result of an increase in stratification of the water column at this time. The layer of warm surface water resulting from this stratification in the summer probably extends south of the nutrient expression of the STC in the Tristan da Cunha region, and would therefore substantially increase the potential primary production of the frontal zone during summer by overlying an area of high nutrients which would otherwise be poorly accessible to organisms preferring warmer water.

This hypothetical mechanism would be in full agreement with all presently available data. It would also agree with the traditional perception of the fishermen of Tristan da Cunha that catches improve during the summer due to changes in the marine environment at this time. Increased primary production during this period may attract pelagic species into the vicinity of the islands during the summer months as suggested by an increase in catch per unit of effort in the local line fishery (data from present study). This hypothesis would not require the major subsurface expression of the STC to meander meridionally on a seasonal basis.

From a biogeographical perspective it may therefore be concluded that, although the STC is more diffuse in the mid-Atlantic than, for example south of Africa, the Tristan da Cunha group and Gough Island fall within its zone of influence. This may be through cast-off eddies (Lutjeharms and Valentine, 1988), filaments and other protrusions throughout the year. It is also likely that the surface expression of the STC retreats southward with the onset of summer and with increased vertical stratification in the water column at Tristan da Cunha. Such increased stratification may lead to increased primary productivity at this time of the year and therefore temporally affect the biology and distribution of fishes in the STC region.


Most of the species reported in this paper were collected by one of the authors (TGA) during the 18-month stay on Tristan da Cunha. Collections were made on each fishing day when the rock lobster fleet (usually 18-22 two-man powerboats) put to sea. Fish were obtained in lobster traps and with hook and line in depths of 50 - 200 m. Collecting was also done with hook and line from the shore and with rotenone ichthyocide in the intertidal and shallow sub-tidal region (max. depth 10 m). Stranded specimens were also collected, and one specimen was collected after being dropped by an Antarctic tern (Sterna vittata). Extensive underwater observations were carried out with SCUBA, especially at Tristan Island, to study fish behaviour and assess some of the collecting methods.

The fishes of Nightingale, Inaccessible and Gough Islands were sampled with hook and line and lobster traps from the MFV TRISTANIA II. Fishes were also collected from the shore at Nightingale by one of us (TGA) and at Gough by a member of the 1988/1989 South African Weather Team (R. Matthews).

During the 18-month period over 4000 specimens were examined in order to obtain biological information. Representative specimens of most species were preserved and are lodged in the collection of the J.L.B. Smith Institute of Ichthyology. Ten species that were reported from Tristan by previous authors, that were not collected during the present study, are also listed below.


The neritic fishes and some mesopelagic species that were collected within one mile of the islands are listed below, with brief notes on the biology and ecology of some species. Many species have local common names that differ from those recognised elsewhere. In these cases, the species name is followed by the internationally recognised common name and the Tristan common name is given in parentheses. The synonymies given here comprise the original nominal species and other synonyms or references that are pertinent to the taxonomy of these species. Sizes of specimens examined are generally listed as total length (TL) although standard length (SL) and fork length (FL) are used in some cases. Detailed investigations on some of the more common species have been published elsewhere. RUSI catalogue numbers are given for voucher specimens housed in the J.L.B. Smith Institute of Ichthyology. SAM catalogue numbers are given for specimens examined at the South African Museum in Cape Town.



Notorynchus cepedianus, broadnose sevengill shark, (rock shark), Fig. 8

    Figure 8. Nototynchus cepedianus, 209 cm mature female, South Africa (from Bass et al., 1986).

Squalus cepedianus Peron, 1807:337, (Tasmania).
Notorynchus cepedianus:
Compagno, et al., 1989; Bass et al., 1986a.

DIAGNOSIS: One dorsal fin; 7 pairs of gill slits; 6 rows of large teeth on each side of lower jaw; horizontal diameter of eye much less than distance between nostrils and 1.6-3.0 times in least depth of caudal peduncle; no inner (nictitating) lower eyelid; snout length 1.1-1.3 times internostril distance and less than distance between lower ends of first gill slits.

Colour: Body pale grey dorsally, white below; dorsal surface of body and fins speckled with black spots.

MATERIAL: Tristan: RUSI 35991, 139 cm TL (jaws); RUSI 35992, 145 cm TL (jaws); RUSI 35995, 167 cm TL (jaws); RUSI 35997, 188 cm TL (jaws). In addition, numerous specimens (120 to 225 cm) were examined at the islands.

REMARKS: Although it has not been reported from these waters before, the sevengill shark is the most common shark on the shelf areas of Gough and the Tristan Group islands. Notorynchus cepedianus was almost certainly mistaken as "Hexanchus sp."in Roscoe's (1979) record, as the sixgill shark is extremely rare in the depths fished by the local fishermen; however, the sevengill shark is commonly caught and used for bait. The local Notorynchus population is probably self sustaining, as all size ranges are found in the region (D. Ebert peps. comm.). The species feeds predominantly on fish and cephalopods.

DISTRIBUTION: The sevengill shark is found worldwide in temperate waters, except in the North Atlantic, and is restricted to relatively shallow coastal waters (Bass et al., 1986a).

Hexanchus griseus, sixgill shark, Fig. 9

    Figure 9. Hexanchus griseus, 115 cm immature male, South Africa (from Bass et al., 1986a).

Squalus griseus Bonnaterre, 1788:9 (Mediterranean).
Hexanchus griseus: Compagno, 1984; Bass et al., 1986a.

DIAGNOSIS: One dorsal fin; 6 gill slits on each side of head; lower jaw with a row of 6 large teeth on each side; mouth length contained 1.6-2.7 times in its width; snout length 4-7% TL, 2.0-2.8 times in mouth width; length of dorsal fin base contained 1.1-1.4 times in distance from dorsal base to origin of upper caudal fin lobe. The species attains a length of 5+ m and a weight of 730+ kg.

Colour: grey.

MATERIAL: Gough: 182 cm TL, 26 kg immature male, caught in 25-35 m, 6 October 1989.

REMARKS: This specimen was examined by Dr D.A. Ebert at the South African Museum in 1990.

DISTRIBUTION: H. griseus is a benthic species that is found at depths of 35-1500 m in all oceans.


Etmopterus cf. princeps, great lanternshark, Fig. 10

    Figure 10. Etmopterus princeps, 780 mm female, Inaccessible Island.

Etmopterus princeps Collett, 1904:3 (Eastern North Atlantic, Faeroe Bank, south of Faeroe Islands; depth 7501200 m); Bigelow et al., 1953; Compagno, 1984.

DIAGNOSIS: Two dorsal fins, both with distinct spines, the origin of first dorsal fin behind tips of pectoral fins by about spiracle length; no anal fin; distance from pelvic fin base to origin of lower caudal fin lobe slightly greater than distance from tip of snout to spiracle and about 1.9 times in distance from pectoral to pelvic fin bases; snout length about 1.5 times in width of head; 5 pairs of gill slits, length of the first 3 subequal to maximum diameter of spiracle and about 1/2 length of eye; upper jaw teeth mostly with 5 cusps, the median cusp much the largest but slender, the outermost cusps minute; lower jaw teeth with cusps directed laterally; lateral trunk denticles thorn-like, with moderately high, nearly erect, slender and slightly curved, conical tips; denticles fairly widely-spaced but most not arranged in longitudinal rows (weakly formed rows on sides of caudal peduncle); head with conspicuous lines of pores (openings for ampullae of Lorenzini?) and a transverse row across the chest joining the pectoral fin origins. Vertebrae: monospondylous 47, precaudal diplospondylous 17, caudal 26, total 90.

Colour: Uniformly dark blackish-brown; no apparent dark markings on the flanks or ventral surface of the body except for a faint black horizontal bar on the lower caudal peduncle just anterior to lower caudal origin; no pale spot between the eyes.

MATERIAL: Inaccessible: SAM 33071, 780 mm TL; caught in a crayfish trap set in 220 m, 3 Dec. 1993 by Capt. E. Stoffberg, of the M/V HEKLA.

REMARKS: We are grateful to Dr L.J.V. Compagno of the South African Museum for bringing this specimen to our attention and providing vertebral counts, photographs and a provisional identification for this shark. This specimen is provisionally identified as Etmopterus princeps, a species previously known from the North Atlantic and recently reported (Nakaya, 1982) from the Kyushu-Palau Ridge in the western Pacific. Dr Compagno has informed us that two specimens of E. princeps from Iceland that he x-rayed have 83 and 86 total vertebrae, fewer teeth in lower jaw, and teeth with higher side cusplets.


Prionace glauca, blue shark, (bottlenose shark), Fig. 11

    Figure 11. Prionace glauca, 279 cm mature male, South Africa (from Bass et al., 1986b)

Squalus glaucus Linnaeus, 1758:235 ('European Ocean").
Glyphis glaucus:
Sivertsen, 1945.
Prionace glauca: Compagno et al., 1989; Bass et al., 1986b.

DIAGNOSIS: Two dorsal fins, the first dorsal just in front of pelvic fins; 5 pairs of gill slits; pectoral fins long and falcate; no spiracles or ridge between dorsal fins; inner (nictitating) lower eyelid present; lip folds minute; teeth unicuspid, strongly serrated and broadly curved in upper jaw; lower jaw teeth narrower.

Colour: Brilliant dark blue above, white below; fades to grey after death.

MATERIAL: Tristan: RUSI 33584, 120 cm TL. Inaccessible: RUSI 35996, 250 cm TL (jaws).

REMARKS: Sivertsen (1945) first reported the presence of this shark in Tristan waters. Blue sharks are caught all the year round at all the islands, from the surface down to at least 150 meters. Our two specimens were caught on the surface at Inaccessible and at 150 m at Tristan Island. These sharks of 2.5 m and 1.2 m suggest that all size ranges are found in these waters. This species is commonly caught, but only smaller specimens are landed and used as bait. The temperature regime at Gough would suggest that the shark reported by fishermen to be common there is the blue shark.

DISTRIBUTION: The blue shark is found around the world in tropical and temperate waters, but is most common in waters of 12 - 16 C (Bass et al, 1986b).


Isurus oxyrinchus, shortfin mako, (bottlenose shark), Fig. 12

    Figure 12. Isurus oxyrinchus, 250 cm female, South Africa (from Bass, 1986)

Isurus oxyrinchus Rafinesque, 1810: 12, Pl. 13 (Sicily). Compagno et al., 1989.

DIAGNOSIS: Two dorsal fins; caudal fin lunate, the lower lobe almost as big as the upper lobe; 5 pairs of large gill slits; origin of first dorsal fin over or behind inner corner of pectoral fin; pectoral fins markedly shorter than distance from snout to pectoral fin origin; teeth lanceolate, without small basal denticles.

Colour: Dark blue above, white below.

MATERIAL: Tristan: RUSI 35994, 267 cm TL (jaws).

REMARKS: A 267 cm TL mako caught on a longline set in over 100 m is the fast record for Tristan. Makos are said to be common in Tristan waters, although seldom landed; they are occasionally seen leaping from the water, and are known to display aggressive behaviour towards boats.

DISTRIBUTION: Worldwide in tropical and temperate waters, both inshore and in the open ocean (Bass, 1986). Reported by Blanc and Paulian (1957:326 as "Isurus glaucus") from St. Paul and Amsterdam Islands.

Carcharodon carcharias, great white shark, Fig. 13

    Figure 13. Carcharodon carcharias, 177 cm immature female, South Africa (from Bass, 1986)

Squalus carcharias Linnaeus, 1758:235 (Europe).
Carcharodon carcharias:
Bass, 1986: 98, Compagno, 1984.

DIAGNOSIS: Two dorsal fins, the first much bigger than the second; origin of first dorsal fin in front of inner corner of pectoral fin; caudal fin lunate; tail with a single, large keel along each side of peduncle (no secondary keels on caudal fin); anal fin origin under or behind axil of second dorsal fin; teeth triangular, flattened, with broad, coarsely serrate cusp (in sharks less than 150 cm, most teeth are lanceolate, smooth-edged, with small basal cusplets).

Colour: Dark grey to brown or black above and white below.

MATERIAL: A 4.3 m FL, 4.9 m TL great white shark was caught at Gough Island in a depth of 51 m on July 1, 1993 by Captain Ben Herwig and Mike Saunders of the MFV TRISTANIA 11. The jaws and dorsal fin of this specimen were examined by Dr Leonard J.V. Compagno of the South African Museum.

REMARKS: Dr Compagno informed us of this capture of a great white shark at Gough Island. This is the first documented occurrence of the species at Gough or the Tristan islands. According to Captain Herwig, the stomach of this shark contained seals and 7 "bluefish" [Hyperoglyphe antarctica].

DISTRIBUTION: Carcharodon carcharias occurs worldwide in tropical and temperate seas.


Torpedo nobiliana, Atlantic electric ray,

Torpedo nobiliana Bonaparte, 1835: Fig. (pages not numbered; Italy); Compagno, 1986; Compagno et al., 1989.

DIAGNOSIS: Head, body and pectoral fins forming a round soft flabby disc which is wider than long; tail short, with two moderate-sized dorsal fins (first distinctly larger than second and placed anterior to rear edge of pelvic fins) and a large triangular caudal fin (caudal fin height about equal to distance from upper origin of caudal to origin of first dorsal fin); no papillae on edge of spiracles.

Colour: Dorsal surface brown, with small irregular blackish spots and specks; underside white. In South African waters, T. nobiliana is described as "uniform shiny black or dark grey above, underside white" (Compagno et al. 1989). Stehmann and Burkel (1984) described northeastern Atlantic specimens as "usually plain dark violet/brown above, sometimes with indistinct darker dots and white spots; underside white to cream, often with dusky margins to disc and pelvic fins".

MATERIAL: Tristan: RUSI 35993, 89 cm TL.

REMARKS: Another Torpedo (SAM 25245) was collected at Tristan in May 1969 and donated by the Division of Sea Fisheries to the South African Museum. Unfortunately, this specimen cannot now be located at the Museum. Our specimen, collected by a Tristan fisherman with hook and line in about 150 m of water, was sent to us by Mr Jimmy Glass. It looks very similar to the Torpedo macneilli (Whitley, 1932) from St Paul Island that was illustrated by Duhamel (1989). It would not surprise us to learn that T. macneilli (described originally from Australia) and T. fairchildi Hutton, 1872 (from New Zealand) are synonyms of T. nobiliana.

DISTRIBUTION: T. nobiliana is known from the Mediterranean and Eastern Atlantic Ocean (from Scotland to South Africa), the western North Atlantic (from Cuba to Nova Scotia) and along the south coast of South Africa to Algoa Bay. A specimen identified as T. nobiliana by B. Seret was reported from the "banc Austral" (34 degrees 54'S, 53 degrees 14'E) by Duhamel (1989).



Bassanago nielseni, hairy conger (eel), Fig. 14

    Figure 14. Bassanago nielseni, 664 mm TL (Nightingale Island); left - upper jaw, right - lower jaw.

Pseudoxenomystax nielseni Kannovskaya, 1990: 8, Figs. 1 c and 2d (Southeastern Pacific: Nazca Ridge, 22 degrees 05'S, 81 degrees 15'W).

DIAGNOSIS: A congrid eel with the tail longer than head and trunk combined; skin covered with minute, slender, darkly-pigmented villi, which are scattered and difficult to see in juveniles, but profuse in adults and give a distinctive "hairy" appearance to the body. Head length 15-19% TL; distance from tip of snout to anus 39-42% TL; predorsal distance 16-19% TL. Eye diameter 13-17% head length; snout fleshy, longer than eye and projecting in front of lower jaw; rear nostrils in front of middle of eye; front nostrils are short tubes directed forward at front of snout; gape extends to below pupil; a fleshy labial flange or lip on each side of lower jaw, but not continued across tip of jaw; no labial flange on upper jaw; cardiform teeth in several rows on jaws; intermaxiilary tooth patch about as broad as long and not separated from maxillary or ethmovomerine teeth; vomer with a band of short conical teeth, the band extending posteriorly about to middle of maxillary tooth band. Lateral line a raised ridge running midlaterally along body, the pores opening via short, white, fleshy tubes along ventral edge of the ridge; 39-43 pores from head to vertical at front of anus; top and sides of head with numerous, short fleshy papillae (superficial neuromasts). Median fins continuous around tip of tail; pectoral fin rays 13-16. Preanal vertebrae (anterior to anus) 42-45 (n= 8); total vertebrae 150-159 (4 specimens with intact tail).

Colour: Brownish dorsally, pale ventrally; median fins dark with pale margin; some specimens distinctly darker than others.

MATERIAL: Tristan: RUSI 33577: 2, 58-60 cm TL; RUSI 33579, 84 cm; RUSI 33580 2, 57-63 cm. Nightingale: RUSI 33578 2, 58-60 cm; RUSI 33581, 66 cm. Inaccessible: RUSI 33604, 61 cm.

REMARKS: Although the hairy conger is the most common eel at Tristan, it has not previously been reported from Tristan or Gough (Contrary to Duhamel, 1989, the species was not reported from Tristan by Penrith, 1967.). All our eels were collected in traps set deeper than 70 m. B. nielseni are often captured in traps overnight, suggesting that they are nocturnal feeders. An adult male captured off Tristan Island in June 1989 was in a ripe running condition, indicating a winter spawning season for the species.

According to David Smith (1989), 'The species treated as Pseudoxenomystax by Castle .... - Congermuraena albescens Barnard 1923, Bassanago bulbiceps Whitley, 1948, and Pseudomystax hirsutus Castle, 1960 - seem quite different from P. dubius Breder [the type of the genus Pseudoxenomystax] and probably should be placed in Bassanago Whitley." Castle (1986) intimated that P. hirsutus (from New Zealand) may be a synonym of B. albescens; and, having examined two specimens of B. bulbiceps from Australia, we believe that this species should also be considered a synonym of albescens. The supposed differences between B. bulbiceps and hirsutus in the numbers of dorsal and anal fin-rays and lateral-fine pores (as given by Castle, 1960) are subsumed in the counts given by Castle (1968) for B. albescens from South Africa.

Specimens of Bassanago from the Tristan Group have fewer vertebrae than B. albescens from South Africa (with 4 5-48 preanal vertebrae [n= 10 ] and 169-174 total vertebrae [7 specimens with intact tails]). We tentatively assign the Tristan species to B. nielseni, which was recently described from the Nazca Ridge in the southeastern Pacific Ocean. According to Karmovskaya (1990), B. nielseni has a total vertebral count of 152-153, which falls within our range for Tristan specimens. The counts of 34-39 vertebrae before the anal fin reported by Karrnovskaya (1990) in B. nielseni are lower than our preanal counts of 42-45 for Tristan specimens. The taxonomic significance of this apparent difference in preanal vertebrae cannot be assessed with the limited material presently available to us.

B. albescens is common off Cape Point, South Africa, and has also been reported from southern Argentina and southern Chile (Nakamura et al., 1986 [as "Pseudoxenomystax albescens"]; Pequeno, 1989). If B. bulbiceps and B. hirsutus are synonyms of B. albescens, the species also occurs at Australia and New Zealand. B. hirsutus and B. bulbiceps were both listed from the Chatham Islands east of New Zealand by Roberts (1991), but he gave no vertebral counts for his specimens, so his identifications are dubious.

DISTRIBUTION: B. nielseni is, at present, known only from the Tristan Group and the Nazca Ridge in the southeastern Pacific Ocean. Duhamel (1989) reported Bassanago albescens from Amsterdam Island, but the vertebral count for this specimen (148) indicates that it is probably B. nielseni, (which was described after Duhamel's paper was published).

Conger wilsoni, conger eel, Fig. 15

Gymnothorax wilsoni Bloch & Schneider, 1801:529 (Australia).
Conger wilsoni:
Castle, 1986.

DIAGNOSIS (data from SAM specimen in parentheses): Tail longer than head and trunk combined; snout longer than eye; rear nostril in front of middle of eye; lips with well developed flange; teeth in 1 or 2 rows on jaws, the outer teeth longer, compressed, forming a continuous cutting edge; dorsal fin origin over or behind pectoral fin tip; a sensory pore behind and in line with comer of mouth; head 7.0-8.2 (8.0) times in SL; lateral-line with 37-40 (37) pores before anus; pectoral fin rays 13-17 (15); vertebrae 140-142 (144).

Colour: Dorsal two-thirds of bead and body dark olive brown, ventral parts creamish white; dorsal and anal fins with black margin.

MATERIAL: Gough Island: SAM 29632, 488 mm TL.

REMARKS: This South African Museum specimen (collected in Tranvaal Bay with hook and line by T. Gosliner and W. Liltred) represents the first record of Conger wilsoni from any of the West Wind Drift islands.

DISTRIBUTION: C. wilsoni is known from South Africa, Australasia, New Zealand and possibly (if C jordani Kanazawa, 1958 is a synonym) Japan. Castle (1986) estimated the larval life of Conger species as "about a year", and this would account for the wide distribution of C. wilsoni.

Gnathophis capensis, southern conger (eel), Fig. 16

    Figure 16. Gnathophis capensis, 290 mm TL, South Africa (from Castle, 1986).

Leptocephalus capensis Kaup, 1856:153 (Cape of Good Hope; larva).
Congermuraena australis
Barnard, 1923:442 (Namibia, Cape Peninsula, False Bay, Tristan da Cunha).
Ariosoma australis:
Penrith, 1967.
Gnathophis capensis:
Castle, 1968; 1986; Duhamel, 1989.

