GENETIC STRUCTURE OF THE PINYON PINE BEETLE, IPS CONFUSUS (LECONTE)
DURING AN OUTBREAK
By
Liu Yang

A THESIS
Submitted to Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Entomology
2012

ABSTRACT
THE GENETIC STRUCTURE OF THE PINYON PINE BEETLE, IPS CONFUSUS (LECONTE)
DURING AN OUTBREAK
BY
Liu Yang
Genetic structure of phylophagous insects are formed under many factors, such as coevolutionary effect with hosts, geographic distribution, or migration which is impacted by
climatic fluctuations or natural disturbances. To investigate the impact of 2003 pinyon pine
beetle outbreak on its genetic structure, we sampled in total 244 individuals from 28
populations across six states in Southwest of United States in 2001 and 2003, constructed a
phylogenetic tree, compared genetic diversity within each populations before and during
outbreak, calculated genetic differentiation among populations, tested genetic variations on
different hierarchical levels, and performed mantel tests to test isolation-by-distance. The
diversity analysis and haplotype network did not demonstrate significant differences
among populations before and during outbreak. Thus the outbreak had little impact on the
genetic structure of Ips confusus. Spatial patterns of haplotype distribution, diversity trend,
AMOVA and Mantel tests indicated that the genetic structure was closely associated with
geography. These results suggest that multiple short-distance dispersals among proximal
populations rather than dispersal among distant populations, have shaped the genetic
structure of I. confusus despite greater potential for long distant dispersal during outbreaks.

ACKNOWLEDGEMENTS
I thank Dr. Anthony Cognato for providing the sequences of Ips confusus, tutoring me with
phylogeny and haplotype network constructions, and helping edit this thesis; I also thank
Dr. Mike Kaufman for tutoring me with PCA analysis and Dr. Kim Scribner for advices with
statistical analyses. Amanda Lorenz, Nick Barc and Rachel Olson provided valuable
comments and great friendship during the whole time of writing.

iii

TABLE OF CONTENTS

LIST OF TABLES................................................................................................................... v
LIST OF FIGURES ................................................................................................................ vi
Introduction........................................................................................................................ 1
Methods and Materials ...................................................................................................... 4
Results ................................................................................................................................. 9
Discussion ......................................................................................................................... 14
APPENDICES...................................................................................................................... 18
Appendix A (Tables) ...................................................................................................................... 19
Appendix B (Figures)..................................................................................................................... 37

REFERENCES ..................................................................................................................... 43

iv

LIST OF TABLES

Table 1a. Sampling information. ................................................................................................... 19
Table 1b. Sampling information. ................................................................................................... 20
Table 2. Molecular diversity information. ..................................................................................... 21
Table 3. Haplotype frequency. ...................................................................................................... 23
Table 4. PC scores derived from haplotype frequencies of each populations. ............................. 24
Table 5. Multivariate analysis of variance (MANOVA) .................................................................. 25
Table 6a. Analysis of molecular variance (AMOVA) in one group. ................................................ 25
Table 6b. Analysis of molecular Variance (AMOVA) in two groups............................................... 25
Table 6c. Analysis of molecular variance (AMOVA) in three groups. ............................................ 25
Table 6d. Analysis of molecular variance (AMOVA) in outbreak effect. ....................................... 26
Table 7a. Pairwise Fsts among populations on outbreak effect. .................................................. 27
Table 7b. pairwise Fsts among populations on geography. .......................................................... 32
Table 8. Three Mantel tests........................................................................................................... 36

v

LIST OF FIGURES
Figure 1 Haplotype network. ........................................................................................................ 37
Figure 2. The map of sampling localities....................................................................................... 38
Figure 3 Principal component analysis. ........................................................................................ 40
Figure 4 Regression plot of haplotype diversity (H) against latitudes. ......................................... 41
Figure 5 Mantel tests. ................................................................................................................... 42

vi

Introduction
Insect populations often experience abnormal growth including the sudden increase in
the number of populations and dispersal events during outbreaks (Christiansen et al.
1987). As food sources are depleted emigration of individuals to new food sources is
expected. This dispersal has potential genetic consequences for the metapopulation
(Vandergast et al. 2004; Eckert et al. 2008). A new population founded by many individuals
from a nearby outbreak would retain much of the genetic variation found in the source
population (Ibrahim et al. 1996). However, dispersal of few individuals to distant resources,
without an established population, would result in a genetic bottleneck given that
additional emigration to the new population would be likely rare. Empirical data
documenting the effect of outbreaks the on genetic structure of pest species are limited to a
few studies (e.g. Ibrahim et al. 2000, 2001; Chapius et al. 2008, 2009; Fonseca et al. 2010;
James et al. 2011; Kobayashi et al. 2011; Ronnas et al. 2011; Tao et al. 2012). The majority
of these studies suggest that outbreaks do not result in an increase in gene flow among
nearby populations (James et al. 2011; Kobayashi et al. 2011; Ronnas et al. 2011). This
observation may depend on the species and the regional scope of the study. For example,
homogenizing effect of outbreak events on genetic variation (lower population
differentiation in outbreak than non-outbreak) was not observed in worldwide populations
of migratory locust (Chapius et al. 2008) however at the regional scale, homogenizing effect
was found on the population structure among local populations (Chapius et al. 2009). Thus,
the additional study of species prone to outbreaks could potentially reveal common genetic
consequences of epidemics. A recent outbreak of a bark beetle, Ips confusus (LeConte)
presents opportunity to document this species population genetic structure before and
1

during an outbreak.
Ips confusus occurs in Western US including Arizona, California, Colorado, Nevada, New
Mexico, Oregon, Texas, Utah, Wyoming and approximately overlaps with the distribution of
pinyon pines (Wood 1982). This bark beetle, as adults and larvae, mainly feeds on two pine
species, Pinus edulis Engelman and Pinus monophylla Torrey & Fremont and they are
considered as host specific despite their occasional use of other conifers (Lanier 1970;
Cognato et al., 2003). The beetle’s lifecycle is dependent on the tree. Colonization of a
suitable host usually starts with the male beetles excavating an entrance tunnel followed by
a nuptial chamber. While the male bores in the tree, it produces pheromones that attract
more conspecifics and thus initiates a mass attack (Wood et al., 1967). Two- five females
join the male in the nuptial chamber, mate, and each construct a gallery in which they lay
eggs. After the eggs hatch, the larvae feed on inner bark, until they pupate under the bark
and emerge (Furniss and Carolin 1977, Wood 1982, Eager 1999, Negron and Wilson, 2003).
Ips confusus usually has three, sometimes four generations a year. It is not uncommon that
after first emergence, the parent beetles with re-infest the tree and produce similar size
broods (Wood 1982).
Ips confusus is not the most aggressive bark beetle species. It rarely feed on healthy
trees and usually attacks stressed or dying individuals. Nevertheless, outbreaks can occur
during times of severe drought or some other natural disaster (Furniss and Carolin 1977;
Negron and Wilson, 2003). Recently, a large outbreak of I. confusus occurred during a
prolong drought (2001–2004) in Arizona, New Mexico, Nevada, Utah, and southwestern
Colorado. Ecological damage and economical loss were extensive. In 2003, during peak
outbreak, approximately 15–30% of pinyons were killed throughout >1.6 million hectares

2

(USDA–Forest Service 2004, Breshears et al. 2005, Williams et al., 2010). By 2007, beetle
populations and pinyon mortality declined to endemic levels (Williams et al., 2010; Halsey
et al., 2011).
Population genetics for I. confusus was characterized for 100 individuals from ten
populations collected in 2001 at the beginning of the outbreak (Cognato et al. 2003, Halsey
et al., 2011). Using partial mitochondrial COI DNA sequences, Cognato et al. (2003) revealed
much haplotype diversity, which was partitioned into two clades corresponding to
geographic regions (California and the Rocky Mountains). These clades likely developed in
Pleistocene refugia. Interestingly, of the 15 observed haplotypes, 11 were unique to a
population and distributed among Californian and Rocky Mountain populations. This
amount of haplotypic endemism appears typical for bark beetles (Cognato et al. 1999,
Stauffer et al. 1999, Cognato et al. 2005a,b, Menard and Cognato 2007). Gene flow among I.
confusus populations was recurrent throughout the species history however the amount of
gene flow among present day populations is not known (Cognato et al. 2003). Bark beetles
in general are able flyers and Ips species have a maximum potential of 50 km/generation
unaided by wind or human transport (Jactel and Gaillard 1991). Dispersal ability is
influenced by environmental conditions especially wind (Gara and Vite 1962; Byer 2000).
Air currents can potentially carry beetles hundreds of kilometers (de la Giroday et al. 2011,
Safranyik et al. 2010) and consequently cause genetic bottlenecks as observed for the
mountain pine beetle ( James et al. 2011). Hence, during the outbreak of I. confusus long
distance dispersal may have increased due to the greater numbers of individuals and if so,
the rare haplotypes observed in Cognato et al. (2003) could have potentially become fixed
in other locations.

3

This study investigates the diversity of mitochondrial cytochrome oxidase I
haplotypes among I. confusus individuals collected before and during a region-wide
outbreak. We test the hypothesis that I. confusus haplotypes are not fixed for populations
thus indicating little barrier to gene flow. Also, we hypothesize that rare haplotypes
observed at the beginning of the outbreak are rare among populations sampled during the
outbreak.

