ª 2004 The American Genetic Association
Journal of Heredity 2004:95(2):144–153
DOI: 10.1093/jhered/esh018
Genetic Consequences of a Severe
Population Bottleneck in the Guadalupe
Fur Seal (Arctocephalus townsendi)
D. S. WEBER, B. S. STEWART,
AND
N. LEHMAN
From the Department of Biological Sciences, University at Albany, State University of New York, Albany, NY 12222 (Weber
and Lehman) and Hubbs-SeaWorld Research Institute, San Diego, CA 92109 (Stewart). N. Lehman is currently at the
Department of Chemistry, Portland State University, P.O. Box 751, Portland, OR 97207. D. S. Weber is currently at the
Molecular Systematics Lab, American Museum of Natural History, Central Park West at 79th Street, New York, NY 10024.
We thank J. Heyning and D. Janiger for access to archaeological bone samples from the collection at the Los Angeles County
Museum of Natural History, G. Bernardi for providing DNA sequences from modern Guadalupe fur seals, A. Jacklet, R.
Osuna, and D. McKeon for providing facilities for analysis of ancient DNA, and D. Decker, J. Fahy (Qiagen), J. deKoning, T.
Pratt, D. Pryor, J. Schienman, C.-B. Stewart, K. A. Walker, K. L. Walker, J. VanderKelen (Qiagen), H. Weber, and P.
Yochem for technical assistance and suggestions. This work was supported by grants from Sigma Xi, the University at Albany
Benevolent Society, the University at Albany Graduate Student Organization, and the Hubbs-SeaWorld Research Institute.
Address correspondence to Niles Lehman, Department of Chemistry, Portland State University, P.O. Box 751, Portland,
OR 97207, or e-mail: [email protected].
Abstract
Population bottlenecks may lead to diminished genetic variability and correlative effects on fitness. The Guadalupe fur seal
was nearly exterminated by commercial sealers during the late 18th and early 19th centuries. To determine the genetic
consequences of this population bottleneck, we compared the variation at a 181 bp section of the mitochondrial DNA
(mtDNA) control region from the bones of 26 prebottleneck fur seals versus variation in the extant population. We found
25 different mtDNA genotypes in the prebottleneck fur seals and only 7 genotypes among 32 extant fur seals, including only
one of the ancient genotypes. These data demonstrate a substantial loss of genetic variability correlating with the recent
population bottleneck. We also found from several genetic measures that the prehistoric population of Guadalupe fur seals
was robust and that it had been increasing at some time during the late prehistoric period. Continued recovery of this
species may, however, owe more to more immediate nongenetic factors, such as poaching and local availability of food
resources during the breeding season and consequent effects on pup survival, than on the reduced genetic variability.
Anthropogenic disturbances to wildlife populations by direct
harvesting or habitat destruction often result in substantial
reductions in population size (population bottlenecks) and
loss of genetic variability (genetic bottlenecks). Lowered
allelic diversity, greater frequency of linkage disequilibrium,
and reduced heterozygosity are some of the observed or
inferred consequences of those bottlenecks (Hartl and Clark
1997; Hedrick 2000). The consequences of genetic bottlenecks vary from species to species and are often strongly
reflective of demographic factors, though it has been
generally held that substantial genetic variation is essential
for population vitality and persistence (Amos 1996; Caro
and Laurenson 1994; Hedrick 1996; O’Brien 1994; O’Brien
et al. 1987). The frequency of rare, deleterious recessive
alleles—the genetic load—can increase after a population
decline, resulting in reduced fitness through inbreeding
144
depression (cf. Amos 1996). However, both theory (Nei et al.
1975) and experiment (Miller and Hedrick 2001) have
suggested that the impacts of severe bottlenecks may be
minimized if subsequent recovery is rapid, and that fitness
may actually increase in some instances. Fitness may often
decline, however, in populations that recover more slowly,
owing to greater reductions in heterozygosity (e.g., Bouzat et
al. 1998; Westemeier et al. 1998).
The populations of most species of seals (family
Phocidae) and sea lions (family Otariidae) were reduced
substantially by commercial sealers and whalers in the 18th
and 19th centuries, and some may have been reduced
centuries earlier by coastal aborigines (Lyons 1937; Reeves
and Stewart 2003; Reeves et al. 1992, 2002; Stewart et al.
1993; Wynen et al. 2000). In fact, most species of fur seals
were presumed to be extinct by the late 19th century. The
Weber et al. Genetic Bottleneck in Guadalupe Fur Seal
Figure 1. The Pacific coastline showing the boundaries between southern California (USA), Baja California (Mexico), and
the islands off both areas in the Pacific Ocean. The Guadalupe fur seal only breeds today on Isla Guadalupe and Isla Benito
(in Mexican waters), however, an occasional individual does establish a territory on the Channel Islands (located off California).
Bones used in this study came from midden sites at Point Mugu and San Nicolas Island. Map adapted from National
Geophysical Data Center, NOAA Satellites and Information website.
prehistoric and even presealing distribution and abundance
of Guadalupe fur seals (Arctocephalus townsendi) are poorly
known (cf. Etnier 2002; Gustafson 1968; Stewart et al. 1993).
At least 52,000 Guadalupe fur seals had evidently been killed
at several islands off the Pacific coasts of Mexico and the
United States from the late 1700s to 1848, and the last few
were harvested in Mexican waters in the late 1800s (Townsend 1931). This species was then presumed extinct (Townsend 1916) until a small breeding group was discovered at Isla
de Guadalupe (Figure 1) in 1928 (Townsend 1928, 1931).
