Pseudocheirodon terrabae Bussing 1967

Cheirodon terrabae (Bussing)

Pseudocheirodon species.—Miller, 1966:784 [listed].

Pseudocheirodon terrabae Bussing, 1967:212 [original description; Rio Grande de Terraba, Costa Rica], 241, fig. 1 [photograph].—López, 1972:101 [listed from Terraba and Pirris river basins; in her reprints López has indicated that the Turrubares basin is an error and should be Pirris (PI)], 105 [compared to C. affinis], 117, fig. 2 [distribution map], 121, fig. 6 [photograph], 123, fig. 7 [distribution map], 129, fig. 14 [scales].

MATERIAL EXAMINED

(All specimens from Costa Rica)

*LACM 2951, paratypes, Costa Rica: San Jose Province, tributary to Rio General, W. A. Bussing, 28 Dec. 1961; 12 specimens 30.0–47.1 mm (1 cleared and stained).

*USNM 194236, Rio Union, IAH near San Isidro, Albert A. Greenberg, 20 Jan. 1963; 4 specimens 30.4–36.0 mm (1 cleared and stained).

*ANSP 121988, with same data as USNM 194236, 1 specimen 35.4 mm.

*CAS 16061, with same data as USNM 194236, 1 specimen 31.1 mm.

*BMNH 1973.2.16;2, with same data as USNM 194236, 1 specimen 35.4 mm.

*USNM 194234, Rio Alia Nuo, near IAH, Albert A. Greenberg, 21 Jan. 1963; 2 specimens 30.2–30.4 mm.

*USNM 194220, with same data as USNM 194234; 1 specimen too damaged to measure.

*UMMZ 194214, overflow pond of Rio Grande de Terraba, Palmar Norte, R. M. Bailey and party, June 1973; 32 specimens 39.7–50.5.

Specimens in description are included in tables.

DESCRIPTION.—Standard length of specimens examined 29.8–49.8 mm. Body deep, compressed laterally; greatest body depth 32.9–42.6. Predorsal body profile somewhat convex with a slight concavity at nape. Between posterior dorsal-fin termination and anterior adipose-fin base, body profile very slightly convex. Some specimens with a slight rise in body profile at base of adipose fin; between posterior adipose-fin termination and dorsal procurrent caudal-fin rays, body profile slightly concave. Distance from eye to dorsal-fin origin 37.2–43.0; distance from dorsal-fin origin to posterior end of caudal peduncle 51.0–58.0. Ventral body profile rounded from jaws to anus; steepest inclination ventral to jaws. Ventral profile protrudes ventrally its greatest distance at point just anterior to or just posterior to pelvic-fin insertion. Body profile along anal-fin base straight or slightly convex; body profile concave between posterior anal-fin termination and procurrent caudal-fin rays. Caudal peduncle depth 11.0–13.2; peduncle length 11.3–15.0.

Head length 22.8–26.6. Eye diameter 8.7–10.4. Snout 5.4–6:5. Least bony interorbital width 7.6–9.2. Maxillary short, sloping ventrally and posteriorly, forming an angle of 50–60 degrees to longitudinal body axis; upper-jaw length 6.0–7.7. Upper-jaw teeth with 6–10, usually 8–9, cusps (except smallest maxillary teeth, which have 5–6 cusps), all in a single series. Maxillary teeth 3–6, usually 4–5, with a rounded cutting edge and with outer cusps equidistant from tooth base. Premaxillary teeth 5 in all specimens with a rounded cutting edge, cusps on anterior cutting edge begin nearer to tooth base than cusps on posterior cutting edge. Dentary with 8–10 teeth, with 6–10, usually 8–9, cusps, anterior teeth larger, becoming smaller posteriorly (posterior 2 teeth may be tricuspid or conical); teeth with an even cutting edge, median cusp not elongate, long axis of anterior dentary teeth extending dorsoanteriorly or may be nearly horizontal in some specimens. No teeth present on vomer, palatines, or pterygoids.

Fontanel moderate, length of that part anterior to epiphyseal bar slightly less than width of fontanel posterior to bar. Gill rakers moderate, 19–22, usually 21. Circumorbital bones well ossified, infraorbital 3 wide, contacting preopercle ventrally and with a naked area posteriorly.

