Papiliochromis (RAMIREZI)

Kullander, Sven O., 1980, A redescription of the South American cichlid fish Papiliochromis ramirezei (Myers & Harry, 1948) (Teleostei: Cichlidae), Studies on Neotropical Fauna und Environment 15, pp. 91-108 : 93-106

PAPILIOCHROMIS RAMIREZI

Apistogramma ramirezi Myers & Harry in Anon., 1948: 77 (informal description; no types designated; type-locality unknown, “evidently from one of the tributaries of the Rio Apuré or Rio Meta in the states of Guarico, Portuguesa, or Apuré ”) - Myers & Harry, 1948: 1 (detailed description of “holotype” and three “paratypes ”; type-locality unknown, “evidently one of the tributaries of the lower Rio Apure [or possibly one of the lower Rio Meta] in the state of Guarico or Apure ”) -— Wickler, 1956: 310, Abb. 11 (egg morphology) - Fischer, 1968: 9 (Rio Orituco; Canon Caballo; Rio Guarico; no preserved specimens) - Luengo, 1970: 32 (in list; Palenque, Edo. Guárico [= misinterpretation of type-locality]; LOS Llanos [cop. Mago]) — Mago, 1971: 102 (in list; vernacular name)

Microgeéphagus ramirezi Frey, 1959: 391 (possible new genus, possibly to contain Apistogramma ramirezi ; generic name not available)

“ Apistogramma ” ramirezi Wickler, 1960: 327 , l fig. (systematic position discussed; compared with other Apistogramma spp., and Geophagus )

Geophagus ramirezi Klee, 1971: 47 (systematic position discussed; compared with other Apistogramma spp., and Geophagus ; new comb.)

Microgeophagus ramirezi Scheel, 1972: 5 (note on karyotype; generic name not explained)

Papiliochromis ramirezi Kullander , 1 977: 253 (type-species of new monotypic genus; new comb.)

Etymology. — ramirezi from a pseudoscientifíc commerical name, “ Ramirezi ” or “ Ramirezi Cichlid ”, coined by Blass (in Anon., 1947). This in its turn was given for Mr Manuel Vicente Ramirez, a Venezuelan fish breeder who travelled with Blass on the journey during which P. ramirezi was first discovered.

Syntypes. — Stanford University 14845 (2: 37.3, 30.9 mm SL), 14846 (31.0 mm SL), 14847 (27.6 mm SL) (Myers & Harry, 1948: Tab. 2; on p. 2 a 26.6 mm “paratype” mentioned)

Despite that no types were designated in Anon. (1948), and Myers & Harry (1948) consider that paper as the original description, they still speak of a “holotype” and three “paratypes” in the subsequent publication. The only material reference made in Anon. (1948) reads: “the fish [species] is described principally from an ‘ adult male collected I am uncertain about the meaning of “principally ”, a word not used by Myers & Harry (1948), but the sentence seems to suggest that Myers & Harry had more than one specimen, and that one of them was an adult male. These would then be syntypes, although it is in no way apparent in the protolog which these are or how they should be recognized. Myers & Harry (1948) seem to have recognized the syntypical series, since they indicate that both their descriptions are based on at least in part the same material. The specimens described by Myers & Harry (1948) should be regarded as syntypes, provided that they were all used for the diagnosis in Anon. (1948).

Type-locality. - The typical specimens were collected by Manuel V. Ramirez and Herman Blass somewhere in the Llanos of the Orinoco basin, south of Palenque, Venezuela, in April 1947 (Myers & Harry, 1948). Blass could not tell where they caught ramirezi , but they did so only at one locality along a route of 500 miles, as indicated by the wording in his description of the collecting trip (in Anon., 1947). After having crossed two shallow rivers, the party turned back at “ a very large river ”, which may have been the R. Meta (Myers & Harry, 1948). It seems strange, though, that Blass should have, as he says (in Anon., 1947), “forgotten” such a simple name. The origin of the types, it seems, will remain unknown, and no type-locality can be stated, only a “type-area ”, between Palenque and the R. Meta in Venezuela.

The below described material comes from the Colombian Llanos, and thus is not from the type-area and is unsuitable for a restriction, were at present anything gained from such an action. The occurrence of P. ramirezi in the Venezuelan Llanos has been confirmed by later collections (D. C. Hicks, pers. comm.).