Methods and Materials

Approximately 10 live Ips confusus adult beetles were collected from host trees (P.
edulis and P. monophylla) of 28 localities across the six states in the Southwest of United
States (CA, NV, UT, CO, AZ and NM) (Figure 1, Table 1a). Each beetle was removed from the
tree by using a knife and forceps, and in total 267 individuals were collected and stored in
100% ethanol for later use (details in Cognato et al., 2003).
Total genomic DNA was extracted from beetle thoraces, with a silica-based spin
column procedure, following the manufacturer’s tissue protocol as described in Cognato et
al., 2003. Mitochondrial cytochrome oxidase I DNA (399bp) was amplified for total 267
Individuals via polymerase chain reaction (PCR) with primers C1-J-2183 and C1–N-2611
following the methods of Cognato et al. (2003). Each PCR reaction consisted of: 35 ul
ddH2O, 5 ul 10X TaqDNA polymerase buffer (Promega, Madison, WI), 4 ul 25 mM Promega
MgCl2, 1 ul 40 mM deoxynucleotide triphosphates (dNTPs), 2 ul of each 5 mM
oligonucleotide primer, 0.2 ul of Promega TaqDNA polymerase, and 1 ul of DNA template.
The PCR was performed on a thermal cycler (MJ Research, Cambridge, MA) under the
4

following conditions: one cycle for 3 min at 95℃, .75min at 45℃, 1 min at 72℃, followed
by 34 cycles of .5min at 94℃, .75 min at 45℃, 1 min at 72℃, and a final elongation cycle of
5 min at 72℃.
Unincorporated dNTPs and oligonucleotides primers were removed from PCR with a
Qiaquick PCR Purification Kit (Qiagen) following the manufacture’s instructions and were
directly sequenced on an ABI 377 automated sequencer as described in Coganto et al.
(2003). Both sense and antisense strands were sequenced for all individuals. Consensus
sequences were arranged and inspected for nucleotide ambiguities in Sequencher (Ann
Arbor, MI) and resulted in 244 sequences that were used for subsequent analyses.
Sequences were complied in MacClade (Maddison& Maddison 2005) and submitted to
GenBank (upon publication).
We used parsimony and Bayesian analyses to investigate potential phylogenetic signal
among the beetle mitochondrial haplotypes. PAUP* ver. 4.0b10 (Swofford 2003) was used
to conduct the parsimony analysis with the following settings: heuristic searches with 10
replicates and tree-bisection-reconnection branch swapping. We attempted a Bayesian
analysis of these data (MRBAYES3; http:// morphbank. ebc. uu. Se/ mrbayes/) using a time
reversible (GTR + C + I) model; four Metropolis-Coupled Markov chain Monte Carlo
searches (one cold, three heated) which were run twice simultaneously for 20 million
generations, each with sampling every 100th iteration. These runs did not complete or
approach to stationarity after two weeks of computation. Hence, Bayesian analysis was not
pursued.
A haplotype network was created with TCS with default algorithm (Clement et al., 2000)
because little resolution found in the parsimony tree. Ambiguous characters were

5

considered as missing data and a limit of nine mutational steps were considered for the
95% plausible set of alternative parsimony networks (Clement et al., 2000).
Molecular diversity for mtDNA sequences was analyzed by estimating haplotype
diversity (H) (Nei 1981) and nucleotide diversity (π) (Nei 1987)in two groups: populations
(1-10) from 2001 and populations (11-28) from 2003. Haplotype diversity (H) was
calculated as H=n/(n-1)(1-∑pi2) where n is the number of gene copies in the population
and pi is the frequency of the ith haplotype (Nei, 1987). For codominant marker, it is the
same formula for calculating expected heterozygosity; nucleotide diversity (π) measures
the average nucleotide differences between all pairs of DNA sequences randomly chosen
from the population. It is calculated as π =n/(n-1)(Σxixjdij) where n is the sample size, xi
and xj are the frequencies of haplotype i and j, and dij is the fraction of the number of
nucleotide differences between two haplotypes out of total nucleotide number per
haplotype (Tajima, 1993; Excoffier et al., 2005). Besides haplotype diversity and nucleotide
diversity, the number of unique haplotypes, the number of pairwise differences, and their
means and confidences were calculated (Table 2). The number of polymorphic sites in each
population was calculated and the loci were identified (Table 1a). We calculated the means
and variances of the pre-outbreak populations and the during-outbreak populations on
molecular diversity indices (Table 2) and haplotype frequencies (Table 3), and used the ttests (on means) and F-tests (on variances) to compare the means and variances between
pre and during-outbreak in order to access the effect of the outbreak (see the result). We
also investigated the relationship between haplotype diversity (H) and latitudes by plotting
the regression graphs (Figure 4).

6

Gene flow and genetic drift usually result in changes in species’ spatial genetic structure
which can be assessed by measuring the changes of haplotype frequencies before and
during the outbreak. We calculated the haplotype frequencies for each haplotype in each
populations, then converted the haplotype frequencies into percentages and log-ratio
transformed, which were treated as variables in the Principal Component Analysis (PCA). In
PCA, the variables were transformed into lower dimensional space (in our case, three
dimensional, thus three Principal Components PC1, PC2 and PC3). The PC scores were
produced under each category (geographic regions and time) using software JMP (a SAS
product) (Table 4). PCA created axes from the variables and assigned them along the axes,
so to explain the distribution of sample values (Figure 3). Then we conducted a Multivariate
analysis of variance (MANOVA) of the values categorized by pre-outbreak, during-outbreak
and geographic populations (Table 5).
The genetic structure among populations was analyzed by computing the hierarchal
analysis of molecular variance (AMOVA) based on estimated Fsts, the exact test of
population differentiation, and a Mantel test with the software, Arlequin ver. 3.0 (Weir and
Cockerham 1984; Excoffier et al., 2005).
AMOVA test was initially performed on different levels of genetic variation and
associated F-statistics for testing corresponding significance levels. The total variance (σ2)
was partitioned into covariance components(σ2a, σ2b and σ2c) due to differences among
groups, differences among populations within group and differences within populations,
respectively (Rousset 2000; Excoffier et al., 2008). Since the same framework could be
extended to the fixation index FST, which is identical to the F-statistics over loci (Michalakis

7

and Excoffier 1996), the significance of F-statistics was tested to interpret the significance
of the fixation indices, by using non-parametric permutation approach with 1,000
iterations (Excoffier et al., 1992). In our study, the hierarchical variation analysis (AMOVA)
was first conducted in one group consisting entire 28 populations 42 haplotypes, and thus
σ2a and FST were tested by permuting haplotypes among populations; then the AMOVA
test was conducted in several smaller groups with different combinations of populations,
σ2c and within populations (FST), σ2b and among populations within groups (FSC), and σ2a
and among groups (FCT) were tested respectively in each case (Table 6. a-c)
The exact test of population differentiation was conducted to test whether populations
were significantly different from each other by comparing the pairwise genetic distances
(Excoffier et al., 2005). This exact test was designed to test the null hypothesis of random
distribution of k haplotypes among r populations (k=42, r=28 in our case) (Raymond and
Rousset 1995), which was extrapolated from Fisher’s exact test of 2x2 contingency table.
The P-value was calculated by summing up the probabilities of all contingency tables that
have same or smaller probabilities and with same row and column sums (Raymond and
Rousset 1995).
Mantel tests were performed to test for correlation between genetic and geographic
distances by evaluating the correlation coefficient (r) and the statistical significance (Pvalue) (Mantel 1967; Sokal & Rohlf 1995). The genetic distances between populations were
estimated as Fst/(1-Fst) (Slatkin’s Distance) (Slatkin 1995) under the Tamura & Nei’s
substitution model (Tamura & Nei 1993), at the permutations of 5000, significant level of
0.05, and Gamma value of 0. We created the geographic distance matrix by calculating the

8

great-circle distances among the 28 populations using the on-line geographic distance
calculator (http://www.movable-type.co.uk/scripts/latlong.html). If the P-value was
smaller than 0.05, then the null hypothesis of no relationship between two distance
matrices was rejected. Isolation-by-distance model was tested and described by plotting
pairwise Fst/(1-Fsts) against geographic distances in the eastern and western group
respectively (Figure 5).

Results
Parsimony resulted in a single tree (not shown here), which was mostly unresolved
except for one large clade, which corresponded 12 haplotypes that were mostly associated
with CA and NV localities Haplotype network showed similar relationship among
haplotypes as the parsimony tree but provided additional information on the reticulation
among haplotypes (Figure 1). Eight haplotypes out of 42 were common (haps 3, 6, 7, 10, 15,
17, 24, 28) and occurred in more than one population. Haplotype 6 (Figure 1) was most
common and was shared by 154 individuals throughout 26 populations. Haplotypes 7 and
10 were the second and third most frequent and shared by 25 and 13 individuals,
respectively. Two haplotype networks were centered around haplotypes 7 and 10, which
were clustered in two distinct groups of closely located populations, respectively (Figure 1,
2). Among the 19 haplotypes that occurred in pre-outbreak populations, only five
haplotypes (haps 6, 7, 10, 15, 17) were observed in outbreak populations (Figure 1, Table
2). None of the unique haplotypes (13) observed in the pre-outbreak populations occurred
in the outbreak populations.
Diversity analysis indicated that pre-outbreak populations had a higher level of
9

genetic variation. Mean values of the number of unique haplotypes, haplotype diversity(H),
nucleotide diversity(π) and the number of pairwise differences from pre-outbreak
populations(1.3, .5229, .0032 and 1.2667, respectively) were all higher compared to the
outbreak populations(1.2, .4002, .0018 and .7024, respectively), but the differences were
not statistically significant (Table 2). The average number of unique haplotypes per
population from the two time periods was similar (1.3 and 1.2). The results suggested that
this genetic diversity was endemic, and there was no obvious sign of founder from preoutbreak populations in outbreak populations, because there was no significant genetic
diversity reduction in outbreak. Interestingly, the highest diversity indices all occurred in
the localities in the same region (populations 3, 10 and 14), which suggested that genetic
variation was associated with geography.
The table 1b and Figure 2 showed the trend of haplotype diversity. In the central
range, populations (24-28) were dominated by the more common haplotypes whereas
starting from population 2 towards northwest haplotype diversity increased gradually, to
population 20 where unique haplotypes comprise 50% of the haplotypes. Northeastern
populations (3, 10, 13, 14) also exhibited greater haplotype diversity. There was a strong
linear relationship of genetic diversity with higher latitude (Figure 4) supported the above
pattern of haplotypes. What’s interesting was haplotype 3 only appeared in three distant
populations (1, 2 and 9) that were scattered at three different states (CA, AZ and NW,
respectively), but seemingly pointed three different clusters (west, center and east). It is
possible that two groups of Ips beetles dispersed from northwest and northeast separately,
and eventually jointed in the south centerline area (where populations 2,4,5 located). .
PCA and, T-test and F-test were conducted to investigate the outbreak effect and

