Austrolebias

Wilson J. E. M. Costa, 2006, The South American annual killifish genus Austrolebias (Teleostei: Cyprinodontiformes: Rivulidae): phylogenetic relationships, descriptive morphology and taxonomic revision., Zootaxa 1213, pp. 1-162 : 139-149

Key to species of the genus Austrolebias

Species assemblages of Austrolebias , as defined above, are in great part diagnosed by osteological features. In order to make possible a quick identification of species, an artificial key based on external morphological characters is provided below. Reference to the distribution maps and color photographs of specimens appearing in figures 7-52 should also be useful in the identification process.

1 a. Scales absent above anal-fin base, at least in females; scales absent from anterior portion of frontal region ................................................................................ 2

1b. Scales always present above anal-fin base and on anterior portion of frontal region ............................................................................................................................ 6

2a. (1a) Body scales overlapping, present in most portion of flanks and opercle; pelvicfin bases medially separated by short interspace, shorter than pelvic-fin base, or united; urogenital papilla cylindrical and narrow in males ................................. 3

2b. Body scales very small, not overlapping and restricted to the median portion of flanks and opercle; pelvic fins medially separated by wide interspace, much wider that pelvic-fin base; urogenital papilla conical and broad in males ............ ..................................................................................... A. cheradophilus (Fig. 17) (Isolated river basins of southeastern Uruguay) (Fig. 18)

3a. (2a) 50-75 scales in longitudinal series; prominent contact organs on most pectoralfin rays in males .................................................................................................. 4

3b. 31-35 scales in longitudinal series; minute contact organs on uppermost pectoral-fin ray and absent in other rays in males .................. A. wolterstorffi (Fig. 19) (Laguna dos Patos basin, Brazil and Uruguay) (Fig. 20)

4a. (3a) Minute scales on the pectoral-fin base and most anterior half of caudal fin in older males ........................................................................................................... 5

4b. No scales on pectoral-fin base and caudal-fin squamation restricted to basal portion of fin in both sexes .................................................... A. prognathus (Fig. 21) (Lagoa Mirim basin, Brazil and Uruguay) (Fig. 18)

5a. (4a) 16-18 dorsal-fin rays in males; gray bars on flank in males ................................. ........................................................................................... A. monstrosus (Fig. 22) ( Río Paraguay basin, in northern Argentina, Paraguay and Bolivia) (Fig. 18)

5b. 18-20 dorsal-fin rays in males; no bars on flank in males .................................... ..................................................................................... A. elongatus (Figs. 23-24) ( Río de La Plata basin, northeastern Argentina and Uruguay) (Fig. 18)

6a. (1b) Venter without scales; suborbital and supraorbital bars absent in live specimens, sometimes barely perceptible in preserved specimens ........................................ 7

6b. Venter covered with scales; conspicuous dark grey to black supraorbital and suborbital bars, both in live and preserved specimens.. .................................... 8

7a. (6a) Pectoral-fin length 23.9-26.0 % SL in males, 24.9-26.6 % SL in females; dorsal-fin origin on vertical through pelvic-fin base, or anterior to it, in males ......... ........................................................................................ A. gymnoventris (Fig. 7) (Laguna Merin basin, Uruguay) (Fig. 8)

7b.. Pectoral-fin length 26.3-29.2 % SL in males, 27.8-29.9 % SL in females; dorsal-fin origin on vertical between pectoral-fin base and anus in males ................. .................................................................................................. A. jaegari (Fig. 9) (Laguna dos Patos basin. Brazil) (Fig. 8)

8a. (6b) Dorsal-fin origin slightly anterior or slightly posterior to anal-fin origin in males; posterior and anterior portion of flank with similar color patterns in males ............................................................................................................................. 9

8b. Dorsal-fin origin well anterior to anal-fin origin, on vertical between base of 4th and 7th anal-fin rays in males; posterior and anterior portion of flank with distinct color patterns in males, anterior portion pale brown, posterior portion silver with dark gray to black bars (Fig. 10) .................... A. luteoflammulatus (Fig. 10) (Laguna Merin basin and adjacent isolated basins to south, Uruguay) (Fig. 11)

