MOLECULAR
PHYLOGENETICS
AND
EVOLUTION
Molecular Phylogenetics and Evolution 23 (2002) 171–188
www.academicpress.com
Extraordinary haplotype diversity in haplodiploid inbreeders:
phylogenetics and evolution of the bark beetle genus Coccotrypes
Bjarte H. Jordal,a,* Benjamin B. Normark,b Brian D. Farrell,c and Lawrence R. Kirkendalld
a
School of Biological Sciences, University of East Anglia, NR4 7TJ, Norwich, UK
Department of Entomology, Fernald Hall, University of Massachusetts, Amherst, MA 01003, USA
Museum of Comparative Zoology, Harvard University, 26 Oxford St., Cambridge, MA 02138, USA
d
Department of Zoology, University of Bergen, Allegt. 41, N-5007 Bergen, Norway
b
c
Received 9 July 2001; received in revised form 19 November 2001
Abstract
Regular inbreeding by sib-mating is one of the most successful ecological strategies in the bark beetle family Scolytinae. Within
this family, the many species (119) in Coccotrypes are found breeding in an exceptional variety of untraditional woody tissues
different from bark and phloem. Species delineation by morphological criteria is extremely difficult, however, as in most other
inbreeding groups of beetles, perhaps due to the unusual evolutionary dynamics characterizing sib-mating organisms. Hence, we
here performed a phylogenetic analysis using molecular data in conjunction with morphological data to better understand morphological and ecological evolution in this sib-mating group. We used partial DNA sequences from the nuclear gene EF-a and the
mitochondrial genes 12S and CO1 to elucidate patterns of morphological evolution, haplotype variation, and evolutionary pathways
in resource use. Sequence variation was high among species and far above that expected at the species level (e.g., 19% for CO1 within
Coccotrypes advena). The tendency for exhaustive sequence variation at deeper nodes resulted in ambiguous reconstructions of the
deepest splits. However, all results suggested that species with the broadest diets were clustered in a single derived position—another
piece of evidence against specialization as a derived evolutionary feature. Ó 2002 Elsevier Science (USA). All rights reserved.
Keywords: 12S; CO1; Dryocoetini; EF-1a; Maximum-likelihood; Maximum-parsimony; Morphology; RASA; Scolytinae; Seed feeding; Specialization
1. Introduction
Our empirical understanding of the origin and evolution of haplodiploid clades is poor. Most extant haplodiploid clades are very ancient and divergent from
their closest relatives, and our understanding of their
origins is hampered by uncertainty about phylogenetic
topology (e.g., pinworms, iceryine scale insects), by uncertainty about reconstruction of ancestral ecological
characters (e.g., Hymenoptera, monogonont rotifers), or
by both sources of uncertainty at once (e.g., mite clades,
Thysanoptera, whiteflies) (Normark et al., 1999). An
exception to this generalization is the haplodiploid, inbreeding clade of scolytine beetles within the tribe
*
Corresponding author. Fax: +44(0)1603592250.
E-mail address: b.jordal@vea.ac.uk (B.H. Jordal).
Dryocoetini. We now have a precise understanding of its
basal relationships (Jordal et al., 2000, 2002; Normark
et al., 1999), across which many important ecological
characters remain constant, yielding an exceptionally
clear picture of the ecological context in which haplodiploidy arose in this case. However, important sources
of uncertainty remain in our understanding of the evolution of the clade.
We now have clear evidence that the basal lineage in
the clade is the genus Ozopemon Hagedorn (Jordal et al.,
2002), and the preponderance of the evidence suggests
that the remainder of the clade consists of two monophyletic sister taxa: (a) the tribe Xyleborini and (b) the
genus Coccotrypes Eichhoff (including Dryocoetiops; see
Jordal et al., 2000, 2002). Two of these taxa are fairly
easy to characterize ecologically: all Ozopemon are
phloem-feeders and all Xyleborini feed on ambrosia
fungi cultured on the walls of tunnels bored into xylem.
1055-7903/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S 1 0 5 5 - 7 9 0 3 ( 0 2 ) 0 0 0 1 3 - 1
172
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
However, Coccotrypes is not so easily characterized.
Though it has barely more than 10% of the number of
species found in Xyleborini, the range of their feeding
habits is vastly greater. Some of them, like Ozopemon,
have retained the ancestral character of feeding in
phloem, but almost all of them have dwarf males like
Xyleborini. Clearly, Coccotrypes is a critical group for
understanding the evolution of this haplodiploid clade.
Coccotrypes is a fairly large genus of sib-mating
beetles, with currently 119 species assigned to it (Bright
and Skidmore, 1997; Wood and Bright, 1992). Most
species are distributed in humid tropical forests of Asia
and Africa, and recent introductions of several species
into the Neotropics has made the genus circumtropical
(Bright and Peck, 1998; Wood, 1982). Among the most
widespread species, we find a few species that also are
well adapted to drier conditions in warm temperate areas, where they feed and breed in small seeds, in particular palm seeds. The best known species, the ‘‘ivory
button’’ beetle Coccotrypes dactyliperda (Fabr.) is well
known as a pest of palm seeds (Blumberg and Kehat,
1982; Hamilton, 1993; Herfs, 1950).
Overall, the species of Coccotrypes have adopted a
wide range of unusual resources for feeding and breeding. They have been collected from many different fruits
and seeds (Browne, 1961), leafstalks (Beaver, 1979a;
Jordal and Kirkendall, 1998), phloem, and the pith of
twigs (Browne, 1959). Coccotrypes is the only bark
beetle genus known to breed in ferns (Browne, 1973;
Gray, 1970, 1972) and in mangrove radicles (Browne,
1961; Woodruff, 1970). In contrast to the narrow specialization of most scolytines on a particular host tissue—usually phloem, as the common name ‘‘bark
beetle’’ implies—fewer than 10 species of Coccotrypes
use phloem exclusively (Browne, 1959). In contrast,
many species of Coccotrypes are host-tissue (and hosttaxon) generalists, capable of breeding in two or more
widely different tissues, usually in a wide range of tree
species.
The reproductive biology of the entire haplodiploid
clade (1400 spp) is characterized by strongly femalebiased offspring sex ratios, in which a single dwarfed
male develops in close association with his many sisters
(Kirkendall, 1993). In all cases where experimental or
cytological studies have been reported, the males are
produced by arrhenotokous parthenogenesis (Entwistle,
1964; Herfs, 1950; Ueda, 1997) and are haploid (Takenouchi and Takagi, 1967; B.B.N. and B.H.J., unpublished flow cytometry data for C. dactyliperda). In
conjunction with regular sib-mating, haplodiploidy
contributes an advantage to colonizing females, which
can either mate with a brother before dispersal, or
produce haploid sons parthenogenetically, one of which
she can eventually mate with (Kirkendall, 1993). Sibmating species of Scolytinae contribute a disproportionately high fraction of the bark and timber beetle
faunas on tropical islands, and the wide distribution and
ecological success for sib-mating species might be explained by their extraordinary reproductive biology
(Jordal et al., 2001).
Inbreeding species complexes often tend to be taxonomically difficult to resolve, and inbreeding groups of
Scolytinae are no exception (Wood, 1982). Until recently, Coccotrypes and Poecilips Schaufuss were considered two genera difficult to distinguish (Browne,
1961) and the latter was synonomized when complete
character intergradation was found (Wood, 1973). Long
lists of generic and specific synonymies (Wood and
Bright, 1992) reflect the problematic classification of this
genus, whose species are regarded particularly difficult
to identify (Wood, 1986). A similar pattern of ambiguous morphology and taxonomic synonymies is also observed in the inbreeding Cryphalini and Xyleborini
(Wood and Bright, 1992), demonstrating a general taxonomic problem with sib-mating species.
The genus Coccotrypes is of great interest due to its
remarkable sib-mating habits, its extremely great lability
of ecological characters, both within and between species, compared with other scolytine beetles, and its pivotal position in the evolutionary history of a
haplodiploid clade. Here we present the first phylogenetic study of the genus, using characters from mitochondrial and nuclear DNA, morphology, and
behavior.
2. Material and methods
2.1. Specimens
The species analyzed, the GenBank accession numbers, and their geographical sources and main food resources are listed in Table 1. Samples were collected into
100% ethanol or acetone by the senior author or persons
credited in the acknowledgements. Vouchers are deposited at the Museum of Comparative Zoology, Harvard University or at the Department of Zoology,
University of Bergen. We included 26 specimens from 19
species of Coccotrypes which were selected to represent
the overall variation in morphological and ecological
characters. Two species of Dryocoetiops, which are
known to be part of a paraphyletic Coccotrypes (Jordal
et al., 2000, 2002, and the outgroups Ozopemon and
Dryocoetes were also included.
