Eumegistus brevorti

Carvalho-Filho, Alfredo, Marcovaldi, Guy, Sampaio, Cláudio L. S., Paiva, Isabel G. & Duarte, Luiz A. G., 2009, First report of rare pomfrets (Teleostei: Bramidae) from Brazilian waters, with a key to Western Atlantic species, Zootaxa 2290, pp. 1-26 : 5-14

Eumegistus brevorti ( FIG 2 )

Tropical Pomfret, Olho-de-Ovo.

The genus Eumegistus is composed of two species, E. illustris Jordan & Jordan, 1922, from the Pacific, and E. brevorti from the Atlantic, known from West Africa (larvae), Cuba, Bahamas, Virgin Islands, Puerto Rico, Colombia, and offshore in the Gulf of Mexico and the Caribbean. We found few reports on E. brevorti , all preserved in institutions around the world. Additionally, we found other preserved specimens in museum internet sites, all listed in Table 3 together with those of the literature and of this study. This meso/ bathypelagic species occurs between 384 and 1317 m and may reach 481 mm SL; adults are known usually from of offshore islands ( Thompson & Russel, 1996; Hernández-Hamón et al., 1999; Bunkley-Williams & Williams, 2004; Moteki & Mundy, 2005; Richards, 2006).

We examined 30 fishes ( Table 1.A–B), ranging from 394 to 550 mm SL and weighing from 1.65 to 4.80 kg. Two females are the largest and heaviest specimens ever reported for the species. From this sample, 20 were males (425–523 mm SL), six were females (447–550 mm SL, Table 4, in part) and four were young adults of indeterminate sex, still having keels in the caudal peduncle’s horizontal rows of scales (394, 420, 443 and 443 mm SL, weighing 1.29, 1.65, 1.72, and 1.73 kg). Females were heavier (4.22–4.80 kg) than males (2.30 and 3.91 kg). The TAMAR vessel’s captain, also a very experienced fisherman, reported that only about 10% of the hooked Eumegistus were females bearing maturing eggs, which was confirmed by us in the field.

Diagnostic characters: Based on examined specimens and literature data ( Table 1.A–B).

Body compressed, rather high, the greatest depth 41–52% of standard length; dorsal profile of head arched and rounded; lower edges of mandibles not touching along entire length, a good part of the isthmus visible between them; pre-caudal groove absent; dorsal-fin rays 29–33; anal-fin rays 20–24; pectoral fin extends beyond lobe of anal fin, with 17–21 rays; horizontal series of scales 47–57; gill rakers slat-like, 3–4 + 6–7, not counting rudiments; scales on sides of body cycloid with a median soft keel, those in the posterior part and on caudal peduncle with a strong antrorse spine in young adults (see discussion below); naked area above and behind the eyes; scales on caudal fin base abruptly smaller than those on caudal peduncle. Color varies from almost black or dark brown with bronze to gold or silver sheen, to silvery white; dorsum always darker, belly somewhat lighter; most of lateral body scales outlined with thin dark lines (pale individuals) or silvery-white lines (dark individuals); dorsal fin usually dark with its anterior portion silvery-white to silverydusky; anal fin silvery-white to silvery-dusky, outlined with a dark band except in the anterior lobe; anterior lobe of dorsal and anal fins tips sometimes whitish in young adults; pectoral and pelvic fins dark interiorly and with a pale posterior large border; caudal fin dark with conspicuous posterior white median margin, split in two by black central rays in young adults ( FIG 4-A ). The differences between adults and young adults in the caudal fins’ white and black marks, as well as the whitish lobes of dorsal and anal fins (which are more conspicuous in young adults), are also displayed by the Pacific Eumegistus illustris Jordan & Jordan, 1922 ( FIG 4-B and 4-C ), leading us to assume that this pattern is a character, among others, of the genus.

