Hybrids, not a new rheophilic morph of Vieja hartwegi

Abstract

Gómez-González et al. (2018) have presented specimens which they identify as a new rheophilic morph of Vieja hartwegi (Taylor & Miller, 1980). Below we question this species determination and instead demonstrate that they are most likely natural hybrids of Vieja hartwegi and the syntopic Chiapaheros grammodes (Taylor & Miller, 1980). First, we present evidence from the reported specimens, then we compare the evidence with a documented case of hybridization of these two species (Buchhauser, 1999) and other hybrids, and finally we provide additional discussion.

Full text

1
Hybrids, not a new rheophilic morph of Vieja hartwegi
Oldřich Říčan1*, Rico Morgenstern2
1 University of South Bohemia, Faculty of Science, Department of Zoology, Branišovská 31
(CZ-370 05), České Budějovice, Czech Republic.
2 09619 Mulda, Germany
* Correspondece: oldrich.rican@prf.jcu.cz
Gómez-González et al. (2018) have presented specimens which they identify as a new rheophilic
morph of Vieja hartwegi (Taylor & Miller, 1980). Below we question this species determination and
instead demonstrate that they are most likely natural hybrids of Vieja hartwegi and the syntopic
Chiapaheros grammodes (Taylor & Miller, 1980). First, we present evidence from the reported
specimens, then we compare the evidence with a documented case of hybridization of these two
species (Buchhauser, 1999) and other hybrids, and finally we provide additional discussion.
The first and most striking observation regarding the reported specimens is that none can be
determined as V. hartwegi. The specimen in their Fig. 4A has no attributes of V. hartwegi whatsoever
and looks like a small-mouthed C. grammodes and has a much larger head than any of the other
shown specimens. The other three specimens (their Fig. 4B-D) and those in their Fig. 6 are
indeterminable to species level and thus cannot be unequivocally determined as V. hartwegi.
The second striking observation is the high amount of shape and coloration-pattern variation
between the supposedly rheophilic specimens. The specimens in their Figs. 4A, C, D and Fig. 6 are
elongated while the specimen in Fig. 4B is rather deep-bodied. The specimen in their Fig. 4A reminds
closely of C. grammodes in overall coloration, showing a midlateral and caudal blotch as well as the
Chiapaheros-typical rusty spots on body and vermiculations on head, which are, however, much
coarser than in C. grammodes and do not extent onto upper snout and forehead. Similar markings are
found in all specimens in their Fig. 4 and are the only unifying feature of them; these spots and
vermiculations are intermediate between the corresponding markings of C. grammodes and V.
hartwegi (which has relatively coarse, irregular spots on head and body). Moreover, none of the
specimens (their Figs 4, 6) has the Vieja-typical course of the midlateral stripe (in a lowered position;
Říčan et al., 2005, 2016; their Figs 3, 5 and 9), but all have the plesiomorphic position (ancestrally
shared with C. grammodes). None of the specimens (their Figs 4 and 6) has the apomorphic
neomorphic condition of the midlateral blotch of Vieja (diffuse, elongated by fusion from two or more
vertical bars, shifted in position dorsally; Říčan et al., 2005, 2016; their Figs 3, 5 and 9) but all have
the plesiomorphic condition also found in C. grammodes. None of the specimens has the Vieja-typical
elongated caudal fin-base blotch (extending onto caudal peduncle and often continous with the

