General enquiries on this form should be made to:
Defra, Science Directorate, Management Support and Finance Team,
Telephone No. 020 7238 1612
E-mail:
[email protected]
SID 5



Research Project Final Report
Note
In line with the Freedom of Information
Act 2000, Defra aims to place the results
of its completed research projects in the
public domain wherever possible. The
SID 5 (Research Project Final Report) is
designed to capture the information on
the results and outputs of Defra-funded
research in a format that is easily
publishable through the Defra website. A
SID 5 must be completed for all projects.
1.
Defra Project code
2.
Project title
This form is in Word format and the
boxes may be expanded or reduced, as
appropriate.
3.
ACCESS TO INFORMATION
The information collected on this form will
be stored electronically and may be sent
to any part of Defra, or to individual
researchers or organisations outside
Defra for the purposes of reviewing the
project. Defra may also disclose the
information to any outside organisation
acting as an agent authorised by Defra to
process final research reports on its
behalf. Defra intends to publish this form
on its website, unless there are strong
reasons not to, which fully comply with
exemptions under the Environmental
Information Regulations or the Freedom
of Information Act 2000.
Defra may be required to release
information, including personal data and
commercial information, on request under
the Environmental Information
Regulations or the Freedom of
Information Act 2000. However, Defra will
not permit any unwarranted breach of
confidentiality or act in contravention of
its obligations under the Data Protection
Act 1998. Defra or its appointed agents
may use the name, address or other
details on your form to contact you in
connection with occasional customer
research aimed at improving the
processes through which Defra works
with its contractors.
SID 5 (Rev. 3/06)
Project identification
MF0230
Spatial and temporal genetic structuring of Edible crab
populations
Contractor
organisation(s)
Royal Holloway University of London
School of Biological Sciences
Egham
Surrey
TW20 0EX
54. Total Defra project costs
(agreed fixed price)
5. Project:
Page 1 of 21
£
177025
start date ................
01 January 2005
end date .................
31 March 2008
6. It is Defra’s intention to publish this form.
Please confirm your agreement to do so. ................................................................................... YES
NO
(a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They
should be written in a clear and concise manner and represent a full account of the research project
which someone not closely associated with the project can follow.
Defra recognises that in a small minority of cases there may be information, such as intellectual property
or commercially confidential data, used in or generated by the research project, which should not be
disclosed. In these cases, such information should be detailed in a separate annex (not to be published)
so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report
without including references to any sensitive or confidential data, the information should be included and
section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No"
answer.
In all cases, reasons for withholding information must be fully in line with exemptions under the
Environmental Information Regulations or the Freedom of Information Act 2000.
(b) If you have answered NO, please explain why the Final report should not be released into public domain
Executive Summary
7.
The executive summary must not exceed 2 sides in total of A4 and should be understandable to the
intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together
with any other significant events and options for new work.
The objective of the present project was to define genetic variation among populations of brown crab, C.
pagurus, sampled from multiple locations and times, and to interpret these patterns in terms of
population/stock structuring in relation to habitat discontinuities and species biology. The primary focus
was the English Channel crab population, however samples from around the British Isles were included to
provide a UK-wide perspective. Large sample sizes (58-162 individuals per site) of adults from 32 sites
were screened for genotypic variation at 8 microsatellite DNA loci and sequence variation in one
mitochondrial DNA gene region. For 7 locations repeat samples collected 5-7 years apart were tested for
temporal variation in gene frequencies. Three samples of C. pagurus larvae collected from the western
and eastern Channel were screened for 4 microsatellite loci and used to test genetic assignment methods
for their ability to assign larvae back to their population of origin. To test the genetic mating system in
brown crab, broods of offspring from 18 females were DNA fingerprinted to determine the presence of
multiple or single paternity. A number of technical developments were made during the project:
optimisation of a new Chelex-based procedure for DNA extraction from single larvae or eggs; to achieve
statistically robust tests for differentiation 12 new C.pagurus-specific microsatellite loci were isolated and
new C.pagurus-specific primers for the amplification of three mtDNA genes were designed.
Simulation analysis indicated that the microsatellite markers used were capable of detecting very low
levels of population structure. Across the area sampled the microsatellite dataset revealed consistently
high levels of genetic diversity within populations (i.e. no indication of large genetic population size
declines) and low levels of genetic differentiation between sites. Global tests for differentiation indicate
significant genetic structuring of the C. pagurus population around UK (and near neighbour) coasts. Within
regions there is evidence for genetic differences between samples, but these differences are small and in
many cases probably do not indicate significant sub-structuring of populations (at least in the long term):
this was particularly the case for comparisons within the North Sea coast of the UK, although there was
indication of more genetic differentiation among samples in the Channel.
Within the region of focus, the English Channel, there was evidence of low but significant genetic
differentiation among samples. Within the Channel the most extensive and significant differentiation
results were displayed by two samples (Newlyn inshore and Brittany inshore) that were both from
locations (bays) with the potential for restricted exchange with neighbouring water bodies. If these two
“inshore” samples are removed from the Channel dataset then the remaining samples display fewer
(proportionally) significant differentiation results, almost all of which involve very low levels of genetic
differences and become non-significant if significance levels are adjusted for multiple tests (Bonferroni
correction), and which display no obvious association with geography or hydrographical features. Within
the North Sea samples Sweden stands out as showing significant differentiation from all UK samples
(although again this may be due to the sample coming from an enclosed water body – Gulmarsfjord), but
SID 5 (Rev. 3/06)
Page 2 of 21
there is no evidence for significant and consistent differentiation among UK sites. Samples collected from
the same area but 5-7 years apart (testing temporal stability of gene frequencies within populations)
indicated no significant differences (i.e. temporal stability) for all sites tested in the North Sea UK coast,
but significant differences (i.e. temporal instability) for some sites in the Channel. Tests of genetic
assignment methodologies confirmed that levels of genetic differentiation among locations was too low to
allow meaningful assignment of larvae or adults to populations of origin.
MtDNA analysis similarly found no evidence for significant isolation-by-distance patterns, but did reveal
higher levels of population structure overall. This pattern of more pronounced population structure for
mtDNA than for microsatellites was more obvious among samples from the North Sea than the Channel.
Hierarchical AMOVA of mtDNA data revealed a strong regional separation between samples from the
Channel and North Sea coasts of the UK. Haplotype frequencies are similar among samples from the
region of the Channel, Celtic Sea, southern Irish Sea and SW coast of Ireland, whereas this grouping is
distinctly different (lower frequencies of haplotype B) from samples spanning the region from the north
coast of Ireland into the western and eastern coasts of the North Sea (although there is much more
variation among North Sea sites).
Overall we would conclude that there is evidence for some genetic patchiness within the C.pagurus
population around the UK, but that genetic differences are small and show no geographical association
within regions (i.e. at local scales), genetic patchiness is less pronounced in the nuclear genome
(microsatellites) than the mtDNA genome, and patchiness is less pronounced (both spatially and
temporally, certainly for nuclear genes) among North Sea populations than Channel populations.
Data from both nuclear and mtDNA markers suggest a degree of genetic differentiation among regional
populations such that the Channel, and western and eastern North Sea regions should be considered as
distinct populations. Crab populations within regions effectively are single genetic populations, with levels
of interbreeding high enough to prevent significant genetic differentiation, certainly on longer timescales
(10s of years). However, low levels of genetic patchiness (particularly within the Channel region) indicate
that interbreeding (i.e. successful adult and larval dispersal) may not be extensive in the short term, yearon-year. Samples from areas with potentially semi-isolated hydrographic regimes (bays) exhibited the
most extensive genetic differentiation, indicating that larval dispersal is likely to be the most important
factor in gene flow. The greater large-scale differentiation observed for mtDNA than nuclear
(microsatellite) markers may result from a combination of historical effects of post-glacial colonisation by
this species of the area studied and male-biased gene flow. Male-biased gene flow could occur in C.
pagurus due to: (a) larval drift back to maternal natal areas; and/or (b) return migrations of females to natal
areas. In both cases females may migrate and mate with males from other areas but due to adult and/or
larval philopatry mtDNA (female) gene flow may be reduced compared to nuclear gene flow, as the female
genes are returned to the same site they originated but male genes are transferred between sites. Female
contranatent (against the current) migrations in a number of areas have been suggested, which is
compatible with the larval drift model. Also, recent data suggesting return migrations of females would
support the female philopatry model. We consider that low level nuclear genetic patchiness is most likely
to result from random and localised effects of skewed reproductive success among breeding adults
generating “sweepstakes” larval recruitment patterns. Genetic patchiness may be more prevalent within
the Channel population (compared to the North Sea) due to the more complex hydrography, generating
potential for localised effects of eddies and frontal systems to exacerbate spatio-temporal patchiness in
larval survival and recruitment.
Microsatellite DNA fingerprinting demonstrated that for all broods tested single paternity was confirmed. In
single paternity mating systems the number of females imposes a strict constraint on the number of males
that can breed, and thus the genetic population size.
At the outset of this project the aim was to describe stock structure in a temporal and/or spatial context
amenable to standard stock assessment protocols. The genetic data reveal a complex pattern of shallow
population structures among UK crab populations. The general pattern of high intra-sample genetic
diversity and low inter-sample genetic differentiation suggests that microgeographic management of the
resource is unnecessary: the Channel and North Sea coast crabs may each be regarded as single large
genetic populations. Genetic differentiation of the Newlyn and Brittany inshore samples, however, confirms
that there is potential for isolated population units which may require specific management. Future
population genetics studies should aim to include samples from any area potentially subject to isolating
hydrodynamic regimes. The greater genetic patchiness and temporal instability in Channel crabs, whilst
not indicating that this population is highly fragmented, does suggest that sub-populations within this
region may be largely self-recruiting and subject to short-term fluctuations in recruitment success, i.e.
localised over-depletion should be avoided. The UK North Sea coast crab population appears to be more
stable genetically, with more mixing of recruits between areas (though whether gene flow is bi-directional
or unidirectional along the coast is not known), although the mtDNA data again suggest some caution in
over-depletion of local populations. In light of what is known about the species biology the genetic data
suggest that females play a key role in shaping population structure/connectivity and are therefore of
paramount importance to population sustainability. Management efforts should focus on data collection
pertaining to females.