10

geographic distribution on spatial genetic structure. PCA explained the sample values
distribution along two perpendicular axes: PC1 always explains most of the variance, and
PC2 explains more variance and PC3 explains less variance than PC1 and PC2. Sometimes,
PCA can calculated more and illustrate in 3D plot (in which new axis is perpendicular to
previous) but in our case, PC1 has explained 53.1% of total variance, PC2 explained
additional 22.5% and PC3 explained 5.6%. So overall, 81% of total variance was explained
and thus we demonstrated the results in 2D graph (Figure 3). From the graph, we can see
that the separation of the sample values was not observed along PC1, which simply means
that the sample groupings were not associated with the most efficient way to address the
variance among the haplotypes. However, the haplotype variance was more associated with
geographic regions seen from PC2. Although it’s not usual to see this, it’s not surprising
given the large overall variance in our dataset with high number of unique haplotypes.
T-tests and F-tests were given on the means and variances of each haplotype
frequencies (haps 6, 7, 10, 15 and 17) respectively (Table 3). All five tests (on each of the
five haplotypes) failed to reject the null hypotheses that there were no significant
differences between the means (or variances) and compared populations. In other words,
the outbreak did not effect spatial genetic structure of I. confusus. Similar tests were
conducted in the grouping based on geographic regions on the same haplotypes (west vs
center, west vs east, center vs east), and haps 6, 7 and 10, which were more common
showed strong association with geographic locations (Table 3).
MANOVA on PC values indicated significant difference on geography (Table 5),
therefore the haplotype patterns were different between each other at least two of the
geographic regions. There was no significance effect on outbreak solely or the interaction

11

between outbreak and geographic location (Table 5). Populations from east and west were
different based on the T-tests results on haplotype frequencies (Table 3) and subsequent
tests (results not shown).
The hierarchical variation analysis, we conducted both standard global AMOVA tests
and locus-to-locus AMOVA (because of the missing data), two types of AMOVA tests were
not statistically significantly different thus we only present the results from the standard
AMOVA tests here. The results of the one group (28 populations) analysis showed 55.68%
and 44.32% of genetic variation was attributed to the variance within populations and the
variance among populations (Table 6a.) and the global Fst was .4432(p <.0005). We
conducted two-group AMOVA to investigate the difference between western and eastern
geographic regions (Table 6b.). The results of the two-group AMOVA showed that within
population contributes about half (49.86%) of the total variation, and the remaining
variance is explained by among groups (19.97%) and among populations (30.18%). All
three levels were highly significant (p <.0005). The p-value of Fct and Va (among groups)
was lower than the significance level, indicating that there were significant differences
between the two groups, suggesting genetic structure among the 28 populations. A cluster
of populations dominated by pure haplotype (6) was observed in the central region (Figure
2, oval area), so we separated those populations from the rest of Western group and formed
a three-group AMOVA (Table 6c.). All three levels were highly significant (p-value< .0005)
as we expected, which confirmed that genetic structure strongly associated with geography.
A fourth AMOVA analysis was also performed to test genetic differentiation between preoutbreak and outbreak populations (Table 6d.). Genetic variation between groups was very
small (little differentiation) and not significant, which suggested genetic differentitation

12

among populations was not associated with outbreak effect. The genetic variation among
populations and “outbreak” differentiation contributed 44.32% and 45.44% of total
variance, versus the variation among populations between geographic regions was smaller
but significant (30.18%, P-value <.0005),
The result of the exact test of population differentiation was shown as the matrix of
pairwise Fsts (Table 7). Significant genetic differentiation Fsts were common all over the
place, there were 140 out of 378 (37%) pairwise comparisons in whole 28 populations
found genetically significant different (P<.05). No obvious trend of Fsts was observed
between pre-outbreak and outbreak populations (Table 7a), even though the average
number of significantly differentiated population in each grouping was slightly different
(12.20 of pre-outbreak vs. 8.89 in outbreak), suggesting that population differentiation was
not related with the outbreak. Among geographic regions however, both within Western
grouping (except populations 1,6) and within central grouping showed very low genetic
differentiation (Table 7b). Except populations 1,2,3,6,10,13,14 and 21, which all
coincidentally located at the edge of our sampling area, the rest populations were not
significantly differentiated from each other (P>.05), suggesting some gene flow. Populations
13, 16, and 27 (Fst=.8881, P-value<.05), and populations 7 and 22(Fst=-.1133, P-value>.86)
were the most and the least differentiated populations respectively (Table 7), suggesting
the possibility of isolation by distance, because the former two were located at the two
corner of the area whereas latter two were very near to each other.
The global Fst (.4432) across entire 28 populations was significant indicating that
populations were highly differentiated from each other and there was restriction of gene
flow among all populations, whereas global Fsts of pre-outbreak and outbreak populations

13

were similar (.3480 and .5172 respectively). In the AMOVA separating pre- and during
outbreak populations (Table 6d), Fct (among pre-outbreak and outbreak populations) was
only .0042, showing little differentiation, which agreed the similar global Fst values that
outbreak has very limited impact on populations differentiation. Also, there were no
obvious geographic patterns of significant differentiation.
Mantel tests were performed to test isolation-by-distance (Table 8, Figure 5). The
first Mantel test was performed on all 28 populations, and there was no significant
correlation between two matrices (r=.1102, P=.1316); then we performed Mantel tests on
Western, Central and Eastern group corresponding to geographic regions respectively as
before, and found strong indication of isolation-by-distance in Western region(r=.5188,
P=.0004) and partially in Eastern region (r=.4321, P=.0434), but not in central area
(r=.2131, P=.1910) (Table 8.).

Discussion
Ips confusus is not the most aggressive bark beetle and thus its appearance is related
to the presence of weakened or dying host trees. Drought produces large amount of
stressed trees, which would provide perfect habitats and food sources for the beetles, in
other words, create the conditions for bark beetle outbreaks. Along with the increasing
drought severity index from 2000, the annual area killed by bark beetle started to increase
dramatically in 2001, peaked in 2003 and dropped to an endemic level by 2007 (Williams
et al., 2010). During that period of time, I. confusus outbreak occurred in the six states
(Breshear et al., 2005; USDA-Forest Service 2004; Williams et al., 2010).
Our study investigated the outbreak effect on the genetic structure of this species by
14

building the haplotype network, comparing the genetic diversity and genetic variation
distribution between pre-outbreak and outbreak populations, and rejected the hypothesis
that outbreak impacted the genetic structure of I. confusus. The six states we sampled
covered the geographic range of pinyon pines of United States (Little 1971), and on each
tree we sampled an individual from separate broods (leading by one single male beetle) so
not to bias the sampling of mtDNA haplotypes .
Changes in haplotype frequency are reliable indicator of gene flow that could result
in alterations of spatial genetic structure. The statistical tests in our study showed no
significant differences between haplotype frequency changes before and during outbreak.
In addition, most haplotypes were unique and those from pre-outbreak populations were
not found in greater abundance in outbreak populations: Forty-two haplotypes were found
in 244 individuals; 34 haplotypes were unique which was consistent with the haplotype
diversity observed among other scolytine and some insect species (Menard and Cognato,
2007, Kobayashi et al., 2011).