Scales moderately large, cycloid, with concentric circuli and about 5–8 radii on exposed posterior field. Lateral line complete, extending as a ventral curve on side of body to below adipose fin, then continuing below midline as a straight line to caudal-fin base, sometimes continuing slightly onto middle caudal-fin rays. Lateral-line scales approximately 33–34. Scales above lateral line 6–7, usually 6; scales below lateral line 5–6. Predorsal scales number about 11. Scale sheath at base of anal fin of about 3–6 scales. Axillary scale present dorsal to pelvic-fin insertion; caudal fin scaleless.

Dorsal fin with ii,9 rays in all specimens. Dorsal-fin origin anterior to anal-fin origin, posterior to pelvic-fin origin, nearer eye than base of caudal fin. Distance from tip of snout to dorsal-fin origin 50.0–57.6. Third or fourth ray of dorsal fin longest, with posterior rays shorter, forming a slightly convex posterior margin to fin; length of longest ray 27.8–34.4.

Anal fin with 4 (rarely 5) unbranched rays and 19–23, usually 22 or 23, branched rays. First unbranched ray not always visible except in radiographs. Anal-fin origin posterior to midpoint of standard length 61.3–69.5. Fourth through seventh or eighth anal rays longer, with successive rays shorter, making ventral margin of anal fin concave. Dorsally recurved anal-fin hooks present in males only (hooks were not noted in original description) on last unbranched ray and first through eleventh to seventeenth branched rays, usually only 1 hook or bilateral pair of hooks per bony ray segment (rarely with 2 hooks on same side of segment); fin often fleshy around hooks.

Pectoral fin with 1 unbranched ray and 11–12, usually 11, branched rays. Pectoral fins reach from just anterior to or slightly beyond pelvic-fin insertion. Distance from snout tip to dorsal end of pectoral-fin base 24.0–29.0 and length of pectoral fin from base to tip of longest ray 19.4–24.1.

Pelvic-fin rays i,7 in all specimens, distal end reaching to or just anterior to anterior basal termination of anal fin. Distance from tip of snout to pelvic-fin origin 45.1–51.1; pelvic-fin length 16.7–21.3. In males, first through seventh branched rays with antrorse hooks usually on ventral surface of ray segments; 1 hook, rarely 2, present per bony ray segment.

Principal caudal-fin rays 10/9 (1 specimen with 9/9, 1 with 9/8); fin forked, not split to base; no hooks present.

Precaudal vertebrae 16; total vertebrae 33–34, usually 33.

Color in alcohol: Ground color pale brown (gray brown in freshly preserved specimens), back and nape darker brown. Small melanophores present on sides bordering scale pockets, more numerous dorsally, forming a reticulate pattern above midline, loosely following myomere junctions below midline above anal fin, present sparingly across belly. Melanophores sometimes slightly concentrated at “pseudotympanum,” forming a weak humeral spot. Melanophores concentrated at end of caudal peduncle, forming a large caudal spot that extends along middle caudal rays. Freshly preserved material with silvery cheeks and with a broad silvery lateral stripe; this silvery coloration less evident in other specimens examined. Dorsal fin with melanophores on interradial membranes. Caudal fin with melanophores on interradial membranes, usually concentrated near middle caudal rays. Anal fin with melanophores on interradial membranes (occasionally on rays) proximally, but melanophores usually lacking on distal one-half to one-fourth of fin. Pectoral fin with melanophores ventrally on rays, lacking on interradial membranes. Pelvic fins with melanophores primarily on interradial membranes, concentrated distally.

Range: Bussing (1967) states that C. terrabae has been found in the basin of the Rio Grande de Terrabae. López (1972) lists it also from the Rio Jicote, of the Rio Pirris basin, also in Costa Rica. Bussing in personal communication has suggested that the Rio Alla Nuo might be Saloma Nuevo, 8°15′N, 83°13′W.

COMPARISONS.—Cheirodon terrabae is very closely related to C. affinis of Panama. It differs from that species in having a greater number of dentary teeth (8–10, usually 9, vs. 7–9, usually 7, in C. affinis), greater modal number of maxillary teeth (3–6, usually 5–4, vs. 3–4, usually 3, in C. affinis), and slightly greater average number of anal-fin rays (22.1 vs. 19.6 in C. affinis). Other apparent differences occur in the number of scales below the lateral line and number of gill rakers (see tables). There may also be a greater number of hooked anal-fin rays in C. terrabae males. In some populations of C. terrabae (e.g., LACM 2951), there may be a modal number of 3 maxillaty teeth, leading to possible confusion with C. affinis.