Collections for zoos and museums and poaching evidently
eliminated that colony soon after discovery, and the species
was again thought to be extinct until a single male was seen at
San Nicolas Island in 1949 (Bartholomew 1950). A small
breeding group of 14 seals was then discovered at Isla de
Guadalupe in 1954 (Hubbs 1956). The colony has grown
exponentially (r ¼ 13.7%) since then (Fleischer 1978; GalloReynoso 1994), and numbered between 7400 and 12,000 in
the late 1990s (Gallo-Reynoso 1994; Stewart 1997). Another
small breeding colony became established on Isla Benito del
Este in 1997 (Maravilla-Chavez and Lowry 1999).
Despite repeated, substantial reductions in population
size during the past few centuries, genetic diversity of the
Guadalupe fur seal is relatively high in the extant population.
Seven genotypes, with 18 polymorphic nucleotide sites, were
found in a 313 bp segment of the mitochondrial DNA
145
Journal of Heredity 2004:95(2)
Table 1.
a
Guadalupe fur seal bones from which DNA products were successfully obtained
Sample name
Indian midden
locationa
Type of bone
mtDNA bp
obtained
Repeats: bones ground/
bones extracted
FS-1
FS-2
FS-4
FS-6
FS-7
FS-10
FS-12
FS-16
FS-18
FS-19
FS-24
FS-25
FS-29
FS-30
FS-34
FS-37
FS-40
FS-41
FS-43
FS-46
FS-48
FS-49
FS-50
FS-51
FS-54
FS-55
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAb
Pt. Mugu, CAc,d
San Nicolas Islandc,e
San Nicolas Islandc,e
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right femur
Right humerus
Left humerus
Right humerus
177
177
178
178
182
178
178
217
217
218
178
218
177
218
178
179
166
178
177
179
178
217
181
180
219
179
2/2
2/2
2/2
5/2
2/2
1/2
2/2
2/2
2/2
2/2
2/2
1/2
1/2
1/2
1/2
2/2
1/2
2/2
2/2
2/2
2/2
2/2
4/2
2/2
2/2
2/2
Samples obtained from the Natural History Museum of Los Angeles County. Catalog numbers (e.g., LACM 31337) available upon request. All middens
were from Chumash settlements that must have predated their missionization, which coincided with the European sealing efforts between 1772 and 1822,
and resulted in abandonment of their sealing practices in favor of agriculture and animal husbandry (Museum of Santa Barbara County Archives).
b
From the L. H. Miller UCLA collection; shell mound collector was G. Lyons (Lyons 1937).
c
FS-51, FS-54, and FS-55 were from different collections than all other bones, eliminating the possibility that any of the femurs and humeri were from the
same individual. In addition, the humeri provided unique haplotypes and the only bones that had the same sequence were both right femurs.
d
Shell mound collector was R. W. Frasier.
e
Shell mound collector was J. Sanger.
(mtDNA) control region in 25 extant Guadalupe fur seals
(Bernardi et al. 1998). Nonetheless, we hypothesized that
some loss of genetic variability should have occurred during
the population bottleneck in the 19th century, owing to its
severity and duration before recovery began in the mid-20th
century. Our goals in this study were (1) to document
prebottleneck patterns of genetic variability in Guadalupe fur
seal and (2) to evaluate the effects of the recent population
bottleneck on various measures of genetic variability. To
achieve those goals, we extracted DNA from the bones of 26
Guadalupe fur seals recovered from archaeological sites in
southern California and assayed a 181 bp segment of the
mtDNA control region containing most of the variation
recently detected in seals from the extant population
(Bernardi et al. 1998).
Materials and Methods
Sample Collection
We obtained bones of 57 Guadalupe fur seals collected from
prebottleneck (i.e., estimated age predating commercial
146
exploitation) archaeological kitchen sites (middens) from
Point Mugu, California, or from the nearby southern
California Channel Islands (Figure 1). These bones were
archived at the Natural History Museum of Los Angeles
County. We attempted DNA extraction from 50 right femur
samples to ensure uniqueness of individual genotypes, as well
as from 7 humerus samples. The successfully analyzed
humeri bones (FS-51, FS-54, and FS-55) were from different
collections than all other bones, eliminating the possibility
that any of the femurs and humeri were from the same
individual (Table 1). We determined genetic variation in the
extant population from (1) mtDNA sequences of 313 bp
from 25 seals reported by Bernardi et al. (1998), and (2) tissue
samples that we obtained from seven additional seals found
injured or deceased along the southern California coast
during the 1980s and 1990s.
DNA Extraction, Amplification, and Sequence Analysis
First, we extensively sanded the outer surface of each bone.
Then we removed a small sample (4–5 g) from each bone,
ground it into a powder with liquid nitrogen, and decalcified
Weber et al. Genetic Bottleneck in Guadalupe Fur Seal
it in 0.5 M ethylenediaminetetraacetic acid (EDTA) for 5–14
days. Decalcification was confirmed by the absence of
a precipitate in the supernatant when a solution of saturated
ammonium oxalate was added. We obtained DNA from the
bone powder using the DNeasy tissue kit (Qiagen) and the
protocol for isolation of genomic DNA from compact bone
(Qiagen User Developed Protocol DY01; Yang et al. 1998).
We modified this procedure in two ways to maximize the
overall DNA yield: first, we lysed the bone pellet overnight
(instead of a few hours), and second, we eluted with only 50
ll (rather than 200 ll) of water to increase the final DNA
concentration.
