PROC. BIOL. SOC. WASH. 101(2), 1988, pp. 444-465
MINIATURIZATION IN SOUTH AMERICAN FRESHWATER FISHES; AN OVERVIEW AND DISCUSSION
Stanley H. Weitzman and Richard P. Vari
Abstract.— MiniatuTQ species within the South American freshwater fish fau- na are hsted and their phylogenetic and ecological distribution is analyzed. Criteria used in identifying miniature South American freshwater fishes are briefly discussed and eighty-five such fishes of five ordinal groups, Clupei- formes, Characiformes, Siluriformes, Cyprinodontiformes and Perciformes, are listed. Approximately eighty-eight percent of these miniatures are characiforms and siluriforms, all primary division ostariophysan fishes. Phylogenetic infor- mation on the relationships of these miniatures indicates that such species have arisen in the freshwaters of South America about sixteen times within the Characiformes, twelve in Siluriformes, four in the Cyprinodontiformes and once each in the Clupeiformes and Perciformes. Miniatures of most groups occur in lentic or slow-flowing waters, but the one pimelodid and some tricho- mycterid catfish miniatures are cryptobenthic in swift-flowing streams.
Resumo.—Feixes miniatura de agua doce da America do Sul sao definidos e uma lista contendo 85 especies miniatura, pertencentes a cinco diferentes ordens de teleosteos, e fornecida. Peixes miniatura de agua doce sul americanos sao arbitrariamente definidos como aqueles que, ate onde se sabe, nao excedem 26 milimetros de comprimento padrao, sendo que a maioria atinge a matu- ridade sexual com menos de 20 milimetros de comprimento padrao. As especies incluidas entre estes limites tem em comum varios caracteres pedomorficos (redutivos) quando comparadas com taxons relacionados. Duas especies mi- niatura sao conhecidas dentre os Clupeiformes, 49 dentre os Characiformes, 26 dentre os Siluriformes, cinco dentre os Cyprinodontiformes e tres dentre os Perciformes. Aproximadamente 88% destes peixes miniatura sao Ostariophysi da divisao primaria. A informagao disponivel sobre as relacoes filogeneticas da maioria dos peixes miniatura e de qualidade questionavel; no entanto, nos provisoriamente esti- mamos que especies miniatura surgiram independentemente pelo menos 34 vezes na fauna de peixes de aqua doce da America do Sul; 1 6 dentre os Cha- raciformes, 12 dentre os Siluriformes, quatro dentre os Cyprinodontiformes e uma unica vez dentre os Clupeiformes e dentre os Perciformes. A informa^ao sobre a ecologia de peixes miniatura e tambem escassa mas, como seria de se experar no caso de peixes diminutos, a maioria ocorre em aguas lenticas ou de fluxo lento. No entanto, um pimelodideo miniatura e alguns tricomicterideos miniatura sao criptobenticos em riachos de fluxo ra- pido. VOLUME 101, NUMBER 2 445
The South American freshwater fish fau- species known from South American drain- na is one of the most diverse in the world, ages. with numerous species unknown to science being continually discovered, and a large Miniature Fishes in South percentage of the fauna still undescribed American Freshwaters (Bohlke et al. 1978). The early collectors and researchers engaged in studies of the The range in body sizes among fishes, ichthyofauna of that continent failed to ap- bony and cartilaginous, is remarkable, cov- preciate the degree of this diversity, partic- ering the spectrum from diminutive fishes ularly among smaller species, and as a con- under 26 mm standard length (SL), such as sequence often focused on the more obvious those discussed in this paper, to species such species of larger adult body size (Myers 1 947: as the whale shark which achieves lengths 13-14). Subsequent research by various au- of over 12 m and possibly up to 21.4 m thors has resulted in a better appreciation (Campagno 1981). Whereas the larger of the large number of species of smaller species of fishes are typical of open waters, body size, but it is only in the last few de- particularly in marine environments, small- cades that a number of species of very small er species, especially miniatures, are usu- size as adults have been described. The pace ally, although not exclusively, associated of description of such miniature species has with more complex substrates and ecosys- quickened in the last decade and recent col- tems. This correlation is reflected in the lecting efforts in various regions of South ichthyofaunas of South American streams America revealed the existence of many in- and rivers, and of Indo-Pacific coral reefs, teresting miniature species (Stewart 1985, which are the most diverse in the freshwater
1 986; M.Weitzmanl 985; Weitzman& Fink and marine environments respectively. 1985; Gery & Uj 1986; Weitzman 1986, Correlated with that high species diversity 1987; Fernandez & Weitzman 1987; Weitz- is the presence in those habitats of a large man & Vari 1 987). The comparative studies number of species of small fishes of ap- associated with the description of some of proximately 25 to 70 or 80 mm SL as adults, those species (Weitzman & Vari 1987) dis- together with many miniature species. In closed that the number of miniature species Table 1 we list 85 nominal species not known is much greater than commonly recognized. to exceed 26 mm SL, many of which are Indeed, during the twelve decades that have known to mature at under 20 mm SL. These passed since the Thayer Expedition about are considered to be miniature fishes in this 85 freshwater South American species (Ta- paper. Although the term "miniature" im- ble 1) that are miniature as adults, or not plies a very small body size, our primary known to exceed 25 to 26 mm in standard criterion for distinguishing miniature fishes length in nature, have been described from is not solely size related. Rather it involves the freshwaters of South America. the presence in such miniatures of a large In the present paper we discuss the phy- number of apparently paedomorphic mor- logenetic distribution of miniatures in the phological features including the degree of Neotropical freshwater ichthyofauna, and development of the laterosensory canal sys- provide an estimate of the number of min- tem of the head and body, reductions in the iaturization events within this fauna. The number of fin rays and body scales, and a available information on the ecology of diminution of the sculpturing on the surface
South American miniature fishes is sum- bones of the head. Such paedomorphic fea- marized, and we evaluate a previous con- tures in fishes are commonly referred to as cept of the predominance of secondary reductive in the systematic ichthyological freshwater fishes among the miniature literature and are considered to be due to — 1 H I( _
446 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON
tn N NO <^ '53 r~-" c ^ in 00 Cn) >o t^ t^ cfl 00 , 1^ l/^ ^ (N ON fN fN '* * o o o O oo „ in ^^ ^^ f— »— e T— NO NO N 00 r- 00 ,__, iri tri >ri Cr, Cr, in m c ( t~-' t-~-' ri 03 ON On 00 00 00 00 00 'C 00 00 oo 00 m ON On ON ON ON ON ON e On on c^ N ^ On on" o o X! X3 m '^ o 00 X3 c -d 66 cp 00 fN O O c c o 13 c c c '^ in O 00^ t-~-' fN U u > > II! (I! ti; oo fN t-^ NO 1^ 1^ 00 -H • oo ON —( On ON ON o -H -H -H On ^- On 00 ON ^ =a ^ ^ ^ =S5 =y =^ in ^ ^ <^ <^ — On ^^ o o N c c c c c c c ON c c c c c r(U %(U \(3J 3 ti^ ^^ ^^ ^^a^ 2 ^^^^^ OOOOQOQ^^