SID 5 (Rev. 3/06)
Page 3 of 21
Project Report to Defra
8.
As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with
details of the outputs of the research project for internal purposes; to meet the terms of the contract; and
to allow Defra to publish details of the outputs to meet Environmental Information Regulation or
Freedom of Information obligations. This short report to Defra does not preclude contractors from also
seeking to publish a full, formal scientific report/paper in an appropriate scientific or other
journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms.
The report to Defra should include:
 the scientific objectives as set out in the contract;
 the extent to which the objectives set out in the contract have been met;
 details of methods used and the results obtained, including statistical analysis (if appropriate);
 a discussion of the results and their reliability;
 the main implications of the findings;
 possible future work; and
 any action resulting from the research (e.g. IP, Knowledge Transfer).
1. INTRODUCTION
Project objectives (as stated in original proposal for MF0203):
1. To provide a detailed analysis of genetic population structuring of the edible crab (Cancer pagurus) from
geographically localised samples (e.g. within the English Channel), and to define the scale of local populations.
2. To assess relationships between spatial and temporal genetic differentiation of adult crab populations, and
spatial and temporal genetic patchiness in populations of larvae and juveniles.
3. To test assignment methods for genetic identification of population-of-origin of larvae within the water column.
4. To relate genetic structuring in crab populations to prevailing oceanographic, ecological and biological factors,
to attempt to identify environmental factors important in structuring and temporal dynamics.
5. To assess the implications of these results for stock structure of crabs in the Channel, and thus contribute to
stock assessments and management advice (for example on impact of aggregate dredging proposals on
spawning stocks).
Populations constitute interbreeding units with more or less autonomous dynamics and recruitment and are
frequently defined as harvest stocks in relation to fisheries management (Carvalho & Hauser 1994). In terrestrial
and freshwater environments, populations are often well delimited by conspicuous physical barriers to mixing and
interbreeding (e.g. Avise 2000). However, in the marine environment distinct populations are more difficult to
detect and for many marine species it is unclear to what degree distinct populations exist at all, or whether they
are organised into larger panmictic units (McQuinn 1997). This distinction is critical, in particular for exploited
species, as different populations may possess different genetic, physiological, behavioural or other characteristics
that may cause differences in life history traits such as fecundity and mortality rates and ultimately production and
abundance (Gold & Richardson 1998). As recruitment and sustainability may be properties specific to individual
populations failure to identify, and independently manage, distinct populations can lead to local overfishing and
ultimately to severe declines or stock collapse (Hutchings 2000; Knutsen et al., 2003).
Genetic markers represent powerful tools for examining population structure that in the marine
environment might otherwise be undetected due to difficulties in implementing standard ecological methods such
as mark-recapture or behavioural observation (Shaklee & Bentzen 1998). Furthermore, genetic markers provide a
unique ability to distinguish between non-reproductive dispersal and effective dispersal (i.e. those individuals that
survive and breed in the new population). Populations that are not linked by effective dispersal (i.e. exchanging
genes/gene flow) may accrue different gene frequencies. Therefore, by characterising the geographical
distribution of genetic variation population units can be identified. Recent developments of increasingly sensitive
genetic markers and associated statistical methods now permit hypothesis testing at the level of individuals as
well as populations. Such studies can be grouped into two broad categories: population assignments and
parentage assignments, the former involving the assignment of individuals to their populations of most likely
origin. Such estimates can be used to track the movements of individuals and more precisely quantify gene flow
among populations with obvious applicability to stock identification and discrimination (Abaunza et al., 2008).
Parentage assignments permit the characterisation of many aspects of breeding systems such as social
structure, mating patterns, kinship and quantification of reproductive success.
Many marine species have the capacity to disperse over vast geographical areas by passive drifting of
eggs and larvae and/or active migrations of juveniles/adults. These intrinsic characteristics of ‘classical’ marine
species (i.e. wide distributions, large population sizes, high fecundity, pelagic eggs and larvae) combined with the
SID 5 (Rev. 3/06)
Page 4 of 21
apparent lack of barriers imposed by the sea leads to the expectation of weak or no population sub-structuring
(Palumbi 1994; Ward et al., 1994). This expectation has often been supported by genetic studies indicating that
population structure in marine fish is best explained by a low structure high gene flow model (Ward et al., 1994;
Waples 1998). However, some recent studies have revealed population subdivision in marine species on limited
geographical scales, ranging from tens of kilometres (Knutsen et al., 2003; Taylor & Hellberg 2003) to a few
hundred kilometres (Ruzzante et al., 1998; Nielsen et al., 2001), and provide evidence that marine species may
exhibit more genetic structuring than predicted from dispersal potential. Adult/juvenile/larval life history
characteristics and behaviours, oceanographic features, habitat discontinuity, historical vicariance and selection
all might contribute, either singly or in combination, to the differentiation of populations (Hemmer-Hansen et al.,
2007). This emphasises the need for population structure to be empirically assessed.
The brown crab, Cancer pagurus L., is distributed continuously in shallow shelf waters of the NE Atlantic
Ocean from the Lofoten Islands (northern Norway) to Morocco. It is commercially exploited throughout its
geographical range with total landings of 46,280 t reported for 2005. The major fisheries are around the British
Isles and adjacent French coasts, and off Norway and Sweden. The most important crab fishery is in the English
Channel, which is fished primarily by vessels from England, Wales, France and the Channel Islands. It is worth an
estimated £11 million annually with over 70% currently landed into England and Wales. This economic
importance has prompted extensive research of C. pagurus biology and fisheries throughout its range with regard
to reproduction, ethology, growth and movement (reviewed in Bennett 1995). Data acquired from studies of C.
pagurus have been readily incorporated into the progressive design of management strategies (Bennett 1995).
Currently the fisheries are managed primarily by regional minimum landing sizes through both the European
Union and UK national legislation. However, available information has been difficult to interpret in terms of stock
structure relationships, knowledge of which is now regarded as a prerequisite for defining the scale at which
management is applied.
C. pagurus exhibits a number of characteristics that would be expected to promote a high degree of
population connectivity across large geographical areas, and consequently low levels of sub-structuring. These
characteristics include: (i) high fecundity, with 0.5-2.9 million eggs produced per female (Edwards 1979; Ungfors
2007); (ii) a prolonged pelagic larval phase lasting 2-3 months (Eaton et al., 2003); (iii) extensive adult migration,
particularly by mature females (reviewed in Bennett 1995). However, some indirect evidence suggests the
possibility of population structure at various geographic scales. For example crabs from the English Channel and
North Sea exhibit characteristic differences in growth, moult frequencies and female migration patterns (Bennett
1995). In addition, the region around the Thames estuary is predicted to be an area of unsuitable habitat which
could serve as a barrier to gene flow between English Channel and North Sea populations. Therefore it has been
suggested that crabs from the English Channel and North Sea may represent separate stocks. In the North Sea
female crabs have been recorded making extensive northward migrations while in the English Channel the
similarly extensive female migrations are in a westerly direction. In both regions these migrations are against the
dominant residual currents and thus it has been suggested that they may be contranatent behaviours with larvae
drifting towards natal areas. If larvae do return to their parents’ natal areas then further population differentiation
between areas could be generated. A previous Defra project (Edible crab spawning grounds in the English
Channel) reported areas of larval production in the western and eastern regions of the Channel with a central
area (between 0oW and 2oW) showing little larval production. This suggests that larvae may be retained in both
the western and eastern areas with little exchange between the two regions. Another Defra project (Edible crab
spawning grounds off eastern England) provided evidence of a jet like circulation in the North Sea known as the
Flamborough front. It has been suggested that this circulation feature might isolate areas north and south of the
front during the main spawning periods and that contrary to previous hypotheses there may be an asymmetric
contribution of southern populations to recruitment north of the front.
DEFRA project MF0226 (Population genetics of shellfish in British waters) reported the first investigation
of population genetic structure in C. pagurus. The data, based on three highly polymorphic (nuclear) microsatellite
loci, revealed weak but statistically significant population structure around UK coasts, even in some cases over
short distances (e.g. 10s of kilometres). No correlation of genetic differentiation with geographical distance or any
particular hydrographic features was found. The most likely biological explanation for such apparent ‘chaotic’
genetic heterogeneity is the hypothesis of “sweepstakes reproductive success” whereby due to stochastic
processes many individuals fail to contribute to recruitment. Such variance may then result in seemingly ‘chaotic’
changes in gene frequencies over space and time. This finding suggests that the effective (genetic) population
size of local crab populations may be highly variable over time and space. However, a concern with the
employment of highly variable genetic markers such as microsatellite loci is that the desired levels of variation
may also confer an elevated level of sampling error which may introduce ‘noise’ into estimates of genetic
differentiation. Where the true levels of population differentiation are weak (as in many marine species) such
noise may assume relatively greater importance (Waples 1998). Therefore the investigators reported the
requirement for further studies of crab population structure using additional genetic markers, larger sample sizes,
as well as a denser sampling scheme within local areas, to provide a more accurate description of population
stock structure in C. pagurus.
The fundamental objective of the present project was to define patterns of genetic variation among C.
pagurus sampled from multiple locations and times, and to interpret these patterns in terms of population/stock
structuring in relation to habitat discontinuities and species biology. The primary focus was the English Channel,
SID 5 (Rev. 3/06)
Page 5 of 21
however samples from around the British Isles were included to provide a UK-wide perspective for the Channel
population study.
Different mating systems can affect the ability of populations to respond to changing environmental
conditions and selection pressures such as fishing (Rowe & Hutchings 2003, Frankham 2005). Declining
population sizes reported for a number of marine species have emphasised the need for information on mating
systems to be incorporated into conservation and management strategies, particularly for exploited species.
Despite extensive study many important aspects of the breeding system of C. pagurus are unknown. During the
population sampling a number of berried female crabs were obtained, so the opportunity was taken to apply the
same genetic methods (microsatellite DNA fingerprinting) to answer a key question of C.pagurus reproductive
biology: are broods of offspring sired by single or multiple males?
A prerequisite to the application of genetic analysis is the possession of suitable genetic markers.