Ancient DNA is normally severely degraded and in small
fragments, but our samples were potentially even more so,
having originated from southern California beaches, not
from cold or low-ultraviolet (UV) environments (higher
latitudes, caves, etc.). Therefore, prior to polymerase chain
reaction (PCR), we accomplished template reconstruction
via a PCR-like method without primers (‘‘primerless
template reconstruction’’; Stemmer 1994). We found here,
as elsewhere (Weber et al. 2000), that our success at
amplifying ancient DNA (aDNA) from sources such as these
is greatly enhanced with this technique. In each case, we
performed primerless template reconstruction on 8 ll of
bone extract in a 50 ll reaction volume with 1.25 U
AmpliTaq DNA polymerase (Perkin-Elmer), the manufacturer’s supplied standard buffer with 1.5 mM MgCl2, and 0.2
mM deoxynucleotide triphosphates (dNTPs). An initial
denaturation for 3 min at 948C was followed by 25 cycles
of 928C (1 min), 508C (1 min), 728C (1 min), and a final
extension step of 728C for 10 min. We then achieved
amplification of the resulting template reconstruction via
a two-step (‘‘nested’’) procedure of the 59 segment of the
mtDNA control region. Ten microliters of the primerless
product served as the template for the external (first step)
PCR in a 50 ll reaction volume with 10 pmol of each primer
and a standard PCR mix as outlined above. The cycles were
adjusted for a profile of 3 min at 958C, followed by 40 cycles
of 928C (30 s), 488C (30 s), 728C (30 s), and a final extension
at 728C for 10 min. Typically 1 ll of the resulting solution
served as the template for the internal (second step) 50 ll
PCR with parameters of 4 min at 948C followed by 25 cycles
of 928C (1 min), 538C (1 min), 728C (1 min), and a final
extension at 728C for 10 min. The use of external primers
(ONL-4a: 59-GTTGCTGGTTTCTCGAGGC-39 and GFS15a: 59-CTGTGCCACCATAGTATC-39) followed by
internal primers (fully nested: ONL-16: 59-TCGCGGGCTAGGTGAATTGC-39 and ONL-17: 59-CCCTATGTACGTCGTGC-39 or seminested using one of the external
primers) provided a stepping-stone process to obtain intact
DNA fragments from bones.
We obtained nucleotide sequences of the purified PCR
products (QIAquick PCR purification kit, Qiagen) by one or
both of the following methods. First, we performed
bidirectional manual sequencing using [a-35S]dATP and
Sequenase 2.0 (US Biochemical) with resolution on 6%
polyacrylamide/8 M urea gels and visualization by autoradiography. Second, we performed bidirectional automated
sequencing using the BigDye Terminator technology
(Applied Biosystems) with resolution on a 373A DNA
genetic analyzer (Applied Biosystems). Modern seals provided a 212 bp segment that included all of the mtDNA
segment provided by the bone samples. We obtained
genomic DNA from the tissue and blood samples with the
DNeasy tissue kit (QIAGEN) and performed a single PCR
procedure using the external primers and procedure listed
above. After purification we obtained nucleotide sequences
bidirectionally by the automated method described above.
Contamination Controls
Patterns of historical population genetics can be examined
with samples of aDNA (e.g., Cooper and Wayne 1998), but
adequate samples are often difficult to obtain from
weathered specimens and may be contaminated in the
process by modern DNA (e.g., Austin et al. 1997).
Prevention of contamination during the amplification process is the key concern in studies of aDNA (cf. Hagelberg
1994; Hagelberg and Clegg 1991; Hummel and Herrmann
1994). Consequently rigorous controls have been established
to minimize and prevent the risk of contamination, and to
ensure the integrity and purity of aDNA samples, we took
the following precautions. We processed all bones for aDNA
in laboratory facilities at the University at Albany (State
University of New York) that are dedicated to studies of
aDNA and that are spatially isolated by two floors from
facilities used for DNA amplification. We then transported
bone DNA extracts to a dedicated PCR setup room with
a UV light–containing hood (CBS Scientific) separate from
facilities where subsequent purification and sequence determination of the products occurred. For nested PCR, we
manipulated amplified DNA in a separate hood in yet
another room from the room with the hood used for the
PCR setup. Because the samples were subjected to three
different protocols using Taq DNA polymerase, we
conducted numerous controls to ensure that we obtained
accurate sequences.
We processed two aliquots per bone extract as separate
samples from start to finish, that is, from DNA extraction
through sequence analysis. We compared the two resulting
sequences in order to ascertain that the aliquots did indeed
give the exact same sequence for that individual. In addition,
for each of the two bone extracts for each individual, we
processed the two aliquots a second time in parallel
beginning from the primerless template reconstruction stage
and taken through nested PCR and sequence analysis to
ensure that identical sequences were obtained both times in
the two aliquots. Consequently, for each successful bone
extract, four separate sequences were obtained and compared. We also obtained second, independent extracts from
the bones of three of the seals (FS-12, FS-50, and FS-6) and
amplified and compared the DNA sequences to those
resulting from the original extracts (Table 1). We attempted
amplifications of negative control tubes open during the
extraction process in all cases, and negative controls were
included in every PCR amplification performed, usually
147
Journal of Heredity 2004:95(2)
several. Lastly, we minimized contamination from modern
fur seals by amplifying the bones prior to the amplification of
any of the seven modern tissue samples.
Data Analysis
We inspected sequences obtained from the ABI 373A using
the program SEQUENCHER (version 5.2; Gene Codes
Corp.). We then aligned all sequences with CLUSTALX 1.81
(Thompson et al. 1997) and GeneDoc 2.6.002 (Nicholas et
al. 1997). We endeavored to obtain general phylogenetic
relationships between the ancient and modern genotypes, in
comparison to other fur seal mtDNA sequences. The limited
number of phylogenetically informative sites in our samples
precluded an unambiguous phylogenetic reconstruction.