DIAGNOSIS: A congrid eel with the tail considerably longer than head and trunk combined; pectoral fins present; snout slightly longer than eye and projecting in front of lower jaw; lips not well developed; several rows of small teeth on jaws and roof of mouth, those on vomer blunt; no hair-like epidermal papillae on body; lateral-line pores and vertebrae 128-135.

Colour: Body olive brown above, paler below; head silvery; dorsal and anal fins with dusky margin.

MATERIAL: Tristan: RUSI 33627:372 mm; SAM 12782: 186 mm; SAM 24551:374 mm.

REMARKS: G. capensis appears to be rare in Tristan waters compared to Bassanago nielseni. Perhaps the southern conger prefers deeper waters and is therefore more difficult to collect. Our single specimen was collected in a trap set overnight in 150 m.

DISTRIBUTION: G. capensis is known from False Bay to Plettenberg Bay, South Africa (Castle, 1986) and from St. Paul and Amsterdam islands (Duhamel, 1989). According to Castle (1968), the duration of larval life is 10-12 months; during this period, the larvae could easily drift from Tristan to South Africa or from South Africa to St. Paul Island.


Vinciguerria poweriae

Gonostoma poweriae, Cocco, 1838: 167, Pl. 5, Fig. 2 (Messina).
Vinciguerria poweriae:
Schaefer et al., 1986.

DIAGNOSIS: Small, compressed silvery fishes with light organs (photophores) along ventral surface of head and body; total gill-rakers 14-16; dorsal fin rays 13-15, the last ray above or behind anal fin origin; anal fin base not longer than dorsal fin base; anal fin rays 12-14; adipose fin over anal fin; pectoral rays 9-11; branchiostegal rays 8; 12-15 photophores from anal fin origin to caudal fin base; no photophore at tip of lower jaw; one photophore in front, and another behind, lower edge of eye; premaxillary teeth uniserial, with 3 or 4 long teeth and a few smaller ones; lower jaw teeth biserial in front; vomer and palatines toothed; pterygoids and tongue toothless; maxilla with 2 supramaxiliae; vertebrae 38-39.

Colour: Head and body dark dorsally, the flanks silvery; dark streak at tip of lower jaw and over premaxilla; fins colourless.

MATERIAL: Tristan: RUSI 33467:49 mm TL.

REMARKS: Our only specimen was found floating dead on the surface near Tristan Island in December 1989. During the day, this mesopelagic species is found at depths of 300-600 m, and at night it rises to 50-350 m.

DISTRIBUTION: V. poweriae has a circumglobal, subtropical distribution (Schaefer et al, 1986).


Maurolicus inventionis

Maurolicus pennanti (non Walbaum, 1792): 1925: 151. Barnard,
Maurolicus muelleri (non Gmelin, 1789): Klages et al., 1988.
Maurolicus inventionis Parin & Kobyliansky, 1993: 83, Fig. 3b (42 degrees 18'S, 00 degrees 08'W [seamount southeast of Gough Island]).

DIAGNOSIS: Small silvery fishes with light organs (photophores) along ventral surface of head and body; head and body compressed; long low adipose fin above rear end of anal fin; eyes large and round; teeth in jaws minute, uniserial or with a short outer row at front of lower jaw; dorsal fin base mostly in front of anal fin; dorsal rays 10-11; anal rays 22-26; pectoral rays 17-19; gill-rakers (7-9) + 1 + (21-25) total 30-35; vertebrae 33-35; 2 long groups of photophores along ventral edge of tail preceded by a single elevated light organ; a large light organ just behind symphysis of lower jaw.

Colour: Brilliant silvery, the dorsal surface dark greenish blue.

MATERIAL: SAM 23950:12 specimens; locality, "Tristan or Gough Island". None collected by us.

REMARKS: Parin and Kobyliansky (1993) revised the genus Maurolicus and described four new species from the South Atlantic. Barnard (1925) recorded "Maurolicus pennanti (Walbaum)" from Tristan and Pann and Kobyliansky (1993) listed M. pennanti in the synonymy of Maurolicus muelleri (Gmelin, 1789) an eastern North Atlantic species. Klages, et al. (1988) recorded "M. muelleri" from the stomach contents of rockhopper penguins (Eudyptes chrysocome moseleyi) at Gough Island.

DISTRIBUTION: M. inventionis is known only from the Tristan Group, Gough Island and the seamount type-locality southeast of Gough.


Electrona paucirastra

Electrona paucirastra Bolin, in Andriashev, 1962:280 (west of Amsterdam Islands; 39 degrees 30'S, 71 degrees 16'E); Miller, 1982; Hulley, 1986.

DIAGNOSIS (from Hulley, 1986): Mouth terminal; maxilla distinctly expanded posteriorly and reaching to vertical at rear edge of eye; anal fin base longer than dorsal fin base; posterodorsal margin of operculum rounded and without serrations; caudal peduncle depth less than 2.5 times in its length; dorsal fin rays 13-15; anal fin rays 20-21; pectoral fin rays 14-16; gill-rakers (5-6)+(15-18) = 21-24.


REMARKS: Miller (1982) collected this species close to Gough Island in midwater trawls from less than 50 m during October 1980. E. paucirastra is a high-oceanic and mesopelagic species, occurring from the surface to 100 m at night.

DISTRIBUTION: Circumglobal between the STC and the Antarctic Polar Front (Hulley, 1986).

Electrona subaspera

Scopelus subaspera Gunther, 1864:411 (43 degrees 00'S, 123 degrees 00'E off SW Australia).
Electrona subaspera: Miller, 1982; Hulley, 1986.

DIAGNOSIS (from Hulley, 1986): Mouth terminal; maxilia distinctly expanded posteriorly and reaching to vertical at rear edge of eye; anal fin base longer than dorsal fin base; posterodorsal margin of operculum rounded and smooth; caudal peduncle depth less than 2.5 times in its length; dorsal fin rays 13-15; anal fin rays 20-22; pectoral fin rays 13-16; gill-rakers (7-9)+(18-22) = 27-31.


REMARKS: Miller (1982) collected E. subaspera near Gough Island in October 1980. It is a high-oceanic and mesopelagic species, occurring from the surface down to 200 m at night.

DISTRIBUTION: Circumglobal between STC and the Antarctic Polar Front (Hulley, 1986).

Lampadena dea

Lampadena dea Fraser-Bruuner, 1949: 1097, Fig. 1101, Pl. 1A (48 degrees 27'S, 22 degrees 7-10'W [about 1400 km SW of Gough Island]); Hulley, 1981.

DIAGNOSIS: A charcoal black lantern fish with large mouth; maxilla extending well behind eye, not expanded posteriorly; anal fin base slightly shorter than dorsal fin base; dorsal fin origin above or slightly anterior to pelvic fin origin; large, black-edged light organ on dorsal surface of caudal peduncle and another one on ventral surface; smaller photophores arranged on body below lateral line. Dorsal fin rays 13-15; anal fin rays 13-15; pectoral fin rays 13-16; gill-rakers (5-6)+(3-6).

MATERIAL: Inaccessible Id: RUSI 34036:52 mm TL.

REMARKS: One specimen was collected with a net between 80 m and the surface. L. dea is amesopelagic species of the open ocean; it is found at depths of 350-2390 m during the day and 150 m at night.

DISTRIBUTION: Circumglobal in the vicinity of the STC (25 degrees to 49 degrees S latitude; Hulley, 1986).


Gaidropsarus novaezelandiae, comb rockling, Fig. 17

    Figure 17. Gaidropsarus novaezelandiae, 210 mm TL, Tristan.

Motella novazealandiae Hector, 1874: 107, Pl. 18, Fig. 76b (Cape Campbell, New Zealand); Svetovidov, 1986.
Gaidropsarus insularum Sivertsen, 1945: 8, Fig. 6 (Tristan da Cunha); Penrith, 1967; Cohen, 1986; Svetovidov, 1986; Duhamel, 1989.
Gaidropsarus parini
Svetovidov, 1986: 122, Fig. 2 (Nazca Ridge).

DIAGNOSIS: Body elongate, with long low dorsal and anal fins that reach to the rounded caudal fin; the dorsal fin is preceded by a row of short filaments and a longer ray on the nape; pectoral fins with 20-21 rays; 2 barbels on snout (from anterior nostrils) and one on chin; pelvic fins with 7 rays; dorsal fin with 62-64 rays; anal fin with 50-54 rays; teeth in bands on jaws and vomer. Vertebrae 44-48.

Colour: Head, body and fins dark reddish brown, purplish grey ventrally.

MATERIAL: Tristan: RUSI 31484:142 mm TL; RUSI 33473: 3, 145-160 mm; RUSI 33477: 3, 89-140 mm; RUSI 33482: 2, 130-180 mm; RUSI 33478: 6, 98-110 mm; SAM 24930:147 mm SL. Gough: RUSI 31486:193 ram; RUSI 13293:300 mm, collected in 1978 by Gideon Rossouw; RUSI 33601: 3, 270-335 mm. Plus 25 additional RUSI specimens and numerous specimens (31-343 mm) examined at the islands, but not preserved.

REMARKS: Juveniles and small adults (less than 20 cm TL) are common in rockpools and in shallow subtidal areas at all the islands; larger adults are found down to at least 50 m. Prejuveniles (less than 30 mm TL) are silvery and probably pelagic. This cryptic predator feeds on crustacea and small fishes.

The specific distinction of G. insularum, G. novaezelandiae and G. parini is dubious. According to Svetovidov's (1986) key to species of Gaidropsarus, these three species differ in number of dorsal fin-rays (insularum with 66-70 and novaezelandiae with 62-65), pectoral fin-rays (novaezelandiae with 20-21 and parini with 23-24), length of first dorsal fin base, and eye diameter. But these data are based on only 3 specimens of insularum, 10 novaezelandiae, and 2 parini. The dorsal fin-ray counts given in Svetovidov's text for novaezelandiae do not agree with those in his key and overlap the counts that he gives in the text for insularum. Furthermore, Duhamel (1989) has shown that the range in dorsal fin-ray numbers of specimens from St. Paul and Amsterdam (62-76) includes the ranges for all three of these supposedly allopatric species.

DISTRIBUTION: G. novaezelandiae is known from Tristan, Gough, the west coast of South Africa, St. Paul and Amsterdam Islands, Australia, New Zealand, Nazca Ridge, and the Bromley Plateau (31 degrees S, 34 degrees W).


Physiculus karrerae, Fig. 18

    Figure 18. Physiculus karrerae, 237 mm TL, Nightingale Island.

Physiculus karrerae Paulin, 1989: 114, Fig. 10 (off Rio Grande, Brazil).

DIAGNOSIS (from Paulin, 1989; data from our specimens given in parentheses): Body elongate, cylindrical anteriorly, tapering to a narrow caudal peduncle and small separate caudal fin; snout rounded, mouth terminal; dorsal fin divided into a short anterior part of 7-8 (8) rays and a long posterior part of 68-76 (66-69) rays; anal fin with 73-78 (73-78) rays; pectoral fin rays 26-27 (25-27); about 150 vertical scale series from upper end of gill opening to caudal fin base; light organ on belly indicated by round black spot of naked (scaleless) skin in front of anus; diameter of light organ 16-18% of distance from anal fin origin to a interventral-line joining origins of pelvic fins; fight organ closer to interventral-line than to anus; chin barbel present; no teeth on vomer; gill-rakers long and slender, 1-2 + 7-9; vertebrae 56 (58-59).

Colour: head and body mostly uniform reddish brown; abdomen dark bluish black; fins and tips dark brown.

MATERIAL: Nightingale Id: RUSI 33611: 2, 235-237 nun TL. Inaccessible Id: RUSI 33585:11,215-255 mm

REMARKS: Physiculus karrerae is common in the shelf waters from 50 m to at least 150 m at the northern islands, but it was not found at Gough. All specimens were collected in traps baited with fish.

DISTRIBUTION: P. karrerae has a wide anti-tropical distribution in the Atlantic Ocean. It was previously known from Bermuda, the Caribbean, southern Brazil and St. Helena; depth range 250-800 m (Paulin, 1989). Our specimens are the most southerly and the shallowest records for this species.


Cheilopogon pinnatibarbatus, smallhead flyingfish, (flying fish), Fig. 19

    Figure 19. Cheilopogon pinnatibarbatus, 300 mm SL, Mozambique (from Heemstra & Parin, 1986).

?Exocoetus exiliens (non Linnaeus): 494. Carmichael, 1819:
Exocoetus pinnatibarbatus Bennett, 1831:146 (Atlantic coast of North Africa).
Exocoetus lineatus Valenciennes, in Cuvier & Valenciennes, 1846:92 (Goree [Dakar] Senegal).
Cypselurus lineatus:
Penrith, 1967.
Cypselurus pinnatibarbatus: Gibbs & Stalger, 1970.

DIAGNOSIS: (from Gibbs and Staiger, 1970; data from our material given in parentheses.) Pectoral fins greatly enlarged, reaching past base of anal fin; pelvic fins also enlarged, reaching past origin of anal fin; pelvic fin origins closer to anal fin than to base of pectoral fins; anal fin origin under or behind base of 3rd dorsal fin ray; jaws subequal; teeth microscopic, mostly unicuspid (conical); dorsal fin rays 11-14 (11-12); anal fin rays 9-12 (10); pectoral fin rays i, 13 - 15 (i, 13 - 14), the first ray unbranched, the rest branched; vertebrae 49-51 (49); predorsal scales 39-47 (40, 45); head length contained 4.4-5.0 (4.4, 4.5) times in SL.

Colour: body dark iridescent blue dorsally, silvery white ventrally; pectoral fins dusky, with narrow pale transverse band and the rear margin pale.

MATERIAL: Tristan: RUSI 13015: 375 mm SL. Nightingale: RUSI 33597: 2; 46 cm TL.

REMARKS: Flying fish are common in the waters around the Tristan Group in the warmer summer months (November to April). Theft presence coincides with sea surface temperatures of 15.5 to 20 C; according to Parin (1959), C. pinnatibarbatus comprises 6 subspecies, which are found circumglobally in waters of 15 to 25 C. At Tristan C. pinnatibarbatus are one of the major prey items for the hammerhead shark, Sphyrna mokarran, and the snoek, Thyrsites atun. Two specimens of 460 mm TL were found after landing aboard ship at night. Carmichael (1819: 494) mentioned a specimen of "Exocoetus exiliens that dropped on board a ship while at anchor, and which measured eighteen inches in length." He gave no descriptive details of the specimen, but judging from the length, it is likely that it was C pinnatibarbatus.

DISTRIBUTION: According to Gibbs and Staiger (1970), this species appears to be associated with islands or proximity to land and generally avoids tropical latitudes. It is known from the eastern North Atlantic (from the Azores and Spain to Liberia, including Madeira, Canary Islands and Cape Verde Islands); in the South Atlantic it has been reported from Ascension, St Helena, South Africa and southern Brazil. Our specimens from the Tristan Group represent the most southerly record for the Atlantic Ocean. Angot (1951) reported C. pinnatibarbatus (as "Exocoetus altipinnis") from St Paul and Amsterdam Island. Gon and Klages (1988) recorded a specimen from the vicinity of Marion Island at 46 degrees 45'S in the Indian Ocean.

Hirundichthys rondeletii, subtropical flying fish, (flying fish).

Exocoetus rondeletii
Valenciennes, in Cuvier & Valenciennes, 1846:115 (Mediterranean).
Hirundichthys rondeletii:
Parin, 1984.

DIAGNOSIS: (Data from our material given in parentheses.) Pectoral fins reaching past base of anal fin; pelvic fins reaching past origin of anal fin; pelvic fin origins closer to anal fin than to base of pectoral fins; anal fin origin under origin of dorsal fin; jaws subequal; dorsal fin rays 10-12 (10); anal rays 11-12 (11); pectoral fin rays ii, 15-18 (ii, 16), the first 2 rays unbranched, the rest branched; predorsal scales 27-32 (29).

Colour: body dark iridescent blue dorsally, silvery white ventrally; dorsal and caudal fins greyish; pectoral fins dark, with narrow pale margin along rear (ventral) edge of fin, other fins hyaline.

MATERIAL: Gough Id: RUSI 17709: 235 mm SL.

REMARKS: Our single specimen was collected in 1972 by G. Basson.

DISTRIBUTION: H. rondeletii is a subtropical species of the upper epipelagic ecosystem (Parin, 1968). Gough Island is on the southern limit of the range for this species.


Scomberesox saurus scomberoides, saury, Fig. 20

    Figure 20. Scomberesox saurus, 350 mm SL, South Africa (from Smith, 1986).

Esox saurus Bloch & Schneider, 1801:399 (part referring to J.R. Forster's ms. pertaining to New Zealand).
Sauris scomberoides Richardson, 1842:26 (on Esox saurus above and G. Forster's illustration of New Zealand specimen).
Scomberesox saurus:
Sivertsen, 1945.
Scomberesox saurus scomberoides: Hubbs & Wisner, 1980.

DIAGNOSIS: Body elongate, compressed, covered with small thin scales; dorsal and anal fins set on rear half of body with 5-7 finlets behind; jaw produced to form long slender beaks, with the lower jaw longest; dorsal fin rays (9-12) + 5 or 6 finlets; anal fin rays (11-14) + 6 or 7 finlets; pectoral fin rays 12-15; gill-rakers 39-51; vertebrae 64-70.

Colour: greenish brown to dark blue dorsally, separated by a silvery lateral stripe from the silvery-white ventral parts; fins hyaline.

MATERIAL: Tristan: SAM 28236:186 mm BL. Inaccessible: RUSI 33473: 5, 23-25 cm TL

REMARKS: Five specimens of approximately 24 cm TL were removed from the gut of a snoek, Thyrsites atun off Inaccessible Island in February 1990. The saury was recorded from the region by Sivertsen (1945). It is likely that S. saurus are present in the surface waters of the region during the warmer summer months as none were observed at sea during the winter. Hagan (1952) reported this species from the stomach contents of the rockhopper penguin, the yellownosed albatross (Diomedea chlororhynchos) and the brown noddy (Anous stolidus stolidus), which were collected in the summer of 1937/1938.

DISTRIBUTION: Tristan Group; the saury has a circumglobal distribution in the southern hemisphere (Hubbs and Wisner, 1980).


Lophotus lacepede crestfish,

    Pl. 1, Fig. B: Lophotus lacepede 120 cm SL, South Africa; from Heemstra, 1986)

Lophotus lacepede Giorna, 1809: 179, P1. II, Fig. 1 (type locality, probably Mediterranean Sea); Heemstra, 1986a.

DIAGNOSIS: Body elongate, markedly compressed, opening evenly to a minute caudal fin; dorsal fin origin in front of eyes, the anterior head profile vertical or inclined forward; dorsal fin-rays 220-263, the first ray produced; anal fin with 12-20 short rays placed near the caudal fin; pectoral fin-rays 14-17; pelvic fins rudimentary, situated below and behind pectoral fin base; no scales; swimbladder present; ink sack posterior to swimbadder, empties into cloaca.

Colour: Head and body silvery; fins red.

MATERIAL: Tristan: SAM 28234:136 cm TL.

REMARKS: The specimen was caught on a line in December 1977 by Mr E. Viljoen and Capt. E. Stoffberg of the MFV HEKLA.

DISTRIBUTION: The crestfish inhabits the lower epipelagic zone in tropical and subtropical waters of all oceans.


Beryx decadactylus beryx, Fig. 21

    Figure 21. Beryx decadactylus, 260 mm SL, Madeira, drawn by Mary Ann Holloway (from Woods & Sonoda, 1973).

Beryx decadactylus Cuvier, in Cuvier & Valenciennes, 1829:222 (Lisbon?); Penrith, 1967; Heemstra, 1986b.

DIAGNOSIS: Body compressed and deep, the depth 4450% SL; scales ctenoid; cheeks and operculum scaly; preopercle serrate, but no single large spine; mouth large, the maxilla expanded posteriorly, with 2 supramaxillae; dorsal fin with 4 spines and 16-20 rays; anal fin with 4 spines and 25-29 rays; pelvic fins with 1 spine and 10 rays; pectoral fin rays 15-17; lateral-line scales 56-61; gill-rakers (6-7) + (17-19).

Colour: Head and body scarlet dorsally, silvery white below; fins red.

MATERIAL: Tristan: SAM 21809: 362 mm SL.

REMARKS: Penrith (1967) received one specimen from Tristan.

DISTRIBUTION: B. decadactylus is found in subtropical and temperate waters of all oceans; records include the North Atlantic, South Africa, Australia, New Zealand, Japan and Hawaii (Heemstra, 1986b).


Centriscops obliquus, banded snipefish, (piper),

    Pl. 1, Fig. C: Centriscops obliquus, above,male.153 mm TL; below, female, 138 mm TL. both from Gough.

Centriscops humerosus obliquus Waite, 1911: 170, PI. 26 (New Zealand?).
Centiscops obliquus: Menni & Miquelarena, 1979; Heemstra, 1986c; Duhamel, 1989.

DIAGNOSIS: Body very compressed; snout produced into a long tube at the end of which is the small toothless mouth; scales minute but visible to the naked eye, like shark denticles with sharp ridges and spines; orbit rim spiny; bony plates on body above pectoral fins and along ventral midline; pelvics fins small, with a spine-like ray and 4 branched rays; dorsal fin with 7 spines (2nd spine greatly enlarged and somewhat movable, the other spines fixed) and 16-18 segmented rays; anal fin with 17-21 rays; pectoral fin rays 15-17; snout length contained twice in body length; orbit diameter contained 3.0-3.2 times in snout length.

Colour: Males white or pale orange, with 6 oblique dark red bands on body; females with several horizontal, wavy dark stripes.

