THE genus Arapaima Müller, with type species A. gigas (Schinz, in Cuvier, 1822), has been considered to be a monotypic genus for the past 145 years (Ferraris, 2003). That perspective follows from a half-page account by Günther (1868), who summarily listed three species described by Valenciennes (in Cuvier and Valenciennes, 1847) in synonymy of A. gigas without rationale or analysis. For two centuries, Arapaima have been among the most important commercial fishes in freshwaters of South America (Castello and Stewart, 2010), but until this year, no taxonomist has questioned Günther's opinion about these iconic fishes. A review of the scant literature on species-level taxonomy of Arapaima (e.g., Cuvier, 1822; Spix and Agassiz, 1829; Schomburgk, 1841; Cuvier and Valenciennes, 1847; Günther, 1868) and the existing holotypes for two species of Arapaima in Museum national d'Histoire naturelle, Paris, revealed that the four nominal species of Arapaima recognized by Valenciennes appear to be valid (e.g., Stewart, 2013). In the course of my studies on alpha taxonomy of Arapaima, I also encountered what appears to be a new, fifth species of Arapaima represented by a single specimen from the central Amazon in Brazil.
Recognition of this distinctive fish will allow field biologists to begin assessing its status and, ultimately, that could lead to focused management and conservation efforts. The Purus River and surrounding areas presently are being fished heavily to supply market demands in Manaus and other cities of the central Amazon (L. Castello, Woods Hole Research Center, pers. comm.). It is thus urgent to look more closely at Arapaima being harvested in that area before populations are seriously depleted. Aquaculture of Arapaima also is being developed farther upstream in the Purus basin, and in time, that may lead to undesirable translocations of non-native species if possible taxonomic diversity is not considered (e.g., Castello and Stewart, 2010). In museums of the world, the total number of preserved arapaima from the general area where this new species was found can be counted on one hand, and there are none from the middle and upper Purus River. Our understanding of the taxonomy and distributional ecology of Arapaima is substantially hindered by that paucity of study materials.
Recent observations on morphological variation in a reference population of Arapaima from Mamirauá Reserve, Brazil, provide a comparative framework for evaluating those morphometric and meristic features that appear to be diagnostic for the new species. Examination of the holotypes of A. gigas and A. mapae, freshly collected topotypical material for A. arapaima from Guyana, and the published description of A. agassizii (Spix and Agassiz, 1829; see re-description by Stewart, 2013) also revealed various characters that can distinguish the new taxon relative to those nominal species. Detailed re-descriptions of other previously described taxa will be published elsewhere, but characteristics important for a clear, differential diagnosis of this new species are presented here.
MATERIALS AND METHODS
Comparative materials.—Analyses of the taxonomy of Arapaima involve meristic, morphometric, osteological, and live-color observations on a wide size range of individuals from Mamirauá Reserve in central Brazil (e.g., C. Arantes, unpubl. data, morphometrics for 28 individuals). Because of logistic constraints, most large individuals from Mamirauá were measured in the field and not retained; preserved voucher specimens are listed below. Other non-type museum specimens studied were from about 33 localities in the Amazon basin (n = 75, from between Belém, Brazil and Pucallpa, Perú) and from the Essequibo River basin in Guyana. Further analyses of those data will be published elsewhere, but they provide a broad context for understanding morphological variation among populations of Arapaima; data for particular characters are integrated into the diagnosis of A. leptosoma by the phrase "distinguished from all other." Institutional abbreviations are as listed at http://www.asih.org/node/204, and the Centre for the Study of Biological Diversity at the University of Guyana is ‘CSBD/UG.’
Type materials.—Type specimens examined for A. gigas and A. mapae are as follows; status of the types for A. arapaima and A. agassizii also are considered; see Stewart (2013) for additional historical notes.
Arapaima gigas (Schinz, in Cuvier 1822): holotype (also = holotype of Vastres cuvieri), MNHN a-8837, 203 cm standard length (SL), dried and stuffed mount, Brazil, Pará State, lower Amazon basin at Villa de Santarém, A. R. Ferreira, between 1783 and 1787 (Ferreira, 1903). This species is known only from the holotype.
