7838

• UCREFRP Catalog Number: 7838
• Author: DeMarais, B. D., et al.
• Title: Origin of
• USFW Year: 1992
• Copyright Material: YES

'. l1Q;j. lJeJ{qJaJ~ eA--~(.

.
I

Proc. Natl. Acad. Sci. USA
Vol. 89, pp. 2747-2751, April 1992
Evolution

1<6'3'1S

Origin of Gila seminuda (Teleostei: Cyprinidae) through
introgressive hybridization: Implications for evolution
and conservation

(speciation! morphological variation! genetic variation! mtDNA)

BRUCE D. DEMARAIS*, THOMAS E. DOWLlNG*, MICHAEL E. DOUGLAS*, W. L. MINCKLEY*, AND
PAUL C. MARSHt

-Department of Zoology, Arizona State University, Tempe, AZ 85287-1501; and teenter for Environmental Studies, Arizona State University, Tempe, AZ
85287-3211

Communicated by Wyatt W. Anderson. December 23, /99/

ABSTRACT Morphological and genetic characters from
cyprinid fIShes of the genus Gila were examined to assess a
hypothesized hybrid origin of Gila seminuda from the Virgin
River, Arizona-Nevada-Utah. Tbe presumed parents, Gila
robusta robusta and Gila elegans, are clearly differentiated
from one another based on morphology, allozymes, and
mtDNA haplotypes, G. seminuda is morphologically interme-
diate and polymorphic at allozyme loci diagnostic for the
parental species. Restriction endonuclease analysis of mtDNA
showed G. seminuda nearly identical to G. elegans. These
results support an origin of the bisexual taxon G. seminuda
througMntrogressivehybridization. Tbe Gila population in the
Moapa River, Nevada, also appears to be of hybrid origin and
is considered a distinctive population of G. seminuda. inter-
specific hybridization is potentially an important mode of
evolution among western North American fishes, and valid
species of hybrid origin may exist in other groups as wen.
Consideration of this mode of evolution argues for the need to
conserve entire species colilplexes~

Interspecific hybridization is potentially an important mode
of evolution in plants and animals. Botanists have long
recognized its role in producing new taxonomic entities and
in incorporating genetic variation into existing taxa (1, 2).
Zoologists have shown less interest in this phenomenon,
primarily because introgressive hybridization seems compar-
atively uncommon in animals, and hybrid taxa are rare.
Whether or not hybrid speciation plays a significant role in
animal evolution remains an open question. Most animal
evolutionists believe it does not (3). However, we argue in
our discussion that hybridization may have played an impor-
tant role in the evolution of fishes in western North America.
Among vertebrates, fishes exhibit the greatest number of
.proposed hybrid taxa. Yet, most ar~ unisexuals (reviewed in
ref. 4), although some bisexual forms have been proposed
(5-9). :MorphologiCal interniediacy tietween two ~presun1ed
parental species fostered the latter suggestions. However,
when allozymes and/or mtDNA were evaluated, most puta-
tive examples have been either strongly questioned or refuted
(e.g., see refs. 10-12). Clearly, morphological intermediacy
alone is insufficient to adequately test hypotheses of hybrid
origin.
Morphology has nevertheless remained the yardstick with
which to measure species, and,' in fact, most species are
described on a traditional morphological basis. Yet, mor-
phology presents difficulties in separating genetic and non-
genetic components, For example, cyprinid fishes of the
genus Gila that inhabit the Colorado River basin of the

The publication costs of this article were defrayed in part by page charge
payment, This article must therefore be hereby marked "advertisement"
in accordance with 18 U.S.C, U734 solely to indicate this fact.

southwestern United States display a complicated array of
phenotypes, certain ones of which seem correlated with
differences in habitat (13, 14), These phenotypes have long
puzzled ichthyologists. Enlarged fins, streamlined bodies,
shallow caudal peduncles, and high numbers of reduced,
embedded scales are typical of taxa in large, turbulent rivers.
More than one species may occur sympatrically in the largest
streams, where macrohabitats are sharply demarcated. Small
streams usually support a single phenotype, with small fins,
thick body, deep caudal peduncle, and low numbers of large,
overlapping scales. Moderate-sized rivers may be occupied
by fish of intermediate morphology.
Historically, most hypotheses to explain this habitat-
related pattern of morphological variation in Gila have
stressed local ecotypic/ecophenotypic responses to specific
environments (15, 16). More recently, hybridization between
distinct species has been forwarded to explain morphologi-
cally intermediate populations (17). Smith et al. (18) hypoth"
esized that selective forces strong enough to maintain sepa-
rate taxa in large rivers were insufficient to do so in moderate-
sized streams. In the latter, selection was thought to shape
phenotypes to resemble mosaics of attributes from large-river
taxa, perhaps assisted by genetic variation incorporated via
hybridization from several distinct morphs.
Smith et al. (18) suggested that the nominal Gila seminuda
Cope and Yarrow (19) from the Virgin River, Nevada-
Arizona-Utah (long considered Gila robusta seminuda),
originated through hybridization between G. robusta robusta
and Gila elegans. G, seminuda is the only species of its genus
ever collected in the Virgin River, a moderate-sized tributary
of the much larger Colorado River, where G, r. robusta, G,
elegans, and Gila cypha occur sympatrically.
We report results of a test in which the putative hybrid
origin of G. seminuda was examined by using morphological
and genetic characters. Taken together, our data provide
. strong support for the origin of G, seminuda through intro-
gressive hybridization involving G, robusta and G, elegans,

MATERIALS AND METHODS

Preserved and frozen materials were obtained for G. elegans,
G, seminuda, G. r, robusta (three populations), and a stock
from the Moapa River, Nevada, previously identified as G,
robusta (hereafter referred to as MRN). Sample sizes used for
morphologic, allozymic, and mtDNA analyses, respectively,
are as follows: G. elegans, 23, 20, and 4; G. seminuda, 24,17,
and 4; G. r. robusta (Bill Williams River basin, Arizona), 26,
20, and 3; G. r. robusta (Salt River basin, Arizona), 26, 20,
and 3; G. r. robusta (Verde River, Arizona), 18, 19, and 4;

Abbreviations: MRN, Gila robusta stock from Moapa River, Ne-
vada; PC, principal component.