Pronounced genetic structure and low genetic diversity in European
red-billed chough (Pyrrhocorax pyrrhocorax) populations
19/04/12
Abstract
The red-billed chough (Pyrrhocorax pyrrhocorax) is of conservation concern in the British Isles and continental Europe,
with historically declining populations and a highly fragmented distribution. We quantified the distribution of genetic
variation within and among European populations to identify isolated populations that may need to be managed as
demographically independent units, and assess whether individual populations are denuded of genetic diversity and so
may show reduced viability. We genotyped 326 choughs from ten wild populations and 22 from one captive population
at 16 nuclear microsatellite loci, and sequenced 34 individuals across three mitochondrial regions to quantify genetic
structure, diversity and phylogeography. Microsatellite diversity was low (often less than 4 alleles per locus), but pairwise
population differentiation was high (often Dest > 0.1), with a signature of isolation-by-distance. Bayesian-inferred a
posteriori genetic clusters coincided with a priori populations, supporting strong genetic structure. Microsatellites also
allowed us to identify the probable origin of the captive choughs and one recently founded wild population. Mitochondrial
DNA sequence diversity was low (π = 0.00103). Phylogeographic structure was consequently poorly resolved, but
indicated that sampled continental-European populations are ancestral to British Isles populations, which comprised a
single clade. Our data suggest that British Isles chough populations are relatively isolated with infrequent gene flow and
relatively genetically depauperate, potentially requiring genetic management. These findings should be integrated into
conservation management policy to ensure long-term viability of chough populations.
Marius A. Wenzel 1 , Lucy M. I. Webster 1 ,
Guillermo Blanco 2 , Malcolm D. Burgess 3 ,
Christian Kerbiriou 4 , Gernot Segelbacher 5 ,
Stuart B. Piertney 1,§ and Jane M. Reid 1,§,*
5
Department of Wildlife Ecology and Management,
University of Freiburg, Tennenbacher Str. 4, D-79106
Freiburg, Germany
* corresponding author. email address:
1
jane.reid@abdn.ac.uk
Institute of Biological and Environmental Sciences,
§ joint last authors
University of Aberdeen, Zoology Building, Tillydrone
Avenue, Aberdeen AB24 2TZ, UK
2
Department of Evolutionary Ecology, National
Museum of Natural History (MNCN-CSIC), c/ José
Gutierrez Abascal 2, 28006 Madrid, Spain
Introduction
1
Primary goals of conservation genetics are to quantify
2
demographic and genetic connectivity among and geCentre for Research in Animal Behaviour, College of netic diversity within populations of conservation conLife & Environmental Sciences, University of Exeter,
cern, consider the consequences for population viability
3
3
EX4 4QG, UK
4
5
and apply appropriate management action (Frankham,
6
1995, 2010a). Small, isolated populations can have
7
Muséum National d’Histoire Naturelle CERSP
increased extinction risk due to demographic, envi-
8
UMR 7204 MNHN-CNRS-UPMC, 61 rue Buffon,
ronmental and genetic stochasticity, whereas frequent
9
75005 Paris, France
dispersal and gene flow can counteract these stochas-
10
4
1
11
tic effects and decrease extinction risk (Lande, 1998;
cern with “amber status” (second most critical status)
60
12
Tallmon et al, 2004). Management intervention may in the United Kingdom (Eaton et al, 2009) due to
61
13
consequently be required to alleviate stochastic loss declining population sizes and contracting European
62
14
of genetic diversity and increase long-term adaptive distributions, particularly in the British Isles during
63
15
potential in small, isolated populations (Reed and the 19th and early 20th centuries (Holloway and Gib-
64
16
Frankham, 2003; Frankham, 2005, 2010b). Appropri-
bons, 1996). Its current Western European distribu-
65
17
ate translocation of wild individuals, or introduction tion is fragmented and restricted to coastal areas of the
66
18
of captive-bred individuals, can successfully increase
British Isles (the Scottish islands of Islay and Colon-
67
19
population viability in such cases (reviewed by Fischer say, the Isle of Man, Wales, Cornwall and Ireland) and
68
20
21
22
23
24
and Lindenmayer, 2000; Frankham, 2005). In this con- Brittany, and to parts of the Alps, Spain and Portutext, quantifying the pattern and degree of population gal (Monaghan, 1988; Carter et al, 2003; Johnstone
connectivity and genetic diversity can identify the pop- et al, 2011). Current published taxonomy recognises a
ulations and spatial scales on which conservation man- nominate Atlantic coast subspecies P. p. pyrrhocorax
agement may need to focus.
(British Isles and Brittany) and a Continental Euro-
69
70
71
72
73
28
Connectivity can be inferred from patterns of ge- pean subspecies P. p. erythrorhamphos (Vaurie, 1954;
netic structure and diversity within and among pop- Monaghan, 1988), although this distinction was based
ulations, assuming that weak genetic structure and on few morphological data from unverified museum
near parity in genetic diversity primarily reflect the specimens. The closely-related Alpine chough Pyrrho-
29
homogenising effect of gene flow (e.g. Nichols et al,
corax graculus occurs in mountain regions in Southern
78
30
2001; Segelbacher et al, 2003; Funk et al, 2007; Techow
and Central Europe, particularly the Alps (Delestrade
79
31
et al, 2010). Genetic structure and diversity are influ-
and Stoyanov, 1995).
80
32
enced by both recent and historic processes, so compre-
Multiple censuses of red-billed chough populations
81
33
hensive characterisation of demographic interactions were conducted across the British Isles and Brittany
82
34
and evolutionary relationships requires consideration
from 1963 to 2002 (Johnstone et al, 2007 and references
83
35
of multiple temporal and spatial scales. The distribu- therein). These suggested slight increases in most pop-
84
36
tion of variation in neutral nuclear markers, such as
ulation sizes after severe decreases prior to the 1950s
85
37
microsatellite length polymorphisms, indicates genetic
(Holloway and Gibbons, 1996).
Nevertheless, most
86
structure and diversity arising from contemporary connectivity (Balloux and Lugon-Moulin, 2002). These
populations remained small in 2002: Ireland held the
87
largest population (445–838 breeding pairs), followed
88
25
26
27
38
39
40
41
42
43
patterns can be used to consider the need to translo- by Wales (228–262 pairs), Isle of Man (128–150 pairs),
cate individuals among wild or captive-bred popula- Scotland (71–83 pairs, including 56–64 on Islay), Brittions and identify appropriate source populations and tany (48–58 pairs) and England (Cornwall) and Norththe origin of recent natural colonisation events (IUCN, ern Ireland (Rathlin) with only one pair each. Since
74
75
76
77
89
90
91
92
45
1998; Frankham, 2008, 2010a). In contrast, genetic the last UK-wide census in 2002, the population on Isstructure inferred from mitochondrial DNA sequence lay declined to c. 45 breeding pairs (Reid et al, 2011).
46
variation reflects long-term demographic processes as-
These small and decreasing population sizes are caus-
95
47
sociated with historic geological events such as tectonic
ing heightened conservation concern (Kerbiriou et al,
96
48
movement of land masses, floods or glaciation (Taber-
2005; Johnstone et al, 2007).
97
49
let et al, 1998; Hewitt, 2000). Phylogeographic anal-
44
Most European populations are the focus of some
93
94
98
100
52
ysis of mitochondrial sequence variation (Avise et al, degree of conservation action and demographic study,
1987) can elucidate evolutionary heritage among pop- involving monitoring of breeding success, survival and
ulations, clarify taxonomic uncertainties and identify movements of colour-ring marked individuals. This
53
evolutionarily significant units (ESUs; Moritz, 1994)
work has identified intrinsic and extrinsic constraints
102
54
for the management of evolutionary diversity in cryptic
on population growth rate (e.g. Blanco et al, 1998a;
103
55
species complexes, subspecies and ecologically isolated
Kerbiriou et al, 2006; Reid et al, 2004, 2006, 2008),
104
56
populations (e.g. Burbrink et al, 2000; Hebert et al,
and highlighted the key role of human impacts in the
105
57
2004; Segelbacher and Piertney, 2007).
chough’s decline, involving historic persecution (Mon-
106
50
51
58
59
The red-billed chough (Pyrrhocorax pyrrhocorax, aghan, 1988; Carter et al, 2003), contemporary tourism
Corvidae) is a Species of European Conservation Con- pressure (Kerbiriou et al, 2009) and agricultural land2
99
101
107
108
109
use change (Blanco et al, 1998b; Whitehead et al, 2005; therefore suitability for release into wild populations
110
Kerbiriou et al, 2006).
(IUCN, 1998; Frankham, 1995, 2010a).
158
159
111
Colour-ring resightings also indicate that choughs
To provide the genetic information required to in-
160
112
in northwestern Europe are typically sedentary and
form chough conservation management policy, we con-
161
113
philopatric as long-distance dispersal between popu-
ducted a large-scale analysis of genetic structure, ge-
162
115
lations is very rarely observed (Carter et al, 2003; Reid netic diversity and phylogeography across British Isles
et al, 2003, 2008; Moore, 2008). Nevertheless, oc- chough populations and a sample of populations from
116
casional long-distance movements are observed, most
114
Continental Southwestern Europe. Our objectives were
118
notably between North Wales and the Isle of Man to 1) quantify genetic differentiation among and geduring 1997–2004 (c. 100 km; Moore, 2006, 2008). netic diversity within populations using microsatellite
119
Furthermore, unringed choughs of unknown origin re-
117
120
121
loci (Wenzel et al, 2011); 2) infer the phylogeographic
colonised Cornwall in 2001 after the chough had been structure of the sampled populations from nucleotide
extinct there since at least 1973 (Carter et al, 2003). variation across mitochondrial DNA regions; and 3)
123
Aided by nest protection and habitat management, this identify the likely origins of the choughs that recently
small population has persisted since and comprised five recolonised Cornwall and of the ancestors of the captive
124
breeding pairs in 2011 (Johnstone et al, 2011). The
125
colonisers are speculated to have originated from the
126
nearest wild populations in Wales or Brittany (Carter
127
et al, 2003). This has not been proven, but is of con-
128
siderable interest in the context of future genetic man-
129
130
122
131
132
133
140
141
142
143
144
145
176
Overall, it remains unclear whether long-distance red-billed chough populations at eleven locations across
dispersal is as rare as suggested by ringing studies, the British Isles and Continental Europe, including a
or occurs more frequently but goes undetected by di- single sample from the sole breeding pair in Northern
The low observed dispersal rates
Ireland (Figure 1). This single sample is not useful
for estimation of genetic diversity and differentiation
the possibility that many or all remaining populations for Northern Ireland, but inclusion in phylogeographic
have low and declining genetic diversity, potentially analysis can indicate evolutionary relationships with
constituting an additional threat to population per- other populations and inform management decisions.
sistence that conservation management has not yet In addition, 22 samples were collected from the captive
identified and integrated into priorities. Genetic di- population at Paradise Park. Finally, one sample each
versity has not been comprehensively quantified across was also collected from Alpine choughs (P. graculus)
all relevant chough populations and molecular mark-
in the French Alps and Corsica to use as a phylogeoers, with only two previous small-scales studies (Mon- graphic outgroup.
aghan, 1988; Kocijan and Bruford, 2011). If genetic
Samples were obtained non-invasively and oppordiversity within the British Isles populations is indeed tunistically from moulted feathers, bones, legs or liver
148
considered (Burgess et al, in press), taking into ac-
shells and membranes from nests, avoiding sampling
of known full siblings. The Alpine choughs were blood
count genetic compatibility between source and target sampled. Samples were collected over several years for
population (Frankham, 2010a). For this potential pur- most populations (Table 1).