2.2. Morphological and behavioral characters
Adults of each species were extensively examined
and compared, and 28 morphological characters were
coded for phylogenetic analysis. Because larval characters are only marginally informative for scolytine
genera (Lekander, 1968), larvae were not examined. In
173
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Table 1
Collecting area for the species sequenced in this study, along with their main food resource (monophagous resources in boldface), and GenBank
accession numbers
Species
Alias
Collection
Food source
Acc. EF-1a
Acc. 12S
Acc. CO1
Coccotrypes aciculatus Schedl
Coccotrypes advena Blandford
dry16
scd24
Equador
Uganda
—
AF186664
AF444047
AF259818
AF444059
AF187116
Coccotrypes advena Blandford
dry11
Uganda
AF444072
AF444044
AF444056
Coccotrypes advena Blandford
scd32
Japan
AF186668
AF259821
AF187120
Coccotrypes advena Blandford
scd36
Borneo
AF259868
AF259824
AF444068
Coccotrypes advena Blandford
dry20
Costa Rica
AF444076
AF444050
AF444062
Coccotrypes cardamomi Schaufuss
scd38
Japan
AF259869
AF259826
AF375308
Coccotrypes carpophagus (Hornung)
Coccotrypes cf. cardamomi
scd43
dry12
USA, Florida
PNG
AF259872
AF444073
AF259829
AF444045
AF444069
AF444057
Coccotrypes cyperi (Beeson)
scd11
Costa Rica
AF259863
AF259813
AF375309
Coccotrypes cyperi (Beeson)
dry15
Queensland
Seed
seed, fruit,
petiole, phloem
seed, fruit,
petiole, phloem
seed, fruit,
petiole, phloem
seed, fruit,
petiole, phloem
seed, fruit,
petiole, phloem
seed, fruit,
petiole, fern,
phloem
palm seed
petiole, palm seed,
more?
seed, fruit,
petiole, phloem,
pith
seed, fruit,
petiole, phloem,
pith
palm seed
palm seed
palm seed
petiole, more?
mangrove radicle
mangrove radicle
seed, fruit,
petiole, pith
small seed
small seed
mangrove radicle
phloem (also
under bark of
petioles)
petiole
phloem, petiole,
fruit
petiole
fruit, seed,
petiole, phloem
phloem
phloem
pith
pith (also pith
of petiole)
phloem
phloem
AF444074
AF444046
AF444058
AF186659
AF444078
AF444075
AF444071
AF259867
AF259815
AF444052
AF444048
AF444043
AF444053
AF444049
AF259823
AF187111
AF444064
AF444060
AF444055
AF444066
AF444061
AF375310
AF259866
AF259874
AF259864
AF259871
AF259820
AF259831
AF259814
AF259828
AF438513
AF444070
AF444065
AF438515
AF186669
AF259875
AF259822
AF259832
AF187121
AF438518
AF444077
AF259865
AF444051
AF259819
AF444063
AF444067
AF259817
AF259873
AF439741
AF186670
AF259817
AF259830
AF438492
AF259825
AF187113
AF444054
AF438507
AF187122
AF259870
AF439740
AF259827
AF438491
AF438514
AF438506
Coccotrypes
Coccotrypes
Coccotrypes
Coccotrypes
Coccotrypes
Coccotrypes
Coccotrypes
dactyliperda (Fabricius)
dactyliperda (Fabricius)
cf. distinctus (Motschulsky)
sp. near rhizophorae
fallax (Eggers)
fallax (Eggers)
gedeanus (Eggers)
scd14
dry22
dry17
dry09
scd13
dryl8
scd35
Argentina
USA, Florida
Costa Rica
PNG
Queensland
Bangladesh
Borneo
Coccotrypes
Coccotrypes
Coccotrypes
Coccotrypes
graniceps (Eichhoff)
impressus Eggers
litoralis (Beeson)
longior (Eggers)
scd31
scd54
scd12
scd40
Japan
Thailand
Bangladesh
Borneo
Coccotrypes marginatus (Browne)
Coccotrypes medius (Eggers)
scd34
scd55
Singapore
Singapore
Coccotrypes petioli (Browne)
Coccotrypes variabilis (Beeson)
dry21
scd30
Borneo
Japan
Dryocoetes affaber (Mannerhein)
Dryocoetes autographus (Ratzeburg)
Dryocoetiops cf. eugeniae (Schedl)
Dryocoetiops coffeae (Eggers)
scd20
scd53
dry19
scd37
USA, NH
Japan
Malaysia
Borneo
Ozopemon brownei Schedl
Ozopemon uniseriatus Eggers
scd39
dry04
Borneo
PNG
—
—
Note. Alias refers to voucher specimens (see Section 2).
addition, 1 developmental character and 2 behavioral
characters were compiled from the literature and
checked against the location, structure, and occupancy
of the gallery systems from which the beetles were
collected (Appendix A). These characters were included
in the ‘‘morphology’’ data partition. The evolution of
each character was traced on the total-evidence trees,
and consistency indices for characters from different
body regions were tabulated and compared. Particular
attention was paid to the evolution of character 30,
resource use.
2.3. DNA sequencing and alignments
DNA was extracted and amplified using primers
and protocols described elsewhere (Jordal et al., 2000;
174
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Normark et al., 1999). Sequences were assembled and
edited in Sequencher 3.1 (Gene Codes Corp., Ann
Arbor, MI). Of the two different copies of EF-1a, we
used the copy that has only one intron, between
coding positions 753 and 754 in scolytine beetles
(Danforth and Ji, 1998; Normark et al., 1999). The
intron was aligned and used in the analyses. For CO1
and the EF-1a coding region, alignments were unambiguous due to the lack of any insertions or deletions in the coding region. Alignments of 12S and the
EF-1a intron were done in ClustalX (Thompson et al.,
1997) under four different gap cost ratios (2, 4, 8, and
16) with transitions and transversions weighted equally
and with opening and extension gaps equally costly.
Maximum-parsimony (MP) trees were inferred by
heuristic search (see below) for each alignment. For
the EF-1a intron, the strict consensus tree for each
alignment was compared by inspection with the strict
consensus tree for the EF-1a coding region, and the
alignment yielding the most congruent consensus tree
(gap cost 4) was chosen for further analysis. Eight
base pairs consisting of AT-repeats unique to C.
litoralis were deleted from the EF-1a intron matrix.
For 12S, an elision matrix was constructed (Wheeler
et al., 1995) and the strict consensus trees for the individual alignments were compared with the strict
consensus tree for the elision matrix. The alignment
yielding the most congruent consensus tree (gap cost
8) was chosen for further analysis. Two inserted sites
unique to C. longior were deleted; the matrix was reanalyzed (again with gap cost 8), resulting in a 12S
matrix with no gaps occurring in that region. After
alignment, the combined matrix was 2239 bp in length,
including 324 aligned sites of 12S and 65 aligned sites
of EF-1a intron. Due to the high degree of consistency between different alignments, the few remaining
gaps were each treated as a fifth character state.
2.4. Optimal outgroup analysis
A preponderance of molecular and morphological
evidence suggests that Coccotrypes + Dryocoetiops
constitutes a monophyletic group whose sister group is
the tribe Xyleborini (Jordal, 2001). However, the
precise relationship between Xyleborini and Coccotrypes + Dryocoetiops is unclear; one may be paraphyletic with respect to the other (Jordal et al., 2000).
We are sufficiently persuaded of the monophyly of
Coccotrypes + Dryocoetiops to treat them as monophyletic in this paper, but for rooting their phylogeny we have conservatively chosen the unequivocal
outgroup Ozopemon (Jordal et al., 2000, 2002; Farrell
et al., 2001). As a further outgroup to test for sensitivity to outgroup composition in a functional ingroup/outgroup analysis (Watrous and Wheeler,
1981), we have selected the sister group of the
haplodiploid clade, the genus Dryocoetes (Normark
et al., 1999).
The program RASA (Lyons-Weiler, 1998) was used
to explore putative plesiomorphy contents of the different outgroup genera (Lyons-Weiler et al., 1998),
separate and in combination, using all data unweighted.
The outgroup composition that results in the highest
increase in phylogenetic signal for the ingroup in a
rooted analyses is supposed to provide the best estimate
of the ingroup root.
2.5. Phylogenetic analysis
MacClade was used for editing matrices, translating
to amino acids, translating to standard format, and
calculating substitution frequencies. Paup* 4.0 (Swofford, 1999) was used for calculating pairwise sequence
transitions and transversions, base composition, sequence divergences, and tree statistics and for all
phylogenetic analyses. Under the maximum-parsimony
criterion, we performed 100 random-addition replicates
of heuristic searches for each of the data partitions and
the combined matrix. Bootstrap support for individual
nodes was assessed by 100 bootstrap replicates (Felsenstein, 1985) of 10 random-addition heuristic searches
each. For the least variable data partition, the morphology matrix, we set maxtrees to 1000 during bootstrapping.
Maximum-likelihood (ML) was used on the molecular data only, with parameters estimated on the topology resulting from the weighted parsimony analysis.
We used a general time reversible model in which the
proportion of invariable sites and a gamma distribution of rates of substitution were estimated. Several
iterations were done to optimize these parameters, until
the topology did not change.
We used the incongruence difference length (ILD)
test (Farris et al., 1995a,b) to measure level of incongruence between data sets. Although we would ultimately combine all data in a total-evidence approach
(Kluge, 1989, 1998), we used the ILD test to evaluate
eventual lack of resolution or topological conflicts.
TreeRot (Sorenson, 1999) was used to calculate partitioned Bremer support indices (Baker and DeSalle,
1997; Baker et al., 1998) showing the relative contribution from each data partition to the combined
analysis. We used heuristic searches with 20 randomaddition sequences for each node in each of the partitions.
We further explored the relative performance of each
data partition using RASA to measure phylogenetic
signal (Lyons-Weiler et al., 1996) in the ingroup for each
of the data partitions and compared these measures to
the overall bootstrap support and saturation plots for
the same partitions. We used this information to inform
differential weighting between partitions.