MATERIAL: Tristan: RUSI 33608:145 mm TL. Gough: RUSI 33496: 2, 138-153 mm TL.

REMARKS: Our specimens represent the first record of C. obliquus from the region. Although it is found at all the islands, C. obliquus is seldom encountered because it is caught only in lobster traps set at 50 m and deeper. One specimen was picked up on the beach after a storm. Males and females differ considerably in colour pattern (Menni and Miquelarena, 1979). Dr G. Duhamel has informed us that C. obliquus is pelagic until it reaches 15 cm TL. However, the size of our specimens which were collected near the bottom in deeper than 50 m indicates that the pelagic phase can be terminated at a smaller size than this. This range in the size at settlement could be expected, as it is probable that juveniles will remain drifting in the pelagia until a suitable substrate on which to settle is reached. This is a phenomenon observed in other fish families and is especially common in the pleuronectiformes (Ahlstrom et al., 1984).

DISTRIBUTION: C. obliquus is known from Tristan, Gough Island, South Africa, St Paul/Amsterdam, Australia, New Zealand, Chile and Argentina (Heemstra, 1986c).

Notopogon lilliei, round bellowsfish, (piper)

Notopogon lilliei Regan, 1914:14 (New Zealand); Sivertsen, 1945; Heemstra, 1986c; Pequeno et al., 1992.

DIAGNOSIS: Body strongly compressed; snout produced into a long tube at the end of which is the small toothless mouth; scales microscopic, lanceolate and erect, giving the body a velvety feel; orbit rim not spiny; body depth (vertical at origin of pelvics) more than snout length and contained 1.6-1.9 times in distance from eye to end of caudal fin; distance from eye to origin of second dorsal spine distinctly less than distance from eye to caudal fin base; distance from base of second dorsal spine to first dorsal ray not more than length of soft dorsal fin base; eye diameter contained 2.7-4.0 times in snout length; snout length 1.6-1.9 times in body length (from rear edge of gill cavity to caudal fin base); dorsal fin with 7 spines (2nd spine greatly enlarged and somewhat movable, the other spines fixed) and 14 or 15 segmented rays; anal fin rays 17-19; pectoral fin rays 16-17; pre-dorsal profile of body evenly convex; adults with a patch of short bristles above operculum.

Colour: Body orange, with silvery white patches anterior and posterior to pectoral fins; soft dorsal, caudal and anal fins with white bands.

MATERIAL: Tristan: RUSI 33490:163 mm TL; SAM 26697:86 mm BL; SAM 27479:122 mm BL. Inaccessible: RUSI 33616: 2, 167-167 mm TL. Nightingale: RUSI 33492:226 mm TL; RUSI 33623:222 mm TL. Gough: RUSI 33617:200 mm TL.

REMARKS: N. lilliei is common in depths greater than 50 m at all the islands, and is eaten by Polyprion oxygeneios. All the specimens were collected in lobster traps set at depths of 40-200 m. N. lilliei is sometimes found washed up on beaches after storms, and it is presumed that the specimens Hagan (1952) found in the stomachs of black- browed albatross (Diomedea melanophris) were scavenged from beaches.

Penrith (1967) suggested that the species originally recorded by Sivertsen (1945) as N. lilliei was in fact N. macrosolen Barnard, 1925. However, the relatively large collection of specimens that we now have available indicates that the species is N. lilliei. According to Heemstra (1986c), N. macrosolen has a smaller eye (4.2-5.2 times in snout length), longer snout (1.3-1.5 times in body length), body depth 1.9-2.2 times in distance from eye to end of caudal fin, and the predorsal profile is straight (or nearly so) from the eye to the first dorsal spine.

DISTRIBUTION: Tristan Group and Gough Island; N. lilliei is also known from South Africa, New Zealand and South Australia (Heemstra, 1986c).


Helicolenus mouchezi, (soldier)

Sebastes mouchezi Sauvage, 1875:988 (St Paul Island).
Helicolenus tristanensis Sivertsen, 1945:17 - 19, Pl. 1, Fig. 12 and Pl. 6, Figs. 34 Tristan da Cunha); Eschmeyer, 1969.
Helicolenus mouchezi: Eschmeyer & Hureau, 1971.
Helicolenus dactylopterus gouphensis
Barsukov, 1979: 12, Fig. 2 (Gough Island).
Helicolenus dactylopterus mouchezi:
Duhamel, 1989.

DIAGNOSIS: Dorsal fin with 12 spines and 12-13 rays; anal fin with 3 strong spines and 5 or 6 rays; head spiny, but the suborbital ridge with or without a small spine below front edge of eye; pectoral fin rays 18-20, the upper part of fin truncate, the lower 8 rays unbranched and thickened with the tips free; lateral-line scales 27-30; vertical scale rows (counted above lateral line from cleithral spine to base of caudal fin) 75-84; gill-rakers 7-8 + 17-19, total 25-27; vertebrae 25 or 26. Head length (measured from front of upper jaw to tip of upper opercular spine) 37-39% SL; body depth 31-35% SL; orbit diameter 9.9-13.1% SL. Attains 42 cm total length.

Colour: Ground colour variable, from brown to pale red; head, body and dorsal fin covered with conspicuous, small, dark brown spots; dark spots also on other fins; belly white; peritoneum and inside of operculum black; tongue and posterior pan of buccal cavity dark grey.

MATERIAL: Tristan: RUSI 33619:230 mm TL. Inaccessible: RUSI 33500: 3, 138-165 mm TL. Nightingale: RUSI 33590: 8, 159-253 mm TL. Gough: RUSI 7682:293 mm SL; RUSI 18244: 2, 370-410 mm SL; RUSI 18248: 2, 310-340 mm SL. Plus 14 RUSI specimens and numerous fish 128-420 mm TL examined at the islands.

REMARKS: At Tristan, the soldier inhabits deeper water (40 to at least 200 m) than the other scorpaenid (Sebastes capensis), and smaller size classes are not found close to shore. Because of its preference for deeper water, H. mouchezi is less common than S. capensis in the catch of local fishermen. H. mouchezi is a fecund live bearer (approximately 300 000 embryos counted in the ovary of a 400 mm TL specimen) spawning during the summer months from October through to April. The larvae are born at a size of 3.5-4.0 mm TL. Moser (1967) reports that Sebastodes paucispinis in Southern California is ovoviviparous with large females producing up to two million larvae in one brood. H. mouchezi feeds almost exclusively on fishes.

This species was thought to be endemic to the Tristan islands by Sivertsen (1945). Eschmeyer and Hureau (1971) suggested that H. mouchezi from Amsterdam and St Paul Islands is conspecific with H. tristanensis, and we accept their suggestion. H. mouchezi differs from H. dactylopterus of South Africa in having a wider interorbital (13.8-15.6% HL, [n= 9, 179 - 278 mm SL] versus 11.2-13.9% HL [n= 8, 203 - 226 mm SL] in dactylopterus), shorter dorsal fin spines (3rd dorsal spine 13-16% SL, versus 16-18% SL in dactylopterus), fewer gill-rakers (lower-limb rakers 17-19 versus 20-22), more numerous scales (75-84 vertical scale rows, versus 60-67 rows) and in colour pattern (head and body covered with conspicuous, small, close-set dark brown spots, versus dorsolateral parts of body with a few small, faint, irregular dark spots).

Barsukov (1979) described a new subspecies (Helicolenus dactylopterus gouphensis) from Gough Island based on 5 specimens. Three of the four characters that Barsukov used to differentiate gouphensis from other species or subspecies of Helicolenus, namely body depth (28.730.8% SL), the number of anal fin rays (5 or 6) and the colour pattern do not differ significantly from the condition in H. mouchezi from Tristan and St Paul/Amsterdam islands. The fourth character, 26 vertebrae is also not useful to distinguish specimens from Gough Island; of 4 specimens from Tristan that we x-rayed, 1 has 26 vertebrae and the other 3 fish have 25 vertebrae.

DISTRIBUTION: This species is endemic to the Westwind Drift Islands Province: Tristan Group, Gough, St Paul and Amsterdam, and the Austral and Sapmer seamounts of the southwestern Indian Ocean (Duhamel, 1989).

Sebastes capensis, false jacopever, (soldier)

Scorpaena capensis Gmelin, 1789:1219 (Cape of Good Hope); Carmichael, 1819.
Sebastichthys capensis: Sivertsen, 1945.
Sebastes capensis: Eschmeyer, 1969; Eschmeyer & Hureau, 1971; Eschmeyer, 1986.

DIAGNOSIS: Dorsal fin with 13 spines and 13-14 rays; anal fin with 3 strong spines and 6 rays; head spiny, but the suborbital ridge without a small spine below front edge of eye; pectoral fin rays 18-19, the lower 9-10 unbranched, the fin wedge-shaped; lateral-line scales 34-41; gill-rakers, including rudiments, 8-9 + 19-21; vertebrae 25 or 26. Head length 37-39% SL; body depth 33-36% SL.

Colour: Variable, ranging from gold, through red, to orange with 5 or 6 pale spots on body at base of dorsal fin.

MATERIAL: Tristan: RUSI 33473:148 mm TL; RUSI 33582:318 mm TL; RUSI 33621: 255 mm TL; SAM 26696:153 mm SL.. Inaccessible Id: RUSI 33489:165 mm TL; RUSI 33589:121 mm TL; RUSI 33592: 3, 125-233 mm TL. Nightingale Id: RUSI 33497: 3, 225-247 mm TL; RUSI 33589: 7, 153-264 mm TL. Gough: RUSI 33563: 330 mm TL; RUSI 33564:435 mm TL. SAM 27489:259 mm SL; SAM 29626-29631: 29, 117-398 mm SL; SAM 3044: 3, 150-258 mm SL; SAM 30445: 3, 195-266 mm SL. Plus 30 additional RUSI specimens and numerous fish of 107-455 mm TL examined at the islands.

REMARKS: S. capensis occurs at all the islands from 2 to about 150 m, but is rare at depths greater than 100 m. Small individuals (less than 20 cm TL) are found in rockpools and shallow sub-tidal habitats to about 10 m. Adults are found in deeper water. This species is the second most abundant fish (after the five-finger) in the catch of the Tristan fishermen. It is used only as bait for lobster traps. S. capensis is highly territorial, even as juveniles, with individuals occupying caves and clefts. It is ovoviviparous and highly fecund (215 000 embryos counted in a 405 mm TL, 1070 g fish). The embryos hatch inside the ovary and are released at a length of 3.0 - 4.0 mm TL) as free-swimming larvae. This ambush predator feeds mainly on fishes, but will also take benthic invertebrates. (Andrew and Heeht, in press)

DISTRIBUTION: Chen (1971) considered the 11 nominal species that have been described from South America. Tristan da Cunha, Gough Island and South Africa as conspecific. According to Eschmeyer and Hureau (1971), who apparently were not aware of Chen's (1971) work, S. capensis may prove to be synonymous with S. oculatus Cuvier, 1833 of Chile and Argentina. Kong U (1985) synonymized 'S. capensis' with 7 nominal species from the coasts of Chile and Argentina. The colour photo of the specimen from South America that was published by Nakamura (1986a) is very similar to S. capensis, but the count of 55-66 lateral-line scales given by Nakamura is much higher than in S. capensis or in the South American specimens examined by Chen (1971). Perhaps Nakamura was counting scale series above the lateral line, rather than pored lateral-line scales. Nakamum's count of "D XII, 12-14;" for the dorsal fin spines of Sebastes oculatus is also a mistake; this should have been written D XIII, 12-14 (i.e., 13 dorsal fin spines). Sebastes capensis was reported from St Paul/Amsterdam Islands by Blanc (1961), but Eschmeyer and Hureau (1971) reidentified these specimens as Helicolenus mouchezi.


Polyprion oxygeneios, wreck fish, (steambras)

Epinephelus oxygeneios Bloch & Schneider, 1801:301 (New Zealand).
? Polyprion americanus (non
Bloch and Schneider): Rowan and Rowan, 1955.
Polyprion oxygeneios: Roberts, 1989.

DIAGNOSIS: Body oblong, somewhat compressed, covered with small strongly ctenoid scales; body depth contained 3.3-3.8 times in SL (for fish larger than 25 cm SL); dorsal fin with 11 or 12 strong spines and 11 or 12 soft rays; anal fin with 3 spines and 8-10 rays; pectoral fins shorter than pelvics; caudal fin truncate; head, except for snout scaly; maxilla exposed, scaly, with well-developed supramaxilla; rear edge of opercle with 2 spines, the lower one at the end of a distinct horizontal ridge; preopercle with large spines in juveniles, serrate in adults; teeth villiform, in bands on jaws, vomer and palatines and in a small oval patch on tongue. Vertebrae 13+ 14.

Colour: Body distinctly bicoloured, the dorsal half dark bluish grey and the ventral half abruptly pale.

MATERIAL: Tristan: SAM 28235, 100 cm SL; 19, 57-145 cm TL examined at Tristan, but none preserved by us. Gough: SAM 30450, 460 mm SL.

REMARKS: According to Roberts (1989) the genus Polyprion comprises two species, P. oxygeneios and P. americanus (Bloch and Schneider, 1801); both species are widely distributed in the Southern Hemisphere, P. americanus also occurs in the North Atlantic; it differs from oxygeneios in having a deeper body (body depth 2.3-3.0 times in SL for fish larger than 25 cm SL) and a mottled or uniformly dark colour pattern (body not distinctly bicoloured).

Rowan and Rowan (1955) examined a "23 inch, 6 lb" steambras from Tristan, which they identified as "Polyprion americanus". All of the Polyprion that we have seen from Tristan (some 50 specimens) were P. oxygeneios, but P. americanus might also occur there, as it is known from Vema Seamount (Penrith, 1967; SAM 24300, 457 and 581 mm SL; SAM 31701, 530 mm SL) and is common off South Africa and the coast of Argentina.

P. oxygeneios occurs at all of the Tristan Group islands and at Gough on high relief, rocky bottom at depths of 80 to at least 200 m. It is not often caught by the local fishermen (about 100 individuals a year), and no juveniles smaller than 50 cm TL were seen. Juveniles are pelagic until at least 30 cm TL and often associate with floating objects (Roberts, 1989). An increase in gonad weight towards the end of summer indicates that the species spawns at this time (February-April). P. oxygeneios feeds mainly on fish and lobsters.

DISTRIBUTION: The species has been reported from depths of 100 - 1000 m along continental coasts and off oceanic islands in most subtropical and temperate regions of the Southern Hemisphere (Roberts, 1989), including St Paul and Amsterdam Islands (Duhamel, 1989).


Lepidoperca coatsii

Caesioperca coatsii Regan, 1913: 237, Pl. 6, Fig. 1 (Gough Island).
Lepidoperca coatsii Duhamel, (1989).

DIAGNOSIS: Dorsal fin with 10 spines and 16-18 rays, the interspinous membranes deeply incised; anal fin with 3 spines and 7-8 rays; pectoral fins with 16-17 rays; caudal fin truncate with 15 branched rays; lateral line scales 43-48; 2 or 3 rows of scales between lateral line and 5th dorsal fin spines; gill-rakers 9-12 + 22-25. Body depth contained 2.7- 3.0 times in SL; head length 2.6-2.9 times in SL; eye diameter 2.4-3.0 times in head length (HL); bony interorbital width 4.0-5.2 times in HL; pectoral fins extend to between anus and base of 2nd anal fin ray, their length contained 1.1-1.2 times in HL; pelvic fins extend to between anus and anal fin origin. Head, including snout, maxilla, lower jaw and gular area scaly; maxilla with or without a sprint-like supramaxilla; jaws with a row of small slender teeth and a pair of widely-spaced, exserted, small canines at front; 2 or 3 small canines at middle of each dentary; chevron-shaped band of small sharp teeth on vomer and 4 or 5 series of similar teeth on palatines; no teeth on tongue. Vertebrae 10+16; supraneural bones 3.

Colour: Head and body pinkish orange fading to white ventrally; 4 or 5 distinct purple lines running longitudinally along the body below pronounced lateral line. Fins yellowish, except spinous part of dorsal which has dark blotches on interspinous membranes.

MATERIAL: Nightingale: RUSI 31523: 3, 110-113 mm SL; RUSI 33606, 111 mm. Gough: BMNH 1912.7.172-6: 4, 70-78 mm (syntypes of Caesioperca coatsii Regan); RUSI 13294:82.5 mm; RUSI 33609:138 mm.

REMARKS: Lepidoperca coatsii is a relatively rare species found at all the islands. The species was originally described from specimens collected at Gough Island in 100 fathoms by the SCOTIA (Regan 1913). The 6 specimens collected in the present study were captured at Gough and Nightingale islands in lobster traps set in about 50 m of water.

The two specimens that we examined from the Indian Ocean Austral Seamount (MNHN 1984-82 and 1984-83) are indistinguishable from the Gough Island material.

DISTRIBUTION: In addition to Gough and the Tristan Group, L. coatsii has also been reported from St Paul and Amsterdam Island (Duhamel, 1989) and the Austral Seamount (Duhamel, 1984).


Pseudopentaceros richardsoni, pelagic armourhead, Fig. 22

    Figure 22. Pseudopentaceros richardsoni, 455 mm TL, Tristan.

Pentaceros richardsoni Smith, 1844:51 (pages not numbered), PI. 21 (Cape of Good Hope); Smith, 1964; Penrith, 1967.
Pseudopentaceros richardsoni:
Hardy, 1983; Heemstra, 1986d.

DIAGNOSIS: Body oblong, compressed; dorsal fin continuous, the spinous part much longer than soft-rayed part; dorsal fin spines 14 or 15, rays 8-10; anal fin spines 4 or 5, rays 7 or 8; pectoral fin rays 17-18; caudal fin truncate or slightly concave. Bones of head mostly exposed, rugose; mouth

small, the maxilla not reaching past front edge of eye; lateral-line scales 74-86; gill-rakers (5-7) + (14-17). Vertebrae 13+13.

Colour: Dark bluish dorsally, pale ventrally.

MATERIAL: Tristan: RUSI 33566, 465 mm TL; RUSI 33599, 455 mm TL; SAM 21791, 80 mm SL.

REMARKS: P. richardsoni was recorded from Tristan by Penrith (1967) who obtained a juvenile specimen from the Islands. This species is likely to be a summer spawner, as a 455 mm TL female with active gonads and a 465 mm TL male with ripe testes were collected with hook and line in 100 m during September 1989 and February 1990 respectively. P. richardsoni is probably common on the shelf edges surrounding the islands. Hardy (1983) described ontogenetic changes in the serrated ridges and spines on the head of small juveniles.

DISTRIBUTION: P. richardsoni has been collected from depths of 50 - 600 m off South Africa, Walters Shoal and other seamounts of the Indian Ocean, Australia, New Zealand, Sala y Gomez Ridge, and South America (Heemstra, 1986d; Penrith, 1967, Parm, 1991). Previous records from the North Pacific (Smith, 1964; Penrith, 1967) were reidentified as two new species by Hardy (1983). The absence of this widely-distributed species at the Amsterdam Islands (Duhamel, 1989) is puzzling.


Brama australis, pomfret, Fig. 23

Brama australis Valenciennes, 1840: Pl. 44, Fig. 1 and unnumbered text (Valparaiso, Chile); Pavlov, 1991: 6-9.
Brama sp. Mead, 1972: 81, Pl. 4; Duhamel, 1989.

DIAGNOSIS (modified from Pavlov, 1991; data from our specimen in parentheses): Body oblong, distinctly compressed; body depth 2.1-2.4 times in SL, the greatest body width 2.7-3.9 times in distance between snout and pelvic fin origin (The body depth cannot be measured on our specimen because of missing epaxial muscles.). Dorsal profile of head distinctly convex, the distance between eye and nearest point on dorsal midline more than haft eye diameter. Dorsal and anal fins scaly, dorsal fin rays 31-36 (dorsal fin damaged); anal fin rays 26-29 (28; last ray split to its base, which is contiguous with base of 27th ray); pectoral fin rays 20-22 (21); caudal fin lunate; lateral scale series 78-85 (about 79); total developed gill-rakers on one side of first gill arch 18-24 (5 + 13 = 18). Vertebrae 14 + 25 = 39 (16 + 24 = 40).

Colour: Head and body mottled metallic silvery-black, danker dorsally; scaleless snout and tips of median fin rays also black; inside of mouth black.

MATERIAL: Tristan: RUSI 34593, 490 mm SL.

REMARKS: Only one specimen was collected; it was stranded on Halfway Beach on Tristan Island. This is a new distributional record for the species. According to the Islanders, this fish is washed up fairly regularly after storms. It is therefore probable that the species is common in waters not fished with hook and line (i.e., deeper than 200 m).

Our identification of this specimen is somewhat tentative. It was found damaged, and most of the dorsal fin and epaxial muscles are missing. Except for the vertebral count, all of the menstic data fit the diagnosis given by Pavlov (1991) for Brama australis and rule out other species that are known from the Atlantic Ocean. According to Mead (1972), B. brama has more vertebrae and more anal fin rays; B. dussumieri has more vertebrae and fewer gill-rakers; B. caribbea has fewer vertebrae, fewer gill-rakers and fewer lateral scale series; B. orcini has fewer lateral scale series and fewer gill-rakers; B. myersi has fewer pectoral fin rays and more vertebrae.