Although a previous project (MF0226) had developed 3 microsatellite marker loci, robust testing of hypotheses of
population structure and paternity required the development of additional microsatellite loci. For the population
structure analysis mitochondrial DNA (mtDNA) sequencing was also employed. Mitochondrial DNA (mtDNA) has
a number of features which can complement nuclear markers, these include: (i) mtDNA is passed through the
female line only and may reveal sex-specific patterns in gene flow; (ii) mtDNA effective population size is one
quarter that of nuclear loci which may make it sensitive to population differentiation and (iii) mtDNA diversity
retains more information on past (historical) patterns of population structure and so can inform interpretation of
contemporary structure.
2. METHODS
2.1. Genetic marker development
This project employed a combination of nuclear microsatellite DNA and mtDNA markers. The analysis of these
genetic markers requires suitable DNA primers permitting amplification of the target region by polymerase chain
reaction (PCR). Initially, microsatellte primers designed for the Dungeness crab (C. magister) were tested in C.
pagurus, but these yielded unsatisfactory results. Therefore a microsatellite enriched genomic library was
constructed for C. pagurus, from which 12 microsatellite loci were isolated and optimised for routine screening.
Details of the 12 loci are given in Table 1, and a full description of procedures can be found in McKeown & Shaw
(2008).
Primers permitting amplification of three different mtDNA gene fragments (Cytochrome oxidase I (COI),
NADH dehydrogenase 1 (ND1) and adenosine triphosphatase 6 (ATPase6) in C.pagurus were designed from
conserved sequence regions within published crustacean sequences. As a preliminary investigation DNA
sequencing of each of the three regions was performed for crab samples collected from Sweden and Brittany
(approximately the northern and southern limits, respectively, of the study). No sequence variation was detected
in either the ND1 or ATP6 fragments but sequence variation suitable for a population genetic study was detected
within the COI fragment. MtDNA variation was therefore assessed using direct sequencing of a 915 base pair (bp)
portion of the COI.
2.2 Population structure analysis
2.2.1 Sampling and DNA extraction
As part of an arrangement with Cefas staff large samples of C. pagurus adults were collected from a number of
sites around the UK covering the major fishing areas. In most cases sampling was from fishing boats using
standard crab pots. Samples from France, Ireland, Orkney, Shetland and Scandinavia were obtained from
research collaborators, again mostly using crab pot samples. Additional samples from within the English Channel
were obtained from local fishers. Due to the sampling gear used most samples comprise mature adult individuals.
Most sample sites were 1.5-15 km off open coasts, but three samples are of note for being taken from “nonoffshore” sites: the Newlyn 2007 sample (“Newlyn inshore”) was collected well within Mounts Bay; the Brittany
2000 sample (“Brittany inshore”) was taken from an inshore site near Brest; the Sweden 2000 sample was taken
from the mouth of the Gulmarsfjord. Sample details are presented in Table 2, and location shown in Fig.1. For a
number of samples sex data was available permitting males and females to be analysed separately. All tissue
samples were preserved in 100% ethanol. Total DNA was extracted using a standard CTABchlorofrom/isoamylalcohol method (Winnepenninckx et al., 1993).
Ethanol preserved larvae collected at a number of stations in the western and eastern regions of the
English Channel were provided by a Cefas operated larval survey (Table 2, Fig.1). As DNA extraction from larvae
using the CTAB method proved unsuccessful, a Chelex based method (modified from Estoup et al., 1996) was
developed for C.pagurus. Briefly, individual larvae were placed in a 0.2ml PCR tube and incubated at 37 oC until
ethanol evaporated. Then 100 l of 10% chelating resin (Chelex: Bio-Rad) and 10 l of proteinase K (10 mgl-1)
were added to each sample. The mixture was incubated at 55 oC overnight and then boiled (100 oC) for 20 min.
Samples were centrifuged at 3000xg/min for 30 s. The supernatant containing nucleic acids was then removed
and stored at -20 oC.
Table 1 – see Appendix
SID 5 (Rev. 3/06)
Page 6 of 21
Table 2. Sample names/locations, date of collection and numbers of individuals analysed for microsatellite and
mtDNA variation. Sample number corresponds with numbers on Figure 1.
Microsatellite genotyped
Sample site
Irish Sea/Celtic Sea
Aberystwyth
Newquay
Pendeen
N. Cornwall
English Channel
Newlyn
Newlyn-inshore
Brittany-North
Brittany-inshore
Brittany-South
Jersey
Jersey
Guernsey
PortScatho
Plymouth
Start Point
Lyme Bay
Swanage
Brighton
Hastings
Hastings
North Sea
Harwich
Harwich
Norfolk
Bridlington
Bridlington
Northumberland
Northumberland
Orkney-Hoy
Orkney-Sanday
Sweden-Gulmarsfjord
Larval samples
Lands End (Lizard)
Mounts Bay
Dover Straits
mtDNA only
Ireland-Southeast
Ireland-Northwest
Ireland-Southwest
Shetland-Northwest
Shetland-East
Shetland-Southwest
Norwegian Sea
Kattegat
Skaggerak
SID 5 (Rev. 3/06)
Sample
number
Collected
Male
Female
1
2
3
4
Aug-00
Jun-06
Jun-06
Sep-07
8
43
51
61
51
51
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
19
Sep-00
Oct-07
Oct-06
Sep-00
Jul-06
Sep-00
Sep-07
Sep-07
Apr-05&Jun-06
Oct-00
Jul-06
Jul-07
Jun-06
Sep-07
Aug-00
Oct-06
1
20
20
21
22
22
23
23
24
24
25
Jun-00
May-05
Jun-00
Aug-01
Jun-06
Jun-00
Sep-05
Jun-02
Jun-02
Jun-02
33
34
35
Jun-04
Jun-04
Jun-04
26
27
28
29
29
29
30
31
32
Jul-07
Jul-07
Jul-07
Jun-07
Jun-07
Jun-07
Dec-04
Jul-07
Sep-06
Gender
unknown
55
83
81
56
58
40
102
32
58
9
62
52
71
44
11
5
54
67
108
47
101
39
44
50
48
56
43
15
58
39
40
56
46
9
40
41
39
84
80
133
2
52
65
1
2
15
38
mtDNA
Total
69
94
102
55
84
81
114
58
102
72
84
80
133
63
133
52
55
65
72
162
63
159
80
84
106
94
65
98
38
80
20
22
21
30
27
34
31
30
27
23
21
29
29
33
24
27
28
20
27
80
80
80
30
29
31
15
20
13
21
28
24
Page 7 of 21
Figure 1. Map of the continental shelf of NW Europe illustrating sample sites. Black dots denote locations were
adults were collected, small black boxes in the English Channel insert denotes approximate locations of larval
surveys. Numbers correspond to those in Table 2.
SID 5 (Rev. 3/06)
Page 8 of 21
2.2.2 Microsatellite and mtDNA genotyping
Adults were genotyped at eight microsatellite loci, and larvae were genotyped at a subset of four of these loci.
PCRs were carried out in a total volume of 10 l, containing 50 ng of C. pagurus DNA, 3 pmol of each primer
(forward primer labelled with Cy5 flourescent dye), 2.0 mM MgCl 2, 1X PCR buffer, 0.2 mM dNTPs, 0.2 U of Taq
DNA polymerase (Bioline, UK). Amplifications involved an initial denaturation step (95 oC for 3 min) followed by
35 cycles of 30s at 95 oC, 30s at the locus specific annealing temperature (Table 1) and 30s at 72 oC. Genotyping
of larvae followed the same procedure with the modification that PCRs involved 55 cycles. PCR products were
analysed on an ALFexpressII automated DNA sequencer (Amersham Pharmacia Biotech) with different alleles
recorded against internal standard size markers using FRAGMENT MANAGER version 1.2 (Pharmacia Biotech).
MtDNA PCRs were carried out in a total volume of 50 l, containing 5-50 ng of C. pagurus DNA, 1mM
each primer, 0.2 U of Taq DNA polymerase (Bioline, UK), 1 X the supplied PCR buffer, 2.0 mM MgCl 2 and 0.2
mM dNTPs, and involved an initial denaturation step (95 oC for 3 min) followed by 35 cycles of 30 s at 95 oC, 30 s
at 54 oC and 60 s at 72 oC. PCR products were ExoSAP purified and sequenced using an ABI 377 sequencer.
2.2.3 Statistical analysis - Microsatellites
For each locus in each sample the number of alleles (N a), allelic richness (A; El Mousadik & Petit 1996 - the
number of alleles estimated by rarefaction if all samples were equal to the smallest sample size), allele
frequencies, observed (HO) and expected (HE) heterozygosity (Nei 1978) were calculated using FSTAT version
2.9.3.2 (Goudet 1995). Linkage equilibrium between pairs of loci and deviations from Hardy-Weinberg equilibrium
(HWE) expectations of genotype proportions were tested by exact tests conducted using a Markov chain method
(Guo & Thompson 1992) in GENEPOP 3.4 (Raymond & Rousset 1995).
Genetic differentiation among samples was quantified by FST (Wright 1951), the coefficient of intersample genetic variation, which ranges from 0 (identical gene frequencies) to 1 (samples fixed for different
genetic variants). Single locus and multilocus estimates of FST were calculated globally (over groups of samples)
and between sample pairs in GENEPOP 3.4. The significance of FST estimates were tested by permutation
(following Goudet et al., 1996) whereby individuals/genotypes were shuffled among samples (10 000 iterations for
each FST tested) and FST recalculated. Permutations were performed for each locus separately and for all loci
simultaneously and the probability of the null hypothesis (FST = 0) was taken as the proportions of replicates that
yielded a value of FST that was equal to, or higher than the observed value. The significance of multilocus
estimates of FST was also tested by bootstrapping over loci to generate 95% confidence intervals using FSTAT.
Genetic differentiation was qualitatively tested by global and pairwise exact tests of allele frequency differences
implemented in GENEPOP 3.4
To test for correlations between genetic and geographic distances (Isolation by distance – IBD) partial
Mantel tests were employed. IBD analyses were carried out over all samples and for the English Channel and
North Sea samples separately to distinguish scenarios where gene flow may be more influential at different
geographical scales. Geographical distances were defined as the shortest sea distance between sampled
locations. Genetic (pairwise FST) and geographical distances were log transformed. A hierarchical analysis of
molecular variance (AMOVA) was used to estimate the proportion of genetic variation distributed (i) between the
English Channel and North Sea, (ii) between samples within both regions, and (iii) within samples.