The genetic diversity of pre-outbreak and outbreak I.

confusus populations was similar to the diversity observed with endemic and epidemic
populations of mountain pine beetle (Chapuis et al., 2008 and 2009). Also, I. confusus
populations were isolated by distance. The distribution of the observed genetic diversity
suggested that Pleistocene geology shaped genetic structure and short-distance dispersal
accounted for beetle migration.
Although, we did not observe an association between the distribution of genetic
variation and outbreak status, genetic variation was associated with geography. The three
genetic clusters revealed by phylogenetic tree, AMOVA analyses (Table 3) and Mantel tests
(Table 8, Figure 5) are associated with the western, southwestern, and eastern range of the

15

beetle as observed in Cognato et al. (2003). A significant association between interpopulation variance in haplotype frequency and geographic distance (Isolation by distance)
was observed within western region (P-value=.0004)

and part of eastern region (P-

value=.0434). The wide distribution of common haplotypes (i.e. haplotypes 6 and 7) and
the pattern of genetic variability association with geography showed in AMOVA and Mantel
tests is likely due to Pleistocene geologic events observed for other North American
scolytine species (Cognato et al. 1999, Kelly et al. 1999, Cognato et al. 2005, James et al.,
2011; Tsui et al., 2012). It is not well understood how the Pleistocene effected the
distribution of genetic variation among I. confusus populations. However, beetle
populations likely followed the distribution of their tree hosts to lower altitudes and
latitudes in the colder climate (Cognato et al. 2003).
Bark beetles’ attack or new colonization occurs when a single male beetle
successfully bores into a host tree and produces pheromone, which attracts female and
other male beetles fly and join. After the mating, younger generation finishes its growth in
the inner bark, then bores out of the bark and fly to next targeting host tree. Therefore
habitat connectivity helps to mediate beetle colonization (Robertson et al., 2009), despite
short-distance dispersal, which may explain the similar frequencies of common haplotypes
in proximal populations (populations 6, 7, 21, 22, Figure 2). However landscape features
(e.g., mountains and treeless areas) likely impact beetle migration (Aukema et al., 2008;
Chen and Walton 2011). For example, the Shoshone Mountains lying between populations
19 and 20 are a possible barrier to beetle movement evidenced by the different haplotypic
composition in these proximal populations. Long-distance dispersal (< 50 km), as observed
with other bark beetles (Chen and Walton 2011; Lowe 2009) is possible but it is likely

16

uncommon given that distribution of haplotypes is better explained by the influence of
Pleistocene geography. Landscape features throughout time and Western North America
(de la Giroday et al., 2011) likely influenced the dispersal of I. confusus by curbing longdistance dispersal, and shaped the current population genetic structure.
There is much evidence for the effect of climate change on insect populations (e.g.
Carroll et al. 2004; Robinet and Roques 2010).

Increased favorable environmental

conditions (e.g. increase of stressed trees in our case promote an increase in the population
size, which increases the likelihood of an outbreak. Our study, as others suggest that
drought promotes multiple independent outbreaks among some herbivorous insects (e.g.,
Ronnas et al. 2011). However, intrinsic factors, such as physiology and behavior, mostly
influence the dispersal ability of the insects and hence short distance dispersal mediates
gene flow. Extrinsic stochastic factors, such as wind and humans, may become more
important to long distance dispersal once outbreak populations grow to a critical size and
number (e.g., Safranyik et al. 2010, de la Giroday et al. 2011). As a consequence, long-term
drought or global warming will likely promote increased gene flow among I. confusus.

17

APPENDICES

18

Appendix A (Tables)
Table 1a. Sampling information.
Population ID, Location (county names), Latitude, Longitude, elevation, host tree, the
number of individuals(NI) collected in each population, the number of haplotypes (NH) in
each population and the number of polymorphic sites (NP) (the number of loci that has
more than one allele per locus) were showed in this table. White Pine= WP, Little
Antelope summit= l.a. sum.
Po Location
Latitude Longitude
p
ID
Pre-outbreak populations (from 2001)
1
San Bernardino, 34°18'N 116°49'W
CA
2
Greenlee, AZ
33°10’N 109°23’W
3
Dolores, CO
37°45’N 108°00’W
4
Greenlee, AZ
33°38’N 109°20’W
5
Gila, AZ
33°36’N 110°15’W
6
Mono, CA
38°05’N 119°10’W
7
Inyo, CA
37°15’N 118°10’W
8
Otero, NM
32°53’N 105°30’W
9
Sandoval, NM
36°01’N 106°57’W
10 Montezuma, CO 37°28’N 108°29’W
Outbreak populations (from 2003)
11 Huerfano, CO
37°30’N 104°42’W
12 Fermont, CO
38°22’N 105°41’W
13 Rio Blanco, CO
39°41’N 108°48’W
14 Duchesne, UT
40°08’N 110°29’W
15 Tooele,UT
40°00’N 112°17’W
16 WP, nr Baker, NV 39°01’N 114°12’W
17 WP, nr Ely, NV
39°03’N 114°37’W
18 WP, nr l.a. sum.
39°24’N 115°28’W
19 Lander, NV
39°27’N 116°45’W
20 Churchill, NV
39°15’N 117°48’W
21 Douglas, NV
38°48’N 119°44’W
22 Esmeralda, NV
37°25’N 117°38’W
23 Clark, NV
36°16’N 115°32’W
24 Washington, UT 37°26’N 113°30’W
25 Iron, UT
37°40’N 113°00’W
26 Washington, UT 37°17’N 113°06’W
27 Coconino, AZ
36°51’N 112°16’W
28 Coconino, AZ
35°24’N 111°35’W

19

Elevatio Host tree
n (ft)

1976
1948
2122
1953
1740
1989
2206
2279
1980
1828
1648
2030
1788
2055
1885
1500
1879
2106

NH

NP

P. monophylla

>3030

NI

9

6

5

P. edulis
P. edulis
P. pungens
P. edulis
P. monophylla
P. monophylla
P. edulis
P. edulis
P. edulis

8
7
9
10
10
9
6
10
8

4
4
2
2
3
3
1
3
6

7
6
1
1
2
3
0
5
5

P. edulis
P. edulis
P. edulis
P. edulis
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. monophylla
P. edulis
P. monophylla
P. edulis
P. edulis

10
10
10
9
7
9
10
10
10
10
8
6
9
8
7
8
9
8

3
3
5
5
3
1
2
3
2
6
4
3
3
1
1
2
1
3

3
2
4
10
2
0
1
2
1
4
3
2
2
0
0
2
0
6

Table 1b. Sampling information.
Population ID, haplotypes and the haplotype number (in the bracket) in each populations
were listed in the table.
Pop ID
Haplotype (#)
1
Hap 1 (1); hap 2 (3); hap 3 (1); hap 4 (1); hap 6 (2); hap 7 (1)
2
Hap 3 (3); hap 6 (3); hap 9 (1); hap10 (1)
3
Hap 3 (1); hap 5(1); hap 6 (3); hap 10(2)
4
Hap 6 (8); hap 8 (1)
5
Hap 6 (9); hap 11(1)
6
Hap 7 (8); hap 6 (1); hap 12 (1)
7
Hap 6 (5); hap 7 (3); hap 13 (1)
8
Hap 6 (6)
9
Hap 6 (8); hap 10 (1); hap 15 (1)
10
Hap 10 (1); hap 14 (2); hap 16 (2); hap 17 (1); hap 18 (1); hap 19 (1)
11
Hap 6 (8); hap 17 (1); hap 20 (1)
12
Hap 6 (8); hap 21 (1); hap 22 (1)
13
Hap 10 (6); hap 23 (1); hap 24 (1); hap 25 (1); hap 39 (1)
14
Hap 6 (2); hap 10 (2); hap 24 (1); hap 26 (3); hap 27 (1)
15
Hap 6 (5); hap 28 (1); hap 29 (1)
16
Hap 6 (9)
17
Hap 6 (9); hap 28(1)
18
Hap 6 (8); hap 7 (1); hap 30 (1)
19
Hap 6 (6); hap 7 (4)
20
Hap 6 (5); hap 7 (1); hap 31 (1); hap 32 (1); hap 33 (1); hap 34 (1)
21
Hap 6 (1); hap 7 (5); hap 35 (1); hap 36 (1)
22
Hap 6 (3); hap 7 (2); hap 37 (1)
23
Hap 6 (7); hap 15 (1); hap 38 (1)
24
Hap 6 (8)
25
Hap 6 (8)
26
Hap 6 (7); hap 40 (1)
27
Hap 6 (9)
28
Hap 6 (6); hap 41 (1); hap 42 (1)

20

Table 2. Molecular diversity information.
The number of haplotypes, the number of unique haplotypes, and a list of the unique haplotypes in each populations were
summarized in this table, and diversity indices including gene diversity(H), nucleotide diversity(π) and the number of
pairwise differences with means and 95% CI were listed in the table as well. μ represented mean value.
Pop # of #
of Unique
Haplotype diversity (H)
ID
haps unique haps
Mean
95% CI
haps
Before-outbreak populations (from 2001)
1
6
3
h1, 2, 4
0.8889
[0.7979, 0.9799]
2
4
1
h9
0.7857
[0.6730, 0.8984]
3
4
1
h5
0.8095
[0.6797, 0.9393]
4
2
1
h8
0.2222
[0.0560, 0.3884]
5
2
1
h11
0.2000
[0.0459, 0.3541]
6
3
1
h12
0.3778
[0.1965, 0.5591]
7
3
1
h13
0.6389
[0.5131, 0.7647]
8
1
0
N
0.0000
0
9
3
0
N
0.3778
[0.1965, 0.5591]
10
6
4
h14, 16, 0.9286
[0.8442, 1.0130]
18, 19
μ
3.4
1.3
0.5229
[0, 1.0130]
During-outbreak populations (from 2003)
11
3
1
h20
0.3778
[0.1965, 0.5591]
12
3
2
h21, 22
0.3778
[0.1965, 0.5591]
13
5
3
h23,25,
0.6667
[0.5034, 0.8300]
39
14
5
2
h26, 27
0.8611
[0.7739, 0.9483]
15
3
1
h29
0.5238
[0.3152, 0.7324]
16
1
0
N
0.0000
0
17
2
0
N
0.2000
[0.0459, 0.3541]
18
3
1
h30
0.3778
[0.1965, 0.5591]
19
2
0
N
0.5333
[0.4386, 0.6280]
21

Nucleotide diversity(π)
mean
95% CI

# of pairwise differences
mean
95% CI

0.0050
0.0059
0.0082
0.0006
0.0005
0.0010
0.0024
0
0.0025
0.0060

[0.0015, 0.0086]
[0.0019, 0.0010]
[0.0027, 0.0137]
[-0.0003, 0.014]
[-0.0003, 0.0013]
[0.0000, 0.0022]
[0.0004, 0.0044]
0
[0.0004, 0.0047]
[0.0018, 0.0103]