Mead (1972) described two specimens from South Africa (434 mm SL) and New Zealand (468 mm SL) which he assigned to an undescribed species (designated "Brama sp."). According to Mead (1972: 81), Brama sp. differs from B. brama and B. japonica in that "the gill filaments (not the rakers) appear to be much longer (longer than distance between anterior nostril and orbit, nearly as long as snout; while in B. brama, and in B. japonica of somewhat shorter length, these filaments are about equal to the distance from anterior nostril to orbit)". In our specimen, the length of the longest gill filaments on the first arch are 26 mm and the distance from anterior nostril to eye is 17 mm; Mead (1972) did not record these measurements for the two specimens that he examined.

B. australis is very similar to B. japonica, which is known only from the North Pacific. Pavlov (1991) mentioned that the dorsal head profile of B. japonica is more steeply convex than in B. australis, and this is also shown on Mead's (1972: Pls. 3 and 4) illustrations.

DISTRIBUTION: Brama australis (including Brama sp. Mead, 1972) is also known from the southeastern Pacific Ocean, Chile, New Zealand, South Africa and Amsterdam Island (Duhamel, 1989).


Emmelichthys nitidus, southern rover, Fig. 24

    Figure 24. Emmelichthys nitidus, 300 mm SL, South Africa.

Emmelichthys nitidus Richardson, 1845: 47, Pl. 29, Figs. 7 and 8 (west coast of Australia); Heemstra and Randall, 1977; Heemstra, 1986e.

DIAGNOSIS: Body elongate, subcylindrical; body depth 4.1-5.2 times in SL, and less than head length; spinous dorsal, fin separated from soft dorsal fin by a gap with 2-4 isolated spines; soft dorsal and anal fins with scaly basal sheath; dorsal fin with 13 spines and 9-11 soft-rays; anal fin with 3 spines and 10 rays; pectoral fin rays 20-23; caudal fin forked; upper jaw very protrusile; maxilla scaly, not covered by preorbital bone when mouth is closed; supramaxilla long and slender; lateral-line scales 87-98; gill-rakers (10-12) + (27-31). Vertebrae 10+14.

Colour: Dark bluish grey dorsally, silvery reddish on ventral parts.

MATERIAL: Tristan: RUSI 34592:285 mm TL.

REMARKS: E. nitidus occurs near the bottom in waters deeper than 100 m. The only specimens collected by us came from the stomach of the deep water predator Polyprion oxygeneios. This is a new record for the region.

DISTRIBUTION: The southern rover is also recorded from South Africa, Australia, New Zealand, St Paul and Amsterdam Islands (Heemstra, 1986e).


Naucrates ductor, pilotfish, Fig. 25

    Figure 25. Naucrates ductor, 300 mm SL, South Africa.

Gasterosteus ductor Linnaeus, 1758:295 ("in Pelago").
Naucrates ductor:
Smith-Vaniz, 1986.

DIAGNOSIS (modified from Smith-Vaniz, 1986): Body elongate, subcylindrical; caudal fin forked; first dorsal fin with 7 spines (first minute), second dorsal with 1 spine and 30-35 rays; anal fin with 2 small spines preceeding fro, a slender spine closely bound to first ray, and 19-22 rays; pectoral fins shorter than pelvics; length of dorsal-fin lobe in adults not much longer than pectoral fin, 11-13% fork length; gill-rakers (7-10) + (15-20) = 22-29. Maxilla reaches about to vertical through anterior edge of pupil. Caudal peduncle with a low fleshy keel on each side.

Colour: Body silvery blue dorsally, with 6 broad dark bars, including one at base of caudal fin; caudal fin lobes with white tips.

MATERIAL: Tristan: SAM 28931: 2, 312-341 mm FL.

REMARKS: N. ductor is a member of the pelagic, open-- ocean fauna, and juveniles are often associated with jellyfishes or flotsam of terrestrial origin. The pilotfish got this name from its habit of riding the bow wave of ships or large fishes. This commensal relationship is adopted with several species of sharks, rays and other large fishes.

DISTRIBUTION: The pilotfish is known from tropical and temperate waters of all oceans.

Seriola lalandi, giant yellowtail, (yellowtail),

Seriola lalandi Valenciennes, in Cuvier & Valenciennes, 1833:208 (Brazil); Smith-Vaniz, 1986.

DIAGNOSIS (modified from Smith-Vaniz, 1986): Caudal fin forked; first dorsal fin with 7 spines (first minute), second dorsal with 1 spine and 30-35 rays; anal fin with 2 small spines preceeding fin, a slender spine closely bound to first ray, and 19-22 rays; pectoral fins shorter than pelvics; length of dorsal-fin lobe in adults not much longer than pectoral fin, 11-13% fork length; gill-rakers (7-10) + (15-20) = 22-29. Maxilla reaches about to vertical through anterior edge of pupil. Caudal peduncle with a low fleshy keel on each side.

Colour: Blue to olivaceous above, shading to silvery white below; sometimes with narrow bronze stripe along midside of body; caudal fin olivaceous yellow.

REMARKS: Seriola lalandi was first recorded by Rowan and Rowan (1955) from Tristan. It seems that this fish is extremely rare in Tristan waters. A specimen of 60 cm TL was caught with hook and line in February 1989; it was not kept. One other specimen was observed underwater at about 20 m during March 1989. This subtropical species probably enters Tristan waters (possibly a fringe area of the species' range) during the warmer summer months.

DISTRIBUTION: The yellowtail is considered to be circumglobal and restricted to subtropical waters (Smith-Vaniz, 1986).

Trachurus longimanus, southern horse mackerel, (mackerel)

Decapterus longimanus Nonnan, 1935b: 255 (Tristan da Cunha); Penrith, 1967.
Trachurus longimanus: Duhamel, 1989; Collette & Parin, 1991.

DIAGNOSIS (data from 6 specimens, 355-492 mm FL): Body depth 4.4-5.0 times in FL; head length 3.7-3.9 times in FL; greatest height of scales in curved part of lateral line 3.0-3.8% FL, the greatest height of scales in straight part 4.0-4.6% FL. Caudal fin forked; first dorsal fin with 8 spines, second dorsal with 1 spine and 31-34 rays; anal fin with 2 distinct spines preceding fin, a slender spine closely bound to first ray, and 26-29 rays; last ray of dorsal and anal fins enlarged, connected by membrane to penultimate ray but set about 50% farther apart than other rays; pectoral fins with 22-23 rays (2 dorsal-most rays unbranched);curved part of lateral line with 48-53 scales (these scales vertically elongate but mostly hidden by overlying skin and small accessory scales); straight part of lateral line with 44-50 scutes (including those on caudal fin); gill-rakers (15-17) + (40-42) = 56-59.

Colour: Body silvery-grey, dark bluish green dorsally, silvery white ventrally; a small black spot on upper rear margin of operculum.

MATERIAL: Tristan: RUSI 34594:395 mm FL; RUSI 34595:492 mm FL. Inaccessible: RUSI 33567:355 mm FL. Nightingale: RUSI 33569:460 mm FL. Gough: RUSI 11084:448 mm FL; SAM 30443:400 mm FL.

REMARKS: T. longimanus shoals from the surface down to about 50 m and is present at Gough as well as the northern islands. At Tristan, it spawns from October to March. As expected from the long numerous gill-rakers, planktonic organisms such as amphipods and small pelagic fishes are the major prey. Collette and Parin (1991) reported euphausids (Euphausia recurva) as the primary food item in specimens from Walters Shoal.

DISTRIBUTION: Once thought to be endemic to Tristan (Sivertsen, 1945), T. longimanus has since been recorded from Vema Seamount (Penrith 1967), St Paul and Austral Seamount (Duhamel, 1989) and Walters Shoal (Collette and Parin, 1991).


Remora remora, shark remora, Fig. 26

    Figure 26. Remora remora, 230 mm SL, South Africa.

Echeneis remora Linnaeus, 1758:260 (Indian Ocean).
Remora remora:
Heemstra, 1986f.

DIAGNOSIS: Body elongate; top of head covered by a long flat sucking disk; body depth contained 5-8 times in SL; dorsal and anal fin bases less than twice head length; disc with 14-20 ridges; dorsal fin rays 22-26; anal fin rays 22-24; pectoral fin rays 26-30; gill-rakers (5-6)+(26-28).

Colour: Dark brownish grey.

MATERIAL: Tristan: RUSI 7296: 2, 102-167 mm SL.

REMARKS: As the name implies, the shark remora is usually attached to sharks. Our 2 specimens were collected in June 1977 by Peter Waffen, captain of a lobster-fishing vessel operating at the islands.

DISTRIBUTION: Worldwide except in polar seas.


Acantholatris monodactylus, (fivefinger),

    Pl. 2, Figs C & D: C) Acantholatris monodactylus, benthic adult, 200 mm TL; Tristan. D) A. monodactylus, 57 mm TL; Tristan.

Chaetodon monodactylus Carmichael, 1819: 500, Pl. 24 (Tristan da Cunha).
Cheilodactylus Carmichaelis Cuvier, in Cuvier & Valenciennes 1830:360 (based on Chaetodon monodactylus Carmichael).
?Cheilodactylus aspersus Richardson, 1850:64 ("Port Arthur, Van Diemen's Land" [= Tasmania], erroneous).
Acantholatris monodactylus: Gill, 1863; Sivertsen, 1945; Penrith, 1967; Duhamel, 1989; Collette & Parin, 1991.
Chilodactylus monodactylus:
Sauvage, 1879; Regan, 1913; Norman, 1935a.

DIAGNOSIS: Caudal fin forked; dorsal fin with 16-18 spines and 24-27 rays; anal fin with 3 spines and 11 or 12 rays; pectoral fins with 14-15 rays, the upper 2 rays and the lower 5-6 rays unbranched, the other rays branched; lateral line scales 50-54; gill-rakers 5-7 + 16-18. Measurements from 18 specimens, 226-323 mm SL: Body depth at pelvic fin origin 38-46% SL, head length 28-34% SL, and snout to dorsal fin origin 28-34% SL; orbit diameter 4.9-9.5% SL, interorbital 5.6-9.5% SL, caudal peduncle depth 7.8-8.9% SL, longest pectoral ray 33-43% SL.

Colour: Body usually bronze (but the colour vanes from silver through bronze to orange) with 5-6 dark brown or black vertical bars on dorsal part of body, the first at dorsal fin origin and the last at base of caudal fin; dorsolateral parts of head and body with numerous small black spots.

MATERIAL (all lengths are SL): Tristan: RUSI 33570: 253 mm; RUSI 33603:284 mm. Nightingale Id: RUSI 33575: 2, 200-218 mm. Inaccessible Id: RUSI 33586: 2, 282-290 mm; RUSI 33587: 4, 214-252 mm. Gough Id: RUSI 26185:260 mm; RUSI 33573:250 mm; RUSI 33568:323 mm; RUSI 33576:281 mm; RUSI 33583:226 mm; RUSI 33593:241 mm; RUSI 33600:274 mm; RUSI 33602:280 mm.

REMARKS: The fivefinger is the dominant species in terms of biomass on the shelf areas of the Tristan islands, and it is also the most abundant species captured for food and bait. It occurs in all well-covered habitats (i.e., algal covered reefs and kelp stands) from 1.5 m to at least 150 m. The species is an opportunistic carnivore, feeding on benthic invertebrates, planktonic salps, amphipods, and fishes (Andrew and Hecht, 1992). It is normally restricted to within 10 m of the seabed. Larval and paperfish stages of this species are pelagic, while juveniles are usually seen swimming in shoals of up to 15 individuals within 2 metres of the bottom. Tagging studies done at Tristan show that larger individuals are territorial, and have a relatively small home range. Adults are often seen resting on the bottom in caves or crevices. The species attains 65 cm TL and is relatively slow growing. Our age and growth studies (using otoliths) show that large adults attain ages in excess of 25 years. The fivefinger spawns in summer from early February to late April.

Specimens from Gough Island appear to have a smaller head than those from the Tristan group: head length 28-29% SL (7 fish of 226-323 mm SL) versus head length 30-34% SL (10 fish of 200-290 mm SL). The colour patterns of specimens from Gough do not differ significantly from those at Tristan (or those that we have seen from St Paul and Amsterdam Islands). In view of the small sample of fish that we have from Gough and the absence of any other differences in the specimens that we have examined, we consider the Gough population of A. monodactylus to be conspecific with the population at Tristan.

DISTRIBUTION: In addition to the Tristan Group and Gough Island, A. monodactylus has been recorded from St Paul and Amsterdam Islands (Sauvage, 1879; Kner, 1865; Huteau, 1969; Duhamel, 1989), Austral Seamount in the southwest Indian Ocean (Duhamel, 1984), Walters Shoal (400 miles south of Madagascar, Collette and Parin, 1991) and from Vema Seamount (Penrith, 1967).


Latris lineata, striped trumpter, (funnyfish),

Cichla lineata Forster, in Bloch & Schneider, 1801:342 (New Zealand).
Latris lineala: Penrith, 1967; Last et al., 1983; Duhamel, 1989.

DIAGNOSIS (data from our specimen given in parentheses): Body elongate, compressed, the depth contained 2.9-3.1 times in SL; caudal fin forked; dorsal fin with 18 spines and 34-36 rays, the fin divided almost to the base between spiny and soft-rayed parts; anal fin with 3 spines and 25-27 (28) rays; pectoral fins with 18 (17) rays, the lower 8 or 9 rays unbranched, the middle rays longest; gill-rakers 6 + 16.

Colour: Head and body dark olive green dorsally, with 3 silvery bands from head to tail; belly silvery yellow; fins greenish yellow.

MATERIAL: Tristan: RUSI 33595:745 mm TL. Gough: SAM 22623: (skin) 90 cm SL: SAM 22891:90 cm SL.

REMARKS: One specimen was collected at Tristan Island in 80 m. This species is seldom hooked, but it is known from all the islands including Gough (Jimmy Glass, Tristan da Cunha, pers. comm.).

DISTRIBUTION: L. lineata is also known from the Amsterdam Islands (Duhamel, 1989), Australia, New Zealand (Last et al., 1983; Ayling and Cox, 1982) and the sub-Antarctic Auckland Islands (Kingsford et al. 1989).

Mendosoma lineatum

    Pl. 2, Figs F & G: F) Mendosoma lineatum, adult,225 mm TL; Tristan; G) M. lineatum, juveniles. 83 & 86 mm TL; Tristam

Mendosoma lineatum Guichenot, 1848:213 (Valparaiso, Chile); Gon & Heemstra, 1987; Duhamel, 1989.

DIAGNOSIS: Body elongate, moderately compressed, the depth contained 3.5-3.9 times in SL; caudal fin forked; dorsal fin with 22-24 spines and 24-27 rays, the fin divided between spinous and soft-rayed parts; anal fin with 3 spines and 17-21 rays; pectoral fin rays 16-18, the upper 2 and lower 6-8 rays unbranched, but not thicker or longer than other rays; mouth very protrusile, the upper jaw with 1 row of small conical teeth; no teeth on lower jaw, vomer or palatines; body covered with cycloid scales; lateral-line scales 69-77; lower-limb gill rakers 20-22. Vertebrae 4246.

Colour: Bluish green dorsally, fading to silver-yellow below lateral line; fins pale yellow, the caudal fin darker yellow.

MATERIAL: Tristan: RUSI 33471: 2, 83-86 mm TL; RUSI 33625:225 mm TL. Nightingale Id: RUSI 33613: 218 mm TL. Gough Id: RUSI 26047, 175 mm SL; RUSI 33626: 4, 210-235 mm TL.

REMARKS: M. lineatum is common at all the islands and shoals in groups of similar-sized individuals. Adults are epibenthic below about 20 m, but juveniles and small fish shoal close to the surface where they are easy prey for birds. Klages et al. (1988) reported this species from the stomachs of rockhopper penguins at Gough Island. Small shoals of juveniles occasionally become trapped in rockpools. These pelagic juveniles are compressed and silvery (PI. 2, Fig. G) like the "paperfish" stage of cheilodactylids. M. lineatum is mainly planktivorous, but it will also feed on small pelagic fishes. The species forms an important prey item for the snoek, Thyrsites atun.

DISTRIBUTION: Gon and Heemstra (1987) reported the first Atlantic record of this species from Gough Island. M. lineatum is also known from Chile, New Zealand, Tasmania, St Paul and Amsterdam Islands (Gon and Heemstra, 1987; Dnhamel, 1989) and the sub-Antarctic Auckland Islands (Kingsford et al., 1989).


Nelabrichthys ornatus Tristan wrasse (concha)

    Pl. 2, Figs H & I: H) Nelabrichthys ornatus, female, 110 mm TL; Tristan; I) N. ornatus, terminal male, 200 mm TL, Tristan.

Labrus ornatus Carmichael, 1819: 502, PI. 27 Tristan da Cunha).
Labrichtys Lantzii
Sauvage, 1875:988 (St Paul Island).
Labrichthys isleanus
Sauvage, 1875:988 (St Paul Island).
Platyglossus robinsoni
Gilchrist & Thompson, 1914:86 ("Natal" [specimen evidently mislabelled]).
Labrichthys ornatus:
Sivertsen, 1945.
Nelabrichthys ornatus: Russell, 1983; Duhamel, 1989; Collette & Parin, 1991.

DIAGNOSIS: Body elongate, compressed; body depth contained 3.9-5.3 times in SL; head length 3.3-4.2 times in SL; snout 3.7-4.4, eye diameter 4.5-6. 1, interorbital width 5.3-6.8 and suborbital depth 5.9-9.2 times in head. Caudal fin rounded; dorsal fin with 9 spines and 12-13 rays; anal fin with 3 slender spines and 11-12 rays; pectoral fin rays 12-14; lateral-line complete, with 26-27 scales; lateral line tubes branched; total gill-rakers 15-25; branchiostegal rays 6; cheeks scaly; preopercle entire. Vertebrae 9 + 17-18.

Colour: Sexually dichromatic; three colour phases are distinguished. Juveniles: predominantly yellow with indistinct markings of initial phase females. Initial (female) phase: head and body olive brownish with azure lateral stripes on body. Juveniles and females with 3 distinct black spots: one between first two dorsal fin spines, the second near base of dorsal fin between last 3 fin rays, and the third at middle of caudal fin base. Terminal (male) phase: Head and body brownish, the body scales marked with violet or red; lines on head and body violet or blue; dorsal fin purplish brown tinged with red, with 3 violet longitudinal stripes and a black spot between first and third spines (no black spot at rear end of dorsal fin or base of caudal fin; anal fin yellowish, violet edged; caudal fin reddish yellow, with concentric violet stripes; paired fins hyaline.

MATERIAL: Tristan: RUSI 33468: 2, 40-50 mm TL; RUSI 33483:188 mm TL; RUSI 33607: 7, 90-175 mm TL; RUSI 33615: 6, 90-180 mm TL; SAM 10503 (5, not measured); SAM 10504 (6, not measured); SAM 24070 (1, not measured); SAM 25284 (1, not measured); SAM 23764 (1, not measured). Nightingale: RUSI 33481:155 mm TL. Gough: SAM 29920: 9, 91-165 mm SL; SAM 29922: 6, 155-172 mm SL; SAM 29923: 4, 84-168 mm SL.

REMARKS: N. ornatus is the most numerous species on the shelf areas of the three northern islands. It was not seen at Gough Island during visits by the senior author (T.G.A.), but there are three lots of specimens in the South African Museum (see above) that were collected at Gough Island with spearguns by W. Liltved and T. Gosliner in September 1981.

N. ornatus inhabits waters from 1.5 m to at least 50 m and is rarely seen more than 2 metres above the bottom. The species favours areas with plentiful cover, either in the form of reef or kelp beds. The diet comprises mainly epiphytic organisms.

N. ornatus is a protogynous hermaphrodite with a lek-like mating system. All fish begin life as females, and some change to males at a size of 16-17 cm TL. The mechanism inducing this change is not clearly understood, but is thought to be related to size, as adult males establish temporary territories. Open areas surrounded by adequate cover are favoured for establishing territories, and larger males occupy more desirable sites that are close to cover and more sheltered. Secondary males of 17 cm TL and larger will set up territories about 2 m in diameter each evening throughout the spawning season (November to March). Vigorous displays are directed at any females moving into the territory of a particular male. If a female accepts the advances, the pair swim rapidly upwards and spawn in midwater. A detailed description of the mating system is currently in press. (Andrew et al., in press)

DISTRIBUTION: N. ornatus is known from Tristan, Gough, Vema Seamount (Duhamel, 1989), St Paul and Amsterdam Islands (Duhamel, 1989). The colour pattern of the 38 mm SL juvenile from Walters Shoal that was illustrated by Collette and Parin (1991) shows a dark red spot on each body scale, but similar-sized juveniles from Tristan have a pale lilac-colored spot on each of the ventrolateral body scales. The sexually dimorphic colour patterns of fresh males and females from St Paul Island (as described by Sauvage, 1879) also differ considerably from adults at Tristan (photographed by T.G. Andrew). In view of the subtle colour differences between many wrasse species, we are dubious about the conspecific status of the wrasses at Tristan, St Paul and Walters Shoal that are currently identified as "Nelabrichthys ornatus". If these wrasses are not conspecific, they certainly represent a monophyletic group of species.


Bovichtus diacanthus

Callionymus diacanthus Carmichael, 1819: 501, P1. 26 (Tristan da Cunha).
Bovichthys diacanthus: Regan, 1913; Penrith, 1967.

DIAGNOSIS: Body elongate, subcylindrical, the depth contained about 5 times in SL; head length about 3 times in SL; two dorsal fins joined at the base; first dorsal fin with 8-9 spines; second dorsal fin with 20-21 rays; anal fin rays 15-16; dorsal and anal fin-rays unbranched; pectoral fin rays 14-15; pelvic fins with 1 spine and 5 rays, the fin origin distinctly anterior to pectoral fin base; caudal fin truncate or slightly convex; head and body naked; a large spine on opercle; a single nostril on each side of snout.

