MASTERS IN AQUACULTURE AND FISHERIES

1
Part 5 Genetics and Fisheries Management

• Genetic variation in fish stocks
• Use of molecular tools
• Estimation of effective population size and
  population dynamics' parameters

2
1 - Overview of the applications of population
genetics to marine resources management
Fisheries Management the application of
scientific knowledge to the problems of providing
the optimum yield of commercial fisheries
products or angling pleasure (Everhart Youngs
1981)
3
Historical Perspective

• Vast majority of fish stock harvested through
  fisheries is wild a conservation problem
• Until recently there was very limited application
  of basic genetics principles to fisheries
  management
• Field dominated by taxonomists that care little
  about differences between individuals
• Fish are more difficult to observe and so the
  study of the relationships between phenotype and
  genotype is more difficult
• Fish are the last major food source captured
  from wild stocks, with hard to define boundaries
• The amounts of phenotypic variation found in fish
  are very wide when compared with other food
  source animals
• First genetic results often contradicted
  ecological and ethological understanding of
  species
• So, the long-term perspective that genetic
  analysis provides has been missing from Fisheries
  Management theory
• The result is that many fisheries stocks were the
  victim of over-exploitation

4
Genetic and Phenotypic Variation in Fish Species
5
Heritability in Fish Species
6
Genetic Divergence Between Populations

• Classic problem in Fisheries Management
• FROM the identification of a fishing stock
• BECOMES the identification of genetically
  meaningful management units
• Many ecological and behavioral differences are
  due to environmental differences
  (freshwater/seawater, different spawning times
  and places, etc.)
• A considerable part of the difference between
  populations is also due to environmental
  differences - similar environmental conditions
  mold variation within populations and so the
  differences between populations tend to be due to
  differences in the environment of each one The
  role of genetics in the differentiation has been
  unclear until recently
• Markers not available until the mid 70s

7
First Biochemical Studies

• Identification of previously unrecognized
  systematic groups
• Population units of Pacific herring in the
  Alaskan peninsula (Grant Utter, 1984)
• Reproductively isolated sympatric populations of
  brown trout (Ferguson Mason, 1981)
• Identification of new species of rockfish (Seeb,
  1986)
• Inconsistencies with previous assumptions of
  genetic divergence
• Conspecificity of anadromous and landlocked forms
  of char in North America (Kornfield et al., 1981)
• No apparent genetic divergence between Fall and
  Spring spawning Atlantic herring (Ryman et al.,
  1984)
• Conspecificity (and local random breeding) of
  distinct morphological types of Ilyodon
  previously considered separate species (Turne
  Grosse, 1980)
• Conspecificity of sympatric but trofically
  specialized forms of Mexican cichlids (Kornfield
  et al., 1982)

8
Management Goals

• Conservation of genetic variation between and
  within natural populations
• Maintenance of the genetic characteristics of
  stocks that are artificially propagated in
  hatcheries (e.g. Pacific salmon species in the
  Northwest USA)
• Stock enhancement
• Selective breeding for production traits

9
The Role of Genetics in Fisheries Management

• Differential Harvesting Among Populations
• Differential Harvesting Within Populations
• Hatchery Populations
• Release of Hatchery Fish

10
2 - Basic concepts in population and molecular
genetics
11
Gene Frequencies and Hardy-Weinberg Equilibrium

• Simplified model for character determination
• Applicable to simple traits, such as blood groups
  or disease resistance
• We will assume diploidy

12
Phenotypes and Genotypes

• Phenotypes are the appearances of characters
  (visible/measurable)
• Genotypes are the genetic compositions that cause
  the phenotypes
• Often one phenotype is the product of many genes
  For example FCR depends on digestive enzymes in
  the gut, on hundreds of metabolic enzymes in
  cells, etc., in total it is estimated that
  hundreds or thousands of genes will determine FCR
  (a phenotype we can measure)

13
Recessive, Dominant and Codominant Alleles

• In diploid organisms each gene (locus) is present
  in two copies (alleles)
• If both are equal we say the locus is homozygous,
  otherwise it is termed heterozygous
• Recessive alleles do not express themselves in
  the presence of a dominant one. Difficult to
  detect the problem with disease carriers
• Codominance means that to a degree the phenotype
  of the heterozygote is the product of the
  expression of both alleles

14
Describing the Genetic Make-up of a
PopulationGenotype Frequencies

• When we know the relationship between phenotypes
  and genotypes we can estimate the genotype
  frequencies in a population
• E.g. blood group typing, eye color in fruit
  flies, etc.