2.0118
2.3464
3.2812
0.224
0.2011
0.4023
0.9548
0
1.0133
2.2329

[0.7652, 3.2584]
[0.9189, 3.7739]
[1.3648, 5.1977]
[-0.0654, 0.5126]
[-0.0689, 0.4711]
[-0.0020, 0.8066]
[0.2390, 1.6706]
0
[0.2739, 1.7528]
[0.8614, 3.6044]

0.0032

[0, 0.0173]

1.2667

[-0.0654, 5.1977]

0.0015
0.0005
0.0024

[3.5E-05, 0.0030]
[-0.0003, 0.0013]
[0.0004, 0.0044]

0.6036
0.2012
0.9632

[0.0823, 1.1249]
[-0.0688, 0.4712]
[0.2496, 1.6767]

0.0106
0.0015
0.0000
0.0005
0.0010
0.0013

[0.0040, 0.0172]
[-6.2E-05, 0.0030]
0
[-0.0003, 0.0013]
[-0.0004, 0.0022]
[-2.5E-05, 0.0027]

4.2437
0.5752
0
0.2011
0.4027
0.5365

[1.9220, 6.5654]
[0.0522, 1.0982]
0
[-0.0689, 0.4711]
[-0.0019, 0.8072]
[0.0531, 1.0199]

Table 2 (cont’d)
20
6
4
21
22
23
24
25
26
27
28
μ

4
3
3
1
1
2
1
3
2.83

2
1
1
0
0
1
0
2
1.2

h31, 32,
33, 34
h35, 36
h37
h38
N
N
h40
N
h41, 42

0.7778

[0.6404, 0.9152]

0.0029

[0.0005, 0.0053]

1.0764

[0.3046, 1.8482]

0.6429
0.7333
0.4167
0.0000
0.0000
0.2500
0.0000
0.4643
0.4002

[0.4588, 0.8270]
[0.5781, 0.8885]
[0.2260, 0.6074]
0
0
[0.0698, 0.4302]
0
[0.2643, 0.6643]
[0, 0.9483]

0.0019
0.0023
0.0012
0.0000
0.0000
0.0013
0.0000
0.0032
0.0018

[0.0001, 0.0036]
[0.0002, 0.0044]
[-0.0001, 0.0025]
0
0
[-8.9E-05, 0.0026]
0
[0.0006, 0.0057]
[-6.2E-05, 0.0172]

0.7541
0.8737
0.4463
0
0
0.5055
0
1.2598
0.7024

[0.1381, 1.3701]
[0.1690, 1.5784]
[0.0117, 0.8809]
0
0
[0.0306, 0.9805]
0
[0.3766, 2.1431]
[-0.0689, 2.1431]

22

Table 3. Haplotype frequency.
The haplotype frequencies of five haplotypes that occurred in pre- and during- outbreak.
Pop
h6
h7
h10
h15
h17
ID
1
0.2222
0.1111
0
0
0
2
0.3750
0
0.1250
0
0
3
0.4286
0
0.2857
0
0
4
0.8889
0
0
0
0
5
0.9000
0
0
0
0
6
0.1000
0.8000
0
0
0
7
0.5556
0.3333
0
0
0
8
1.0000
0
0
0
0
9
0.8000
0
0.1000
0.1000
0
10
0
0
0.1250
0
0.1250
11
0.8000
0
0
0
0.1000
12
0.8000
0
0
0
0
13
0
0
0.6000
0
0
14
0.2222
0
0.2222
0
0
15
0.7143
0
0
0
0
16
1
0
0
0
0
17
0.9000
0
0
0
0
18
0.8000
0.1000
0
0
0
19
0.6000
0.4000
0
0
0
20
0.5000
0.1000
0
0
0
21
0.1250
0.6250
0
0
0
22
0.5000
0.3333
0
0
0
23
0.7778
0
0
0.1111
0
24
1
0
0
0
0
25
1
0
0
0
0
26
0.8750
0
0
0
0
27
1
0
0
0
0
28
0.7500
0
0
0
0
Pop ID
Pre
outbreak
During
outbreak
West
Center
East

Average
Variance
Average
Variance
Average
Variance
Average
Variance
Average

Hap 6
0.5270
0.1283
0.6869
0.0944
0.5303
0.0980
.9004
.0113
.5481

Hap 7
0.1244
0.0676
0.0866
0.0322
0.2803
0.0733
0
0
0

23

Hap 10
0.0636
0.0092
0.0457
0.0219
0
0
0
0
.1325

Hap 15
0.0100
0.0010
0.0062
0.0007
0
0
.0159
.0018
.0091

Hap 17
0.0125
0.0016
0.0056
0.0006
0
0
0
0
.0205

Table 4. PC scores derived from haplotype frequencies of each populations.
Population
ID
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
27
28

Outbreak
Pre
Pre
Pre
Pre
Pre
Pre
Pre
Pre
Pre
Pre
During
During
During
During
During
During
During
During
During
During
During
During
During
During
During
During
During
During

Geo
region
West
East
East
East
Center
West
West
East
East
East
East
East
East
East
East
West
West
West
West
West
West
West
Center
Center
Center
Center
Center
Center

PC1

PC2

PC3

0.1096
0.0574
0.0514
-0.083
-0.0851
0.2187
0.0381
-0.1045
-0.0529
0.185
-0.0633
-0.0645
0.2087
0.1106
-0.0468
-0.1045
-0.0855
-0.0489
0.0349
0.0216
0.1958
0.0349
-0.0802
-0.1045
-0.1045
-0.0802
-0.1045
-0.0539

0.0111
0.0714
0.0859
0.0033
0.0028
-0.1591
-0.0822
-0.0018
0.0297
0.1031
0.0093
0.0077
0.1609
0.0973
0.0124
-0.0018
0.003
-0.0264
-0.1007
-0.0106
-0.1276
-0.1007
0.004
-0.0018
-0.0018
0.004
-0.0018
0.0104

0.0718
0.0059
-0.0395
0.0013
0.0008
-0.0131
-0.0118
-0.0035
-0.0044
0.1203
0.0144
0.0058
-0.0844
-0.0443
0.0118
-0.0035
0.0017
-0.0044
-0.0182
0.0137
-0.0048
-0.0182
0.002
-0.0035
-0.0035
0.002
-0.0035
0.009

24

Table 5. Multivariate analysis of variance (MANOVA)
Model
Outbreak
Geo region
Outbreak*
geo region

Wilk’s Lambda
Pillai’s Trace
Value Prob>F Value Prob>F
F test: Prob>F=.5871
.3602 .0016* .7322 .0028*
.8602 .7885
.1428 .7761

Hotelling-Lawley
Value
Prob>F

Roy’s Max Root
Value
Prob>F

1.5194
.1591

1.3259
.1333

.0010*
.8014

.0004*
.4419

Table 6a. Analysis of molecular variance (AMOVA) in one group.
Variance among populations (Va) and variance within populations (Vb), percentages of
each variation (%) and associated F-statistics, with significance level (P-value).
Source
of Variance
Percentage
Fst
P-value
variance
components
of variation
Among
.3641 Va
44.32
.4432
<.0005*
populations
Within
.4574 Vb
55.68
populations
total
.8218
Table 6b. Analysis of molecular Variance (AMOVA) in two groups.
Variance among groups (Va), variance among populations within groups (Vb) and variance
within populations (Vc), and corresponding fixation indices Fct, Fst, Fsc, respectively, and
the associated F-statistics with p-values.
Source of
Variance
Percentage
Fixation
P-value
variance
components
of variation indices
Among groups
.1832 Va
19.97
.2000 Fct
<.0005*
Within groups
.2768 Vb
30.18
.3771 Fsc
<.0005*
among
populations
Within
.4574 Vc
49.86
.5014 Fst
<.0005*
populations
Table 6c. Analysis of molecular variance (AMOVA) in three groups.
Variance among groups (Va), variance among populations within groups (Vb) and variance
within populations (Vc), and corresponding fixation indices Fct, Fst, Fsc, respectively, and
the associated F-statistics with p-values.
Source
of Variance
Percentage Fixation
P-value
variance
components
of variation indices
Among groups
.1156 Va
20.64
.2064 Fct
<.0005*
Within groups
.0750 Vb
13.37
.1685 Fsc
<.0005*
among
populations
Within
.3696 Vc
65.99
.3401 Fst
<.0005*
populations
25