As an alternative descriptor of population structure and to provide additional information in their own right
assignment tests were performed whereby an individual is assigned to the population sample from which it has
the highest probability of ‘belonging’ based on analysis, in this case using the Bayesian method of Rannala &
Mountain (1997), of the individual’s genotype and the allele frequencies of the potential source populations.
Assignment tests were applied to investigate (i) self classification of adult crab samples, whereby an individual
crab could be assigned to the its original sample (in which case it was regarded as correctly self classified) or to
any of the other population samples, and (ii) assignment of larvae to adult samples. To complement this analysis
a simulation was performed to generate 10 000 genotypes for each population sample. From these genotypes a
distribution of random assignment values was generated. This distribution permitted rejection zones to be defined
whereby if the observed probability of ‘belonging’ was below a given threshold level (5% in this case) the
population sample could be rejected as a possible source (as opposed to being regarded as simply less likely
than another population in the empirical analysis).
Population bottlenecks or substantial reductions in effective population size, can severely compromise
population sustainability and genetic diversity. From a technical perspective bottlenecks may also generate
statistically significant estimates of genetic differentiation at microsatellite loci that are biologically meaningless
(Hedrick 1999). Data were therefore analysed for signatures of bottlenecks following Cornuet & Luikart (1996),
and estimates of effective population size obtained using the linkage disequilibrium approach of Waples & Do (in
press).
The statistical power of the microsatellite loci and sample sizes employed here to detect population
structure was assessed using a power analysis (Ryman & Palm 2006). The method simulates the divergence of
populations to various degrees of genetic differentiation (inferred by the user) and then tests if the loci and sample
sizes would be able to detect this ‘true’ level of population structure using various statistical tests. The power
analysis was employed for FST ranging from 0-0.1. 1000 replicates were run for each level of differentiation and
SID 5 (Rev. 3/06)
Page 9 of 21
results were reported as the proportion of tests that indicated significant differentiation at P < 0.05. For simulated
FST = 0 this therefore indicates the probability of Type I errors.
2.2.3 Statistical analysis – mtDNA
For each individual sequence data for 690 nucleotides was obtained. Different sequences were designated as
distinct haplotypes. Intrasample variation was described using number of haplotypes and haplotype diversity (i.e.
the probability that any two randomly chosen haplotypes are different). The mutational relationships among
haplotypes (phylogeny) was visualised by means of a minimum spanning parsimony tree. Genetic differentiation
analysis (FST, IBD and hierarchical AMOVA) were performed as for the microsatellite data. In all cases where P
values were calculated critical significance levels were adjusted for multiple tests using the Bonferroni correction
(Rice 1989).
2.3 Paternity analysis
2.3.1 Sampling
Ovigerous crabs overwinter without feeding and are rarely caught in traps (Edwards 1979). However, 18 berried
females were opportunistically captured as part of a Cefas operated beam trawl survey in the western English
Channel. Carapace width was measured to the nearest millimetre. Tissue samples were obtained from each
female by stimulating autotomy of a cheliped. Each cheliped was then stored in 95% ethanol. Fertilised eggs
(hereafter referred to as eggs) were brushed off each female and placed in a vial containing 95% ethanol.
Females were returned to the sea after sampling.
2.3.2 Technical and statistical analysis of paternity
Total DNA was isolated from female cheliped muscle tissue using the CTAB method. DNA was isolated from 40
individual eggs from each brood. Due to the small amount of tissue available a Chelex extraction method was
used for individual eggs. Using a pipette, individual eggs were transferred to 0.2 l PCR tubes and incubated at
37 oC until ethanol evaporated. Then 50 l of 5% chelating resin (Chelex: Bio-Rad) and 3 l of proteinase K (10
mgl-1) were added to each sample. The mixture was incubated at 55 oC for 3 hr and boiled (100 oC) for 10 min.
Samples were then centrifuged at 12000xg/min for 5 min, and the supernatant removed and stored at -20 oC. As
C. pagurus is a highly fecund species with 0.5-2.9 million eggs produced per female (Edwards 1979; Ungfors
2007), DNA was also extracted from pooled batches of eggs. This pooling strategy permitted the study of a
greater number of offspring. Briefly, eggs which were floating freely in ethanol were randomly sampled by
pipetting ~ 750 l of ethanol/egg ‘soup’ into an eppendorf tube. This was then incubated at 37 oC until all the
ethanol had evaporated. DNA extraction from the egg pool then followed the standard protocol as for adults. Each
pool contained between 300 - 500 eggs and 4 pools were extracted per brood. For the maternal, individual egg
and pooled egg samples PCR was used to amplify three microsatellite loci (Cpag4; Cpag5D8 and Cpag6C4B) as
described previously.
For each locus the maternal genotype was determined directly from DNA extracted from the female’s
tissue. The minimum number of fathers contributing to fertilisation of each brood was then inferred using two
methods. First, a single locus manual assignment approach whereby the minimum number of sires was estimated
to be half the number of non-maternal alleles recorded at a locus (based on both individual and pooled egg allele
counts), rounded to the next upper integer (Toonen 2004). Multiple paternity was therefore concluded if three or
more non-maternal alleles were detected at any locus among offspring. The second method was a multilocus
approach performed using the program GERUD version 2.0 (Jones 2005). While this is more sophisticated than
allele counting as it uses multiple loci simultaneously it was only applicable to data derived from individuals.
Following removal of maternal alleles from each offspring genotype, the minimum number of fathers (and their
genotypes) siring each brood was inferred using an exhaustive search algorithm. The probability (power) of the
loci to detect multiple paternity in a sample of offspring (PrDM) was estimated using the program PrDM v.1 (Neff
& Pitcher 2002).
3. RESULTS
3.1. Genetic marker development
As stated in the Methods section, marker development was successful with PCR primers obtained permitting
amplification of 12 microsatellite and three mtDNA gene fragments, for C.pagurus. The microsatellite primers are
published (McKeown & Shaw 2008) and have been used by other researchers studying C.pagurus population
structure in Scandinavian (Ungfors et al., submitted) and Irish (Prof E Gosling, Galway University) fisheries.
3.2.1 Population structure analysis – microsatellites
The number of alleles segregating at each locus ranged from 5 to 33, with a total of 135 alleles detected over all
samples. Only 10 alleles were restricted to single samples (private alleles) with the highest frequency of such
alleles being 0.012. For each locus levels of genetic variation within samples, as estimated by N a, A, HO and HE
were similar across all samples.
There was no evidence of significant linkage disequilibrium between any pair of loci within or across
samples, indicating that each locus represents an independent descriptor of genetic variation. No locus or sample
SID 5 (Rev. 3/06)
Page 10 of 21
was marked by systematic departures from HWE indicating: (a) that each locus was free from technical artefacts,
such as null alleles, that could compromise results; and (b) genotype frequencies within each sample conformed
to expectations of random mating (i.e. no inbreeding).
The global (over all samples) multi-locus estimate of population structure was low (FST = 0.004) but
significant (95% CI 0.002-0.006; P < 0.0001) (Table 3). Individual locus global FST’s ranged from 0.002 to 0.009
and were significant at 7 out of the 8 loci. The global multilocus Exact Test also revealed highly significant
population structure (P < 0.001), with significant differentiation in allele frequencies at the same 7 (out of 8) loci as
for FST. A similar pattern of low but significant population structure was observed for the English Channel
samples, but for the North Sea samples an FST of 0.001 (not significantly greater than zero by bootstrap and
permutation analysis) indicated no significant sub-population structuring (Table 3). Although multilocus exact tests
of allele frequency differences did indicate significant structure among North Sea samples non-significant singlelocus tests results were observed at 4 of the 8 loci and if Bonferroni correction was applied across loci (critical
significance level of 0.05/8 = 0.006) all single locus values were non-significant.
Table 3. Single and multilocus tests of global population differentiation using FST (and the corresponding
probability that rejection of the null hypothesis that FST = 0 is false) and exact tests of allele frequency differences
across all samples, and within the Channel and North Sea areas. 95% confidence intervals for multilocus FST
obtained by bootstrapping.
Locus
Cpag5D8
Cpag4
Cpag6c4B
Cpag3A2
Cpag2A5-2
Cpag15
Cpag1B9
Cpag3D7
All
loci
95%
CI
FST
All samples
P that
Exact
FST = 0
P
English Channel
P that
Exact
FST
FST = 0
P
FST
North Sea
P that
Exact
FST = 0
P
0.002
< 0.001
< 0.001
0.003
0.004
0.004
0.001
0.419
0.020
0.002
< 0.001
< 0.001
0.001
0.025
0.008
0.002
0.196
0.038
0.009
< 0.001
< 0.001
0.006
0.001
< 0.001
0.004
0.026
0.042
0.008
< 0.001
< 0.001
0.009
0.001
< 0.001
-0.001
0.979
0.697
0.002
0.133
0.097
0.001
0.027
0.120
0.001
0.876
0.500
0.004
< 0.001
0.001
0.003
0.001
0.013
0
0.504
0.421
0.002
< 0.001
0.005
0.001
0.089
0.030
0
0.016
0.126
0.006
< 0.001
< 0.001
0.006
< 0.001
< 0.001
0.002
0.052
0.011
0.004
< 0.001
< 0.001
0.004
< 0.001
< 0.001
0.001
0.069
0.008
0.0020.006
0.0020.006
0-0.002
Pairwise tests of genetic differences between samples within the English Channel (Table 4) revealed a
number of significant results. Values of FST ranged from zero to 0.0144, with 57 of 120 pairwise estimates
statistically significant. Applying a sample specific (SS) Bonferroni correction whereby the critical significance
level was adjusted according to the number of simultaneous tests for each sample (i.e. k=15 for the English
Channel samples, so corrected significance level = 0.05/15 = 0.0033) then only 25 comparisons remained
significant. If a table-wide (TW) Bonferroni correction (i.e. the total number of simultaneous tests considered) was
applied (k=120, 0.05/120 = 0.0004) 13 remained significant. Notably, 15 of the 25 significant comparisons at
P<0.0033 involved the Newlyn and/or Brittany “inshore” samples, while 11 of the 13 significant comparisons at
P<0.0004 also involved these two “inshore” samples. 82 out of 120 pairwise Exact tests were significant at the
0.05 level, of which 44 remained significant after SS Bonferroni correction (20 involving the two “inshore”
samples), and 20 remained significant after TW Bonferroni correction (13 of which involved the two “inshore”
samples). Excluding the two ‘inshore’ samples the average pairwise FST among Channel samples was 0.0039.