DESCRIPTION OF COLOMBIAN MATERIAL

NRM 11250 - 2 ♂ ♂, 29.0, 29.6 mm; 7JA 1 ijij, WR. O-30.3 mm. Colombia (Meta), R. Yucao right bank caño ZRJ6D20 ' N 72°08 ' W). 1972.03.05. Leg. T. Hongslo (Sta. VIT 55).

NRM 11251 - 7 QQ, 24.6 - 29.3 mm. Colombia (Vichada), R. Guarrojo, Finca Boca de Guarrojo, laguna (4°06 'N 70°49 ' W). 1972.03.08. Leg. T. Hongslo (Sta. VIT 56).

NRM 11252 - 2 ♂♂, 32.2, 33.9 mm; 19 ♀♀, 25.1 - 33.9 mm. Colombia (Vichada), R. Guarrojo, Finca Boca de Guarrojo, cafio (4°06 ' N 70°49 'W). 1972.03.09. Leg. T. Hongslo (Sta. VIT 58).

The material is in fine condition. I find it difficult to determine sex by the shape of the genital papilla. A difference is apparent only with the two 11250 males, which also. have much produced ventral fins and anterior dorsal fin lappets. The two 11252 males show less marked secondary sex characteristics.

At my disposal has also been several aquarium raised P. ramirezi and notes made on two living import specimens (origin uncertain, now deposited in the British Museum [Natural History], London) in 1973. I have not used this material since experiences with aquarium material of above all species of Apistogramma Regan and Cichlasoma Swainson indicate that such is usually not comparable with wild material, whether aquarium raised or aquarium reared imports (cf. Kullander, 1974; Barlow & Munsey, 1976).

The VIT 55 locality is said by Hongslo to be a caño emptying in the R. Yucao, an affluent of the R. Meta. The collecting site is about 500 m westwards along the road Puerto Lopez - Puerto Gaitán from a Iaguna located midway between the R. Yucao and the R. Manacacías, some 300 m from the road. The two other stations are close to the mouth of the R. Guarrojo into the R. Vichada.

Description. - Little variation in body shape. Elongately ovoid, deeper anteriorly, moderately compressed. Deepest just in advance of ventral fins. Dorsal contour continuous with predorsal contour, straight horizontal or ascending to below about one-third of spinous dorsal contour, which longer than the ventral, straight, slightly sloping to caudal fin base or horizontal. Chest contour continuous with preventral head contour, little curved or straight, sloping; abdominal contour concave or straight, about horizontal. Anal fin base contour about straight, slanting, discontinuous with ventral caudal peduncle contour, which about straight or slightly concave, of steeper slope than the dorsal caudal peduncle, less than the anal base contour.

Head rather deep, outline of a broad, more 0 r less pointed triangle. Moderately compressed. Predorsal contour little curved or straight to anterior margin of orbit where is an interorbital groove producing a notch in the profile. Anterior to this notch straight descending or rostrad increasingly descending. No nuchal hump. The interorbital notch may be very inconspicuous. Preventral head contour little arched, sloping, less steep than predorsal contour.

Snout rounded from above and in profile. Appearing produced in specimens with pronounced frontal groove. Contours straight, about equal or dorsal more strongly decurved at tip. The dorsal part dominates due to the subinferior mouth. Mouth subinferior (subterminal), small, narrow. Jaws equal. Maxillary at about 65 - 70° angle to horizontal, covered by preorbital or aboral tip slightly exposed, to nearer orbit than nostril. Lips moderate in size, of about same width or lower somewhat wider. Lower thickening laterally, fold continuous; upper laterally covered by preorbital and lower lip. Orbit circular, subtangential, lower edge at horizontal level of lower edge of lower lip or below; more than one half of horizontal diameter in anterior part of head length. Interorbital area slightly convex except for groove at upper edge of frontal reception of premaxillary processes which reach anterior margin of orbit.

Preoperculum even-edged or serrated along entire 0r part of posterior edge of vertical limb. Serrations 0-40; absent inl 4 of 37 specimens, more than 10 in 14, more than 20 in 3. No apparent correlation with sex, size or locality in frequency. Gill-cover edge rounded, gill-membrane laterally not reaching pectoral fin base; no pointed flap dorsally. 5 branchiostegal rays.