Table 6d. Analysis of molecular variance (AMOVA) in outbreak effect.
Variance among groups (Va), variance among populations within groups (Vb) and variance
within populations (Vc), and corresponding fixation indices Fct, Fst, Fsc, respectively, and
the associated F-statistics with p-values.
Source
variance

of Variance
components

Among groups
Within groups
among
populations

0 Va
.3702 Vb

Percentage
of
variation
0
45.44

Within
populations

.4574 Vc

56.14

P-value

0 Fct
.4473 Fsc

.6110
<.0005*

.4386 Fst

26

Fixation
indices

<.0005*

Table 7a. Pairwise Fsts among populations on outbreak effect.
* p-value<.05.
Pop Pre-outbreak
ID
1
2
3
4
5
6
1
0
2
0.1323 0
3
0.2907 0.0118 0
4
0.4643* 0.0899* 0.2696* 0
5
0.5172* 0.1769* 0.3422* 0.0006 0
6
0.4440* 0.4560* 0.4994* 0.7164* 0.7273* 0
7
0.3589 0.1500* 0.2647 0.125
0.1375 0.3039*
8
0.4607 0.1068 0.2613 -0.0511 -0.0588 0.7534*
9
0.4317* 0.0771 0.1652 -0.0441 0
0.5333*
10
0.5111* 0.2832* 0.0322 0.4768* 0.5111* 0.6026*
11
0.4707* 0.1215* 0.2513 -0.0053 0
0.6154*
12
0.4929* 0.1620* 0.3227 -0.0033 0
0.6667*
13
0.7803* 0.6770* 0.4126 0.8615* 0.8744* 0.8768*
14
0.4903 0.2926 0.0696 0.4463* 0.4842* 0.5485*
15
0.3924 0.0563 0.235
0.0157 0.0238 0.6318*
16
0.5263* 0.1761* 0.3421* 0
-0.0112 0.7902*
17
0.5172* 0.1769* 0.3422* 0.0006 0
0.7273*
18
0.4668* 0.1620* 0.3227 -0.0033 0
0.6078*
19
0.3979* 0.2201* 0.3479* 0.2537 0.2667 0.3137
20
0.2705 0.0774 0.2366 0.0736 0.0972 0.2912*
21
0.3831 0.3857* 0.4315* 0.6167* 0.6322* -0.0382
22
0.2606 0.0563 0.1988 0.1254 0.1712 0.3333
23
0.4762* 0.1396 0.2888 -0.0385 0.0045 0.6553*
24
0.5069* 0.1558* 0.3183* -0.0141 -0.0242 0.7793*
25
0.4852* 0.1331 0.2917 -0.0307 -0.0396 0.7672*
26
0.4583* 0.1319 0.2821 0.0058 0.0119 0.6436*
27
0.5263* 0.1761* 0.3421* 0
-0.0112 0.7902*
28
0.4012 0.093
0.2107 0.0107 0.0207 0.5125*

27

7

8

0
0.0866
0.0588
0.4217*
0.0765
0.1165
0.8034*
0.4153*
0.0894
0.15
0.1375
0.0294
-0.0701
-0.0383
0.2255
-0.1133
0.1071
0.1316
0.1108
0.098
0.15
0.073

0
-0.0588
0.4434*
-0.0588
-0.0588
0.8665*
0.4114*
-0.0244
0
-0.0588
-0.0588
0.25
0.04
0.6263
0.1333
-0.0511
0
0
-0.0403
0
-0.0403

Table 7a (cont’d)
Pre-outbreak
Pop
ID
9
10
1
2
3
4
5
6
7
8
9
0
10
0.3371* 0
11
-0.0526 0.4166*
12
0
0.4903*
13
0.7672* 0.3397*
14
0.3602* 0.0696
15
-0.012
0.4310*
16
-0.0112 0.5145*
17
0
0.5111*
18
0
0.4858*
19
0.1482 0.5003*
20
0.057
0.4072*
21
0.4560* 0.5401*
22
0.0454 0.4027*
23
-0.0521 0.4565*
24
-0.0242 0.4935*
25
-0.0396 0.4700*
26
-0.0086 0.4466*
27
-0.0112 0.5145*
28
-0.0199 0.3543*

During outbreak
11
12

13

14

15

16

0
0
0.8232*
0.4231*
-0.0011
-0.0112
0
0
0.1905
0.0778
0.5305*
0.0988
-0.0017
-0.0242
-0.0396
-0.0024
-0.0112
0.0097

0
0.1723
0.8335*
0.8881*
0.8744*
0.8558*
0.8481*
0.7843*
0.8464*
0.8092*
0.8453*
0.8818*
0.8747*
0.8415*
0.8881*
0.7769*

0
0.4141
0.4792*
0.4842*
0.4701*
0.4747*
0.4219
0.4973*
0.3747
0.4442*
0.4591*
0.4367*
0.4349*
0.4792*
0.3812*

0
0.0382
-0.0611
0.0083
0.1966
0.0601
0.5285*
0.0908
0.0043
0.0204
0
0.0013
0.0382
-0.007

0
-0.0112
-0.0112
0.3156
0.0966
0.6834*
0.2174*
0
0
0
0.0156
0
0.0156

0
0.8558*
0.4716*
-0.0555
-0.0112
0
0
0.2222
0.0864
0.5771*
0.1282
0.0006
-0.0242
-0.0396
0.003
-0.0112
0.0144

28

Table 7a (cont’d)
During outbreak
Pop
ID
17
18
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
18
0
0
19
0.2667 0.1026
20
0.0972 0.0212
21
0.6322* 0.5117*
22
0.1712 0.0286
23
0.0045 -0.0265
24
-0.0242 -0.0242
25
-0.0396 -0.0396
26
0.0119 0.003
27
-0.0112 -0.0112
28
0.0207 0.0144

19

20

21

22

23

24

0
-0.0336
0.2287
-0.0916
0.2127
0.2962
0.2746
0.2039
0.3156
0.1443

0
0.2237
-0.0837
0.0553
0.0805
0.062
0.0446
0.0966
0.0588

0
0.2122
0.5571*
0.6667*
0.6478*
0.5455*
0.6834*
0.4286*

0
0.1156
0.1928
0.1651
0.103
0.2174*
0.0505

0
-0.0141
-0.0307
0.0009
0
-0.0643

0
0
0
0
0

29

Table 7a (cont’d)
During outbreak
Pop
ID
17
18
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
18
0
0
19
0.2667 0.1026
20
0.0972 0.0212
21
0.6322* 0.5117*
22
0.1712 0.0286
23
0.0045 -0.0265
24
-0.0242 -0.0242
25
-0.0396 -0.0396
26
0.0119 0.003
27
-0.0112 -0.0112
28
0.0207 0.0144

19

20

21

22

23

24

0
-0.0336
0.2287
-0.0916
0.2127
0.2962
0.2746
0.2039
0.3156
0.1443

0
0.2237
-0.0837
0.0553
0.0805
0.062
0.0446
0.0966
0.0588

0
0.2122
0.5571*
0.6667*
0.6478*
0.5455*
0.6834*
0.4286*

0
0.1156
0.1928
0.1651
0.103
0.2174*
0.0505

0
-0.0141
-0.0307
0.0009
0
-0.0643

0
0
0
0
0

30

Table 7a (cont’d)
During outbreak
Pop
ID
25
26
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0
25
-0.0182 0
26
0
0.0156
27
-0.0182 0
28

27

28

0
0.0156

0

31

Table 7b. pairwise Fsts among populations on geography.
* p-value<.05
Pop Western Region
ID
1
6
7
16
17
1
0
2
0.4440* 0
3
0.3589 0.3039* 0
4
0.5263* 0.7902* 0.15
0
5
0.5172* 0.7273* 0.1375 -0.0112 0
6
0.4668* 0.6078* 0.0294 -0.0112 0
7
0.3979* 0.3137 -0.0701 0.3156 0.2667
8
0.2705 0.2912* -0.0383 0.0966 0.0972
9
0.3831 -0.0382 0.2255 0.6834* 0.6322*
10
0.2606 0.3333 -0.1133 0.2174* 0.1712
11
0.5172* 0.7273* 0.1375 -0.0112 0
12
0.4762* 0.6553* 0.1071 0
0.0045
13
0.5069* 0.7793* 0.1316 0
-0.0242
14
0.4852* 0.7672* 0.1108 0
-0.0396
15
0.4583* 0.6436* 0.098
0.0156 0.0119
16
0.5263* 0.7902* 0.15
0
-0.0112
17
0.4012 0.5125* 0.073
0.0156 0.0207
18
0.1323 0.4560* 0.1500* 0.1761* 0.1769*
19
0.2907 0.4994* 0.2647 0.3421* 0.3422*
20
0.4643* 0.7164* 0.125
0
0.0006
21
0.4607 0.7534* 0.0866 0
-0.0588
22
0.4317* 0.5333* 0.0588 -0.0112 0
23
0.5111* 0.6026* 0.4217* 0.5145* 0.5111*
24
0.4707* 0.6154* 0.0765 -0.0112 0
25
0.4929* 0.6667* 0.1165 -0.0112 0
26
0.7803* 0.8768* 0.8034* 0.8881* 0.8744*
27
0.4903 0.5485* 0.4153* 0.4792* 0.4842*
28
0.3924 0.6318* 0.0894 0.0382 -0.0611

32

18

19

20

0
0.1026
0.0212
0.5117*
0.0286
0
-0.0265
-0.0242
-0.0396
0.003
-0.0112
0.0144
0.1620*
0.3227
-0.0033
-0.0588
0
0.4858*
0
0
0.8558*
0.4701*
0.0083

0
-0.0336
0.2287
-0.0916
0.2667
0.2127
0.2962
0.2746
0.2039
0.3156
0.1443
0.2201*
0.3479*
0.2537
0.25
0.1482
0.5003*
0.1905
0.2222
0.8481*
0.4747*
0.1966

0
0.2237
-0.0837
0.0972
0.0553
0.0805
0.062
0.0446
0.0966
0.0588
0.0774
0.2366
0.0736
0.04
0.057
0.4072*
0.0778
0.0864
0.7843*
0.4219
0.0601