175
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
3. Results
3.1. Sequence structure and divergence
There was a slight (EF-1a coding region) to strong
(mitochondrial and EF-1a intron) AT bias for all gene
fragments (Table 2). This bias was also associated with
generally high substitution rates, in particular in the
mitochondrial 12S and CO1 third codon positions and
in the nuclear noncoding EF-1a intron (Fig. 1). While
there was no significant heterogeneity in base composition between taxa, among-site rate variation was pronounced within the protein-encoding EF-1a and CO1
(Table 3), reflecting different rates of substitution in the
three codon positions.
A direct comparison between the EF-1a coding region and the CO1 highlights the threefold higher divergence in the latter, for the ingroup (Table 3). The
maximum intrageneric CO1 divergence was 20.4% (uncorrected), and most pairwise comparisons had more
than 15% divergence, with slightly lower levels for 12S.
Even between individuals of the morphologically nearly
invariable C. advena complex, we observed a remarkable
19.0% divergence in the CO1 sequences. High variation
was also observed between individuals of the mangrove
beetle, C. fallax (9.5%). In contrast, the morphologically
distinguishable species C. cyperi and another mangrove
beetle C. litoralis were only 2.6% divergent, indicating
more rapid morphological and ecological evolution in
this clade.
The high rates of substitution in CO1 third positions
resulted in high levels of homoplasy (Fig. 1). More than
half of the informative characters for these sites were
changing eight times or more over the total-evidence tree.
This same subpartition also revealed evident saturation
from the plot of transitions vs transversions, as did 12S
(Fig. 1). On the contrary, we observed a complete absence of saturation of transitions in the least variable
molecular data partition, EF-1a, for the ingroup. More
than 80% of the variable characters from the latter partition changed fewer than four times through the totalevidence tree. Despite the high sequence divergence
found in the much more rapidly evolving EF-1a intron,
this subpartition mimicked the saturation plot for the
coding region (Fig. 1), albeit with a much lower relative
rate of transitional substitutions (Table 3).
Saturation in both of the mtDNA gene partitions was
also evident from exhaustive sequence divergence observed at deeper nodes (Table 3). This trend was not
observed in the EF-1a intron which instead revealed
considerable increase in sequence divergence also from
the ingroup to the primary outgroup Ozopemon. The
exhaustive sequence variation and marked saturation in
transitional substitutions in the two mtDNA gene partitions also coincided with the lowest phylogenetic signals
measured by the RASA regression method (Table 4).
Table 2
Base composition for partial gene sequences
Locus
Position
A
C
G
T
AT-bias
EF-1a
1
2
3
1+2+3
Intron
0.29
0.30
0.23
0.27
0.28
0.17
0.22
0.26
0.22
0.13
0.34
0.15
0.13
0.21
0.08
0.20
0.32
0.40
0.30
0.52
0.49
0.62
0.63
0.57
0.80
CO1
1
2
3
1+2+3
0.30
0.19
0.48
0.33
0.18
0.24
0.16
0.19
0.26
0.16
0.03
0.15
0.25
0.41
0.33
0.33
0.55
0.60
0.81
0.66
12S
All
0.39
0.19
0.09
0.33
0.72
3.2. Separate analyses
As a means to assess the relative contribution from
each data partition to the total-evidence hypothesis, we
analyzed each partition separately (Fig. 2). Clearly, the
EF-1a data, which showed the lowest substitution frequency (Fig. 1) and the second highest RASA signal
(Table 4), had also the highest overall bootstrap support
and the highest proportion (52%) of partial Bremer
support values (Table 5). Adding the EF-1a intron to
the analysis of the coding region resulted in higher resolution with increased bootstrap support. Although
CO1 also contributed much to the total Bremer support,
half of this support was confined to two nodes involving
conspecifics (nodes 10 and 23 in Fig. 3).
EF-1a resulted in a consensus topology most similar
to the consensus resulting from morphological data
(Fig. 2). With the inclusion of the mtDNA gene fragments, which had the lowest RASA signal and the
fewest bootstrap-supported nodes, the overall support
did not improve. If Ozopemon was used as the sole
outgroup, the number of supported nodes actually decreased with the addition of mtDNA data.
Separate analyses of amino-acid-translated CO1 sequences did not perform much better than CO1 nucleotides (cf., Pruess et al., 2000; Simmons, 2000).
However, when CO1 amino acids were substituted for
nucleotides in the combined analysis, the number of
ingroup nodes with more than 50% bootstrap support
increased from 12 to 15.
3.3. Combined analyses: optimal outgroup analysis
The ingroup alone contained significant and high
phylogenetic signal (tRASA ¼ 11.82) for all data combined (Table 4). The outgroup designation with the
highest increase in signal for the ingroup was the
phylogenetic sister group Ozopemon alone (tRASA ¼
19.48), followed by all outgroups simultaneously
(tRASA ¼ 17.54). However, because both of the
Ozopemon species had significantly long branches as
176
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Fig. 1. DNA substitution patterns of the three gene fragments analyzed. Numbers of nucleotide changes were estimated over the tree topology
depicted in Fig. 3. Plots of transitions vs transversions reflect pairwise divergences. For EF-1a, open bars and filled dots indicate third positions for
the exon and hatched bars and open dots indicate the intron. For CO1, open bars and dots indicate first + second codon positions and hatched bars
and filled dots indicate third positions only.
indicated by the taxon variance ratio plot (Lyons-Weiler
and Hoelzer, 1997), we should be cautious with interpreting topologies where only Ozopemon are used as
outgroups, and hence we continue here with exploring
the effect of outgroup composition on the ingroup
rooting, with preferentially all outgroups included.
177
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Table 3
Sequence variation for different subsets of the genes used in this study
Locus
Position
Total
characters
Informative
characters
ti/tv ratiosa
Max
divergence
C. advena
Max
divergence
Coccotrypes
Max
divergence
Cocc + Ozop
Max
divergence
Cocc + Dryo
EF-1a
1
2
3
All coding
Intron
285
285
284
854
65
9
6
139
154
44
14.3
0.6
4.3
–
1.1b
1.1
1.4
10.6
4.3
15.2
2.4
1.4
17.7
6.2
34.8
2.1
1.4
27.9
10.5
41.3
3.3
2.2
27.3
10.6
41.3
CO1
1
2
3
All coding
aa
332
332
332
996
332
71
14
295
380
38
3.5
1.3
1.0
–
–
9.3
1.8
48.0
19.0
3.6
12.3
2.7
49.6
20.4
6.6
11.1
2.4
48.2
20.5
6.3
13.3
2.4
45.0
19.5
7.5
12S
All
324
112
1.5
11.1
18.7
19.9
19.5
Note. Cocc, Coccotrypes; Dryo, Dryocoetes; Ozop, Ozopemon; aa, amino acids
a
ti/tv signifies the relative proportions of transitional to transversional substitutions.
b
ti/tv ¼ 1.8 when averaged over MP tree. For other subsets, tree-based ratios differed marginally from the average pairwise estimates in the table.
Table 4
Phylogenetic signal for 25 ingroup taxa and eventual outgroups as measured by RASA and the number of ingroup nodes supported by 50 and 90%
bootstrap support
Partition
bobs
b0
tRASA
df
> 50 BP
> 90 BP
EF-1a, ingroup
12S, ingroup
CO1 nuc, ingroup
CO1 aa, ingroup
Morphology, ingroup
All data, ingroup
—rooted all outgroups
—rooted Dryocoetes
—rooted Ozopemon
7.46
7.77
5.73
9.26
11.04
8.95
15.32
13.04
13.99
3.41
4.82
3.70
6.58
6.83
5.29
10.48
8.70
8.39
13.26
9.68
7.36
8.91
14.55
11.82
17.54
14.16
19.48a
272
272
272
272
272
272
272
272
272
16
8
6
8
7
–
14
16
15
7
2
3
1
3
–
10
10
9
a
Outgroup with the highest putative plesiomorphy content.
3.4. Combined analyses: ingroup rooting
Choice of outgroup affected the placement of the
ingroup root. Each of the three outgroup compositions
resulted in a different root. In the unweighted MP
analysis of 29 taxa the ingroup root consisted of one
large basal polytomy (indicated by Bremer support of
zero in Fig. 3). The topological conflict between the
two most parsimonious trees may be a result of conflict among outgroups. When using only Dryocoetes
species as outgroups, C. longior was the basal taxon
supported by 71% bootstrap support. When only
Ozopemon species were used as outgroups, a clade of
seed feeders (C. cf. distinctus, C. impressus, C. carpophagus, and C. dactyliperda) was basal, supported by
58% bootstrap support, a result similar to the ML
analysis with both outgroup genera present (Fig. 4).
When the rapidly evolving mtDNA gene fragments
were downweighted by a factor 0.5, C. longior, C.
petioli, and C. marginatus made up the basal clade (all
outgroups). If weighted 0.3 or less, C. longior again
recovered as the sole basal taxon—a result also consistent with amino acid translation of CO1 (in combination with remaining data). Using C. longior as a
functional outgroup resulted in functional ingroup
rooting by the seed-feeding clade.
3.5. Combined analyses: ingroup topology
Some relationships were consistent across the various
outgroup configurations, including the following. The
four species feeding on small seeds (the seed-feeding
clade) made up the best-supported larger clade, and a
fifth seed-feeding species, C. aciculatus, grouped with
these in the mtDNA and morphology analyses (Fig. 3).