2
midlateral stripe). Instead all specimens have the plesiomorphic blotch (rounded and exclusively on
caudal fin; their figs 4 and 6) found also in C. grammodes. To sum up, the melanin coloration patterns
of the specimens are thus not those diagnostic of Vieja but are all as found in C. grammodes.
The high amount of shape variation and notable differences between Vieja species and the
reported specimens here suggested as of hybrid origin is evident in the reported proportional values
(Table 2 in Gómez-González et al., 2018). In all genera of the Theraps-Paraneetroplus clade of
herichthyines, the head is (irrespective of their ecomorphology) short and deep, and especially
foreshortened in the rheophilic genera. The reported specimens have none of these characteristics of
the rheophilic herichthyines and instead have long heads and comparatively large mouths (not evident
from the jaw length in Table 2 because there given only in relation to head length), a unique
combination among their containing group. Table 2 in Gómez-González et al. (2018) reports a notable
difference in head length (34.3% SL vs. 28.3) between the examined specimens of the normal morph
of V. hartwegi (vs. 28.3–30.4 in the two examined Vieja species), but the table contains errors that
make these and other values hard to trust, because the mean 34.3 for the putative hybrids is listed as
derived from a range of 30.7–31.1 (in Table 2 also the minimum and maximum values are switched
for the paratypes of V. hartwegi, their head length and head depth stray much from the examined
specimens, and the Max-column for the ‘rheophilic morph’ repeats the data for the normal morph for
all values). Also, the listed SD (5.4) is much higher than the SDs in the normal species (1.2, 1.3, 1.5)
demonstrating a much higher variability among the specimens of the putative hybrids (see above).
The mean head length of the reported specimens (34.3% SL) is intermediate between the short-headed
Vieja species values (28.3–30.4) and those of the long-headed C. grammodes (based on Taylor and
Miller, 1980) (39.6% SL).
The reported specimens also have a much narrower head (as judged from the interorbital distance)
of 25.4% HL vs. 40.8 in V. hartwegi and 39.6–45.4 in the other two reported Vieja species. Again, the
interorbital distance of the reported specimens (25.4% HL) is thus completely outside the values of
Vieja (39.6–45.4) but very similar to C. grammodes (based on Taylor and Miller, 1980) (about 25.0%
HL).
The optically much larger mouth of the reported specimens has not been addressed in the study
but it does not fit to a rheophilic morphology. All rheophilic species of herichthyine cichlids of the
Theraps-Paraneetroplus clade have small, subterminal mouths, in size thus comparable to the lentic
Vieja, differing only in their inferior position on the head (Říčan et al., 2016).
In light of the above observations the much larger head, mouth and much more sharp heads of the
reported specimens (their Figs 4 and 6 and Table 2) than found in V. hartwegi and in other Vieja (their
Figs 3, 5 and 9) are easily explained as averaged values of the extreme morphologies found in the two
proposed parental species, V. hartwegi and C. grammodes.
Additional arguments for the hybrid origin of the reported specimens are found in their meristic
characters. Vieja species have a combination of modally 15 abdominal and 15 caudal vertebrae (a total