(Table 7b cont’d)
Western Region
Pop
ID
21
22
1
2
3
4
5
6
7
8
9
0
10
0.2122 0
11
0.6322* 0.1712
12
0.5571* 0.1156
13
0.6667* 0.1928
14
0.6478* 0.1651
15
0.5455* 0.103
16
0.6834* 0.2174*
17
0.4286* 0.0505
18
0.3857* 0.0563
19
0.4315* 0.1988
20
0.6167* 0.1254
21
0.6263 0.1333
22
0.4560* 0.0454
23
0.5401* 0.4027*
24
0.5305* 0.0988
25
0.5771* 0.1282
26
0.8464* 0.8092*
27
0.4973* 0.3747
28
0.5285* 0.0908

Central Region
5
23

24

25

26

27

0
0.0045
-0.0242
-0.0396
0.0119
-0.0112
0.0207
0.1769*
0.3422*
0.0006
-0.0588
0
0.5111*
0
0
0.8744*
0.4842*
0.0238

0
0
0
0
0
0.1558*
0.3183*
-0.0141
0
-0.0242
0.4935*
-0.0242
-0.0242
0.8818*
0.4591*
0.0204

0
-0.0182
0
-0.0182
0.1331
0.2917
-0.0307
0
-0.0396
0.4700*
-0.0396
-0.0396
0.8747*
0.4367*
0

0
0.0156
0
0.1319
0.2821
0.0058
-0.0403
-0.0086
0.4466*
-0.0024
0.003
0.8415*
0.4349*
0.0013

0
0.0156
0.1761*
0.3421*
0
0
-0.0112
0.5145*
-0.0112
-0.0112
0.8881*
0.4792*
0.0382

0
-0.0141
-0.0307
0.0009
0
-0.0643
0.1396
0.2888
-0.0385
-0.0511
-0.0521
0.4565*
-0.0017
0.0006
0.8453*
0.4442*
0.0043

33

Table 7b (cont’d)
Central
Pop
ID
28
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
18
0.093
19
0.2107
20
0.0107
21
-0.0403
22
-0.0199
23
0.3543*
24
0.0097
25
0.0144
26
0.7769*
27
0.3812*
28
-0.007

Eastern Region
2
3

4

8

9

10

11

0
0.0118
0.0899*
0.1068
0.0771
0.2832*
0.1215*
0.1620*
0.6770*
0.2926
0.0563

0
-0.0511
-0.0441
0.4768*
-0.0053
-0.0033
0.8615*
0.4463*
0.0157

0
-0.0588
0.4434*
-0.0588
-0.0588
0.8665*
0.4114*
-0.0244

0
0.3371*
-0.0526
0
0.7672*
0.3602*
-0.012

0
0.4166*
0.4903*
0.3397*
0.0696
0.4310*

0
0
0.8232*
0.4231*
-0.0011

0
0.2696*
0.2613
0.1652
0.0322
0.2513
0.3227
0.4126
0.0696
0.235

34

Table 7b (cont’d)
Eastern Region
Pop
ID
12
13
14
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
0
26
0.8558* 0
27
0.4716* 0.1723 0
28
-0.0555 0.8335* 0.4141

15

0

35

Table 8. Three Mantel tests.
Group A(all 28 populations), W(Western populations) and E(Eastern populations) with
correlation coefficient(r) and p-values. Permutation=5000, * indicated the significance level
(P< .05) between two matrices. E’ is without pop 11, 12 which were on the edge of
sampling area.
Group
ID

populations

Correlation
coefficient(r)

p-value

A

All 28

.1102

.1316

W

1,6,7,16-22

.5188

.0004*

C

5, 15, 23-28

.2131

.1910

E

2-4, 9-14

.1785

.1982

E’

2-4, 9-10,13-14

.4321

.0434*

36

Appendix B (Figures)

Figure 1 Haplotype network.
Each circle represents a haplotype. The square in the center represented the most common
haplotype. The sizes of circles indicated the frequencies of each haplotype, the larger the
more frequent. The line connecting circles and squares represent mutational steps, which is
nucleotide substitutions. The nodes represented hypothetical unsampled haplotypes, either
because they extinct or not sampled. The red solid bars under some sequences indicated
the sequences from pre-outbreak populations (1-10) and red dash bars indicated the
sequences from pre-outbreak populations and still maintained in outbreak populations. For
interpretation of the references to color in this and all other figures, the reader is referred
to the electronic version of this thesis.

37

Figure 2. The map of sampling localities

38

Figure 2. (cont’d). Each teardrop icon located the 28 populations we sampled on the map of western and southwestern US.
Population IDs were marked on top of each icons. Stars on tear-drop icon indicated pre-outbreak populations from 2001. The
solid and dashed curve lines and the oval in the center divided the sampling area into three geographic regions.

39

Figure 3 Principal component analysis.
Black, green and red represent three different geographic regions respectively, and solid
shapes and hollow shapes differentiate during and pre-outbreak populations respectively.

40

gene diversity (H)

Western side populations
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0

pop20
pop22
pop7
pop21
pop23

pop5

pop6

pop4
32

34

36
latitudes

38

40
y = 0.0931x - 2.9106
R² = 0.7126

gene diversity (H)

Eastern side populations
1

pop 10

0.8

pop 3

0.6
0.4

pop 14
pop 13

pop 9

0.2
pop 8

0

30

32

pop 4
34

36
latitudes

38
40
y = 0.1108x - 3.5249
R² = 0.762

Figure 4 Regression plot of haplotype diversity (H) against latitudes.
Western group included populations 4-7 and 20-23 and the Eastern group included
populations 3,4,8,9,10,13,14.

41

Fst/(1-Fst)

Western Group
4.0000
3.5000
3.0000
2.5000
2.0000
1.5000
1.0000
0.5000
0.0000
0

200
400
600
Geographic Distance (km)

800

pop 1
pop 6
pop 7
pop 16
pop 17
pop 18
pop 19
pop 20
pop 21
pop 22

Fst/(1-Fst)

Eastern Group
7.0000
6.0000
5.0000
4.0000
3.0000
2.0000
1.0000
0.0000

pop 2
pop 3
pop 4
pop 9
pop 10
pop 13
0

200
400
600
800
Geographic Distance (km)

1000

pop 14

Figure 5 Mantel tests.
The relationship between geographic and genetic distances was compared for individuals
in West and East geographic regions. P-values indicate a significant linear relationship
between the two matrices.

42

REFERENCES

43

REFERENCES
Aukema, B. H., Carroll, A. L., Zheng, Y., Zhu, J., Raffa, K. F., Dan Moore, R., Stahl, K. and Taylor,
S. W. (2008) Movement of outbreak populations of mountain pine beetle: influences of
spatiotemporal patterns and climate. Ecography 31: 348–358.
Breshears, D. D., N. S. Cobb, P. M. Price, C. D. Allen, R. G. Balice, W. H. Romme, J. H. Kastens, M.
L. Floyd, J. Belnap, J. J. Anderson, O. B. Myers, and C. W. Meyer (2005) Regional vegetation
die-off in response to global change-type-drought. PNAS 102: 15144-15148.
Byers, J. A. (2000) wind-aided dispersal of simulated bark beetles flying through forests.
Ecological modeling 125: 231-243.
Carroll, A.L.; Taylor, S.W.; Regniere, J.; Safranyik, L. (2004) Effects of climate change on range
expansion by the mountain pine beetle in British Columbia. The Bark Beetles, Fuels, and Fire
Bibliography. Page 195.
Chapuis, M.P., Lecoq, M., Michalakis, Y., Loiseau, A., Sword, G. A., Piry, S. and Estoup, A.
(2008) Do outbreaks affect genetic population structure? A worldwide survey in Locusta
migratoria, a pest plagued by microsatellite null alleles. Mol. Ecol. 17: 3640–3653.
Chapuis, M.P., Loiseau, A., Michalakis, Y., Lecoq, M., Franc, A. and Estoup, A. (2009)
Outbreaks, gene flow and effective population size in the migratory locust, Locusta
migratoria: a regional-scale comparative survey. Mol. Ecol. 18: 792–800.
Chen, H., and Walton, A. (2011) Mountain pine beetle dispersal: spatiotemporal patterns
and role in the spread and expansion of the present outbreak. Ecosphere 2:art66
Christiansen, E., Waring, R.H., Berryman, A.A. (1987) Resistance of conifers to bark beetle
attack: searching for general relationships. For. Ecol. Manage. 22: 89-106.
Clement, M., Posada, D. and Crandall, K.A.(2000) TCS: a computer program to estimate gene
genealogies. Mol. Ecol. 9, 1657-1659.
Cognato, A.I., Seybold, S.J. and Sperling, F.A.H. (1999) In complete barriers to mitochondrial
gene flow between pheromone races of the North American pine engraver, ips pini (Say).
Proc. R. Sco. Lond. B 266: 1843-1850.
Cognato, A.I. and Harlin, A.D. and Fisher, M.L. (2003) Genetic structure among pinyon pine
beetle populations (Scolytinae: Ips confusus). Environmental Entomology 32: 1262-1270.
Cognato, A. I., Gillette, N. E., Bolan os, R. C. and Sperling, F.A.H. (2005a) Mitochondrial
phylogeny of pine cone beetles (Scolytinae, Conophthorus) and their affiliation with
geographic area and host. Mol. Phylogenet. Evol. 36: 494-508.
Cognato, A. I., Sun, J.H., Anducho, M. and Owen, D. (2005b) Genetic variation and origin of