Cheirodon terrabae is distinguished from C. dialepturus, C. gorgonae, and C. mitoptems in lacking modified caudal squamation and in its possession of asymmetrical premaxillary teeth and dentary teeth without an enlarged median cusp. Young Carlana eigenrnanni, which have teeth similar to those of C. terrabae, can be quickly identified by the much higher number of anal-fin rays; other differences between these species may be found under C. eigenmanni and in the tables. Number of anal-fin rays, position of the dorsal fin, and dentition are among the numerous differences between C. terrabae and P. macrolepis (see P. macro lepsis description below).

Bussing (1967) remarked:

Although it has been proposed that the present genera Odontostilbe, Holoshesthes, and Saccoderma have retained the complete lateral line of an Odontostilbe-like ancestor and that the general trend in minute species is toward reduction or suppression of the lateral line…, such is not the case with P. terrabae. If it is assumed that P. terrabae is a descendant species of P. affinis, then the new species has attained a complete lateral line and certain populations of P. affinis are actively mutating in the same direction.

To assume that one existing species is the ancestor of another as Bussing does is presumptuous and probably fallacious (cf. Hennig, 1966). In all probability these two species, which seem closest to one another, had a common ancestor not too different from either existing species. We cannot see any basis for predicting the characters of the ancestors in this case, especially not by the use of such apparently labile characters as the length of a lateral line. This is the primary reason why we have not accepted the genus Odontostilbe (see discussion above under the generic name Cheirodon). We do, however, fully agree with Myers (1958) that loss of a lateral line is a general trend in small characids. We believe that it probably happened repeatedly and independently in different groups of species and genera, but we agree with Bussing's suggestion that occasionally the line may have reappeared, confusing the attempt to establish relationships by use of this character.

Carlana Strand

Carlia Meek, 1914:108 [type-species Cheirodon eigenmanni Meek, 1912, by original designation].

Carlana Strand, 1926:54 [type-species Cheirodon eigenmanni Meek, 1912, replacement name for Carlia Meek, 1914, preoccupied by Carlia Gray, 1854].

Adults with teeth along entire ventral maxillary margin; dentary teeth broad with 9–12 cusps; number of branched anal-fin rays 26–31 (in Carlana eigenmanni, only known species), compared with 15–23 or 45–50 in other genera in geographical area covered here. Weak spinules present on pelvicand anal-fin rays of adult males (see discussion below).

Meek (1914) indicated that his new genus Carlia was closely related to Rhoadsia Fowler, which is currently included in a subfamily Rhoadsiinae. Eigenmann and Myers (1929) included Carlia in Rhoadsia but stated that the species C. eigenmanni was very likely generically distinct from other species of Rhoadsia. Juvenile Rhoadsia have a single series of premaxillary teeth and gain a second row when approaching adulthood [reaching up to 168 mm according to Eigenmann and Myers (1929)]. The maxillary in juvenile Rhoadsia has but 2 multicuspid teeth, while the maxillary in adults has, in addition, numerous conical teeth along its entire ventral margin. We have 4 adult males (as evidenced by anal- and pelvic-fin spinules) of Carlana eigenmanni ranging from 52.8–53.9 mm, all of which lack an outer premaxillary tooth row but have a fully toothed maxillary. We are confident that Carlana does not have an outer row of premaxillary teeth. According to the present classification system, this feature alone would remove Carlana eigenmanni from Rhoadsia and place it in the Cheirodontinae.

We have examined several specimens of Rhoadsia as follows: R. minor (CAS[IUM] 13419, holotype; CAS[IUM] 13420, 92 specimens, paratypes; CAS[SU] 49795, 4 specimens; CAS[SU] 49796, 1 specimen); R. altipinna (FMNH 55907, 1 specimen, paratype; USNM 83543, 1 specimen; USNM 83536, 1 specimen). Adult specimens have dentary teeth quite different from C. eigenmanni in that they are not nearly as broad and have 4 large cusps with 1 or 2 smaller cusps (compared to very broad teeth with 9–12 cusps in C. eigenmanni). Also, all adults have a well-defined outer row of 2 or 3 premaxillary teeth in contrast to C. eigenmanni, which has only 1 tooth row.

While these differences in dentition are, according to the currently accepted system, of generic significance, it is obvious that a complete review of the so-called Rhoadsiinae, sensu lato (including Parastremma Eigenmann, Rhoadsia, and possibly Carlana), is needed. There are many characters in all these species that possibly indicate relationships, including overall similarity of body and fin shape and ontogenetic dentition changes. Despite the fact that Carlana, because of its single row of premaxillary teeth, must currently be defined as a so-called cheirodontin, it is possible that further investigation will place all the species now in Parastremma and Carlana in Rhoadsia.