3
of 30; Říčan et al., 2016). This diagnosis of Vieja has been questioned by Gómez-González et al.
(2018) but on the wrong assumption. The ancestral count of vertebrae in Middle American cichlids is
overwhelmingly robust and is modally 13 abdominal plus 16 caudal (13+16). The diagnosis of Vieja
by 15+15 thus states that there was an increase in modal number of abdominal vertebrae from the
ancestral condition to at least 15 and a decrease of caudal vertebrae to at least 15 (modally 15,
extreme range 14–16), and this combination is unique among Middle American cichlids. That e.g. V.
bifasciata has only 14 caudal vertebrae as reported by Gómez-González et al. (2018) is thus not a
negation of the diagnosis, but is fully in agreement with it. The 15+15 vertebrae counts thus agree
with those reported for V. hartwegi by Gómez-González et al. (2018). The specimens here interpreted
as hybrids have 15 abdominal and 15(1)–16(2) caudal vertebrae, total 30(1)–31(2) according to
Gómez-Gonzáles et al. (2018). The specimen in their Fig. 8B actually has 17 caudal vertebrae.
Chiapaheros grammodes has modally 13+17(18) vertebrae. The elevated number of caudal vertebrae
can thus again be explained by hybridization with C. grammodes, since the specimens have Vieja-like
numbers of abdominal vertebrae (15) and Chiapaheros-like numbers of caudal vertebrae (up to 17).
Middle American cichlids with a combination of 15+17 include only three species (Říčan et al., 2016
based on x-ray examination of >1100 heroine specimens): Rheoheros, Talamancaheros and
Tomocichla tuba; all easily distinguishable from the reported specimens; a combination of 15+16 is
found addtionaly only in Amphilophus citrinellus, Cincelichthys, Chuco, and Vieja bifasciata.
A likely decisive difference between the reported specimens and Vieja species is evident in Fig. 8
of Gómez-González et al. (2018). The difference concerns the oral teeth. The specimen of normal V.
hartwegi in Fig. 8A shows typical Vieja teeth (type F or E sensu Říčan et al., 2016) that are
characteristically flattened anteroposteriorly, while the specimen in Fig. 8B, the likely hybrid, has
clearly visible conical, pointed teeth, additionally with an enlarged pair of teeth along the symphysis
of the upper jaw and much smaller remaining teeth, a tooth morphology sensu Říčan et al. (2016) of
type B or C, typical of predatory heroines (including C. grammodes) and not a rheophilic-morph tooth
morphology of species in the Theraps-Paraneetroplus clade of herichthyines (type F or E; see Říčan
et al., 2016 for photos and a review of all tooth-types in heroine cichlids and their association with
feeding and habitat ecology). This important difference between the oral teeth in Fig. 8A and B has
not been addressed in the paper and the teeth are not described as markedly different, which they are
even from what is visible in the two photographed specimens in Fig. 8. No herichthyine in the
Theraps-Paraneetroplus clade has type B or C oral teeth (Říčan et al., 2016).
Another important difference is also visible in Fig. 8 of Gómez-González et al. (2018). The
anguloarticular bone of the lower jaw in the two specimens in Fig. 8 has a completely different shape
and proportions that are beyond variation found conspecifically in Middle American cichlids (Říčan,
unpublished data). The anguloarticular bone acts as two opposing lever systems in the cichlid lower
jaw, one opening the jaw, and the other closing it, and these are governed by the sizes and shapes of
the retroarticular process and the coronoid process of this bone (see Albertson & Kocher, 2006 for a

4
review). The putative hybrid specimen in Fig. 8B has a long ventral base of the retroarticular process
and a very sharp coronoid process, the specimen of V. hartwegi in Fig. 8A has a short ventral base of
the retroarticular process and a robust coronoid process. These two types of the anguloarticular bone
are based on examined material of Middle American cichlids (Říčan, unpublished data) basically on
the opposite ends of the spectrum of variation. The more a species is predatory/piscivorous the longer
is the ventral base of the retroarticular process and the sharper and shorter its coronoid process (the
extreme is found in the highly specialized piscivores, i.e. Parachromis, Petenia, Kronoheros and
Caquetaia) while on the opposite end herbivorous species have a short ventral base of the
retroarticular process and a robust and long coronoid process (see Fig. 8 in Říčan et al., 2016 for the
variation in length of the ventral base of the retroarticular process). The likely hybrid specimen in Fig.
8B thus has a shape of the anguloarticular bone resembling generalized predatory species, while the
one in V. hartwegi in Fig. 8A agrees with other Vieja species and other detritivorous/herbivorous
genera.
Based on their oral teeth and jaw morphology, head and mouth shape and size the reported
specimens cannot be placed among the rheophilic guild (ecomorph) since all rheophilic herichthyines
have a markedly different morphology as explained above. The reported specimens also cannot be
placed in the rheophilic guild based on field data. The specimens have been collected in syntopy with
V. hartwegi and the streams have not been surveyed along an elevational gradient. The whole
argument for interpretation of the specimens as a rheophilic morph is thus completely without any
evidence.
Gómez-González et al. (2018) have tried to demonstrate the conspecifity of the supposed
rheophilic morph by molecular phylogeny, but unfortunately have only used a uniparentally inherited
mtDNA marker. Recombinant nuclear DNA multilocus markers are now widely used in Middle
American and Neotropical cichlid studies (e.g. Říčan et al., 2016; Ilves et al., 2017; Burress et al.,
2018a,b) and should be used on these specimens in question to provide the definitive test of their
origin.
We now turn to documented cases of hybrids in Middle American cichlids which further
underscore our interpretation of the reported specimens as of hybrid origin. The high amount of shape
and coloration-pattern variation between the reported specimens (see above) is a typical feature of low
generation-specimens of hybrid origin. Based on this large variation in the reported specimens by
Gómez-González et al. (2018) we deduce that the specimens are of a recent hybridization event(s).
Říčan et al. (2005) reported specimens from a cross between a female Parachromis managuensis and
male Amatitlania nigrofasciata. The hybrids remained fertile at least to generation F4 (after which the
experiment was discontinued) and already since generation F1 showed a large amount of shape and
coloration variation that could be grouped into two forms, each more resembling one of the parental
species. Similar large variation in shape and coloration of specimens has also been obtained in e.g.
hybrids of Vieja x Parachromis, or in hybrids of Amatitlania nigrofasciata x Thorichthys meeki. In all