However, using PAUP* (version 4.0b10; Swofford 2002),
we were able to estimate the general relationships of ancient
and modern mtDNA lineages. In PAUP*, we employed the
distance matrix-based method using a maximum-likelihood
variant of the HKY85 model with minimum evolution
optimality criterion. The HKY85 model allows for different
transversion and transition rates and different base frequencies, both of which are the two most significant sources of
bias. In order to test if the modern samples are a subset of
the ancient samples, we found the best tree under the
constrained null hypothesis of monophyly for all modern
samples relative to all ancient and outgroup taxa. We
calculated likelihoods for both the constrained (null; modern
samples constrained to form a monophyletic clade with
respect to all other taxa) and unconstrained trees, using an
equivalent substitution model (HKY85 variant). We performed the likelihood-based Shimodaira-Hasegawa test to
determine whether the data significantly distinguish between
the two tress. The P value was calculated using a simulated
distribution of test statistics by 1000 RELL bootstrap
resampling in PAUP* 4.0b10. Additional sequences for the
phylogenetic analysis included the northern elephant seal
(Weber et al. 2000), plus the following Arctocephalus species
(Wynen et al. 2001): Juan Fernández fur seal (A. philippii;
GenBank accession no. AF384405), Australian fur seal (A.
pusillus doriferus; AF384395), sub-Antarctic fur seal (A.
tropicalis; AF384384), South American fur seal (A. australis;
AF384402), Antarctic fur seal (A. gazella; AF384377), and
Galapagos fur seal (A. galapagoensis; AF384386).
We estimated several population genetic parameters for
ancient and modern mtDNA: (1) number of distinct control
region genotypes, (2) number of polymorphic sites, (3)
haplotype (gene) diversity (h ¼ probability that two mtDNA
sequences chosen randomly from the sample are different),
(4) nucleotide diversity (p ¼ number of nucleotide differences between randomly chosen pairs of sequences), (5)
proportion of variable nucleotide (segregating) sites (p), (6)
Tajima’s D test statistic, (7) Fu’s FS test statistic, (8) average
number of nucleotide differences (k ¼ variance; Tajima
1993), (9) raggedness index (r ¼ a statistical measure
indicating distributions of stationary/expanding/contracting
populations; Harpending et al. 1993), and (10) h, both hS
(estimates of nucleotide polymorphism using h ¼ s/a) and
148
hp. The population genetic program DnaSP (version 3.51;
Rozas and Rozas 1999) provided h, p, D, k, and the graphs
for the mismatch distribution (also called distribution of
pairwise differences), while ARLEQUIN (version 2.000;
Schneider et al. 2000) provided FS, h, and r, and determined
the means of the mismatch distribution (pairs of genotypes
analyzed for the distribution of observed number of
differences; Schneider et al. 2000).
Results and Discussion
We obtained mtDNA sequences from 26 prebottleneck
Guadalupe fur seals. The sequences began approximately 86
bp from the terminus of the tRNA-Pro on the 59 end of the
mtDNA control region and spanned 181 bp. We found 25
ancient genotypes among those 26 prebottleneck Guadalupe
fur seals (indicated as FS1 through FS55; Figure 2). Twentyfour of the 25 genotypes have not yet been identified in
extant Guadalupe fur seal populations, whereas one (FS25)
matched one of the 7 genotypes reported by Bernardi et al.
(1998) from 25 seals (Figures 2 and 3). In contrast, in our
modern seal samples (N ¼ 7) we found three genotypes, all
of which match one of the previously published seven
modern genotypes. These data indicate a substantial historical loss of mtDNA genotypes in the Guadalupe fur seal.
Lineages from only two mtDNA clades appeared to have
survived the bottleneck, one containing modern genotypes
1–4 and 7, and the other containing genotypes 5 and 6
(Figure 3). Because many bootstrap values in the reconstruction shown in Figure 3 are low, we performed tests
of specific hypotheses to determine if the data significantly
support this conclusion. The Shimodaira-Hasegawa test
(using RELL bootstrap replicates) indicated that when the
constrained tree has only the northern elephant seal as an
outgroup and all other fur seal species are placed with the
ancient samples, the data do in fact reject the null hypothesis
that the modern samples are monophyletic (P ¼ .034). We
then tested a more conservative analysis in which the tree is
further constrained, such that all non-Guadalupe fur seals are
forced into the outgroup, and found that we again could not
reject the monophyly of the modern samples (P ¼ .003).
Both tests support our contention that the modern seals are
indeed a subset of the prebottleneck seals and that they
possess two general surviving mtDNA lineages.
The population bottleneck also resulted in a substantial
loss in genotypic diversity. We found 51 variable nucleotide
sites (26.5%) in the prebottleneck samples versus 18 (5.7%)
in the extant population. However, 15 of the variable ancient
sites were the result of only single nucleotide changes found
in single genotypes. We confirmed that these were not
artifacts caused by Taq polymerase error by comparing each
of these genotypes against all other genotypes at other
nucleotide sites and found no matches. Moreover, Taq has an
introduced error rate of about 1.1 3 104 nucleotides
polymerized (Tindall and Kunkel 1988), such that the
likelihood that the same error would occur in our duplicate
amplifications is less than 1 3 108.
Weber et al. Genetic Bottleneck in Guadalupe Fur Seal
Figure 2. Comparison of variable positions in the mtDNA control region in Guadalupe fur seals. Modern individuals are
identified by a numerical sign followed by a number (e.g., #1 through #7); total numbers of each modern genotype discovered
are indicated (Bernardi et al. 1998, plus current study). Prebottleneck bone samples are represented with an ‘‘FS’’ prior to the
identification number for each individual seal (e.g., FS1). Modern Guadalupe fur seal sequence 1 is the representative
sequence and all modern and prebottleneck sequences sharing the same nucleotide at that position are indicated by a dot.
One prebottleneck individual (FS25, indicated by a w) has a genotype found in seals today. Question marks designate
missing data; dashes indicate inferred gaps.