Flanc scales ctenoid, of moderate size, smaller towards body contours. Lateral line scales equal in size to adjacent scales. Of two types above anal fin origin: A) with very large nucleus and few peripheral circuli, low number of cteni and short radii, B) (examined) with central or nearly centric nucleus, 36- 58 circuli, 38 - 55 cteni and 8 - 13 long (to nucleus) radii. Predorsal scales ctenoid or anteriormost cycloid. Usually posterior median in single series, diverging anteriorly. Squ. prd. 8 - 12, rarely as many as (12- 12 in one) or more than (12- 11 in one) squ. prv. Lateral chest scales ctenoid; distal pectoral axilla naked, proximally cycloid scales. Median chest scales in a middle series, ctenoid, interventral and one adjacent cycloid. Throat scales irregular or in a middle series, small, cycloid. Squ. prv. 11 - 14; squ. P-V 3 (14) or 4 (23). Naked just behind ventral fin bases, otherwise abdominal scales ctenoid, in a middle series or more or less cross-wise along middle. Cheek covered by scales. Scales behind orbit and dorsally on operculum may have few cteni, otherwise cheek and gill-cover scales all cycloid; preoperculum naked. Subopercular scales in one (usually) or two series. Scales around caudal peduncle 16 in all. Squ. tr. 8 + 1 + 2 (36) or 8 + 1 + 2 ' /2 (1). Scales between dorsal fin and last upper lateral line scale 1 (31) or 11 /2 (6). Lateral lines usually non-overlapping, the upper to last dorsal fin spine or at most 5 th ray, the lower to 2 nd or 3 rd anal fin ray. Upper line with 0- 4 terminal pores, 22 specimens without pores; subserial pores very rare (2 on one side in 2 specimens). Lower line without pores in 3 specimens only. On caudal fin 3 short sequences of canals: one on middle and one on each lobe near the base; in varying combinations of canal numbers: 0 - 3 / 0 -2 /0- 5. Fins naked, except the spinous anal fin base covered by distal portions of body scales and proximal scaly sheath of caudal fin. Caudal fin scaled area at most occupying one half of fin length; all caudal scales ctenoid, or outer cycloid; squ. caud. 3 - 5.

Cephalic lateralis pores: 1 before nostril, 9 around orbit, 2 preorbital, 1 frontal, 5 on preoperculum, 4 on lower jaw, 2 extrascapular. Epibranchial lobe very small, 3 -4 rakers along edge in randomly inspected specimens. Ceratobranchial rakers 5 -7, small, knoblike, anteriormost minute. No gill-rakers on lower pharyngeal tooth-plate. Distinct, but small teeth internally on fourth ceratobranchial. Gill-raker like structure on dorsal edge of interoperculum. Lower pharyngeal tooth-plate triangular in dorsal aspect; posteriorwards bicuspid, rather regularly arranged and toward middle of tooth-plate large teeth, anteriorwards more conical, more scattered teeth.

Dorsal fin spines increasing in length rapidly from the first through the 4 th, from the 5 th decreasing to the 7 th or 10th, from which subequal, the last as long as or little longer than the penultimate, shorter than the 4 th. Lappets 2- 3, 2-4, or 2 - 5 produced, the 3 rd longest, may be longer than its spine and reach spinous dorsal end adpressed to the body; not produced from the 6 th, pointed, little beyond spine tips. Produced lappets in both sexes, but longer in males than in any of the females. Soft dorsal fin bluntly pointed, 4th to 6th or 5 th to 7th rays contributing to point, reaching at most little beyond caudal fin base. D. XII.9. i (1), XII. 10 (2), XIII.8. i (2), XIII.9.í (7), XIII.10 (1), XIV. 6 (1; aberrant), XIV.8. i (6), XIV. 9 (15), XIV.9. i (2); Dm. 20 (1), 22 (5), 23 (29), 24 (2). Anal fin origin opposite soft dorsal fin origin (last spine or first ray). Spines increasing in length to last. Lappets moderately long, little beyond spine tips, pointed. Soft fin bluntly pointed, rays 5 - 7 or 4 - 6 forming point, reaching at most to little beyond caudal fin base. A. constantly 3 spines; rays 6. i (2), 7 (4), 7. i (10), 8 (17), 8. i (3), 9 (1); Aw, 10 (6), 11 (27), 12 (4).

Pectoral fin roundish acuminate; rays increasing in length to 3 rd, then decreasing, 2 - 5 forming point; tip to above vent. P. 11 (29), 12 (8). Ventral fin pointed, first ray longest, inner gradually shorter. In two males of NRM 11250 extremely long first ray, beyond anal fin base; in other males and females to spinous or soft anal fin origin, Spine tip about halfway to anal fin origin. Caudal fin obliquely truncate, upper lobe slightly longer. 16 principal rays in all except one specimen with only 14.