The two petiole-feeding specialists C. petioli and C.
marginatus clustered as expected. The two pith-feeding
Dryocoetiops species clustered together with high support and clustered further with two, primarily petiolefeeding, species of Coccotrypes. However, a Kishino–
Hasegawa (parsimony) variance test (Kishino and Hasegawa, 1989) could not reject monophyly of Coccot-
178
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Fig. 2. Separate unweighted MP analyses of the 25 ingroup taxa and 4 outgroup taxa having no missing data partitions. Bootstrap support values are shown on nodes with more than 50% support.
Collecting areas are indicated by their initials for the nonmonophyletic Coccotrypes advena (see Table 1). Morphology: strict consensus of 3 trees of length 87 steps each, CI ¼ 0.57, RI ¼ 0.80.
mtDNA: 1 tree of length 3040 steps, CI ¼ 0.28, RI ¼ 0.37. EF-1a: strict consensus of 17 trees of length 628 steps each, CI ¼ 0.56, RI ¼ 0.62.
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Table 5
Partitioned bremer support indices for each of the four data partitions
Node
EF-1a
12S
CO1
Morphology
All
combined
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Total
34.5
2.5
11.2
0.5
15.5
0.5
6.6
7.5
8.2
15.2
0
7.5
7.8
15
)0.5
)0.5
0
6.5
13.5
)0.5
2.5
4.5
21.5
8.5
12.5
1.5
202
14.5
)2.5
1.3
1.5
1.5
1.5
0.4
0
3.8
13.5
0
)0.5
)1.5
)2.5
)1.5
)1.5
0
)2.5
)2.5
)1.5
)2
)1.8
1
)1.5
2.5
)1.5
18.2
7
9
8.7
)1
9
)1
2.6
2.2
4.1
48.3
0
)4
6.3
6.5
1
1
0
1.5
30
1
)0.5
)1.4
27
)3
)7
5
152.3
7
)1
3.8
)1
7
)1
3.4
)3.7
)2.1
)3
0
1
1.4
5
1
1
0
0.5
2
1
1
)0.3
3.5
)3
)4
)1
18.5
63
8
25
0
33
0
13
6
14
74
0
4
14
24
0
0
0
6
43
0
1
1
53
1
4
4
391
29
0.64
708
0.55
Inf. ch.
198
Tot/Inf. ch.
1.02
112
380
0.16
0.40
Note. Values are estimated from a topology based on combined
data for 25 ingroup taxa and 4 outgroups, with no missing partitions
(see Fig. 3).
rypes, to the exclusion of Dryocoetiops. C. gedeanus
grouped with C. cyperi and C. litoralis. The latter species
had haplotypes very similar to those of C. cyperi (EF-1a
coding region was identical, but intron differed by one
11-bp insertion) and was nested within the latter in
many analyses, including some analyses of partial gene
fragments, and in the ML analysis. On the contrary, and
perhaps the most remarkable result of the analyses, was
the consistent nonmonophyly and long branches of
C. advena, a widespread resource generalist. The latter
result holds for all gene partitions when analyzed separately, but not morphology.
The general absence of well-supported deep nodes
was not associated with incongruence among data partitions. Only when each of the mtDNA data partitions
was tested separately against EF-1a was incongruence
significant (P ¼ 0:03 [12S], 0.01 [CO1]). However, because the mtDNA gene partitions are linked in the mitochondrial genome, they must ultimately be combined,
and as such, the combined mtDNA partition was not
significantly different from any other partition (P ¼ 0:20
[EF-1a], 0.74 [morphology]).
179
3.6. Morphology and behavior
Table 6 shows the relative performance of morphological and behavioral characters on the topology depicted in Fig. 3. Among those characters with maximum
CI values, only characters 1, 16, 23, and 27 could potentially vary within the ingroup. The phylogenetic
utility of proventricular characters (Nobuchi, 1969) is
little known, but we note here that the shape of the
anterior plate and the placement of marginal bristles
define groupings also suggested by our current best estimate of the phylogeny (Fig. 3: node 7, char. 5; node 15,
char. 1). We also note that the characters with the worst
performance were all describing cuticular granules and
ridges (char. 12, 14, 15, and 17). Resource use (char. 30)
performed equally to the average for all characters
combined.
4. Discussion
4.1. Unusual evolutionary dynamics under sib-mating
Perhaps the most striking result from our study is
the unusually high sequence divergence and morphological stasis in many lineages. Variation in the genus
Coccotrypes in fact equals that reported for the entire
Prodoxidae (Brown et al., 1994) or Silphidae (Dobler
and M€
uller, 2000) for mtDNA or for the Noctuidae
(Mitchell et al., 1997) for EF-1a. Not only is the observed CO1 divergence among the highest reported for
a genus, but the 19% uncorrected divergence among
the five C. advena haplotypes (22.3% Kimura two-parameter (k2) corrected) is the highest intraspecific CO1
variation known to us for any insect or most organisms (but see Hedin, 2001; Rocha-Olivares et al.,
2001). Such high intra specific variation is not unique
to C. advena, however. Although lower variation was
found between the two C. fallax haplotypes (9.5%),
considerable variation was found between ecologically
and morphologically near-identical species, such as C.
petioli and C. marginatus (18.1%) and C. dactyliperda
and C. carpophagus (16.1%). Some other sib-mating
scolytine genera contain similar genetic structure as
revealed by unpublished CO1 haplotypes of Xylosandrus morigerus (Blandford) (16.0%) and Hypothenemus
eruditus Westwood (16.7%) (B.H.J. and L.R.K., unpublished).
One might argue that such high levels must reflect
morphologically cryptic species, in particular because
outbreeding scolytine beetle species—and even species
complexes—typically encompass less than 5% sequence
divergence at COI (Cognato and Sperling, 2000; Kelley
and Farrell, 1998). The consistent nonmonophyly of C.
advena further supports this view, suggesting that the
Ugandan populations constitute a separate species. On
180
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Fig. 3. Most parsimonious tree for 25 ingroup taxa and 4 outgroup species, with EF-1a and morphology weighted 2 and mtDNA weighted 1. Length
4540 steps, CI ¼ 0.34, RI ¼ 0.46. Numbers in gray refer to node numbers (see Table 5). Values above nodes are bootstrap support values and below
are Bremer support values for weighted/unweighted analyses. Bremer support of zero indicates an unresolved or a slightly conflicting node. Arrows
point to the placement of taxa for which only mtDNA sequences were available, if included in the equally weighted analysis. Vertical bars show
resource preferences.
the other hand, none of the markers used were congruent with respect to the topological placement of the
different C. advena specimens, pointing toward possible
events of introgression and incomplete lineage sorting.
Only further sampling of other populations and additional nuclear markers can resolve this ambiguity.
However, there are reasons to expect sib-mating lineages
to show more incongruence and higher levels of intraspecific DNA sequence divergence than outbreeders.
Sexual selection is thought to be one of the major
components driving species diversification in a variety of
outbreeding organisms (Schluter, 2000), and the loss of
sexual selection in sib-mating scolytines may reduce divergence rates between species. This also involves the
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
181
Fig. 4. Most likely tree (score )17489.89) found in the ML analysis of all molecular data partitions combined for 29 taxa (and no missing data
partitions). Parameter settings used with the GTR + C + I model were empirical base frequencies, I ¼ 0.496, and C ¼ 0:487 with four rate categories
and substitution frequencies A $ G 11.46, A $ C 4.85, A $ T 8.92, T $ G 1.00, C $ T 66.90, and C $ G 1.93. Parameters were approached
through iterations of parameter estimates, starting with the topology depicted in Fig. 3. To the right are drawings from the internal side of one of the
eight proventricular plates from the following species, top to bottom: Dryocoetes affaber, Ozopemon uniseriatus, Coccotrypes cf. distinctus, Dryocoetiops coffeae, and Coccotrypes gedeanus. Vertical bars indicate similarity in proventricular characters 1 and 5 (see Table 6, Appendix A).
loss of specific mate recognition characters which define
most outbreeding species. Indeed, the characters most
commonly used to distinguish outbreeding bark beetle
species are secondary sexual characters (Jordal, 1998)
which are presumably important for exchange of tactile
cues during mate selection (Page and Willis, 1982). The
slowdown in morphological evolution in species of
Coccotrypes and other sib-mating beetles is therefore
largely expected. Still, morphological differences could
develop between shallowly diverged sib-mated lineages
because, under zero outcrossing, the fixation of mutations in any lineage is virtually instantaneous. This may
182
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Table 6
The relative fit of morphological and behavioral characters as measured by the consensus index (CI) over tree topology in Fig. 3
Proventriculus
1
1.00
2
0.5
Head
3
0.33
4
0.33
5
0.75
0.58
7
1.0
8
1.0
9
1.0
10
0.67
21
1.0
22
0.5
23
1.0
24
1.0
11
0.5
12
0.17
13
0.4
14
0.25
25
0.6
26
0.5
27
1.0
28
0.5
0.67
Pronotum
15
0.29
6
1.0
Elytra
16
1.0
17
0.25
0.61
18
1.0
19
0.5
20
0.4
Legs
0.73
0.72
Behaviour and development
29
1.0
Average morphology: 0.63
0.89
30
0.67
31
1.0
Average all: 0.66
Note. Indices vary marginally between alternative tree topologies. Numbers in boldface are average CI for each body section and for behavior.