Colour: Generally olive or blackish green with white blotches; belly silver white; lower border of paired fins, also anal and caudal fins edged with white. B. diacanthus is able to adopt a reddish brown or blackish colour in order to match the colour of its habitat.

MATERIAL: Tristan: RUSI 33484: 3, 135-162 mm TL; RUSI 33487: 6, 120-185 mm TL; RUSI 33488: 9, 115-152 mm TL; SAM 10505 (14); SAM 26991 (4); SAM 23951 (1). Gough: RUSI 33614:190 mm TL; SAM 27478 (1); SAM 27481 (1); SAM 29925 (1); SAM 29926 (3). Plus 20 additional RUSI specimens and numerous fish of 50-250 mm TL examined at the islands.

REMARKS: B. diacanthus is extremely common in rockpools 10-200 cm deep and in subtidal areas down to about 20 m. This species is able to withstand a wide temperature range. Sun-warmed rockpools where the species occurs reach summer midday temperatures of up to 27.4 C, and winter surface ocean temperatures can drop as low as 10 C at Gough Island.

The pelagic larva develops into a silvery post-larval stage, and at 5-6 cm TL, the post-larvae move into shallow water where they transform to the benthic olive-coloured juveniles. The pelagic post-larvae occur in large shoals and are important prey for predators such as Thyrsites atun. Adults become scarce or absent from the shallow subtidal region during February and March. Gonad development prior to this seaward movement suggests that it is a spawning related event. B. diacanthus feeds mainly on epiphytic fauna and amphipods.

DISTRIBUTION: B. diacanthus is endemic to the Tristan Group and Gough Island. Other species of Bovichtus are known from St Paul and Amsterdam Islands (Duhamel, 1989), the Antarctic Peninsula and the southern coasts of South America, Australia and New Zealand (Fischer and Hureau, 1985), and also the sub-Antarctic Auckland Islands (Kingsford et al, 1989).


Thyrsites atun snoek Fig. 27

Scomber atun Euphrasen, 1791:315 (Cape of Good Hope).
Thyrsites atun:
Penrith, 1967; Nakamura, 1986b.

DIAGNOSIS: Body distinctly elongated and compressed, the greatest depth contained 7-9 times in SL; mouth large, with fearsome teeth; lateral line single, abruptly bent downwards below posterior dorsal fin spines; caudal fin forked; dorsal fin divided to the base before soft-rayed part, with 18-21 spines followed by a short fin of 10-12 rays and 5-7 separate finlets; anal fin of 2 or 3 minute spines, 8-11 rays and 5-7 finlets; pelvic fins small, with 1 spine and 5 rays; pectoral fin rays 14-15; no fleshy keels on caudal peduncle or belly; 2 nostrils on each side of snout.

Colour: Silvery with iridescent purplish hue when fresh; fins dark grey.

MATERIAL: Tristan: RUSI 33480:43 mm TL; SAM 10502 (3); SAM 28233 (2). Gough: SAM 30439 (skull); SAM 30448 (1). Plus numerous specimens (maximum size 126 cm TL) examined at the islands.

REMARKS: This mesopelagic predator occurs year round in the waters of all the islands, including Gough. The population at the islands appears to be self sustaining with larvae, juveniles and breeding adults present. Spawning occurs in shallow water throughout the summer months (October to April). Adult snoek feed primarily on fish, mainly pelagic juveniles of Bovichtus diacanthus and Mendosoma linealum; during the summer months, exocoetids are also a common prey item.

DISTRIBUTION: T. atun is found close to shore throughout temperate waters of the southern hemisphere (Grant et al., 1978). It was not previously recorded from Gough Island.


Lepidopus caudatus, buttersnoek, Fig. 28

    Figure 28. Lepidopus caudatus, 510 mm SL, South Africa.

Trichiurus caudatus Euphrasen, 1788: 52, Pl.9, Fig. 2 (Cape of Good Hope, South Africa).
Lepidopus caudatus:
Nakamura, 1986c.

DIAGNOSIS: Body greatly elongated and compressed, the depth contained 11-15 times in SL; dorsal fin base runs entire length of body; no notch between dorsal fin spines and rays; dorsal fin with 9 slender spines and 88-100 rays; anal fin with 2 small spines (the second, plate-like, much shorter than pupil diameter) and 61-64 rays, the first 40 or so rays embedded; pectoral fin rays 12; caudal fin small and forked; pelvic fins reduced to a scale-like spine inserted close behind pectoral fin base; a single nostril on each side of snout; interorbital area concave or flat, the dorsal head profile concave; lateral line single; scales absent; gill-rakers 17-20. Vertebrae 105-114.

Colour: Silvery; fins pale.

MATERIAL: Tristan: RUSI 33596:127 cm TL. Nightingale: RUSI 18251:150 cm TL.

REMARKS: Only two specimens of this benthopelagic species were collected at the Tristan Group. This is a new record for the region. One other specimen was reported to have been caught by an islander in the past.

DISTRIBUTION: The species has a wide anti-tropical distribution extending down the eastern Atlantic from Iceland and Norway to the Mediterranean, South Africa, St Paul and Amsterdam Islands, Australia and New Zealand (Duhamel, 1984; Nakamura, 1986c).


Allothunnus fallai, slender tuna Fig. 29

    Figure 29. Allothunnus fallai, 265 mm TL, Nantibia.

Allothunnus fallai Serventy, 1948: 132, Fig. 1 (South Island, New Zealand); Collette, 1986.

DIAGNOSIS: Body elongate, torpedo shaped, the depth contained 4.5-6.5 times in FL; head length 3.7-3.9 times in FL; first dorsal fin of 17-18 slender spines depressible into a groove; second dorsal fin with 12-13 rays followed by 6-8 little finlets; anal fin of 13-14 rays, followed by 6-7 finlets; caudal fin forked; pectoral fin rays 24-26; gill-rakers (21-25) + (49-53) = 71-78. Vertebrae 20+19.

Colour: Bluish dorsally, white ventrally.

MATERIAL: Tristan: RUSI 31527, 265 mm TL.

REMARKS: One juvenile was collected on the surface at Tristan during 1989. The slender tuna was not previously recorded from the region, but it has a worldwide distribution between 20 degrees and 50 degrees S. (Collette, 1986).


Hyperoglyphe antarctica, southern butterfish, (bluefish), Fig. 30

    Figure 30. Hyperoglyphe antarctica, 500 mm SL, Tristan.

Perca antarctica Carmichael, 1819: 501, P1. 25 (Tristan da Cunha).
Seriollella antarctica: Regan, 1913.
Seriolella antarctica:
Sivertsen, 1945.
Palinurichthys porosus:
Barnard, 1948, Smith, 1949.
Mupus perciformis:
Smith, 1949.
Palinurichthys antarcticus:
Penrith, 1967.
Hyperoglyphe antarctica: Haedrich, 1967, 1986; Duhamel. 1989.

DIAGNOSIS (based on 16 specimens, 390-884 mm SL): Body stout, little compressed, the depth 32-37% SL; head length 32-35% SL; least depth of caudal peduncle 47-61% of its length (from anal fin base to ventral end of caudal fin base); dorsal fin divided to the base, the front part low, with 7 or 8 short stout spines, not increasing in length posteriorly and less than half length of anterior soft-rays in second dorsal fin; dorsal fin origin on or slightly behind a vertical at upper end of pectoral fin base; second dorsal fin with 1 hidden spine and 17-20 rays; anal fin with 3 spines (hidden at front margin of fin) and 14-16 rays; pectoral fin long and falcate, reaching vertical at anal fin origin (pectoral length 27-32% SL), the rays 19-22; top of head naked except for a small patch of scales on each side above and behind eyes; lateral-line arched anteriorly, curving to mid-lateral position below middle dorsal fin rays, thence running straight to tail base; about 80-89 pored lateral-line scales; gill-rakers 6-7 + 16-18; vertebrae 25.

Colour: Body dark blue-grey fading to grey below; fins dusky.

MATERIAL: Tristan : RUSI 33565:495 mm SL; RUSI 34597:425 mm SL; RUSI 34598:528 mm SL; RUSI 34599:440 mm SL. Gough: RUSI 18246: 2, 385-420 mm SL; SAM 23311: 432 mm SL; 30449 (1). Many specimens from the Tristan islands and Gough were examined in the field but not preserved.

REMARKS: Hyperoglyphe antarctica occurs year round at Tristan in depths of 40 to over 150 m. Fish more than about 45 cm TL are epibenthic in shoals; juveniles are pelagic and not captured by hook and line. They feed predominantly on cephalopods and colonial salps, but they will also take fish prey.

In New Zealand, H. antarctica reaches approximately 100 cm TL and 14 kg in 9 years (Horn, 1988) and attains a maximum size of at least 137 m FL and a weight of 36 kg (McDowall, 1982). At Tristan, specimens of 80 cm TL are sexually mature, and spawning occurs during January and February.

DISTRIBUTION: H. antarctica has a wide Southern Hemisphere distribution. In addition to the Tristan Group, this species is known from Gough Island and Vema Seamount (Penrith, 1967), South Africa (Haedrich, 1986), the Mozambique Plateau and Southwest Indian Ridge (Shcherbachev, 1987), St Paul/Amsterdam Islands (Duhamel, 1989), southern Australia, New Zealand, Kermadec Islands (McDowall, 1982), and the Chatham Islands east of New Zealand (Horn, 1988; Roberts, 1991).

Schedophilus velaini, oval driftfish, (stumpnose). Fig. 31

    Figure 31. Schedophilus velaini, 430 mm SL, Tristan.

Seriollela velaini Sauvage, 1879: 32, Pl. 1, Fig. 2 (St Paul Island).
Leirus moselii Cunningham, 1910: 95, Pl. 5 (St Helena Island).
Seriolella christopherseni Sivertsen, 1945: 23, Fig. 16, Pl. 1 (Tristan da Cunha)
Mupus imperialis (non Cocco, 1840): Smith, 1949; Penrith, 1967.
Schedophilus ovalis (non Cuvier, 1833): Barnard, 1948; Haedrich, 1967; 1986; Stehmann & Lenz, 1973; Edwards & Glass, 1987; Duhamel, 1989.
Schedophilus labyrinthicus
McAllister & Randall, 1975: 1, Fig. 1 (Easter Island)
Seriolella labyrinthica:
McDowall, 1982:117, Fig. 7.

DIAGNOSIS (compiled from Bamard, 1948; Stehmann and Lenz, 1973; McDowall, 1982; and Duhamel, 1989; data from our specimen in parentheses): Body firm, greatest depth 2.6-3.1 (2.9) times in SL; head length 3.2-3.8 (3.5) times in SL; a single dorsal fin with 6-8 (7) weak spines graduating in length to the soft-rays and 26-29 (28) rays; anal fin with 3 spines (hidden at front margin of fin) and 19-21 (19) rays; pectoral fin elongate and pointed in adults, with 21-22 (22) rays; lateral-line scales 90-110 (90); gillrakers 6-7 + 17 (7+17). Vertebrae 10+15 (11+15); supraneural (predorsal) bone formula (of Ahlstrom et al., 1976) 0/0/2+1+1/1+1/.

Colour: Olive green dorsally, silvery on sides and below.

MATERIAL: Tristan: RUSI 34596, 430 mm SL. Numerous specimens of 45-73 cm TL examined at the islands but not preserved.

REMARKS: Schedophilus velaini is relatively common at the Tristan Group, and is often caught amongst bluefish. It is generally epibenthic, but occasionally feeds higher up in the water column; stomach contents include planktonic salps, cephalopods and small fish. S. velaini is seldom caught in waters shallower than about 40 m. Spawning occurs in late summer. According to Haedrich (1986), this species attains 100 cm TL, but the largest specimen caught at these islands is only 73 cm TL.

McAllister and Randall (1975) pointed out that specimens of S. ovalis (Cuvier, 1833) from the North Atlantic Ocean have more anal fin-rays (21-23) compared with those from the Southern Hemisphere (18-20). This difference is also supported by data supplied by Dr C. Karrer: 14 specimens from the North Atlantic with anal fin counts of 20(1), 21(5) and 22(8). McAllister and Randall's (1975) distinction of their new species (S. labyrinthicus) from S. christopherseni and S. velaini based on a supposed difference in the number of anal fin-rays (18-19 in labyrinthicus [n= 4], vs 20 in the latter two species [n= 2]) does not hold up in the light of our data and those presented by Duhamel (1989) for additional specimens from St. Paul/Amersterdam Islands. We conclude therefore, that S. labyrinthicus and S. christopherseni are synonyms of S. velaini and agree with Dr Karrer (pers. corem.) that S. velaini is distinct from S. ovalis.

Like Hyperoglyphe perciformis (and most other stromateioid fishes), the postlarvae and small juveniles of Schedophilus velaini are associated with flotsam and pelagic jellyfishes in the open ocean (Ahlstrom et al., 1976). This habit accounts for the worldwide distribution of these species.

DISTRIBUTION: S. velaini is known from St Helena, Vema Seamount, South Africa, St. Paul/Amsterdam Islands, Australia, Lord Howe, New Zealand, Rapa Island, the Juan Fernandez Islands and Chile (Haedrich, 1986; Duhamel, 1989).


Arnoglossus capensis, Cape flounder, (sole), Fig. 32

    Figure 32. Arnglossus capensis, 113 mm SL, Nightingale Island.

Arnoglossus capensis Boulenger, 1898:1 (False Bay, South Africa); Hensley, 1986.

DIAGNOSIS (supplied by D.A. Hensley from our specimens): Dorsal fin rays 98-104; anal fin rays 76-82; pectoral fin rays (ocular side) 11; gill-rakers 0 + 10-12; lateral line ca. 61-63. Body depth 2.1-2.2, head length 3.9-4.2 times in SL. Scales of ocular side ctenoid; head profile without pronounced notch at level of upper border of lower eye; interorbital region concave; dorsal fin rays 2-5 slightly longer than adjacent rays; maxillar reaching slightly posterior to front edge of lower eye; upper jaw length 3.1-3.4 times in head length.

Colour: Head, body and fins with small indistinct dark spots.

MATERIAL: Nightingale: RUSI 29061, female, 94 mm SL. Gough: RUSI 18238, males, 113 and 114 mm SL; RUSI 29062, female 129 mm SL.

REMARKS: The specimen from Nightingale was collected by Robert Tarr on a sand bottom. The two males from Gough were collected by Wyn and Uys from the stomach of Helicolenus mouchezi, and the female was caught in 42 m with a dredge. Our specimens were identified by D.A. Hensley; they are the first record of a bothid from the Tristan Group and Gough Island.

DISTRIBUTION: Arnoglossus capensis is known from southern Africa (from Natal to Angola), St Helena and Ascension Island. According to Hensley (1986), it may also be conspecific with A. entomorhynchus Stauch, 1967 from the Gulf of Guinea.



Sphyrna rookarran great hammerhead, (hammerhead shark), Fig. 33

    Figure 33. Sphyrna mokarran, 187 cm, immature male, South Africa (from Bass, 1986c).

Zygaena mokarran Ruppell, 1837: 66, Pl. 17 (Red Sea).
Sphyrna mokarran:
Compagno, 1984.

DIAGNOSIS: Head flattened and greatly expanded laterally with the eyes on the ends of the lateral expansions; front edge of head with a median indentation; rear edge of pelvic fins markedly concave; second dorsal fin relatively large, its height greater than length of 3rd gill slit; teeth distinctly serrate.

REMARKS: This warm-water species occurs in Tristan waters only from November to April when water temperatures are at their highest (16 - 20 C). It has not been recorded from the islands before. During these months this shark is commonly seen feeding on the surface often in pursuit of exocoetids, which are also only present in the region during the warm water months. S. mokarran is wide ranging in tropical seas close inshore and in the open ocean (Compagno, 1984). No hammerhead was captured during our field work. Identification of this species is based on numerous observations at sea. This was possible because S. mokarran has a high second dorsal fin that is easily visible when the shark swims at the surface. The other two Sphyrna species that are known from the South Atlantic have a much lower second dorsal fin that would not project much above the surface of the water (Compagno, 1984).


Mola sp.

Ocean sunfish have been sighted on three occasions around the Tristan islands. The specimens were too large to capture. Ocean sunfish are found in all oceans, but not in polar seas (Heemstra, 1986g).



?Alepisaurus ferox, lancetfish, Fig. 34

    Figure 34. Alepisaurus ferox, 100 cm SL (after Gibbs, 1960).

?Alepisaurus ferox Lowe, 1833:104 (seas off Madeira); Rowan & Rowan (1955); Heemstra & Smith, 1986.

DIAGNOSIS: Body greatly elongated, compressed, without scales; fleshy keel along each side of caudal peduncle; dorsal œm high and long, with about 40 rays; anal fin much shorter, with 13-18 rays; dorsal adipose fin above posterior  anal fin rays; caudal fin forked; mouth large, with fearsome


MATERIAL: None collected by us.

REMARKS: Rowan and Rowan (1955) reported "Alepisaurus ferox" from Tristan, but they gave no diagnostic information on their specimen. This fish could be either A. ferox or A. brevirostris Gibbs, 1960. The latter species is common at the sub-Antarctic islands (Gon, 1990), and both species belong to the lower epipelagic fauna of tropical and temperate waters of all oceans (Heemstra and Smith, 1986).


?Symbolophorus barnardi

Myctophum humboldti: Brauer, 1906:192 (in part); Barnard, 1925:243 (in part); Sivertsen, 1945.
Myctophum humboldti barnardi
Taning, 1932:128 (off Cape Point, South Africa).
Symbolophorus barnardi:
Hulley, 1986.

DIAGNOSIS (from Hulley, 1986): Mouth subterminal, the maxilla moderately expanded posteriorly and reaching well past eye; anal fin base longer than dorsal fin base; pectoral fin unpigmented; dorsal fin rays 13-15; anal fin rays 21-23; pectoral fin rays 12-14; gill-rakers (4-6)+(12-14) = 17-20.

REMARKS: Sivertsen's (1945) record of "Myctophum Humboldti" was based on a specimen dropped by a tern (Sterna vittata), which had been feeding close to Tristan Island. His identification was taken from descriptions by Bamard (1925) and Brauer (1906), and according to Hulley (1986), these descriptions apply (at least in part) to Symbolophorus barnardi. It is a high-oceanic mesopelagic species that comes to the surface at night.

DISTRIBUTION: Circumglobal between 30 degrees and 40 degrees S.


?Trachipterus trachypterus, peregrine ribbonfish

Cepola trachypterus Gmelin, 1789:1187 (Adriatic Sea).
Trachipterus trachypterus:
Heemstra & Kannemeyer, 1984, 1986.

DIAGNOSIS (from Heemstra and Kannemeyer, 1986): Body extremely elongate and compressed, tapering evenly to a minute caudal fin that projects dorsally from the slender tip of the tail; dorsal fin with 145-184 rays, the fin base extends from head to caudal fin; anal fin absent; scales absent, but ventral edge of body with enlarged pointed tubercles; no membranous flaps on dorsal and pelvic fin rays of juveniles; pelvic fins absent in adults; pectoral fin with 13-14 rays; eye diameter 8.4-10% of distance from snout to anus and contained 1.4-1.6 times in length of lower jaw.

Colour: Head and body silvery; front of head, interorbital area, dorsal part of upper jaw and tip of lower jaw are black; young with 3-5 dark blotches on sides of body; dusky blotch usually visible on body at bases of 20-28th dorsal fin rays.

REMARKS: One specimen (length not given) was taken close to Gough Island in an RMT-2 midwater trawl in October 1980 (Miller, 1982). Our efforts to locate this specimen have been unsuccessful. This offshore pelagic species is rare but widely distributed from the Mediterranean to South Africa, Japan and New Zealand (Heemstra and Kannemeyer, 1986).



1) Eleginops maclovinus (Cuvier, 1830).

E. maclovinus is known from Southern Chile, Argentina and the Falkland Islands (Nakamura, 1986d). It attains 60 cm SL and is a popular food fish at the Falkland Islands. One specimen, 231 mm SL, SAM 22892, reported from Tristan by DeWitt et al. (1990: 285). There is no collector or date of collection listed in the museum register for this specimen, and it is the only one known from the island. Some of the Tristan fishermen who have been to the Falkland Islands are familiar with E. maclovinus, and they say that this species does not occur at Tristan. We suspect that the South African Museum specimen may have been mislabeled, and that it actually was caught at the Falkland Islands.

2) Notothenia coriiceps Richardson, 1844.

A similar case is the specimen of Notothenia coriiceps (SAM 22892), which is also listed in the Museum's register from Tristan, but without collector or date. N. coriiceps occurs at sub-Antarctic islands all round the Southern Ocean (DeWitt et al., 1990). It is a common shallow-water fish that is easy to see and to collect. We did not see any specimens of N. coriiceps at Tristan, and the fishermen there also have not seen this species.


Most of the 51 fish species reported from the Tristan Group and Gough Island are widely-distributed species that are known from several continental and insular localities in warm-temperate waters of the Southern Hemisphere. Situated within the Subtropical Convergence Zone, the Tristan Group represents the southern-most "way station" for many wide-ranging temperate and eurythermic subtropical species. A knowledge of the fish fauna from a frontal zone such as the STC yields valuable information about the factors affecting general dispersal and distribution of fishes in the oceans. Comparison of the fish fauna from the Tristan da Cunha Group with that of Gough Island, which lie on the northern and southern limits of the STC respectively, is interesting, because the proximity of the islands and the different temperature regimes at the two localities enable us to determine how temperature affects the distribution of fishes in this oceanic region. Twenty-six of the 28 neritic species known from the region were recorded from the Tristan Group, but only 19 occur at Gough (Table 1).