A single supraneural spine. Vertebrae 25 (2), 26 (26), 27 (9).

Jaw dentition; 2 series in both jaws, the inner restricted to the symphyseal parts, the outer extended along the entire jaw rim. The upper jaw is much shorter than the lower and thus its outer series comprises only 9 - 13 teeth on each jaw half. The outer series of the lower jaw comprises 16- 21 teeth on each jaw half. Teeth number is positively correlated with SL. In several specimens is formed kind of a 3 rd, middle series, of only few teeth, in the lower jaw. Teeth moderate in size, shaped much like inApistogramma, slightly recurved apically, becoming smaller lateroposteriorly in the jaw, the outer larger than the inner. No canines.

Coloration: Very little variation. No apparent sexual dimorphism. Countershaded, dorsum pale brownish, gradually fading to whiteish abdominally. Very narrowly along dorsal fin base a darker stripe. Sometimes a more stepwise transformation from darker to lighter pale brown along the level of the lower lateral line. Colour pattern simple, markings dark brown. Markings frequent in Apistogramma spp., such as preorbital stripe, midventral stripe, abdominal stripes, pectoral spot or caudal spot, absent. No trace of an ocellus.

A superorbital stripe not as wide as the orbit run obliquely dorsad to occiput before dorsal fin origin, only almost connected with that of the opposite side. A broad suborbital stripe, of about pupil width dorsally, runs obliquely ventrad, narrowing, to the rim of the corner of the preoperculum, sometimes shorter.

The vertical bars, part of the ground colour rather than markings, vary in intensity between samples. In NRM 11252 they are very faint and the last is not developed. In NRM 11250 and 11251 the barring is still weak, but more prominent, and the last bar is present in several specimens, but not in the largest. Countershading is less evident in the lots 11250 and 11251 suggesting that this effect in lot 11252 is rather due to more dispersed bar pigment. The bar pigment intensity decreases ventrad. The anteriormost bar runs straight vertically ventrad from the dorsal fin origin to behind operculum. The second bar originates near the middle of the spinous dorsal fin base. The third bar runs from the posterior portion of the spinous dorsal fin base. The fourth bar is below the soft dorsal fin and the fifth originates at the end of the soft dorsal base /beginning of caudal peduncle. The sixth bar is on about the middle of the caudal peduncle. The lower limit of the bars is about at the level of the lower lateral “liriey-although both veritiicailhand horizontal borders are unsharp.

Flanc spots are of two types: A) a large dominating blackish blotch in the second bar (lateral spot), B) faint spots at the lower limits of the bars which might, with some imagination, be taken for representing an interrupted lateral band. The lateral spot has its centre between the upper lateral line and the level of the lower lateral line, and the horizontal position is about above the distal half of the pectoral fin. It is always deeper than long, in smaller specimens nearly roundish, in larger specimens of variable size. The other spots are quite weak and apparent only in the well-barred lots NRM 11250 and 11251. The first is rounded and lies just behind the gill-cover. The next is that of the third bar and it is more quadratic as that of the fourth bar. The spots of the fifth and sixth bars are again more roundish. Spot continuity in two NRM 11250 specimens results in a narrow, one scale deep longitudinal band from the fourth spot to the caudal fin. The bar pigmentation becomes very weak before the terminating spot, this granting spot integrity. Although no actual caudal base spot develops, lots NRM 11250 and 11251 show a diffuse pigmentation of the lower caudal fin lobe basally.

Unpaired fins dusky. At least in NRM 11252 males a few light spots especially near the base in soft dorsal and caudal fins. At least males, but obviously also most females with dusky outer edge of ventral fins. Ventral fins otherwise whiteish to colourless. The three anteriormost dorsal fin membranes black or almost black.

No apparent sexual dimorphism is found in the coloration of preserved specimens. In live wild and aquarium raised material studied, sexes differ slightly in that the males are on the whole more intensely coloured. Also the iridescent blue dots on the flancs (absent in preserved material) become irregular on the lateral spot in males, but are regularly arranged throughout the flancs in females. A detailed description of life colours, presumably of the syntypes, was given by Myers & Harry (1948).

