BI 3063

1
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
The development of a "modern", science-based
fisheries biology early in the 1900s soon lead to
the recognition that unless the fishery
management was administred to the real
reproduction units, any management measure (like
quotae, mesh size regulations, season closures,
minimum fish size) would be unprecise and might
have unpredictable effects. Hence, much effort
was laid down in identifying those reproduction
units, or "stocks" as they were (and still are)
rather unprecisively called. The classical tools
(like tagging and recapture, monitoring of the
location of the fishing fleet, landing
statistics) were of course used for the purpose
but in addition, genetic traits assumed to be
population characteres were employed. (This was
in the days prior to modern laboratory techniques
like electrophoresis and DNA technology, and even
prior to much of the population genetic
theory). One methodology that was extensively
applied in the 1920ies and 1930ies was the use of
frequencies of meristic characteres i.e.
characteres like individual number of vertebrae,
fin rays, or gill rakers. It was observed that
many stocks differ in their average values for
such traits. Certainly, the variability of such
traits have a genetic basis. Hence, large studies
were devoted to plotting mean values for meristic
characters in different fish stocks.
2
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
A pioneer in this new science of meristics was
Johannes Schmidt (left) at the Carlsberg
Laboratory in Denmark, the same man who after a
massive work 1905 -1930 had established the early
life history of the European and American
freshwater eels (picture to the right).
Armed with the tools of meristics, Schmidt and
contemporaries explored the stock structure of
many of our commercial fish species, e.g. the
atlantic herring and the atlantic cod. Schmidt's
opinion of the genetic structure of the cod was
published in the 1930ies in form of tables of
mean number of vertebrae and fin rays in cod
samples from different parts of its distribution
area in the Atlantic, and maps connecting groups
with similar characteristics. The "stock map"
produced in this way showed a very tight
correlation with ocean temperatures, and it was
later shown experimentally (Tåning) that the
environmental temperature during embryogenesis
affected the number of vertebrae and fin rays to
develop in the individual. Hence the "stock map"
might reflect temperature ecotypes rather than
genetically distinct stocks or populations.
Meristics was soon discontinued as a key to
genetic population structure.
3
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08

• Types of markers in modern population genetics
  studies
• the various marker types have their advantages
  and disadvantages.
• Serum protein loci
• Isozymes
• mtDNA
• mini- and microsatellites (VNTR)
• nuclear and mtDNA sequences
• SNPs

4
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Genetic stock ID studies in the Atlantic cod The
first extensive use of electrophoresis and
protein genetics was published in Nature in 1961
by the danish population geneticist Knud Sick. He
used allele frequencies at a simple haemoglobin
polymorphisms (HbI) in cod (and whiting) to
demonstrate the existence of different
populations with very different allele
frequencies. His studies included the entire
north Atlantic ocean. Actually, the stock
structure indicated by the HbI locus was
remarkably similar to that resulting from the
previous meristic studies by Schmidt. After
having been used for several decades in cod
management, it was realised that allele
frequencies at the HbI locus was influenced by
environmental tempereratures, and therefore could
not be used as signs of reproductive isolation.
An attempt to use blood group techniques for
stock identification i cod was terminated quite
quickly. The genetic basis for the blood group
variation was never established. Thereafter,
studies using isozymes, mtDNA haplotypes,
microsatellites, nuclear DNA RFLP, and SNP have
succeeded each other but still, many questions
are unclear about the genetic population
structure of the Atlantic cod.
5
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
GSI Genetic Stock Identification studies in the
Atlantic cod, cont'd Various types of genetic
markers have given very different estimates of
the general level of ifferentiation in cod. While
markers like haemoglobins, one microsatellite and
the PanI (a nuclear RFLP marker) indicate
substantial differentiation and some very abrupt
changes in allele frequencies over short
geographical distances, serum proteins, isozymes,
most microsatellie, most nuclear RFLPs as well as
"silent" substitutions of mt Cytb sequences do
not indiate substantial differentiation at at.
Rather, the latter support an "isolation by
distance" model of differentiation. The figures
on the next slides show results from a
distribution-wide study of genetic
differentiation in cod based on 13 polymorphic
isozymes loci (Mork et al. 1985). The results are
quite similar to those obtained based on a set of
nuclear RFLPs. A trait that seems to be common
in many studies, is that the Baltic Sea cod stand
out as being the genetically most differentiated
among current stocks. Also, the western and
eastern Atlantic stocks appear to form subgroups
with some degree of genetic differentiation (only
at common population level though).
6
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Cod sampling sites (of Mork et al. 1985)
7
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
UPGMA dendrogram, and isolation by distance for
cod (Mork et al. 1985)
8
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Wrights Fst for cod compared to some other
species
9
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08