Mead (1972) reported the presence of an antrorse spine on posterior body and caudal peduncle scales on the specimen he examined with 259 mm SL, the lectotype of the junior synonym E. saussuri (Lunel, 1865) . Hernández-Hamon et al. (1999), reported 5 rows of keeled scales on caudal peduncle for an individual 246 mm SL. Moteki & Mundy (2005) reported that a juvenile (23.0 mm) had head and body covered by scales with a few spines. Thompson & Russell (1996) did not report any keel or spines on body scales on three adult specimens (425, 475 and 481 mm SL). The four young adults examined by us presented the following distinguishing features in the caudal peduncle:

Counts M e a d Thompson & Moteki & M o te k i Hernández- Hamón Present Study

(1972) Russell (1996) Mundy (2005) (2005) et al. (1999)

(N) specimens 1 3 2 2 1 30

Dorsal fin rays 33 31–33 31–33 31–33 ---- 29–32

Anal fin rays 22 22–23 23–24 23–24 22–23(?) 20–23

Pectoral fin rays 20 20 20 20 20 17–21

Scales, horizontal ca. 50 55–57 ----- 50** 54–57(?) 47–56

Upper gill rakers 3 3 ----- ----- ---- 3–4

Lower gill rakers 6 7 6–7

(?): The authors examined one specimen, so the gap in counts is not plausible.

* Type of E. saussuri (Lunel) 259 mm, considered by us a young adult.

** Three adult females.

(1): According to Moteki & Mundy (2005), the smaller specimen is so damaged that is impossible to determine the species, identified as E. brevorti by Mead (1972).

(2): Misidentified by Mead (1972) as Taractichthys longipinnis ; this misidentification persists in the collection records of the MCZ to date.

(3): Lectotype of B. saussurii.

(4): Dried end mounted syntype of B. saussurii.

(5): Stanford University Collection in California Academy of Sciences; assumed as E. brevorti .

TAMAR 0 0 35, 394 mm SL; MZUSP 95974, 420 mm SL: 6–12 scales with antrorse spine on each of the four central rows of scales, more sharp and strong at the middle of the row;

UNESP-SV 1A, 443 mm SL: 4–7 scales with antrorse spine on each of the four central rows of scales, more sharp and strong at the middle of the row;

ZUEC 6313, 443 mm SL: 3–5 scales with a moderate keel on each of the four central rows of scales, higher at the middle of the row.

Thus, it is very probable that the development and the reduction of the spines on scales follows the age of the fish, developing as a few on each scale in juveniles, transforming into one strong, antrorse, sharp spine in young adults, becoming a low keel on older fish, and finally disappearing almost completely, being just a smooth low keel, on large adults. To confirm this hypothesis we would need specimens in the 50–180 mm SL range, unknown to date. The Pacific species seems to present the same pattern ( FIG 4-B ).

Concerning the dorsal, anal and pectoral fins, as well as the horizontal series of scales and the proportional measurements, our data widen several parameters of previous studies ( Table 1-A and 1-B).

Bunkley-Williams & Williams (2004) suggested that more than one species of Eumegistus may exist in the Caribbean and Atlantic due to the differences among several morphometric measurements reported by previous studies ( Mead, 1972; Thompson & Russell, 1996; Hernández-Hamon et al., 1999). We disagree with this suggestion, since: 1) Bramidae , as a rule, presents different proportions with growth, and 2) there were not enough specimens at that time (2004), to support such assertion. Our data show a wider range of morphometric proportions than previously reported and we assume that this wide-ranging species might differ somewhat when separate populations are analyzed. The meristic and morphometric data of the examined specimens, compared to previous descriptions are in Table 1-A and 1-B.

Distribution: Tropical Atlantic Ocean, from West Africa to Bahamas and the Caribbean area, Gulf of Mexico, Colombia and Venezuela, and Fernando de Noronha Archipelago and Bahia in Brazil. The present study extends the occurrence of the tropical pomfret southwards about 3,500 kilometers, since it was known only from the tropical North Atlantic.

Proposed Brazilian name: “Olho-de-Ovo” (= egg-eye). This name was adopted by the Fishing Team due to the fact that many of the collected specimens arrived at surface with bulged eyes due to sudden decompression.