u-i o f^ -; ^_ 00 o in "". °°. 1 00 NO 1^ in NO in On On q q q q q"' q ON iri NO rn ro r~-' NO -! rsi in (N r-' O On rn r<-> d NO d r-' — ON 00 rn P-' -H ^H —1 (N (N (N fS (N (NoofNOinNOOOin 'itr-^ONNdrnr'^fN-^rn C^- C^- C^- (>• 00 i-t ON On C O 5^ O ii; in 0> O fT . -rH b ^ ON 00 On 00 00 — m m ^_; ' C On ^ c ON ON ^ .« «a 03 c ^^ ON 0^ •P N N S r2 r2 OnrN '— C .2P o~> ss^ -^J^ ON '53 C (1) N S2 ^ op c — XI S 03 ^ 5 iS 05 s: c c s; ^i (J VOLUME 101, NUMBER 2 447 B e B s o o o u £3 O^ ;r - q — — NO (N i/-> 00 ON —1 NO OS vo u, t: t^ o S o ^ On On r-j (^ vO — OI^^Ot^OOONOOOt^fNNOminONiONO in'^tON—'COOi/^O fN o NO O —; O r-; q p q ^' ON NO t~~' r~-' r-' ON •Ti- I--' r-^ NO r<-i o — —' (N V — ON ON — On On on i-. it o ON — CC (^ — NO C C ON ON C (U en —< „^ CQ c - O ^ N x: -- ^ (U C« 1- ON "p J 5 Ph S <^ Z2 c S S; Si) O S S ^ > ?u .~ .s: S k C -^ o "^ ^ s S ^ So 5 S -2 c P O -2 21 S s; s: s: S -^ o o -2 ^ S" -S -s: ^j ON On !U 00 00 (N ON OO —I ^ 00 DO OO t^ C —H -r. i?; 'sf — fN fN vi> vi) vo •• NO i5 r~-' c c o: r- I- -i °° (2 ON On on ON g; ON 00 I it s s -- 6 --:i:-_^ c c lU _o s x; B » ** 3 o q q q q Al q q ON NO r<^ Tt (N (N ^^ —' r^ E -^ '^ o •n fN f<-) j^. j^. fv. 00 <^- t^. ir^ 00 00 On C - X) X3 — NO NO ^ P3 ^ t^ t^ C NO C3 On On ON OJ (^ c Zh Z. « 00 (S HJ ^^ S (u u Z ON Tt -J-; >n ^ ON K *= ^ o :3 :3 U NO W NO E u. Ui c :3 ON 2 ON ^ <^ 'S X) XI S J "J^ s ' t« « =0 ON ON U5 > f2 XI -^ U > ^ ^ e < 55 55 rt a 's I Si. ci, c > ^ a a a a ?^ iJ ?? 5 ^ V ^ ^ bo 03 o C) o o ^ =^ c '^ OJ O "O Si -& -Si -Si 3 P s -a ?a nj us X. V v^ *-> ^ ^ ^ ^ JS X! s (u a « Q c o o o o c -^ ^ a, a, w fi; §^ •3 T3 Q, ^ e; t^ f; ^ CO UUU 13 u ' VOLUME 101, NUMBER 2 449 ^ m m C! O t^' r~^ C rN (N >^ OS— OS ' — C - 03 «n i« S X . >. C U s s o ON OS "^ 3 ^ ^ 5i -3 OS i2 ;i >. T3 £ g X 6 03 cat: 5 O -5"^ §1 a Q 5 'S 1 1 2 £ c '-' 2 o •S c (u ,_, .g a a ^C •a ^ o 03 .n ^ ^ c ? n •fi 2 00 'C 'G 03 ITi a. o Q, OJ 5: c ^ E 3. % ^^ Cu d: U # * 2 * U 2* 450 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON the loss of terminal stages in the develop- natural habitat further limits the adult size mental sequence. Myers (1958:29) dis- of the species. Most miniatures probably cussed various characters commonly dem- live less than one year in the wild, perhaps onstrating paedomorphic features in South considerably less. American freshwater fishes, and those fea- The apparent paedomorphic nature of tures will be discussed and evaluated in miniature species in Table 1 at maturity, greater detail in a future paper. For the pur- and the reductive nature of some characters poses of this paper, the miniature South are recognized within the context of the de- American fishes listed in Table 1 satisfy velopmental sequences of the characters in either oftwo criteria. They are either species apparently related outgroups maturing at which mature sexually at under 20 mm SL longer lengths. The phylogenetic hypotheses and may sometimes reach somewhat larger available for most taxa are of an imprecise sizes, or species for which maturity data are nature. Thus the documentation of paedo- unknown but which are not reported to ex- morphosis with respect to proximate sister ceed 25 to 26 mm SL in the wild. Species groups involves general rather than specific examined by us or reported in the literature phylogenetic comparisons. Precise infor- as getting larger than that size, even though mation on morphology and size at maturity they may mature at lengths somewhat is not available for some taxa, a number of shorter than 20 mm SL are not included. which are known only from limited series. This cut-off point for "miniatures" is ar- In those instances we extrapolate that the bitrary and the listing is meant as a prelim- correlations between a body size under 25 inary guide to future studies of miniature to 26 mm SL and the presence of paedo- fishes in South America rather than as a morphic features that are characteristic of definitive enumeration of such species. other more inclusive taxa also applies to the The size limitation used in the present species we have been unable to examine study was adopted because all listed species critically. It is also necessary to note that that have been examined in detail which maturity data are absent for a number of mature under 20 mm SL, and do not grow the listed species. It is possible that future to longer than 25 to 26 mm SL, have in studies may reveal that adults of some of common numerous apparently paedo- these taxa achieve sizes greater than the cut- morphic morphological reductions. A few off" used to delimit miniatures in this paper. species in Table 1 are known to somewhat The size criterion for miniatures in this exceed 26 mm SL when raised and kept in study is not completely satisfactory when aquaria. These include Corydoras habrosus applied to some unlisted species with elon- Weitzman, Paracheirodon innesi (Myers), gate bodies. For example the trichomycterid Hyphessobrycon flammeus Myers, Hyphes- catfishes Typhlobelos ternetzi Myers and sobrycon griemi Hoedemann, and Hyphes- Glanapteryx anguilla Myers, have small sobrycon loretoensis Ladiges. Some of these heads displaying many paedomorphic fea- species may exceed this length in nature, but tures of miniatures, although the fishes ex- that remains unconfirmed. Geisler & An- ceed 30 mm SL. Similarly Tomeurus graci- nibal (1984:77-78) have discussed aquari- lis Eigenmann, an elongate poeciliid, is in um versus wild lengths for Paracheirodon many ways a miniature fish, but we have axelrodi (Schultz), a fish that exceeds our examined both males and females of 26.5 size limit, but which displays paedomorphic mm SL from Surinam (USNM 225463), and features (Weitzman & Fink 1983:354-357). as a consequence do not include the species Geisler & Annibal believed that individuals in our list. These three species are in ac- of this species rarely live more than one year tuality elongate miniatures in which the head in the wild, and that poor nutrition in its length, but not body length agrees with that VOLUME 101, NUMBER 2 451 of the species in Table 1. The concept of an last 40 years. Bohlke & Robins (1968), Jew- elongate miniature is perhaps even more ett & Lachner (1983), Lachner & Kamella appropriate for the South American fresh- (1978, 1980), Winterbottom (1983), and water needlefishes of the genus Belonion Winterbottom & Emery (1981, 1986) pro- Collette. Collette (1966:9, 15-21) in his de- vide many examples of miniature marine scription of the genus and contained species gobies, and references in those publications detailed numerous reductive characters for will lead to other descriptions of such the genus relative to outgroup needlefishes species. Springer (1983), in describing a new involving the epipleurals, fourth upper pha- miniature species of western Pacific goby, ryngeal tooth bones, laterosensory canal has documented several skeletal modifica- system and other features. Although Belo- tions, of which some appear paedomorphic nion dibranchodon Collette matures at 24.5 to us. More recently Whitehead and Teugels mm SL and members of the genus reach (1985) have described apparently paedo- 41.8 mm SL (Collette 1966:12-13), the morphic features in Sierrathrissa leonensis species are very elongate. In fishes of such Thys van den Audenaerde a species of Af- shapes it might be more appropriate to com- rican freshwater clupeid that reaches a pare species on the basis of the bony head slightly greater length, 26.7 mm SL, than length (exclusive of the elongate lower jaw the 26 mm SL limit on miniature fishes. in Belonion) rather than focusing on stan- Finally, Roberts (1986:332-340) reported dard length. In such a framework more many similar reductions in a miniature species such as Belonion dibranchodon Asian cyprinid, Danionella translucida would be listed as miniatures. A criterion Roberts (1986:233). of size using such a portion of the body with numerous paedomorphic features could be Phylogenetic Lineages of South American similarly used to delimit miniature species Freshwater Fishes regardless of their actual standard length. Although the data from Belonion and The large number of species of miniature some characiform groups suggests that some fishes within the South American freshwa- alternative standard for miniature fishes ter ichthyofauna is striking, but a true sense might be appropriate, morphological data of the possible significance of this phenom- on paedomorphic features are unfortunately enon can only be achieved by examining largely absent for many relatively small miniaturization within a phylogenetic species of some major groups of South framework. In other words, in which major American freshwater fishes (e.g., Gymno- groups of fishes is miniaturization found, tiformes, Cichlidae) slightly longer than 26 and how many times has it independently mm SL. Similarly there is a paucity of data evolved? on how and whether paedomorphic char- Not surprisingly, the two largest groups acters correlate with standards such as ab- of South American fishes, the ostariophysan solute