See Appendix A for description of characters 1–31.
explain the distinct morphological differences between
the genetically close C. cyperi and C. litoralis. Another
well-studied ecological mate recognition character is
pheromone production, where outbreeding species find
conspecific mates by sorting diverse pheromone plumes
during dispersal. Pheromones are not known from sibmating species, as expected, because females mate with
their brother before dispersal. Consequently, the huge
genetic divergences that we observe in some sib-mating
lineages rarely represent ecologically segregated cryptic
species.
While close relatives in Coccotrypes sometimes demonstrate distinct ecological segregation (see below), the
overwhelming impression is one of ecological niche
overlap in sympatry. It is not unusual to find closely
related species breeding in the same unit, for instance the
petiole specialists C. petioli and C. marginatus or the
seed specialists C. dactyliperda and C. carpophagus
(B.H.J. and L.R.K., pers. obs.). We also expect to find
the same pattern in species of Dryocoetiops where all use
pith of tiny twigs or of petioles and in species of Ozopemon where all breed under bark of large logs. The
observation of a marked increase in niche overlap for
many sib-mating sister species contrasts with the sometimes clear evidence for allopatric distribution or
sympatric resource division between closely related
species in general (Barraclough and Vogler, 2000) and
for outbreeding bark beetles in particular (Cognato and
Sperling, 2000; Jordal, 1998; Kelley and Farrell, 1998).
If the overall outcrossing rate for a population is low,
and if hybridization is physically possible, the relaxed
selection on mate recognition characters may also permit more frequent matings between these ecologically
similar and sympatric ‘‘species.’’ Judged by field observations (B.H.J., L.R.K.), broods from multiple colonizations sometimes coalesce, providing opportunities for
outcrossing. The genetic outcome of this process could
sometimes mimic the principles of artificial breeding
programs in agriculture, in which a cross between inbred
‘‘pure strains’’ creates hybrids, and then the new hybrid
strain is inbred until it ‘‘breeds true,’’ i.e., until it is
homozygous. A long history of such low-frequency reticulation events may be the cause of the extreme
haplotype diversity and incongruence between morphological characters and gene trees for some Coccotrypes lineages. We also note that hybridization in
nature of genetically highly diverged individuals seems
only possible for subclonally reproducing lineages. For
instance hybridizing Daphnia species in the galeata
complex are as much as 19% diverged (K2 corrected) in
their CO1 sequences (Schwenk et al., 2000), consistent
with our view that hybridization may also be possible
between the different lineages of C. advena and related
species.
We are currently working on assessing more closely
the intrapopulation genetic variation within and between sites, but it is clear that variation is considerable
for most sib-mating lineages studied (B.H.J. and
L.R.K., unpublished). Whatever is causing the extreme
haplotype variation in Coccotrypes, is also resulting in
certain drawbacks of phylogeny reconstruction as discussed below.
4.2. Gene utility and phylogeny reconstruction
The most rapidly evolving mitochondrial genes
showed high levels of substitutional saturation and
homoplasy between the most divergent lineages. Although high substitution rates rarely contribute mere
noise in a data matrix (Wenzel and Siddall, 1999; Yang,
1998) and often provide the majority of phylogenetic
signal in terms of node support (Allard et al., 1999;
Baker et al., 2001; Bj€
orklund, 1999; K€allersj€
o et al.,
1999), the negative effect of noise from mtDNA was in
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
this study demonstrated by the reduced number of
bootstrap-supported ð> 50%Þ nodes (16 vs l2) when
mtDNA nucleotides were added to the combined data
matrix. By comparison, using amino-acid-translated
CO1 data, or downweighting mtDNA by a factor of
two or more, we increased the number of supported
ingroup nodes to 15 or more. Taken together with the
many bizarre relationships found in the mtDNA phylogeny, the lack of support and negative contribution to
the total-evidence tree suggest that the mitochondrial
genes are far too rapidly evolving to be useful for
phylogeny reconstruction above species-group level in
Coccotrypes.
A commonly used method to accommodate poor
resolution is downweighting rapidly evolving sites—
third positions or transitions—to elevate signal from
more conservative sites (Swofford et al., 1996). However, combined analysis of first and second positions
yielded only slightly better results than third positions
alone for our CO1 data, and transversions performed
no better than transitions for all mtDNA data. Despite
the appeal of this type of differential weighting, studies
showing that such character weighting does not increase
node support or decrease incongruence (Baker and
DeSalle, 1997; Baker et al., 2001; Broughton et al.,
2000; Durando et al., 2000; Fu, 2000; Garin et al.,
1999), and that it performs even worse when second
positions are especially favoured (Xia, 1998), are now
accumulating.
The sometimes limited utility of mitochondrial genes
is in contrast to the usefulness of certain rapidly
evolving nuclear genes (e.g., Campbell et al., 2000).
Nuclear genes usually outperform mtDNA genes in
combined analyses (Baker and DeSalle, 1997; Baker
et al., 2001; Durando et al., 2000; Remsen and DeSalle,
1998), especially at deeper splits (Springer et al., 1999),
and the value of nuclear protein-coding genes to the
resolution of deeper nodes of insect genera cannot be
overstated (Durando et al., 2000). Also in this study,
the nuclear gene contributed most of the phylogenetic
signal within a combined data matrix which otherwise
consisted of highly diverged mitochondrial sequences.
Thus, higher weights could be given to the best-performing gene to also recover the deepest splits (see
Reed and Sperling, 1999). With higher weights applied
to EF-1a and morphology, topologies converged toward the one in Fig. 3 or the one with C. longior as a
single basal taxon when mtDNA data were excluded
altogether.
As noted previously (Jordal et al., 2000, 2002),
Dryocoetiops appear to be nested within Coccotrypes.
Morphological differences between the two genera are
small and not consistent, in particular the separation
of the procoxae and shape of the pronotum (Wood,
1986; but see Appendices A and B). Although more
nuclear markers and species of Dryocoetiops must be
183
sampled before reaching a final conclusion, synonymy
of this genus seems likely when such data become
available.
The two preferred topologies (Figs. 3 and 4) are also
similar to our previous study where C. longior, C.
marginatus, and the seed-feeding clade were among the
most basal taxa (Jordal et al., 2000). However, we could
not confidently place the basal node of the ingroup due
to a presumably strong conflict among the outgroups.
Because there is no consistent rule on how to root
phylogenetic trees, and because general recommendations differ among authors (Lyons-Weiler et al., 1998;
Smith, 1994; Watrous and Wheeler, 1981), it is unclear
which rooting should be preferred. Hence we prefer to
err on the side of inclusion by using both Dryocoetes
and Ozopemon as outgroups. The two ingroup roots
found in the MP and ML analyses using both outgroup
genera (Figs. 3 and 4) are not very different, and both
alternatives place the seed-feeding clade close to the
base in each topology. Furthermore, the topology depicted in Fig. 3 is identical to the Coccotrypes topology
reconstructed as a part of an analysis of 80 dryocoetine
and xyleborine EF-1a sequences (B.H. Jordal, unpublished), demonstrating topological stability with increased sampling. On the other hand, we note in the ML
analysis that the resulting basal position of the seedfeeding clade is similar to the result obtained with Ozopemon as the sole outgroup in the MP analysis. Despite
these ambiguities, the combined data still provided evidence for certain evolutionary patterns in resource use
that we discuss below.
4.3. Implications for resource-use evolution
The expansion of niches and extreme resource generalism in sib-mated lineages is a paradox, given the
presumed loss of genetic variation due to inbreeding.
However, we have demonstrated huge amounts of
haplotype diversity in populations, hinting at a broad
array of genotypic variation conducive to ecological
adaptability and evolutionary change. The resolution
of the genetic variation paradox may lie in the ability
of sib-mating Coccotrypes to regain ‘‘lost’’ genetic
variation through occasional outcrossing/hybridization
events. During the long periods of inbreeding that
separate such events, any beneficial mutations enabling
use of a particular resource will be immediately fixed in
the inbred lineages on that resource. If that resource is
locally abundant such lineages may proliferate, and
when outcrossing/hybridization events do occur, such
alleles may be captured by other inbred lineages in the
same or different species. Selection for ecological
adaptability in opportunistic generalists such as Coccotrypes may act to promote distant outcrossing and to
eliminate any vestiges of a specific mate recognition
system.
184
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
The clear preference for small resources set some sibmating groups apart from most outbreeding bark and
timber beetles (Kirkendall, 1993). Smaller resources
such as seeds and petioles are scattered in the forest and
even more ephemeral than the logs and branches used
by other scolytines (Beaver, 1979b). Because larger,
longer-lived resources attract more beetle individuals
over a longer time period than do smaller resource units,
the probability of finding a mate for an outbreeding
individual decreases with decreasing resource size. Small
resources such as petioles (Beaver, 1979b; Jordal and
Kirkendall, 1998) and palm seeds (Kirkendall, 1993)
have usually one or a few conspecific colonizing females
(inbreeders) or colonizing pairs (outbreeders) per resource unit. Because colonizing females of inbreeders
carry with them sperm from their natal nest, these breed
independently of colonization densities. This is the most
obvious reason that outbreeders are generally less frequent in small resources. Another important factor is
the near twofold advantage of sib-maters’ female-biased
offspring sex ratios, allowing demographic viability in
less productive resource units (Beaver, 1979a; Jordal and
Kirkendall, 1998).