Size. - The largest specimen recorded here is 33.9 mm SL, 44.5 mm TL. The sexes appear not to be differentiated in maximum size. The smallest female, 24.6 mm SL, contains immature ova. The ”holotype” is the largest recorded specimen of this species (37.3 mm SL, suggesting that it had been kept in aquarium before it was preserved). Aquarium specimens may grow as large as 4- 5 cm SL (pers. obs.) or possibly larger. “Supersize” P. ramirezi are imported to Europe from East Asia, where they are probably raised in ponds. These fishes have but little resemblance to the wild imports from South America.

Ecology.- The localities represented by the present material are in savannah regions. The brook at the mouth of the R. Guarrojo (VIT 58) had clear, slow-running water and mud-bottom, and was relatively rich in helophytes and submerse vegetation. The nearby laguna (VIT 56) had turbid water, of beige colour. The temperature was 29.5 °C in the air and 31 °C in the water, pH 5.5. Sympatric smaller cichlids were here Apistogramma hongsloi Kullander , and Crenicara filamentosa Ladiges. Females from all stations contained not fully ripe ova.

Diagnosis. - The original diagnosis failed somewhat in comparing P. ramirezi with two very different species of a very distinct genus ( Apistogramma 0 rnati~ pirınis Ahl, and A. riterısis [Haseman]), but is otherwise still useful. An improved brief characterization of the species may, however, be proposed on basis of the above description:

1. No pharyngeal gill-rakers.,

2. A small lobe on the first epibranchial, with rakers along the edge.

3. Narrow preorbital depth (3.3 -4.8% in SL, 21.6 - 36.5% in orbit diameter).

4. 25- 27 vertebrae.

5. A single supraneural spine.

In this combination of characters P. ramirezi differs from all other known cichlids. Characters 1 and 2 place it with Biotodoma Eigenmann & Kennedy , Geophagus Heckel , and Gymnogeophagus A. de Miranda Ribeiro, with which genera it also shows most superficial and meristic similarities. For field identification the small adult size and the development of the anterior dorsal fin should be useful. From the species of Biotodoma it is further distinguished by having the lateral spot anteriorly instead of posteriorly on the flancs. It has fewer (5 -7) gill-rakers than any Geophagus species (8- 22). In Gymnogeophagus species the pharyngeal teeth are moderately close-set, the posteromedial tending to be blunt, in P. ramirezi they are rather scattered and pointed. From individual species of the other genera there are in each case readily recognized differences. Data on Biotodoma , Geophagus , and Gymnogeophagus here is chiefly from Gosse (1976).

DISCUSSION

Comments on Myers & Harry (1948). - The above description of Colombian specimens agrees fairly well, when comparable, to that given of the types of P. ramirezi (Myers & Harry, 1948) . There are, however, some discrepancies, which probably to a large extent can be explained as differences in methods. Some of these are commented below with some inconsistencies in the text of the 1948 description that need observation.

1. Dentition. According to Myers & Harry: “Teeth of jaws in three rows, each series uniform in shape and length. Teeth very slender, brown tipped, the inner rows almost setiform, the outer series slightly enlarged. Premaxillary and mandibular teeth with an outer row of slightly enlarged teeth, separated from two closely spaced inner rows.” This is in considerable contrast to the condition found in the Colombian P. ramirezi . A reexamination of the types appears necessary. It is difficult to believe in the observation, but also that Myers, used to examinations of finer details in characoid dentition, would have made a mistake.

2. Squamation. Myers & Harry found “all scales ctenoid” They probably then simply overlooked cycloid head, gill-cover, throat and pectoral axilla scales. They did not find any interopercular scales, but these may be difficult to detect without manipulation. The transverse scale count is given as 111 /2 scales on p. 2, but in Table 1 is given as “scales above lateral line” 11 /2 scale and as “scales below lateral line” 10 scales. The sum is 11 ‘/z, but in this sum the lateral line scale is exluded whilst it should be included in the count given on p.2, according to the method described: “between origin of anal obliquely upwards and forward to dorsal base” Scales in a “lateral series” are 23 - 26 according to statement on p. 2, but in Table 1 the “longitudinal scale series” count is 26-29. Maybe these counts are not the same by definition. None of these differences/ inconsistencies appears important.