Blood group Blood group Blood group Number of individuals
M MN N Number of individuals
Frequency () Greenland 83.5 15.6 0.9 569
Frequency () Iceland 31.2 51.5 17.3 747
Mourant(1954) in Falconer 1989
15
Describing the Genetic Make-up of a
PopulationGene Frequencies
Gene Gene
M N
Frequency () Greenland 91.3 8.7
Frequency () Iceland 57.0 43.0
16
MN blood Group Genotypes
17
Hardy-Weinberg Equilibrium

• Assumes
• Large population
• Random mating
• No selection
• No migration
• No mutation
• Predicts
• Stable gene frequencies from generation to
  generation
• Simple relationship between gene frequencies
  (allele frequencies) and genotype frequencies.

18
H-W Equilibrium
Genes in parents Genes in parents Genes in progeny Genes in progeny Genes in progeny
A1 A2 A1A1 A1A2 A2A2
Frequencies p q p2 2pq q2
19
Applications of the H-W Law

• Determination of the gene frequency of a
  recessive allele
• Frequency of carriers
• Test of H-W equilibrium

20
Changes in gene frequency

• Random drift
• Migration
• Mutation
• Selection
• Assortative Mating
• Computer simulation - PopG

21
Assortative Mating

• When mated pairs are of the same phenotype more
  often than expected by chance (common in humans,
  e.g. stature, intelligence, etc.)
• The opposite is called disassortative mating
  (common in plants with self-sterility systems)
• Increases frequency of homozygotes (although
  most characters are multiple loci coded)

22
Migration

• Thus teh rate of change depends on
• Immigration rate
• Difference of gene frequencies between immigrants
  and natives

q1 m qm (1 m) q0 m (qm q0) q0 ?q
q1 q0 m (qm q0)
23
Mutation

• Non-recurrent mutation low chances of mutant
  allele survival
• Recurrent mutation more relevant
• If u is mutation rate from A1 to A2
• and v is mutation rate for reciprocal mutation
  (A2 to A1)
• and p0 and q0 are frequencies of A1 and A2
• then
• The change in gene frequency in one generation is
• ?q up0 vq0
• And at equilibrium
• q u / (u v)

24
Selection

• Occurs when some genotypes are more fit than
  others
• Degrees of dominance with respect to fitness
• No dominance
• Partial dominance
• Complete dominance
• Overdominance.

25
Changes in Gene Frequency Under Selection
If s is the selective advantage of the A1
dominant allele (p), then the frequency of A2
after one generation (q1) is
So, the change of gene frequency from one
generation of selection ?q q1 q is
and substituting p (1 q)
26
Polymorphism Possible Causes

• Heterozygote advantage
• Frequency-dependent selection
• Heterogeneous environment
• Transition stages in evolution (due to
  environment changes, for example)
• Neutral mutation (allele)
• Heterozygosity a measure of the amount of
  polymorphism
• Proportion of polymorphic loci
• Frequency of heterozygotes averaged over all loci
  tested average Heterozygosity

27
Small Populations

• Random drift
• Sampling effects
• Variance in the change of gene frequency
• Variance of gene frequency among lines

28
Random Drift and N
29
Inbreeding

• The mating together of individuals that are
  related to each other by ancestry.
• In a bisexual population the number of ancestors
  of a individual t generations ago is 2t
• So, in a small population the relatedness of
  individuals will be greater than in a large
  population