152
pose, a captive chough population has been sustained
in Paradise Park Wildlife Sanctuary, Cornwall (here-
153
after: “Paradise Park”) since the late 1970s (Burgess
156
157
172
A total of 327 DNA samples were collected from wild
or release of captive-bred individuals, may need to be
155
171
et al, 2011).
147
154
170
175
samples from choughs found dead, or remnant egg-
151
169
Sample collection
low, translocation of individuals among populations,
150
168
agement of the small Cornish population (Johnstone
146
149
167
174
among the small remaining chough populations raise
139
166
Materials and Methods
135
138
165
173
rect observation.
137
164
Paradise Park population.
134
136
163
DNA extraction
et al, in review). Documentation and anecdote suggest DNA was extracted from a 3-5 mm clipping of the lower
that at least some ancestors of the captive population feather calamus, or scrapings of bone/leg tissue, shreds
came from North Wales (Burgess et al, in press). How- of liver tissue, fragments of egg-shell and membrane,
ever, some uncertainty remains over their origin and or 50 µl of well-mixed blood, using Proteinase K diges3
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
Figure 1: Sampling locations for red-billed chough populations classified by published taxonomy as nominate
Atlantic coast subspecies Pyrrhocorax pyrrhocorax pyrrhocorax (black circles). For comparison, two Continental
European populations (subspecies P. p. erythrorhamphos; grey circles) and a captive population at Paradise
Park Wildlife Sanctuary, Cornwall, were also sampled.
202
tion, ammonium acetate precipitation of cell debris and Sigma-Aldrich) and 20-500 ng of template DNA. An
219
203
DNA recovery by ethanol precipitation as described in initial denaturation step at 95 ºC for 5 min was followed
220
204
Hogan et al (2008). DNA quality and quantity were
by 30 cycles of annealing at 49 ºC for 30 s, elongation at
221
205
assessed with a NanoDrop ND-1000 spectrophotome- 72 ºC for 30 s and denaturation at 95 ºC for 30 s, a final
222
206
ter.
annealing step at 49 ºC for 1 min and a final elongation
223
step at 72 ºC for 5 min. PCR products were checked
224
207
and scored on 3 % agarose-TBE gels run at 6 V cm
Molecular sexing
208
In order to test whether DNA was of sufficient quality
209
for genotyping PCR (Hogan et al, 2008), PCR-based
-1
225
and stained with WebGreen DNA stain.
226
Microsatellite genotyping
227
211
sex determination was attempted for all individuals, using the P2 (5’-TCTGCATCGCTAAATCCTTT-
212
3’) and P8 (5’-CTCCCAAGGATGAGRAAYTG-3’)
213
primers (Griffiths et al, 1998). PCRs were performed billed chough (Wenzel et al, 2011). A subset of 31
230
214
in an MJ Research PTC-100 or Thermo Hybaid Px2
individuals, selected to cover the entire sampled geo-
231
215
thermocycler. The total reaction volume was 20 µl and graphic range and as many different alleles as possi-
232
216
contained 2.5 mM MgCl2 , 16 mM (NH4 )2 SO4 , 67 mM
ble, was genotyped twice to estimate genotyping error
233
217
Tris-HCl, 0.2 mM of each nucleotide, 0.5 µM of each
rates. PCRs were performed in simplex following Wen-
234
218
primer, 0.5 U of Taq DNA polymerase (Bioline or
zel et al (2011), but using TouchDown gradients from
235
210
All individuals were genotyped at 16 microsatellite loci
(Ppy-001 to Ppy-016) developed specifically for red-
4
228
229
Table 1: Collection years and total and genetically sexed (male, female or unknown) sample sizes of presumed
a priori red-billed chough populations.
Population
Collection years Total Male Female Unknown
Colonsay
Islay
Isle of Man
Northern Ireland
South Ireland a
North Wales
South Wales
Cornwall (wild)
Brittany
French Alps
Spain
Paradise Park (captive)
a
2005–2011
2004–2011
2004–2011
2010
2010
2009–2011
2011
2003–2011
2005–2010
2008–2010
2010
2003–2011
40
77
41
1
26
73
11
9
18
14
17
22
Total
349
Beara and south coast; hereafter “Ireland”
15
29
23
–
9
29
6
1
7
1
4
11
6
13
3
–
5
5
–
5
2
6
2
2
165
135
49
Genetic differentiation
268
either 6-FAM, HEX, NED or PET, and genotypes were
Global and pairwise genetic differentiation among
269
resolved on an automatic ABI 3730 Capillary DNA se-
eleven a priori red-billed chough populations (includ-
270
236
60 ºC to 50 ºC for all loci except for locus Ppy-007,
237
where a 55 ºC to 45 ºC gradient was used. The 5’ end
238
of each forward primer was fluorescently labelled with
239
240
19
35
15
1
12
39
5
3
9
7
11
9
242
quencer (DNA Sequencing & Services, MRCPPU, Col- ing Paradise Park but excluding the single Northern
lege of Life Sciences, University of Dundee, Scotland, Ireland sample) was estimated using the statistics D
243
(Jost, 2008) and FST (Wright, 1951). The software
273
spade (Chao and Shen, 2010) was used to calculate
274
an adjusted estimator for global and pairwise D (Dest )
275
with 95 % confidence intervals constructed from 1,000
276
bootstrap replicates using a percentile method and re-
277
241
www.dnaseq.co.uk).
244
Genotypes were scored by eye using genemarker
245
1.4 (SoftGenetics). The dataset was checked for geno-
246
typing errors and to estimate null-allele frequencies per
247
population using microchecker 2.2.3 (van Ooster-
248
hout et al, 2004). gimlet 1.3.3 (Valiere, 2002) was
249
used to calculate the unbiased probability that two un-
250
related individuals drawn at random from each popu-
251
lation (or the overall dataset) will have the same geno-
252
type (probability of identity PID ; Waits et al, 2001).
253
254
255
centering (Chao and Shen, 2010). Global and pairwise FST estimates (Weir and Cockerham, 1984) were
These probabilities were used to screen the dataset for
duplicate samples from the same individual (genotypegrouping function in gimlet), which were removed.
271
272
278
279
calculated in fstat 2.9.3.2 (Goudet, 1995, 2002) with
280
a 95 % CI for global FST constructed from 15,000
281
bootstrap replicates over loci and significance tests for
282
pairwise FST performed by randomising multi-locus
283
genotypes between each population pair (1100 permu-
284
tations; strict Bonferroni-corrected significance level
285
α = 0.00091).
286
256
Observed (HO ) and expected (HE ) heterozygos-
257
ity at each locus were calculated in genalex 6.4
Both Dest and FST pairwise estimates of popu-
287
258
(Peakall and Smouse, 2006). Using an MCMC ap-
lation differentiation (excluding Paradise Park) were
288
259
proach (1000 dememorisations, 100 batches, 1000 itera-
then used to test for isolation by distance (Wright,
289
260
tions), genepop 4.0.10 (Raymond and Rousset, 1995;
261
Rousset, 2008) was used to test for deviations from
1943; Slatkin, 1993) using the software ibd 1.52 (Bohonak, 2002). A Mantel test with 1,000 randomisa-
262
Hardy-Weinberg equilibrium per locus by performing
tions was performed to test for correlation between
2
264
global χ tests across population-specific FIS (Wright, Dest or FST /(1–FST ) and logarithmic Euclidean geo1951) estimates (Fisher’s method) and to test for link- graphic distance as proposed for two-dimensional habi-
265
age disequilibrium between each of 120 locus combi-
263
266
nations
267
tests).
( 21
· 16 · 15) in each of 11 population (= 1320
5
290
291
292
293
294
tat (Rousset, 1997). A linear regression line was con-
295
structed using a Reduced Major Axis (RMA) method
296
(Hellberg, 1994).
297
298
Bayesian inference of genetic structure
299
The software structure 2.3.3 (Pritchard et al, 2000;
300
Falush et al, 2003) was used to implement Bayesian
301
302
303
Mitochondrial DNA sequencing
344
A 1,205 bp segment of the mitochondrial control re-
345
Markov Chain Monte Carlo (MCMC) inference of a gion was amplified in three individuals per population
posteriori genetic clusters to detect any cryptic genetic (chosen to represent a broad geographic area within
structure unidentified by the assumed a priori popula- populations) using the primers JCR03 (Saunders and
305
tions (Mank and Avise, 2004). The number of assumed Edwards, 2000) and H1248 (Tarr, 1995). The single ingenetic clusters (K) was set from 1 to 11, and 15 runs dividual from Northern Ireland was included, as were
306
were performed for each K with 200,000 MCMC iter-
307
ations (a precursory burn-in of 10,000 iterations was
308
found sufficient) using the admixture ancestry model
304
two Alpine choughs as an outgroup.
PCRs were performed in a G-Storm GS1 or MJ
310
with correlated allele frequencies. The full analysis was Research PTC-100 thermocycler. The total reacthen repeated with the same parameters, but also in- tion volume was 25 µl and contained 2.5 mM MgCl2 ,
311
cluding a priori sampling locations as prior information
309
16 mM (NH4 )2 SO4 , 67 mM Tris-HCl, 0.2 mM of each
346
347
348
349
350
351
352
353
354
355
313
(LOCPRIOR setting) to detect any further structure nucleotide, 0.5 µM of each primer, 0.625 U of Taq DNA
unidentified by the standard model (Hubisz et al, 2009; polymerase (Bioline or Sigma-Aldrich) and 50-200 ng of
314
358
315
Barlow et al, 2011). To test for spurious results caused
by individuals with missing genotype data, all analyses
template DNA. A denaturation step at 95 ºC for 2 min
was followed by 20 TouchDown cycles from 60 ºC to
359
316
were repeated after excluding individuals with partially
50 ºC in 0.5 ºC decrements (denaturation at 95 ºC for
360
317
missing data.
312
45 s, annealing for 45 s, elongation at 72 ºC for 1 min),
361
362
320
321
Replicate runs for each K were aligned and averaged in
319
322
323
324
325
326
products were checked on 1 % agarose-TBE gels stained
-1
clumpp 1.1.2 (Jakobsson and Rosenberg, 2007), using with WebGreen DNA stain (run at 9 V cm ) and puthe Greedy alignment algorithm with 10 randomised in- rified using the QIAquick PCR Purification Kit (QIput orders, and visualised using distruct 1.1 (Rosen- AGEN) according to the manufacturer’s instructions.
berg, 2004).
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
363
364
365
366
367
368
Using the same primers, DNA sequencing was per-
369
formed by Eurofins MWG GmbH, Ebersberg, Germany
370
or Beckman Coulter Genomics, Takeley, UK.
