Astyanax cocibolca , a new characid ( Pisces : Ostariophysi ) from Lake Nicaragua , Central America

Astyanax cocibolca, a small filter-feeding species is a new characid species endemic to Lake Nicaragua. It is distinguished from the syntopic Astyanax aeneus by its high gill raker count, slender body, upturned snout, long maxillary bones, conical outer premaxillary teeth, lower number of scales above lateral line, more intense black lateral band, smaller size, and trophic niche partitioning. At present the new species utilizes the open-water habitat at least as a nursery where young can feed on the rich plankton population. Adults feed on copepoda, cladocera, fruits, seeds and algae. The bulk of the adult population presumably resides inshore. Rev. Biol. Trop. 56 (3): 1361-1370. Epub 2008 September 30.

wrote "In the difficult and highly interesting group of characins there is no more difficult nor more interesting genus than Astyanax."They further noted that a widely variable group of related species in Central America and Mexico, centering upon Astyanax rutilus, would require much larger series to determine their status.Now, 100 years later, little has changed in Middle American Astyanax taxonomy.Astyanax rutilus (=aeneus) and A. mexicanus still present a mosaic of difficult-to-define local geographic populations in Mexico and Central America.Several of these local forms have been recognized as distinct species especially if shown to be sympatric with another Astyanax species (Schmitter Soto 1998).Some distinctive populations with reduced or restricted distributions also merit recognition (Lozano-vilano and Contreras-Balderas 1990).The subject of the present study is a presumptive case of sympatric speciation of a semipelagial species in Lake Nicaragua.Spatial and trophic niche partitioning between the new species and the syntopic A. aeneus is indicated.Astorqui (1972Astorqui ( , 1975)), based on collections made in the Nicaraguan Lakes basin in 1960, proposed that two species of Astyanax were syntopic in the basin, but did not formally describe the new form.In 1971 biologists from the Fisheries Division of the Instituto de Fomento Nacional (INFONAC) in collaboration with FAO/UN, U.S. Peace Corp and investigators from several universities initiated a three-year study to evolve a rational plan for the development and management of the Lake`s fishery resources (INFONAC 1974, Davies 1976).On the initial cruises of the Program on the M/N Gaspar, I was able to capture fishes from throughout the lake and at diverse shore localities.This and other material collected by Jaime villa in 1966 and 1968 provided the specimens of Astyanax for this study.