Habitat, Depth, Temperature and Presumed Schooling Pattern: All specimens were caught from October 2006 to June 2008, at 300–900 m depth, and between 30 and ~100 meters from the ocean floor. This data, alone, is enough to partially confirm the supposition made by Mead (1972) and Thompson & Russell (1996), about its benthopelagic and probable demersal habits. Temperature probably has an important role in the habitat preferences of the tropical pomfret, as all individuals were collected between 8 and 10o C measured on occasions. In several fishing trips 3–10 individuals were caught in the same site and depth, very probably at the same time, which indicates grouping or schooling behavior.

Diet: We examined stomach contents or vomit of 30 specimens. Of these, 40% (N=12) were empty and had not vomited; 46.7% (14) preyed only on fishes and cephalopods, and 13.3% (4) presented a diet of fishes, crustaceans ( Oplophoridae , Scyllaridae and Squillidae ), and cephalopods (5–7). The data confirm the benthopelagic habitat proposed by Mead (1972) and Thompson & Russell (1996), since some of the prey are considered bottom-dwelling species, e.g., the oplopherid shrimp and the scyllarid Parribacus antarcticus (Lund, 1793) (M. Tavares, pers., comm.), the squillid mantis shrimp, and the ogcocephalid batfish Halieutichthys aculeatus (Mitchill, 1818) ; the small batfishes are the main support of this statement, since all five specimens (20.7–25.3 mm SL) were diagnosed as juveniles, with a naked ventral surface, elongated rays in the caudal fin, and similar meristics as the adults ( Richards & Bradbury, 2006).

3. Paralepididae , Lestidium atlanticum . 4. Chaliodontidae, Chauliodus sloani .

5. Ogcocephalidae , Halieutichthys atlanticus . 6. Caristiidae , Caristius sp.

7. Anguilliformes , unknown leptocephalus. 8. Crustacea, Decapoda , Oplophoridae .

The prey most often observed among fishes were Chauliodus sloani Bloch & Schneider, 1801 (adults, including a ripe female), Diaphus garmani Gilbert, 1906 , Scopelarchus sp., Lestidium atlanticum Borodin, 1928 , young Halieutichthys aculeatus , young Gempylidae (Neopinnula americana (Grey, 1953) and Gempylus serpens Cuvier, 1829 ) and Trichiuridae ( Lepidopus sp.). The largest recorded prey was the steindachnerid fish Steindachneria argentea Goode & Bean, 1896 , weighing 18.31 g and measuring 135.8 mm SL. The largest stomach content included fishes, crustaceans, cephalopods and worms, and weighed 38.28 g. In about 25% of the examined stomachs we found a few parasitic worms (Nematoda and Trematoda).

As side benefits, several new or rare records of fishes and cephalopods for Brazilian waters, and even for the Southwestern Atlantic Ocean, were found in the vomit and stomach contents, such as Caristius sp., Steindachneria argentea , Asarcenchelys longimanus McCosker, 1985 , and the squid Spirula spirula (Linnaeus, 1758) , which will be dealt with in another paper.

2. Cephalopoda, Enoploteuthidae . 3. Trichiuridae , Lepidopus altifrons 4. Scopelarchidae , Scopelarchoides sp.

5. Steindachneridae, Steidachneria argentea . 6. Myctophidae , Diaphus garmani . 7. Gempylidae, Neopinnula americana. 8. Gempylidae , Gempylus serpens .

Spawning season: The ovaries of four large females were analyzed, all of them ripe. The eggs were transparent (an indication of maturity in Bramidae according to Omori et al., 1997, and Yatsu & Nakamura, 1989), with a diameter of 0.63–0.66 mm, and numbering up to about 350,000. All females had one of the ovaries smaller than the other ( FIG 8 ), and in the case of the specimen caught in September, the difference of size was most evident: the right ovary weighed only 10.06 g, and the left one 75.81 g. (See Table 2 for an overview of all data, except for the specimen caught in June 2008 that had its viscera accidentally discarded before a proper analysis).