9a. (8a) Anteromedian rays of anal fin long in females, anal fin nearly triangular fin ..................................... 10

9b. Anteromedian rays of anal fin short in females, anal fin rounded .................... 22

10a. (9a) Single transverse row of scales on anal-fin base in males; minute papillate contact organs on rays of anterior half of anal fin in males, often absent; 3-4 + 9-13 gill-rakers on first branchial arch ...................................................................... 11

10b. Three to five transverse rows of scales on anal-fin base in males; prominent papillate contact organs on rays of anterior half of anal fin in males; 5 + 14-16 gill-rakers on first branchial arch ................................... A. vandenbergi (Fig. 43) ( Río Paraguay basin, in northern Argentina and Paraguay) (Fig. 11)

11a. (10a) Urogenital papilla attached to anal fin in males ............................................... 12

11b. Urogenital papilla free from the anal fin ........................................................... 15

12a. (11a) Single anterior rostral neuromast; 0-3 black spots on anterocentral portion of flanks in females ................................................................................................ 13

12b. Two anterior rostral neuromasts; 5-16 black spots on anterocentral portion of flanks in females (Fi38) .................................................. A. melanoorus (Fig. 38) ( Río Negro basin, Uruguay) (Fig. 11)

13a. (12a)17-18 supraorbital neuromasts .......................................................................... 14

13b. 20-24 supraorbital neuromasts ................................................. A. apaii (Fig. 42) ( Río Uruguay basin, Uruguay) (Fig. 11)

14a. (13a) Body depth 36.6-39.8% SL in males; pre-pelvic length 47.0-49.4% SL in males; length of dorsal-fin base 40.1-44.4% SL in males; 11-18 rows of bright dots on the flanks in males ........................................ A. univentripinnis (Fig. 39) (Rio Jaguarão drainage, lagoa Mirim basin, Brazil) (Fig. 11)

14b. Body depth 40.1-46.1% SL in males; pre-pelvic length 42.7-46.3% SL in males; length of dorsal-fin base 34.1-41.2% SL in males; 7-12 rows of bright dots on the flanks in males.. ........................................... A. bellottii (Figs. 40-41) ( Río de La Plata basin, Argentina) (Fig. 11)

15a. (11b) Pectoral fin with a conspicuous black stripe on ventral margin of pectoral fin in males ................................................... 16

15 b. No black stripe on pectoral fin in males .................................. A. viarius (Fig. 45) (River basins of southeastern Uruguay) (Fig. 20)

16a. (15a) Juveniles and adult females, sometimes adult males, with a pair of black spots arranged vertically in close proximity, sometimes coalesced to form an 8-shaped blotch, on posterior portion of caudal peduncle.. ............................................. 17

16b. No distinctive black mark on caudal peduncle (Fig. 44)............................. A. arachan (Fig. 44) ( Río Negro and laguna Merin basins, Uruguay) (Fig. 18)

17a. (16a) Black bar on parietal series of neuromasts.. ................................................... 18

17b. No black mark on parietal series of neuromasts ................... A. charrua (Fig. 46) (Lagoa Mirim basin, Brazil and Uruguay) (Fig. 52)

18a. (17a) No distinctive black bar on the dorsal and anal fins in males; dots or elongate light spots on dorsal-fin base in males.. ........................................................... 19

18b. Broad black bar on the posterior portion of the dorsal and anal fins; light blue bars on dorsal fin in males.. ............................................................................. 20

19a. (18a)Dorsal-fin origin posterior to anal-fin origin in both sexes; dark bars on caudal peduncle narrower than light interspace (Fig. 48).. ................ A. adloffi (Fig. 48) ( Río Guaíba drainage, laguna dos Patos basin, Brazil) (Fig. 52)

19b. Dorsal-fin origin posterior to anal-fin origin in males, anterior to anal-fin origin in females; dark bars on caudal peduncle wider than light interspace.. ............... .............................................................................................. A. minuano (Fig. 47) (Coastal plains adjacent to laguna dos Patos, Brazil) (Fig. 52)

20a. (18b) No light bars on anal fin in males.. ................................................................ 21

20b. Light blue bars on anal fin in males ............................................................................... A. nigrofasciatus (Fig. 49) (Canal de São Gonçalo drainage, laguna dos Patos basin, Brazil) (Fig. 52)