371
Genetic diversity
In addition, two mitochondrial protein coding re-
327
357
15 standard cycles (denaturation at 95 ºC for 45 s, an-
structure harvester 0.6.7 (Earl, 2011) was used
to collate the results and infer the statistically best sup- nealing at 50 ºC for 45 s, elongation at 72 ºC for 1 min)
ported K using the ΔK statistic (Evanno et al, 2005). and a final elongation step at 72 ºC for 10 min. PCR
318
356
Genetic diversity was calculated per population using gions were PCR amplified using primers designed
a variety of statistics. Mean allele numbers and al- in primer3 (Rozen and Skaletsky, 2000) based
lelic richness (allele numbers rarefacted to a minimum on conserved regions of the consensus sequence of
sample size of 4 across all populations in the dataset; three mitochondrial genomes of species closely related
Mousadik and Petit, 1996) were calculated in fstat. to red-billed chough (retrieved from genbank usAllele frequencies as calculated by fstat were used to ing the Basic Local Alignment Search Tool blast
count private alleles and to calculate the effective num- [www.ncbi.nlm.nih.gov/blast/]: rook Corvus frugilegus
ber of alleles per population (Kimura and Crow, 1964; accession Y18522, Hume’s ground-tit Pseudopodoces
Jost, 2008). Observed (HO ) and expected (HE ) het- humilis accession HM535648, and Eastern Orphean
erozygosity were calculated in genalex.
FIS was calculated per population and tested for
373
374
375
376
377
378
379
380
381
warbler Sylvia crassirostris accession NC_010229).
382
Fragment CHMT06 corresponded to a 922 bp segment
383
statistical significance by randomising alleles within of the NADH1 gene; and fragment CHMT17 contained
populations (3520 randomisations; strict Bonferroni- the final 612 bp of the NADH5 gene, a 9 bp non-coding
corrected significance level α = 0.00028) in fstat in segment and the first 607 bp of the CYTB gene. PCR
order to identify deviations from Hardy-Weinberg equi- amplification conditions were the same as described
librium and potential substructuring within popula- above, but with different TouchDown temperature gradients (Appendix Table 5).
tions (Wahlund, 1928).
6
372
384
385
386
387
388
389
390
Inference of phylogeography
391
Sequences were checked by eye and then aligned in
392
mega4. Resolved haplotypes were deposited in gen-
393
bank for each fragment separately.
394
sequences of the three fragments were concatenated
395
into one aligned dataset for phylogeographic analyses.
396
Overall haplotype diversity (h) and nucleotide diversity
397
(π) were calculated in dnasp v5 (Librado and Rozas,
398
2009). A statistical parsimony haplotype network was
399
constructed using tcs v1.21 (Clement et al, 2000).
The ingroup
400
The software jmodeltest 0.1.1 (Guindon and Gas-
401
cuel, 2003; Posada, 2008) was used to find the optimal
for 147 out of 1320 loci combinations in 11 populations,
435
but in no case was any combination out of equilibrium
436
consistently across all populations, suggesting no phys-
437
ical linkage of loci (results not shown).
438
Evidence was found for allelic drop-out at some loci
439
from replicate genotyping of 31 individuals. Of 496
440
replicated genotypes (31·16 loci), seven cases (= 1.4 %)
441
occurred where either the original or the replicate geno-
442
type was heterozygous whereas the other was homozy-
443
gous. In these cases, the heterozygote genotype was re-
444
tained. Occurrence of allelic drop-out was not system-
445
atic for particular loci or populations and restricted to
446
412
individuals where PCR quality was low overall, probof 88 models of nucleotide evolution for the sequence
ably caused by contamination of the template DNA
data (including outgroup sequences) using the Akaike
extract as apparent from a low spectrophotometric
information criterion (AIC; Akaike, 1974). The opti260 nm : 230 nm ratio in these cases.
mal model (ln likelihood = –5632.84; AIC = 11415.67)
The probability of identity (PID ) for two individuwas defined as HKY+G (Hasegawa-Kishino-Yano +
als drawn at random from the final dataset (348 indigamma rate distribution) with base frequencies A =
viduals) decreased from 9.04 · 10−2 (most informative
0.2977, C = 0.2889, G = 0.1339 and T = 0.2795, tranlocus Ppy-011) to 2.53 · 10−10 (all 16 loci), indicatsition/transversion ratio = 6.8537 and gamma shape
ing a high power to discriminate between individuals.
= 0.0140. This model was used for a Maximum LikeWithin populations, the highest PID was observed in
lihood analysis implemented in paup* 4.0b10 (SwofColonsay and decreased from 1.96 · 10−1 to 6.60 · 10−6 .
ford, 2000), using a heuristic search with tree bisection
413
and reconnection (TBR) as the branch-swapping al-
414
gorithm. Bootstrapping was performed 10,000 times
415
using the Neighbour-Joining method on the same evo-
416
lutionary model.
402
403
404
405
406
407
408
409
410
411
417
418
419
Results
Characterisation of microsatellite loci
The number of alleles per microsatellite locus ranged
426
451
452
453
454
455
456
457
459
billed chough populations was Dest = 0.241 (95 % CI:
460
0.222, 0.259) and FST = 0.208 (95 % CI: 0.179, 0.245).
461
Pairwise Dest and FST estimates were highly signif-
462
icantly correlated (r = 0.70; p < 0.001) and were
463
greater than 0.10 for most population pairs (Table 2).
464
Cases of weak differentiation were Islay vs. Colonsay
465
(Dest = 0.017; FST = 0.047), Cornwall vs. Ireland
466
heterozygosity ranged from 0.05 to 0.66 and 0.07 to
425
450
Global genetic differentiation among all eleven red-
422
423
449
458
421
424
448
Genetic differentiation
from three (locus Ppy-015) to 14 (loci Ppy-010) (Ap- (Dest = 0.020; FST = 0.053) and Paradise Park vs.
pendix Table 6). Observed (HO ) and expected (HE ) North Wales (Dest = 0.065; FST = 0.069). The only
420
447
non-significant Dest estimate was Cornwall vs. Ireland
0.71, respectively. Significant deviations from Hardy- (95 % CI: 0.000, 0.080; bounded by zero). All FST esWeinberg equilibrium (α = 0.05) based on pooled timates were significant at the 5 %-level, but some espopulation-specific FIS estimates were found in loci timates involving populations with small sample sizes
Ppy-003, Ppy-005, Ppy-008, Ppy-012 and Ppy-016 were not significant after strict Bonferroni correction
(Table 2).
467
468
469
470
471
472
473
427
(Appendix Table 6). Heterozygote deficiency identified
428
by microchecker suggested that null alleles might be
There was a highly significant correlation between
475
429
present at some of these loci (Appendix Table 6; neg-
geographic distance and genetic differentiation both for
476
430
ative null-allele frequencies are a software artefact and Dest (r = 0.81; p < 0.001) and FST (r = 0.59; p <
477
431
can be interpreted as zero). However, this was not con- 0.001). The RMA regression lines for Dest and FST ex-
478
432
sistent across populations for any locus, suggesting that plained 65.0 % and 35.3 % of the variation respectively
479
433
heterozygote deficiency was not due to null-alleles. Sig-
(Figure 2). When the Continental European popula-
480
434
nificant linkage disequilibrium (α = 0.05) was detected tions Spain and French Alps were removed, the corre-
481
7
474
land and the Isle of Man) and a southern group (Ire-
510
land, Wales, Cornwall, Brittany and Paradise Park).
511
At K = 4, Isle of Man became separated from Scot-
512
land, and the southern group became subdivided into
513
Ireland, Cornwall and Brittany versus Wales and Par-
514
adise Park (Figure 3). At K = 5, Brittany became
515
separated from Ireland and Cornwall. At higher K,
516
the delineation of genetic clusters coincided well with
517
the a priori populations.
518
A small number of individuals were assigned to a
519
different cluster to that of most other individuals in
520
their a priori population, using the standard admix-
521
ture model. However, most of these cases were not
522
apparent in the LOCPRIOR models. Overall, no difFigure 2: Relationships between geographic distance
and genetic differentiation (isolation by distance), us- ferences in cluster distribution at any K or the best
ing Dest (dots, dotted line) and FST /(1–FST ) (trian- supported number of clusters were observed when indigles, dashed line).
viduals with partially missing genotypes were excluded.
482
lations for both Dest (r = 0.50; p = 0.007) and FST
483
(r = 0.52; p = 0.002) were still significant and the re-
484
gression lines explained 25.0 % and 27.0 % respectively.
485
486
Bayesian inference of genetic structure
487
Based on the ΔK statistic, the best supported number
488
of a posteriori genetic clusters was K = 3 for the stan-
489
dard admixture model and K = 2 for the LOCPRIOR
490
model (ΔK = 73 and 83 respectively; Appendix Table
491
7). For K = 3, the first cluster comprised Spain and
492
the French Alps, the second cluster comprised Ireland,
493
Wales, Cornwall, Brittany and Paradise Park, and the
494
third cluster comprised Scotland and the Isle of Man
495
(Figure 3).
523
524
525
526
Genetic diversity
527
Table 3 summarises the genetic diversity statistics for
528
each a priori population. The Continental European
529
populations Spain and French Alps had highest diver-
530
sity and the northernmost populations Colonsay, Islay
531
and the Isle of Man had lowest diversity. Ireland and
532
Wales had highest diversity in the British Isles. De-
533
viations from Hardy-Weinberg equilibrium (heterozy-
534
gote deficiency) were apparent in Colonsay (FIS =
535
0.131), Ireland (FIS = 0.130) and the French Alps (FIS
536
= 0.167) at the 5 % level, but only the latter value
537
remained significant after strict Bonferroni correction
538
(Table 3).
539
540
Phylogeography
541
(subspecies P. p. erythrorhamphos sensu Vaurie, 1954)
A total of 3,355 base pairs could be resolved un-
542
498
were so strongly differentiated from all other popu-
ambiguously across three PCR amplicons in in-
543
499
lations (subspecies P. p. pyrrhocorax sensu Vaurie, group sequences.