Some of the highly unusual resources used by
Coccotrypes are seemingly evolutionary novelties with
little or no subsequent diversification. Perhaps the most
unusual such novelty is bracken fern (Gray, 1970,
1972), with three associated species of Coccotrypes. We
were not able to obtain DNA from the single monophagous species in ferns, C. pteridophytae (Schedl), but
judged by morphological characters, this species seems
closely related to C. cardamomi, a generalist species
with preference for seeds (Browne, 1961) and occasionally found in bracken fern (Gray, 1972). A third
species found in ferns (Asplenium), C. confusus (Eggers), is most probably the closest relative of two other
generalists, C. medius and C. cylindricus (Eggers) (see
Browne, 1973). Based on the widely separated phylogenetic placement of C. medius and C. cardamomi, it
seems likely then that fern feeding evolved at least
twice in Coccotrypes.
The seeds and radicles of viviparous mangroves also
have a limited number of associated beetles (Beeson,
1961; Browne, 1961; Woodruff, 1970) and three species
of Coccotrypes breed exclusively in such tissues. In at
least two of these (C. litoralis and C. rhizophorae
[Hopkins]), the mangrove association has evolved from
the use of fruits and seeds as part of their more generalized diet. With the lack of EF-1a data, we were not
able to assess the placement of C. fallax in this study.
However, judged by proventricular and other morphological characters in conjunction with mitochondrial
data, C. fallax is not associated with the ‘‘generalist’’
clade (node 15 in Fig. 3). Hence, two or three independent origins of mangrove seed feeding occurred. Because
there is no example of speciation within each mangrove-
feeding lineage, and because close relatives of C. litoralis
and C. fallax have diverged little in genome and morphology, the association with mangroves appear to be
recent in each case.
Seeds of various size and form are clearly the most
productive resources in terms of brood size (Kirkendall,
1993) and the number of species capable of breeding in
this type of resource (Browne, 1961). Despite its productivity, seed feeding has only evolved (in Scolytinae)
in two inbreeding clades (Coccotrypes and Hypothenemus) and in a few outbreeding species. Nonetheless,
more than half of the species in Coccotrypes show slight
or strong preferences for seeds, although only a minor
fraction are specialists on this resource. These specialists
feed on small seeds (primarily palm seeds) and have
followed the recent expansion of palms throughout the
tropics. Although we have sampled only a minor fraction of putative seed specialists, a majority of the species
seems to constitute a monophyletic group (cf. Figs. 3
and 4) and as such demonstrates evolutionary consistency of this habitat.
Given the high-energy endosperm provided by large
seeds (e.g., dipterocarp seeds) and the frequent but
facultative use of larger seeds by species of Coccotrypes,
why is there apparently so little specialization upon this
resource? The most important sources of large seeds are
those of dipterocarp trees growing in southeast Asia, the
area where Coccotrypes originated during the Miocene
(Jordal et al., 2000). Since dipterocarps exhibit multispecies gregarious flowering and mast fruiting (Ashton
et al., 1988), dipterocarp seeds are not available every
year, and it is not possible for multivoltine insects such
as Coccotrypes to specialize on them. They can be utilized only opportunistically, by resource generalists. To
specialize, a species is constrained to use predictably
available resources—for instance, in the case of Coccotrypes, palm seeds, mangrove radicles, Moraceae petioles, or bracken ferns. When dipterocarp seeds are
artificially made permanently available, as in ilipe nut
(oil-seed) plantations, many different Coccotrypes species converge to destroy much of the crop (Browne,
1973), suggesting preference for this type of resource
when available.
Regardless of the ambiguity associated with our
phylogenetic hypothesis, all reconstructions indicate an
evolutionary trend from specialization upon one kind of
host tissue toward a more opportunistic behavior using
many different tissue types (Figs. 3 and 4). And other
generalist species, not sampled for DNA—C. cinnamomi
(Eggers), C. nitidus (Eggers), and C. papuanus (Eggers)
(Browne, 1961)—have morphological features indicating
a clear relationship to the generalist-dominated sub
clade (node 15). Our study thus adds to the increasing
bulk of evidence that specialization does not necessarily
evolve from more generalized lifestyles (Thompson,
1994).
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Acknowledgments
We especially thank Roger A. Beaver for identification of specimens and James Lyons-Weiler for help
with RASA analyses and permission to use a prerelease of RASA 3.0 Turbo. We also greatly appreciate
the additional specimens collected by Akira Ueda,
Hideaki Goto, Andrea Sequeira, Geoffrey Morse, and
Swee Peck Quek. B.H.J. was funded by a Norwegian
185
Research Council Grant (123588/410), BDF by an
USDA National Research Initiative grant (97-353024226), and LRK by a Norwegian Research Council
SUP grant (128388/420). Field work was supported by
the Putnam Expedition Fund of the Museum of
Comparative Zoology, and we thank members of the
parataxonomist training centre in Madang, Papua
New Guinea for helpful assistance with collecting and
logistics.
Appendix A. Morphological and behavioral characters coded for Coccotrypes, Dryocoetiops, and outgroups
Proventriculus
1.
Anterior margin of anterior plate: (0) straight or slightly curved; (1) angular.
2.
Anterior plate with tubercles: (0) absent; (1) present.
3.
Median suture: (0) oval; (1) linear or diverging.
4.
Median denticles: (0) pointed, sharp; (1) small, rounded; (2) absent.
5.
Marginal bristles: (0) tuft of few; (1) longitudinal rows of > 10; (2) longitudinal rows of widely spaced,
soft bristles; (3) absent.
Head
6.
7.
8.
9.
10.
11.
12.
13.
14.
Labial palpus segment 1: (0) shorter than 2 + 3 combined; (1) longer than 2 + 3 combined, barrel shaped.
Labial palpus, segment 3: (0) longer than segment 2, narrow; (1) subequal to segment 2, broad.
Maxillary fringe with setae: (0) few and coarse; (1) many, medium coarse; (2) numerous fine and soft.
Maxillary palpus segment 3: (0) shorter than 1 + 2 combined; (1) longer than 1 + 2 combined.
Antennal club: (0) more than basal 1/3 corneous; (1) less than basal 1/4 corneous; (2) hairy to base.
Setose area of anterior face of club with transverse suture: (0) visible (1) not visible.
Epistoma with granules: (0) present; (1) absent.
Aciculation in frons: (0) absent; (1) narrowly, from below level of eyes; (2) broadly, from vertex.
Median, longitudinal carina in frons: (0) absent; (1) present.
Pronotum
15.
16.
17.
18.
19.
Pronotal asperities on anterior half: (0) absent; (1) low or granulate; (2) sharply elevated.
Row of asperities along apical margin: (0) absent; (1) present.
Asperities on basal disc: (0) present; (1) laterally only; (2) absent.
Pronotal disc: (0) shining; (1) reticulate.
Dorsum: (0) flat; (1) lightly curved; (2) dome-shaped in middle; (3) dome-shaped posteriorly.
Elytra
20.
21.
22.
Interstrial setae: (0) thin; (1) stout; (2) spatulate.
Strial punctures (0) in rows; (1) confused at least on declivity.
Strial setae: (0) absent; (1) tiny, recumbent; (2) erect, at least half length of interstrial setae.
Legs
24.
25.
26.
27.
28.
Probibiae with: (0) 4 socketed teeth, (1) 5 teeth; (2) 7 teeth.
Mesotibiae with: (0) 4 socketed teeth; (1) 5 teeth; (2) 6 teeth; (3) 7 teeth; (4) 8 teeth.
Metatibiae with: (0) 4 socketed teeth; (1) 5 teeth; (2) 6 teeth; (3) 7 teeth; (4) 9 teeth.
Procoxae: (0) contiguous, prosternal process pointed; (1) separated, prosternal process broad/blunt.
Mesocoxae separated: (0) approximately width of scapus; (1) twice or more the width of scapus.
Males
29.
Male development: (0) similar to females: (1) smaller, with adult characters (2) larviform.
Behavior
30.
31.
Feeding on: (0) phloem; (1) pith; (2) leafstalks; (3) small, hard seeds; (4) fruits; (5) mangrove radicles.
Mating system: (0) outbreeding; (1) inbreeding.