Table 1. Fishes recorded from the Tristan Group (T) and Gough Island (G). New records for Tristan or Gough are indicated with an asterisk.


I) Shallow-water (neritic) species T G 1) * Notorynchus cepedianus + + 2) * Carcharodon carcharias + 3) * Torpedo nobiliana + 4) * Bassanago nielseni + 5) * Conger wilsoni + 6) Gnathophis capensis 7) Gaidropsarus novaezelandiae + 8) * Physiculus karrerae 9) Beryx decadactylus 10) * Centriscops obliquus + 11) Notopogon lilliei + + 12) Helicolenus mouchezi + + 13) Sebastes capensis + + 14) * Polyprion oxygeneios + + 15) * Lepidoperca coatsii + + 16) Pseudopentaceros richardsoni + 17) * Emmelichthys nitidus + 18) Seriola lalandi + 19) Trachurus longimanus + + 20) Acantholatris monodactylus + + 21) * Latris lineata + + 22) * Mendosoma lineatum + + 23) Nelabrichthys ornatus + + 24) Bovichtus diacanthus + + 25) Thyrsites atun + + 26) Hyperoglyphe antarctica + + 27) Schedophilus velaini + 28) * Arnoglossus capensis + + II) Oceanic and mesopelagic species 29) * Hexanchus griseus + 30) * Etmopterus princeps + 31) Prionace glauca + +? 32) * Isurus oxyrinchus + 33) * ?Sphyrna rookarran + 34) * Vinciguerria poweriae + 35) Maurolicus inventionis + + 36) Electrona paucirastra + 37) Electrona subaspera + 38) Lampadena dea + 39) ? Symbolophorus barnardi + 40) ? Alepisaurus ferox + 41) Cheilopogon pinnatibarbatus + 42) * Hirundichthys rondeletti + 43) Scomberesox saurus + 44) Lophorus lacepede + 45) Trachipterus trachypterus + 46) * Brama australis + 47) * Naucrates ductor + 48) * Remora remora + 49) * Lepidopus caudatus + 50) * Allothunnus fallai + 51) * Mola sp + -----------------------------------------------

The STC has been noted to act as a substantive biogeographic bamer (Lutjeharms, 1987). In addition, it has been observed that the STC itself is an area of enhanced primary productivity and phytoplankton density (Allanson et al., 1981) which might influence organisms in higher trophic levels through the food chain. Recent research (Stramma and Peterson, 1990) has furthermore shown that a strong current, the South Atlantic Current, flows along the full longitudinal extent of the STC and it is proposed that this current enhances the dispersal of pelagic stages of fishes in the STC region beyond that of the general eastward drift.

The differences between the fauna at Gough and Tristan are also evident in the oceanic species that move into the Tristan area from the north during the warmer summer months. Of the 15 epipelagic species that have been reported from Gough and Tristan, only 3 or 4 have been found at Gough. Although sampling at Gough was less intense than at Tristan during the present study, it is apparent that several resident neritic species (eg. Schedophilus velaini and Physiculus karrerae) are only found at the northern islands. This indicates that there is a strong isolating factor preventing certain species from reaching Gough Island, only 200 miles to the south of Tristan. Mean monthly water temperatures at Gough are on average 3 C lower than Tristan during all months (Fig. 5). The subantarctic islands of the Prince Edward and Crozet groups, which are at 46 degrees S latitude in the Indian Ocean and just south of the Antarctic Convergence, have no fishes in common with the areas close to the STC (Gon and Heemstra, 1990). The sea surface temperatures at these subantarctic islands range from 4 to 8 C. That 21 of the 51 species known from Tristan and Gough are found only within or north of the STC indicates that this subtropical frontal zone is an effective barrier to the north-south dispersal of animals in the Southern Ocean. As suggested by Sivertsen (1945) and Penrith (1967), water temperature appears to be the primary factor affecting the north-south distribution of fishes in the region.

For the purposes of zoogeographical analysis, the fishes of Tristan and Gough (T-G fishes) were divided into a neritic (coastal) and an oceanic category; the latter category includes mesopelagic and offshore epipelagic fishes that have been taken within a mile of these islands. Penrith (1967) used the same categories, although some of the fishes that he considered oceanic species are here included in the neritic group, due to biological information that is now available. Our neritic group comprises those species that appear to be restricted to the islands during their adult stage. The oceanic group includes species that are wide ranging and common, as adults, in the pelagic zone of the oceans. Once again it is stressed that, although some oceanic species are included in our list of T-G species (Table 1), only neritic species are used in the zoogeographic analysis that follows.

Table 2. Neritic fishes from the Tristan Group and Gough island, showing occurrence at other areas close to the Subtropical Convergence. T = Tristan Group, G = Gough, V = Vema Seamount, AF = southern Africa, P/A = St Paul and Amsterdam Islands, ANZ = Australia and New Zealand, SA = South America. * = new record for Tristan Group or Gough Island. + = species present; g = genus present.

     NERITIC SPECIES            T    G    V    AF  P/A   ANZ   SA 
* Notorynchus cepedianus        +    +         +          +    + 
* Carcharodon carcharias             +         +    +     +    + 
* Torpedo nobiliana             +              +    g     g    g 
* Bassanago nielseni            +              g    +     g    g 
* Conger wilsoni                     +         +          +    g 
Gnathophis capensis             +              +    +     g    g 
Gaidropsarus novaezelandiae     +    +         +    +     +    + 
* Physiculus karrerae           +                              + 
Betyx decadactylus              +              +          +    + 
* Centriscops obliquus          +    +         +    +     +    + 
Notopogon lilliei               +    +         +    g     +    g 
Helicolenus mouchezi            +    +         g    +     g    g
Sebastes capensis               +    +         +               g 
* Polyprion oxygeneios          +    +    g    g    +     +    + 
Lepidoperca coatsii             +    +              +     g    g 
Pseudopentaceros richardsoni    +              +          +    + 
* Emmelichthys nitidus          +              +    +     +    + 
Seriola lalandi                 +         +    +    +     +    + 
Trachurus longimanus            +    +    +    g    +     g    g 
Acantholatris monodactylus      +    +    +         +          g 
Latris lineata                  +    +              +     + 
Mendosoma lineatum              +    +              +     +    + 
Nelabrichthys ornatus           +    +    +         + 
Bovichtus diacanthus            +    +              g     g    g 
Thyrsites atun                  +    +    +    +    +     + 
Hyperoglyphe antarctica         +    +    +    +    +     + 
Schedophilus velaini            +         +    +    +     +    + 
*Arnoglossus capensis           +    +         +
The zoogeographic affinities of the T-G fish fauna are indicated by the number of neritic species and genera that are shared with other areas (Table 3).

Table 3. The number of species, and the percentage species and genera of Tristan + Gough neritic fishes at other areas close to the STC. T-G = Tristan and Gough; V = Vema Seamount; AF = southern Africa; P-A = St Paul and Amsterdam Islands; ANZ = Australia and New Zealand; SA = South America.

                       T-G      V       AF      P-A      ANZ       SA 
T-G/Total              28      6/17   18/500+   18/29   16/600+   14/300+ 
%T-G species/% Total  21/35    64/4   64/62     57/3    50/5 
% Genera               27       77     67        79      80

Distributional records of neritic species occurring at Gough and the Tristan Group were compiled in order to characterise a shallow-water Subtropical Convergence ichthyofauna (Table 2). These records were taken from the present study, as well as from published checklists of fishes from other shallow-water areas in the Southern Ocean. These included Vema Seamount (Penrith, 1967; Heemstra, ms.), South Africa (Smith and Heemstra, 1986), St. Paul and Amsterdam Islands (Angot, 1951; Huteau, 1969; Beurois, 1976; Duhamel, 1989), Walters Shoal (Collette and Parin, 1991), Austral Seamount (Duhamel, 1984), South Australia (Maxwell, 1980; Last et al., 1983), New Zealand (Ayling and Cox, 1982) and the southern parts of South America (Menni et al., 1984; Nakamura et al., 1986).

Heemstra (ms.) has discussed the zoogeography of the fishes at Vema Seamount, a submerged volcano located at 31 degrees 38'S, 08 degrees 20'E, about 650 km off the west coast of South Africa. Although the fish fauna of Vema is poorly sampled, the zoogeographic affinities of the 17 neritic species that are presently known from Vema, like those of the crustaceans (Kensley, 1980) and other invertebrates (Berrisford, 1969), are primarily with Africa. Vema is well north of the Subtropical Convergence Zone, and the water temperatures on the plateau (18-21 C at depths of 26-73 m) are distinctly warmer and less variable than those of the Tristan Group. The fish fauna at Vema is a mixture of tropical and warm temperate species. Although Vema has 6 (21%) of the neritic species recorded from Tristan and Gough, it has 13 species in common with southern Africa. Three of the 6 T-G species shared with Vema (Seriola lalandi, Hyperoglyphe antarctica, and Schedophilus velaini) are widespread species that also occur along the African coast and elsewhere in the Southern Hemisphere. The remaining three species (Trachurus longimanus, Acantholatris monodactylus, and Nelabrichthys ornatus) are known only from the Westwind Drift Islands, Walters Shoal and the Austral Seamount.

Collette and Pann (1991) reported 20 species of neritic fishes from Walters Shoal, a shallow seamount located 740 km south of Madagascar and 1111 km east of South Africa. Based on the presence of Serranus novemcinctus, T. longimanus, A. monodactylus, and N. ornatus (which also occur at St Paul and Amsterdam Islands) and their misinterpretation of the zoogeographic affinities of Walters Shoal and Vema Seamount, Collette and Parin (1991) concluded that 'The fish fauna of Walters Shoals links the Tristan-Gough Province of Briggs's (1974) southern South America Cold Temperate Region with the Amsterdam-St Paul Province of his Southern African Warm Temperate Region into a single zoogeographic province, herein named the West Wind Drift Islands Province. There are no Antarctic or subantarctic species and there is little similarity to South African fishes."

The fish faunas of Vema Seamount and Walters Shoal are mixtures of tropical and warm-temperate species; 11 of the 20 species reported by Collette and Parin are tropical coral-reef species that are not known from the Westwind Drift Islands (sensu stricto). In fact, Walters Shoal shares more species with South Africa (9) than it does with the Westwind Drift Islands (5 or 6).

The "cladistic" or vicariance method of biogeographic analysis (Humphries and Parenti, 1986) assumes that disjunct distributions are the result of "vicariance events" (development of geological or climatological barriers) that fragment ancestral species ranges. This method also requires acceptable cladistic classifications of the relevant taxa. Since the Tristan Group and Gough are relatively young (6-18 m.y., Baker, 1973) volcanic islands that originated at or near the mid-Atlantic Ridge, the zoogeographic affinities of the T-G fishes cannot be the result of some vicariance event occurring more than 20 m.y. ago. We assume, therefore, that the present neritic fish fauna is a product of transilient ("jump") dispersal of species from another (older) fauna and subsequent evolutionary divergence.

The circumpolar distribution of many neritic T-G fishes is indicative of complex dispersal mechanisms. Species that are bound to shallow waters throughout their life cycle, or at least in the adult form, have evolved mechanisms to ensure dispersal to distant regions while maintaining a breeding population in the source region. It is hypothesised that dispersal via eggs, larvae, and (in some cases) pelagic juveniles, has resulted in the ichthyofaunal similarities at these widely-separated localities. In the case of the T-G species, this implies that their larvae and juveniles occur in the pelagic zone for long periods (from two months to a year) before metamorphosis into the benthic form occurs. At least 13 of the widely-distributed T-G species have prolonged pelagic stages (Table 4). The relatively large size that these species attain before adopting a benthic or epibenthic lifestyle allows time for dispersal over considerable distances.

The three congrid eels, Bassanago nielseni, Conger wilsoni and Gnathophis capensis, the first recorded from Tristan and the Nasca Ridge, the second from Gough, South Africa and Australasia, and the third from Tristan, South Africa and Amsterdam Island, have relatively long-lived leptocephalid larvae (Castle, 1986), which would facilitate dispersal within the Antarctic Circumpolar Current (ACC) between these distant points.

Mundy (1990) illustrated the larvae and pelagic juveniles of both species of Beryx. The pelagic juveniles of B. decadactylus attain at least 23 mm SL. Dr G. Duhamel (pets. comm.) suggests that Centriscops obliquus remains pelagic until a relatively large size.

Some species of Sebastes have a pelagic juvenile stage that lasts for up to a year (Moser and Boehler, 1991), and the widespread scorpaenid species, Helicolenus mouchezi and Sebastes capensis are likely to have similarly long pelagic stages (Moser et al., 1977; Sanchez and Acha, 1988).

Polyprion oxygeneios, another suprabenthic species, has pelagic juveniles that occur near floating objects until they attain at least 30 cm TL (Roberts, 1989). This T-G species is also found off St Paul and Amsterdam, Southern Australia, New Zealand and South America. The two widely-distributed centrolophid species, Hyperoglyphe antarctica and Schedophilus velaini, are also pelagic until at least 30 cm TL (Haedrich, 1986; Horn, 1988).

Relatively large pelagic stages of Acantholatris monodactylus (6 cm TL), Mendosoma lineatum (9 cm TL), and Gaidropsarus novaezelandiae (3 cm TL), species which are also widely distributed, were encountered during the present study. Bovichtus diacanthus has a pelagic juvenile stage that is terminated at a size of approximately 6 cm TL. Adults of all these species and congenerics of B. diacanthus occur at widely separated areas of shallow water in the Southern Ocean close to the STC.

Table 4. Reproductive characteristics of benthic or epibenthic shallow-water fishes at Tristan and Gough Island. BS = broadcast spawner. A "long" pelagic stage is of two months or longer duration.

Species              Reproductive   Fecun-  Egg   Pelagic   References 
                            mode     dity   size  stages 
Bassanago nielseni           BS      high   small  long   Thresher 1984 
Conger wilsoni               BS      high   small  long   Castle 1986 
Gnathophis capensis          BS      high   small  long   Castle 1986 
Gaidropsarus novaezelandiae   ?       ?       ?    long   present study 
Physiculus karrerae           ?       ?     small   ?     present study 
Beryx decadactylus            ?      high   small  long   Mundy 1990 
Centriscops obliquus          ?      high   small  long   Duhamel (pers. 
Notopogon lilliei             ?      high   small   ?     present study 
Helicolenus mouchezi    live bearer  high   small  long   Thresher 1984;    
                                                              present study

Sebastes capensis       live bearer  high   small  long   Moser et al.      
                                                        1977; present study

Polyprion oxygeneios         BS      high   small  long   Roberts, 1989 
Lepodoperca coalsii           ?       ?       ?     ? 
Pseudopentaceros richardsonii ?      high   small   ?     present study 
Emmelichthys nitidus         BS?      ?       ?     ?     Thresher 1984 
Acantholatris monodactylus   BS      high   small  long   present study 
Latris lineata                ?      high   small   ?     present study 
Mendosoma lineatum            ?      high   small  long   present study 
Nelabrichthys ornatus   protogynous, high   small   ?     present study 
Bovichthus diacanthus        BS?     high   small  long   present study 
Hyperoglyphe antarctica      BS      high   small  long   Haedrich 1986;    
                                                             present study

Table 4 lists the reproductive characteristics of neritic T-G species that have benthic or epibenthic adults. The reproductive biology of only a few species is known in detail; for the other species, common trends for the family or genus are taken from the literature. These species display a wide range of reproductive strategies including live-bearing, protogynous hermaphroditism, and pelagic broadcast spawning. Sinclair and Tremblay (1984) have shown that an increase in fecundity occurred with an increase in the duration of the larval stage in different populations of Atlantic herring. They suggested that an increase in fecundity is necessary to compensate for the greater losses of larvae from the spawning area that occur as a result of longer periods in the pelagia, especially in poor retention areas. Therefore, observations from the present study were used to broadly categorise fecundity and egg size of the various species as being high or low, and small or large, respectively. The presence or absence of an extended pelagic stage is listed in Table 4 from observations based on the present study or from the literature.

Sixty-five percent of the benthic or epibenthic species listed in Table 4 have an extended pelagic stage. This is a high percentage, given that the reproductive biology of most of the remaining 35% of these species is unknown. All of the species with extended pelagic stages have a high fecundity and small egg size, and 80% of all the species listed above display this characteristic. Because high fecundity and small egg size are correlated with extended pelagic stages, we predict that many of the neritic T-G fishes whose early life-history is unknown will also be found to have similar prolonged pelagic stages. It is apparent that most of the benthic oriented T-G species have a long pelagic stage, and that this common trait is present in unrelated species displaying a wide range of reproductive modes. There is strong evidence, therefore, that an extended pelagic stage is important for the occurrence and survival of the majority of the benthic shallow-water species at the islands.

It is unlikely that the life cycles of T-G species include a complete circumpolar cycle. That is, for spawning to occur at one point and settlement to occur at the same place, after the pelagic stages have circled the southern ocean in the ACC. It is also improbable that the pelagic stages are retained in the enormous (and relatively slow-moving) oceanic gyres of the South Atlantic or Indian oceans, and that they return to the area where they were spawned via these circulations; this remote possibility is not considered viable for the maintenance of breeding populations.

The literature on early life history stages of oceanic island fish populations often places more emphasis on the retention (rather than the dispersal) of propagules in the vicinity of spawning areas. Authors such as Boden (1952), Boden and Kampa (1953) and Sale (1970) have concluded that physical oceanographic phenomena such as current eddies and upwellings often enhance retention rather than dispersal from the source areas. Leis (1982) stressed that diverse mechanisms are utilised by the pelagic stages themselves, in conjunction with oceanographic phenomena, to ensure retention in the source area. At present, not enough information exists to define localised current patterns in the Tristan da Cunha and Gough Island region, or to determine the capacity that pelagic stages have to actively utilise these currents to ensure retention. It is suggested, however, that the majority of propagules remain in the source area, and that dispersal to other distant shallow-water areas has occurred through the occasional alteration of these local hydrographic conditions (or perhaps the regular escape of a small number of larvae and juveniles from the system).

Sinclair (1988) cautioned against the use of surface oceanographic phenomena as explanations for pelagic-stage dispersal and emphasised that movements of propagales within a current system may greatly affect the speed and hence distance of dispersal. Nevertheless, we believed that it would be useful to determine the distance that pelagic stages of T-G species could theoretically disperse passively in the eastward-moving surface current systems if they became separated from local retention areas. Trajectories of FGGE drifting-buoy experiments in the Southern Ocean have shown that surface current speeds of the ACC at the STC are about 26 cm/s (Hoffman, 1985). Tins current velocity would allow pelagic stages to be transported passively up to 4088 km in six months. If these drift rates are considered in relation to the relatively long-lived larval and juvenile stages of T-G fishes, it appears that dispersal over the great distances (e.g., from South America to Tristan, 4023 kin, or Tristan to St Paul, 4680 kin) between shallow-water areas in the Southern Ocean could be possible, if the species are tolerant of other physical oceanographic conditions, such as temperature. Localities such as St Paul and Amsterdam Islands have remarkably similar environmental conditions to Tristan da Cunha and Gough Island. Apart from the similarities in fish fauna, St Paul and Amsterdam also support dense beds of Macrocystis pyrifera, large populations of the lobster Jasus paulensis (Duhamel, 1989), and display sea temperatures very similar to Tristan and Gough (David and Guerin-Ancey, 1990).

We propose that the benthic T-G fish fauna was rounded by means of pelagic larvae and/or juveniles that drifted via the Westwind Drift (ACC and South Atlantic Current) to Tristan from source populations along the Atlantic coast of South America. This hypothesis predicts that (1) pelagic larvae and/or juveniles of the species that occur at South America and Tristan will be found "upstream" (west) as well as "downstream" (east) of Tristan, (2) pelagic larvae and/or juveniles of T-G species that do not occur at South America will be found only downstream (east) of Tristan, (3) the fish fauna of St Paul and Amsterdam Islands (which are downstream of Tristan) are derived from the fauna at Tristan and South Africa, hence the sister taxa of T-G taxa will be found at St Paul and Amsterdam Islands, axed the sister taxa of these two (T-G + P-A) taxa will be found at South America, Vema or South Africa.

Although there are no cladistic studies of T-G fishes, there are some indications of relationships in the currently accepted classifications of the genera Helicolenus and Acantholatris. According to Eschmeyer and Hureau (1971), H. mouchezi, which is known only from T-G, P-A and the Austral Seamount west of P-A, is more closely related to H. dactylopterus lahillei from Uruguay than to H.d. dactylopterus of South Africa. The present generic classification of cheilodactylid fishes is confused (Allen and Heemstra, 1976; Smith, 1980), but based on superficial features (length of caudal peduncle and number of anal fin rays), it appears that Cheilodactylus bergi Norman, 1937 from the southern coast of South America may be the sister species of Acantholatris monodactylus and A. vemae.