500
1954) that more subtle genetic structure among these morphic with only two transversions:
G↔T at
545
501
other populations may not have been detected. When site 403 (control region) and T↔A at site 1,474
546
502
Spain and the French Alps were excluded from the
(NADH1). The polymorphic sites defined ten haplo-
547
503
analysis to clarify genetic structure within the remain-
types, with haplotype diversity h = 0.750 ± 0.068 SD
548
496
However, the Spain and French Alps populations
497
504
505
506
Of these, 19 sites were poly-
ing nine populations (running K = 1 to 9), the best sup- and overall nucleotide diversity π = 0.00103 ±
ported number of clusters was K = 2 (ΔK = 534 and 0.00019 SD (Table 4). These haplotypes are stored
108 respectively; Appendix Table 7), but with a strong in genbank at accessions JQ924832–JQ924841 (con-
544
549
550
551
507
secondary peak at K = 4 (ΔK = 180 and 46 respec-
trol region), JQ924842–JQ924851 (CHMT06) and
552
508
tively; Appendix Table 7). The two main clusters di-
JQ924852–JQ924861 (CHMT17). The resolved Alpine
553
509
vided the geographic range into a northern group (Scot-
chough outgroup sequences for the three fragments
554
8
9
a
0.148
(0.114, 0.189)
0.113
(0.091, 0.136)
0.186
(0.129, 0.250)
0.172
(0.104, 0.239)
0.156
(0.113, 0.199)
0.533
(0.460, 0.605)
0.506
(0.450, 0.565)
0.219
(0.171, 0.267)
(0.155, 0.281)
0.213
(0.141, 0.284)
0.138
(0.096, 0.180)
0.552
(0.480, 0.621)
0.494
(0.434, 0.553)
0.218
(0.172, 0.266)
0.121
(0.097, 0.147)
0.047**
Islay
0.172
(0.132, 0.213)
0.114
(0.089, 0.142)
0.217
0.017
(0.003, 0.031)
0.099
(0.073, 0.130)
interval bounded by zero
Paradise Park
Spain
French Alps
Brittany
Cornwall
South Wales
North Wales
Ireland
Isle of Man
Islay
Colonsay
Colonsay
(0.498, 0.637)
0.493
(0.434, 0.549)
0.244
(0.194, 0.293)
(0.138, 0.268)
0.245
(0.178, 0.322)
0.236
(0.190, 0.288)
0.568
0.195
(0.154, 0.239)
0.171
(0.139, 0.205)
0.202
0.205**
0.177**
Isle of Man
(0.352, 0.504)
0.385
(0.322, 0.451)
0.191
(0.141, 0.244)
(0.074, 0.195)
0.020
(0.000, 0.080)a
0.156
(0.105, 0.209)
0.430
0.103
(0.071, 0.139)
0.130
0.247**
0.227**
0.232**
Ireland
(0.355, 0.495)
0.428
(0.371, 0.485)
0.065
(0.034, 0.103)
(0.075, 0.181)
0.096
(0.044, 0.157)
0.146
(0.109, 0.189)
0.426
0.126
0.101**
0.191**
0.150**
0.144**
North Wales
(0.365, 0.535)
0.404
(0.318, 0.490)
0.141
(0.081, 0.214)
0.162
(0.079, 0.251)
0.221
(0.150, 0.294)
0.452
0.126*
0.137*
0.256*
0.261**
0.274**
South Wales
(0.374, 0.561)
0.453
(0.359, 0.541)
0.191
(0.125, 0.265)
0.153
(0.078, 0.238)
0.468
0.215*
0.127**
0.053*
0.352*
0.312**
0.332**
Cornwall
(0.409, 0.567)
0.417
(0.349, 0.486)
0.184
(0.132, 0.240)
0.491
0.206*
0.221*
0.148**
0.175**
0.296**
0.241**
0.214**
Brittany
0.230
(0.150, 0.309)
0.434
(0.357, 0.506)
0.287*
0.252*
0.239*
0.248**
0.254**
0.409**
0.430**
0.406**
French Alps
0.430
(0.362, 0.496)
0.088**
0.228**
0.219*
0.199*
0.228**
0.209**
0.339**
0.380**
0.344**
Spain
0.217 **
0.240**
0.193**
0.212**
0.146*
0.069**
0.184**
0.284**
0.281**
0.270**
Paradise Park
Table 2: Pairwise genetic differentiation among eleven a priori red-billed chough populations based on 16 microsatellite loci. Jost’s Dest with 95 % confidence intervals is
given below the diagonal, Wright’s FST with annotated significance at the 5 % level (*) and strict Bonferroni-corrected level (α = 0.00091**) is given above the diagonal.
Figure 3: Individual membership coefficients derived from Bayesian inference of genetic structure across all
eleven red-billed chough populations (top four plots) and Atlantic coast populations only (bottom six plots).
Each individual is represented by a single vertical line. Black lines demarcate a priori populations. Coefficients
are averaged across 15 replicate runs or from the single most likely replicate for K = 5, due to multiple solutions
among replicates, using the standard admixture model or including sampling locations as prior information
(LOCPRIOR).
555
are stored at JQ963890–JQ963892 (French Alps) and levels of polymorphism.
556
JQ963893–JQ963895 (Corsica).
557
A statistical parsimony network of ingroup haplo-
558
types illustrates two major haplotype groups: Conti-
569
Discussion
570
We quantified genetic structure, genetic diversity and
571
560
nental Europe (Spain, French Alps and Brittany) and
the British Isles, diverged by five transitions (Figure 4).
phylogeography among red-billed chough populations
572
561
A Maximum Likelihood phylogram with Alpine chough across the British Isles in comparison to a sample
573
562
as outgroup defined two clades, separating the Conti-
of Continental European populations, in order to in-
574
563
nental European populations Spain, French Alps and
fer population connectivity, identify management units
575
564
Brittany from all populations in the British Isles (Fig-
and assess the potential need for management interven-
576
565
ure 5). Within the British Isles, a further lineage was tion to increase genetic diversity. Our microsatellite
577
566
apparent, consisting of Ireland, Cornwall and South
loci were robust and provided a dataset with high res-
578
567
Wales (two individuals only). None of these major
olution to identify individuals within populations and
579
568
groups were bootstrap supported, reflecting low overall detect significant genetic differentiation among a priori
580
559
10
Table 3: Genetic diversity statistics (means ± 1 SD) derived from 16 microsatellite loci across 348 red-billed
choughs from eleven populations. Population size (n) is given alongside the average percentage of missing
genotype data, number of alleles (na ), allelic richness (ar ), effective number of alleles (ne ), number of private
alleles (np ), observed heterozygosity (HO ), expected heterozygosity (HE ) and Wright’s FIS with significance
indicated at the 5 % level (*) and strict Bonferroni-corrected level (α = 0.00028**).
Population
n
Missing data (%)
na
Colonsay
Islay
Isle of Man
Ireland
North Wales
40
77
41
26
73
4.85
6.51
5.95
6.31
7.18
2.88
3.13
3.13
3.63
3.38
±
±
±
±
±
1.15
1.31
1.50
1.50
1.54
1.95
1.97
1.98
2.53
2.46
±
±
±
±
±
0.58
0.67
0.67
0.89
0.79
1.60
1.62
1.63
2.16
2.29
±
±
±
±
±
South Wales
Cornwall
Brittany
French Alps
Spain
Paradise Park
11
9
18
14
17
22
6.55 ± 6.27
19.78 ± 25.04
3.67 ± 4.67
13.07 ± 18.43
4.35 ± 10.22
3.86 ± 8.77
2.69
2.25
2.81
4.88
6.38
2.81
±
±
±
±
±
±
1.01
0.77
1.17
1.67
2.55
1.05
2.34
2.05
2.23
3.55
4.11
2.38
±
±
±
±
±
±
0.70
0.62
0.70
0.86
1.17
0.67
2.03
1.75
1.96
3.33
4.51
2.15
±
±
±
±
±
±
Total
348
6.57 ± 11.58
–
±
±
±
±
±
9.01
10.63
10.85
10.30
12.48
ar
ne
–
–
np
HO
HE
FIS
0.52
0.62
0.55
0.95
0.83
1
2
4
1
1
0.30
0.40
0.44
0.52
0.40
±
±
±
±
±
0.05
0.04
0.05
0.03
0.05
0.33
0.49
0.58
0.65
0.52
±
±
±
±
±
0.06
0.05
0.04
0.04
0.05
0.131*
0.024
0.019
0.130*
–0.009
0.63
0.55
0.74
1.30
1.92
0.61
0
0
1
5
28
0
0.49
0.45
0.45
0.49
0.42
0.37
±
±
±
±
±
±
0.06
0.06
0.06
0.06
0.07
0.06
0.51
0.45
0.47
0.46
0.45
0.39
±
±
±
±
±
±
0.06
0.06
0.06
0.06
0.06
0.06
–0.078
–0.031
–0.040
0.167**
–0.038
–0.063
–
–
–
–
581
populations. Sequencing large portions of three mito-
among red-billed chough populations was therefore no-
611
582
chondrial regions provided good characterisation of mi-
tably high and demonstrates strong genetic structure.
612
583
tochondrial polymorphism and hence phylogeographic
584
585
Genetic differentiation between population pairs was
structure. We demonstrate strong genetic differentia- strongly correlated with geographic distance; the lattion among most populations, low nuclear and mito- ter explained 25–65 % of the variation in the former.
613
614
615
586
chondrial genetic diversity, and weak phylogeographic
Geographic distance rarely explains more than 20 %
616
587
structure across the sampled populations.
of variation in genetic differentiation in bird popu-
617
lations (e.g. Johnson et al, 2003; Funk et al, 2007;
618
588
Genetic structure and dispersal
Techow et al, 2010). Notable exceptions include 27 %
619
589
Genetic differentiation is generally deemed moderately
in European shags (Barlow et al, 2011) and 38.4 % in
620
590
high when Dest or FST is greater than 0.10–0.15 (Bal-
orange-crowned warblers Vermivora celata in Canada
621
591
loux and Lugon-Moulin, 2002). The observed differen-
592
593
594
595
596
597
598
599
and Alaska across a large spatial scale of up to 4,000 km
tiation among most red-billed chough population pairs, (Bull et al, 2010). Genetic structure among chough
separated by up to 1,700 km, exceeded 0.10. This is populations was apparent even on a relatively small
high compared to recent avian studies. Barlow et al geographic scale. The North and South Wales popula(2011) report weak differentiation among philopatric tions were considerably and significantly differentiated,
European shag Phalacrocorax aristotelis populations even though they are not separated by sea. The Scot(global D
= 0.066 compared to D
= 0.241 in tish islands of Colonsay and Islay are only 10 km apart,
est
est
choughs). Segelbacher et al (2003) report moderate dif- yet there was detectable small genetic differentiation
ferentiation among fragmented European capercaillie between them. The strong genetic structure among
601
Tetrao urogallus populations (global FST = 0.102 com- chough populations was therefore at least partially expared to F = 0.208 in choughs). However, genetic plicable by geographic distance and implies very low
602
differentiation similar to that observed in choughs has
600
603
604
ST
rates of successful long-distance dispersal and gene flow
been reported in house sparrow Passer domesticus with across the British Isles, even among relatively proxipairwise D
of 0.07–0.33 among European popula- mate populations.
est
622
623
624
625
626
627
628
629
630
631
632
633
634
635
607
tions (Schrey et al, 2011). Stronger differentiation has
This conclusion concurs with ringing data. Only
also been reported at very large spatial scales, e.g. pair- six ringed individuals have been observed to disperse
wise FST = 0.362 in snowy plover Charadrius alexan- between Islay and Colonsay in over twenty years (al-
608
drinus across 4,000 km (Funk et al, 2007) and pairwise
though Colonsay was probably colonised from Islay
639
609
Dest = 0.260 in giant petrel (Macronectes spp.) across
in the late 1960s, Reid et al, 2003, 2008).
Never-
640
610
7,000 km (Techow et al, 2010). Overall, differentiation
theless, field observations show that choughs do oc-
641
605
606
11
636
637
638
Figure 4: Statistical parsimony network of ten resolved haplotypes in 34 red-billed choughs from twelve locations.
Haplotype names (e.g. H1.1) and frequencies (n) are given within circles. Circle areas are proportional to
haplotype frequencies. Empty circles represent inferred, unsampled haplotypes. Transversion mutations are
indicated by bold lines. Branch lengths are arbitrary.
642
casionally disperse over long distances. At least nine
no longer apparent when sampling location was incor-
666
643
choughs moved between North Wales and the Isle of porated as prior information. They may therefore be
667
644
Man (c. 100 km) during 1997–2004, and two of them
668
645
were proven to have bred (Moore, 2006, 2008). Fur-
erroneous initial assignments due to partially missing
647
genotype data, small population size or local violation
thermore, the recolonisation of Cornwall in 2001 is as- of the Hardy-Weinberg equilibrium assumption rather
sumed to reflect natural long-distance dispersal from than true long-distance migrants (Pritchard et al, 2000;
648
other wild populations (Johnstone et al, 2011). The Evanno et al, 2005; Latch et al, 2006).
649
colonisers are speculated to have originated in Brittany
650
or South Wales (Carter et al, 2003). However, our ge-
651
netic data show that the colonisers do not match these
652
populations, or the local captive population in Paradise
653
Park, but suggest they probably originated in Ireland.
