9411

• UCREFRP Catalog Number: 9411
• Author: Williamson, J. H., D. C. Morizot and G. J. Carmichael.
• Title: Biochemical Genetics of Endangered Colorado Pikeminoow from the Green, Yampa, Colorado, and San Juan Rivers.
• USFW Year: 1998.
• USFW - Doc Type: \
• Copyright Material: NO

9
72 C. For routine screening, PCR products were fractionated on 1.5% agarose gels and visualized by
staining with ethidium bromide. Appropriately sized bands were plugged with Pasteur pipets and used as
templates for a second round of 30 PCR cycles which differed only in an increased annealing temper were
purchased from GIBCOBRL (Grand Island, NY), New England BioLabs (Beverly, MA), Promega ature
of 52 C. To more accurately estimate the size of small fragments, representative samples were fractionated
on 3.5% non-denaturing acrylamide gels (Sambrook et al., 1989).
DNA from the second PCR amplifications was digested with restriction enzymes which
(Madison, WI), or U. S. Biochemicals (Cleveland, OH). Reactions were buffered and incubated according
to manufacturers' instructions. Restriction enzymes used were Bbs I, Bsu 36 1, Dde 1, Hae III, Hha 1,
Hinc II, Hind 111, Hinf 1, Msp 1, Nla 111, Rsa 1, Sau 96 1, Tfi 1, and Taq I (Appendix 2; Tables 1 and
2). Products were visualized with ethidium bromide on agarose gels.
Findings and Discussion
The purpose of this work is to describe analyses of biochemical genetic polymorphisms and to
report on a PCR-based assay of potential Mendelian polymorphisms from anonymous single-copy nuclear
DNA techniques. The intent had been to examine or repeat full analysis of 89 loci on samples of YOY.fish
from each designated location, but several circumstances (see below) prevented completion of these
analyses.
Not fully sampling all populations for all loci violates a basic statistical assumption of the
BIOSYS-1 population genetic analyses. The decision to analyze only polymorphic loci was due to limited
funding. Also, field notes and laboratory records indicate that samples or bodies of YOY Colorado
pikeminnow from the upper Green River and lower Green River were incomplete. Fish'heads' were
removed to obtain otoliths. This removal of otoliths or heads containing otoliths inadvertently resulted in
viscera removal, particularly liver, leaving muscle and fin only. The result was loss of resolution of liver
specific loci and brain/eye specific loci, and loss of this resolution precluded potential examination of fish
for many loci that would have allowed calculation of population genetic parameters for indigenous fish
1 populations.
All biologists involved with the Colorado pikeminnow genetics study design and collections agreed
that adult and juvenile fish not be killed to obtain necropsy samples needed for some loci. Statistical
assumptions necessary to compute the population genetic analyses were not met. These unmet statistical
assumptions include but are not limited to the following: 1) Lack of randomness,z Because of budget
limitations and the decision by the Biology Committee to not use necropsy tissue samples on adult
Colorado pikeminnow, we had to limit analyses on most adult populations to biochemical loci that had been
determined to be polymorphic and that could be taken by biopsy. This limitation meant that the samples
were biased toward variation which would greatly inflate values of population structure or variability.
These values were predetermined to be biased and not random, by budgetary and study design; 2) Highly
varied sample size. Sample sizes for various locations were inconsistent because of tissue losses and
depauperate populations; 3) Lack of independence. Some collection sites overlapped for some YOY,
juvenile, and adults sampled. This inadvertent overlap makes comparison between sites moot because of
the need for independent samples. Interpretation of data from juveniles was problematic or ambiguous
because of potential hatchery or wild origins; 4) Lack of population definition. Samples were collected in
multiple years for budgetary and logistical reasons. Analyses of relatedness of YOY to adults on spawning
sites was lost when YOY were collected in different years from purported adult spawners; 5) Lack of
statistical power. Population statistical analyses were designed on data from balanced populations. This
makes population parameters powerful for detecting change from healthy populations. However, in our
opinion the use of these tools on known unbalanced, depauperate, populations does little to define
population structure. We are convinced of the robustness of the tests if statistical assumptions were met,
which they were not.
The BIOSYS-1 analyses were thus allowed for our use as limited tools only and are not completely
valid as analyses to determine basic population genetic characters such as average heterozygosities,