9698

• UCREFRP Catalog Number: 9698
• Author: Borley, K., and M.M. White.
• Title: Mitochondrial DNA variation in the endangered Colorado pikeminnow
• USFW Year: 2006.
• USFW - Doc Type: a comparison among hatchery stocks and historic specimens.
• Copyright Material: YES

~

DNA VARIATION IN COLORADO PIKE MINNOW

ltion observed
ltchery brood-
ill number of
tic diversity is
,ulations when
II (Vrijenhoek
JW has under-
size and range
decline would
1.
~gests that low
m decline and
ninnow in the
en subjected to
00 years ago
ght is thought
faunas, whose
ies (Matthews
hypothesis that
~rm impacts on
concluded that
e flannelmouth
::olorado River
ought scenario.
lpid population
e end of the
: affected the
Jacted genetic
other Colorado
attributed low
Gila elegans to
10 evidence of
J G. cypha or

TABLE 3.-Primers used to amplify and sequence the ND-4
and ND-4-L gene regions in museum samples of Colorado
pikeminnow and three year-classes of Colorado pikeminnow
maintained at Dexter National Fish Hatchery and Technology
Center, New Mexico.

Primer name

Sequence

ArgBL
ND4LH
ND4LB(L)
NAP2(H)

5'-CAAGACCCTTGA TTTCGGCTCA-3'
5' -TCTGTCGCT AGGT A T AGA TT-3'
5~CAAAACCTTAATCTYCTACAATGCT~'
5' - TGGAGCTTCT AC G TG RGC TTT -3'

1989; Morizot et al. 2002). Three of these alleles were
found in wild Green River fish, and four were found in
the upper Colorado River. It is possible that other
basin-specific mtDNA haplotypes are not present in the
hatchery strains because of the small number of
founders. However, the observation that the hatchery
stock exhibiting variation was founded from fish
obtained from two different rivers (Green and Colo-
rado) suggests that there may be additional haplotypes
present in the wild.
Low mtDNA variation in DNFHTC Colorado pike-
minnow is attributable to an apparently low level of
such variation in the wild source populations, mode of
mtDNA inheritance, and small founding sizes of the
captive populations. Due to its maternal mode of
inheritance, mtDNA is more prone to bottlenecks than
nuclear DNA. The effective population size for
mtDNA is one-fourth of the autosomal effective
population size (Ballard and Whitlock 2004), making
mtDNA more susceptible to genetic drift. The
maximum number of mtDNA haplotypes that will be
present in a hatchery strain is equal to the. number of
females included in the original broodstock. The 81 YC
stock was created using three females (Morizot et al.
2002). The sex ratio of 74YC founders is unknown
(Morizot et al. 2002). The 91YC stock was founded
with five females (J. H. Williamson, DNFHTC,
personal communication).

lrone to genetic
;175). Previous
:minnow found
n and Morizot

'ado pikeminnow
logy Center, New
mia Academy of

Management Recommendations

Genetic management of the Colorado pikeminnow
should focus on preserving the existing levels of
variation. Morizot et al. (2002) suggested the mainte-
nance of three hatchery stocks derived from the
Yampa, Green, and Colorado rivers. They proposed
using hatchery-raised individuals derived from 74YC
for the Yampa River strain along with periodic
incorporation of wild fish. The Green and Colorado
River broodstocks would be established from 20
(Green River) to 50 (Colorado River) wild individuals.
Additional wild individuals would be introduced
periodically. Inclusion of wild individuals would

Iaplotype

A
A
A
A
A
A
A
A
A
A
A

919

increase the effective population size of the founder
populations (Hedrick et al. 2000).
An alternative would be to maintain a single hatchery
strain for basinwide stocking. The mtDNA data support
the presence of only a single ESU. There have been
numerous interbasin stockings (e.g., Yampa 74YC to
Colorado River, Yampa 74YC to White River, and
Green-Colorado 81YC to San Juan River) that would
tend to homogenize allele and haplotype frequencies
throughout the drainage, creating a single MU. Crist and
Ryden (2003) suggested that the size of the wild San
Juan River stock would preclude establishment of a San
Juan River broodstock and recommended using the
nearest available broodstock, probably either the Green
or Colorado River stock, to supplement. Combining the
broodstocks would have the advantage of more cost-
efficient maintenance and would maximize the available
variation. However, the absence of variation among the
stocks should not necessarily be used as justification for
homogenization of stocks (Crandall et al. 2000).
Carefully designed breeding schemes, establishment of
multiple populations in separate facilities, and augmen-
tation of the stock with wild-caught individuals would
be critical for the success of this strategy.
The choice of Jnanagement strategy depends on
whether the past. stockings have resulted in adults
displaying normal behavior and successful reproduc-
tion. There is site fidelity, both to the spawning areas
and to the postspawning home ranges (Tyus 1985,
1986; Osmundson et al. 1998). Age-O Colorado
pikeminnow are known to drift downstream and have
been observed at high densities below the confluence of
the Green and Colorado rivers (USFWS 2002), where
population mixing could occur (Tyus 1990). However,
Irving and Modde (2000) suggested that hatchery fish
do not exhibit typical wild fish behavior. If these fish
do not exhibit normal movements, it would complicate
management (Tyus 1991), alter existing patterns of
gene flow, and probably delay recovery of the Colorado
pikeminnow. Additional studies of the hatchery and
wild populations with more rapidly evolviTIg markers
(i.e., mtDNA control region and microsatellites) might
allow greater resolution of the patterns of variation and
clarify future hatchery management strategies.

Acknowledgments

We would like to thank the following individuals for
their help in obtaining specimens: G. Carmichael, J. H.
Williamson, C. Keeler, A. Fuller, and M. Ulibarri
(USFWS) and D. Morizot (University of Texas). This
manuscript was significantly improved by comments
from G. Carmichael, J. H. Williamson, J. Faber, and
two anonymous reviewers. Laboratory facilities and
expertise during the initial stages of this project were