654
Although inference is constrained by the small sample
655
size (9 individuals), the only case of non-significant ge-
656
netic differentiation was Ireland vs. Cornwall. These
657
populations also shared a mitochondrial haplotype and
658
an a posteriori genetic cluster. Assuming that this re-
659
colonisation was unassisted, the genetic date therefore
660
show that successful long-distance dispersal can occur.
646
669
670
671
672
Phylogeography
673
Phylogeographic structure within the British Isles was
674
poorly resolved due to low mitochondrial DNA se-
675
quence polymorphism. Observed polymorphism sug-
676
gested weak diversification of haplotypes sampled in
the British Isles from those sampled in Continental Eu-
677
678
rope. The phylogeographic tree placed the Continental
679
European populations in Spain, French Alps and Brit-
680
tany ancestral to all British populations, which is con-
681
sistent with a classic northward pattern of postglacial
682
Some individuals were initially assigned to differ- recolonisation from refugia in southern Europe (Taber-
683
662
ent a posteriori genetic clusters than most other in- let et al, 1998; Hewitt, 2000). No evidence for coloni-
684
663
dividuals from the same a priori population, imply-
sation by more than one lineage (e.g. Celtic fringe
685
664
ing some dispersal among Wales, Ireland, Scotland scenario; Searle et al, 2009) was found, as all British
686
665
and Brittany. However, most such assignments were
687
661
populations formed a single clade. The single sample
12
Figure 5: Maximum Likelihood phylogram of 34 red-billed choughs based on sequencing of three mitochondrial
regions. Two Alpine choughs were used as an outgroup (branch clipped to clarify ingroup branching). The scale
bar represents 0.001 nucleotide substitutions per site. The values on nodes are bootstrap support values (only
> 50 % are shown) derived from 10,000 iterations using the Neighbour-Joining construction method.
688
from Northern Ireland did not share the same haplo- chondrial DNA, caused by a smaller effective popula-
707
689
type and clade as Ireland and was more similar to the tion size of mitochondrial versus nuclear DNA (Avise
708
690
UK populations.
et al, 1987; Birky et al, 1989). Higher mutation rates
709
in nuclear microsatellite loci are likely to amplify this
710
discrepancy (Balloux and Lugon-Moulin, 2002). These
711
explanations comply with the known decline of chough
712
populations during the 18th–20th centuries and conse-
713
quent bottlenecks (Holloway and Gibbons, 1996).
714
691
Weak mitochondrial genetic structure contrasted
692
with strong nuclear genetic structure.
Whilst mi-
693
crosatellite genotypes showed genetic differentiation
694
even between Colonsay and Islay, almost the entire
695
UK population shared a single mitochondrial haplo-
696
type. Such discrepancies in genetic structure are fre-
697
quently reported for avian species (e.g. Johnson et al,
Our current aim was to link the phylogeography
715
700
2003; Caparroz et al, 2009; Hefti-Gautschi et al, 2009) of chough populations in the British Isles with samand are often attributed to sex-biased dispersal where a pled Continental European populations, rather than to
weaker mitochondrial structure would indicate female- compile a full Continental European phylogeography.
701
biased dispersal. This is unlikely to be the case in
Sampling was therefore restricted to only one location
719
702
choughs. Although females disperse slightly further
in Spain and two locations in France. While including
720
703
than males within individual populations (Reid et al,
relatively few samples per location is not unusual (e.g.
721
704
2006; Moore, 2008), long-distance dispersal is rarely Taberlet et al, 1998; Questiau et al, 1999), future anal-
722
705
observed in either sex. A more likely explanation is yses could compile the full chough phylogeography by
723
706
increased propensity to genetic stochasticity in mito-
724
698
699
sampling a greater range of populations.
13
716
717
718
in small, isolated populations due to stochastic loss
727
of alleles.
The observed strong genetic structure
728
among small chough populations indicates low popula-
729
tion connectivity and consequently predicts low within-
730
population genetic diversity.
731
Most British Isles chough populations had fewer
732
than 4.0 alleles per locus, whereas the sampled Con-
733
tinental European populations had slightly higher di-
734
versity (c. 5.0–7.0 alleles). Observed heterozygosity
735
was also low, ranging from 0.30 to 0.52. Colonsay,
736
Ireland and French Alps were significantly deficient in
737
heterozygote genotypes (positive FIS ), which might in-
738
dicate some within-population sub-structuring caused
739
by wrongly delineated a priori populations (Wahlund,
740
1928). However, a posteriori genetic clusters did not
741
show sub-structuring in these populations, suggesting
742
that heterozygote deficiency is not due to a Wahlund
743
effect.
744
.
T
T
.
.
A
G
745
experienced population bottlenecks typically have less
746
than 3.0–4.0 alleles per locus, for example 4.0 in golden
747
eagle Aquila chrysaetos (Bourke et al, 2010), 3.0 in
748
Galapagos penguin Spheniscus mendiculus (Nims et al,
749
2008) and 1.9 in Madagascar fish-eagle Haliaeetus vo-
750
ciferoides (Johnson et al, 2009). Similarly, heterozy-
751
gosity is typically below 0.50, for example 0.44 in caper-
752
caillie Tetrao urogallus (Segelbacher et al, 2003), 0.20
753
in black robin Petroica traversi (Ardern and Lambert,
754
1997) and 0.10 in Mauritius kestrel Falco punctatus
755
(Nichols et al, 2001). At the other end of the spectrum
756
are widely-dispersed, high-abundance species such as
757
house sparrow Passer domesticus with 13.6 alleles per
758
locus and heterozygosity of 0.83 (Schrey et al, 2011).
759
In comparison, all chough populations had relatively
760
low genetic diversity.
761
.
.
.
.
.
Threatened bird populations that are known to have
A
.
Spain02
Spain03
.
726
.
.
Neutral genetic diversity is expected to be reduced
.
.
JQ924841; JQ924851; JQ924861
Spain01
725
Within the British Isles, the northerly populations
762
Colonsay, Islay and Isle of Man had lower genetic di-
763
versity than the more southerly populations. The new
764
population in Cornwall had lower genetic diversity than
its most likely source population in Ireland, which is
H8
G
.
.
C
.
.
.
.
T
T
T
T
.
A
.
.
A
A
G
G
.
.
.
.
G
.
A
A
T
.
A
A
.
.
.
G
.
.
JQ924839; JQ924849; JQ924859
JQ924840; JQ924850; JQ924860
H7
H6
FrenchAlps01, FrenchAlps03
FrenchAlps02
.
.
.
.
.
C
.
.
T
T
T
T
.
.
.
.
A
A
G
G
.
.
.
T
.
.
.
.
.
.
A
A
.
.
.
.
.
T
JQ924837; JQ924847; JQ924857
JQ924838; JQ924848; JQ924858
H5
H4
SouthWales01, SouthWales03
Ireland01, Ireland02, Ireland03, Cornwall01, C
Brittany01, Brittany02, Brittany03
.
.
.
.
T
T
.
.
A
G
A
.
.
.
.
A
.
.
.
JQ924836; JQ924846; JQ924856
H3
.
.
.
.
.
.
.
C
.
.
.
.
.
.
C
C
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
JQ924834; JQ924844; JQ924854
JQ924835; JQ924845; JQ924855
H2.2
H2.1
Colonsay01, Colonsay02, Colonsay03, Islay01,
IsleOfMan01
.
A
T
.
.
T
T
.
.
C
C
.
.
G
T
.
.
G
A
.
.
T
C
.
.
A
G
.
.
G
G
.
T
C
A
.
.
C
JQ924832; JQ924842; JQ924852
JQ924833; JQ924843; JQ924853
H1.2
H1.1
Individual
1197
3324
3093
966
681
2808
2741
614
360
2487
2011
860
563
1768
1585
380
347
1552
1540
335
269
25
1474
Haplotype
Accessions
47
137
296
352
403
1040
1049
1230
CHMT17 (NADH5/CYTB)
CHMT06 (NADH1)
Control region
Table 4: Polymorphic nucleotide sites and defined haplotypes in mitochondrial DNA sequences of 34 red-billed choughs. Dots denote the same nucleotide as the reference
sequence. Nucleotide positions are given for each of three sequence fragments separately as well as combined. genbank accessions are given for each fragment separately.
Genetic diversity
14
765
766
not surprising because there were only 3–7 founders
767
(Carter et al, 2003; Johnstone et al, 2011). The low ge-
768
netic diversity in the north might be a consequence of
769
founder effects during post-glacial south-north coloni-
770
sation events, but lack of resolution within the phy-
771
logeographic tree precludes assessment of colonisation
772
773
routes within the British Isles. Furthermore, as there
sensu Vaurie, 1954). Strict application of the phylo-
820
774
are no historic nuclear genetic diversity data available genetic species concept based on reciprocal monophyly
821
775
to compare to contemporary diversity, it is not possi-
(Donoghue, 1985) would classify Brittany’s choughs as
822
776
ble to ascertain whether the observed patterns of ge-
part of the Continental European subspecies. How-
823
777
netic diversity reflect more recent population contrac- ever, given the weak statistical support for the phylo-
824
778
tion and isolation.
Notwithstanding the underlying geographic groups, the microsatellite data may provide
825
779
causes, nuclear genetic diversity in most chough popu-
a more credible structure and therefore concur with
826
780
lations was notably low.
Vaurie’s taxonomy.
827
781
782
783
Compared with recent avian studies, mitochondrial
Similarly, if evolutionarily significant units (ESUs)
genetic diversity was also low, even in the hypervariable are based solely on reciprocal monophyly (Moritz,
control region (e.g. Piertney et al, 2001; Segelbacher 1994), the weakly supported chough phylogeography
785
and Piertney, 2007; Barbanera et al, 2009). A recent divides the sampled populations into three broad units:
study that quantified mitochondrial genetic diversity the Continental European populations in Spain, the
786
in choughs did not find any polymorphism in a 365 bp
784
787
788
French Alps (and possibly Brittany); the populations
control region segment among 23 extant Welsh choughs in Ireland, Cornwall and South Wales; and all other
and 19 museum specimens from across the British Isles, British Isles populations. However, given the high mi-
790
and concluded that all extant choughs in the UK form crosatellite differentiation among populations within
a single matrilineage (Kocijan and Bruford, 2011). We these three units (Moritz, 1994), each population may
791
confirm overall low mitochondrial diversity and that
789
792
793
794
795
need to be managed separately as each is to some ex-
North Wales is monomorphic across 3,355 bp, but we tent a distinct genetic unit. The individual popularesolved an additional haplotype in South Wales. We tions within the British Isles are already monitored and
resolved four haplotypes across the British Isles over- managed largely separately (Finney and Jardine, 2003;
all, although one haplotype was much commoner than Gray et al, 2003; Kerbiriou et al, 2005; Whitehead et al,
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
797
the other three. Low mitochondrial diversity is not un- 2005; Moore, 2008; Johnstone et al, 2011). Our data
usual (e.g. Waits et al, 2003; Roques and Negro, 2005; suggest that this is an appropriate strategy to conserve
798
Cadahia et al, 2007). Given the decline in chough pop-
799
ulation size and range during the 18th–20th centuries,
There is growing evidence that reduced genetic di-
846
800
bottlenecks in the early 20th probably caused losses of
versity can increase long-term extinction risk (Reed
847
801
mitochondrial as well as nuclear genetic diversity (Hol-
and Frankham, 2003; Frankham, 2005, 2010b), even
848
802
loway and Gibbons, 1996).
when reduced fitness is not immediately apparent (e.g.