30
Measuring Inbreeding

• The coefficient of Inbreeding is F
• F1 is the inbreeding coefficient of generation 1
  and F2 that of generation 2

Probability that a gamete pairs with another of
different sort
Probability that a gamete pairs with another of
the same sort
31
General Inbreeding Coefficient for individuals in
generation t

New Inbreeding or the rate of inbreeding
32
Effective Population Size Ne

• Real-life populations are not ideal
• Structured
• Assortative / disassortative mating
• Breeding structure
• Etc.
• Ne is the number of individuals that in an ideal
  population would give rise to the calculated
  sample variance or rate of inbreeding.
• So, in an ideal population Ne1/(2 ?F)

33
Formulas for Ne

• Self-feritilization
• Sib-mating
• Diferent numbers of males and females
• Unequal numbers in successive generations
• Non-random distribution of family size

34
Complicating Matters

• In real-life populations populations are far from
  ideal.
• Small effective sizes, various mating structures,
  mutation, selection and migration, all act
  together to produce the phenotypes and genotypes
  we see or measure.

35
Main Tools Available

• Allozymes
• Mitochondrial DNA (mtDNA)
• Nuclear DNA (nDNA)
• Microsatellites and minisatellites

36
Allozymes

• Allozymes are electrophoretically distinguishable
  protein variants
• First used in fish stocks in late 60s

• DISADVANTAGES
• Needs relatively large amounts of tissue in order
  to yield enough proteins for visualization
• Many enzyme systems available (gt75) although for
  each study usually only a small fraction shows
  polymorphism
• Potentially subject to selection pressures

• ADVANTAGES
• Simple to use and applicable to all species (just
  needs a source of soluble proteins)
• Standard protocols that require only minor
  adjustments from species to species

37
(No Transcript)
38
Mitochondrial DNA

• First studies (80s) revealed high levels of
  sequence diversity
• Useful for inter and intra-specific analysis
• Occurs in multiple copies per cell (gt1000)
• Uniparental transmission, no recombination
• Transmitted via the maternal line useful in the
  analysis of sex-specific gene-flow patterns
• Evolves faster than coding regions of nDNA

39
Techniques to Analyze mtDNA

• RFLPs
• Restriction fragment length polymorphisms
• Uses restriction enzymes to cut mtDNA in pieces
  and the separates them using agarose or
  acrylamide gel electrophoresis
• PCR
• Polymerase Chain Reaction
• Amplification of the number of copies of a target
  sequence defined by two flanking primers
• DNA sequencing

40
Nuclear DNA

• RFLPs
• nDNA sequences
• Tandemly repeated DNA
• VNTRs -DNA fingerprints (minisatellite sequences)
• STRs - Microsatellites
• RAPDs
• Randomly amplified polymorphic DNA
• AFLPs
• Amplified fragment length polymorphisms
• QTLs
• Quantitative trait loci

41
Summary of the Common Genetic Markers Used for
Fisheries Management Studies
42
Information Provided by Molecular Tools

• Population structure
• Deviations from H-W Law
• Inbreeding, migration, selection, mating
  structure, demographic history, phylogeographic
  history, etc.
• Genealogical and phylogenetic relationships

43
Statistical Methods Used

• Allozymes, RFLPs, AFLPs, DNA sequences, mini and
  microsatellites all provide us with allele
  frequency data (with different levels of
  polymorphism)
• Most statistical analysis are based on the
  analysis of genetic variation and its partition
  into the various hierarchical levels of
  structure, from families to demes, to
  sub-populations and the population as a whole.
• Most are base in Wrights F-statistics (FST),
  adapted by Nei for molecular data (GST) and
  further adapted to newer markers.