3. On p. 3 the caudal fin is said to have 13 principal rays and two simple rays above and below these. In Table 1 the “principal caudal rays” count is given as 13 for two specimens and as 14 for two other specimens. The inconsistency may be a lapse, the low number is probably due to vagueness in the concept of “principal caudal ray” Normally all cichlids have 16 principal caudal rays, including one or two unbranched marginal rays in each lobe. Exceptions occur, and one of the Colombian P. ramírezi has only 14 principal caudal rays. The only cichlid species known to me that is characterized by having 14 principal caudal rays is a Nannacara species (unpubl.).

4. The discussion on relationships may today seem quite in vain. Apistogramma ornatipinnis Ahl and A. rítensis (Haseman) were pointed out as the most closely related species. I have examined the holotype of the former and find it to be a specimen of A. steindachneri (Regan) , a true Apistogramma species. Myers 8t Harry had a specimen of A. steindachneri , but obviously did not consider it representing a close relative of P. ramírezí . Despite repeated efforts, I have been unable to locate the holotype of A. rítensís for reexamination, but I have little doubt that this species is closely related to or identical with A. borellii (Regan) . There is nothing to suggest relationship with P. ramirezi , the least Haseman’s (1911) retouched photograph. The failure of the discussion probably lies very much in that the authors relied almost exlusively on literature. It is not clear, however, what Myers & Harry really believed about relationship with A. ornatipinnis. The diagnosis says: “Most closely related to Apistogramma ornatipinne Ahl 1936....”, but in the discussion we read: “...we believe that ornatipinne is not close to ramirezi , ... This obscurity and other inconsistencies probably stem from a repeatedly apparent rewriting of the manuscript.

Notes on morphology. - Considering morphological characteristics of P. ramirezi it is near at hand to compare with species of the related genera. There is a danger in this, however, that some morphological characters are no doubt correlated with size. The narrow preorbital and the low scale counts are probably such character states. Unfortunately, the morphology of small Biotodorm: and Geophagus is little known, so it is difficult to say where comparisons would be fruitful. Some more spectacular characteristics may be discussed more easily, however.

1. The small subinferior mouth with a reduced upper jaw dentition is no doubt specialised. The snout is not produced following a broadening of the lacrimal bone as in Geophagus . Actually, the mouth and snout shapes approach those of Biotodoma , which I believe is otherwise a relatively ancestral group within the larger geophagines‘). The pharyngeal dentition is relatively unspecialised, not unlike that of the Bi otodoma spp. The gill-arch epithelium is unspecialised, the lobe is very small and the gill-rakers quite short. Dentition is present on the fourth ceratobranchial. In all, the oral and pharyngeal structures suggest a relatively generalised species with adaptation chiefly in the shape of the mouth parts. The diet probably consists in smaller invertebrates taken from exposed stubstrates.

P. ramirezi occurs sympatrically with at least one similar-sized Apistogramma species ( A. hongsloí ). Apístogramma spp. have a similar pharyngeal dentition, but more teeth, larger lobe, rakers on the lower pharyngeal tooth-plate, no ceratobranchial teeth, and few gill-rakers. This probably reflects some specialisation in the mastication process, the jaw morphology and dentition being very general (unpubl.). The difference suggests niche separation. The sympatric Crenicara filamentosa has a similar tooth-plate and dentition, a moderate number of gill-rakers, but no lobe, no ceratobranchial teeth, and no pharyngeal rakers. The mouth, however, is much lower in position, and the oral dentition stronger (Kullander, in prep. b).

2. The preoperculum serrations vary in presence and development between individuals-T hat is true also forApistogramma spp. (incl. Pinloichthys Fowler) in which the character is found (unpubl.). In Crenícara Steindachner , Batrachops Heckel , and Crenicichla Heckel , the only other cichlids with this character, all individuals have serrated preoperculum, and the serrations are well developed, very regular in Crenicara , more irregular in the other genera (obs. on NRM material). The functional and systematic significance of the character is uncertain. It seems very unlikely that it would reflect close relationship between the genera mentioned. It appears rather to be a plesiomorph trait which is retained, actually or as a potential, in some lineages of small or elongate cichlids. The possibility of convergent development cannot be discarded, however. Whatever the case, this character is hardly deserving the importance it has had in neotropical cichlid systematics, independent as it is of graver relationship criteria.