849
Jamieson et al, 2006; Johnson et al, 2009). Genetic
850
796
genetic diversity and evolutionary potential.
843
844
845
804
Implications for conservation manage- diversity was comparatively low in all chough populations, indicating that concern over individual fitness,
ment
evolutionary potential and population persistence may
853
805
Current published chough taxonomy (Vaurie, 1954)
be warranted, particularly for the Colonsay, Islay and
854
806
was based on morphology and has not been verified
Isle of Man populations. However, genetic diversity in
855
803
808
genetically. Subspecies taxonomy based on morphol- neutral microsatellite markers may not be a good meaogy alone may be misleading if phenotypic variation sure of adaptive genetic diversity (Moss et al, 2003).
809
does not reflect evolutionary splits (e.g.
807
810
811
812
In fact, as microsatellite loci evolve faster than single
Burbrink
et al, 2000; Piertney et al, 2001; Segelbacher and Piert- nucleotide polymorphisms (SNPs) in genes, neutral geney, 2007). Microsatellite-based genetic differentiation netic diversity may overestimate genome-wide adaptive
and a posteriori genetic clusters matched current pub- genetic diversity (Väli et al, 2008). If adaptive diversity in choughs is low, as suggested by neutral diversity,
813
lished taxonomy in that the Brittany population clus-
814
tered with the British Isles population (equating to the
concern over long-term adaptability may be justified
851
852
856
857
858
859
860
861
862
863
816
nominate subspecies P. p. pyrrhocorax sensu Vaurie, and consideration of translocations to increase genetic
1954). However, the haplotype network and phylo- diversity in particularly depauperate and isolated pop-
817
geographic tree suggested that the Brittany popula-
818
tion is more closely related to the Continental Euro-
as
867
819
pean populations (equating to P. p. erythrorhamphos demonstrated for example in adders Vipera berus and
868
815
ulations may be warranted.
15
Translocation can aid population recovery,
864
865
866
man Coulter Genomics, Takeley, UK.
869
gray wolves Canis lupus (reviewed in Tallmon et al,
870
2004; Frankham, 2005), but many such projects fail
871
(Fischer and Lindenmayer, 2000; Tallmon et al, 2004).
872
Successful translocation programmes require consider-
This study was funded by the Royal Society (JMR),
916
873
able planning and effort to satisfy IUCN guidelines
the Philip Leverhulme Prize (JMR) and the Nuffield
917
874
(IUCN, 1998). The source population must be genet-
Foundation Undergraduate Research Bursary (MAW).
918
875
ically similar to the target population to avoid out-
876
breeding depression, although Frankham et al (2011)
877
argue that concerns over outbreeding depression may
References
919
878
879
880
We thank two anonymous reviewers for helpful comments on this manuscript.
be exaggerated for populations that became fragAkaike H (1974) A new look at the statistical model
mented relatively recently. The chough populations
identification. IEEE T Automat Contr 19(6):716–723
in Ireland and North Wales hold the greatest genetic
913
914
915
920
921
884
diversity and are only moderately differentiated from Ardern S, Lambert D (1997) Is the black robin in gethe northern populations. They may therefore be suitnetic peril? Mol Ecol 6:21–28
able sources for translocations. The genetic data confirmed that the ancestors of the captive choughs in Par- Avise JC, Arnold J, Ball RM, Bermingham E, Lamb
885
adise Park most probably originated from North Wales
T, Neigel JE, Reeb CA, Saunders NC (1987) In-
925
886
(Burgess et al, in press). They may be suitable for rein-
traspecific phylogeography: The mitochondrial DNA
926
887
troduction, but are more substantially differentiated
bridge between population genetics and systematics.
927
888
from the northern populations. In any case, given the
Annu Rev Ecol Syst 18:489–522
928
889
very small census sizes of some populations, thorough
890
evaluation of the consequences of removing individuals
891
from these populations will be necessary. Not least,
892
appropriate habitat management and restoration will
893
be required before any useful translocations could take
Barbanera F, Marchi C, Guerrini M, Panayides P,
932
894
place. Indeed, improved habitat quality might even
Sokos C, Hadjigerou P (2009) Genetic structure
933
895
facilitate natural dispersal and hence genetic connec-
of Mediterranean chukar (Alectoris chukar, galli-
934
896
tivity among populations (Johnstone et al, 2011).
formes) populations: conservation and management
935
implications. Naturwissenschaften 96:1203–1212
936
881
882
883
897
Appendix
898
See tables 5, 6 and 7.
899
Acknowledgements
900
We are extremely grateful to everyone who contributed
901
samples, most particularly Caitlin, Eric & Sue Big-
902
nal, Maria Bogdanova, Rob Colley, Tony Cross &
903
Adrienne Stratford (Cross & Stratford Welsh Chough
904
Project), Anne Delestrade, Annie & Bob Haycock,
906
Jane Hodges, David Jardine, Ian Johnstone, Davy McCracken, Allen Moore, Greg Morgan, Claire Muck-
907
low, Mike Peacock, Tom Pennycott, Chris Sharpe, Vic
908
Simpson, Mike Trewby, Gareth Watkins and David
909
910
905
922
923
924
Balloux F, Lugon-Moulin N (2002) The estimation of
929
population differentiation with microsatellite mark-
930
ers. Mol Ecol 11:155–165
931
Barlow EJ, Daunt F, Wanless S, Alvarez D, Reid JM,
937
Cavers S (2011) Weak large-scale population genetic
938
structure in a philopatric seabird, the European shag
939
Phalacrocorax aristotelis. Ibis 153:768–778
940
Birky C, Fuerst P, Maruyama T (1989) Organelle gene
941
diversity under migration, mutation, and drift: equi-
942
librium expectations, approach to equilibrium, ef-
943
fects of heteroplasmic cells, and comparison to nu-
944
clear genes. Genetics 121(3):613–627
945
Blanco G, Fargallo JA, Cuevas JA, Tella JL (1998a)
Effects of nest-site availability and distribution on
946
947
density-dependent clutch size and laying date in the
948
chough Pyrrhocorax pyrrhocorax. Ibis 140:252–256
949
Woolcock. We acknowledge the work of DNA Sequenc-
Blanco G, Tella JL, Torre I (1998b) Traditional farming
950
ing & Services (MRCPPU, College of Life Sciences,
and key foraging habitats for chough Pyrrhocorax
951
911
University of Dundee, Scotland, www.dnaseq.co.uk),
pyrrhocorax conservation in a Spanish pseudosteppe
952
912
Eurofins MWG GmbH, Ebersberg, Germany and Beck-
landscape. J Appl Ecol 35:232–239
953
16
Table 5: Characterisation of three primer pairs to amplify mitochondrial DNA regions in red-billed chough.
Fragment sizes are given alongside PCR TouchDown annealing temperature gradients (Ta ) and genbank
accessions of resolved haplotypes.
Locus
Primer name
Primer sequence (5’–3’)
Fragment size (bp)
Ta (ºC)
genbank accessions
Control region
JCR03
H1248
CHMT06-F
CHMT06-R
CCCCCCCATGTTTTTACR
CATCTTCAGTGTCATGCT
AGGTTCAAATCCTCTCCCTAGC
AACCATCATTTTCGGGGTATG
1205
60→50
JQ924832–JQ924841
922
65→55
JQ924842–JQ924851
CHMT17-F
CHMT17-R
AACCTAGCCCTAATAGGAAC
AGTAGTATGGGTGGAATGG
1228
55→45
JQ924852–JQ924861
NADH1
NADH5/CYTB
Table 6: Characterisation of 16 microsatellite loci for red-billed chough. Statistics (± 1 SD) were calculated from
348 individuals in eleven populations. The microsatellite repeat unit is given alongside TouchDown annealing
temperature gradient (Ta ), number of alleles (na ), allele range (bp), observed (HO ) and expected (HE ) heterozygosity, the probability of Hardy-Weinberg equilibrium (pHWE ) and null allele frequency (van Oosterhout
et al, 2004). See Wenzel et al (2011) for full characterisation.
Locus name
Repeat unit
Ppy-001
TACA
Ppy-002
ATCT
Ppy-003
AGAT
Ppy-004
AGAT
Ppy-005
TATC
Ppy-006
CATC
Ppy-007
GATA
Ppy-008
GATA
Ppy-009
AAGT
Ppy-010
CA
Ppy-011
TAGA
Ppy-012
TAGA
Ppy-013
GATA
Ppy-014a
GATG
Ppy-015a
TATG
Ppy-016
GGAT
a
locus also isolated by Jaari
Ta
na
60 → 50 4
60 → 50 4
60 → 50 11
60 → 50 8
60 → 50 7
60 → 50 8
55 → 45 9
60 → 50 10
60 → 50 6
60 → 50 14
60 → 50 10
60 → 50 13
60 → 50 10
60 → 50 5
60 → 50 3
60 → 50 13
et al (2008)
Allele range
HO
HE
pHWE
Null-allele frequency
151–179
151–179
292–344
173–239
226–250
139–175
161–193
221–265
222–242
108–146
163–191
210–266
197–221
239–275
152–158
200–244
0.46 ± 0.06
0.33 ± 0.05
0.50 ± 0.04
0.40 ± 0.03
0.25 ± 0.04
0.05 ± 0.03
0.61 ± 0.03
0.55 ± 0.03
0.58 ± 0.05
0.51 ± 0.05
0.66 ± 0.05
0.46 ± 0.07
0.58 ± 0.02
0.34 ± 0.03
0.06 ± 0.04
0.52 ± 0.03
0.45 ± 0.04
0.36 ± 0.06
0.58 ± 0.04
0.46 ± 0.02
0.30 ± 0.06
0.11 ± 0.06
0.69 ± 0.02
0.66 ± 0.02
0.59 ± 0.02
0.50 ± 0.04
0.71 ± 0.02
0.61 ± 0.03
0.68 ± 0.02
0.36 ± 0.02
0.07 ± 0.04
0.60 ± 0.04
0.148
0.993
< 0.001
0.183
0.028
0.729
0.425
0.018
0.420
0.187
0.190
< 0.001
0.493
0.615
0.120
0.022
−0.04 ± 0.13
−0.05 ± 0.12
0.01 ± 0.11
−0.04 ± 0.13
−0.02 ± 0.14
0.00 ± 0.05
0.00 ± 0.08
−0.02 ± 0.17
−0.06 ± 0.11
−0.11 ± 0.17
−0.08 ± 0.13
0.00 ± 0.23
0.01 ± 0.07
0.02 ± 0.08
−0.04 ± 0.15
0.02 ± 0.07
17
Table 7: Likelihood statistics of Bayesian inference of genetic clusters in structure. The mean logarithmic
likelihood (± SD) of 15 runs at each K is given alongside the ΔK statistic by Evanno et al (2005). Peak values
for ΔK are indicated in bold.