44
Variation

• Usually measured by polymorphism (P, the
  proportion of polymorphic loci) and
  heterozygosity (H, the proportion of heterozygous
  individuals)
• If the population is in H-W equilibrium then H
  can be calculated as

Where NAa is the number of heterozygote
individuals in the sample and N is the sample
size and n is number of loci, including the
monomorphic ones
Where xi is the frequency of the ith allele
45
Genetic Distance

• It is a measure of the gene diversity between
  populations expressed as a function of genotype
  diversity
• According to Nei,
• For 2 populations X and Y, the probability of
  identity of 2 randomly chosen genes at a single
  locus (jk) is
• The probability of identity of a gene at the same
  locus in populations X and Y is
• The normalized identity between populations X and
  Y with respect to all loci is
• The Genetic Distance between populations is then

46
Inter-Population Diversity

• Determine jk for each population and then the
  gene identity within all populations (JS)
• The average gene diversity within populations is
• The gene identity for the total population is
• And the inter-population gene diversity is
• The coefficient of differentiation can be defined
  as

47
Demographic History

• Aims at uncovering past population bottlenecks or
  booms
• Uses DNA sequence data and plots histograms of
  pair-wise differences between sequences
• Needs estimate of molecular clock in order to
  time events

48
Genealogy Theory

• Uses genealogy data, mainly from DNA sequences
  but also from microsatellites, RFLPs etc., to
  reconstruct the genealogies of alleles.
• Can provide information about population
  structure as well as past demographic events

49
Analysis of Phylogeny

• Using different data types but preferably DNA
  sequences together with specific algorithms for
  grouping the data (maximum likelihood, parsimony,
  etc.), it estimates the relationships between
  samples or alleles
• Software
• PAUP Phylogeny Analysis Using Parsimony
• PHYLIP
• MacClade
• TreeView
• CAIC - Comparative Analysis of Independent
  Contrasts
• Etc (http//evolution.genetics.washington.edu/phy
  lip/software.html)

50
Some Applications
51
Conservation
52
Determining Family Relationships and Ne

• Herbinger, C.M., R.W. Doyle, C.T. Taggart, S.
  Lochmann, A.L. Brooker, J.M. Wright and D. Cook.
  1997. Family relationships and effective
  population size in a natural cohort of cod
  larvae. Can. J. Fish. Aquat. Sci. 54
  (Suppl-1)11-18. Abstract
• Sibship relationships within a naturally spawned
  cohort of Atlantic cod (Gadus morhua) larvae on
  the Western Bank of the Scotian Shelf were
  investigated by a likelihood ratio method that
  estimates relationships is among individuals
  using microsatellite (DNA fingerprint)
  information. We found no evidence of any temporal
  or spatial family structure among the larvae from
  seven different sample collections taken at
  sequential time intervals during a 21-d period of
  sampling the larval cohort. There was no evidence
  that larvae were more related within sample
  collections than across sample collections.
  Within each sample collection, there was no
  evidence of a family structure within or among
  the depths sampled. Similarly, there was no
  apparent change in the potential occurrence of
  sibship with time (successive sample
  collections), or in association with the passage
  of a storm during the sampling period. This
  cohort of cod larvae appears to have been a
  fairly homogeneous mixture of larvae that were
  not siblings and came from a large genetic pool.
  The minimum estimate of the inbreeding effective
  population size is 2800 individual spawners.

53
 Mixed Stock Management
To determine what impact the winter fishery that
takes place in the 3Pn and 3Ps regions has on
migratory groups (4T,4R,4S,4Vn and 4Vs) we are
looking at the genetic makeup of fish from 3Pn
and 3Ps areas during the mixed period and
comparing this to the genetic make up of the
stocks that are thought to migrate here. From
these comparisons it will be possible to
determine the level of exploitation of the
migratory stocks and to have some estimate of the
impact the winter fishery has on the migratory
stocks and how this will effect their ability of
these stocks to recover. The graphs below
illustrates two possible mixing profiles as they
may exist. In Figure A the outcome of the fishery
would be a harvest which would be mostly focused
on fish which come from the inner Gulf region
(4T,4R and 4S) Figure B would be a fishery which
was mostly focused on resident populations from
3Pn and 3Ps. With either case, consideration of
the winter fishery impact must be taken and
adjustments made as to how the affected stocks
are managed.
54
Assigning Individuals to Populations- Forensics