149
Journal of Heredity 2004:95(2)
Clearly the population bottleneck resulted in a substantial
loss of genetic variability in the Guadalupe fur seal, similar to
that reported for the sympatric northern elephant seal, which
experienced a population bottleneck during the same time
period (Weber et al. 2000). However, because many
evolutionary factors may influence patterns of genetic
variation in populations (Fu 1997), we also considered
several other indices of polymorphism (Table 2). We found
significantly fewer (P , .001, t9 test) nucleotide differences
between pairs of genotypes (k) of the prebottleneck versus
modern seals (Table 2). Accordingly, nucleotide diversity (p)
was significantly greater (P , .001, t test) for prebottleneck
(0.055 6 0.004) versus postbottleneck seals (0.025 6 0.003).
Nucleotide diversity in the presealing Guadalupe fur seal
population is similar to that reported for extant populations
of other species of fur seals (A. tropicalis, p ¼ 0.048; A. gazella,
p ¼ 0.032; A. forsteri, p ¼ 0.051; A. philippii, p ¼ 0.030)
(Goldsworthy et al. 2000; Wynen et al. 2000), all of which
were reduced to near-extinction levels by commercial sealers
in the 18th and 19th centuries. Other species impacted by
humans (hunting or habitat reduction) show much lower
values of nucleotide diversity today, such as the whooping
crane (0.0045 for both pre- and postbottleneck) (Glenn et al.
1999) and the northern elephant seal (Table 3) (Weber et al.
2000). Because pairwise DNA sequence divergence estimates
are influenced by the length of the sequence (Tajima 1993),
we also measured haplotype (gene) diversity (h), which is the
equivalent of heterozygosity for haploid loci (Nei 1987).
Haplotype diversity was significantly greater (P , .001, t9
test; Table 2) in the ancient (0.997 6 0.012) versus the
modern seals (0.80 6 0.038). For comparison, the sea otter
has apparently lost nuclear heterozygosity as a result of fur
trapping (HE ¼ 0.82 pretrapping versus 0.47 post-trapping),
but this is not mirrored in a loss of mtDNA haplotype
diversity (h ¼ 0.37 pretrapping versus 0.41 post-trapping)
(Larson et al. 2002).
The genetic consequences of founder events or population bottlenecks that result in a loss of genetic variability
may be evaluated by comparing estimates of nucleotide
diversity (hS) and heterozygosity (hp) in populations before
and after those events (Galtier et al. 2000; Hedrick 2000).
Estimates of nucleotide diversity are strongly influenced by
rare alleles, whereas estimates of heterozygosity are relatively
insensitive to them. Consequently hp hS will be negative in
large prebottleneck populations under an infinite-sites model
of mutation drift equilibrium. In contrast, postbottleneck
populations should be in disequilibrium as a result of
a reduction in the number of sites segregating for different
alleles, and hp hS should be positive. Assuming neutrality
for the mtDNA control region, our data for the Guadalupe
fur seal are consistent with an equilibrium state in the
prebottleneck population (hp hS ¼ 1.544) and with a
population in disequilibrium in the modern population (hp hS ¼ 1.815). We further evaluated the status of the ancient
and modern fur seal populations by constructing mismatch
distributions of pairwise differences among nucleotide
sequences (Harpending 1994; Harpending et al. 1993; Hartl
and Clark 1997). We found significant differences in
150
Figure 3. Phylogenetic reconstruction of the relationships
between the pre- and postbottleneck Guadalupe fur seal
mtDNA genotypes. Prebottleneck bones are designated by
‘‘FS’’ and an identification number. The 32 modern seals
possess seven different genotypes (shown as bold lines) that fall
into two fundamental clades in the overall tree, one containing
genotypes 1, 2, 3, 4, and 7 and encompassing clades A and C as
defined by Bernardi et al. (1998), and the other containing
genotypes 5 and 6, from clade B as defined by Bernardi et al.
(1998). One genotype found in ancient seals (FS25) is identical
to a genotype found in modern individuals (#5); these are
placed on different lines for clarity. Other species used in the
phylogenetic analysis include the Arctocephalus species (Wynen
et al. 2001): Juan Fernández fur seal (A. philippii; ‘‘APH’’
genotypes), Australian fur seal (A. pusillus doriferus), subAntarctic fur seal (A. tropicalis), South American fur seal (A.
australis), Antarctic fur seal (A. gazella), and Galapagos fur seal
(A. galapagoensis). The more common (of two) mtDNA
genotype of the northern elephant seal (‘‘NESc’’; Weber et al.
2000) is the designated outgroup. Phylogenetic relationships
were determined by a distance matrix-based method using
a maximum-likelihood model; numbers at nodes, when present,
represent bootstrap values (in percent) from 200
pseudoreplicates; lack of a number indicates that that node is
supported in less than 60% of pseudoreplicates. The robustness
of the two clades, indicated with bootstrap values with asterisks
(*), were confirmed with Shimodaira-Hasegawa tests of
significance (see text).
Weber et al. Genetic Bottleneck in Guadalupe Fur Seal
Table 2.