Comparisons involving the Newlyn inshore sample produced an average FST of 0.0070. The corresponding value
was even higher for the Brittany Inshore sample (FST = 0.0083). The only other sample that was markedly
differentiated was the Swanage sample: 12 of 15 FST comparisons significant, of which 6 remain significant at
SID 5 (Rev. 3/06)
Page 11 of 21
P<0.0033, and 2 at P<0.0004, with average FST = 0.0070 (excluding the two “inshore” samples). Aside from the
Swanage and two ‘inshore’ samples the remaining English Channel samples were all genetically homogenous.
Table 4 – see Appendix
Pairwise FST values among North Sea samples ranged from zero to 0.0116 (Table 5) with 16 of 45
pairwise comparisons significantly greater than zero. After SS Bonferroni correction (i.e. k=9 tests, so corrected
significance level = 0.05/9 = 0.0056) 8 comparisons remained significant while after TW Bonferroni correction
(k=45, 0.05/45 = 0.0011) 6 remained significant. These six highly significant outcomes all involved the Swedish
(fjord) sample (average pairwise FST = 0.0149; range 0.0057-0.0116), which is significantly different form all UK
samples, and accounts for 6 of the 8 significant results at P<0.0056. Average pairwise FST between North Sea
samples, excluding the Swedish sample, was 0.0013. 20 out of 45 pairwise Exact tests were significant at the
0.05 level, 15 remained significant after SS Bonferroni correction (including all comparisons with the Swedish
sample), and 9 remained significant after TW Bonferroni (all involve the Swedish sample).
Table 5. Tests of genetic differentiation at microsatellite loci between samples from the North Sea. Pairwise FST
(below diagonal) and Exact Test probability of genic differentiation (above diagonal). FST values significantly
greater than zero underlined and probability indicated where Exact Test results significant. Values remaining
significant after Bonferroni correction at k=9 in italics, at k=45 in bold. ns=non-significant). Comparisons between
temporal replicates in boxes.
Harwich
‘00
Harwich
‘05
Norfolk
Brid ‘01
Brid ‘06
Northum
‘00
Northum
‘05
OrkneyHoy
OrkneySanday
Sweden
-
NS
NS
NS
0.0135
NS
NS
NS
NS
<0.0001
-0.001
-
0.0013
NS
NS
NS
NS
NS
NS
<0.0001
0.0019
0.0061
-
NS
0.0043
0.0435
0.0137
NS
0.0024
<0.0001
Brid ‘01
-0.0016
0.0013
-0.0003
-
0.0493
NS
0.0047
NS
0.0024
<0.0001
Brid ‘06
0.0026
0.0022
0.0061
0.0015
-
0.0241
NS
0.0153
0.0223
<0.0001
0.0007
-0.0006
0.0023
-0.001
0.0032
-
NS
NS
NS
<0.0001
-0.0003
-0.0017
0.0044
0.0017
0.0021
0.001
-
NS
NS
<0.0001
-0.0007
0.0002
0.0014
-0.0009
0.0038
-0.0013
-0.0011
-
NS
<0.0001
-0.0013
-0.0014
0.0082
0.0018
0.0027
0.0029
-0.0009
0.0031
-
<0.0001
0.007
0.0083
0.0116
0.0075
0.0086
0.0093
0.0057
0.0076
0.0057
-
Harwich
‘00
Harwich
‘05
Norfolk
00
Northum
‘00
Northum
‘05
OrkneyHoy 02
OrkneySanday
02
Sweden
00
Hierarchical AMOVA indicated that 99% of genetic variation was contained within samples (i.e. very little
differentiation between samples) and there was no indication of genetic differentiation at the regional level (i.e.
between English Channel and North Sea, or between these regions and the Celtic/Irish sea samples). With the
exception of the Newlyn and Brittany inshore samples and the Swedish fjord sample microsatellites revealed
overall genetic homogeneity. As a result there was no power to test for Isolation-By-Distance (IBD), with no
geographic-based distance differentiation observed within or between any of the regions,
The power analysis indicated that the microsatellite loci and sample sizes employed had a high
probability (94%) of detecting significant genetic differences as low as FST = 0.001. Power analysis also indicated
a low probability (3%) of Type I error, i.e. falsely inferring genetic structure.
For the three samples of larvae tested, no significant deviations from HWE were detected at any of the 4
loci screened. Exact tests revealed a significant global value of differentiation among larval samples (P = 0.038),
but this was due to a significant result at a single locus only. Global FST per locus ranged from 0 to 0.004, with a
combined estimate indicating zero genetic differentiation. All pairwise tests (Exact, FST) yielded non-significant
outcomes.
Low levels of self classification were observed among the adult crab samples with no population sample
exhibiting more than 10% correct self classification. Each individual had an almost identical probability of coming
from each potential source population indicating that even for correctly self-classified individuals differences in
assignment probabilities to different populations (i.e. the strength of assignment) were marginal. Assignment of
larvae was concordant with this pattern, with each individual larva having an almost identical probability of
belonging to any of the population samples, indicating that the origin could not be reliably discerned. The
SID 5 (Rev. 3/06)
Page 12 of 21
weakness of the assignment power was confirmed by the simulation analysis which indicated that for any
individual (adult or larval) no population could be excluded as a potential source at the 5% significance level.
There was no evidence of bottlenecks in population size as detected by the suite of analyses described
by Cornuet & Luikart (1996) for any of the samples (adult and larvae). Median estimates of effective population
size were large (100s-1000s) for most samples, with upper 95% confidence intervals at infinity indicating effective
populations that are so large the analysis struggles to define a realistic upper limit. For the Newlyn inshore
sample the median estimate of effective population size was 99 (95% CI 72.5 - 147.5), and for the Brittany
inshore sample was 17.5 (95% CI 14.2 - 21.7). The estimate for the Jersey (2001) sample was 138.2 (95% CI
81.1 – 357.7) while the corresponding confidence interval for the Jersey sample collected in 2006 was infinity.
Figure 2: (A) Haplotype network depicting genetic relationships among the 9 mtDNA types found in C.pagurus.
Bars denote single mutations separating haplotypes, and size of discs denotes relative frequencies in the whole
dataset. White = haplotype A, Grey = haplotype B and Black = haplotype C. (B) Geographical distribution of
haplotype frequencies in samples of C.pagurus across NW Europe. Colours in pie charts correspond to the
respective haplotypes in network (multiple sample sites are pooled within western and eastern Channel).
3.2.2 Population structure analysis – mtDNA
A total of nine different mitochondrial types (haplotypes) were detected. Analysis of evolutionary relationships
among haplotypes indicated that 7 of the 9 radiated by single mutations from one central haplotype (A - Fig. 2A),
which was also by far the most common (Table 6). The most divergent haplotype was separated from its closest
relative by two mutations (and three from the central haplotype).
MtDNA haplotypes frequencies for each sample are given in Table 6 and represented graphically (with
some pooling of the English Channel samples (for clarity) in Fig.2B. The total number of haplotypes per sample
ranged from two to six (Shetland). Haplotype diversity ranged from 0.190 to 0.697, with higher diversity among
North Sea samples (h=0.500-0.677) than among the Channel samples (0.190-0.548).
Over all samples mtDNA revealed a much higher level of genetic structure than microsatellites with an
overall FST of 0.1384. Among the English Channel samples the global FST was 0.066. Of the 55 pairwise
comparisons among the Channel samples 23 were significant, 13 after SS Bonferroni correction and 6 after TW
Bonferroni correction (Table 7). Prominent among these significant outcomes were the Guernsey and Brighton
samples. Unlike the microsatellite data, mtDNA revealed a high level of population structuring in the North Sea
with an overall FST of 0.1384. Of the 45 pairwise comparisons 23 were significant, 14 after SS Bonferroni
correction, 9 after TW Bonferroni (Table 8).
For both the Channel and North Sea samples there was no obvious geographic pattern to the observed
genetic structure, particularly among the North Sea samples. No significant IBD was detected in either region or
over the entire sampled range. However, unlike the microsatellites, regional differences were observed in mtDNA
diversity as hierarchical AMOVA indicated that 16.14% of the total genetic variation was attributable to differences
between the Channel and North Sea regions, which was greater than the amount of variation distributed among
samples within both regions (11.83%).
SID 5 (Rev. 3/06)
Page 13 of 21
Table 6. Geographical distribution of 9 mtDNA COI haplotypes and corresponding haplotype diversity (H
diversity)
Haplotype frequency
Location
A
Ireland-NW
Ireland-SW
Ireland-SE
Aberystwyth
Pendeen
Newlyn inshore
Brittany inshore
Brittany-North
Jersey
Guernsey
Plymouth
Start Point
Lyme Bay
Swanage
Brighton
Hastings
Harwich
Norfolk
Bridlington
Northumberland
Orkney
Shetland
Sweden
Skaggerak-G
Skaggerak-LV
Norway
11
26
25
16
8
16
18
27
30
15
24
22
20
17
19
27
17
3
9
3
10
26
9
18
11
12
B
C
D
E
F
18
5
4
4
7
2
8
2
1
1
6
G
H
H
diversity
0.487
0.279
0.296
0.336
0.697
0.419
0.484
0.190
0.219
0.548
0.331
0.333
0.245
0.338
0.507
0.133
0.619
0.590
0.621
0.548
0.611
0.619
0.661
0.500
0.677
0.586
I
1
7
2
1
1
1
3
15
3
2
3
8
2
5
15
4
18
8
8
7
1
4
1
2
11
6
14
6
2
11
2
9
8
7
1
1
1
1
1
1
1
1
1
1
1
Table 7. Tests of mtDNA genetic differentiation between samples from the Channel. Pairwise FST (below
diagonal) and Exact Test probability of genic differentiation (above diagonal). Probability indicated where Exact
Test results significant. Values that remain significant after Bonferroni correction at k=10 in italics, and at k=55 in
bold. ns=non-significant). Signifcant FST values in bold.