21a. (20a) Dorsal-fin origin on vertical between base of 2nd and 4th anal-fin rays in males; snout length 12.6-15.3 % HL; head width 66.6-71.4 % HL ................................ .................. A. nachtigalli (Fig. 50) (Rio Grande and rio Jaguarão drainages, laguna Mirim basin, Brazil) (Fig. 52)

21b. Dorsal-fin origin on vertical through 1st or 2nd anal-fin ray, sometimes slightly anterior to anal-fin origin; snout length 16.4-18.6 % HL; head width 61.4-66.6 % HL.. .................................................................................... A. salviai (Fig. 51) ( Río Tacuarí drainage, laguna Merin basin, Uruguay) (Fig. 52)

22a. (9b) Short filamentous rays on the distal margin of the dorsal fin in males; supraorbital bar longer than infraorbital bar.. .............................................................. 23

22b. Short filamentous rays on the distal margin of the dorsal fin in males; supraorbital bar longer than infraorbital bar ................................... A. patriciae (Fig. 37) ( Río Paraguay basin, Paraguay and Argentina) (Fig. 52)

23a. (22a) Pectoral fins dark gray with bright blue iridescence in males.. ....................... 24

23b. Pectoral fins hyaline in males.. ......................................................................... 33

24a. (23a) Alternating dark gray and light bars with similar width on flank in males; a black spot on anterior portion of anal fin .......................................................... 25

24b. Flank predominantly dark brown to black in males; no distinctive black spot on the anterior portion of anal fin.. ........................................................................ 26

25a. (24a) Double vertical rows of bright blue dots on flank in males; anterior three bars of flank darker than posterior bars; anal-fin base 44.7-48.7 % in males .................. ........................................................................................... A. alexandri (Fig. 25) ( Río Uruguay floodplains, Argentina, Uruguay and Brazil) (Fig. 52)

25b. No bright dots on flank in males; anterior bars of flank as dark as posterior bars; anal-fin base 36.4-42.8 % SL in males ............................ A. ibicuiensis (Fig. 26) ( Río Ibicuí dranage, rio Uruguay basin, Brazil) (Fig. 52)

26a. (24b) Infraorbital bar wide, its dorsal portion approximately so wide as orbital diameter.................................................. 27

26b. Infraorbital bar narrow, conspicuously narrower than orbit .............................. 28

27a. (26a) Dorsal-fin origin in a vertical anterior to anal-fin origin ................................................ A. nigripinnis (Fig. 27) ( Río de La Plata and río Uruguay basins, Argentina and Uruguay) (Fig. 20)

27b. Dorsal-fin origin in a vertical posterior to anal-fin origin ............................................................................................................................. A. paranaensis (Fig. 28) ( Río Paraná floodplains, Paraguay) (Fig. 20)

28a. (26b) Infraorbital neuromasts 1-2 + 19-23; preopercular neuromasts 15-21.. ........ 29

28b. Infraorbital neuromasts 2 + 25-27; preopercular neuromasts 23-26 ................... A. affinis (Fig. 29) (Upper río Negro drainage, río Uruguay basin, Uruguay) (Fig. 20)

29a. (28a) Posterior margin of pectoral fin in males reaching a vertical between urogenital papilla and base of second anal-fin ray ............................................................. 30

29b. Posterior margin of pectoral fin in males reaching between base of 2nd and 6th anal-fin rays ....................................................................................................... 31

30a. (29a) Flank with vertical rows of blue dots in males, except sometimes narrow illdefined bars on anteriormost portion of flank; preopercular neuromasts 18-21; mandibular neuromasts 11-13 .......................................... A. duraznensis (Fig. 30) (Lower río Negro drainage, río Uruguay basin, Uruguay) (Fig. 20)

30b. Flank with well-defined bright blue bars on flank in males; preopercular neuromasts15-18; mandibular neuromasts 9-10 ....................... A. juanlangi (Fig. 34) (Upper río Negro drainage, río Uruguay basin, and lagoa Mirim basin, Brazil and Uruguay) (Fig. 20)

31a. (29b) Flank with small round spots in females, spots always smaller than eye and separated among themselves by a broad space ....................................................... 32