head size in most groups. Thus for orders Characiformes and Siluriformes, ac- the sake of consistency we prefer to use stan- count for the vast bulk (88%) of the mini- dard rather than head length or another atures, while the Clupeiformes, Cyprini- measurement as the benchmark for our dis- dontiformes and the Perciformes which are cussion until more data are available. less speciose on that continent cumulatively If the concept of miniature fishes of this include only about 12% of the miniatures. paper can be extended to other groups of These five orders are each monophyletic and fishes, then the majority of the miniature the evolution of miniatures in each of these fishes, in both freshwater and marine en- clades has clearly proceeded independently. vironments, have been described within the The question of how many times miniatur- 452 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. L Amazonsprattus scintilla Roberts, a miniature clupeiform fish, family Engraulidae, from the Amazon basin, AMNH 57371, 24.3 mm SL; Brazil, state of Amazonas, Rio Urubaxi, tributary of Rio Negro, south of east end of Ilha Tamaquari (Or32'N, 64°59'W). ization has evolved within each hneage can- Interestingly no miniature species are not be answered with the same precision, known to occur in ten of the twelve Neo- but we will address the issue to the extent tropical characiform families recognized by possible within the context ofavailable phy- Greenwood et al. (1966:395-396), although logenetic data. We must preface our general in at least one, the Gasteropelecidae, there discussion with the observation that there are several species of Carnegiella Eigen- are relatively few rigorous phylogenetic mann displaying numerous apparently pae- studies of components of the South Amer- domorphic features (Weitzman 1954, ican freshwater fish fauna, and many tra- 1960b). Many of the characiform families ditionally accepted taxa have not been crit- without known miniatures demonstrate ically reviewed with respect to their possible marked morphological adaptations for spe- monophyly. In light of this lack of precision cialized feeding methods (e.g., Curimatidae, we conservatively do not hypothesize mul- Prochilodontidae, Anostomidae and Chilo- tiple miniaturization events within a genus dontidae, Vari 1983:7-33, 43-46; in press), unless phylogenetic data indicate that re- which may be incompatible functionally duced size and its attendant paedomorphic with a pronounced reduction in overall body features have independently evolved more size. Miniatures are rather found only in the than once in the taxon. Lebiasinidae and Characidae among the Only two species of miniature clupei- South American characiforms. Within those formes are described from South American families the miniatures are distributed freshwaters. Roberts (1984:321) withheld among eighteen genera, one lebiasinid and judgement on the phyletic associations of seventeen characid, most ofwhich are poor- Amazonsprattus Roberts (Fig. 1), but fur- ly understood in terms of intrageneric phy- ther study by Gareth Nelson (pers. comm.) logenetic relationships. indicates that it is a member of the Engrau- Lebiasinids show a considerable range in lidae, as is Anchoviella Fowler, the other overall standard lengths with Pyrrhulina South American clupeiform genus with a Valenciennes and Lebiasina Valenciennes described miniature species. A phylogenetic each having species reaching 100 mm SL, hypothesis that would provide insight into and species of Piabucina Valenciennes the relationships ofthe two miniature South growing to about 200 mm SL. The three American engraulids is not available, and species of Nannostomus Giinther listed in currently it is not appropriate to hypoth- Table 1 are members ofa genus of 1 5 species esize more than a single evolutionary min- that includes numerous species of about 25 iaturization event in the group. to 35 mm SL and it is not surprising that VOLUME 101, NUMBER 2 453 several miniatures are found in that taxon. 25 to 35 mm SL size range. Thus, the mi- According to Weitzman (1964:152-153) nute size of the Priocharax species is con- Nannostomus forms a derived clade within sidered to be the consequence of a single the Lebiasinidae. That hypothesis along with ancestral miniaturization. the larger body sizes in the lebiasinid out- The next two subfamilies, the Glandu- groups, and the small but not minute body locaudinae and Tetragonopterinae (includ- size of other Nannostomus species indicates ing Cheirodontinae), both of which contain that the minute size of the miniature Nan- many miniatures, are apparently closely re- nostomus is derived. Attempts to resolve lated groups, neither of which has, however, the relationships within Nannostomus have been shown to be monophyletic (Weitzman not proved feasible (Weitzman & Weitzman & Fink 1983:341-346; Weitzman et al. in 1982:417-419; Fernandez & Weitzman Weitzman & Fink 1985:1 12-1 17). The phy- 1987:165). In the absence of that data we logenetic relationships of the miniature are unable to reasonably hypothesize glandulocaudine genera lotabrycon Rob- whether the miniaturization process in that erts, Xenurobrycon Myers & Miranda-Ri- genus has occurred one, two, or three times, beiro, Scopaeocharax Weitzman & Fink, but conservatively assume that it took place and Tyttocharax Fowler (Table 1 ) were ana- only once. lyzed phylogenetically by Weitzman & Fink The absence of reliable hypotheses of (1985:12-70) who hypothesized with con- relationships among the vast majority of siderable evidence that they form a mono- subfamilies and genera in the Characidae phyletic unit. The sister taxon to that as- similarly does not permit us to advance any semblage, the genus Argopleura Eigenmann, critically evaluated hypothesis on the exact has adults that range from 35 to 53 mm SL number of evolutionary events represented and the miniaturization in the four cited by the occurrence of miniatures in 1 7 genera xenurobryconin genera is apparently the within the family. These genera are cur- product of a single evolutionary event. rently assigned to four subfamilies, Chara- Few tetragonopterine genera have been cinae, Glandulocaudinae, Tetragonopteri- the subject of comparable studies, and for nae (including Cheirodontinae) and most it is only possible to tentatively pre- Characidiinae. We prefer not to recognize a sume monophyly for multispecific taxa. One Cheirodontinae separate from the Tetra- exception is the genus Paracheirodon Gery gonopterinae in light of the difficulties in hypothesized to be monophyletic by Weitz- advancing an unambiguous diagnosis of the man & Fink (1983:347-357), but of uncer- Cheirodontinae that were discussed by Fink tain phylogenetic relationships within the & Weitzman ( 1 974: l),Gery( 1977:543) and Tetragonopterinae. The genus consists of particularly Weitzman & Fink (1983:341- three species, P. simulans Gery and P. in- 353). nesi, both miniatures, and P. axelrodi which The uncertainties associated with hy- reaches an adult size of 33.1 mm SL in the potheses of the exact phylogenetic relation- wild, although Geisler & Annibal (1984:32) ships of Priocharax Weitzman & Vari with- report an old aquarium specimen that in the Characinae were discussed by reached 51.0 mm SL. Thus the genus evi- Weitzman & Vari (1987:648-651). None- dently represents a single evolutionary shift theless, the data demonstrate that the two to reduced body size with subsequent min- species form a monophyletic group within iaturization. The hypothesis of the mono- the Characinae, a putative group of typically phyly of Paracheirodon as proposed by larger size fishes, although some taxa Weitzman & Fink has been challenged by (Gnathocharax Fowler, Heterocharax Ei- Gery (1984:59) and Gery & Mahnert (1986: genmann and Hoplocharax Gery) are in the 49) who maintain that P. simulans belongs 454 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON to the large tetragonopterine genus Hyphes- tures (Gery 1973:117) as is the case with sobrycon Durbin. A reanalysis of the data Axelrodia (Gery 1965a:31-35, 1966:111- of Gery (1984) and Gery & Mahnert (1986) 118, 1973:111-115) and Oxybrycon (Gery by Weitzman & Fink (1987) has recon- 1964:15). Microschemobrycon, in contrast, firmed the placement of P. simulans in the contains a number of species of 30 to 35 monophyletic clade Paracheirodon. mm SL (Eigenmann 1915:53; Gery 1973: The lack of information on phylogenetic 86-101, 1977:598). relationships within and among the re- The last characid subfamily to be consid- maining tetragonopterine taxa containing ered and one containing eight described miniatures precludes any indepth critical miniatures in three genera (Elachocharax discussion of the number of times that min- Myers, Klausewitzia Gery and Jobertina iaturization arose among these taxa. We Pellegrin) is the Characidiinae. Once again prefer to take a conservative course and as- the information on relationships within the sume that miniatures in genera with two or subfamily is limited, with only Elacho- more species arose from single ancestral charax having been the subject of phylo- miniaturization events. That assumption genetic treatment (Weitzman & Gery 1981; may be disproved when the necessary phy- Weitzman 1986). Elachocharax is hypoth- logenetic studies are undertaken, indicating esized to be monophyletic with four known multiple independent miniaturization species, all miniatures, and the reduced adult events in some of the taxa. Alternatively, size assumed to be a consequence of a single such analyses may also reveal presently un- evolutionary event. The situation in Klause- suspected sister group relationships be- witzia is more equivocal, with the two de- tween some of the genera, with a consequent scribed species, both of which are minia- reduction in the number of hypothesized tures, perhaps united by a single character, miniaturization events. The involved tetra- the presence of a maxillary tooth, a char- gonopterine genera and the number of in- acter that may be either plesiomorphous or cluded species (in parentheses) are as fol- a derived reversal (Weitzman & Kanazawa lows: Hyphessobrycon (12), Tyttobrycon 1977:157-159). Gery (4), Axelrodia Gery (3), Oxybrycon The phylogenetic problems associated Gery (1), Megalamphodus Eigenmann (1), with Jobertina and the evidently most closely Microschemobrycon Eigenmann (1), Chei- related genus, Characidium Reinhardt are rodon Girard (1), Hemigrammus Gill (1), complex and we neither have data on and Spintherobolus Eigenmann(l) (Fig. 2). whether Jobertina is monophyletic nor on Although we only count the miniature whether the two known miniatures are sister species of Hyphessobrycon as being the re- species. We can only note that some of the sult ofa single ancestral miniaturization, the species of Characidium achieve adult body genus is relatively speciose (over 60 nomi- sizes of over 1 10 mm SL and that the re- nal species) and includes a number of species duced size of the two Jobertina species is in the 26 to 35 mm SL range (Weitzman apparently a derived feature. Miniaturiza- 1977:355-356; M. Weitzman 1985:805) in- tion in the Characidiinae is thus counted as creasing the likelihood that miniaturization having occurred three times, once each in may have occurred multiple times in the Elachocharax, Klausewitzia and Jobertina. genus. It has also been suggested that Hy- We note, however, that most of the mini- phessobrycon may be polyphyletic (Weitz- ature species of these genera have in com- man & Fink 1983:342), a possibility that mon various features (e.g., elongate pectoral further complicates the question of the fin rays, Weitzman & Kanazawa 1978:160) number of miniaturization events in the ge- that may represent synapomorphies for nus. All species of Tyttobrycon are minia- those taxa. If that is the case, miniaturiza- VOLUME 101, NUMBER 2 455 Fig. 2. Spintherobolus broccae Myers, a miniature characiform fish, family Characidae, from southeastern Brazil, USNM 287324: (A) mature female, 22.0 mm SL, (B) mature male, 18.5 mm SL; Brazil, state of Rio de Janeiro, small tributary of Rio Macacu, about 1-2 km from town of Cachoeiros de Macacu (approx. 22°29'S, 42°41'W). tion may have arisen only once in the Cha- placed it in the monotypic Scoloplacidae. racidiinae, but the problem requires more That shift fails to elucidate the phylogenetic study. relationships ofthe single contained species, Phylogenetic analyses of South American but that classification is followed for the freshwater siluriform families and genera are purposes of this discussion since the alter- very rare. According to present classifica- native proposed alignment is no more sat- tions (Myers 1944, Myers & Weitzman isfactory. 1 966, Bailey & Baskin 1 976, Isbriicker 1 980, The eight genera of miniature Tricho- Stewart 1 985) the miniature siluriforms list- myteridae are found in five subfamilies, ed in Table 1 are distributed among six fam- Trichomycterinae {Trichomycterus Valen- ily group taxa, although some authors dis- ciennes), Glanapterygiinae {Pygidianops agree on the placement of Scoloplax Bailey Myers), Vandelliinae {Paravandellia Miran- and Baskin. Under all the classificatory da-Ribeiro), Tridentinae {Tridensimilis schemes, miniatures are found in the As- Schultz, Miuroglanis Eigenmann & Eigen- predinidae, Callichthyidae, Loricariidae, mann, and Tridentopsis Myers (Myers 1 944: Pimelodidae and Trichomycteridae. Sco- 599) (Fig. 3)) and Sarcoglanidinae (Sarco- loplax, originally placed in the Loricariidae glanis Myers & Weitzman, and Malaco- by Bailey & Baskin (1976), was removed glanis Myers & Weitzman (Myers & Weitz- from that family by Isbriicker (1980) who man 1966:279)). The Trichomycterinae 456 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 3. Tridentopsis sp., a miniature South American siluriform fish, family Trichomycteridae, 23.8 mm SL; aquarium specimen, no definite collecting locality infiDrmation. contains species ranging from 16.0 to over to be conservative and to hypothesize only 260 mm SL as adults (Eigenmann 1918b: a single miniaturization within each genus. 286, 326; Fowler 1932:368). The phyloge- Miniaturization is consequently considered netic relationships of the three described to have arisen at least twice in the Cal- miniatures of that subfamily are unresolved lichthyidae. and we assume that they represent a single Two miniatures, one each in Hoplomyzon miniaturization event. Four subfamilies Myers and Dupouyichthys Schultz, occur in contain only miniatures (Sarcoglanidinae) the Aspredinidae. According to Stewart or mostly miniatures and small species (1985:4) the species occur in separate (Glanapteryginae, Vandelliinae, Tridentin- monophyletic taxa containing other small ae) but belong to separate and divergent species. Thus miniatures are presumed to clades within the Trichomycteridae (J. N. have arisen twice in the Aspredinidae. There Baskin, pers. comm.) and thus represent is only one described miniature in the Pime- distinct miniaturization events. Miniaturi- lodidae, Horiomyzon retropinnatus Stewart zation in the Trichomycteridae thus arose (1986), and a single such species is tenta- at least five times. tively identified for the Loricariidae, Mi- Miniature species in the Callichthyidae crolepidogaster lophophanes (Eigenmann & are limited to three species ofAspidoras von Eigenmann). As noted above, the phyloge- Ihering and four species of Corydoras La- netic relationships of Scoloplax, the only cepede. Although Weitzman & Balph (1979: genus of the Scoloplacidae, are the subject 16-18) raised questions about the distinc- of some controversy, but under either pro- tiveness of these genera, they noted (1979: posed classification it probably represents 20) that reduced size in Corydoras and As- an independent miniaturization event since pidoras was apparently independently de- no miniatures are known in likely sister rived. The relationships of the four minia- groups to that genus. In summary minia- ture Corydoras species are unresolved and tures have arisen at least twelve times in the the genus is very speciose raising the pos- South American members of the Siluri- sibility that miniaturization arose several formes. times in that taxon. We prefer nonetheless The five miniature species in the Cyprin- VOLUME 101, NUMBER 2 457 Fig. 4. Fluviphylax pygmaeus (Myers & Caravalho), a miniature cyprinodontiform fish, family Poeciliidae, from the Amazon basin, USNM llllliO: (A) mature male, 10.3 mm SL, (B) mature female, 1 3. 