• GSI Genetic Stock Identification studies in the
  Atlantic herring
• Not being quite as controversial as that of the
  cod, the genetic population structure of the
  Atlantic herring show some traits which are not
  much disputed
• 1 There are numerous smaller stocks tied by
  their life history to local areas like fjords,
• coastal areas and brackish oceanic areas
  (White Sea, Baltic Sea).
• 2 The genetic differentiatiation across the
  Atlantic is actually smaller than what can
• often be found between two neighbouring
  Norwegian fjords.
• 3 Previous keys to population isolation, like
  being spring- or autumn spawners, are not
• valid. Spring- and autumn spawners may be
  found in one and the same local
• population. Of course, the realization of
  this fact led to a simplification of previous
• models of stock structure.
• 4 In general, the herring appears, to a much
  higher degree than many other fishes,
• to be characterised by a so-called
  metapopulation structure (local populations are
• transient they come and go depending on
  environmental conditions, general
• abundance etc).

10
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Statistical analyses A typical population
genetic study proceeds through three phases. The
first stage involves testing the validity of
using allele frequencies to describe the
genotypic variation within samples or
populations. These are genotypic analyses. The
second stage usually involves exploring data for
patterns. During this stage, scientists employ a
variety of taxonomic type analyses, which can
provide quantitative, objective descriptions of
the pattern(s) of differences among groups. The
third stage focuses on hypothesis-based analyses,
whereby one hypothesizes that genetic differences
are structured in a specific way, and then try to
quantify or statistically test for those
differences. Almost always, the structuring that
is suspected and can be tested is geographic.
11
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08

• Measures of genetic differences (the most
  commonly used)
• Genetic Identities and Distances (Masatoshi Nei
  1972)
• FST of Sewall Wright (or ? of Weir
  Cockerham), GST of M. Nei
• Cavalli-Sforza chord distance
• Hierarchical genetic analyses (Amova)

• Cluster analysis and dendrogram drawing
• UPGMA / WPGMA
• Various algorithms and tree analyses

Statistical tests The most basic are the
chi-squared tests for Hardy-Weinberg genotypic
proportions and RxC contingency table tests of
homogeneity.
12
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Stock management and Conservation Establishig a
Genetic Inventory is a necessary starting point
for a meaningful, genetically based management or
conservation program. Usually, this is best
achieved by performing initial studies that
include the entire distribution range of a
species. Depending on the findings in an
inventory study, the principles for management
can be layed down. Various fish species differ
greatly in how their total gene pools are
structured within and between poulations. It has
been convincingly showed that one of the most
important keys to structuring is the possibility
of a gene flow between populations within species
(next slides).
13
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
The relationship between the species' general
biology and the degree of differentiation between
its populations

• Several review studies have shown that in fishes,
  the degree of genetic structuring
• (i.e. the genetic differences between populations
  within species) is consistently
• depending on whether the actual fish species is
• Limnic
• Anadromous, or
• Marine
• It is very likely that this pattern is a result
  of the general biology of the species.
• For example, to which degree it tends to develope
  local adaptations, how well
• its biology allows for a gene flow between
  populations, and to which degree
• its way of life results in a genetic adaptation
  to local environmental factors.
• These are factors that obviously differ between
  limnic, anadromous and marine
• species (examples on next slides).

14
Gadus morhua
Cod (Gadus morhua). Found in the entire North
Atlantic. Very large populations with relatively
similar environmental conditions. Can undertake
extensive migrations, but do not show particulary
accurate homing. Pelagic egg- and larvae stages
which can last for several months. Thrives at
cold and temperate latitudes. Extensive genetic
investigations have shown limited genetic
differentiation. Salmon (Salmo salar).
Distributed over most of the North Atlantic.
Anadromous spawns in fresh water. Relatively
small river populations which can differ
considerably environmentally. Benthic egg and fry
stages. Thrives at cold and temperate latitudes.
Can undertake extensive migrations, and show
extremely accurate homing. Extensive genetic
studies have indicated a moderate level of
genetic differentiation. N. lamellosa is
distributed on the US West Coast. Relatively
small populations. Lives in the littoral, and
hence under very variable micro-geographic
habitat and milieu conditions. Not very mobile,
but shows a clear tendency for homing. Benthic
egg capsules where the larvae develop into a
"micro-individual" before hatching. No pelagic
stages. Thrives in cold and temperate latitudes.
Studies of genetic structure have shown a
substantial degree of genetic differentiation on
both a small and a large geographic scale.
Salmo salar
Nucella sp.
15
Distribution and migrations of cod, salmon and
Nucella sp.
North east Arctic Cod
Cod stocks and currents
Atlantic salmon
Nucella sp.
16
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
17
(No Transcript)
18
Genetic and biologic stock management
BI 3063 J. Mork H08
Hallerman Ch. 13 Genetic Stock Identification and
Risk Assessment
19
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
20
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08