31b. Flank with vertically elongated dark gray spots in females, usually as large as eye, sometimes forming short bars .................................................................................... A. periodicus (Fig. 33) (Rio Ibicuí and rio Quaraí drainages, río Uruguay basin, Brazil and Uruguay) (Fig. 20)

32a. (31a) Flank approximately black with 12-22 vertical rows of bright blue dots in males; bright dots on entire unpaired fins, often forming subdistal series of close dots on dorsal fin .................................................................... A. cyaneus (Fig. 31) (Lower rio Jacuí drainage, laguna dos Patos basin) (Fig. 20)

32b. Flank bluish gray, with 8-15 vertical rows of bright blue dots in males; bright dots of unpaired fins in males restricted to basal half of fins ....... A. litzi (Fig. 32) (Upper rio Jacuí drainage, laguna dos Patos basin) (Fig. 20)

33a. (23b) Flank with dark bars alternating with bright areas in larger males; 3-4 + 9 gill rakers on first branchial arch; pelvic-fin bases medially separated by short interspace.............................................................. 34

33b. Flank dark gray without bars or bright marks in larger males; 3-4 + 11 gill rakers on first branchial arch; pelvic-fin bases medially united ............................. 35

34a. (33a) 21-22 anal-fin rays in males .............................................. A. carvalhoi (Fig. 35) (Rio Iguaçu drainage, rio Paraná basin, Brazil) (Fig. 8)

34b. 23-26 anal-fin rays in males .................................................. A. varzeae (Fig. 36) (Upper rio Uruguay basin, Brazil) (Fig. 8)

35a. (33b) Flank with vertically elongated spots and bars in females; one or two large, round dark gray to black spot on caudal peduncle end in females .................... 36

35b. Flank with rounded spots and no bars in females; no distinctive spot on caudal peduncle in females ........................................................................................... 37

36a. (35a) Dorsal-fin origin on vertical between base of 1st or 3rd anal-fin ray in males; flanks brownish gray in male, usually with narrow light gray vertical lines ........ ......................................................................................... A. vazferreirai (Fig. 13) ( Río Negro and laguna Merin basins, Uruguay) (Fig. 8)

36b. Dorsal-fin origin on vertical through base of 3rd or 4th anal-fin rays in males; flank bluish gray in male, without light gray vertical lines .. A. cinereus (Fig. 12) ( Río de La Plata basin, Uruguay) (Fig. 8)

37a. (35b) Dorsal-fin origin on vertical through base of 2nd or 3rd anal-fin ray in males, slightly anterior to anal-fin origin in females ................ A. robustus (Figs. 14-15) (Isolated coastal basins of northeastern Argentina, between mouth of río de la Plata and Mar del Plata) (Fig. 8)

37b. Dorsal-fin origin on vertical between base of 4th and 7th anal-fin rays in males, between base of 1st and 4th anal-fin rays in females ... A. nonoiuliensis (Fig. 16) (Upper río Salado drainage, lower río de La Plata basin, northeastern Argentina) (Fig. 8)

Discussion

The phylogenetic hypothesis derived from the analysis of morphological characters supports monophyly of the clade comprising Austrolebias and Megalebias , as defined by Costa (2002a). This clade is also supported by molecular studies (e. g., García et al., 2002). However, while Megalebias is strongly supported by bootstrap values, monophyly of Austrolebias sensu Costa (1998a) is supported by neither morphological nor molecular data ( García et al., 2002). To make the generic classification consistent with the most recent studies, Megalebias and Austrolebias sensu Costa (1998a) are united into a single genus, Austrolebias , for which monophyly is strongly supported and is easily diagnosed both by external morphological and osteological characters (see earlier generic diagnosis).

Austrolebias was separated by Cynolebias on the basis of a series of morphological characters (Costa, 1998a), which are consistent with the present study. However, García et al. (2002) did not accept the validity of Austrolebias , claiming that a previous study ( García et al., 2000) “... supports the monophyly of the genus [ Cynolebias ] reinforcing the original taxonomical hypothesis” ( García et al., 2002: 49). Surprisingly, no specimen of Cynolebias sensu Costa (1998a, 2001) was analysed by García et al. (2000) and García et al. (2002), making this justification inconsistent. In fact, studies by García et al. (2000, 2002) strongly support the monophyly of Austrolebias . In a more recent study, Loureiro, Azpelicueta & García (2004) recognized the validity of Austrolebias , placing a new species in this genus.