Comparison of pre- and postbottleneck mtDNA diversity in Guadalupe fur seals
Prebottleneck
Postbottleneck
n
No. of
genotypes
Proportion
of variable
sites, p
Haplotype
diversity,
hb (SD)
Nucleotide
diversity,
pb (SD)
Tajima’s
D statisticb
Fu’s FSc
hp hSc
kb,
26
32a
25
7
0.282
0.057
0.997 (0.012)
0.798 (0.038)
0.055 (0.004)
0.025 (0.003)
0.865
1.35
16.576
4.423
1.544
1.815
8.905
5.238
d
rc,
e
0.009
0.101
a
Modern samples included 25 from G. Bernardi and 7 from this study.
b
The measures h, p, D, and k were obtained with DnaSP (version 3.51; Rozas and Rozas 1999) using 163 bp of the mtDNA control region in the
prebottleneck samples and 212 bp in the postbottleneck (modern) samples.
c
The measures FS and r were obtained with ARLEQUIN (version 2.000; Schneider et al. 2000) using 181 bases of the mtDNA control region in the
prebottleneck samples and 313 bp in the postbottleneck (modern) samples.
d
k is the average number of nucleotide differences (Tajima 1989).
e
r is the raggedness index (Harpending et al. 1993).
population trends between the pre- and postbottleneck
populations (P , .001, t9 test), with a smooth distribution of
sequence matches (raggedness statistic; r ¼ 0.009) for the
prebottleneck population, characteristic of an equilibrium
status, but a substantial mismatch in the extant population
(r ¼ 0.101), characteristic of a recently bottlenecked,
nonequilibrium population (Figure 4).
The inference that the prebottleneck Guadalupe fur seal
population was at mutation drift equilibrium raises the issue
of whether the Guadalupe fur seal population was stable or
increasing prior to the 19th century bottleneck, which may
have implications for continued recovery of the extant
population. We used Tajima’s D test to evaluate the
presealing population status of Guadalupe fur seal, as
negative D values are indicative of historical growth (Hartl
and Clark 1997; Rogers et al. 1996). We found that the
prebottleneck Guadalupe fur seal populations were in an
expansion mode (D ¼ 0.865), with the caveat that not all of
our Guadalupe fur seal bones were likely to be from the same
temporal population. The large positive value of D (1.35) for
the extant population is consistent with a population that has
been interrupted by a recent, substantial population
bottleneck (Schneider and Excoffier 1999). Fu (1997)
Table 3.
Comparison of modern levels of mtDNA diversity in the Guadalupe fur seal and the northern elephant seal
Species
Guadalupe fur seal
Northern elephant
sealc
a
suggested the FS statistic as an alternative measure of
detecting historical population growth. Large negative FS
values are indicative of an excess of rare alleles, suggesting an
increase in recent mutations during population expansion.
The prebottleneck FS of –16.576 (highly significantly
different from zero; P , .0001) is again consistent with
historical population growth. Conversely the postbottleneck
FS value of 4.423 (P ¼ .94) implies that the modern
population has few rare alleles as a consequence of the
bottleneck. The consensus of our computed metrics of
genetic variability is that there was a substantial loss of
genetic variability in the Guadalupe fur seal during the
population bottleneck caused by commercial sealing in the
19th century, and that the population had been growing at
some time prior to that bottleneck.
Amos (1996) suggested that substantial loss of genetic
variation generally requires near extinction or a prolonged
period of low population abundance. Yet purging of
deleterious recessive alleles during such reductions may
sometimes minimize reductions in fitness otherwise associated with low genetic variability. In contrast, delays in
population recovery can cause substantial losses in fitness due
to the buildup of a genetic load of mildly deleterious alleles
No. genotypes
(no. of individuals)
No. of polymorphic
sites (no. of base pairs)
Haplotype
diversity, h
Nucleotide
diversity, p
No. of modern
genotypes found
in ancient bones
hn hS
7 (32)a
18 (313)b
0.798 (0.038)
0.025 (0.003)
2
1.815
1g
0.644
2 (150)
8 (1223)d
0.406 (0.032)e,
f
0.0068 (0.0005)e,
f
Including 25 sequences from Bernardi et al. (1998) and 7 sequences from this study.
b
The sequences from this study were 212 bases in length, which were combined with those from Bernardi et al. (1998).
c
Analysis taken from Weber et al. (2000) and unpublished data.
d
Exclusive of the hypervariable region that displays variable number of tandem repeats (VNTR) variation and heteroplasmy.
e
Measures h and p were obtained with DnaSP (version 3.51; Rozas and Rozas 1999) using 179 bp of the mtDNA control region in 150 northern elephant
seal postbottleneck (modern) samples.
f
Standard deviations for h and p are provided in parentheses.
g
A second modern genotype in northern elephant seals was found in the two skins taken in 1892 at the point of the known demographic bottleneck (Weber
et al. 2000).
151
Journal of Heredity 2004:95(2)
recover very soon after a population bottleneck versus
similarly small populations where recovery is delayed.
Theoretically, low levels of genetic variation may substantially
increase the probability that a species may go extinct—the
‘‘central dogma of conservation genetics’’ (Lehman 1998). Yet
several species, particularly marine mammals, are now robust
and evidently doing very well despite substantial population
and genetic bottlenecks and environmental change (including
anthropogenic change and resource extraction). Consequently our data, and those from other recent studies, suggest
caution in labeling any species as having ‘‘high’’ or ‘‘low’’ levels
of variation, as this may give a misleading snapshot of the
genetic health of the species and be of limited relevance in
assessing its likelihood of persistence. Indeed, the continued
viability of the Guadalupe fur seal may depend primarily on
nongenetic factors, such as poaching or the local availability of
food resources during the breeding season and its consequent
effects on pup growth and survival.
References
Amos B, 1996. Levels of genetic variability in cetacean populations have
probably changed little as a result of human activities. Rep Int Whal Comm
46:657–658.
Austin JJ, Smith AB, and Thomas RH, 1997. Paleontology in a molecular
world: the search for authentic ancient DNA. Trends Ecol Evol 12:303–306.
Bartholomew GA, 1950. A male GFS on San Nicolas Island, California.
J Mamm 31:175–180.
Bernardi G, Fain SR, Gallo-Reynoso JP, Figueroa-Carranza AL, and Le Boeuf
BJ, 1998. Genetic variability in Guadalupe fur seals. J Hered 89:301–305.