186
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Appendix B. Data matrix of the 31 morphological characters described in Appendix A
Dryocoetes affaber
Dr. autographus
Ozopemon brownei
O. uniseriatus
Coccotrypes aciculatus
C. advena (B)
C. advena (CR)
C. advena (J)
C. advena (U)
C. advena (U)
C. cardamomi
C. carpophagus
C. cf. carpophagus
C. cf. rhizophorae
C. cyperi (CR)
C. cyperi (Qld)
C. dactyliperda (Arg.)
C. dactyliperda (FL)
C. cf. distinctus
C. falax (Bangl)
C. fallax (Qld)
C. gedeanus
C. graniceps
C. impressus
C. litroralis
C. longior
C. marginatus
C. medius
C. petioli
C. variabilis
Dryocoetiops coffeae
D. cf. eugeniae
0 0
0 0
0
0 0
0 0
0
0 0
0 0
3
0 1
0 0
3
0 0
1 1
2
1 0
1 1
1
1 0
1 1
1
1 0
1 1
1
1 0
1 1
1
1 0
1 1
1
1 0
1 1
1
0 0
1 1
2
1 0
1 1
0
0 0
0 0
0
1 0
1 1
0
1 0
1 1
0
0 0
1 1
2
0 0
1 1
2
0 0
1 1
2
0 1
0 0
0
0 1
0 0
0
1 {01} 1 {01} 0
1 0
1 1
0
0 0
1 1
2
1 0
1 1
0
0 0
0 0
0
0 0
1 1
3
0 0
1 0
0
0 0
1 1
3
1 0
1 1
1
0 1
1 0
0
0 1
1 0
0
00000
00000
10111
10111
11210
11211
11211
11211
11211
11211
11211
11210
11211
11211
11211
11211
11210
11210
11210
11211
11211
11211
11211
11210
11211
11211
11212
11211
11212
11211
11211
11211
References
Allard, M.W., Farris, J.S., Carpenter, J.M., 1999. Congruence among
mammalian mitochondrial genes. Cladistics 15, 75–84.
Ashton, P.S., Givnish, T.J., Appanah, S., 1988. Staggered flowering in
the Dipterocarpaceae: New insights into floral induction and the
evolution of mast fruiting in the aseasonal tropics. Am. Nat. 132,
44–66.
Baker, R.H., DeSalle, R., 1997. Multiple sources of character
information and the phylogeny of Hawaiian drosophilids. Syst.
Biol. 46, 645–673.
Baker, R.H., Wilkinson, G.S., DeSalle, R., 2001. Phylogenetic utility
of different types of molecular data used to infer evolutionary
relationships among stalk-eyed flies (Diopsidae). Syst. Biol. 50, 87–
105.
Baker, R.H., Yu, X., DeSalle, R., 1998. Assessing the relative
contribution of molecular and morphological characters in simultaneous analysis trees. Mol. Phylogent. Evol. 9, 427–436.
Barraclough, T.G., Vogler, A.P., 2000. Detecting the geographical
pattern of speciation from species-level phylogenies. Am. Nat. 155,
419–434.
Beaver, R.A., 1979a. Leafstalks as a habitat for bark beetles (Col.:
Scolytidae). Z. Angew. Entomol. 88, 296–306.
100
100
010
010
112
110
111
111
111
111
101
112
111
112
111
111
112
112
112
100
100
101
101
112
111
010
010
111
010
101
100
100
0
0
1
1
0
0
0
0
0
0
{01}
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
1
2
1
2
2
2
0
0
0
0
0
1
2
1
1
2
2
2
2
2
2
2
2
1
2
2
2
0
1
0
1
2
2
0100
0100
0003
0003
1002
0200
0201
0201
0201
0201
0100
1002
0101
0100
0101
0101
1002
1002
1002
0011
0011
0101
0001
1002
0101
0110
0000
0000
0000
0000
0002
0002
0
0
0
0
1
2
2
2
2
2
0
0
1
0
1
1
0
0
0
0
0
0
0
0
0
{12}
0
0
0
2
0
0
01001
01024
12001
12002
02000
00001
00001
00001
00001
00001
00001
02000
01001
01001
00001
00001
02000
02000
02000
00001
00001
00001
01001
02001
00001
00001
02101
01011
02101
00001
01002
01002
2010
5010
1002
2002
0011
1011
1011
1011
1011
1011
1011
0011
101?
111?
1011
1011
0011
0011
0011
2111
2111
1011
1001
101?
1011
1011
1011
111?
1011
1011
211?
211?
0
0
0
0
3
{024}
{024}
{024}
{024}
{024}
{34}
3
3
2
{01234}
{01234}
3
3
3
5
5
{0124}
3
3
5
{02}
2
{012}
2
{034}
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Beaver, R.A., 1979b. Non-equilibrium ‘island’ communities: A guild of
tropical bark beetles. J. Anim. Ecol. 48, 987–1002.
Beeson, C.F.C., 1961. The Ecology and Control of the Forest Insects
of India and the Neighbouring Countries, 2nd ed. Government of
India, New Delhi.
Bj€
orklund, M., 1999. Are third positions really that bad? A test using
vertebrate Cytochrome b. Cladistics 15, 191–197.
Blumberg, D., Kehat, M., 1982. Biological studies of the date stone
beetle, Coccotrypes dactyliperda. Phytoparasitica 10, 73–78.
Bright, D.E., Peck, S.B., 1998. Scolytidae from the Galapagos Islands,
Equador, with descriptions of four new species, new distribution
records, and a key to species. Koleopt. Rund. 68, 233–252.
Bright, D.E., Skidmore, R.E., 1997. A Catalog of Scolytidae and
Platypodidae (Coleoptera), Supplement 1 (1990–1994). NRC
Research Press, Ottawa.
Broughton, R.E., Stanley, S.E., Durrett, R.T., 2000. Quantification of
homoplasy for nucleotide transitions and transversions and a
reexamination of assumptions in weighted phylogenetic analysis.
Syst. Biol. 49, 617–627.
Brown, J.M., Pellmyr, O., Thompson, J.M., Harrison, R.G., 1994.
Mitochondrial DNA phylogeny of the Prodoxidae (Lepidoptera:
Incurvarioidea) indicates rapid ecological diversification of yucca
moths. Ann. Entomol. Soc. Am. 87, 795–802.
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Browne, F.G., 1959. Notes on two Malayan scolytid bark-beetles.
Malay. For. 22, 292–300.
Browne, F.G., 1961. The biology of Malayan Scolytidae and Platypodidae. Malay. For. Rec. 22, 1–255.
Browne, F.G., 1973. The African species of Poecilips Schaufuss. Rev.
Zool. Bot. Afr. 87, 679–696.
Campbell, D.L., Brower, A.V.Z., Pierce, N.E., 2000. Molecular
evolution of the wingless gene and its implications for the
phylogenetic placement of the butterfly family Riodinidae (Lepidoptera: Papilionoidea). Mol. Biol. Evol. 17, 684–696.
Cognato, A.I., Sperling, F.A.H., 2000. Phylogeny of Ips species
(Coleoptera: Scolytidae) inferred from mitochondrial cytochrome
oxidase I sequence. Mol. Phylogenet. Evol. 14, 445–460.
Danforth, B.N., Ji, S., 1998. Elongation Factor-1a occurs as two
copies in bees: Implications for phylogenetic analysis of EF-1a
sequences in insects. Mol. Biol. Evol. 15, 225–235.
Dobler, S., M€
uller, J.K., 2000. Resolving phylogeny at the family level
by mitochondrial Cytochrome Oxidase sequences: Phylogeny of
carrion beetles (Coleoptera, Silphidae). Mol. Phylogenet. Evol. 15,
390–402.
Durando, C.M., Baker, R.H., Etges, W.J., Heed, W.B., Wasserman,
M., DeSalle, R., 2000. Phylogenetic analysis of the repleta group of
the genus Drosophila using multiple sources of characters. Mol.
Phylogenet. Evol. 16, 296–307.
Entwistle, P.F., 1964. Inbreeding and arrhenotoky in the ambrosia
beetle Xyleborus compactus (Eichh.) (Coleoptera: Scolytidae). Proc.
R. Soc. Lond. A 39, 83–88.
Farrell, B.D., Sequeira, A., O’Meara, B., Normark, B.B., Chung, J.,
Jordal, B., 2001. The evolution of agriculture in beetles
(Curclionidae: Scolytinae and Platypodinae). Evolution 10,
2011–2027.
Farris, J.S., K€
allersj€
o, M., Kluge, A.G., Bult, C., 1995a. Constructing
a significance test for incongruence. Syst. Biol. 44, 570–572.
allersj€
o, M., Kluge, A.G., Bult, C., 1995b. Testing
Farris, J.S., K€
significance of incongruence. Cladistics 10, 315–319.
Felsenstein, J., 1985. Confidence limits on phylogenies: An approach
using the bootstrap. Evolution 39, 783–791.
Fu, J., 2000. Towards the phylogeny of the family Lacertidae—Why
4708 base pairs of mtDNA sequences cannot draw the picture.
Biol. J. Linn. Soc. 71, 203–217.
Garin, C.F., Juan, C., Petitpierre, E., 1999. Mitochondrial DNA
phylogeny and the evolution of host-plant use in Palearctic
Chrysolina (Coleoptera, Chrysomelidae) leaf beetles. J. Mol. Evol.
48, 435–444.
Gray, B., 1970. Observations on the first scolytid, Poecilips pteridophytae (Coleoptera: Scolytidae), to be found in association with a
fern (Pteridophyta). Can. Entomol. 102, 578–585.
Gray, B., 1972. Observations on Poecilips cardamomi (Schaufuss), the
second species of Scolytidae to be found in bracken fern (Col.).
Entomol. Tidskr. 93, 229–237.
Hamilton, W.D., 1993. Inbreeding in Egypt and in this book: A
childish perspective. In: Thornhill, N.W. (Ed.), The Natural
History of Inbreeding and Outbreeding: Theoretical and Empirical Perspectives. University of Chicago Press, Chicago, pp. 429–
450.
Hedin, M.C., 2001. Molecular insights into species phylogeny,
biogeography, and morphological stasis in the ancient spider genus
Hypochilus (Araneae: Hypochilidae). Mol. Phylogenet. Evol. 18,
238–251.