Arapaima mapae (Valenciennes, in Cuvier and Valenciennes, 1847): holotype, MNHN a-8836, 203 cm SL, dried and stuffed mount, "lac Mapa, sur les confins des nouvelles frontières de la Guyane française", purchased by C. Pradier, a French naval officer, 1837. This locality is now called Lago do Amapá, in Amapá State, Brazil. This locality is near the Atlantic coast and outside the Amazon basin. This species also is known only from the holotype.
Arapaima arapaima (Valenciennes, in Cuvier and Valenciennes, 1847): holotype, BMNH 2009年1月19日.1 (not seen, presently missing from BMNH, but efforts to find this huge specimen are continuing, O. Crimmen, BMNH, pers. comm.), 246 cm total length (when freshly collected), dried, stuffed specimen, British Guiana, Sir Robert H. Schomburgk, from his expeditions of 1835–1839; this specimen was mentioned in Schomburgk (1841) and by Valenciennes (in Cuvier and Valenciennes, 1847, within the original description of A. arapaima). If the holotype of A. arapaima cannot be relocated, a neotype from the Rupununi River basin in Guyana will be designated; that specimen is listed below under ‘non-type voucher specimens’ and, for various comparisons, it is referenced here as a ‘topotype.’
Arapaima agassizii (Valenciennes, in Cuvier and Valenciennes, 1847): based on description and illustrations of Spix and Agassiz (1829:31–40, plate 16, anatomy plates B, skeleton, and C, scales only), Brazilian Amazon, Spix, 1817–1820 (Spix and Agassiz, 1829; Tiefenbacher, 1983). This holotype was deposited in Zoologische Staatssammlung München, Munich, Germany, but was destroyed in World War II. See Pethiyagoda and Kottelat (1998) for an English translation of Spix and Agassiz (1829), and Stewart (2013) for detailed re-description of this species, which also is known only from the holotype.
Non-type voucher specimens.—Arapaima arapaima (Valenciennes, in Cuvier and Valenciennes, 1847): topotype, CSBD/UG 1667, 104 cm SL, alcohol specimen, Guyana, Rupununi River basin, Grass Pond, ∼4.5 km SW from Rewa village, going upstream along Rewa River, 3.86769°N, 58.76678°W, L. C. Watson and D. J. Stewart, 12 August 2006.
Arapaima sp. incertae sedis: all of following from Brazil, Amazonas State, near Comunidade Jarauá, Mamirauá Sustainable Development Reserve, approximately 2.833°S, 65.000°W, C. Arantes (see Castello, 2008:fig. 1). INPA 26580, approximately 160 cm SL (tail was missing), dry skull, Paraná do Jarauá, close to Comunidade Jarauá, 13 November 2006; INPA 26581, 126 cm SL, complete dry skeleton, Lago Samaúma, 14 November 2006; INPA 26582, 5, 61.9–73.5 cm SL, 4 alcohol specimens and 1 complete dry skeleton, Ressaca do Curuçá, 17 November 2006; INPA 26583, 3, 86.6–112 cm SL, 2 alcohol specimens and 1 complete dry skeleton, Lago Samaumeirinha do Jaraqui, 25 November 2006; INPA 26584, 96.8 cm SL, alcohol specimen, Lago Cedrinho, 24 November 2006.
Morphometric and meristic characters.—Measurements were made with digital calipers to nearest mm, or for long measurements on specimens over about 1.0 m SL, to nearest 0.5–1.0 cm using a tape measure. Most measurements follow standard procedures for fish systematics and are self-explanatory. Standard length was measured from tip of upper jaw to mid-point of caudal-fin base. Caudal-peduncle length is from posterior end of anal-fin base to mid-point of caudal-fin base. Head length is from tip of upper jaw to tip of opercular flap; postorbital distance is from posterior bony margin of orbit to that same point. Length of fourth infraorbital is from posterior bony margin of orbit to posterior tip of that bone. Length of upper jaw is from mid-point of snout to posterior tip of maxilla. Tooth counts for the dentary are for either the left or right ramus; so total number of teeth on the lower jaw is double the values reported here. All tooth counts include gaps where openings in the skin indicate a missing tooth.