MATERIAL AND METHODS
A variety of seines were used to collect at inshore sites, along beaches and around the Isletas of Granada.A few small individuals of A. cocibolca and other small species were taken incidentally while trawling in offshore waters for cichlids and other commercially important fishes.In order to sample juveniles and small species, a fine-mesh net was placed in the cod end of the try net on several attempts.This produced 60 specimens of juvenile and young adults of the cocibolca sardine (UCR 539-6) and other smaller catches.
Counts and measurements follow the methods given in Fink and Weitzman (1974).An additional measurement is Maxillary Length which refers to that portion of the bone extending from the rictus of the upper jaw to distalmost termination of the maxillary bone.Sex was determined by dissection.Body measurements are given in mm standard length (SL).All proportional measurements are in percentages of SL. vertebral counts were determined from cleared and alizarinstained specimens.The holotype is deposited in the Natural History Museum of Los Angles County (LACM); paratypes are deposited in LACM, Museu de Ciencias e Tecnologia, Porto Alegre, Brazil (MCP) and American Museum of Natural History (AMNH); other paratypes and all comparative material is deposited in the Museo de Zoología, Universidad de Costa Rica (UCR).
Description.Table 1 gives morphometrics of the holotype, 50 adult paratypes and respective data for 30 adult Astyanax aeneus Fig. 2. Astyanax aeneus, UCR 63-1, female, 56.2 mm (upper) and UCR 24-8, male 56.4 mm, Lake de Nicaragua.from Lake Nicaragua.Astyanax cocibolca is a small slender-bodied species of the genus.Predorsal profile convex; straight over nape, nape to origin of dorsal fin straight or slightly rounded.Dorsal profile of body along dorsal-fin base straight.Mouth superior, lower jaw projecting slightly in advance of tip of snout giving an upturned appearance to ventral margin of head.Greatest body depth between origin of dorsal fin and midpelvic fins; only a small overlap with the deeper-bodied A. aeneus (Table 2 and Fig. 2).No evidence of sexual dimorphism was found in meristics or morphometrics.
Live coloration of head and body is bright silvery; an underlying black midlateral band is notable only on the caudal peduncle and base of caudal fin, extending to the tips of the midcaudal rays.Coloration in alcohol is dependant on the duration in formalin which affects the silvery guanine pigment.Beneath the guanine layer the fish are tan above, pale straw color below; a black midlateral band of eye width extends between the upper margin of the opercle to tail base, forming a large caudal spot and continuing as a narrow stripe to the end of the middle rays.A diffuse humeral blotch extends slightly above the lateral band.
Ecology.Thirty-two species of freshwater fishes are found in the Great Lakes proper; fourteen other freshwater species have been collected in affluents to the lakes or tributaries of the River San Juan.The lake covers an area of 8,264 km 2 , the longest axis 160 km; maximum depth is about 70 m, average only 9-13 m (Incer 1973).Oxygen values of surface waters of 74-173% saturation imply intense photosynthesis (Cole 1976).The shallow depth and abundance of phyto-and zooplankton suggest very high productivity throughout this tropical lake.
Planktivorous fishes inhabiting Lake Nicaragua include the shad Dorosoma chavesi and the silverside Atherinella sardina.Although some Dorosoma were collected in open waters, A. sardina was taken only in shallow inshore habitats around the lake.Thus, the relatively empty small-planktivore niche apparently allowed A. aeneus to evolve a filter-feeding specialist, A. cocibolca.At present the new species utilizes the open-water habitat at least as a nursery where young can feed on the rich plankton population.Two small adults from the open water habitat contained 90% copepoda and cladocera.The bulk of the adult population presumably resides inshore where it can gorge opportunistically on plant and other food matter.Inshore specimens contained 90% fruit, 5% cladocera, 5% seed, algae.Twelve A. aeneus from inshore sites contained 90% fruit, husks, etc., 10% chironomid larvae and terrestrial insects.Thus, there is partial trophic niche partitioning between these two syntopic Astyanax species.The adult A.cocibolca female from open water contained ripe eggs, although ten adult females from inshore sites contained only undeveloped eggs.Only one of nine adult A.aeneus showed any egg development.Observations made in the lake during other times of the year presumably would produce different results.
Etymology.The species is named for Lake Cocibolca, the indigenous name for Lake Nicaragua.The name is used as a noun in apposition.
DISCUSSION AND RELATIONSHIPS Eigenmann (1921) considered Astyanax rutilus a synonym of Astyanax fasciatus, both species originally described from Brazil.He included five subspecies in fasciatus, one being A. fasciatus aeneus (Günther 1860) originally described from Oaxaca, Mexico and recently considered part of the "Astyanax fasciatus species complex" by Melo and Buckup (2006).In describing the new species Astyanax armandoi, Lozano-vilano and Contreras-Balderas (1990) reviewed the state of the taxonomy of Astyanax in Mexico.They tentatively recognize two wide-ranging species: Astyanax mexicanus in northern Mexico and Astyanax aeneus southward from River Papaloapan and Oaxaca into Central America.Eigenmann (1921) included Astyanax oerstedii (Lütken 1874), described from River San Juan (the effluent of Lakes Managua and Nicaragua), as a synonym of A. fasciatus; but he recognized as valid species A. nicaraguensis (Eigenmann and Ogle 1907) from "Nicaragua" and Deuterodon nasutus (Meek 1907) from Lake Managua, Nicaragua.Regarding A. nicaraguensis, Eigenmann stated "In general characters the specimens agree with the specimens of A. aeneus, and those with but two maxillary teeth are indistinguishable from them.The fact that such a large per cent of specimens have a large number of maxillary teeth entitles them to a separate name."He thought it possible that specimens with numerous maxillary teeth are all males, although this is not the case.Bussing (1998) considered this Lake Nicaragua