1 mm SL; Brazil, state of Amazonas, Rio Negro, Lago Sirola, Anavilhanas Archipelago (approx. 03°10-15'N, 60°40-45'W). idontiformes belong to two subfamilies of certain relationship to other members of that the Poeciliidae according to Parenti (1981: family, and not shown to be monophyletic, 451, 463). Fluviphylax pygmaeus (Myers & they represent at least one evolutionary Carvalho) (Fig. 4) is the only described miniaturization event. member of the Fluviphylacinae and is an On the basis of present information we independent evolution to small size. The can reasonably estimate that miniaturiza- four other diminutive cyprinidontiform tion within the South American freshwater species are distributed in three tribes of the fish fauna has taken place independently Poeciliinae in the classification proposed by sixteen times in the Characiformes, on at Rosen «fe Bailey (1963:3, 39, 61, 116). The least twelve occasions in the Siluriformes, two species of Poecilia Bloch & Schneider in four instances in the Cyprinidontiformes are placed in the Poeciliini, Neoheterandria and once in each of the Clupeiformes and Henn in the Heterandriini, and Phalloto- Perciformes, for a total of about thirty-four rynus Henn in the Cnesterodontini. If these events. tribes represent monophyletic groups, then miniatures have evolved at least three times Ecology of South American Miniature in the Poeciliidae and a total of four times Freshwater Fishes in the Cyprinodontiformes. The final miniatures known in the South There are currently 85 described species American freshwater ichthyofauna, the of miniature or putatively miniature fishes eleotridids of the genus Micwphylipnus from forested regions of South America Myers (Fig. 5) have not been studied since (Table 1), a number that will undoubtedly described by Myers (1927). Although of un- increase. Fewer miniature fish species have 458 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON B Fig. 5. Microphilypnus sp., a miniature perciform fish, family Eleotrididae, from the Amazon basin, USNM 290836: (A) mature male, 20.8 mm SL, (B) mature female, 21.7 mm SL; Brazil, state of Amazonas, Rio Urubaxi, tributary of Rio Negro, south of east end of Ilha Tamaquari (01°32'N, 64°59'W). been reported from other tropical fresh- catfishes, all diminutive (Amazonian) water regions. If we use the same standard species are secondary freshwater fishes and for a miniature, approximately one dozen belong to groups whose presence can be miniature fish species are known from the thought of as 'marginal'." More recently freshwaters of Africa (Daget et al. 1984, ( 1 984:320) he has further suggested that "In 1986) and a comparable number seem to the Amazon, where community structure of have been reported from the Asiatic tropics, freshwater fishes may be more complex than totals admittedly likely to increase in the anywhere else, minute body size seems to future. Reasons for the current differences have arisen primarily in response to biotic in the numbers of miniature species from factors." He continued that, "this in my the three continental tropical regions are ob- opinion, is the general reason why so many scure. In part the larger number of minia- of the smallest Amazonian fishes belong to tures in South American freshwaters may secondary freshwater fish groups which in simply reflect the richness of the overall fish terms of relative numbers of species rep- fauna on that continent; however, our resent an insignificant fraction of the fau- knowledge of the ichthyofauna is so poor na." He also notes that miniatures in var- (Bohlke et al. 1978:667-670) that it is pre- ious other groups may be discovered in the mature to speculate further at this time. future. Secondary freshwater fishes as de- The diversity and composition of the fined by Myers (1937:344-345) are those miniature fish fauna in South America was fishes belonging to teleost groups ordinarily previously discussed by Roberts (1973a: of marine or euryhaline habit, or those 247-248), who stated that "Excepting the groups with members commonly entering . VOLUME 101, NUMBER 2 459 an euryhaline habitat. In South American freshwater fishes in that region. The reasons freshwaters the known miniatures belong to for this lack of information about minia- five major groups, three of which, the Cy- tures are myriad, many revolving around prinodontifi^rmes, Clupeiformes and Per- the diminutive size of the species in ques- ciformes (only Eleotrididae) are secondary tion. Such species are not often noted while freshwater fish groups according to Myers' making general collections. Those sampling concepts. The other two groups, the Charac- efforts are, furthermore, usually not focused iformes and Siluriformes are primary fresh- at the microhabitat level necessary to delim- water groups, that is teleost groups typically it those ecological parameters possibly idio- of freshwater habits. Of the 85 listed species syncratic to miniature species. of South American miniatures, about 88% Miller (1979:271-274) brieffy discussed belong to the primary freshwater groups, the the habitats of small fishes, but did not spe- characiforms and siluriforms (Table 1). The cifically refer to miniatures as defined herein. secondary freshwater groups, cyprinidon- Our own observations and the limited per- tiforms, clupeiforms, and perciforms in tinent literature reveal one general ecolog- contrast, together represent only ten of the ical parameter common to many minia- 85 miniature freshwater fishes known from tures—their repeated occurrence in still or South America (Table 1). Increased col- slow-flowing shallow waters. This associa- lecting with specialized techniques, and fu- tion is not unexpected given the difficulties ture revisionary studies may change the rel- small or miniature fishes have in maintain- ative contribution of these major taxonomic ing position in a rapidly moving water col- units to the fauna of miniature fishes. Cur- umn. Miniature Neotropical fishes living in rent data indicate, however, that contrary less turbulent habitats are detailed below. to Roberts, it is the primary rather than This listing is based on personal observa- secondary freshwater groups that predom- tions supplemented by personal commu- inate among miniature fishes ofthe Amazon nications and literature information. basin, the area of occurrence of the majority Among the miniature clupeiforms, only of characiform and siluriform fishes listed Amazonsprattus scintilla Roberts is known in Table 1 to live in still or slow-flowing waters. The The degree to which ecological factors (? more speciose characiform miniatures we the biotic factors of Roberts (1984:320)) have seen in, or which have been reported show a correlation with, and may have been from that habitat include the characidiines involved in the evolution of miniature fish- Elachocharax mitopterus Weitzman & es in South American freshwaters is difficult Elachocharax pulcher Myers, the characi- to determine at this time. The possibility nines Priocharax ariel Weitzman & Vari and that acidic waters of low productivity may Priocharax pygmaeus Weitzman & Vari, the result in reduced poring of the laterosensory glandulocaudines Xenurobrycon polyancis- system of the body, and perhaps of overall trus Weitzman (Weitzman 1987:1 18-1 19), body size was noted by Collette ( 1 962: 1 79- species of Tyttocharax Fowler (Saul 1975: 191) relative to the North American percid 111; D. J. Stewart, pers. comm.), and lo- fish Etheostoma fusiforme Girard. Black tabrycon praecox Roberts (Roberts 1973b: acidic waters are quite common in South 501-502), the tetragonopterine Hyphesso- America, and many of the species listed in brycon elachys M. Weitzman (1985:799), Table 1 occur in that environment (pers. and the lebiasinids Nannostomus anduzei observ.). In-depth studies of the ecology of Fernandez & Weitzman and A^. marginatus most freshwater fishes is South America are Eigenmann. The siluriform species Cory- rare and very