To evaluate statistical significance of morphometric measures in the context of allometric trends in various characters, I used comparative measurement data for a wide size range of Arapaima from a reference population from Mamirauá Reserve, Brazil. Taxonomic status of the Mamirauá population is a complex problem involving the superficially similar A. arapaima and beyond the scope of this paper (for example, they share a relatively deep body that contrasts with other nominal species of Arapaima and A. leptosoma). Individual characters were evaluated using bivariate plots versus SL, and 95% confidence ellipses were computed for values of the reference population (using software package PAST Ver. 2.13; Hammer et al., 2001). If the corresponding value for the holotype of A. leptosoma falls outside that confidence ellipse, then it can be inferred that it is different from the reference population. Likewise, if the value of A. leptosoma is outside and values for type specimens of other nominal species fall within the 95% confidence ellipse (or better still, outside in the opposite direction from A. leptosoma), it can be inferred that they also are different from A. leptosoma. It is not possible, however, to directly estimate the statistical probability that two individual type specimens differ; we can only make indirect inferences about such differences.
Arapaima leptosoma, new species
Fig. 1.
Arapaima leptosoma, holotype: (A) lateral view, arrow indicates approximate position of dorsal-fin origin, and (B) ventral view to show relative size and position of pectoral and pelvic fins, scale bar = 10 cm; (C) dorsal view, and (D) ventral view of head, scale bar = 10 cm; and (E) ventrolateral view of head to illustrate angular junction of preopercle (pop) and third infraorbital (io3); arrows indicate where those two bones meet, scale bar = 5 cm.
Fig. 2.
Comparisons of shape for dorsalmost lateralis cavity on preopercle for: (A) A. gigas, holotype; (B) A. arapaima, topotype, Guyana (CSBD/UG 1667); (C) A. mapae, holotype; (D) A. agassizii, holotype; (E) juvenile Arapaima sp. from Mamirauá Reserve, Brazil (INPA 26582, 61.9 cm SL); and (F) A. leptosoma, holotype; scale bars for A, C, E, F = 5 cm; for B, D = 2 cm.
Holotype.—INPA 16847, 77.6 cm SL, Brazil, Amazonas State, Solimões River about 200 km W of Manaus and 21 km SW of confluence with Purus River, near Anori, approximately 3.75583°S, 61.67388°W, Manoel Brandas, 25 January 2001; this is the only known specimen.
Diagnosis.—Distinguished from all other Arapaima by following three characters: 1) dorsalmost lateralis sensory cavity on preopercle extremely slender (Fig. 2F, vs. broadly oval or nearly rectangular in other nominal taxa, Fig. 2A–D); 2) ventrolateral margin of head where third infraorbital meets anterior limb of preopercle strongly angled such that ventral surface of head is almost flat (Figs. 1E, 2F, vs. ventrolateral margin of head relatively rounded, Fig. 2E); and 3) anterior third of dorsal-fin base covered with an enlarged, thickened sheath that hides anterior dorsal-fin rays when adpressed (Fig. 1A, vs. no thickened sheath on anterior base of dorsal fin and distal tips of anterior dorsal-fin rays visible when adpressed). Distinguished from all other Arapaima except A. agassizii by having extremely long fourth infraorbital (Fig. 3A). Distinguished from A. arapaima and all other examined non-type specimens by notably slender body (Figs. 1A, 3B; holotypes of A. gigas, A. mapae, and A. agassizii also have slender bodies). Further distinguished from A. mapae and A. agassizii by having relatively deep caudal peduncle (6.0% SL, vs. ∼4.0 in both of latter two taxa), from A. gigas by having dentary teeth in a single row (vs. dentary teeth in 2–2.5 irregular rows), and from A. agassizii by relatively lower number of teeth on maxilla (28, vs. 43) and on single ramus of dentary (30–32, vs. 44; see Stewart, 2013, for additional characters that can distinguish A. leptosoma from A. agassizii).