• Hierarchical Genetic Analysis
• (often called Amova Analysis of Molecular
  Variance)
• The genetic structure of a species may include
  several hierarchical levels, e.g.
• Regions
• Areas within regions
• Drainages within areas
• Rivers within drainages
• Populations within rivers
• With hierarchical genetic analysis the amount of
  genetic differentiation connected with each level
  can be estimated. Often, but not always, analysis
  show that the largest differentiation is
  connected to the highest levels i.e. that
  genetic differences are larger between regions
  than between lower levels in the hierarchy. Since
  the hierarchy usually is geographical, this would
  support the idea of "isolation by distance".

21
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Hierarchical Genetic Analysis cont'd Among
population geneticists, such hierarchical
analysis is usually referred to as "Amova"
Analysis of molecular variance. This kind of
analysis is very computing-intensive, and is
usually performed with computers. The most
widely used software for such analysis is the
"Arlequin" package. The software is free for
download from the web. http//lgb.unige.ch/arle
quin/ (see next slide)
22
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
23
Genetic and biologic stock management
BI 3063 J. Mork H08
Hallerman Ch. 13 Genetic Stock Identification and
Risk Assessment
Mixed-Stock Fishery Analysis (MSA) Pacific
salmon species have posed special challenges to
management, because in the saltwater phase they
may co-occur in specific areas where they are
exploited jointly by the fishing fleet. In order
to manage the fishery so that no one river stock
is over-exploitated, the composition over rivers
stocks in the mixed-fishery must be
known. Genetic analyses called MSA
(Mixed-Stock-Analysis) have been developed for,
and employed in such situations. The principle
is to first create a detailed baseline of
knowledge of the genetic characteristics of the
potentially represented populations, by sampling
them in their "home river". The more loci
included, the better. Then, by sampling the
mixed-fishery and analysing each individual for
its multi-locus genotype, maximum-likely methods
(MLY) can be employed to sort each individual to
its most propable home population, and then
estimate the relative proportions of the various
river stocks in the mixed fisheries.
24
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
Mixed-Stock Fishery Analysis (MSA) cont'd While
MSA has worked well on Pacific salmon species,
its efficiency in Atlantic salmon is less
impressing. This is mainly due to the fact that
the Atlantic salmon is less genetically
structured than e.g. the chinook salmon. Hence,
the sorting of individuals to their "home
populations" is much less accurate, and
mis-sortings occur more frequently. In fact,
mis-sorting is the one most common problem with
MSA, in that small river stocks tend to be
over-represented in the mixed fisheries, and vice
versa.
25
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
There has been some discussion about what type of
statistical procedures and algoritms to use in
MSA. This discussion emerged with the
introduction of new type of genetic markers
i.e. with the transition from isozyme markers to
microsatellites which have a higher "resolution"
because of higher mutation rates. While the
first approaches used Maximum Likelihood
estimates (in the "isozyme age" of the 1990ies),
there is evidence that Bayesian methods perform
better for microsatellites. The next slide show
a treatment of this topic downloaded from the
web.
26
Genetic and biologic stock management
BI 3063 J. Mork H08
Hallerman Ch. 13 Genetic Stock Identification and
Risk Assessment
27
Genetic and biologic stock management
BI 3063 J. Mork H08
Hallerman Ch. 13 Genetic Stock Identification and
Risk Assessment
Software for mixed-stock analysis is available
for fre download from the Internet. For those who
masters R, an application of MSA can be found at
this site http//cran.r-project.org/web/packages
/mixstock/vignettes/mixstock.pdf Mixed stock
analysis in R getting started with the mixstock
package Ben Bolker July 8, 2008
28
Genetic and biologic stock management
BI 3063 J. Mork H08
Hallerman Ch. 13 Genetic Stock Identification and
Risk Assessment
Box 13.3
29
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
30
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
31
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08
32
Genetic and biologic stock management Hallerman
Ch. 13 Genetic Stock Identification and Risk
Assessment
BI 3063 J. Mork H08