Figure 4. Mismatch distribution for Guadalupe fur seal
individuals showing the observed pairwise nucleotide site
differences (dotted line) and the expected (solid line) obtained
with DnaSP (version 3.51; Rozas and Rozas 1999) using 212
bases of the mtDNA control region in the postbottleneck
(‘‘modern’’) samples and 163 bases in the prebottleneck
(‘‘bones’’) samples. The x-axis shows pairwise differences and
the y-axis shows genotype frequency. (A) Prebottleneck
individuals show a smooth distribution, indicating a population
in mutation drift equilibrium that has expanded historically. (B)
Postbottleneck individuals show a ragged (erratic) distribution,
indicating a population in disequilibrium that is less likely to be
expanding.
and more lethal, but recessive alleles in the heterozygous
condition. Consequently the patterns of genetic variability in
postbottleneck populations may be determined more by the
patterns of population recovery than by the magnitude of the
bottleneck alone. Our empirical evidence with mtDNA
suggests that the presealing population of Guadalupe fur
seals was a robust and outbred one of high fitness, in contrast
to the small, inbred extant population. Moreover, our results
are consistent with theoretical models of the genomic effects
of population bottlenecks (Kirkpatrick and Jarne 2000; Miller
and Hedrick 2001) that predict more substantial loss of
genetic variability in small populations that may begin to
152
Bouzat JL, Cheng HH, Lewin HA, Westemeier RL, Brawn JD, and Paige
KN, 1998. Genetic evaluation of a demographic bottleneck in the greater
prairie chicken. Conserv Biol 12:836–843.
Caro TM and Laurenson MK, 1994. Ecological and genetic factors in
conservation: a cautionary tale. Science 263:485–486.
Cooper A and Wayne R, 1998. New uses for old DNA. Curr Opin
Biotechnol 9:49–53.
Etnier MA, 2002. Occurrences of Guadalupe fur seals (Arctocephalus
townsendi) on the Washington coast over the past 500 years. Mar Mamm Sci
18:551–557.
Fleischer LA, 1978. The distribution, abundance, and population characteristics of the Guadalupe fur seal, Arctocephalus townsendi (Merriam 1897)
(Master’s thesis). Seattle, WA: University of Washington.
Fu Y-X, 1997. Statistical tests of neutrality of mutations against population
growth, hitchhiking, and background selection. Genetics 147:915–925.
Gallo-Reynoso JP, 1994. Factors affecting the population status of
Guadalupe fur seal, Arctocephalus townsendi (Merriam, 1897), at Isla de
Guadalupe, Baja California, Mexico (PhD dissertation). Santa Cruz, CA:
University at California, Santa Cruz.
Galtier N, Depaulis F, and Barton NH, 2000. Detecting bottlenecks and
selective sweeps from DNA sequence polymorphism. Genetics 155:981–987.
Glenn TC, Stephan W, and Braun MJ, 1999. Effects of a population
bottleneck on whooping crane mitochondrial DNA variation. Conserv Biol
13:1097–1107.
Goldsworthy S, Francis J, Boness D, and Fleischer R, 2000. Variation in the
mitochondrial control region in the Juan Fernandez fur seal (Arctocephalus
philippii). J Hered 91:371–377.
Gustafson CE, 1968. Prehistoric use of fur seals: evidence from the
Olympic coast of Washington. Science 161:49–51.
Weber et al. Genetic Bottleneck in Guadalupe Fur Seal
Hagelberg E, 1994. Mitochondrial DNA from ancient bones. In: Ancient
DNA: recovery and analysis of genetic material from paleontological,
archaeological, museum, medical, and forensic specimens (Herrmann B and
Hummel S, eds). New York: Springer-Verlag; 195–204.
Schneider S and Excoffier L, 1999. Estimation of past demographic
parameters from the distribution of pairwise differences when the mutation
rates vary among sites: application to human mitochondrial DNA. Genetics
152:1079–1089.
Hagelberg E and Clegg JB, 1991. Isolation and characterisation of DNA
from archaeological bone. Proc R Soc Lond B 244:45–50.
Schneider S, Roessli D, and Excoffier L, 2000. Arlequin ver. 2000:
a software for population genetics data analysis. Geneva, Switzerland:
University of Geneva.
Harpending HC, 1994. Signature of ancient population growth in a low-resolution mitochondrial DNA mismatch distribution. Hum Biol 66:591–600.
Harpending HC, Sherry ST, Rogers AR, and Stoneking M, 1993. The genetic
structure of ancient human populations. Curr Anthropol 34:483–496.
Hartl DL and Clark AG, 1997. Principles of population genetics, 3rd ed.
Sunderland, MA: Sinauer Associates.
Hedrick PW, 1996. Bottleneck(s) or metapopulation in cheetahs. Conserv
Biol 10:897–899.
Hedrick PW, 2000. Genetics of populations, 2nd ed. Sudbury, Ontario:
Jones and Bartlett.
Hubbs CL, 1956. Back from oblivion. Guadalupe fur seal: still a living
species. Pacific Disc 9:14–51.
Hummel S and Herrmann B, 1994. General aspects of sample preparation.
In: Ancient DNA: recovery and analysis of genetic material from
paleontological, archaeological, museum, medical, and forensic specimens
(Herrmann B and Hummel S, eds). New York: Springer; 59–68.
Kirkpatrick M and Jarne P, 2000. The effects of a bottleneck on inbreeding
depression and the genetic load. Am Nat 155:154–167.
Larson S, Jameson R, Etnier M, Fleming M, and Bentzen P, 2002. Loss of
genetic diversity in sea otters (Enhydra lutris) associated with the fur trade of
the 18th and 19th centuries. Mol Ecol 11:1899–1903.
Lehman N, 1998. Genes are not enough. Curr Biol 8:1078–1081.