• Enormous expense is currently incurred by
  enforcement officers while patrolling the fishing
  grounds with boats and aircraft. This method of
  enforcement is inefficient in terms of time,
  money, and the amount of catch actually
  monitored. With MGPL's DNA Fingerprinting
  capabilities, catch restrictions can be enforced
  at the dock instead of at sea on the basis of
  forensic analysis of the catch.
• The genetic information carried by the animals
  themselves will identify the stock from which
  they were taken.

55
Estimating changes in population size

• Use of coalescent theory and of pair wise
  sequence comparisons in order to study and date
  demographic events
• Use of gene diversity estimates in order to
  detect past bottlenecks in populations
• Etc.

56
Estimating gene flow among populations
57
Software
58
Brief summary of computer programs and data
analysis approaches
http//www.cf.adfg.state.ak.us/geninfo/research/ge
netics/software/anlink.php
LAMARC - Likelihood Analysis with Metropolis
Algorithm using Random Coalescence
Lamarc is a program for doing Likelihood Analysis
with Metropolis Algorithm using Random
Coalescence. Lamarc estimates effective
population sizes, population exponential growth
rates, a recombination rate, and past migration
rates for one to n populations assuming a
migration matrix model with asymmetric migration
rates and different subpopulation sizes. This
version can use DNA or RNA sequence data, SNPs,
microsatellites, or electrophoretic data. The
program can produce estimates of recombination
rate, migration rates between each population
pair, population sizes (assuming constant
mutation rates among loci), population
exponential growth rates, profile likelihood
tables, and percentiles.If you know that there is
no recombination in your data (for example in
mtDNA) you might look also at the other programs
Fluctuate or Migrate.
Whichrun 4.1 A computer program for population assignment of individuals based on multilocus genotype data. Microsatellite DNA provides essentially limitless, highly varied information within species.  That this provides a means for distinguishing not only among populations but also individuals has not escaped current theoretic interest (Smouse and Chevillon 1998, Waser and Strobeck 1998).  Here, we present a C computer program named WHICHRUN that uses multilocus genotypic data to allocate individuals to their most likely source population.
Genetic Mixture Analysis (GMA) Software for
estimating the stock proportions within mixed
stock fisheries
BOTTLENECK version 1.2.02  (16.II.1999)
Bottleneck is a program for detecting recent
effective population size reductions from allele
data frequencies.
59
Structure The program structure is a free
software package for using multi-locus genotype
data to investigate population structure. Its
uses include inferring the presence of distinct
populations, assigning individuals to
populations, studying hybrid zones, identifying
migrants and admixed individuals, and estimating
population allele frequencies in situations where
many individuals are migrants or admixed. It can
be applied to most of the commonly-used genetic
markers, including microsatellites, RFLPs and
SNPs. This method was described in an article by
Pritchard, Stephens Donnelly (2000). Extensions
to the method were published by Falush, Stephens
and Pritchard (2003). An interesting example from
the original paper is shown here.
                                                  
                                                  
                                       
- RST Calc A program to calculate unbiased
estimates of Slatkin's RST and Goldstein et
al's (delta-mu)2 distance for microsatellite data
GENEPOP is a population genetics software package
originally designed by Michel Raymond
(Raymond_at_isem.univ-montp2.fr) and Francois
Rousset (Rousset_at_isem.univ-montp2.fr), at the
Laboratiore de Genetique et Environment,
Montpellier, France.
60
Presentations

• Duration 15, groups of 2-3 (all should
  participate)
• When Friday 17th, 14.00 h
• What Results of selection programs in
• Salmon
• Trout
• Nile Tilapia
• Catfish
• Oyster
• Shrimps
• Structure
• The species and its farming
• Important traits
• Heritabilities
• Structure of selection program
• Performance (response) of program