Although Austrolebias and Cynolebias form a monophyletic group, well supported by morphological characters (Costa, 2002a; and present study), they are considered to be distinct genera based on phylogeny, taxonomy and biogeography. Both genera are well corroborated clades (Costa, 2001; and present study), easily identifiable both by scientists and aquarists; the name Austrolebias is today well established both in the scientific and aquaristic literature; and Austrolebias and Cynolebias , as separate taxonomic entities, may be useful in biogeographic studies, since they exhibit disjunct distribution patterns and are endemic to different climatic regions. Austrolebias is endemic to temperate and subtropical areas in the Chaco and Pampas, and Cynolebias is endemic to the tropical Cerrado and Caatinga (Costa, 2001).

García et al. (2000) analyzed 13 species of Austrolebias , based on the mitochondrial cytochrome b gene (Fig. 53a). Subsenquently, García et al. (2002) incorporated eight additional Austrolebias terminal taxa, and examined a longer stretch of the cyt-b gene and two additional mitochondrial genes, finding a different result (Fig. 53b). However, the analysis by García et al. (2002) shows some indication of errors. The cynolebiatine clade, comprising the genera Austrolebias , Cynolebias , Simpsonichthys , Nematolebias , Leptolebias , Cynopoecilus , and Campellolebias according to Costa’s (1998a) classification, is strongly supported both by different morphological data (Costa, 1990, 1998a) and two independent molecular analyses (Hrbek & Larson, 1999; Murphy et al., 1999). However, according to García et al. (2002), this group would be paraphyletic, with Cynopoecilus melanotaenia (Regan) more closely related to Plesiolebias glaucopterus (Costa & Lacerda) and Trigonectes balzanii (Perugia) than to species of Austrolebias , Simpsonichthys , and Leptolebias . Thus, it would be desirable to perform new molecular analyses, preferentially including all species of Austrolebias and larger samples of different genes, with the goal of testing the present morphological phylogenetic hypothesis.

Despite the discrepancy between data presented by García et al. (2002) and data discussed above, some zones of congruence between García et al.’s (2002) molecular hypothesis and the present morphological analysis may be detected. Clades of Austrolebias defined by morphological and molecular data are compared and discussed below (see also Fig. 54 for a strict consensus tree combining Fig. 1 and 53b).

Austrolebias robustus species group. Clade highly supported by morphology. Only two species considered valid in the present study were included in the molecular analysis, which were placed in a well-supported clade, thus indicating congruence between morphological and molecular data.

Austrolebias elongatus species group. Clade highly supported by morphological data. Also supported by molecular data, although two species were not included in the molecular analysis. This indicates congruence between morphological and molecular data, although presenting highly conflicting hypotheses of relationships among included species.

Austrolebias robustus species group + Austrolebias elongatus species group. Clade supported by morphological data. According to molecular data, a clade comprising the A. robustus species group and the A. elongatus species group also includes species of the A. bellottii group. This indicates a partial congruence between morphological and molecular data.

Austrolebias alexandri species group. Clade supported by morphological data, but corroborated by molecular data only if A. nigripinnis is excluded, which would be the most basal species of Austrolebias . This indicates a partial congruence between morphological and molecular data.

Austrolebias bellottii species group. A subclade including A. arachan , A. viarius and species of the A. adloffi group is supported both by morphological and molecular data. However, according to molecular data, A. apaii would be more closely related to the A. robustus species group than to A. arachan , A. viarius , and species of the A. adloffi group. Therefore, there is a partial congruence between morphological and molecular data.

Austrolebias luteoflammulatus + A. gymnoventris . Clade supported by molecular data. Both species and A. jaegari , which was not included in the molecular analysis, share an anteriorly positioned dorsal fin, which could be interpreted as synapomorphic. However, the morphological analysis does not corroborate monophyly of this assemblage. Thus, there is no congruence between morphological and molecular data.