All populations
Atlantic coast populations only
Std. admixture
LOCPRIOR
Std. admixture
LOCPRIOR
954
K
LnP(K)
ΔK
LnP(K)
ΔK
LnP(K)
ΔK
LnP(K)
ΔK
1
2
3
4
5
6
7
8
9
10
11
−11024 ± 1
−10049 ± 5
−9411 ± 4
−9071 ± 19
−8814 ± 230
−8626 ± 22
−8511 ± 20
−8411 ± 36
−8346 ± 52
−8273 ± 11
−8281 ± 127
–
68
73
4
0
3
1
1
0
7
–
−11024 ± 0
−9960 ± 5
−9339 ± 6
−9011 ± 6
−8733 ± 11
−8633 ± 42
−8558 ± 79
−8484 ± 57
−8425 ± 78
−8465 ± 153
−8378 ± 106
–
83
50
8
16
1
0
0
1
1
–
−8470 ± 0
−7817 ± 1
−7494 ± 21
−7168 ± 1
−7041 ± 21
−6936 ± 30
−6848 ± 49
−6795 ± 103
−6714 ± 52
–
–
–
534
0
180
1
1
1
0
–
–
–
−8470 ± 0
−7787 ± 3
−7456 ± 10
−7141 ± 4
−7021 ± 22
−6947 ± 73
−6911 ± 64
−6798 ± 48
−6783 ± 87
–
–
–
108
2
46
2
1
1
2
–
–
–
Bohonak AJ (2002) IBD (Isolation By Distance): a Caparroz R, Miyaki CY, Baker AJ (2009) Contrasting
982
955
program for analyses of isolation by distance. J
phylogeographic patterns in mitochondrial DNA and
983
956
Hered 93:153–154
microsatellites: Evidence of female philopatry and
984
male-biased gene flow among regional populations
985
of the blue-and-yellow macaw (Psittaciformes: Ara
986
ararauna) in Brazil. The Auk 126(2):359–370
987
957
Bourke BP, Frantz AC, Lavers CP, Davison A, Daw-
958
son DA, Burke TA (2010) Genetic signatures of pop-
959
ulation change in the British golden eagle (Aquila
960
chrysaetos). Conservat Genet 11:1837–1846
961
Bull RD, McCracken A, Gaston AJ, Birt TP, Friesen
962
VL (2010) Evidence of recent population differentia-
963
tion in orange-crowned warblers (Vermivora celata)
964
in Haida Gwaii. The Auk 127(1):23–34
965
966
typic North American rat snake (Elaphe obsoleta):
968
A critique of the subspecies concept. Evolution
969
54(6):2107–2118
970
Burgess M, Woolcock D, Hales R, Hales A (in press) A
971
pilot release of captive-bred red-billed choughs into
972
Cornwall, UK. In: Pritpal S (ed) Global Reintroduc-
973
tion Perspectives: IUCN
975
988
The restoration of the red-billed chough in Cornwall.
989
Br Birds 96:23–29
990
Chao A, Shen TJ (2010) SPADE (species prediction
991
and diversity estimation). Program and User’s Guide
992
published at http://chao.stat.nthu.edu.tw
993
Burbrink FT, Lawson R, Slowinski JB (2000) Mi- Clement M, Posada D, Crandall K (2000) TCS: a comtochondrial DNA phylogeography of the polyputer program to estimate gene genealogies. Mol
967
974
Carter I, Brown A, Lock L, Wotton S, Croft S (2003)
Ecol 9(10):1657–1660
of the red-billed chough (Pyrrhocorax pyrrhocorax).
977
Zoo Biol
995
996
Delestrade A, Stoyanov G (1995) Breeding biology and
997
survival of the Alpine chough Pyrrhocorax graculus.
998
Bird Study 42:222–231
999
Donoghue MJ (1985) A critique of the biological species
1000
concept and recommendations for a phylogenetic al-
1001
ternative. Bryologist 88(3):172–181
1002
Burgess MD, Woolcock D, Hales RB, Waite R, Hales Earl D (2011) Structure Harvester v0.6.1. Available at:
AJ (in review) Captive husbandry and socialization
http://taylor0.biology.ucla.edu/structureHarvester/
976
994
(accessed 24/08/2011)
1003
1004
1005
Eaton M, Brown A, Noble D, Musgrove A, Hearn R,
1006
978
Cadahia L, Negro JJ, Urios V (2007) Low mitochon-
Aebischer N, Gibbons D, Evans A, Gregory R (2009)
1007
979
drial DNA diversity in the endangered Bonelli’s eagle
Birds of conservation concern 3: the population sta-
1008
980
(Hieraaetus fasciatus) from SW Europe (Iberia) and
tus of birds in the United Kingdom, Channel Islands
1009
981
NW Africa. J Ornithol 148:99–104
and the Isle of Man. Br Birds 102:296–341
1010
18
1011
Evanno G, Regnaut S, Goudet J (2005) Detecting the
Gray N, Thomas G, Trewby M, Newton SF (2003)
1052
1012
number of clusters of individuals using the soft-
The status and distribution of choughs Pyrrhocorax
1053
1013
ware STRUCTURE: a simulation study. Mol Ecol
pyrrhocorax in the Republic of Ireland 2002/03. Irish
1054
1014
14:2611–2620
Birds 7:147–156
1055
Griffiths R, Double MC, Orr K, Dawson RJG (1998) A
1015
Falush D, Stephens M, Pritchard JK (2003) Inference
1016
of population structure using multilocus genotype
1017
data: linked loci and correlated allele frequencies.
1018
Genetics 164:1567–1587
1019
Finney S, Jardine D (2003) The distribution and status
1020
of the chough in Scotland in 2002. Scott Birds 24:11–
1021
17
1023
published results of animal relocations. Biol Conserv
1024
96:1–11
1026
1027
1028
1029
1030
1031
1058
curate algorithm to estimate large phylogenies by
1059
maximum likelihood. Syst Biol 52:696–704
1060
Hallwachs W (2004) Ten species in one:
Fischer J, Lindenmayer D (2000) An assessment of the
Frankham R (1995) Conservation genetics. Annu Rev
Genet 29:305–327
Frankham R (2005) Genetics and extinction. Biol Conserv 126:131–140
genetic approaches to biological conservation. Biol
1033
Conserv 143:1919–1927
DNA
1062
1063
ical skipper butterfly Astraptes fulgerator. PNAS
1064
101(41):14,812–14,817
1065
Hefti-Gautschi B, Pfunder M, Jenni L, Keller V, El-
1066
legren H (2009) Identification of conservation units
1067
in the European Mergus merganser based on nuclear
1068
and mitochondrial DNA markers. Conservat Genet
1069
10:87–99
1070
Hewitt G (2000) The genetic legacy of the Quaternary
ice ages. Nature 405:907–913
1034
1035
Frankham R (2010b) Inbreeding in the wild really does
matter. Heredity 104:124
1036
Frankham R, Ballou JD, Eldridge MDB, Lacy RC,
1037
Ralls K, Dudash MR, Fenster CB (2011) Predicting
1038
the probability of outbreeding depression. Conserv
1039
Biol 25(3):465–475
1061
barcoding reveals cryptic species in the neotrop-
Frankham R (2008) Genetic adaptation to captivity in Hellberg ME (1994) Relationships between inferred
levels of gene flow and geographic distance in a
species conservation programs. Mol Ecol 17:325–333
philopatric coral, Balanophyllia elegans. Evolution
Frankham R (2010a) Challenges and opportunities of
48(6):1829–1854
1032
1057
Guindon S, Gascuel O (2003) A simple, fast, and ac-
Hebert PDN, Penton EH, Burns JM, Janzen DH,
1022
1025
DNA test to sex most birds. Mol Ecol 7:1071–1075
1056
1071
1072
1073
1074
1075
1076
Hogan FE, Cooke R, Burridge CP, Norman JA (2008)
1077
Optimizing the use of shed feathers for genetic anal-
1078
ysis. Mol Ecol Res 8:561–567
1079
Holloway S, Gibbons DW (1996) The historical atlas
1080
of breeding birds in Britain and Ireland, 1875-1900.
1081
T. & A.D. Poyser
1082
1040
Funk WC, Mullins TD, Haig SM (2007) Conserva-
Hubisz MJ, Falush D, Stephens M, Pritchard JK (2009)
1083
1041
tion genetics of snowy plovers (Charadrius alexandri-
Inferring weak population structure with the assis-
1084
1042
nus) in the Western Hemisphere: population genetic
tance of sample group information. Mol Ecol Res
1085
1043
structure and delineation of subspecies. Conservat
9:1322–1332
1086
1044
Genet 8:1287–1309
IUCN
(1998)
Guidelines
for
re-introductions.
1087
re-introduction
1088
1045
Goudet J (1995) FSTAT (version 1.2): A computer
prepared
1046
program to calculate f-statistics. J Hered 86(6):485–
specialist
1047
186
http://www.iucnsscrsg.org/download/English.pdf
1048
Goudet
J
(2002)
FSTAT
test
gene
-
a
Available
online:
1089
1090
1091
acterization of 100 polymorphic microsatellite loci
1092
Online:
for the Siberian jay (Perisoreus infaustus). Mol Ecol
1093
Res 8:1469–1474
1094
estimate
and
1050
fixation
indices
1051
http://www2.unil.ch/popgen/softwares/fstat.htm
2.9.3.2.
group.
IUCN/SSC
to Jaari S, Välimäki K, Merilä J (2008) Isolation and char-
1049
version
the
and
program
diversities
by
19
1095
Jakobsson M, Rosenberg NA (2007) CLUMPP: a clus- Kocijan I, Bruford MW (2011) Mitochondrial DNA
1140
1096
ter matching and permutation program for dealing
monomorphism in red-billed choughs Pyrrhocorax
1141
1097
with label switching and multimodality in analysis
pyrrhocorax in the United Kingdom. Bird Study
1142
1098
of population structure. Bioinformatics 23(14):1801–
58(2):213–216
1143
1099
1806
1100
Jamieson IG, Wallis GP, Briskie JV (2006) Inbreed-
1101
ing and endangered species management: Is New
1102
Zealand out of step with the rest of the world? Con-
1103
serv Biol 20(1):38–47
Lande R (1998) Anthropogenic, ecological and genetic
1144
factors in extinction and conservation. Res Popul
1145
Ecol 40(3):259–269
1146
Latch EK, Dharmarajan G, Glaubitz JC, Rhodes Jr
1147
OE (2006) Relative performance of Bayesian clus-
1148
1104
Johnson JA, Toepfer JE, Dunn PO (2003) Contrasting
tering software for inferring population substructure
1149
1105
patterns of mitochondrial and microsatellite popula-
and individual assignment at low levels of population
1150
1106
tion structure in fragmented populations of greater
differentiation. Conservat Genet 7:295–302
1151
1107
prairie-chickens. Mol Ecol 12:3335–3347
1108
Johnson JA, Tingay RE, Culver M, Hailer F, Clarke
1109
ML, Mindell DP (2009) Long-term survival despite
1110
low genetic diversity in the critically endangered
1111
Madagascar fish-eagle. Mol Ecol 18:54–63
1112
Johnstone I, Thorpe R, Moore A, Finney S (2007)
1113
Breeding status of choughs Pyrrhocorax pyrrhocorax
1114
in the UK and Isle of Man in 2002. Bird Study 54:23–
1115
34
1116
1117
1118
1119
1120
1121
Johnstone I, Mucklow C, Cross T, Lock L, Carter I
(2011) The return of the red-billed chough to Corn-
Librado P, Rozas J (2009) DnaSP v5: A software for
1152
comprehensive analysis of DNA polymorphism data.