3. The vertebrae count, 25-27, with a mode of 26, is peculiarly intermediate between that of smaller geophagines ( Apistogramma with 22-24, Apistogrammoies j. Meinken with 24 [unpubl.]), and that of larger geophagines like Geophagus (27-34), Bíotodoma (27-28), and Gymnogeophagus (25- 30), according to Gosse’s data (1976). Probably this reflects size reduction rather than intermediate systematic position. Gosse (1976) noted that endemic species or species with restricted distribution of Geophagus , Biotodoma , and Gymrıogeophagus have lower vertebrae counts than widely distributed species. Such a tendency would then apply very well also to P. ramirezi , as well as to the usually restricted Apistogramma spp.

4. P. ramirezi is one of the smallest cichlid species. The largest specimens recorded here are 33.9 mm SL. The “holotype” is 37.5 mm SL, but its validity as a wild specimen may be questioned. Some Apistogramma spp. may be smaller: A. bitaeniata Pellegrin (largest wild specimen 33 mm SL), A. commbrae (Eigenmann) (28 mm), A. geisleri Meinken (28 mm), A. luelíngí Kullander (29 mm), A. parva Ahl (16 mm), A. viejita Kullander (30 mm), and five undescribed species of lengths of 32, 33, 28, 23 and 23 mm respectively. Some of these largest specimens are decidedly juvenile or subadult, however. (Unpubl. data.)

The Bolivian Papiliochromis . - Meinken (1967) reported a specimen of P. ramirezi from Central Bolivia, near Todos Santos in the R. Mamoré drainage basin. This finding is curious, not least because Meinken never published a description or figure of the specimen, and the little he said about it does not exactly support the determination. The length of the specimen was stated to be 52 mm SL, 72 mm TL, the depth 25 mm. It would then be almost 15 mm longer than any other specimen of P. ramirezi recorded, and have a depth of almost 50 % in the SL, whilst my deepest specimen has a depth of 43.9 %. Unfortunately, it has not been possible for me to reexamine the very specimen, but I find it impossible, on available data, to accept the record as being for the same species as that found in the Orinoco basin.

Geographical distribution. - Blass (in Anon., 1947) reported the species from somewhere in the Llanos of W Venezuela. Fischer (1968) reported it as abundant along the R. Orituco, R. Guarico, and the Caño Caballo in the Edo. Guarico, Venezuela. The collections reported here are from near Puerto Gaitán (Depto. Meta), and the mouth of the R. Guarrojo (Com. Vichada) in Colombia. Thus, collections have so far been made in two disjunct geographical areas, but the distribution probably covers a major portion of the Orinoco basin savannah areas in Colombia and Venezuela. The species is probably absent from the southernmost parts of the Orinoco basin, and from Colombia west of the Cordillera. The fishes of the Puerto Inirida area and the Cafio Bocon are well represented in Hongslo’s collection, but no P. ramirezí are among them. Western Colombia is ichthyologically one of the best known regions of South America, but without record of P. ramirezi . Outside South America the species is apparently released in rice fields on Java (P. Tzang in Socolof, 1976).

The natural endemicity to the Orinoco basin is not unique for a cichlid, but it is nevertheless noteworthy. Of the 33 other species that occur in the Orinoco basin, only ll are endemic. These are Aequidens chimantanus Inger , Ae. marine Eigenmann, Ae. metae Eigenmann, Apistogramma hoignei Meinken , A. hongsloi Kullander , A. iniridae Kullander , A. macmasterí Kullander , A. viejita Kullander , Cichla intermedia Machado-Allison , Cichlasoma axelrodí (Fernández-Yépez), and Pterophyllum altum Pellegrin . (Data based chiefly on Fernandez-Yépez, 1969 & 1971; Fowler, 1942; Gosse, 1976; Inger, 1956; Kullander, in prep. a; Luengo, 1970; Machado-Allison, 1971; Mago, 1971; doubtful records and undescribed species not included.) The endemic Aequidens spp. are very restricted in their distribution (cf. Fowler, 1942; Inger, 1956). Cichla intermedia belongs to a genus of wide distribution and is present also in the Brazo Casiquiare. Cichlasoma axelrodí is a doubtful taxon, known only from the type series, from the R. Aguaro, Edo. Guarico, Venezuela, but with strong resemblance to Amazonian congeneries. Pterophyllum altum is but vaguely distinct from the Amazonian Pt. scalare (Lichtenstein) (cf. Burgess, 1976). Five endemic species are Apístogramma , and these are the only members of this genus in the Orinoco basin, and each has but a restricted range. One species has definite Amazonian relationships ( A. iniridae ), the others are more related to each other than to any other species (Kullander, in prep. a). Papiliochromis is the only endemic Orinoco basin cichlid genus. The distribution is similar only to that of four of the Apistogramma spp., and, perhaps, the endemic Aequidens spp. By the absence of apparent close relationship to non-Orinocoan forms, it is similar only to the Apistogramma spp., but has a wider range than any other of these endemics.