little is known about the habits doras hastatus Eigenmann & Eigenmann or life histories of small and miniature (Myers 1953:269; I. Sazima, pers. comm.) 460 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON and Scoloplax dicra Bailey & Baskin (R. M. live in streams with strong currents, they C. Castro, pers. comm.) also are found in are actually benthic, occupying an area of that habitat. The cyprinodont Fluviphylax reduced current flow. The same is appar- pygmaeus (Myers & Carvalho) has been ob- ently true for the one pimelodid catfish served at or near the water surface by us species in Table 1 (Stewart 1986:48). and others in still waters, attached and de- Although we note an association of min- tached floating meadows, and slow-flowing iature fishes with still or slow-flowing waters, river margins both with and without plant this does not imply that such environmental cover (Weitzman 1982:193-195). An un- parameters have been the prime selective identified species of Microphylipnus of the agents for miniaturization, a process that Eleotrididae has been observed in quiet undoubtedly is the result of many different waters of small and large rivers, usually on factors. Further studies are necessary to de- or near the bottom, both over sand and mud, termine the generality of the correlation be- in areas with and without plant cover (C. J. tween still waters and miniatures, and to Ferraris, Jr., pers. comm.). It is likely that study the degree to which these relate to the remaining miniatures in these groups physical environmental parameters. will also share comparable habitats. The catfish family Trichomycteridae has Acknowledgments many members parasitic on the gills of large fishes, and perhaps as a consequence of that Research associated with this project was unusual habit does not neatly fit the above supported in part by the I. E. S. P. Neo- characterization of the habitats of minia- tropical Lowland Research Project of the tures. During feeding the miniature para- Smithsonian Institution. Assistance on and/ sitic trichomyterids (e.g., Paravandellia ber- or information for this study was provided /o«// Eigenmann (A. Miranda-Ribeiro 1923: by Gareth Nelson and Carl J. Ferraris, Jr. 1 1-12) and P. magdalenae Miles (1947:99) (American Museum of Natural History), attach to the gills of larger fishes, a feeding Peter J. P. Whitehead (British Museum habit that also serves to insulate them from (Natural History)), John G. Lundberg (Duke the currents of the surrounding stream. University), Karsten Hartel (Museum of Trichomycterus santaeritae Eigenmann Comparative Zoology), Ivan Sazima (Uni- (1918b:341; see also Eigenmann 191 1:350, versidade Estadual de Campinas), Donald locality number 56) of the Trichomycteri- J. Stewart (University of Wisconsin), Jon- nae and Malacoglanis gelatinosus Myers & athan N. Baskin (California State Polytech- Weitzman (1966:286) of the Sarcoglanidi- nic University), Bruce B. Collette (National nae both live in relatively still waters, and Marine Fisheries Service), Marilyn J. their morphology indicates that they are not Weitzman, Lynn Norrod, Ann W. Vari, and parasitic. As noted by Eigenmann (1918b: Andrew G. Gerberich (National Museum of 269) most non-parasitic members of the Natural History, Smithsonian Institution), Trichomycterinae (=Pygidiinae of Eigen- Ricardo M. C. Castro (Universidade de Sao mann), in contrast, live in moderately to Paulo, Ribeirao Preto), Mario C.C. de Pin- swiftly flowing streams, but avoid the water na (Universidade Federal do Rio de Janei- currents by being cryptobenthic. Thus on a ro) and especially Naercio A. Menezes (Mu- microhabitat level the two species of Trich- seu de Zoologia da Universidade de Sao omycterus listed in Table 1 can be consid- Paulo). The Portuguese translation of the ered to inhabit water conditions typical of "Resumo" was generously provided by Ri- other miniatures. cardo M. C. Castro. This paper was im- Stewart (1985) noted that although the proved by the comments and suggestions of miniature catfish species of Aspredinidae Ricardo M. C. Castro, Bruce Collette, Carl VOLUME 101, NUMBER 2 461 J. Ferraris, Jr., Naercio A. Menezes, Paula freshwater fishes of Africa. Volume 2. Institut Mabee, Lynne R. Parenti, Scott A. Schaefer, Royal des Sciences Naturelles de Belgique, Brus- sels, 520 pp. Ann W. Vari, Anna Weitzman, Marilyn Durbin, M. L. 1909. II. Reports on the expedition Peter P. Whitehead, Weitzman, J. and two to British Guiana of the Indiana University and anonymous reviewers. the Carnegie Museum, 1908.— Annals of the Carnegie Museum 6(l):55-72. Literature Cited Eigenmann, C. H. 1909. Some new genera and species of fishes from British Guiana.— Annals of the Almeida, V. A. 1971. Descrifao de uma nova especie Carnegie Museum 6:4—54. do genero Characidium (Pisces, Characidae).— . 1911. The localities at which Mr. John D. Papeis Avulsos de Zoologia, Sao Paulo 25(14): Haseman made collections. Pp. 287-314 in J. 111-119. D. Haseman & C. H. Eigenmann, A brief report Anonymous. 1970. Annotated chronological bibli- upon the expeditions of the Carnegie Museum ography of the publications of George S. Myers to central South America together with a list of (up to the end of 1969). — Proceedings of the localities at which Mr. Haseman collected. An- California Academy of Sciences, fourth series, nals of the Carnegie Museum, volume 7, parts 38(2): 19-52. 3 and 4. Bailey, R. M., & J. N. Baskin. 1976. Scoloplax dicra, . 1914. Some results from studies of South a new armored catfish from the Bolivian Am- American fishes. — Indiana University Studies azon. —Occasional Papers ofthe Museum ofZo- 12(l):15-48. ology, University of Michigan 674:1-14. . 1915. The Cheirodontinae, a subfamily of Bohlke, J. E. 1958. Results of the Catherwood Foun- minute characid fishes of South America.— dation Peruvian Amazonian expedition: The Memoirs of the Carnegie Museum 7(1): 1-99. description of two new xenurobryconine char- . 1918a. Descriptions of sixteen new species acids.-Copeia 1958(4):3 18-325. of Pygidiidae. — Proceedings of the American , & C. R. Robins. 1968. Western Atlantic sev- Philosophical Society 56:690-703. en-spined gobies, with descriptions of ten new . 1918b. The Pygidiidae, a family of South species and a new genus, and comments of Pa- American catfishes.— Memoirs of the Carnegie cific relatives. — Proceedings of the Academy of Museum 7(5):259-398. Natural Sciences of Philadelphia 120(3):45-174. , & R. S. Eigenmann. 1888. Preliminary notes , S. H. Weitzman, & N. A. Menezes. 1978. on South American Nematognathi I. — Proceed- Estado atual da sistematica dos peixes de agua ings of the California Academy of Sciences, se- doce do America do Sul.— Acta Amazonica 8(4): ries 2 119-172. 657-677. 1(2): . , & 1889. Preliminary notes on South Campagno, L. J. V. 1981. Family Rhincodontidae. American Nematognathi II. — Proceedings of the (Unpaginated) in Fischer, W., G. Bianchi & W. California Academy of Sciences, (2)2:28-56. B. Scott, eds., FAO species identification sheets Fernandez, J. M., S. H. Weitzman. 1987. new of fishery purposes. Eastern Central Atlantic. & A species of Nannostomus (Teleostei: Lebiasini- Fishing areas 34, 47 (in part). Food and Agri- dae) from near Puerto Ayacucho, Rio Orinoco culture Organization, Rome. drainage, Venezuela. — Proceedings of the Bio- Cervigon, F. 1982. La ictiofauna estuarina del Caiio logical Society of Washington 100(1):165-172. Manamo y area adyacentes. Pp. 205-260 in D. Novoa R., ed., Los recursos pesqueros del rio Fink, W. L.. & S. H. Weitzman. 1974. The so-called Orinoco y su explotacion. Corporacion Vene- cheirodontin fishes of Central America with de- zolana de Guayana, Caracas. scriptions of two new species (Pisces: Charac- Collette, B. B. 1962. The swamp darters of the sub- idae).— Smithsonian Contributions to Zoology genus Hololepis (Pisces, Percidae). — Tulane 172:1-46. Studies in Zoology 9(4): 1 15-21 1. Fowler, H. W. 1913. Fishes of the Madeira River, . 1966. Belonion, a new genus of fresh-water Brazil. — Proceedings of the Philadelphia Acad- needlefishes from South America.— American emy of Sciences 65:517-579. Museum Novitates 2274:1-22. . 