Pollock (1990) suggested that the South Atlantic may have been the source area from which the palinund lobster Jasus paulensis at Amsterdam and St Paul Islands originated. This species and J. tristani from Tristan, Gough and Vema Seamount are closely related and could perhaps still be regarded as conspecific (Pollock, 1990). He suggested that a northward latitudinal shift of from 2 degrees to 4 degrees of the STC combined with an increase in velocity of the ACC during recent glacial periods, as proposed by McIntyre et al. (1976) and Kennett (1982), would increase the chance of dispersal of long-lived larvae in an easterly direction from the Atlantic into the indian Ocean. This explanation might also be true for the wide distribution of STC fishes. A slight shift in the position of the STC to the north would effectively place fish propagules that are now virtually unaffected by the SAC and the ACC, in their path and facilitate a circumpolar dispersal in an easterly direction. St. Paul and Amsterdam islands have been estimated to be in the order of 700 000 years old (Gunn and Abranson, 1971) while the Tristan group and Gough Island are at least 18 million years old (Baker, 1973). For this reason it would seem probable that the Indian Ocean Islands were colonised from the South Atlantic Islands and Southern Africa. Also, as Pollock (op cit.) suggests, a number of glacial periods have occurred during the last 700 000 years with the last drawing to an end as recently as 15 000 years ago. Therefore, the last colonisation of St. Paul and Amsterdam by fish could have been relatively recent. It is probable that conspecific populations at the widely spread areas of shallow water in the STC region are now dependant mainly on the retention of local propagules rather than distant recruitment.

The original source areas of the widespread resident species at the oceanic islands must have been the three ancient continental land masses of Southern Africa, Australasia and South America to a greater or lesser degree. The volcanic islands of Tristan da Cunha, St Paul and Amsterdam and seamounts such as Vema and Walters Shoal were colonised from these areas, and are thought to have acted as stepping stones for dispersal, once breeding populations had established themselves there. The relatively high affinities that all the areas have with each other in terms of common genera could be explained by subsequent speciation after the initial dispersal, perhaps followed by further dispersal between them. The systematic relationship that the various genera have to one another needs to be addressed in order to define source areas with certainty.

The prolonged pelagic stages present in many of the species of the isolated Westwind Drift islands is probably responsible for the relatively low level of endemism at Tristan da Cunha, Gough Island and St Paul and Amsterdam Islands in the Indian Ocean. Gough Island has no endemic fishes, and only Bovichtus diacanthus is endemic to Gough and the Tristan Group. St Paul and Amsterdam Islands have four endemic species (Duhamel, 1989).

It is impossible to be sure of the mechanisms responsible for the widespread distribution of certain species without further information on the distribution of the early life history stages of the species involved. The ideas on dispersal outlined above will hopefully serve as a starting point to which further early life-history information on the species of the STC region can be added.


In zoogeographical analyses, many historical zoogeographers focus on endemic species as the key to understanding animal distributions. Although endemism is an important aspect of zoogeography, it is not the complete picture. To understand the zoogeography of the fish fauna at Tristan and Gough, one should employ an holistic approach that considers not only endemism, but also the geological history of the islands, the geographic proximity of other neritic fish communities, the hydrography of the South Atlantic and southwestern Indian Ocean, the ecology of the marine community at these islands, the possibilities for dispersal, and the cladistic relationships of the T-G fishes with regard to their sister species at other localities.

As indicated in Table 3, the T-G fish fauna has most affinities with the fishes of St Paul and Amsterdam. These two localities (T-G and P-A) and the Austral Seamount comprise the Westwind Drift Islands Province. More than half of the neritic species at each of these three localities is shared with the other two localities. Two species (Helicolenus mouchezi and Lepidoperca coatsii) are endemic to this province. Three species (Trachurus longimanus, Acantholatris monodactylus and Nelabrichthys ornatus) are known only from the Province and the two subtropical seamounts, Vema and Walters Shoal. The large number of species in common between the Westwind Drift Islands is also a reflection of their similar temperate, rocky-bottom, kelp habitat.


Numerous people have made the current program of fish research on Tristan possible and successful. The financial support received from the South African CSIR and the Government of Tristan da Cunha is much appreciated. The Tristan islanders and especially Jimmy Glass have offered their utmost support for work in the field both in gathering information and in the collection of specimens. The captain and crew aboard the cray-fishing vessel MFV TRISTANIA II have been extremely helpful during the course of the research. Permission to work aboard this vessel and MFV HEKLA, granted by the owners, Tristan Investments (Pty) Ltd. is appreciated. We are grateful to Elaine Heemstra, Dave Voorvelt and the late Margaret Smith for their excellent illustrations. Billy Ranchod and Edward Matama of the J.L.B. Smith Institute of Ichthyology are thanked for their efficiency in handling the Tristan collection. We thank Robin Stobbs for his skillful photographic and radiographic work. Joan Wright also provided important assistance. P.A. Hulley, Leonard Compagno and Cedfic Goliath facilitated the examination of specimens in the fish collection of the South African Museum. We are grateful to Richard Rosenblatt and Guy Duhamel for their critical review of the manuscript.


AHLSTROM, E.H, J.L. BUTLER & B .Y. SUMIDA. 1976. Pelagic stromateoid fishes (Pisces: Perciformes) of the eastern Pacific: kinds, distributions and early life histories and observations on five of these from the northwest Atlantic. Bull. mar. Sci. 26(3): 285-402.

AHLSTROM, E.H., K. AMAOKA, D.A. HENSLEY, H.G. MOSER & B.Y. SUMIDA. 1984. Pleuronectiformes: Development. In: Ontogeny and systematics of fishes. Special Publication No. 1. Amer. Soc. Ichthyol. Herp.: 640-670.

ALLANSON, B.R, HART, R.C. & J.R.E. LUTJEHARMS. 1981. Observations on the nutrients, chlorophyll and primary production of the Southern Ocean south of Africa. S. Afr. J. Antarct. Res., 10/11: 3-14.

ALLEN, G.R. & P.C. HEEMSTRA. 1976. Cheilodactylus rubrolabiatus, a new species of morwong (Pisces: Cheilodactylidae) from western Australia with a key to the cheilodactylid fishes of Australia. Rec. Wes. Aust. Mus., 4(4): 311-325.

ANDREW, T.G. & T. HECHT. 1992. The feeding biology of Acantholatris monodactylus (Pisces: Cheilodactylidae) at Tristan da Cunha and Gough Island, South Atlantic Ocean. S. Afr. J. Antarct. Res., 22(1/2): 41- 49.

ANDREW, T.G, C.D. BUXTON & T. HECHT. (In press) Aspects of the reproductive biology of the coneha wrasse, Nelabrichthys ornatus, at Tristan da Cunha. Env. Biol. Fish.

ANDREW, T.G. & T. HECHT. (in press) Age and growth of the soldier, Sebastes capensis (Pisces: Scorpaenidae) at Tristan da Cunha and Gough Island, South Atlantic Ocean. J. Zool. (Lond.), 237

ANDRIASHEV, A.P. 1962. I. Bathypelagic fishes of the Antarctic. I. Family Myctophidae. Biol. rep. Soviet Antarct. Exped. Akad. Nauk zool. Inst., Moscow 1: 216-300.

ANGOT, M. 1951. Observations sur la faune marine et la peche aux iles Saint - Paul et Amsterdam. Mem. Inst. sci. Madagascar. (Ser. A), 6(1): 1-51.

AYLING, T. & G.J. COX. 1982. Collins Guide to the Sea Fishes of New Zealand. Collins, Auckland, Sidney, London. 343 pp.

BAKER, P.E. 1973. Islands of the South Atlantic. (pp 493-553) In: The Ocean Basins and Margins. Vol. 1: The South Atlantic, A.E.M. Naim & F.G. Stehli, eds. Plenum Press, New York. 583 pp.

BARNARD, K.H. 1923. Diagnoses of new species of marine fishes from South African waters. Ann. S. Aft. Mus. 13: 439-45.

BARNARD, K.H. 1925. A monograph of the marine fishes of South Africa. 1. Ann. S. Aft. Mus., 21: 14-18.

-- 948 Further notes on South African marine fishes. Ann. S. Afr. Mus, 36: 341-406.

BARSUKOV, V.V. 1979. Subspecies of the Atlantic blackbelly rosefish Helicolenus dactylopterus (De la Roche, 1809). J. Ichthyol. 19(4): 1-17.

BASS A.J. 1986. Family Lamnidae. (pp 98-101) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

BASS, A.J., P.C. HEEMSTRA & L.J.V. COMPAGNO. 1986a. Family Hexanchidae. (pp 4547) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg, 1047 pp.

-- 1986b. Family Carcharhinidae. (pp 67-87) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

-- 1986c. Family Sphyrinidae. (pp 96-97) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg, 1047 pp.

BENNETT, E.T. 1831. Characters of new genera and species of fishes of the Atlantic coast of northern Africa presented by Capt. Belcher. Proc. zool. Soc. Lond, 1831: 146-148.

BERRISFORD, C.D. 1969. Biology and zoogeography of Vema Seamount: a report on the first biological collection made on the summit. Trans. R. Soc. S. Aft, 38: 387-398.

BEUROIS, J. 1975. Etude ecologique et halieutique des fonds de peche et des especes d'interet commercial (Langoustes et poissons) des iles Saint - Paul et Amsterdam (Ocean Indien). C.N.F.R.A, 37: 1-91.

BIGELOW, H.B.W.C. SCHROEDER & S. SPRINGER. 1953. New and little known sharks from the Atlantic and from the Gulf of Mexico. Bull. Mus. comp. Zool. 109(3): 213-276.

BLANC, M. 1961. Les poissons des Tenes Australes et Antaretiques Francaises. Mere. Inst. sci. Madagascar. (Set. F) 4: 109-159.

BLANC, M. & P. PAULIAN. 1957. Poissons des iles Saint-Paul et Amsterdam. Mem., Inst. sci. Madagascar. (Ser. F) 1: 325-335.

BLOCH, M.E. & J.G. SCHNEIDER. 1801. Systema Ichthyologiae, Iconibas CX Illustratum. 584 pp, 110 pls. Berlin.

BODEN, B.P. 1952. Natural conservation of insular plankton. Nature, 169: 697-699.

BODEN, B .P. & E.M. KAMPA. 1953. Winter cascading from an oceanic island and its biological implications. Nature, 171: 426-427.

BONAPARTE, C.L. 1832-41. Iconografia della Fauna italica per le quattro classi degli Animali Vertebrati, tomo III, Pesci, Roma. 160 puntate [without pagination] in 30 fascicles, 78 pls. 1835: f 12-14 (punt. 59-79), 12 pl.

BONNATERRE, J.P. 1788. Ichthyologie. Tableau encyclopddique et methodique des trois regnes de la Nature, Paris: Ivy + 21500 pp.

BOULENGER, G.A. 1898. The flat-fishes of Cape Colony. Mar. Invest. S. Afr., 1: 1-4.

BRAUER, A. 1906. Die Tiefseefische. 1. Systematischer Teil. Wiss. Ergebn. dt. Tiefsee-Exped "Valdivia". 15(1): 1-432.

BRIGGS, J.C. 1974. Marine Zoogeography. McGraw-Hill Book Co., 475 pp.

CARMICHAEL, D. 1819. Description of four species of fish found on the coast of Tristan da Cunha. Trans. Linn. Soc. London, 12: 500-502. [Published between June 15 and July 2, 1819 (fide, Raphael, S. 1970. Biol. J. Linn. Soc. London 2: 61-76).]

CASTLE, P.H.J. 1960. Two eels of the genus Pseudoxenomystax from New Zealand waters. Trans. Roy. Soc. N.Z, 88(3): 463-472, 2 figs.

-- 1968. The congrid eels of the western Indian Ocean and the Red Sea. lchthyol. Bull. Rhodes Univ. (33): 685-726.

-- 1986. Family Congridae. (pp 161-165) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

CHEN, L-C. 1971. Systematics, variation, distribution, and biology of rockfishes of the subgenus Sebastomus (Pisces, Scorpaenidae, Sebastes). Bull Scripps Inst. Oceanogr. 18: 1-115.

COCCO, A. 1838. Su di alcuni Salmonidi del mare di Messina; lettera al ch. Principe C.L. Bonaparte. Nuovi Ann. Sci. nat., Bologna, 1(2): 161-194.

-- 1840. Su di alcuni nuovi pesci del mare di Messina. Maurolico giorn. di. sc. lett. art. del nuovo periodo, Messina, 4(5): 236-244.

COHEN, D.M. 1986. Family Gadidae. (pp 324) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

COLLETT, R. 1904. Diagnosis of four hitherto undescribed fishes from the depths south of the Faroe Islands. Videns. Selskabs Forhand. Christiania, no. 9:1-7.

COLLETTE, B.B. 1986. Family Scombridae. (pp 831-838) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

COLLETTE, B.B. & N.V. PARIN. 1991. Shallow-water fishes of Walters Shoals, Madagascar Ridge. Bull. Mar. Sci. 48(1): 1-22.

COMPAGNO, L.J.V. 1984. Sharks of the world. FAO Fish. Synop. 125(Vo1.4), Part 1 & 2:655 pp.

-- 1986. Family Torpedinidae. (pp 112-113) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

COMPAGNO, L.J.V., D.A. EBERT & M.J. SMALE. 1989. Guide to the Sharks and Rays of Southern Africa. Struik, Cape Town. 160pp.

CRAWFORD, A.B. 1972. Sea-surface temperature measurements by direct sampling. Maritime Weather Office, Youngsfield, Cape Town, South Africa. 8pp.

CUNNINGHAM, J.T. 1910. On the marine fishes and invertebrates of St. Helena. Proc. zool. Soc. Lond.: 86-131.

CUVIER, G. & A. VALENCIENNES. 1829. Histoire Naturelie des Poissons. F.G. Levrault, Paris. Vol. 3:500 pp.

-- 1830. Histoire Naturelie des Poissons. F.G. Levrault, Paris, Vol. 6:559 pp.

-- 1833. Histoire Naturelie des Poissons. F.G. Levrault, Paris, Vol. 9:512 pp.

-- 1846. Histoire Naturelie des Poissons. F.G. Levrault, Paris, Vol. 18:505 pp.

DAVID & GUERIN-ANCEY. 1990. Les rapports des campagnes a la mer, Mission de Recherche des TAAF, (86-04): 43-76.

DEACON, G.E.R. 1933. A general account of the hydrology of the South Atlantic Ocean. Discovery Rep., 7: 171-278.

-- 1966. Subtropical Convergence. In: The Encyclopaedia of Oceanography, Fairbridge, R.W. ed. Dowden, Hutchinsen and Ross, Stroudsburg: 884-885.

DEWITT, H.H, P.C. HEEMSTRA & O. GON. 1990. Nototheniidae. (pp 279-331) In: Fishes of the Southern Ocean, O. Gon & P.C. Heemstra, eds. J.L.B. Smith Institute of Ichthyology, Grahamstown. 462 pp.

DUHAMEL, G. 1984. Ichtyofaune d'un haut - fond (34 degrees 54'S, 53 degrees 14'E) de l'ocean Indien Sud-Ouest. Cybium, 8(4): 91-94.

-- 1989. Ichtyofaune des iles Saint-Paul et Amsterdam (Ocean Indien Sud). Mesogee, 9: 21-47.

EDWARDS. A.J. & C.W. GLASS. 1987. The fishes of Saint Helena Island, South Atlantic Ocean. J. Nat. Hist., 21: 617-686.

ESCHMEYER, W.N. 1969. A systematic review of the Scorpionfishes of the Atlantic Ocean (Pisces: Scorpaenidae). Occ. Pap. Calif. Acad. Sci. 79: 1-30.

-- 1986. Family Scorpaenidae. (pp. 463478) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

ESCHMEYER, W.N. & J-C. HUREAU. 1971. Sebastes mouchezi, a senior synonym of Helicoienus tristanensis, with comments on Sebastes capensis and zoogeographical considerations. Copeia 1971 (3): 576-579.

EUPHRASEN, B.A. 1788. Beskrifning pa trenne fiskar. K. svenska VetenskAkad Handl, Stockholm, 9: 51-55.

-- 1791. Scamber (Atun) och echneneis (Tropica). K. svenska VetenskAkad. Handl. 12: 315-318.

FISCHER, W. & J.C. HUREAU. (,is) 1985. FAO species identification sheets for fishery purposes. Southern Ocean (Fishing areas 48, 58, 88) (CCAMLR Convention Area). Prepared and published with the support of the Commission for the Conservation of Antarctic Marine Living Resources. Rome, FAO, Vol. 2: 233-470.

FOURNIER, R.O., VAN DET, M., WILSON, J.S. & N.B. HARGREAVES. 1979. Influence of the shelf-break off Nova Scotia on phytoplankton standing stock in winter. J. Fish. Res. Board Can., 36: 1228-1237.

FRASER-BRUNNER, A. 1949. A classification of the fishes of the family Myctophidae. Proc. zool. Soc. Lond., 118:1019-1106, 167 unnumb. fig., pl. 1

GIBBS, R.H. 1960. Alepisaurus brevirostris, a new species of lancet fish from the western Northern Atlantic. Breviora, 123: 1-14.

GIBBS, R.H. Jr. & J.C. STAIGER. 1970. Eastern tropical Atlantic flyingfishes of the genus Cypselurus (Exocoetidae). Studies Trop. Oceanogr., 4(2): 432-466, 9 fig.

GILCHRIST, J.D.F. & W.W. THOMPSON. 1914. Descriptions of fishes from the coast of Natal (Part IV). Ann. S. Afr. Mus. 13(3)(5): 65-95.

GILL, T.N. 1863. Synopsis of carangoids of the eastern coast of North America. Proc. Acad. nat. Sci. Philad. (1862) 14: 430-443.

GIORNA, M.E. 1809. Suite et conclusion du memoire sur des poissons d'especes nouvelles et de genres nouveaux. Memoire Acad. Aci. Torina, 1805-08 [1809]: 177-180.

GMELIN, J.F. 1788(1789). Pisces, in Caroli a Lime, Systema Naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Ed. XIH, 1(3): 1126-1516.

GON, O. 1990. Aleposauridae. (pp 144-145) In: Fishes of the Southern Ocean. O. Gon & P.C. Heemstza, eds., J.L.B. Smith Inst. Ichthyol., Grahamstown. 462 pp.

GON, O. & P.C. HEEMSTRA. 1987. Mendosoma linearurn (Guichenot, 1848), first record in the Atlantic Ocean, with a re-evaluation of the taxonomic status of other species of the genus MeMosoma (Pisces, Latfidae). Cybium, 11(2): 183193.

GON, O. & N.T.W. KLAGES. 1988. The marine fish fauna of the sub-Antarctic Prince Edward Islands. S. Afr. J. Antarct. Res.,18(2): 32-54.

GRANT, C.J, T.R. COWPER & D.D. REID. 1978. Age and growth of snoek, Leionura atun (Euphrasen), in South-eastern Australian waters. Aust. J. Mar. Freshwater Res., 29: 435-44.

GUICHENOT, A. 1848. Notice sur l'etablissement d'un nouveau genre de Chiltodons. Rev. Zool. 11: 12-14.

GUNN, B.H. & C.E. ABRANSON. 1971. Amsterdam Island, an isolated volcano in the southern Indian Ocean. Cantrib. Mineral Petrol., 32: 79-93.

GUNTHER, A. 1864. Catalogue of the fishes in the British Museum 5. Catalogue of the Physostomi containing the families Siluridae, Characinidae, Haplochitonidae, Sternoptychidae, Scopelidae, Stomiatidae in the collection of the British Museum. London, 455 pp.

GUREFSKIY, V.V. 1986. Study of large-scale circulation in the Atlantic sector of the Southern Ocean. Oceanology. 26(2): 131-135.

HAEDRICH, R.L. 1967. The stromateoid fishes; systematics and a classification. Bull Mus. comp. Zool. Harv., 135(2): 31-139.

-- 1986. Family Stromateidae. (pp 842-851) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047pp.

HAGAN, Y. 1952. Birds of Tristan da Cunha. Res. Norw. Sci. Ex. Tristan da Cunha 1937-1938, 20:1-248.

HARDY, G.S. 1983. A revision of the fishes of the family Pentacerotidae (Perciformes). N.Z. J. Zool. 10: 177-220.

HECTOR, J. 1874. Notice of Motella novae-zealandiae, sp. nov.. Trans. Proc. N.Z Inst., 6: 107-108.

HEEMSTRA, P.C. 1986a. Family Lophotidae. (pp402-403)In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds., Macmillan, Johannesburg. 1047 pp.

-- 1986b. Family Berycidae. (pp 409-410) In: Smiths'Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047 pp.

-- 1986c. Family Macroramphosidae. (pp 459-461) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds., Macmillan, Johannesburg. 1047 pp.

-- 1986d. Family Pentacerotidae. (pp 622-623) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047 pp.

-- 1986e. Family Emmelichthyidae. (pp 637-638) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047 pp.

-- 1986f. Family Echeneidae. (pp 662-664) In: Smiths'Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

-- 1986g. Family Molidae. (pp 907-908) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047 pp.

HEEMSTRA, P.C. ms. Epinephelas vemaensis, a new species of grouper (Pisces: Serranidae), and six new records of fishes from Vema Seamount off the west coast of South Africa. (submitted January 1994 to Ann. S. Afr. Mus.).

HEEMSTRA, P.C. & S.X. KANNEMEYER. 1984. The families Trachipteridae and Radiicephalidea (Pisces: Lampriformes ) and a new species of Zu from South Africa. Ann. S. Afr. Mus., 94(2a): 13-39.

-- 1986. Family Trachiptefidae. (pp 399- 402) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, ,is., Macmillan, Johannesburg. 1047 pp.

HEEMSTRA, P.C. & J.E. RANDALL. 1977. A revision of the Emmelichthyidae (Pisces: Perciformes). Aust. J. mar. freshwat. Res. 28(3): 361-396.