Herfs, A., 1950. Studien an dem steinnussborkenk€afer, Coccotrypes
tanganus Eggers. Hofch. Wiss. Praxis. 3, 3–31.
Jordal, B.H., 1998. A review of Scolytodes Ferrari (Coleoptera:
Scolytidae) associated with Cecropia (Cecropiaceae) in the northern Neotropics. J. Nat. Hist. 32, 31–84.
Jordal, B.H., 2001. The origin and radiation of sib-mating haplodiploid beetles (Coleoptera, Curculionidae, Scolytinae). Ph.D. Thesis,
Department of Zoology, University of Bergen, Bergen.
187
Jordal, B.H., Beaver, R.A., Kirkendall, L.R., 2001. Breaking taboos in
the tropics: Inbreeding promotes colonization by wood-boring
beetles. Glob. Ecol. Biogeogr. 10, 345–358.
Jordal, B.H., Beaver, R.A., Normark, B.B., Farrell, B.D., 2002.
Extraordinary sex ratios and the evolution of male neoteny in sibmating Ozopemon beetles. Biol. J. Linn. Soc. 75, 353–360.
Jordal, B.H., Kirkendall, L.R., 1998. Ecological relationships of a
guild of tropical beetles breeding in Cecropia leafstalks in Costa
Rica. J. Trop. Ecol. 14, 153–176.
Jordal, B.H., Normark, B.B., Farrell, B.D., 2000. Evolutionary
radiation of an inbreeding haplodiploid beetle lineage (Curculionidae, Scolytinae). Biol. J. Linn. Soc. 71, 483–499.
K€allersj€
o, M., Albert, V.A., Farris, J.S., 1999. Homoplasy increases
phylogenetic structure. Cladistics 15, 91–93.
Kelley, S.T., Farrell, B.D., 1998. Is specialization a dead end? The
phylogeny of host use in Dendroctonus bark beetles (Scolytidae).
Evolution 52, 1731–1743.
Kirkendall, L.R., 1993. Ecology and evolution of biased sex ratios in
bark and ambrosia beetles. In: Wrensch, D.L., Ebbert, M.A.
(Eds.), Evolution and Diversity of Sex Ratio: Insects and Mites.
Chapman & Hall, New York, pp. 235–345.
Kishino, H., Hasegawa, M., 1989. Evaluation of the maximum
likelihood estimate of the evolutionary tree topologies from
DNA sequence data, and the branching order in Hominoidea. J.
Mol. Evol. 29, 170–179.
Kluge, A.G., 1989. A concern for evidence and a phylogenetic
hypothesis of relationships among Epicrates (Boidae, Serpentes).
Syst. Zool. 38, 7–25.
Kluge, A., 1998. Total evidence or taxonomic congruence: Cladistics
or consensus classification. Cladistics 14, 151–158.
Lekander, B., 1968. Scandinavian bark beetle larvae; descriptions and
classification. Rep. No. 4, Department of Forest Zoology, Royal
College of Forestry, Stockholm.
Lyons-Weiler, J., 1998. RASA for the Mac. Available from http://
loco.biology.unr.edu/archives/rasa/rasa.html.
Lyons-Weiler, J., Hoelzer, G.A., 1997. Escaping from the Felsenstein
zone by detecting long branches in phylogenetic data. Mol.
Phylogenet. Evol. 8, 375–384.
Lyons-Weiler, J., Hoelzer, G.A., Tausch, R.J., 1996. Relative Apparent Synapomorphy Analysis (RASA) I: The statistical measurement of phylogenetic signal. Mol. Biol. Evol. 13, 749–757.
Lyons-Weiler, J., Hoelzer, G.A., Tausch, R.J., 1998. Optimal outgroup
analysis. Biol. J. Linn. Soc. 64, 493–511.
Mitchell, A., Cho, S., Regier, J.C., Mitter, C., Poole, R.W., Matthews,
M., 1997. Phylogenetic utility of Elongation Factor-1a in Noctuidea (Insecta: Lepidoptera): The limits of synonymous substitution.
Mol. Biol. Evol. 14, 381–390.
Nobuchi, A., 1969. A comparative morphological study of the
proventriculus in the adult of the subfamily Scolytoidea (Coleoptera). Bull. Government For. Exp. Sta. 224, 39–110.
Normark, B.B., Jordal, B.H., Farrell, B.D., 1999. Origin of a
haplodiploid beetle lineage. Proc. R. Soc. Lond. B 266, 2253–
2259.
Page, R.E., Willis, M.A., 1982. Sexual dimorphism in ventral
abdominal setae in Scolytus multistriatus (Coleoptera: Scolytidae):
Possible role in courtship behaviour. Ann. Entomol. Soc. Am. 76,
78–82.
Pruess, K.P., Adams, B.J., Parsons, T.J., Zhu, X., Powers, T.O., 2000.
Utility of the mitochondrial Cytochrome Oxidase II for resolving
relationships among black flies (Diptera: Simuliidae). Mol. Phylogenet. Evol. 16, 286–295.
Reed, R.D., Sperling, F.A.H., 1999. Interaction of process partitions
in phylogenetic analysis: An example from the Swallowtail butterfly genus Papilio. Mol. Biol. Evol. 16, 286–297.
Remsen, J., DeSalle, R., 1998. Character congruence of multiple data
partitions and the origin of the Hawaiian Drosophilidae. Mol.
Phylogenet. Evol. 9, 225–235.
188
B.H. Jordal et al. / Molecular Phylogenetics and Evolution 23 (2002) 171–188
Rocha-Olivares, A., Fleeger, J.W., Foltz, D.W., 2001. Decoupling of
molecular and morphological evolution in deep lineages of a
meiobenthic harpacticoid copepod. Mol. Biol. Evol. 16, 1088–1102.
Schluter, D., 2000. The Ecology of Adaptive Radiation. Oxford
University Press, Oxford.
Schwenk, K., Posada, D., Hebert, P.D.N., 2000. Molecular systematics
of European Hyalodapnia: The role of contemporary hybridization
in ancient species. Proc. R. Soc. Lond. B 267, 1833–1842.
Simmons, M.P., 2000. A fundamental problem with amino-acidsequence characters for phylogenetic analyses. Cladistics 16, 274–282.
Smith, A.B., 1994. Rooting molecular trees: Problems and strategies.
Biol. J. Linn. Soc. 51, 279–292.
Sorenson, M.D., 1999. TreeRot.v2. Boston University, Boston.
Springer, M.S., Amrine, H.M., Burk, A., Stanhope, M.J., 1999.
Additional support for Afroteria and Paenungulata, the performance of mitochondrial versus nuclear genes, and the impact of
data partitions with heterogeneous base composition. Syst. Biol.
48, 65–75.
Swofford, D., 1999. PAUP*. Phylogenetic Analysis Using Parsimony
(*and other methods) version 4. Sinauer, Sunderland, MA.
Swofford, D.L., Olsen, G.J., Waddell, P.J., Hillis, D.M., 1996.
Phylogenetic inference. In: Hillis, D.M., Moritz, C., Mable, B.K.
(Eds.), Molecular Systematics. Sinauer, Sunderland, MA.
Takenouchi, Y., Takagi, K., 1967. A chromosome study of two
parthenogenetic scolytid beetles. Ann. Zool. Jpn. 40, 105–110.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins,
D.G., 1997. The ClustalX windows interface: Flexible strategies for
multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res. 24, 4876–4882.
Thompson, J.N., 1994. The Coevolutionary Process. Chicago University Press, Chicago.
Ueda, A., 1997. Brood development of an inbreeding spermatophagous scolytid beetle Coccotrypes graniceps (Eichhoff) (Coleoptera:
Scolytidae). Jpn. J. Entomol. 65, 677–687.
Watrous, L.E., Wheeler, Q.D., 1981. The outgroup comparison
method of character analyses. Syst. Zool. 30, 1–11.
Wenzel, J.W., Siddall, M.E., 1999. Noise. Cladistics 15, 51–64.
Wheeler, W.C., Gatesy, J., DeSalle, R., 1995. Elision: A method for
accomodating multiple molecular sequence alignments with alignment-ambiguous sites. Mol. Phylogenet. Evol. 4, 1–9.
Wood, S.L., 1973. New synonymy in American bark beetles (Scolytidae: Coleoptera). Gt. Basin Nat. 33, 169–188.
Wood, S.L., 1982. The bark and ambrosia beetles of North and
Central America (Coleoptera: Scolytidae), a taxonomic monograph. Gt. Basin Nat. Mem. 6, 1–1359.
Wood, S.L., 1986. A reclassification of the genera of Scolytidae
(Coleoptera). Gt Basin Nat. Mem. 10, 1–126.
Wood, S.L., Bright, D., 1992. A catalog of Scolytidae and Platypodidae (Coleoptera) Part 2: Taxonomic index. Gt. Basin Nat. Mem.
13, 1–1553.
Woodruff, R.E., 1970. A mangrove borer, Poecilips rhozophorae
(Hopkins) (Coleoptera: Scolytidae). Florida Dept. Agric. Entomol.
Circ. 98, 1–2.
Xia, X., 1998. The rate heterogeneity of non-synonymous substitutions
in mammalian mitochondrial genes. Mol. Biol. Evol. 15, 336–
344.
Yang, Z., 1998. On the best evolutionary rates for phylogenetic
analysis. Syst. Biol. 47, 125–133.