Newlyn
Inshore
Brittany
Inshore
North
Brittany
Jersey
Guernse
y
Plymout
Start
Point
Lyme
Bay
Swanag
e
Brighton
Hasting
s
-
NS
NS
NS
0.0033
NS
NS
NS
NS
NS
NS
0.01023
-
NS
0.00125
NS
0.0012
0.0041
0.01405
0.00195
0.0005
0.06368
-
NS
NS
NS
NS
NS
0.02345
NS
0.0559
0.01725
-
0.0189
NS
NS
NS
0.04615
NS
0.22519
0.31116
0.36727
0.30089
-
<0.0000
1
0.0042
0.00165
0.0172
NS
<0.0000
1
0.00065
0.02776
0.02398
0.04082
0.00958
0.00422
0.01849
0.36114
-
0.00585
0.01125
0.0432
0.00315
0.01185
0.05479
0.00697
0.25425
0.02988
-
NS
NS
NS
NS
0.06403
0.02083
0.02801
0.03241
0.28645
0.02839
0.03816
0.0197
-0.0218
0.18561
0.03434
Brighton
0.04163
0.11854
0.1585
0.09518
0.04569
Hastings
0.00813
0.09333
0.02474
-0.0261
0.33969
Newlyn
Inshore
Brittany
Inshore
North
Brittany
Jersey
Guernsey
Plymouth
Start
Point
Lyme
Bay
Swanage
0.00483
0.02714
SID 5 (Rev. 3/06)
<0.0000
1
<0.0000
1
<0.0000
1
0.03693
0.02927
-
NS
NS
NS
0.01918
-
NS
NS
0.15345
0.06687
0.09355
0.01238
-
NS
0.03782
0.02219
0.03416
0.00307
0.13523
-
Page 14 of 21
Table 8. Tests of mtDNA genetic differentiation between samples from the North Sea. Pairwise FST (below
diagonal) and Exact Test probability of genic differentiation (above diagonal). Probability indicated where Exact
Test results significant. Values that remain significant after Bonferroni correction at k=10 in italics, and at k=55 in
bold. ns=non-significant). Signifcant FST values in bold.
Harwich
Norfolk
Bridling
Northum
Orkney
Shetlan
d
Sweden
Kattegat
Skaggera
k
Norway
Harwich
-
0.00075
NS
0.00005
NS
NS
NS
NS
NS
NS
Norfolk
0.2927
-
0.00165
NS
0.03480
0.00065
0.02780
0.00010
0.00430
<0.0000
1
Bridling
0.0000
0.2532
-
<0.0000
1
0.00735
NS
0.00940
NS
NS
0.04820
Northum
0.3367
0.0000
0.2998
-
0.02505
<0.0000
1
0.01325
<0.0000
1
0.00080
<0.0000
1
Orkney
0.0571
0.0844
0.0770
0.1192
-
NS
NS
0.00505
NS
NS
Shetland
0.0000
0.2682
0.0355
0.3084
0.0307
-
NS
NS
NS
0.04980
Sweden
0.0349
0.1080
0.0614
0.1449
0.0000
0.0097
-
0.00395
NS
0.00120
Kattegat
0.0174
0.4632
0.1033
0.5016
0.1958
0.0303
0.1594
-
NS
NS
Skaggera
k
0.0000
0.2563
0.0000
0.3016
0.0352
0.0000
0.0157
0.0249
-
NS
Norway
0.0405
0.4628
0.1233
0.5018
0.2087
0.0502
0.1729
0.0000
0.0428
-
3.3 Paternity analysis
For each of the 18 broods tested the genotypes of the 40 individual eggs revealed the same allelic combinations
(maternal and presumed paternal alleles) as detected in the respective samples of pooled eggs. Allele counts
provided no evidence (i.e. not more than 2 paternal alleles at any locus) that more than a single male had
contributed to any of the 18 broods tested. The multilocus analysis was compatible with the single-locus analysis
for all broods, indicating that mating involving a single male and the genotyped mother was sufficient to account
for the respective brood genotypes in each case.
Overall, the set of microsatellites provided considerable power for detecting multiple paternity (if present)
within broods. As expected the PrDM increased with sample size. A sample of only 40 eggs (equal to the number
of individual eggs assayed per brood here) was sufficient to obtain a PrDM of 99.9% under the assumption of
equal male contribution and 92.7% in the case of a 90:10 paternal skew. With a sample size of 50 eggs (less than
assayed here via the pooling method) PrDM was 99.8-99.9% regardless of paternal contribution.
4. INTERPRETATION
4.1 Population structure
Levels of microsatellite variation were similar across all samples with no sample exhibiting significantly reduced or
elevated levels of diversity. Concordant with this no sample demonstrated signatures of excessively reduced or
declining ‘genetic’ population sizes that could be indicative of overexploitation. The microsatellite loci exhibited
high levels of polymorphism/variation and simulation analysis demonstrated that this variation conferred a high
degree of statistical power, enabling the potential detection of even very weak levels of population structure. Over
all samples microsatellites revealed numerically small but statistically significant population structure. Statistically
significant genetic differentiation was also detected between many sample pairs. Concordant with the low overall
level of genetic structure many of these pairwise differences were small and in many cases became nonsignificant after Bonferroni correction of significance levels for multiple simultaneous tests. Although Bonferroni
correction is almost ubiquitously employed in studies such as the one reported here it has been suggested that
this approach may be overly conservative (resulting in elevated Type II error rates) in situations where as few as
5 or more comparisons are considered (Altman 1981). However Hutchinson et al. (2001) suggested that
significant values that become non-significant after Bonferroni correction should be considered with caution.
Furthermore, in light of the simulation results indicating that the sample sizes and loci employed were unlikely to
produce many Type II errors it would not be prudent to simply disregard the genetic differentiation revealed at the
microsatellite loci.
Interpretation of the biological processes underlying the microsatellite results is complicated. The pairwise
differences among samples exhibited a chaotic pattern with respect to geography. No significant isolation by
distance effect was detected (i) across all samples or (ii) among the English Channel or North Sea samples
separately. Hierarchical AMOVA revealed no significant differentiation between the North Sea and English
SID 5 (Rev. 3/06)
Page 15 of 21
Channel groupings, that is to say that within-region pairwise differences (within English Channel or within North
Sea) were not less than between-region comparisons (English Channel vs. North Sea). It is noteworthy that
microsatellite genetic structuring was less pronounced in the North Sea than the English Channel suggesting a
subtle regional pattern. Furthermore genetic differences between geographically distinct samples were in many
cases of similar magnitude to genetic differences between temporal replicates (i.e. samples collected from the
same locations at different times). This means that estimates of population structure may not simply reflect
geographical differences between samples but also temporal differences, and also that such small differences
between samples may not be stable over time. We would propose that the spatial/temporal genetic patchiness
results from large variance in the reproductive success of individuals within areas, resulting in random (and
possibly temporary) distortions of allele frequencies. Such biased reproductive success may be due to stochastic
processes and/or selection. The related hypothesis of ‘sweepstakes reproductive success’ has been proposed in
many marine species. This posits that due to stochastic processes associated with reproduction and
oceanographic conditions conducive to larval development, retention and recruitment (Hedgecock 1994) many
individuals fail to contribute to recruitment. The findings of the paternity analysis of C. pagurus indicate a mating
system that would be susceptible to large variances in reproductive success among individuals (discussed
below).
A salient feature of the microsatellite data was the greater than average genetic differentiation of the
Newlyn inshore, Brittany inshore and Gulmarsfjord samples, which were the only samples originating from sites
(bays and a fjord) with potentially restricted larval exchange with the open ocean. There has been a recent
paradigm shift in that larval dispersal is increasingly being regarded as directed or restricted rather than random
(Armsworth et al. 2001) and numerous studies have indicated that levels of local larval retention may be high
(Swearer et al. 1999; Jones et al. 1999). In a recent study Oresland & Andre (2008) reported genetic
differentiation of cod larvae sampled within a fjord from larvae outside the fjord. It is possible that oceanographic
features associated with the ‘inshore’ sampled locations may promote retention of locally produced larvae and/or
prevent entry of larvae from populations in the open sea, which would facilitate genetic divergence. This isolation
may also be conducive to more pronounced variance in reproductive success among individuals and hence the
elevated genetic differentiation of these samples. Support for this comes from the fact that the Newlyn and
Brittany inshore samples produced the lowest estimates of effective population size, as variance in reproductive
success among individuals is expected to reduce effective population size.
In contrast to the microsatellite results the mtDNA marker revealed pronounced population structure at
both the macro-(regional) and micro-(within region) geographical scales. Hierarchical AMOVA revealed a strong
partitioning between the English Channel and North Sea, and haplotype frequencies were notably similar among
samples from the English Channel, Celtic Sea, southern Irish Sea and SW coast of Ireland, with this group
distinctly different (lower frequencies of haplotype B) from samples collected from the north coast of Ireland into
the western and eastern coasts of the North Sea (although there was much more variation among North Sea
sites). Despite this geographical pattern there was no significant pattern of IBD at any geographical level. This
lack of an IBD effect was especially apparent among the North Sea and English Channel samples. Again, within
both these regions the mtDNA population structure could be described as chaotic with respect to geography.
The apparent variance in patterns between the two genetic marker types can be explained by several
mutually nonexclusive hypotheses. First, microsatellites may display higher levels of homoplasy (possessing
similar genes by chance rather than descent) than mtDNA because of higher mutation rates in the former.
Homoplasy can inflate apparent similarity among populations and lead to an underestimation of population
structure (Estoup et al., 2002). Second, the large variances in reproductive success resulting in chaotic genetic
patchiness at the nuclear level (microsatellites) would be expected to produce even more pronounced
differentiation of mtDNA. This is because mtDNA (maternally inherited and clonal) is subject to an effective
population size one quarter that of nuclear genes which from a population genetics viewpoint this can make
mtDNA markers more sensitive to population dynamics than nuclear markers resulting in stronger genetic
differentiation (Birky et al. 1989). The similarly ‘chaotic’ pattern of mtDNA diversity (especially within regions) is
certainly compatible with this hypothesis. A third hypothesis is sex-biased gene flow. As mtDNA is inherited
through the maternal line only, male biased gene flow will homogenise nuclear (i.e. microsatellite) gene
frequencies whilst allowing mtDNA differentiation, whereas female biased gene flow will tend to homogenise
mtDNA frequencies along with nuclear gene frequencies (Birky et al. 1989). Typically, male biased gene flow is
reported in situations where males are the migrating sex. However, in C. pagurus it is the females which
undertake extensive migrations while males are more resident (Bennett 1995). An important distinction to be
made here is between dispersal (migration) and gene flow (migration followed by interbreeding and recruitment).