HEEMSTRA, P.C. & M.M. SMITH. 1986. Family Alepisauridae. (pp 280-281 ) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds., Macmillan, Johannesburg. 1047 pp.

HENSLEY, D.A. 1986. Family Bothidae. (pp 854-863) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

HEYDORN, A.E.F. & J.R.E. LUTJEHARMS. 1980. Tristan da Cunha -- the loneliest isle. African Wildlife. 34(6): 12-17.

HOFFMANN, E.E. 1985. The large-scale horizontal structure of the Antarctic Circumpolar Current from FGGE Drifters. J. Geophys. Res., 90: 7087-7097.

HORN, P.L. 1988. Age and growth of bluenose, Hyperoglyphe antarctica (Pisces: Stromateidei) from the lower east coast, North Island, New Zealand. N.Z.J. Mar. Freshwater Res. 22: 369-378.

HUBBS, C.L. & R.L. WISNER. 1980. Revision of the sauries (Pisces, Scomberesocidae) with descriptions of .two new genera and one new species. Fish. Bull. 77(3): 521-566.

HULLEY, P.A. 1981. Results of the research cruises of FRV "Walther Herwig" to South America. LVIII. Family Myctophidae (Osteichthyes, Myctophiformes). Arch. Fischereiwiss., 31(1): 1-300.

-- 1986. Family Myctophidae. (pp 282-321) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

HUMPHRIES, C.J. & L.R. PARENTI. 1986. Cladistic biogeography. Oxford Monog. Biogeog., 2: 1-98.

HUREAU, J.C. 1969. Remarques sur les poissons des iles Saint - Paul et Amsterdam. Bull. Mus. Nail. Hist. Nat. 40(6): 1150-1161.

HUTTON, F.W. 1872. Contributions to the ichthyology of New Zealand. Trans. Proc. N.Z Inst. 1872 5: 259-272.

KANAZAWA, R.H. 1958. A revision of the eels of the genus Conger with descriptions of four new species. Proc. U.S. nat. Mus., 108: 219-267.

KARMOVSKAYA, E.S. 1990. New species of conger eels from south eastern Pacific seamounts. Vopr. ikhtiol. 30(5): 764-772.

KAUP, J.J. 1856. Catalogue of the apodal fish in the collection of the British Museum. London, 163 pp.

KENNETT, J.P. 1982. Marine Geology. Prentice-Hall, Englewood Cliffs, New Jersey. 813 pp.

KENSLEY, B. 1980. Decapod and isopod crustaceans from the west coast of southern Africa, including seamounts Vema and Tripp. Ann. S. Afr. Mus., 83(2): 13-32.

KINGSFORD, M.J., D.R. SCHIEL & C.N. BATTERSHILL. 1989. Distribution and abundance of fish in a rocky reef environment at the subantarctic Auckland Islands, New Zealand. Polar Biol. 9: 179-186.

KLAGES, N.T.W., BROOKE, M. de L. &B.P. WATKINS. 1988. Prey of Northern Rockhopper Penguins at Gough Island, South Atlantic Ocean. Ostrich, 59: 162-165.

KNER, R. 1865-1867. Fische, (1-3), 433 pp, 16 pls. In.' Reise der osterreichischen Fregatte "Novara" ton die Erde in den Jahren 1857-59, unter den Befehlen des Commodore B yon Wullerstorf-Urbain. Zoologischer Theil.

KNOX, G.A. 1960. Littoral ecology and biology of the Southern Ocean. Proc. Roy. Soc. Lond. (Set. B). 152: 577-624.

KONG U, I. 1985. Revision de las especies Chilenas de Sebastes (Osteichthyes, Scorpaeniformes, Scorpaenidae). Estud. Oceanol, 4:21-75.

KREFFT, G. 1974. Investigations on midwater fish in the Atlantic Ocean. Bet. dr. wiss. Kommn Meeresforsck, 23: 226-254.

--1978.Distribution patterns of oceanic fishes in the Atlantic Ocean. Rev. Tray. Inst. Peches marit., 40: 439-460.

LAST, P.R.E.O.G. SCOTT & F.H. TALBOT. 1983. Fishes of Tasmania. Tasmanian Fisheries Development Authority. 562 pp.

LEIS, J.M. 1982. Nearshore distributional gradients of larva/fish (15 taxa) and planktonic crustacean (6 taxa) in Hawaii. Mar. Biol., 72: 89-97.

LINNAEUS, C. 1758. Systema Naturae, ed. X, vol. 1, 824 pp. Nantes & Pisces: 230-338. (Reprint, 1956, London.)

-- 1788. Systema Naturae per regna tria naturae, secundum classes, ordines, genera, species, cunt characteribus, differentiis, synonymis, locis. Ed. XIII, 1(3): 1126-1516.

LOWE, R.T. 1833. Description of a new genus of acanthopterygian fishes, Alepisaurusferox. Proc. zool. Soc. Lond, 1:104.

LUTJEHARMS, J.R.E. 1985. Location of frontal systems between Africa and Antarctica: some preliminary results. Deep-sea res., 32(12): 1499-1509.

-- 1987. Die Subtropiese Konvergensie en Agulhasretrofleksievaart. (SCARC). S. Aft. rydskr. Wet., 83: 454-45.

LUTJEHARMS, J.R.E. & L.H. McQUAID. 1986. Changes in the structure of thermal ocean fronts south of Africa over a three month period. S. Aft. J. Sci., 82: 470-476.

LUTJEHARMS, J.R.E. & H.R. VALENTINE. 1984. Southern Ocean thermal fronts south of Africa. Deep-Sea Res., 31: 1461-1475.

-- 1988. Eddies at the Subtropical Convergence south of Africa. J. Phys. Oceanogr., 18: 761-774.

MARRA J. HOUGHTON, R.W. & C. GARSIDE. 1990. Phytoplankton growth at the shelf break front in the Middle Atlantic Bight. J. mar. Res., 48: 851-868.

MAXWELL, J.G.H. 1980. A field guide to trawlfish from the temperate waters of Australia. CSIRO, Div. Fish. Oceanog. Circular No. 8, 201 pp.

McALLISTER, D.E. & RANDALL, J.E. 1975. A new species of centrolophid fish from Easter Island and Rafa Iti Island in the south Pacific. Pubis. biol. oceanogr. nat. Mus. Can., (8): 1-7.

McDOWALL, R.M. 1982. The centrolophid fishes of New Zealand (Pisces: Stromateoidei). J.R. Soc. N.Z. 12(2): 103-142.

MCINTYRE, A. et al. 1976. The surface of the Ice-Age Earth. Science, 191: 1131-1137.

MEAD, G.W. 1972. Bramidae. Dana-report, 81: 1-166.

MEEUWIS, J.M. 1991. Geographic characteristics of circulation patterns and features in the South Atlantic and South Indian Oceans using satellite remote sensing. PhD thesis, Rand Afrikaans University, South Africa.

MENNI R.C. & A.M. MIQUELARENA. 1979. Dimorfismo sexual y status de Centriscops obliquus maculatus Poi y Bordale. 1936 (Osteichthyes, Macrorhamphosidae). Acta zool. Lilloana, 35: 573-585.

MENNI, R.C., R.A. RINGUELET & R.H. ARAMBURU. 1984. Peces Marinos de la Argentina y Uruguay. Editorial Hemisfero Sur, Buenos Aires. 359 pp.

MILLER, D.G.M. 1982. Results of a combined hydroacoustic and midwater trawling survey of the Gough Island region. S. Aft. J. Antarct. Res., 12:17-22.

MOSER, H.G. 1967. Reproduction and development of Sebastodes paucispinis and comparison with other rockfishes off Southern California. Copeia, 1967 (4): 773-797.

MOSER, H.G.E.H. AHLSTROM & E.M. SANDKNOP. 1977. Guide to the identification of scorpionfish larvae (Family Scorpaenidae) in the Eastern North Pacific, with comparative notes on species of Sebastes and Helicolenus from other oceans. NOAA Tech. Rep. NMFS Cir. 402: 1-71.

MOSER, H.G. & G.W. BOEHLERT. 1991. Ecology of pelagic larvae and juveniles of the genus Sebastes. Env. Biol. Fish., 30: 203-224.

MUNDY, B.C. 1990. Development of larvae and juveniles of the alfonsins, Beryx splendens and B. decadactylus (Berycidae, Beryciformes). Bull. mar. Sci., 46(2): 257-273.

NAKAMURA, I. 1986a. Scorpaeuldae. (pp 180-181) In: Important Fishes Trawled off Patagonia. 1. Nakamura, T. Inada, M. Takeda & H. Hatanaka, eds. Japan Marine Fishery Resource Research Centre, Tokyo, Japan. 369 pp.

-- 1986b. Family Gempylidae. (pp 825-829) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

-- 1986c. Family Trichiuridae. (pp 829-831) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

-- 1986d. Nototheniidae. (pp 246-259) In: Important Fishes Trawled off Patagonia. NAKAMURA, I, T. INADA, M. TAKEDA & H. HATANAKA, eds. Japan Marine Fishery Resource Research Centre, Tokyo, Japan. 369 pp.

NAKAMURA, I, INADA, T., TAKEDA M. & H. HATANAKA. 1986. Important fishes trawled off Patagonia. Japan Marine Fishery Resource Research Centre, Tokyo, Japan. 369 pp.

NAKAYA, K. 1982. Chondrichthyes. In: O. Okamura, K. Amaoka & F. Mitani (eds .) Fishes of the Kyushu-Palau ridge and Tosa Bay. Fisheries Research Conservation Association, Tokyo. 435 pp.

NORMAN, J.R. 1935a. Coast Fishes. Part I. The South Atlantic. "Discovery" Rep., 12: 1-58.

NORMAN, J.R. 1935b. The Carangid fishes of the genus Decapterus (Bleeker). Ann. Mag. nat. Hist. (10)16: 252-264.

-- 1937. Coast Fishes. Part II. The Patagonian region. "Discovery" Rep. 16: 1-150.

PARIN, N.V. 1959. On the resemblance in the geographic distribution of Sardina pilchardus, Sardinops sagax, and Cypselurus pinnatibarbatus. Dokl. Akad. Nauk SSSR, 124(5): 1130-1132.

-- 1968. Ichthyofauna of the Epipelagic zone. Acad. Sci. USSR Inst. Oceanol, 206 pp. (English translation: Israel Program for Scientific Translations, Jerusalem, 1970).

-- 1984. Exocoetidae. In: W. Fischer & G. Bianchi (eds), FAO species identification sheets for fishery purposes, western Indian Ocean. FAO Fish. Dept. Rome.

-- 1991. Fish fauna of the Nazca and Sala Y Gomez submarine ridges, the easternmost outpost of the Indo-West Pacific Zoogeographic region. Bull. mar. sci., 49(3): 671-683.

PARIN, N.V. & S.G. KOBYLIANSKY. 1993. Review of the genus Maurolicus (sternoptychidae, stomiiformes), with reestablishing validity of five species considered junior synonyms of M. muelleri and descriptions on nine new species. Trans. P.P. Shirshov Inst. Oceanol. 128:64-107 (in Russian).

PAULIN, C.D. 1989. Review of the morid genera Gadella, Physiculus, and Salilota (Teleostei: Gadfformes) with descriptions of seven new species. N.Z J. Zool., 16: 93-133.

PAVLOV, Y.P. 1991. Brama australis Valenciennes - a valid species of sea bream (Bramidae) from the southeastern Pacific Ocean. J. Ichthyol. 31(1): 6-9.

PENRITH, M.J. 1967. The fishes of Tristan da Cunha, Gough Island and Vema Seamount. Ann. S. Aft. Mus. 48(22): 523-548.

PEQUENO, G. 1989. Peces de Chile: Lista sistematica y comentada. Rev. Biol. Mar., Valparaiso, 24(2): 1-132.

PERON, F. 1807. Squalus cepedianus (p. 337). In: F. Peron & L.C. Freycinet 1807-1816. Voyage de decouvertes aux Terres Australes .... 1800-1802. Paris, 2 Vols and Atlas.

PETERSON, R.G. & L. STRAMMA. 1991. Upper-level circulation in the South Atlantic Ocean. Prog. Oceanog., 26:1-73.

POLLOCK, D.E. 1990. Palaeoceanography and speciation in the spiny lobster genus Jasus. Bull Mar. Sci., 46(2): 387-405.

RAFINESQUE, C.S. 1810. Caratteri di alcuni nuovi generi e nuove specie di animali (principalmente di pesci) e piante della Sicilia, con varie osservazioni sopra i medissimi. Palermo, 105 pp. (Reprint, 1967, Asher-Amsterdam.)

REGAN, C.T. 1913. The fishes described by Carmichael from Tristan da Cunha. Ann. Mag. Nat. Hist. (8), 11.

REGAN, C.T. 1914. A synopsis of the fishes of the family Macrorhamphosidae. Ann. Mag. nat. Hist., 13(8): 17-21.

RICHARDSON, J. 1842. Report on the present state of the ichthyology of New Zealand. Rep. Br. Ass. Aclvmt Sci., 12th meet.: 12-30.

RICHARDSON, J. 184448. Ichthyology. In: J. Richardson & J.E. Gray, The Zoology of the voyage of HMS 'Erebus' and 'Terror' under the command of Capt. Sir J.C. Ross during .... 1839-43. London, 2(2): 139 pp. (pp 17-52 were published in 1845).

-- 1850. Notices of Australian fish. Proc. zool Soc. Lond. 1850: 58-77.

ROBERTS, C.D. 1989. Reproductive mode in the percomorph fish genus Polyprion Oken. J. Fish Biol.(34): 1-9.

-- 1991. Fishes of the Chatham Islands, New Zealand: a trawl survey and summary of the ichthyofauna. N.Z J. Mar. Freshwat. Res., 25: 1-19.

ROSCOE, M.J. 1979. Biology and exploitation of the rock lobster Jasus tristani at the Tristan da Cunha islands, South Atlantic, 1949 - 1976. S. Aft. Dep. Ind. Sea Fish. Branch, Inv. Rep. (118), 47 pp.

ROWAN, M.K. & A.N. ROWAN. 1955. The fishes of Tristan da Cunha. S. Aft. J. Sci. 52: 129. .

RUPPELL, E. 1837. Fische des rothen Meeres. (pp 53-80 ) In: Neue Wirbelthiere zu der Fauna von Abyssinien gehorig., Frankfurt am Main.

RUSSELL, B.C. 1983. Nelabrichthys, a new genus of labrid fish (Perciformes: Labridae) from the southern Indian and Atlantic Oceans. Spec. Publ. Smith Inst. Ichthyol. (27): 1-7.

SALE, P.F. 1970. Distribution of larval Acanthundae off Hawaii. Copeia, 1970: 765-766.

SANCHEZ, R.P. & E.M. ACHA. 1988. Development and occurrence of embryos, larvae and juveniles of Sebastes oculatus with reference to two Southwest Atlantic scorpenids: Helicolenus dactylopterus lahillei and Pontinus rathbuni. Meeresforsch., 32: 107-133.

SAUVAGE, H.E. 1875. Poissons. Sur la faune ichthyologique de l'ile Saint-Paul. C.R. Acad. Sci., Paris, 81: 987-989.

-- 1879. Le faune ichthyologique des iles Saint-Paul. Arch. Zool. exp. gen. 8: 1-46.

SCHAEFER, S., R.K. JOHNSON & J. BADCOCK. 1986. Family Photichthyidae. (pp 243-247) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds., Macmillan, Johannesburg. 1047 pp.

SERVENTY, D.L. 1948. Allothunnus fallai, a new genus and species of tuna from New Zealand. Rec. Canterbury Mus. 5(3): 131-135.

SHCHERBACHEV, Yu.N. 1987. Preliminary list of thalassobathyal fishes of the tropical and subtropical waters of the Indian Ocean. J. Ichthyol, 1987(2): 37-46.

SINCLAIR, M. 1988. Marine populations: An essay on population regulation and speciation. Washington Sea Grant Program, University of Washington Press, Seattle and London. 251 pp.

SINCLAIR, M. & M.J. TREMBLAY. 1984. Timing of spawning of Atlantic herring (Clupea harengus harengus) populations and the match-mismatch theory. Can. J. Fish. Aquat. Sci., 41: 1055-1065.

SIVERTSEN, E. 1945. Fishes of Tristan da Cunha. Res. Norw. Sci. Ex. Tristan da Cunha 1937 - 1938, 2(12): 1-44.

SMITH, A. 1838-1849. Illustrations from the zoology of South Africa, consisting chiefly of figures and descriptions of the objects of natural history collected during an expedition into the interior of South Africa in 1834-36. London, 5 Vols., 4, Pisces, 1938-47.

SMITH, D.G. 1989. Congridae. In: Fishes of the western North Atlantic. Metn. Sears Found Mar. Res., New Haven, no. 1, part 9, vol. 1: 460-567.

SMITH, J.L.B. 1949. The stromateid fishes of South Africa. Ann. Mag. nat. Hist., 12(2): 839-851.

-- 1964. Fishes of the family Pentacerotidae. Ichthyol. Bull. Rhodes Univ., (29): 547-564.

SMITH, M.M. 1980. A review of the South African Cheilodactylidae fishes (Pisces: Percfformes ), with descriptions of two new species. Ichthyol. Bull Smith Inst. Ichthyol. (42): 1-14.

SMITH, M.M. & P.C. HEEMSTRA. 1986. Smiths' Sea Fishes. Macmillan, Johannesburg. 1047 pp.

SMITH-VANIZ, W.F. 1986. Family Carangidae. (pp. 638-661) In: Smiths' Sea Fishes, M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

STAUCH, A. 1967. Description de'une nouvelle espece de Bothinae: Arnoglossus entomorhynchus n sp (Pisces: Teleostei, Heterosomata). Bull. Mus. natl. Hist. nat., Paris (2) 39: 660-664, 1 fig.

STEHMANN, M. & D.L. BURKEL. 1984. Torpedinidae. (pp. 159-162) In: Fishes of the North-eastern Atlantic and the Mediterranean Vol. I, P.J.P. Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen & E. Tortonese, eds. UNESCO, Paris. 510 pp.

STEHMANN, M. & W. LENZ. 1973. Ergebnisse der Forschungsreisen des FFS "Walther Herwig" nach Sudamerika. XXVI. Systematik und Verbreitung der Artengruppe - Seriolella punetata (Schneider, 1801), S. porosa Guichenot, 1848, S. dobula (Gunther, 1869) - sowie taxonomische Bemerkungen zu Hyperoglyphe Gunther, 1859 und Schedophilus Cocco, 1839 (Osteichthyes, Stromateoidei, Centrolophidae). Arch. Fish Wiss.: 23(3): 179-201.

STRAMMA, L. & PETERSON, R.G. 1990. The South Atlantic Current. Phys. oceanogr., 20(6): 846-859.

SVETOVIDOV, A.N. 1986. Review of the three-bearded rocklings of the genus Gaidropsarus Rafinesque, 1810 (Gadidae) with description of a new species. J. Ichthyol., 26(1): 114-135.

TANING, AV. 1932. Notes on scopelids from the Dana Collections. 1. Vidensk Meddr. dansk naturh. foren, 94:125-146.

THRESHER, R.E. 1984. Reproduction in reef-fishes. T.F.H. Pub- lications. 399 pp.

VALENCIENNES, A. 1840. Les Poissons., in [G. Cuvier's] Le Regne Animal... ed. 3, Vol. 4: 1-329, 120pls. Paris.

WAITE, E.R. 1911. Antarctic fishes. In: Br. Antarct. Exped. 1907-1909, Biol., 2(2): 11-16.

WALBAUM, J.J. 1792. Petri Artedi sueci genera Piscium in quibus systema totum ichthyologiae proponitur cum dassibus, ordinibus, generum characteribus, specierum differentiis, observationibus plurimis. Ichthyologiae Pars III. Grypeswaldiae, (ed. Ii), 723 pp., 3 pls.

WEITZMAN, S.H. 1986. Family Stemoptychidae. (pp 253-261) In: Smiths' Sea Fishes. M.M. Smith & P.C. Heemstra, eds. Macmillan, Johannesburg. 1047 pp.

WHITLEY, G.P. 1932. Studies in ichthyology, No. 6. Rec. Aust. Mus., 18(6): 321-348.

-- 1948. Studies in ichthyology, No. 13. Rec. Aust. Mus., 22(1): 70-91.

WHITLEY, G.P. & W.J. PHILLIPS. 1939. Descriptive notes on some New Zealand fishes. Trans. R. Soc. N.Z, 69(2): 183-281.

WHITWORTH, T & W.D. NOWLIN. 1987. Water masses and currents of the Southern Ocean at the Greenwich Meridian. J. Geophys. Res., 92: 6462-6476.

Copyright 1995 J.L.B. Smith Institute of Ichthyology

The following images related to this document are available:

Photo images

[fb95002a.jpg] [fb95002r.jpg] [fb95002w.jpg] [fb95002k.jpg] [fb95002s.jpg] [fb95002z.jpg]

Line drawing images

[fb95002x.gif] [fb95002j.gif] [fb95002i.gif] [fb95002q.gif] [fb95002f.gif] [fb95002m.gif] [fb95002c.gif] [fb95002t.gif] [fb95002b.gif] [fb95002h.gif] [fb95002v.gif] [fb95002l.gif] [fb95002p.gif] [fb95002y.gif] [fb95002g.gif] [fb95002u.gif] [fb95002n.gif] [fb95002d.gif] [fb95002e.gif] [fb95002o.gif]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2018, Site last up-dated on 11-Dec-2018.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Internet Data Center of Rede Nacional de Ensino e Pesquisa, RNP, Brazil