All females were collected in September and November 2007, April and June 2008, covering a large period of the year: early to late winter, mid-spring and early autumn. The fact that the September female had one of the ovaries much smaller than the other may imply in a peak of spawning pattern; however, this anomaly may be due to other causes. Thus, we provide evidence to support the hypothesis that E. brevorti spawns from early winter to late summer in the Northern Hemisphere, as suggested by Thompson & Russell (1996) from observations in the Caribbean of three large females, all mature with late stage vitellogenic and early hydrated oocytes, indicating a spawning season over much of the year (January–August), or from early winter to late summer in the Northern Hemisphere. Spawning season in Brazilian water may also range over at least 8 to10 months of the year (this paper).

Fishery Resource & Conservation: The high number of specimens of the tropical pomfret collected in Brazilian waters is not atypical as one might believe at a first glance. Bunkley-Williams & Williams (2004) reported that it was also found off Puerto Rico, but did not specify how many. Fishing with the TAMAR fishing sets (appropriate to research sampling of deep fish fauna) indicates that this species is rather abundant in the region. The only other species so often hooked is the oil-fish, Ruvettus pretiosus Cocco, 1833 , a common species around the world ( Parin, 2003). The Bramidae are targeted by fisheries in several parts of the world and regarded as excellent foodfish. Thus, the presence of a probably large population of the tropical pomfret in Bahia may lead to commercial fishery, but its stock and biology need to be better known (otherwise we will deal with another species at the risk of overfishing). An excellent analysis of deep-water fisheries is in Morato et al. (2006); in the abstract the authors state: “Global landings of demersal marine fishes are demonstrated to have shifted to deeper water species over the last 50 years. Our analysis suggest deep-water fish stocks may be at serious risk of depletion, as their life histories render them highly vulnerable to overfishing with little resilience to over-exploitation. Deep-sea fisheries are exploiting the last refuges for commercial fish species and should not be seen as a replacement for declining resources in shallower waters. Instead, deep-water habitats are new candidates for conservation.” With this alert in mind, the major effort now shall be done to collect the data about the tropical pomfret in the Brazilian waters before “go fishing”.

When the above mentioned finding was coming to conclusion, another bramid species, new to Brazilian waters, was collected at the same site off Bahia: a male of the keeltail pomfret, Taractes rubescens , hooked in about 600 meters deep and now preserved in the PROJETO TAMAR collection. It differs from its only congener, Taractes asper , mainly by the presence of a strong keel in the caudal peduncle (absent in T. asper ), as well as by body and fins proportions ( Mead, 1972; Haedrich, 1986; Smith, 1986; Thompson & Russell, 1996).

Figures

FIGURE 2. Eumegistus brevorti, freshly preserved male, 523 mm SL and 2.8 k, caught at 600 m depth 10 miles off Praia do Forte, Bahia, Northeast Brazil.

FIGURE 8. Ovaries of Eumegistus brevorti caught off Bahia, Northeast Brazil: 1: 11.16.2007: A 57.37 g; B: 93.25 g and 2: 09.07.2007: A: 10.06 g; B: 75.82 g.

FIGURE 4 - A. The smallest Eumegistus brevorti specimen, TAMAR 0 0 35 (394 mm SL), showing white-tipped dorsal and anal fins lobes and caudal fin with black central rays, caught at the Praia do Forte, off Bahia, Northeast Brazil.

Tables

TABLE 2. Eumegistus brevorti. Egg and ovary characteristics of three females. Measurements in mm and weights in g.

	September 2007	November 2007	April 2008
SL Egg diameter	447 0.63	514 0.66	550 0.63
Left ovary length	78.4	85.1	73.7
Left ovary max. width Left ovary max. height	43.1 37.5	36.2 27.8	39.1 28.9
Left ovary weight	75.8	57.4	50.7
Number of eggs Right ovary length	178,500 64.8	208,200 88.2	84,400 80.6
Right ovary max. width	17.2	51.8	39.6
Right ovary max. height Right ovary weight	13.6 10.0	32.9 93.3	33.1 51.8
Number of eggs	55,800	138,900	111,900
Total weight of ovaries Total number of eggs	85.8 234,300	150.7 347,100	102.5 196,300