Male biased gene flow could occur in C. pagurus due to (a) larval drift to the maternal natal area and/or (b) return
migrations of females to natal areas. In both cases females may migrate and mate with males from other
‘populations’ but due to larval/female philopatry mtDNA (female) gene flow may be reduced compared to nuclear
gene flow, as the female genes are returned to the same site they originated but male genes are transferred
between sites. The suggestion that female migrations against the prevailing currents may be contranatent
behaviour to compensate for larval drift (Bennett & Brown 1983; Bennett 1995) would support such a process.
Despite extensive mark-recapture studies throughout the English Channel (Bennett & Brown 1983) and North
Sea (Mistakidis 1960; Mason 1962; Edwards 1979) no evidence of apparent return migrations of females has
been reported, although Bennet & Brown (1983) did observe some easterly movements of crabs in the English
Channel. More recently Ungfors et al. (2007) reported evidence of a high proportion (> 40%) of possible ‘return’
SID 5 (Rev. 3/06)
Page 16 of 21
movements by females. The particular life history characteristics of C. pagurus raise the intriguing possibility that
female migrations may enhance male gene flow but constrain female gene flow. We would conclude that greater
mtDNA than nuclear gene differentiation among crab populations is consistent with what is known of the biology
of the species, and should be investigated further with physical and genetic tagging studies.
The patterns reported here, and the proposed process underlying these, do not contradict data for C.
pagurus derived from other Defra projects, and in some cases the various types of information are
complementary. As previously mentioned a Defra project (Edible crab spawning grounds in the English Channel)
identified the central portion of the Channel as an area of little or no larval production. Microsatellite analysis of
larvae sampled from the western (the Lizard and Mounts Bay) and eastern (Dover Strait) Channel flanking this
proposed ‘non-productive’ area revealed no significant differentiation. Under the proposed model of female return
migrations such a lack of differentiation does not require larval transfer between these two areas. In fact, the
recording of limited larval density in this area supports the female return migration model for this region rather
than the larval drift model.
The data pertaining to the North Sea also do not contradict the proposal (from Defra project: Edible crab
spawning grounds off eastern England) that the Flamborough front serves as a barrier to larval exchange
between the areas north and south of this front. An accompanying hypothesis from the same project was that
there may be asymmetric migration (south to north) across this area. Interestingly, the samples north of the front
exhibited a number of mtDNA haplotypes not found among samples south of the front. While this does not
indisputably prove asymmetric migration it does suggest that female gene flow may not be occurring in a north to
south direction across this front (as it would be expected that these northern haplotypes would be found among
the southern samples) therefore any exchange that is occurring is likely to be south to north.
4.2 Paternity analysis
The genetic paternity analysis provided no evidence that female broods were sired by more than one male.
Although multiple paternity in C. pagurus cannot be ruled out the data indicate that single paternity is the
predominant system. An explanation for single paternity is that C. pagurus females are monandrous. While it is
believed that male cancrids are generally polygynous and exhibit female-centred competition (Orensanz &
Gallucci 1988, Orensanz et al., 1995) the range of female mating behaviours have not been fully resolved. C.
pagurus mating requires the female to be recently moulted and when in a soft-shelled receptive state females are
likely to be highly vulnerable to injuries. The ability of C. pagurus to use stored sperm to fertilise multiple
successive egg clutches might benefit the female by reducing the need for additional mating, and therefore
avoiding the need for extra moult cycles. Also, trans-moult sperm retention has been demonstrated in Cancrids
(Orensanz et al. 1995), including C. magister (Shirley & McNutt 1989). If sperm is retained by C.pagurus across
moults, then single paternity of broods suggests that females mate once and then use stored sperm to fertilise
successive clutches of eggs. If females do mate multiple times, then single paternity suggests that either stored
sperm is lost during moulting or that there must be some process of sperm precedence allowing one male’s
sperm to fertilise all eggs. Fresh sperm from a new mating might fertilise the majority of eggs due to depletion
and/or degradation of older stored sperm. Alternatively, single paternity might reflect effective female
postcopulatory sperm choice or sperm precedence mechanisms, such as sperm removal or sperm stratification
processes (Birkhead & Hunter 1990).
4.3 Implications for management
An important consideration when incorporating genetic data into fisheries management is that very limited
exchange of migrants (1-5%) per generation is sufficient to obscure genetic structuring. While such low migration
rates are generally sufficient for population connectivity over evolutionary timescales, they are a negligible force
for replenishing depleted stocks over a timescale of interest to fisheries. Therefore while genetic markers are
powerful tools to detect population structure it must be stressed that they often provide a conservative estimate of
stock structure.
Among highly mobile species it is possible that population structure may be obscured by migration. For
example Nesbo et al. (2000) detected population structure among samples of Atlantic mackerel (Scomber
scombrus) collected at spawning times but not among those collected outside of spawning times. Migrations by
adult crabs during non-spawning periods could theoretically obscure extensive population structure, as samples
could comprise a large proportion of migrants. Recent statistical approaches permitting population clustering
independent of sample of origin offer an analytical approach of teasing out such cryptic population structuring.
Such methods were employed here but did not indicate any marked population units. Furthermore, C. pagurus
males are known to be largely resident and hence samples of males are less likely to include migrants.
Restricting analysis to males did not reveal different or increased levels of population structure. Therefore, while
future studies may benefit from sampling newly settled individuals, which may reduce the sampling noise due to
adult migration, the finding of weak population structuring here is not an artefact of inappropriate sampling.
The microsatellite data revealed weak population structure indicating a high degree of connectivity among
locations, at least within the main regions studied. Such a pattern suggests that if crabs are over harvested from
one location migration and gene flow will contribute to replenishment. However, the mtDNA data suggest that
such an approach to management would be overly-simplistic and dangerous. The degree of population structure
revealed by mtDNA indicates that despite the overall level of connectivity suggested by microsatellites there is
extensive structuring/isolation at the mtDNA level. This means that recruitment in many areas may be largely
SID 5 (Rev. 3/06)
Page 17 of 21
dependent on local females. Therefore while recruits in a particular location may be derived from a male some
distance away their presence in that area is dependent, to some degree, on a local female.
The paternity analysis indicating single paternity of broods emphasises the importance of female
numbers to local populations. The effective population size is a key parameter in conservation and population
genetics, and is essentially the number of breeders in a population. At the population level multiple paternity is
expected to increase effective population size. Therefore single paternity species/populations may be more prone
to loss of genetic diversity if population sizes are reduced (e.g. by over-fishing) than genetically polyandrous
ones. In situations where females are each fertilised by a single male in a given reproductive cycle the number of
females imposes a strict constraint on the number of males that can breed. This is likely to be conducive to large
variances in reproductive success among individuals as certain males may essentially sire all a female’s
offspring. Also, under this type of mating system a decline in female numbers would result in proportional
reductions in effective population size regardless of the number of males. Therefore although the data suggest
that effective population sizes are large and there is no evidence of population bottlenecks, C.pagurus may be
susceptible to decline in effective population size (and therefore loss of genetic diversity) if over-fishing of females
occurs. Female landings per unit effort are considerably higher than those for males for most of the year and are
highest in the 6 month period from June to November (Bennett 1995). Given the important role of females
outlined above, the importance of collecting and monitoring sex-specific landings statistics is emphasised.
Furthermore, while the ban on landing of ovigerous females may seem a little incongruous due to their low
catchability, a continuation of this ban is recommended.
At the outset of this project the aim was to describe stock structure in a spatial and/or temporal context
amenable to standard stock assessment and management protocols. However, the data revealed a complex
pattern of population structure and potential underlying processes. MtDNA revealed distinct genetic differentiation
between the English Channel and North Sea indicating that these two regions are largely independent and should
be managed separately. Despite the extensive sampling and power of the genetic markers employed the present
data did not permit the detection of well resolved spatially separated population units within regions. Overall the
data indicated a high degree of connectivity among samples, suggesting that micromanagement of the resource
at a local scale is not necessary. Interpretation of both the mtDNA and microsatellite data suggests that females
play an essential part in genetic population connectivity and overall population dynamics. The finding of a greater
level of genetic patchiness in the English Channel than the North Sea suggests that conditions within the English
Channel may be more conducive to population fluctuations, possibly due to the more complex hydrography
generating potential for localised effects of eddies and frontal systems to exacerbate spatio-temporal patchiness
in larval survival and recruitment.. The genetic differentiation of the inshore samples confirms that there is the
potential for locally isolated and small population units which may be more vulnerable to population declines and
require independent management. However, in general the genetic data provided no evidence of excessively
reduced or declining population units.
References
Abaunza P, Murta AG, Campbell N, et al. (2008) Stock identity of horse mackerel (Trachurus trachurus) in the
Northeast Atlantic and Mediterranean Sea: Integrating the results from different stock identification
approaches. Fisheries Research, 89, 196-209.
Altman DG (1991) Practical statistics for medical research. Chapman and Hall London.
Armsworth PR, James MK, Bode L (2001) When to press on or turn back: dispersal strategies for reef fish larvae.
American Naturalist, 157, 434–450.
Avise JC (2000) Phylogeography. Harvard University Press, Cambridge, MA.
Bennett DB (1995) Factors in the life history of the edible crab (Cancer pagurus L.) that influence modelling and
management. ICES Marine Science Symposium, 199: 89-98.
Bennett D, Brown C (1983) Crab (Cancer pagurus) migrations in the English Channel. Journal of the Marine
Biology Association of the UK, 63, 371-398.
Birkhead TR, Hunter FM (1990) Mechanisms of sperm competition. Trends in Ecology and Evolution, 5, 48-52.
Birky CW jr, Fuerst P, Maruyama T (1989) Organelle gene diversity under migration, mutation and drift:
Equilibrium expectations, approach to equilibrium, effects ofheteroplasmic cells and comparison to
nuclear genes. Genetics, 160, 741-751.