Fig. 3.
Bivariate plots for: (A) fourth infraorbital length, and (B) body depth at pelvic-fin origin, comparing values for 28 specimens from Mamirauá Reserve (dark dots, and shaded areas = associated 95% confidence ellipses) with holotypes for four species of Arapaima: L = A. leptosoma, G = A. gigas, M = A. mapae, Ag = A. agassizii, and with a topotype of A. arapaima = Ar.
Description.—.—Morphometric characters (as % SL) presented in Table 1, and meristic counts as follows (counts for vertebrae, caudal, dorsal, and anal fins based on x-rays): caudal-fin rays 32 total, approximately 20 branched; dorsal-fin rays vi,34 or vii,33, 40 total; anal-fin rays iii,35, 38 total; pectoral-fin rays i,10; pelvic-fin rays i,5; lateral-line scales 52–53; complete scale rows between pelvic-fin origin and lateral line six; gill rakers (upper limb, first arch) ∼12; gill rakers (lower limb, first arch, including one at angle) ∼29; premaxillary teeth 32; maxillary teeth 28; dentary teeth 30–32; branchiostegal rays 11; pre-caudal vertebrae 41; caudal vertebrae 42; total vertebrae 83, including ural centra and first centrum closely associated with basioccipital.
Table 1.
Morphometric Characteristics (as % SL) for the Holotype of Arapaima leptosoma.
i0045-8511-13-3-470-t01.tifDorsal profile of head and body almost straight from tip of snout to near dorsal-fin origin, slightly upturned posteriorly; dorsal profile of body along dorsal-fin base sloping gently downward to caudal peduncle; as noted in diagnosis, anterior third of dorsal-fin base covered in enlarged, fleshy sheath; ventral profile of head and body as continuous, gentle convex arc from tip of dentary to caudal peduncle (curvature may be partially an artifact of preservation).
Dorsal fin originates just anterior to anal-fin origin, relatively low with longest rays in posterior half of fin; rounded posteriorly, with posterior rays extending beyond caudal-fin base when adpressed. Caudal fin with dorsal and ventral margins distinctly rounded, forming broad vertical ellipse. Adpressed pectoral fin extends about 36% of distance from its origin to pelvic-fin insertion, inserted low on body with base oriented almost horizontally (Fig. 2F; this may be a diagnostic feature that correlates with flattened underside of head and flattened breast; compare to angled pectoral-fin base in a Mamirauá specimen, Fig. 2E); first pectoral-fin ray unbranched, slightly shorter than first branched ray, posterior margin of pectoral fin distinctly rounded with rays 4–6 longest. Pelvic fin inserted about two-thirds of distance between tip of snout and caudal-fin base; outer, unbranched ray shorter than first branched ray; posterior margin of fin rounded with rays 3–4 longest; adpressed pelvic fin extends posterior to about 60% of distance from its origin to anal-fin origin; pelvic fin folds longitudinally between rays 3 and 4 when retracted against ventrolateral margin of body (as is typical of most Arapaima). Anus situated about midway between posterior tip of pelvic fin and anal-fin origin, but slightly closer to pelvic-fin tip. Anal fin inserts just posterior to dorsal-fin origin, with its posterior tip rounded and overlapping caudal-fin base when adpressed.
Mouth distinctly upturned with opening nearly superior, mandible extends beyond tip of snout a distance about one third of orbit diameter; mouth opening large and nearly as wide as snout; maxilla extends posteriorly almost to below center of orbit. Vomerine tooth-patch distinctly separated from tooth-patch on the parasphenoid. Eyes lateral to slightly dorsolateral, centered about one fifth of distance from tip of snout to tip of opercular flap (when seen in lateral profile). Opercular flap extends to posterior margin of cleithrum, not overlapping onto scales posterior to that or onto pectoral-fin base. Gill membranes overlap broadly along ventral midline, free from isthmus. Anterior and posterior nostrils situated on dorsum of snout; anterior nostril about midway between tip of snout and anterior margin of orbit, with short, unpigmented flap or papilla on posterior margin. Dorsal surface of head with numerous, irregular, bony ruggosities, not organized into clear longitudinal or transverse striations; likewise, third infraorbital lacking clearly developed longitudinal striations, and striations only weakly developed on fourth infraorbital (vs. various other Arapaima having distinct longitudinal striations on various bones of head; e.g., Fig. 2A, D).