It may be noted that whilst the four Apistogramma spp., probably have evolved from a single Orinocoan ancestor, P. ramirezi has not split. Indeed, it shows a remarkably slight intraspecific variation in morphology. A comparison with Apistogramma is, however, probably not possible, since this is the largest genus of South American cichlids. It appears that P. ramirezi may have evolved to generic distinctness from the rest of the larger geophagines in a northern isolation. Distant relatives, such as Geophagus daemon Heckel, G. jurupari Heckel, and G. surinamensis (Bloch) (Gosse, 1976; Luengo, 1970), may have entered the Orinoco basin only recently. The environment appears to have favored reduced size, but the pharyngeal apparatus has remained relatively generalised. The combination of apparent absence of cichlid competitors in the very area, a conservative tendency in the larger geophagines against speciation, preference for a common habitat in the region (savannah brooks and lagunas), and a free interconnection between populations may have worked to preserve the species.

Tables

Table 2. Meristics of Apistogramma ramirezi from Colombia (NRM 11250, 11251, 11252)? = mean ", SG) = standard error of the mean; sz = variance.

Count	n	Range	x	SU)	s 2
Squ. long.	37	25 -27	25.9	0.06	0.15
Llı (c)	36	12- 17	15.1	0.21	1.56
Lll (cp)	36	14- 18	15.7	0.14	0.72
Ll 2 (C)	35	4 - 9	5.8	0.25	2.26
Llz(cp)	35	6- 10	8.8	0.20	1.42
Cheek scales	37	2- 3	2.1	0.05	0.08
Squ. op.	22	10 - 17	13.7	0.47	4.87
Squ. sop.	36	3 - 6	4.7	0.13	0.57
Squ. iop.	37	1- 3	2.2	0.09	0.32
D spines	37	12 - 14	13.6	0.11	0.42
Dm.	37	20-24	22.8	0.11	0.42
A spines	37	3
Am,	37	10 - 12	10.9	0.09	0.27
P	37	11 - 12	11.2	0.07	0.17
Serrations	37	0-40	8.6	1.72	108.9
Rakers	36	5 - 7	5.7	0.10	0.33

Table 1. Morphometrics of Papiliochromis ramirezi from Colombia (NRM 11250, 11251, 11252). Range and mean (Y) is in per cent of SL. s (î) = standard error ofthe mean; S 2 = variance. SL range for n = 37: 24.6 - 33.9 mm.

Measurement	n Range	x	sff)	s2
Head length	37 29.2-32.4	30.6	0.11	0.45
Head depth	37 30.0-35.1	32.6	0.19	1.32
Head width	37 13.8 - 17.3	16.1	0.11	0.43
Body depth	37 36.7 -43.9	40.6	0.24	2.19
Predorsal length	36 35.7 -41.0	38.6	0.20	1.46
Preventral length	37 38.6-45.5	41.5	0.24	2.11
Orbit diameter	37 12.9 - 15.0	13.7	0.09	0.25
Snout length	37 5.2 - 6.9	6.2	0.07	0.19
Cheek depth	37 4.4 - 6.0	5.2	0.38	0.21
Postorbital head length	37 9.6- 11.5	10.8	0.06	0.14
Interorbital width	37 9.4- 11.8	10.6	0.09	0.32
Preorbital depth	37 3.3 - 4.8	3.9	0.06	0.12
Upper jaw length	35 6.4- 8.4	7.3	0.08	0.22
Lower jaw length	37 11.5 - 13.4	12.2	0.10	0.33
CP depth	37 15.0- 17.6	16.4	0.11	0.42
CP length	37 12.5 - 16.2	15.1	0.15	0.81
Dorsal base	37 49.0- 54.5	52.2	0.19	1.37
Anal base	37 14.5- 18.9	16.8	0.18	1.24
P length	32 25.3-31.2	28.6	0.23	1.75
V length	37 27.0-45.6	31.7	0.65	15.4
Last D spine	37 13.5- 17.9	15.5	0.17	1.04
Last A spine	37 15.7 -20.0	17.9	0.16	0.94