44

red turpentine beetles (Dendroctonus valens LeConte) introduced to the People’s Republic
of China. Agric. For. Entomol. 7: 87-94.
de la Giroday, H.M.C, Caroll, A.L., Lindgren, B.S. and Aukema, B.H. (2011) Incoming!
Association of landscape features with dispersing mountain pine beetle populations during
a range expansion event in western Canada. Landscape Ecology 26: 1097-1110.
Eager, T.J. (1999) Factors affecting the health of pinyon pine trees (Pinus edulis) in the
pinyon-juniper woodlands of western Colorado. Pages 397-399 In: Monsen, S.B., Richards,
S., Tausch, R.J., Miller, R. F., Goodrich, C., editors. Proc. Ecology and Management of PinyonJuniper Communities within the Interior West. USDA Forest Service, RMRS-P-9.
Eckert, C.G., Samis, K.E. and Lougheed, S.C. (2008) Genetic variation across species’
geographical ranges: the central-marginal hypothesis and beyond. Mol. Ecol. 17: 1170-1188.
Excoffier, L., Smouse, P.E. and Quattro, J.M. (1992) Analysis of molecular variance inferred
from metric distances among DNA haplotypes: application to human mitochondrial DNA
restriction data. Genetics 131: 479-491.
Excoffier, L., Laval, G., Schneider, S. (2005) Arlequin (Version 3.0): an integrated software
package for population genetics data analysis. Ecol Bioinf 1: 47-50.
Excoffier, L. (2008) Analysis of Population Subdivision, in Handbook of Statistical Genetics.
3rd Edition. John Wiley & Sons, Ltd, Chichester, UK.
Fonseca, D.M., Widdel, A.K., Hutchinson, M., Spichiger, S.E. and Kramer, L.D. (2010) Finescale spatial and temporal population genetics of Aedes japonicas, a new US mosquito,
reveal multiple introductions. Mol. Ecol. 19: 1559-1572.
Furniss, R.L., and Carolin, B.M. (1977) Western forest insects. U.S. Dep. Agric. For. Serv. Misc.
Publ. 1339.
Gara, R.I. and Vite, J.P. (1962) Studies on the flight patterns of bark beetles (Coleoptera:
Scolytidae) in second growth ponderosa pine forests. Contributions of Boyce Thompson
Institute 21: 275-290.
Halsey, D., Guyon, J., Knight, J. Wang, S. (2011) Nevada Aerial Detection Survey Damage
Areas. USDA. http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5358303.pdf
Holder, M., and P. O. Lewis. (2003) Phylogeny estimation: Traditional and Bayesian
approaches. Nat. Rev. Genet. 4: 275-284.
Hudson, R. R., Slatkin, M., and Maddison, W.P., (1992) Estimation of levels of gene flow from
DNA sequence data. Genetics 132: 583-589.
Ibrahim, K.M. (2001) Plague dynamics and population genetics of the desert locust: can
45

turnover during recession maintain population genetic structure? Mol. Ecol. 10: 581-591.
Ibrahim, K.M., Nichols, R.A. and Hewitt, G.M. (1996) Spatial patterns of genetic variation
generated by different forms of dispersal during range expansion. Heredity 77: 282–291
Ibrahim, K.M., Sourrouille, P. and Hewitt, G.M. (2000) Are recession populations of the
desert locust (Schistocerca gregaria) remnants of past swarms? Mol. Ecol. 9: 783-791.
Jactel, H., and Gaillard, J. (1991) A preliminary study of the dispersal potential of Ips
sexdentatus (Boern) (Col, Scolytidae) with an automatically recording flight mill. Journal of
Applied Entomology 112:138-145.
James, P.M.A., Coltman, D.W., Murray, B.W., Hamelin, R.C., Sperling, F.A.H. (2011) Spatial
Genetic Structure of a Symbiotic Beetle-Fungal System: Toward Multi-Taxa Integrated
Landscape Genetics. PLoS ONE 6(10): e25359.
Kelley, S.T., Mitton, J.B. and Paine, T.D. (1999) Strong differentiation in mitochondrial DNA
of Dendroctonus brevicomis (Coleoptera: Scolytidae) on different subspecies of ponderosa
pine. Ann. Entomol. Soc. Am. 92: 193- 197.
Kobayashi, T., Sakurai, T., Sakakibara, M. and Watanabe, T. (2011) Multiple origins of
outbreak populations of a native insect pest in an agro-ecosystem. Bulletin of Entomological
Research 101: 313-324.
Lanier, G. N. (1970) Biosystematics of North American Ips (Coleoptera: Scolytidae):
Hopping’s group IX. Can. Entomol. 102: 1139-1163.
Little, E.L., Jr. (1971) Atlas of United States trees: volume 1: conifers and important
hardwoods. U.S. Department of Agriculture Miscellaneous Publication 1146, 9 p., 200 maps.
Lowe, W.H. (2009) What drives long-distance dispersal? A test of theoretical predictions.
Ecology 90: 1456–1462.
Maddison, D.R. and Maddison, W.P. (2005) MacClade 4: Analysis of phylogeny and character
evolution. Version 4.08a. http://macclade.org.
Mantel, N. (1967) The detection of disease clustering and a generalized regression
approach. Cancer res 27: 209-220.
Menard, K.L., Cognato, A.I. (2007) Mitochondrial Haplotypic Diversity of Pine Cone Beetles
(Scolytinae: Conophthorus) Collected on Food Sources. Environmental Entomology 36(4):
962-966.
Michalakis, Y. and Excoffier, L. (1996) A generic estimation of population subdivision using
distances between alleles with special reference for microsatellite loci. Genetics. 142: 10611064.

46

Negron, J.F., Wilson, J.L. (2003) Attributes associated with probability of infestation by the
pinion ips, Ips confusus (Coleoptera: Scolytidae), in pinion pine, Pinus edulis. West. N. Am.
Nat. 63: 440-451.
Nei, M. (1987) Molecular evolutionary genetics. Columbia University Press, New York.
Raymond, M. and Rousset, F. (1995) An exact test for population differentiation. Evolution
49: 1280-1283.
Robertson, C., Nelson, T.A., Jelinski, D.E., Wulder M.A. and Boots, B. (2009) Spatial–temporal
analysis of species range expansion: the case of the mountain pine beetle, Dendroctonus
ponderosae. Journal of Biogeography 36(8): 1446–1458.
Robinet, C. and Roques, A. (2010) Direct impacts of recent climate warming on insect
populations. Integrative Zoology 5: 132–142.
Ronnas, C., Cassel-lundhagen, A., Battisti, A., Wallen, J. and Larsson, S. (2011) Limited
emigration from an outbreak of a forest pest insect. Mol. Ecol. 20: 4606–4617.
Rousset (2000) Genetic differentiation between individuals. Journal of Evolutionary. Biology
13: 58-62.
Safranyik, L., Carroll, A. L., Regniere, J., Langor, D.W., Riel, W.G., Shore, T.L. et al. (2010)
Potential for range expansion of mountain pine beetle into the boreal forest of North
America. Can. Entomol. 142: 415-4442.
Slatkin, M. (1995) A measure of population subdivision based on microsatellite allele
frequencies. Genetics 139: 457-462.
Sokal, R.R. and Rohlf, F.J. (1995) Biometry: The principles and practice of statistics in
biological research. 3rd edition. W.H. freman, New York.
Swofford, D.L. (2003) PAUP*: phylogenetic analysis using parsimony (* and other methods).
Version 4. Sinauer Associates, Sunderland, Massachusetts.
Stauffer, C., Lakatos, E. and Hewitt, G.M. (1999) Phylogeography and postglacial
colonization routes of Ips typographus L. (Coleoptera, Scolytidae). Mol. Ecol. 8: 763-773.
Tajima F (1993) Simple methods for testing molecular clock hypothesis. Genetics 135:599607.
Tamura, K. and Nei, M. (1993) Estimation of the number of nucleotide substitutions in the
control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Ecol. 10: 512526.
Tao, J., Chen, M., Zong, S.X. and Luo, Y.Q. (2012) Genetic structure in the Seabuckthorn

47

Carpenter Moth (Holcocerus hippophaecolus) in China: The role of outbreak events,
geographical and host factors. PLoS ONE 7: 1-7.
Tsui, C.K.M., Roe, A.D., El-Kassaby, Y.A., Rice, A.V., Alamouti, S.M., Sperling, F.A.H., Cooke,
J.E.K., Bohlmann, J. and Hamelin, R.C. (2012) Population structure and migration pattern of
a conifer pathogen, Grosmannia clavigera, as influenced by its symbiont, the mountain pine
beetle. Mol. Ecol. 21: 71–86.
USDA-Forest Service (2004) Forest insect and disease conditions in the southwestern
region, 2004. http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5238440.pdf
Vandergast, A.G., Gillespie, R.G and Roderick, G.K. (2004) Influence of volcanic activity on
the population genetic structure of Hawaiian Tetragnatha spiders: fragmentation, rapid
population growth and the potential for accelerated evolution. Mol. Ecol. 13: 1729-1743.
Weir BC, Cockerham CC (1984) Estimating F-statistics for the analysis of population
structure. Evolution 38: 1358-1370.
Williams, A. P., Allen, C. D., Millar, C. I., Swetnam, T. W., Michaelsen, J., Stilla, C. J. and Leavitt,
S. W. (2010) Forest responses to increasing aridity and warmth in the southwestern United
States. Proceedings of the National Academy of Sciences of the United States of America 107:
21289–21294.
Wood, D.L., Stark, R.W., Silverstein, R.M. and Rodin, J. O. (1967) Unique synergistic effects
produced by the principal sex attractant compounds of Ips confusus (Leconte) (Coleoptera:
Scolytidae). Nature 215: 206.
Wood, S.L. (1982) the bark and ambrosia beetles of North and Central America (Coleoptera:
Scolytidae), a taxonomic monograph. Great Basin Naturalist Memoirs 6: 1359.

48