1 932. Zoological results of the Matto Grosso Daget, J., J. -P. Gosse, & D. F. E. Thys van den Au- expedition to Brazil in 1931. — 1. Fresh water denaerde, eds. 1984. Checklist of the freshwater fishes. — Proceedings of the Academy of Natural fishes of Africa. Volume 1 . Paris Office de la Sciences of Philadelphia 84:343-377. Recherche Scientifique et Technique Outre-Mer, Garman, S. 1895. The cyprinodonts. — Memoirs of 410 pp. the Museum of Comparative Zoology at Har- ,, & . 1986. Checklist of the vard College 19(1): 1-1 79. 462 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Geisler, R., & S. R. Annibal. 1984. Okologie des Hypessobrycon amandae, sp. n. (Pisces, Char- Cardinal-Tetra Paracheirodon axelrodi (Pisces, acoidei).— Tropical Fish Hobbyist 35(5):58-61, Characoidea) im Stromgebiet des Rio Negro/ 67. Brasilien sowie zuchtrelevante Faktoren.— Greenwood, P. H., D. E. Rosen, S. H. Weitzman, & Amazoniana 9(1): 5 3-8 6. G. S. Myers. 1966. Phyletic studies of teleos- Gery, J. 1960. Jobertina eleotrioides n. sp. (Chara- tean fishes with a provisional classifications of cidiinae) from French Guiana with consider- living forms.— Bulletin of the American Mu- ations about the genus and redescription of the seum of Natural History 131:339-455. type species.— Opuscula Zoologica 47:1-10. Henn, A.W. 1916. On various South American poe- . 1961a. Three new South-American chara- ciliid fishes. — Annals of the Carnegie Museum cids, Knodus savannensis, Hyphessobrycon her- 10(l-2):93-142. bertaxelrodi and Megalamphodus sweglesi, with Hoedeman, J. J. 1957. Hypessobrycon griemi, eine a review of some Hyphessobrycon-groups of Farbenprachtige Neuheit der Bifasciatus-Gruppe species.— Tropical Fish Hobbyist 9(9):26-39, ans Gojas. — Die Aquarien-und Terrarien-Zeit- 42-46. schrift (DATZ) 10(4): 8 7-8 9. . 1961b. Notes of the ichthyology of Surinam Ihering, R. von. 1930. Notes ecologicas referentes a and other Guianas, 7, Hyphessobrycon georgetti peixes d'agua doce do Estado de Sao Paulo e sp. nov. a dwarf species from southern Suri- descip9ao de 4 especies novas.— Archivos do nam.— Bulletin of Aquatic Biology 2(22): 121- Instituto Biologico 3:93-104. 128. Isbriicker, I. J. H. 1980. Classification and catalogue . 1963. Hyphessobrycon simulans, anew neon of the mailed Loricariidae (Pisces, Siluri- tetra.- Tropical Fish Hobbyist 1 1(8): 13-60, 70- formes).— Verslangen en Technische Gegevens, 72. Institut voor Taxonomische Zoologie 22: 1-181. . 1964. Poissons characoides de I'Amazonie Jacobs, K. 1969. Livebearing Aquarium Fishes. Lon- peruvienne.— Beitrage zur Neotropischen Fau- don, Studio Vista Limited, 460 pp. + chart. na 5(1): 1-44. Jewett, S. L., & E. A. Lachner. 1983. Seven new . 1965a. Characidae et Crenuchidae del ITga- species of the Indo-Pacific genus Eviota (Pisces: rape Preto (Haute Amazonie). — Senckenber- Gobiidae).— Proceedings of the Biological So- giana Biologica 46(1): 1 1-45. ciety of Washington 96(4):780-806. . 1965b. Characidae et Crenuchidae de ITga- Knaack, J. 1966. Ein Zwergpanzerwels— Corv<3?ora5 rape Preto (Haute Amazonie). — Senckenber- pygmaeus.—T>'\e Aquarien-und Terrarien-Zeit- giana Biologica 46(3): 195-2 18. schrift 19:168-169. . 1966. Axelrodia riesei, a new characoid fish Lachner, E. A., & S. J. Kamella. 1978. Fishes of the from upper Rio Meta in Colombia (with re- genus Eviota of the Red Sea with descriptions marks concerning the genus Axelrodia and de- of three new species (Teleostei: Gobiidae).— scription of a similar, sympatric Hyphessobry- Smithsonian Contributions to Zoology 286:1- co« -species). — Ichthyologica/ the Aquarium 23. Journal 39(3): 11 1-1 20. , & . 1980. Fishes of the Indo-Pacific . 1971. Une sous-famille nouvelle de poissons genus Eviota with descriptions of eight new characoides sud-americains: Les Geisleriinae. — species (Teleostei: Gobiidae). — Smithsonian Vie et Milieu 12(1):153-166. Contributions to Zoology 315:1-127. . 1973. New and little-known Aphyoditeina Ladiges, W. 1938. Drie neue fische der Gattungen (Pisces, Characoidei) from the Amazon basin. — Hypessobrycon und Hemigrammus an dem pe- Studies on the Neotropical Fauna 8:81-137. ruanischen Tiel des Amazons.— Zoologischer . 1977. Characoids of the World. Neptune City, Anzeiger 124:49-52. New Jersey, TFH Publications, 672 pp. . 1950. Microbrycon Cochui spec. nov. eine . 1984. Le Cladisme commente. — Revue neue Art der siidamerikanischen Glandulocau- Franfaise d'aquariologie [et] herpetologie 1 1(2): dinae.— Zoologischer Anzeiger 145(1 1-12):305- 57-60. 309. , & V. Mahnert. 1986. A new rummy-nosed LaMonte, F. 1939. Tridentopsis tocantinsi, a new py- tetra from the Rio Negro, Brazil: Hemigrammus gidiid fish from Brazil.— American Museum bleheri n. sp. (Characidae, Tetragonopterinae) Novitates 1024:1-2. with comments on Paracheirodon.— TroTpical Miles, C. 1943a. Feces de agua dulce del valle del Fish Hobbyist 34(1 1):37, 40-41, 44-45, 48-49, Cauca. Call, Colombia, Publicaciones de la Se- 52. cretaria del Valle del Cauca, 2 1 4 pp. , &A. Uj. 1986. The ember tetra: a new pygmy . 1943b. On three recently described species characid tetra from the Rio das Mortes, Brazil, and a new genus of pygidiid fishes from Colom- — VOLUME 101, NUMBER 2 463 bia. — Revista de la Academia Colombiana de of cyprinodont fishes.— The Tropical Fish Mag- Ciencias Exactas 5(19):367-369. azine 4(4):7. . 1947. Los peces del Rio Magdalena. Bogota, . 1958. Trends in the evolution of teleostean Colombia, Ministerio de la Economia Nacional, fishes.— Stanford Ichthyological Bulletin 7(3):27- Seccion de Piscicultura, Pesca y Caza, 214 pp. 30. Miller, P. J. 1979. Adaptiveness and implications of , & P. Miranda-Ribeiro. 1945. A remarkable small size in teleosts. Pp. 263-306 in P. J. Miller, new genus of sexually dimorphic characid fishes ed., Fish phenology: Anabolic adaptiveness in from the Rio Paraguay basin in Matto Grosso. — teleosts. Symposium of the Zoological Society Boletim do Museu Nacional do Rio de Janiero, of London 44. Nova Serie Zoologia 32:1-8. Miranda-Ribeiro, A. de. 1912. Loricariidae, Callich- , & S. H. Weitzman. 1966. Two remarkable thyidae, Doradidae, Trichomyceteridae.— new trichomycterid catfishes from the Amazon Commissao de Linhas Telegraphicas Estrate- basin in Brazil and Colombia.— Journal of Zo- gicas de Matto-Grosso ao Amazonas. Historia ology, London 149:277-287. Natural, Zoologia Annexo 5:27-29. Nijssen, H., & I. J. H. Isbriicker. 1976a. A new cal- . 1923. Critical notes of Brazilian Zoology. lichthyid catfish, Corydoras gracilis, from Bra- Archivos da Escola Superior da Agricultura e zil.-Tropical Fish Hobbyist 25(1):90-91, 94- Medicina Veterinaria, Nictheroy 6(for 1922) (1- 96. . 2):11-15. , & 1 976b. The South American plated Miranda-Ribeiro, P. de. 1947. Notas para o estudo catfish genus Aspidoras R. von Ihering, 1907, dos Pygidiidae Brasileiros (Pisces — Pygidi- with descriptions on nine new species from Bra- idae— Vandelliinae). — Boletim do Museu Na- zil (Pisces, Siluriformes, Callichthyidae).— tional, nova serie, Zoologia 78:1-9. Bijdragen de Dierkunde 46(1):1 07-131. Myers, G. S. 1924. A new characin fish from Rio de , & . 1986. Review of the genus Cory- Janiero.— The Fish Culturist, Philadelphia 4(3): doras from Peru and Ecuador (Pisces, Siluri- 330-331. formes, Callichthyidae). — Studies on Neotrop- . 1925a. Description of a new cheirodontine ical Fauna and Environment 21(l-2):l-68. characin from Rio de Janiero.— Annals of the Parenti, L. R. 1981. A phylogenetic and biogeograph- Carnegie Museum 16:143-144. ic analysis of cyprinodontiform fishes (Teleos- . 1925b. Tridentopsispearsoni, anew pygidiid tei, Atherinomorpha). — Bulletin of the Ameri- catfish from Bolivia. — Copeia (old series), can Museum of Natural History 168(4):335-557. 1925(148):83-86. Pearson, N. E. 1924. The fishes of the eastern slope . 1927. Description of new South American of the Andes. 1 . The fishes of the Rio Beni basin, fresh-water fishes collected by Dr. Carl Ter- Bolivia, collected by the Mulford Expedition.— netz. — Bulletin of the Museum of Comparative Indiana University Studies 1 1(64): 1-83. Zoology (Harvard College) 68(3): 107-1 35. Regan, C. T. 1907. Descriptions of two new chara- . 