1153
Bioinformatics 25:1451–1452
1154
Mank JE, Avise JC (2004) Individual organisms as
1155
units of analysis: Bayesian-clustering alternatives in
1156
population genetics. Genet Res 84:135–143
1157
Monaghan P (1988) The background of chough studies
1158
in Britain. In: Bignal E, Curtis DJ (eds) Choughs
1159
and Land-use in Europe, Scottish Chough Study
1160
Group
1161
wall: a review of the first 10 years and prospects for Moore AS (2006) Welsh choughs in the Isle of Man.
Peregrine 9(2):146–152
the future. Br Birds 104:416–431
Jost L (2008) GST and its relatives do not measure
Moore AS (2008) How far do Manx choughs travel?
differentiation. Mol Ecol 17:4015–4026
Peregrine 9(4):340–349
1122
Kerbiriou C, Thomas A, Floch P, Beneat Y, Gager
1123
L, Champion M (2005) Statut du crave à bec rouge
1124
Pyrrhocorax pyrrhocorax en Bretagne en 2002. Or-
1125
nithos 12-3:113–122
1126
Kerbiriou C, Gourmelon F, Jiguet F, Viol IL, Bioret F,
1127
Julliard R (2006) Linking territory quality and re-
Moritz C (1994) Defining "Evolutionarily Significant
Units" for conservation. TREE 9(10):373–375
1162
1163
1164
1165
1166
1167
Moss R, Piertney SB, Palmer SC (2003) The use and
1168
abuse of microsatellite DNA markers in conservation
1169
biology. Wildl Biol 9(4):243–250
1170
1129
productive success in the red-billed chough Pyrrho- Mousadik AE, Petit RJ (1996) High level of genetic differentiation for allelic richness among populations of
corax pyrrochorax: implications for conservation
1130
management of an endangered population. Ibis
the argan tree [Argania spinosa (l.) Skeels] endemic
1173
1131
148:352–364
to Morocco. Theor Appl Genet 92:832–839
1174
1128
1171
1172
Kerbiriou C, Viol IL, Robert A, Porcher E, Gourmelon
Nichols RA, Bruford MW, Groombridge JJ (2001) Sus-
1175
1133
F, Julliard R (2009) Tourism in protected areas can
taining genetic variation in a small population: evi-
1176
1134
threaten wild populations: from individual response
dence from the Mauritius kestrel. Mol Ecol 10:593–
1177
1135
to population viability of the chough Pyrrhocorax
602
1178
1136
pyrrhocorax. J Appl Ecol 46:657–665
1132
Nims BD, Vargas FH, Merkel J, Parker PG (2008) Low
1179
1137
Kimura M, Crow JF (1964) The number of alleles that
genetic diversity and lack of population structure
1180
1138
can be maintained in a finite population. Genetics
in the endangered Galapagos penguin (Spheniscus
1181
1139
49:725–738
mendiculus). Conservat Genet 9:1413–1420
1182
20
1183
van Oosterhout C, Hutchinson WF, Wills DP, Shipley Reid JM, Bignal EM, Bignal S, McCracken DI, Bog-
1224
1184
P (2004) MICRO-CHECKER: software for identify-
danova M, Monaghan P (2008) Investigating pat-
1225
1185
ing and correcting genotyping errors in microsatellite
terns and processes of demographic variation: envi-
1226
1186
data. Mol Ecol Notes 4:535–538
ronmental correlates of pre-breeding survival in red-
1227
billed choughs Pyrrhocorax pyrrhocorax. J Anim Ecol
1228
77:777–788
1229
1187
Peakall R, Smouse P (2006) GENALEX 6: genetic
1188
analysis in Excel. Population genetic software for
1189
teaching and research. Mol Ecol Notes 6:288–295
1190
Piertney SB, Summers R, Marquiss M (2001) Mi-
1191
crosatellite and mitochondrial DNA homogeneity
1192
among phenotypically diverse crossbill taxa in the
1193
UK. Proc R Soc B 268:1511–1517
1194
1195
1230
aghan P, McCracken DI (2011) Diagnosing the tim-
1231
ing of demographic bottlenecks: sub-adult survival
1232
in red-billed choughs. J Appl Ecol 48:797–805
1233
Roques S, Negro JJ (2005) MtDNA genetic diversity
1234
and population history of a dwindling raptorial bird,
1235
the red kite (Milvus milvus). Biol Conserv 126:41–50
1236
Rosenberg NA (2004) DISTRUCT: a program for the
1237
graphical display of population structure. Mol Ecol
1238
Notes 4:137–138
1239
Posada D (2008) jModelTest: Phylogenetic model averaging. Mol Biol Evol 25:1253–1256
1196
Pritchard JK, Stephens M, Donnelly P (2000) Infer-
1197
ence of population structure using multilocus geno-
1198
type data. Genetics 155:945–959
1199
Questiau S, Gielly L, Clouetà M, Taberlet P (1999)
1200
Phylogeographical evidence of gene flow among com-
1201
mon crossbill (Loxia curvirostra, Aves, Fringilli-
1203
dae) populations at the continental level. Heredity
83:196–205
1204
Raymond M, Rousset F (1995) GENEPOP (version
1205
1.2): population genetics software for exact tests and
1206
ecumenicism. J Hered 86:248–249
1202
Reid JM, Bignal EM, Bignal S, Bogdanova MI, Mon-
Rousset F (1997) Genetic differentiation and estima-
1240
tion of gene flow from F-statistics under isolation by
1241
distance. Genetics 145:1219–1228
1242
Rousset F (2008) Genepop’007: a complete reimple-
1243
mentation of the Genepop software for Windows and
1244
Linux. Mol Ecol Res 8:103–106
1245
Rozen S, Skaletsky H (2000) Primer3 on the WWW for
1246
general users and for biologist programmers. Meth
1247
Mol Biol 132:365–386
1248
Saunders MA, Edwards SV (2000) Dynamics and phy-
1249
Reed DH, Frankham R (2003) Correlation between fit-
logenetic implications of mtDNA control region se-
1250
1208
ness and genetic diversity. Conserv Biol 17(1):230–
quences in New World jays (aves: Corvidae). J Mol
1251
1209
237
Evol 51:97–109
1252
1207
1210
Reid JM, Bignal EM, Bignal S, McCracken DI,
1211
Monaghan P (2003) Environmental variability, life-
1212
history covariation and cohort effects in the red-
1213
billed chough Pyrrhocorax pyrrhocorax. J Anim Ecol
1214
72:36–46
Schrey AW, Grispo M, Awad M, Cook MB, Mccoy
1253
ED, Mushinsky HR, Albayrak T, Bensch S, Burke
1254
T, Butler LK, Dor R, Fokidis HB, Jensen H, Im-
1255
boma T, Kessler-Rios MM, Marzal A, Stewart IRK,
1256
Westerdahl H, Westneat DF, ZehtindjIev P, Mar-
1257
tin LB (2011) Broad-scale latitudinal patterns of ge-
1258
1215
Reid JM, Bignal EM, Bignal S, McCracken DI, Mon-
1216
aghan P (2004) Identifying the demographic deter-
netic diversity among native European and introduced house sparrow (Passer domesticus) popula-
1217
minants of population growth rate: a case study of
tions. Mol Ecol 20:1133–1143
1218
red-billed choughs Pyrrhocorax pyrrhocorax. J Anim
1219
Ecol 73:777–788
Searle JB, Kotlik P, Rambau RV, Markova S, Herman
JS, McDevitt AD (2009) The Celtic fringe of Britain:
1220
Reid JM, Bignal EM, Bignal S, McCracken DI, Mon-
1221
aghan P (2006) Spatial variation in demography and
1222
population growth rate: the importance of natal lo-
1223
cation. J Anim Ecol 75:1201–1211
1259
1260
1261
1262
1263
insights from small mammal phylogeography. Proc R
1264
Soc B 276:4287–4294
1265
Segelbacher G, Piertney S (2007) Phylogeography of
1266
the European capercaillie (Tetrao urogallus) and its
1267
21
1268
implications for conservation. J Ornithol 148 (Suppl
1269
2):S269–S274
1270
Segelbacher G, Höglund J, Storch I (2003) From con-
1271
nectivity to isolation: genetic consequences of pop-
1272
1273
1274
1275
1276
Slatkin M (1993) Isolation by distance in equilib-
the probability of identity among genotypes in nat-
1312
ural populations: cautions and guidelines. Mol Ecol
1313
10:249–256
1314
38(6):1358–1370
rium and non-equilibrium populations. Evolution Wenzel MA, Webster LMI, Segelbacher G, Reid JM,
47(1):264–279
Piertney SB (2011) Isolation and characterisation
Swofford D (2000) PAUP*. Phylogenetic Analysis Us-
1278
ing Parsimony (*and other methods). Version 4. Sinauer Associates, Sunderland, Massachusetts
1280
Taberlet P, Fumagalli L, Wust-Saucy A, Cosson J
1281
(1998) Comparative phylogeography and postglacial
1282
1311
ulation fragmentation in capercaillie across Europe. Weir B, Cockerham CC (1984) Estimating f-statistics
Mol Ecol 12:1773–1780
for the analysis of population structure. Evolution
1277
1279
Waits LP, Luikart G, Taberlet P (2001) Estimating
colonization routes in Europe. Mol Ecol 7:453–464
1283
Tallmon DA, Luikart G, Waples RS (2004) The allur-
1284
ing simplicity and complex reality of genetic rescue.
1285
TREE 19(9):489–496
1286
Tarr CL (1995) Primers for amplification and determi-
1287
nation of mitochondrial control-region sequences in
1288
oscine passerines. Mol Ecol Notes 4:527–529
Patterson-Fraser D, Quintana F, Ritz M, Thompson D, Wanless R, Weimerskirch H, Ryan P (2010)
1292
Speciation and phylogeography of giant petrels
1293
Macronectes. Mol Phylogenet Evol 54:472–487
1294
Väli Ü, Einarsson A, Waits L, Ellegren H (2008)
1295
To what extent do microsatellite markers reflect
1296
genome-wide genetic diversity in natural popula-
1297
tions? Mol Ecol 17:3808–3817
1298
Valiere N (2002) GIMLET: a computer program for
1299
analysing genetic individual identification data. Mol
1300
Ecol Notes 2:377–379
1301
Vaurie C (1954) Systematic notes on Palearctic birds.
1302
No. 4 the choughs (Pyrrhocorax). Am Mus Novit
1303
1658
1304
Wahlund S (1928) Zusammensetzung von Popula-
1305
tionen und Korrelationserscheinungen vom Stand-
1306
punkt der Vererbungslehre aus betrachtet. Hereditas
1307
11(1):65–106
1308
Waits J, Avery M, Tobin M, Leberg P (2003) Low mito-
1309
chondrial dna variation in double-crested cormorants
1310
in eastern north america. Waterbirds 26(2):196–200
22
1319
1321
3:737–740
1322
Whitehead S, Johnstone I, Wilson J (2005) Choughs
1323
Pyrrhocorax pyrrhocorax breeding in wales select for-
1324
aging habitat at different spatial scales. Bird Study
1325
52:193–203
1326
Ann Eugen 15(4):323–354
1291
1318
(Pyrrhocorax pyrrhocorax). Conservat Genet Res
Wright S (1951) The genetical structure of populations.
1290
1317
1320
138
Techow N, O’Ryan C, Phillips R, Gales R, Marin M,
1316
of 17 microsatellite loci for the red-billed chough
Wright S (1943) Isolation by distance. Genetics 28:114–
1289
1315
1327
1328
1329
1330