Carvalho GR, Hauser L (1994) Molecular genetics and the stock concept in fisheries. Reviews in Fish Biology
and Fisheries, 4, 326-350.
Cornuet JM, Luikart G (1996) Description and power analysis of two tests for detecting recent population
bottlenecks from allele frequency data. Genetics, 144, 2001-2014.
Eaton DR, Brown J, Addison JT, Milligan SP, Fernand LJ (2003) Edible crab (Cancer pagurus) larvae surveys off
the east coast of England: implications for stock structure. Fisheries Research, 65, 191-199.
Edwards E (1979) The Edible Crab and its fishery in British Waters. Fishing News Book Ltd, Farnham Surrey,
England.
El Mousadik A, Petit RJ (1996) High level of genetic differentiation for allelic richness among populations of the
argan tree [Argania spinosa (L.) Skeels] endemic to Morocco. Theoretical and Applied Genetics, 92, 832–
839.
SID 5 (Rev. 3/06)
Page 18 of 21
Estoup A, Jarne P, Cornuet JM (2002) Homoplasy and mutation model at microsatellite loci and their
consequences for population genetics analysis. Molecular Ecology, 11, 1591–1604.
Frankham R (2005) Genetics and extinction. Biological Conservation, 126, 131-140.
Gold JR, Richardson LR (1998) Mitochondrial DNA diversification and population structure in fishes from the Gulf
of Mexico and western Atlantic. Journal of Heredity, 89, 404-414.
Guo SW, Thompon EA (1992) Performing the exact test of Hardy-Weinberg proportions for multiple alleles.
Biometrics, 48, 361-372.
Goudet J (1995) FSTAT (version 1.2): a computer program to calculate F-statistics. Journal of Heredity, 86, 485486.
Goudet J, Raymond M, de Meeus T, Rousset F (1996) Testing differentiation in diploid populations. Genetics,
144, 1933-1940.
Hedgecock D (1994) Does variance in reproductive success limit effective population sizes of marine organisms?
In: Genetic and Evolution of Aquatic Organisms (ed. Beaumont AR), pp. 122-134. Chapman & Hall,
London, UK.
Hedrick PW (1999) Perspective: Highly variable loci and their interpretation in evolution and conservation.
Evolution, 53, 313-318.
Hemmer-Hansen J, Nielsen EEG, Grønkjaer P, Loeschcke V (2007) Evolutionary mechanisms shaping the
genetic population structure of marine fishes; lessons from the European flounder (Platichthys flesus L.).
Molecular Ecology, 16, 3104-3118.
Hutchings JA (2000) Collapse and recovery of marine fishes. Nature, 406, 882-885.
Hutchinson WF, Carvalho GR, Rogers SI (2001) Marked genetic structuring in localised spawning populations of
cod Gadus morhua in the North Sea and adjoining waters, as revealed by mirosatellites. Marine Ecology
Progress Series, 223, 251-260.
Jones AG (2005) GERUD 2.0: a computer program for the reconstruction of parental genotypes from half-sob
progeny arrays with known or unknown parents. Molecular Ecology Notes, 5, 708-711.
Jones GP, Milicich MJ, Emslie MJ, Lunow C (1999) Self-recruitment in a coral reef fish population. Nature, 402,
802–804.
Knutsen H, Jorde PE, Andre C, Stenseth CHR (2003) Fine-scaled geographical population structuring in a highly
mobile marine species: the Atlantic cod. Molecular Ecology, 12, 385-394.
Mason JM (1962) The Scottish crab-tagging experiment 1960-1961. ICES Special Shellfish Symposium CM Nr.
18.
McKeown NJ, Shaw PW (2008) Polymorphic nuclear microsatellite loci for studies of brown crab , Cancer
pagurus L. Molecular Ecology Notes, 8, 653-655.
McQuinn IH (1997) Metapopulations and the Atlantic Herring. Reviews in Biology and Fisheries, 7, 297-329.
Mistakis M (1960) Movements of the edible crab (Cancer pagurus) in English waters. ICES CM Shellfish
Committee NO. 88.
Neff BD, Pitcher TE (2002) Assessing the statistical power of genetic analyses to detect multiple mating in fishes.
Journal of Fish Biology, 61, 739-750.
Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals.
Genetics, 89, 583.
Nesbo CL, Rueness EK, Iversen SA, Skagen DW, Jakobsen KS (2000) Phylogeography and population history of
Atlantic Mackerel (Scomber scombrus L.): a genealogical approach reveals genetic structuring among the
eastern Atlantic stocks. Proceedings of the Royal Society of London, 267, 281-292.
Nielsen EE, Hansen MM, Schmidt C, Meldrup D, Grønkjær P (2001) Determining the population of origin of
individual cod in the Northeast Atlantic. Nature, 413, 272.
Orensanz JM, Gallucci VF (1988) Comparative study of post-larval life-history schedules in four sympatric species
of Cancer (Decapoda: Barchyura: Cancridae). Journal of Crustacean Biology, 8, 197-220.
Orensanz JM, Parma AM, Armstrong DA, Armstrong J, Wardrup P (1995) The breeding ecology of Cancer
gracilis (Crustacea, Decapoda, Cancridae) and the mating system of cancrid crabs. Journal of Zoology,
235, 411-437.
Oresland V, Andre C (2008) Larval group differentiation in Atlantic cod (Gadus morhua) inside and outside the
Gullmar Fjord. Fisheries Research, 90, 9-16.
Palumbi SR (1994) Genetic divergence, reproductive isolation, and marine speciation. Annual Review of Ecology
and Systematics, 25, 547–572.
Rannala B, Mountain JL (1997) Detecting immigration by using multilocus genotypes. Proceedings of the National
Academy of Sciences USA, 94, 9197-9201.
Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and
ecumenicism. Journal of Heredity, 86, 248-249.
Rice WR (1989) Analysing tables of statistical tests. Evolution, 43, 223-225.
Rowe S, Hutchings JA (2003) Mating systems and the conservation of commercially exploited marine fish. Trends
in Ecology and Evolution, 18, 567-572.
Ruzzante DE, Taggart CT, Cook D (1998) A nuclear DNA basis for shelf- and bank-scale population structure in
Northwest Atlantic cod (Gadus morhua): Labrador to Georges Bank. Molecular Ecology, 7, 1663–1680.
Ryman N, Palm S (2006) POWSIM: a computer program for assessing statistical power when testing for genetic
differentiation. Molecular Ecology, 6, 600-602.
SID 5 (Rev. 3/06)
Page 19 of 21
Shaklee JB, Bentzen P (1998) Genetic identification of stocks of marine fish and shellfish. Bulletin of Marine
Science, 62, 589–621.
Shirley TC, McNutt L (1989) Precocious mating and trans-moult sperm retention by female Dungeness crabs.
American Zoologist, 29, 131A (Abstract).
Swearer SE, Caselle JE, Lea DW, Warner RR (1999) Larval retention and recruitment in an island population of a
coral-reef fish. Nature, 402, 799–802.
Taylor MS, Hellberg ME (2003) Genetic evidence for local retention of pelagic larvae in a Caribbean reef fish.
Science, 299, 107–109.
Toonen RJ (2004) Genetic evidence of multiple paternity of broods in the intertidal crab Petrolisthes cinctipes.
Marine Ecology Progress Series, 270, 259-263.
Ungfors A (2007) Sexual maturity of the edible crab (Cancer pagurus) in the Skagerrak and the Kattegat, based
on reproductive and morphometric characters. ICES Journal of Marine Science, 64, 318-327.
Waples RS (1998) Separating the wheat from the chaff: patterns of genetic differentiation in high gene flow
species. Journal of Heredity, 89, 439–450.
Ward RD, Woodwark M, Skibinski DOF (1994) A comparison of genetic diversity levels in marine, freshwater and
anadromous fishes. Journal of Fish Biology, 44, 213–232.
Winnepenninckx B, Backeljau T, DeWachter R (1993) Extraction of high-molecular-weight DNA from molluscs.
Trends in Genetics, 9, 407-407.
Wright S (1951) The genetical structure of populations. Annals of Eugenics, 15, 323-354.
SID 5 (Rev. 3/06)
Page 20 of 21
References to published material
9.
This section should be used to record links (hypertext links where possible) or references to other
published material generated by, or relating to this project.
Published papers
Ungfors A, McKeown NJ, Shaw PW, Andre C Lack of spatial genetic variation in the edible crab (Cancer
pagurus) in the Kattegat-Skagerrak area. Submitted to ICES J. Mar. Sci.
McKeown NJ, Shaw PW (2008) Single paternity within broods of the brown crab (Cancer pagurus L.); a
highly fecund species with long term sperm storage. Marine Ecology Progress Series, in press.
McKeown NJ, Shaw PW (2008) Polymorphic nuclear microsatellite loci for studies of brown crab, Cancer
pagurus. Molecular Ecology Notes, 8, 653-5.
Weetman D, Ruggiero A, Mariani S, Shaw PW, Lawler A & Hauser L (2007) Hierarchical population
genetic structure in the commercially exploited shrimp Crangon crangon identified by AFLP
analysis. Marine Biology,151, 565-575.
Weetman D, Hauser L, Bayes MK, Ellis JR & Shaw PW (2006) Genetic population structure across a
range of geographic scales in the commercially exploited marine gastropod, Buccinum undatum.
Marine Ecology Progress Series, 317, 157-169.
Presentation at conferences / meetings
Shaw PW, McKeown NJ (2008) MF0230 - Spatial and temporal genetic structuring of edible crab
populations. Defra Marine Fisheries Science Advisory Group 3rd Meeting, 8th May 2008.
Shaw PW, McKeown NJ (2007) Population genetics of the brown crab (Cancer pagurus L.) Oral
presentation to the ICES crab working group, Lowestoft, UK.
McKeown NJ, Shaw PW (2006) Population genetics of the edible crab (Cancer pagurus L.). Oral
presentation at the 41st European Marine Biology Symposium, Cork, Ireland.
Shaw PW, McKeown NJ (2006) Spatial and temporal genetic structuring of edible crab populations. MF02
Programme Review, “Shellfish population biology and assessment methodology”, 3rd Oct. 2006
Lowestoft
SID 5 (Rev. 3/06)
Page 21 of 21