5
these hybrids there is a large amount of shape and coloration variation that can again be grouped into
two forms, each more resembling one of the parental species. This type of variation closely matches
the variation shown in Fig. 4 in Gómez-González et al. (2018), where specimen A is very similar to C.
grammodes and the three other specimens are a mix of body shapes between the two putative parental
species, with coloration patterns predominantly of C. grammodes.
The likelihood of hybrid origin for the reported specimens by Gómez-González et al. (2018)
reaches virtual certainty due to reported hybrids (Buchhauser, 1999; Fig. 1) of the two purported
parental species (V. hartwegi x C. grammodes; Fig. 1). The hybrid specimens of Buchhauser (1999)
and the specimens of Gómez-González et al. (2018) share similar intermediate (between V. hartwegi
and C. grammodes) body, head and mouth shape, mouth size and coloration patterns, chiefly
characterized by the C. grammodes-typical rusty spots and body and head and especially the coarse
vermiculations on the cheek and preorbital region, and by the absence of apomorphic Vieja-typical
coloration patterns (see above, i.e. course of the midlateral stripe, condition of the midlateral blotch,
condition of the caudal fin-base blotch).
Rheophilic morphs (or forms) of cichlid species are not known to differ so utterly in coloration
patterns as to render the specimens indeterminable to species level as in the case of the specimens
reported by Gómez-González et al. (2018). Typically such rheophilic forms are just more streamlined
but easily recognizable variants of their species. To demonstrate our case we provide an example (see
Musilová et al., 2015) of a rheophilic (lotic) form of Mesoheros festae in comparison with the typical
lowland (lentic) form (Fig. 1). These two forms really differ in their occupied habitats but share the
same coloration patterns unlike the situation reported in Gómez-González et al. (2018). Similar
variation of morphology with habitat (e.g. water-flow, turbidity, bottom type) is common in many
herichthyine species and is not worth of describing as special morphs, since the variation has the form
of a cline and there are no sharply delineated distinct and habitat-segregated morphs. As an example,
Říčan et al. (2016) and Pérez-Miranda et al. (2017) have reported an ecomorphological dichotomy in
the herichthyine species H. pantostictus that however on a more detailed examination will also likely
prove to be a smooth cline.
We believe to have convincingly demonstrated that the description of the supposed rheophilic
morph of Vieja hartwegi by Gómez-Gonzáles at al. (2018) is based actually on natural hybrids
between V. hartwegi and C. grammodes. Since the data concerning the ‘rheophilic morph’ are given
separately in diagnosis, description and table, they do not distort the information on the ‘normal form’
and its distinction from related species. Therefore, the redescription of V. hartwegi (safe for the partly
flawed table) is nevertheless a useful contribution to the knowledge of the species.
We would like to thank Peter Buchhauser (Tegernheim, Germany) for providing photos and
additional information on the aquarium hybrid specimens of V. hartwegi x C. grammodes and their
parents.