Lyons GM, 1937. Pinnipeds and a sea otter from the Point Mugu shell
mound of California. Publications of the University of California at Los
Angeles in Biological Sciences 1:133–168.
Maravilla-Chavez MO and MSLowry, 1999. Incipient breeding colony of
Guadalupe fur seals at Isla Benito del Este, Baja California, Mexico. Mar
Mamm Sci 15:239–241.
Miller PS and Hedrick PW, 2001. Purging of inbreeding depression and
fitness decline in bottlenecked populations of Drosophila melanogaster. J Evol
Biol 14:595–601.
Nei M, 1987. Molecular evolutionary genetics. New York: Columbia
University Press.
Nei M, Maruyama T, and Chakraborty R, 1975. The bottleneck effect and
genetic variability in populations. Evolution 29:1–10.
Nicholas KB, Nicholas HB Jr and Deerfield DW II 1997. GeneDoc:
analysis and visualization of genetic variation. EMBNEW.NEWS 4:14.
O’Brien SJ, 1994. The cheetah’s conservation controversy. Conserv Biol
8:1153–1155.
O’Brien SJ, Wildt DE, Bush M, Caro TM, FitzGibbon C, Aggundey I, and
Leakey RE, 1987. East African cheetahs: evidence for two population
bottlenecks?. Proc Natl Acad Sci USA 84:508–511.
Reeves RR and Stewart BS, 2003. Marine mammals of the world: an
introduction. In: Walker’s marine mammals of the world (Nowak RM, ed).
Baltimore: Johns Hopkins University Press; 1–64.
Reeves RR, Stewart BS, Clapham P, and Powell J, 2002. Guide to marine
mammals of the world. New York: Knopf.
Reeves RR, Stewart BS, and Leatherwood S, 1992. The Sierra Club
handbook of seals and sirenians. San Francisco: Sierra Club Books.
Rogers AR, Fraley AE, Bamshad MJ, Watkins WS, and Jorde LB, 1996.
Mitochondrial mismatch analysis is insensitive to the mutational process.
Mol Biol Evol 13:895–902.
Rozas J and Rozas R, 1999. DnaSP version 3: an integrated program for
molecular population genetics and molecular evolution analysis. Bioinformatics 15:174–175.
Stemmer WPC, 1994. DNA shuffling by random fragmentation and
reassembly: in vitro recombination for molecular evolution. Proc Natl Acad
Sci USA 91:10747–10751.
Stewart BS, 1997. California pinnipeds: population trends, trans-jurisdictional migrations, and ecological function in large marine ecosystems of the
eastern north Pacific Ocean. In: Pinniped populations, eastern north
Pacific: status, trends, and issues (Stone G, Goebel J, and Webster S, eds).
Monterey: American Fisheries Society; 13–21.
Stewart, BS, Yochem PK, DeLong RL, and Antonelis GA, 1993. Trends in
abundance and status of pinnipeds on the southern California Channel
Islands. In: Third California Channel Islands symposium: recent advances
in research on the California islands, Santa Barbara, CA, 2–6 March 1987
(Hochberg FG, ed). Santa Barbara, CA: Santa Barbara Museum of Natural
History; 501–516.
Swofford DL, 2002. PAUP*: phylogenetic analysis using parsimony (*and
other methods), version 4. Sunderland, MA: Sinauer Associates.
Tajima F, 1993. Measurement of DNA polymorphism. In: Mechanisms of
molecular evolution: introduction to molecular paleopopulation biology
(Takahata N and Clark AG, eds). Sunderland, MA: Sinauer Associates; 37–59.
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, and Higgins DG,
1997. The ClustalX windows interface: flexible strategies for multiple
sequence alignment aided by quality analysis tools. Nucleic Acids Res
24:4876–4882.
Tindall KR and Kunkel TA, 1988. Fidelity of DNA synthesis by the Thermus
aquaticus DNA polymerase. Biochemistry 27:6008–6013.
Townsend CH, 1916. The Guadalupe fur seal. A contribution to its history.
Bull N Y Zool Soc 19:1330–1331.
Townsend CH, 1928. Reappearance of the lower California fur seal. Bull N
Y Zool Soc 31:173–174.
Townsend CH, 1931. The fur seal of the California islands with new
descriptive and historical matter. Zoologica 9:442–457.
Weber DS, Stewart BS, Garza JC, and Lehman N, 2000. An empirical
genetic assessment of the severity of the northern elephant seal population
bottleneck. Curr Biol 10:1287–1290.
Westemeier RL, Brawn JD, Simpson SA, Esker TL, Jansen RW, Walk JW,
Kershner EL, Bouzat JL, and Paige KN, 1998. Tracking the long-term
decline and recovery of an isolated population. Science 282:1695–1698.
Wynen LP, Goldsworthy SD, Guinet C, Bester MN, Boyd IL, Gjertz I,
Hofmeyr GJG, White RWG, and Slade R, 2000. Postsealing genetic
variation and population structure of two species of fur seal (Arctocephalus
gazella and A. tropicalis). Mol Ecol 9:299–314.
Wynen LP, Goldsworthy SD, Insley SJ, Adams M, Bickham JW, Francis J,
Gallo JP, Hoelzel AR, Majluf P, White RWG, and Slade R, 2001.
Phylogenetic relationships within the eared seals: implications for the
historical biogeography of the family. Mol Phylogenet Evol 21:270–284.
Yang DY, Eng B, Waye JS, Dudar JC, and Saunders SR, 1998. Improved
DNA extractions from ancient bones using silica-based spin columns. Am J
Phys Anthropol 105:539–543.
Received November 23, 2002
Accepted September 15, 2003
Corresponding Editor: Stephen J. O’Brien
153