Color in alcohol.—The following description of color patterns is based on the holotype, which was preserved in formalin and subsequently transferred to ethanol for permanent storage. Dorsal surface of head and body from near tip of snout to posterior end of dorsal-fin base dark, dusky gray. Side and ventrum of head below ventral margin of orbit and flank below lateral line lighter colored from mandible posteriorly to about anal-fin origin (perhaps white, silvery or cream-colored in life); flank darker between bases of dorsal and anal fins. Lighter pigment on posterior margins of scales along dorsal- and anal-fin bases, forming narrow streaks that angle anteriorly (Fig. 1A; in most live Arapaima of other species that I have seen, the light-colored scale edges and spots on the median fins range in color from pink to red).
Dorsal fin dark gray with relatively large, vertically elongate spots on posterior third. Caudal fin and peduncle with similar dark gray base color; large, sub-marginal spots arranged in arc around posterior margin of caudal fin; distal tips of middle caudal-fin rays darker, almost black (Fig. 1A). Most of anal fin uniformly dark, almost black, but lighter colored near origin. Both pectoral and pelvic fins lightly pigmented, perhaps white or silvery in life and matching lower flank and belly.
Etymology.—From the Greek leptos, slender, and soma, body, in reference to the relatively slender body form of this species.
DISCUSSION
This is the first new species of Arapaima to be described since 1847. It can be separated readily from the Mamirauá population (e.g., Fig. 2E, F), so it is clear that there are at least two species of Arapaima in the central Amazon. Arapaima gigas is the only freshwater fish in South America listed as an endangered species (i.e., CITES, Appendix II; but ‘Data Deficient’ according to IUCN, 2011). My review of materials of Arapaima in museums did not yield a second specimen of A. gigas, so its known distribution presently is limited to the type locality near Santarém, Pará State, Brazil. An important implication is that neither A. leptosoma nor the Mamirauá population is protected under international conventions; the same is true for A. mapae, A. agassizii, and A. arapaima. Another implication is that, like many other fish genera in the Amazon (e.g., Kullander and Ferreira, 2006), Arapaima may include common (e.g., Mamirauá population) or wide-spread species together with rare or geographically restricted species (Castello and Stewart, 2010; Stewart, 2013). It is time to set aside the idea that the genus Arapaima is represented only by one, panmictic species (e.g., Günther, 1868; Hrbek et al., 2005; Araripe et al., 2013).
The Mamirauá population has been estimated to number in the tens of thousands, but that is among the very few areas in the Amazon where good population count data exist (Castello et al., 2011). That is also an area where populations have recovered following implementation of community-based management programs. Elsewhere, especially around large cities, populations were seriously depleted as early as the 1920s (e.g., Eigenmann and Allen, 1942; Neves, 1995). If species distributions are restricted, then some taxa could be severely threatened, but available data are insufficient to judge such matters.
Fortunately for A. leptosoma, substantial protected areas have been established along the lower Purus River (i.e., Reserva de Desenvolvimento Sustentável Piagaçu-Purus with about 7,900 km2 and the smaller but contiguous Reserva Biológica de Abufari). It seems likely that A. leptosoma will persist in those areas because the extensive várzea habitat with numerous lagoons is ideal for Arapaima. Its rarity in museum collections probably reflects insufficient collecting efforts in the Purus River and adjacent areas of the Solimões River. Collecting adult Arapaima involves considerable difficulties, both with logistics in the field and subsequent storage in museums. As this study demonstrates, however, collecting at least a few voucher specimens can greatly enhance our knowledge of these fascinating fishes. Many more are needed. Arapaima have high economic, cultural, and scientific value, but their diversity has been overlooked for too long.