6
References
Albertson, R. C. & Kocher, T. D. 2006. Genetic and developmental basis of cichlid trophic diversity.
Heredity 97: 211–221.
Buchhauser, P. 1999. Ungewöhnliche Hybriden bei Cichliden. Deutsche Cichliden Gesellschaft-
Informationen 30(7): 130-135.
Burress, E.D., Alda, F., Duarte, A., Loureiro, M.. Armbruster J.W. & Chakrabarty, P. (2018a)
Phylogenomics of pike cichlids (Cichlidae: Crenicichla): the rapid ecological speciation of an
incipient species flock. Journal of Evolutionary Biology 31: 14-30.
Burress, E.D., Piálek, L., Casciotta, J.R., Almirón, A., Tan, M., Armbruster J.W. & Říčan, O. (2018b)
Island- and lake-like parallel adaptive radiations replicated in rivers. Proceedings of the Royal
Society B 285: 20171762.
Gómez-González, A.E, Álvarez, F., Matamoros, W.A., Velázquez- Velázquez, E., Schmitter-Soto, J.J.,
González-Díaz, A.A. & McMahan, C.D. (2018) Redescription of Vieja hartwegi (Taylor & Miller
1980) (Teleostei: Cichlidae) from the Grijalva River basin, Mexico and Guatemala, with
description of a rheophilic morph. Zootaxa 4375 (3): 371–391.
Ilves, K.L., Torti, D. & López-Fernández, H. (2017) Exon-based phylogenomics strengthens the
phylogeny of Neotropical cichlids and identifies remaining conflicting clades (Cichliformes:
Cichlidae: Cichlinae). Molecular Phylogenetics and Evolution 118: 232–243.
Musilová, Z, Říčan, O., Říčanová, Š., Janšta, P., Gahura, O. & Novák, J. (2015) Phylogeny and
historical biogeography of trans-Andean cichlid fishes (Teleostei: Cichlidae). Vertebrate Zoology
65: 333-350.
Pérez-Miranda, F., Mejía, O., Soto-Galera, E., Espinosa-Pérez, H., Piálek, L. & Říčan, O. (2017)
Phylogeny and species diversity of the genus Herichthys (Teleostei: Cichlidae). Journal of
Zoological Systematics and Evolutionary Research 10.1111/jzs.12197.
Říčan, O., Musilová, Z., Muška, M. & Novák, J. (2005) Development of coloration patterns in
Neotropical cichlids (Perciformes: Cichlidae: Cichlasomatinae). Folia Zoologica Monographs 1:
1-46.
Říčan, O., Piálek, L., Dragová, K. & Novák, J. (2016) Diversity and evolution of the Middle American
cichlid fishes (Teleostei: Cichlidae) with revised classification. Vertebrate Zoology 66: 1-102.
Taylor, J.N. & Miller, R.R. (1980) Two new cichlid fishes, genus Cichlasoma, from Chiapas, Mexico.
Occasional Papers of the Museum of Zoology, University of Michigan, 693, 1–16.

7
Figure 1. The figure demonstrates
one of the hybrids obtained from a
cross between Vieja hartwegi and
Chiapaheros grammodes
(Buchhauser, 1999); the two
parental species are also shown
(photos courtesy Peter
Buchhauser). Further, a real
situation of intraspecific
ecomorphological dichotomy is
demonstrated in an example from
Mesoheros festae (collected by
Říčan, 2009; see Musilová et al.,
2015) from a highland lotic
locality with corresponding
morphology (0 23 21.8 S, 79 14
35.5 W; 336 m a. s. l.) and a
lowland lentic locality with
corresponding morphology (1 12
29.7 S, 80 21 34.6 W; 70 m a. s.
l.).