population a case of speciation in progress whereby numerous maxillary teeth provide some advantage to this lacustrine population of A. aeneus.Typical Central American aeneus have 1-3 maxillary teeth, usually 2, but in tributaries of Lake Nicaragua such as River Sapoá they usually have 1-5 teeth, average 2.9 compared with 1-9, average 4.5 in the lake specimens (Table 6).This suggests that the lacustrine aeneus are not a closed gene pool, and some selective pressure maintains a high tooth count in spite of intergrading with fluviatile populations.A large number of maxillary teeth may improve capability of capturing more active and larger prey organisms.The sympatric Bramocharax bransfordi with 6-19 maxillary teeth and long snout (Rosen 1970) is a formidable predator.Eighteen B. bransfordi from the River Puerto viejo, a Costa Rican tributary of the River San Juan, ate 51% fish, the remainder mostly insects (Bussing 1993).Meek (1907) described Astyanax nasutus from Lake Managua and A. aeneus from both Lake Managua and Nicaragua.A comparison of the two species descriptions reveals very little difference: a longer snout and maxillary, a concave profile between nape and end of snout, larger size, and a prominent chin in nasutus.Eigenmann (1921) redescribed three paratypes of nasutus of about 104 mm and included a fine drawing of one of the specimens.Although he stated the stomach was filled with plants, the species is a very large and formidable-looking Astyanax.He placed nasutus in the genus Deuterodon based principally on the teeth of the lower jaw being regularly graduate from in front to the last tooth on the sides as opposed to Astyanax with teeth of the sides of the lower jaw more or less abruptly minute.He also noted a greater head length than that given by .I examined two (80.7 and 81.8 mm) paratypes of nasutus (IU11600), and also found greater head length than typical Lake aeneus (29.6-29.8vs 24.9-25.5% SL).Lucena and Lucena (2002) redefined the genus Deuterodon, and on the basis of three synapomorphies, recognized seven species, all from South America.These synapomorphies of the maxillary bone exclude the nominal D. nasutus, which they considered incertae sedis.Upon request from the Lucenas in 2001, I donated material of what I had considered A. nasutus (UCR 24-16).They determined that the specimens were not a Deuterodon, nor representative of the species nasutus, and considered it an Astyanax sp.A more careful analysis of a large series of this material and the literature bear out their conclusions and resulted in the present study of the previously undescribed Astyanax cocibolca.Astorqui (1972Astorqui ( , 1975) ) reported on a large collection that he and L.R. Rivas made in 1960 in the Great Lakes basin (Lakes Managua and Nicaragua).He concluded there were two valid species of Astyanax in the basin: A. fasciatus (including as synonyms aeneus, nicaraguensis and nasutus) and Astyanax sp.The latter was distinguished from fasciatus by its slender body, upturned snout, and very dark lateral band with diffuse humeral spot when preserved.Astorqui's recognition that Astyanax sp. was not assignable to any known species was correct and not adequately recognized in subsequent studies of Nicaraguan fishes.
No specimens of A. nasutus have been reported since the original description of 12 specimens from Lake Managua and one specimen from Cisplaya, a small nearby lake.The fact that nasutus was found only in Lake Managua and the nicaraguensis-type aeneus and cocibolca are known only from Lake Nicaragua support other faunal evidence that River Tipitapa connecting the two lakes is a virtual barrier to free passage due to a sandstone dropoff, hot springs and lack of running water (villa 1976).Since sometime in the past, waters rarely flow through the Tipitapa which at present is a slow-moving swamp filled with water hyacinth Eichornia azurea (Incer 1973(Incer ,1976)).This isolation has permitted the Lake Managua nasutus to diverge in dentary and other traits.Astyanax nasutus was reported to have 4 to 6 maxillary teeth, but I counted 7-8 maxillary teeth on one of the two paratypes I examined.Thus, like A. aeneus in Lake Nicaragua, it has an increased number of maxillary teeth and modified lower jaw dentition as well.Regrettably Astorqui did not obtain nasutus while collecting at Lake Managua, as I believe he would have distinguished it as distinct from aeneus from the same lake.
Comparative Material: Astyanax aeneus was collected together with A. cocibolca at six stations: UCR 58,1,46.6 mm;22,39,52,45, mm.Astyanax aeneus was also collected at the following Lake Nicaragua stations where no A. cocibolca were present: UCR 63-1, 7, 41.5-62.5 mm, Isletas de Granada, 20 Jun 1966. UCR 546-8, 71, 23.2-70.0 mm, shore mouth of River Tipitapa, 25 Nov 1971. UCR 548-3, 44, 17.3-31.4 mm, shore at Bahía La Pelona, 26 Nov 1971.ACKNOWLEDGMENTS I am grateful to R. Camacho of the Ministry of Agriculture and the biologists of INFONAC, Nicaragua; to the FAO/UN fisheries biologists R. Ellis, S. Lofnes and A. Hagberg who designed the fisheries resources program; to C.A.S. and Z.M.S. Lucena for bringing to my attention that specimens of A. cocibolca were not A. nasutus; to A. Arostegui, R. Beatty, K. Koenig, L. Reyes and J. villa who collaborated in the collection of specimens.Special thanks go to I. Astorqui, S.J. for recognizing Astyanax cocibolca as an undescribed species early on.

TABLE 1
Body proportions for type specimens of Astyanax cocibolca n. sp. and Lake Nicaragua Astyanax aeneus in percent of standard length Frequency distribution of greatest body depth of Astyanax cocibolca and Astyanax aeneus in percent of standard length Frequency distribution of the number of total gill rakers in Astyanax cocibolca n. sp. and Astyanax aeneus Frequency distribution of maxillary length of Astyanax cocibolca and Astyanax aeneus in percent of standard length note serrations on some large males of A. cocibolca and aeneus (Table5, fig.3and fig.4).Maxillary teeth 1 to 4, mean 2.1 teeth (Table6); teeth very small, compressed, with 2 to 5 rounded cusps, central cusp not pronounced, holotype with 1-2 maxillary teeth.

TABLE 6
Frequency distribution of the number of maxillary teeth in Astyanax cocibolca n. sp. and Astyanax aeneus