1936. A new characid fish of the genus T/^'- cinid fishes from Argentina.— Annals and Mag- phessobrycon from the Peruvian Amazon.— azine of Natural History, ser. 7, 19:261-262. Proceedings of the Biological Society of Wash- Roberts, T. R. 1970. Description, osteology and re- ington 49:97-98. lationships of the amazonian cyprinodont fish . 1937. Fresh-water fishes and West Indian Fluviphylax pygmaeus (Myers and Carvalho).— zoogeography.— Smithsonian Report for 1937: Breviora 347:1-28. 339-364. . 1973a. Ecology of fishes in the Amazon and . 1944. Two extraordinary new blind nema- Congo basins. Pp. 239-254 in B. J. Meggers, E. tognath fishes from the Rio Negro, representing S. Ayensu, & W. D. Duckworth, eds.. Tropical a new subfamily of Pygidiidae, with a rearrange- forest ecosystems in Africa and South America: ment of the genera of the family, and illustra- A comparative view. Smithsonian Institution tions of some previously described genera and Press, Washington. species from Venezuela and Brazil. — Proceed- . 1973b. The glandulocaudine characid fishes ings of the California Academy of Sciences. of the Guayas basin in western Ecuador. — Bul- Fourth series 23(40):59 1-602. letin of the Museum of Comparative Zoology . 1947. The Amazon and its fishes. Part 2. The (Harvard University) 144(8):488-514. fishes.— The Aquarium Journal 18(4): 13-20. . 1984. Amazonsprattus scintilla, new genus . 1953. Habits and classification of CorV(^ora5 and species from the Rio Negro, Brazil, the hastatus.—The Aquarium Journal 24(I):268- smallest known clupeomorph fish. — Proceed- 270. ings of the California Academy of Sciences . 1955. Notes on the classification and names 43(20):3 17-321. 464 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON . 1986. Danionella translucida, a new genus asinidae and Erythrinidae with special reference and species of cyprinid fish from Burma, one of to subtribe Nannostomina. — Proceedings of the the smallest living vertebrates. — Environmental United States National Museum 1 16(3499): 127- Biology of Fishes 16(4):23 1-241. 169. Rosen, D. E., & R. M. Bailey. 1963. The poecihid -. 1977. A new species of characoid fish, Hy- fishes (Cyprinodontiformes), their structure, phessobrycon diancistrus, from the Rio Vichada, zoogeography and systematics.— Bulletin of the Orinoco river drainage, Colombia, South Amer- American Museum of Natural History 126(1): ica (Teleostei: Characidae).— Proceedings of the 1-176. Biological Society of Washington 90(2):348-357. Saul, W. G. 1975. An ecological study of fishes in -. 1982. Fluviphylaxpygniaeus,3im.miaivire\a\.- upper Amazonian Ecuador.— Proceedings of the lifish from tropical South American freshwa- Academy of Natural Sciences of Philadelphia ters.— Journal of the American Killifish Asso- 127(12):93-134. ciation 15(5): 190-200. Schultz, L. P. 1944. The catfishes of Venezuela, with -. 1986. A new species of Elachocharax (Te- descriptions of thirty-eight new forms. — Pro- leostei: Characidae) from the Rio Negro region ceedings of the United States National Museum of Venezuela and Brazil.— Proceedings of the 94(3 172): 173-338. Biological Society of Washington 99(4):737-745. Springer, V. G. 1983. Tyson bellos, new genus and -. 1987. A new species of Xe«MroZ?ryco7t (Te- species of western Pacific fish (Gobiidae, Xe- leostei: Characidae) from the Rio Mamore of nisthminae), with discussions of gobioid os- Bolivia. — Proceedings of the Biological Society teology and classification.— Smithsonian Con- of Washington 1 00( 1 ): 1 1 2- 1 20. tributions to Zoology 390:1-40. -, & M. H. Balph. 1979. Some phylogenetic Stewart, D. J. 1985. A review of the South American implications of a discovery of Aspidoras pau- catfish tribe Hoplomyzontini (Pisces, Aspredin- ciradiatus (Pisces: Siluriformes: Callichthyidae) idae), with descriptions of a new species from from the Rio Negro in Brazil.— Proceedings of Ecuador.— Fieldiana, Zoology (25):1-19. the Biological Society of Washington 92(1): 10- . 1986. A new pimelodid catfish from the deep- 22. river channel of the Rio Napo, Eastern Ecuador -, & J. S. Cobb. 1975. A revision of the South (Pisces: Pimelodidae). — Proceedings of the American fishes of the genus Nannostomus Academy of Natural Sciences of Philadelphia Giinther (Family Lebiasinidae).— Smithsonian 138(l):46-52. Contributions to Zoology 186:1-36. Vari, R. P. 1983. Phylogenetic relationships of the -, & S. V. Fink. 1985. Xenurobryconin phy- families Curimatidae, Prochilodontidae, Ano- logeny and putative pheromone pumps in glan- stomidae and Chilodontidae. — Smithsonian dulocaudine fishes (Teleostei: Characidae).— Contributions to Zoology 378:1-60. Smithsonian Contributions to Zoology 421:1- . (In press.) The Neotropical characiform fam- 121. ily Curimatidae (Pisces, Ostariophysi), a phy- -, & W. L. Fink. 1983. Relationships of the logenetic study.— Smithsonian Contributions to neon tetras, a group of South American fresh- Zoology. water fishes (Teleostei, Characidae), with com- Weitzman, M. 1985. Hyphessobrycon elachys, a new ments on the phylogeny of New World Characi- miniature characid from eastern Paraguay forms.— Bulletin of the Museum of Compara- (Pisces: Characiformes).— Proceedings of the tive Zoology 150(6):339-395. Biological Society of Washington 98(4):799-808. -, & . 1987. Neon tetra relationships and Weitzman, S. H. 1954. The osteology and relation- phylogenetic systematics.— Tropical Fish Hob- ships of South American characid fishes of the byist 36(2):72-77. subfamily Gasteropelecinae.— Stanford Ichthy- -, & J. Gery. 1981. The relationships of the ological Bulletin 4(4):2 13-263. South American pygmy characoid fishes of the . 1960a. Figures and descriptions offour South genus Elachocharax, with a redescription of American catfishes of the genus Corydoras, in- Elachocharax junki (Teleostei: Characidae).— cluding two new species.— Stanford Ichthyolog- Proceedings of the Biological Society of Wash- ical Bulletin 7(4): 140-1 54. ington, 93(4):887-9 13. . 1960b. Further notes on the relationships and -, & R. H. Kanazawa. 1977. A new species of classification of the South American characid pygmy characoid fish from the Rio Negro and subfamily Gasteropelecinae. — Stanford Ichthy- Rio Amazonas, South America (Teleostei: ological Bulletin 7(4):239-244. Characidae).—Proceedings of the Biological So- . 1964. Osteology and relationships of South ciety of Washington 90(1): 149- 160. American characid fishes of subfamilies Lebi- -, & . 1978. The South American fish VOLUME 101, NUMBER 2 465 genus Elachocharax Myers with a description general features, visceral anatomy, and osteol- of new species (Teleostei: Characidae). — Pro- ogy.— American Museum Novitates 2835: 1-44. ceedings of the Biological Society of Washington Winterbottom, R. 1983. A review of the gobiid fish 91(1):158-183. genus Trimma from the Chagos Archipelago, , & H. Nijssen. 1970. Four new species and central Indian Ocean, with the description of one new subspecies of the catfish genus Cory- seven new species.— Canadian Journal of Zo- doras from Ecuador, Colombia and Brazil ology 62:695-715. (Pisces, Siluriformes, Callichthyidae).— Beau- , & A. R. Emery. 1981. A new genus and two fortia 18(223): 119-132. new species of gobiid fishes (Perciformes) from , & R. P. Vari. 1987. Two new species and a the Chagos Archipelago, Central Indian Ocean.— new genus of miniature characid fishes (Teleos- Environmental Biology of Fishes 6(2): 139-149. . fish- tei: Characiformes) from northern South Amer- , & 1986. Revision of the gobioid ica.— Proceedings of the Biological Society of es of the Chagos Archipelago, Central Indian Washington 100(3):640-652. Ocean. — Royal Ontario Museum Life Sciences , & M. Weitzman. 1982. Biogeography and Contributions 142:1-82. evolutionary descriptions in Neotropical fresh- water fishes, with comments on the refuge the- Department of Vertebrate Zoology (Fish- ory. Pp. 403-422 in G. T. Prance, ed.. Biological es), National Museum of Natural History, diversification in the tropics, Columbia Uni- versity Press, New York. Smithsonian Institution, Washington, D.C. Whitehead,P.J.P.,&G.G.Teugels. 1985. The West 20560. African pygmy herring Sierrathrissa leonensis: