SEARCH

SEARCH BY CITATION

Keywords:

  • ecotype;
  • gene flow;
  • habitat adaptation;
  • isolation-by-distance

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

Invasions by pest organisms are among the main challenges for sustainable crop protection. They pose a serious threat to crop production by introducing a highly unpredictable element to existing crop protection strategies. The western flower thrips Frankliniella occidentalis (Insecta, Thysanoptera) managed to invade ornamental greenhouses worldwide within < 25 years. To shed light on possible genetic and/or ecological factors that may have been responsible for this invasion success, we studied the population genetic structure of western flower thrips in its native range in western North America. Analysis of nucleotide sequence variation and variation at microsatellite loci revealed the existence of two habitat-specific phylogenetic lineages (ecotypes) with allopatric distribution. One lineage is associated with hot/dry climates, the second lineage is restricted to cool/moist climates. We speculate that the ecological niche segregation found in this study may be among the key factors determining the invasion potential of western flower thrips.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

Invasive species are of growing economic concern and often are recognized as a major threat to biodiversity (Everett, 2000; Morse & Hoddle, 2006). However, despite this awareness, the causes underlying the success and spread of invasive species often remain unknown. With the emergence of molecular methods, evidence is accumulating that the genetic composition of founder populations has profound impacts on their capacity to adapt to new environments and their success as invasive species. For example, it has been hypothesized that high genetic variability of invasive populations may be advantageous for their success (Williamson, 1996). On the other hand, a reduced level of genetic diversity has been observed in populations of successful invasive species compared to their native populations (e.g. Tsutsui et al., 2000). Several mechanisms could cause genetic bottlenecks in the introduced range; low number of colonizers, colonizers originating from a restricted source area and/or selection of favourable genotypes better adapted to the new habitat. Therefore, comparing the population genetic structure in the native and introduced range of an invasive species should reveal important insight into the biology and preconditions of an invasive organism. Furthermore, knowledge of the native origin of an introduced or invasive species is an important asset to inferring evolutionary history, predicting adaptive range and locating sources of invasions.

Thrips comprise a single insect order with about 5500 described species. While most thrips species remain inconspicuous, some show all the features that predispose them to be major pest species by causing direct feeding damage and spreading viral diseases to food, fibre and ornamental crops. In only three decades, the western flower thrips (Frankliniella occidentalis Perg.) has spread worldwide and today is the most economically important thrips species in greenhouse production. The original distribution of F. occidentalis is western North America, mainly west of the Rocky Mountains, from Mexico as far north as Alaska (Fig. 1a). Within this natural range it occupies very diverse ecosystems from sea level to subalpine altitudes and from wet to arid conditions. Also, while most thrips are host-plant specific, F. occidentalis is highly polyphagous feeding on more than 240 plant species in 62 different plant families (Tommasini & Maini, 1995). Furthermore, F. occidentalis is particularly variable in size and colour. While adult males are all pale, female colour forms range from almost white through yellowish to black. Bryan & Smith (1956) considered these colour forms to be different species, but they have now been synonymized (Nakahara, 1997). This ecological and morphological diversity raises the question of genetic variation and local adaptation by different populations (reviewed by Futuyma & Peterson, 1985).

image

Figure 1.  (a) Map showing the sampling sites for Frankliniella occidentalis. The dotted area in North America denotes the native range of the species, and the expanded map shows the sampling sites of native populations. See Table S1 for exact sites and further information on samples. Black and white colour-codes represent the two major habitat associations as identified in Fig. 3. (b) Mitochondrial DNA percentage of individuals in the hot/dry (white) and cool/moist (black) clades, and averaged co-ancestry coefficients for k = 2 and k = 3 from STRUCTURE analyses using variation at microsatellite loci.

Download figure to PowerPoint

The aim of this study was to establish the genetic diversity of native populations of the western flower thrips to better understand the processes involved in the invasion biology of this species. We study, for the first time, the genetic population structure of F. occidentalis within its native range using nucleotide variation of the mitochondrial cytochrome oxydase I (COI) gene and variation at nuclear microsatellite loci. We test the hypothesis that metapopulation structure can be explained with an isolation-by-distance model, which predicts that genetic distance among populations is correlated with geographical distance. Finally, by evaluating congruence of genetic findings with morphology and ecological characteristics we test an alternative hypothesis, that genetic relatedness is determined by habitat limited dispersal, and evaluate the possibility of distinct ecotypes.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

Data collection and species identification

To investigate the population genetic structure of Frankliniella occidentalis, thrips were collected at 12 localities from wild flowers throughout its native range in the western USA (Fig. 1a; Table S1). Collected specimens were first identified morphologically, and, because adult males of F. occidentalis are all pale, only adult females were scored for variability in body coloration and subjected to subsequent genetic analyses. We followed the description of Bryan & Smith (1956) and divided the colour forms into three categories; 1 = pale (white or yellow throughout), 2 = intermediate/bicoloured (orange thorax and brown abdomen), 3 = dark (brown or black throughout).

Total genomic DNA was extracted from single female thrips using the GenElute DNA extraction kit (SIGMA) following the manufacturer’s specifications, and a fragment of the mitochondrial COI gene was amplified via a standard PCR using the primers C1-J-1751 and C1-N-2329 (Simon et al., 1994). DNA sequences were generated using the same two primers on an ABI 3100 automated sequencer (Applied Biosystems, Foster City, CA, USA). All DNA was sequenced in both directions and aligned with the multiple sequence editor clustal X (Thomson et al., 1997). Six microsatellites were specifically developed for F. occidentalis population studies and amplified as described earlier (Brunner & Frey, 2004).

Thrips are minute insects that often require microscopic preparation of specimens for correct species identification by specialists. Assessing nucleotide variation at the COI gene has already proven to be an ideal marker for species differentiation in thrips and a valuable alternative or addition to traditional morphological methods (Brunner et al., 2002). Thus, to avoid analysing a potentially heterogeneous data set of morphological sibling species, we first tested all generated sequences using our well-established molecular thrips key (see Brunner et al., 2002 for a detailed description). In short, all COI sequences from other thrips species were retrieved from GenBank and entered into a template file also containing all known F. occidentalis haplotypes and analysed using PAUP* 4.0b10 (Swofford, 2002). Most parsimonious topologies were recovered by heuristic search with simple addition of haplotypes. When a new sequence clustered within the F. occidentalis haplotype clade with maximal statistical significance (i.e. 100% of 1000 bootstrap replicates), it was assigned to this species and added to the data set.

Colour and habitat differentiation

For each population a data matrix was compiled based on the body colour coding and the habitat parameters. To assess the correlation between habitat and population structure, we compiled information on climatic and vegetation parameters for each sampling location. We used information available at the following sources: USDA-Zones (US Department of Agriculture), range of average annual minimum temperatures; Plant Heat-Zone Map (American Horticultural Society), average annual number of days above 86 °F/30 °C; and several parameters from the United States Department of Commerce; National Oceanic and Atmospheric Administration (NOAA). The parameters provided on the NOAA climatic maps were scored as follows. Climatic parameters are presented as colour zones on the NOAA maps. Each colour represents a specified range/category of the parameter. For example, ‘mean total precipitation’ is illustrated by nine colours/categories (A–I). For the habitat matrix, category A (< 5.01 inches) was scored as ‘1’, category B (5.01–12.0 inches) was scored as ‘2’ and so forth. A summary of scores for all used parameters and further details on samples are listed in Table S1. From the total habitat matrix (see Table S1), a principle component analysis (PCA) was conducted using the program XLSTAT (Addinsoft, New York, NY, USA) to determine the factors influencing the population structure. The colour and habitat matrices were also compared against geographical and genetic distances between populations estimated as pairwise linearized FST (Rousset, 1997). Significance of covariation between matrices was tested by two-way Mantel tests (Douglas & Endler, 1982) using the program package NTSYS-pc 2.10e (Rohlf, 1997).

Mitochondrial COI analysis

All population genetic parameters such as nucleotide diversity (π) and haplotype diversity (h) were calculated using the program DNASP v. 4.10 (Rozas et al., 2003). An amova implemented in ARLEQUIN v3.01 (Excoffier et al., 2005) was used to test the hierarchic genetic structure of the populations. Here, populations were a priori pooled into two groups according to major habitats identified in the previous correspondence analysis (see earlier and Figs 1a and 2).

image

Figure 2.  Principle component analysis of habitat parameters, suggesting that populations cluster into a hot/dry group (HD; white) and a cool/moist (CM; black) group. Abbreviations of populations and parameters as in Table S1.

Download figure to PowerPoint

Phylogenetic relationships among all mitochondrial haplotypes were reconstructed with the maximum likelihood approach implemented in Genetic Algorithm for Maximum Likelihood phylogeny inference(GAML) (Lewis, 1998). We used this program because it allows for a fast estimation of phylogenetic trees based on the ML approach compared to most other programs when many haplotypes are sampled. Furthermore, GAML implements the Hasegawa, Kishino and Yano (Hasegawa et al., 1985) nucleotide substitution model. Preliminary analysis based on Likelihood Ratio Tests (LRTs) implemented in MODELTEST v3.7 (Posada & Crandall, 1998) in conjunction with PAUP* suggested the HKY substitution model with a gamma distribution of rate variation among sites provided the best fit to the F. occidentalis data set. Sequences from Echinothrips americanus, Thrips palmi and Anaphothrips obscurus (Brunner et al., 2002) were included for outgroup comparison. Node supports were calculated out of 1000 nonparametric bootstrap replications.

Microsatellite analysis

Population genetic parameters were calculated using the programs GenAlEx (Peakall & Smouse, 2006) and FSTAT (Goudet, 1995). We used the program STRUCTURE (Pritchard et al., 2000) to estimate population assignment without a priori assumptions of population subdivision. Under the admixture model, structure estimates co-ancestry coefficients for individuals in each of k populations. We ran 10 replicates with a burn-in of 5.0 × 105 followed by 5.0 × 106 iterations for each value of k, ranging from 1 to 12. The most likely value of k was estimated using the Δk method (Evanno et al., 2005). Genetic population differentiation between habitat types was also visualized over multilocus genotypes by applying an individual-based multivariate factorial correspondence analysis implemented in GENETIX ver. 4.04 (Belkhir et al.; http://www.university-montp2.fr/~genetix/genetix. htm).

Migration rates among populations were estimated using two different approaches. First, long-term estimates were obtained by calculating Slatkin’s linearized FST and resultant number of migrants (M values) for each pair of populations. As recommended by Slatkin (1993), isolation by distance was tested by plotting M values against geographical distance. Secondly, recent migration rates were estimated using BAYESASS 1.3, which applies a Bayesian approach to multilocus genotypes (Wilson & Rannala, 2003). 5 000 000 MCMC iterations were run, with a burn-in of 1 000 000 iterations and a sampling frequency of 2000.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

Habitat parameters were analysed with a PCA (Fig. 2). An advantage of using PCA is the ability to characterize the habitats by looking at the influence of each parameter on the principal components (i.e. ‘scores’ in Table S1). The F1 axes is mainly determined by parameters related to temperature. Populations from warm habitats cluster to the left, whereas populations from cooler habitats cluster to the right of the F1 zero-line. The F2 axis is mainly determined by parameters related to precipitation. Populations from dry habitats cluster below, whereas populations from habitats with more precipitation cluster above the F2 zero-line. Populations SEV, NEP, WIN and RIT formed a distinct group in the PCA (Fig. 2). Given its position in the lower left of the diagram, this group can be associated with a hot/dry (HD) habitat. For convenience, we will refer to the other populations as the cool/moist (CM)-associated clade, although this is strictly speaking only true for populations KAL, SIL, COR and BOD.

A total of 192 adult female F. occidentalis were unambiguously identified using morphological characters and the genetic assignment described earlier. Sequencing of the COI gene for 112 individuals yielded a 433 bp long fragment. A total of 43 nucleotide positions (10%) were polymorphic, and these defined 30 distinct haplotypes (Tables 1 and S2; GenBank accessions GU372378-GU372407).

Table 1.   Summary of population genetic parameters (standard deviation in parenthesis) describing the native F. occidentalis populations pooled according to the two major climatic divisions (Figs 1a and 2).
PopulationNCOI nucleotide sequence variationNMicrosatellite variation
HaplotypeshπSΘGene diversityAllelic richnessuHeHWE
  1. h, haplotypic diversity; π, nucleotide diversity; S, polymorphic sites; Θ, mutation parameter used to compare effective population sizes (Θ = 2 Nμ).

HD (hot/dry)3950.563 (0.060)0.002 (0.001)60.003 (0.001)640.833 (0.062)16.800 (4.549)0.872 (0.024)< 0.001
CM (cool/moist)73280.894 (0.023)0.015 (0.002)410.020 (0.003)1280.873 (0.038)19.359 (6.611)0.838 (0.014)< 0.001
TOTAL112300.897 (0.015)0.024 (0.001)430.019 (0.003)1920.853 (0.053)22.204 (7.236)0.855 (0.014)< 0.001

Standard diversity indices of both nucleotide sequence data and microsatellite variation are summarized in Table 1. At face value, diversity values were consistently lower for populations associated with the HD habitat. amova (Table 2) corroborated the previous finding of two habitat-associated groups and attributed 74% of total genetic variation among these groups, suggesting significant genetic structuring on larger geographical scales. Only six per cent of total variation explained differentiation among populations within groups, suggesting substantial gene flow among habitat-associated populations. The remaining 20% was attributed to variation within populations.

Table 2.   Results of hierarchical analysis of genetic variance (amova). Populations were a priori assigned to the two major habitats identified in Fig. 2.
Source of variationVariance componentsPercentage variationFixation indicesProbabilities
Among groups6.52 Va74.10FCT: 0.741< 0.001
Among populations within groups0.52 Vb5.91FSC: 0.228< 0.001
Within populations1.76 Vc19.99FST: 0.783< 0.001

The phylogenetic ML tree revealed that haplotypes clustered into two phylogenetic lineages with good bootstrap support (Fig. 3). Sampling sites of F. occidentalis that clustered together in one of the two major lineages were disjunct. Sampling sites were either located along the Pacific coast (i.e. western slope of the Sierra Nevada Mountains and Cascade Range, respectively) or at the opposite side of the natural distribution range along the western slope of the Rocky Mountains. The region in between (represented by sampling sites SEV, NEP, WIN and RIT) formed the other phylogenetic lineage. All but one population were homogeneous for one lineage. The exception being the mixed population BOD, which in addition also contained haplotypes H_01 and H_02 usually associated with the other lineage (Fig. 3; Table S2). Most notably, populations SEV, NEP, WIN and RIT, which together formed a distinct phylogenetic clade also clustered together in the PCA (Fig. 2), forming a group associated with a hot/dry (HD) habitat.

image

Figure 3.  Maximum likelihood tree of the individual mitochondrial COI nucleotide sequences. Abbreviations of sampling localities as shown in Fig. 1a, and described in Table S1. Node supports calculated out of 1000 nonparametric bootstrap replications are given for the two major clades. The arrow is pointing to the outgroup species Echinothrips americanus, Thrips palmi and Anaphothrips obscurus.

Download figure to PowerPoint

Microsatellite analyses of population structure based on all 192 individuals suggested that the data set was likely composed of two to three populations (Figs 1b and 4a). For Δk = 2, populations clustered according to the two identified habitats. For Δk = 3, the analysis identified BOD and RIO as forming a third distinct population (Fig. 1b). Similarly, the correspondence plot (Fig. 4b) showed samples collected from HD habitats almost completely separated from CM samples on axis 1 (95% of variance).

image

Figure 4.  (a) Δk values from STRUCTURE analyses are shown for k ranging between 2 and 12, for the total number of sampled populations. (b) Multilocus genotypes for depicted along the first three axes of a multivariate factorial correspondence analysis. White represent genotypes collected from hot/dry (HD) habitats, and black dots are genotypes from cool/moist (CM) habitats.

Download figure to PowerPoint

Among population structure was further investigated by statistically contrasting the genetic findings against matrices of body colour differentiation, geographical distance and habitats. Results of the Mantel tests are summarized in Table 3. The only significant correlation was between genetic distance and habitat. Additionally, overall lack of isolation by distance was graphically illustrated by plotting number of migrants vs. geographical distance among populations (Fig. 5). In accordance with the amova results of high among-group differentiation but low among population within group differentiation, number of migrants between sampling sites were significantly higher for pairs associated with the same habitat than for pairs from different habitats. Similarly, Bayesian estimates of recent directional migration rates were nonsignificantly different from zero among populations associated with different habitats. Migration rates from HD into CM habitats were 0.005 (CI; 0.000, 0.017), and 0.007 (CI; 0.000, 0.026) from CM into HD, suggesting no/marginal recent gene flow between habitats.

Table 3.   Matrix comparisons among 12 native F. occidentalis populations using classical two-way Mantel tests. Significance of matrix correlation (P) is computed out of 1000 random permutations. See text and Table S1 for more details on colour morphs and habitat information.
Compared matricesMatrix correlation (r)Probability (P)
Genetic- vs. geographical distance0.00170.3986
Genetic distance vs. colour morph−0.12780.1568
Genetic distance vs. habitat0.49170.0050
Habitat vs. colour morph0.12170.1389
image

Figure 5.  Plot showing the pairwise population correlation of genetic distance (expressed as number of migrants, M) and geographical distance, suggesting no isolation by distance.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

The use of molecular genetic techniques, particularly analysis of mitochondrial DNA, has substantially contributed to a growing appreciation of cryptic genetic differentiation. These molecular approaches are especially helpful in difficult groups such as thrips, which often show a mosaic of diverse ecological traits superimposed onto an often highly variable morphology (Crespi et al., 1998; Brunner et al., 2004). Here, we assessed nucleotide variation at the COI gene and microsatellite variation to investigate the population genetic structure of the western flower thrips throughout its native range. It is worth noting that we only used females for the morphological and genetic analyses. However, we only assumed that males and females have the same dispersal capacity. The implications of this assumption (or violations of it) should be kept in mind with respect to the nuclear microsatellite data vs. the maternally inherited mtDNA data set.

Thrips are tiny insects. Thus, they are only weak active flyers but they are distributed by passive wind dispersal over long distances as an important component of aerial zooplankton (Mound, 1983). Furthermore, suitable habitats for thrips (e.g. food plants) are distributed patchily often over a wide range, thereby reflecting an island model of distribution. Wallace (1880) postulated that populations on islands that are closer together tend to be more similar to each other. Recent genetic studies on various organisms are in support of this idea and showed that this pattern is caused by distance-limited dispersal resulting in an isolation-by-distance pattern (e.g. beetles, Finston & Peck, 1995; Drosophila, DeSalle, 1995).

Our most significant finding was that F. occidentalis populations did not show any isolation by distance pattern. Instead, genetic relatedness among populations was significantly related with habitat. For example, thrips populations sampled from opposite ends of their west-east range of distribution with similar habitats, such as Monte Rio CA and Rock Springs WY, are genetically similar. Geographically more proximate populations that differ in habitat, such as Silverdale WA and Ritzville WA, belong to two distinct genetic clusters that can be attributed to cool/moist or hot/dry habitats, respectively. The congruence of population genetic structure and major habitat divisions is a prediction of the natal habitat-biased dispersal hypothesis (Haughland & Larsen, 2004; Sacks et al., 2005). In vertebrates, for example, several studies have shown that individuals tend to disperse preferentially to habitats similar to their natal home range (e.g. Vogl et al., 2002; Sacks et al., 2004). A logic corollary of this behaviour is less migrants and gene flow between habitat types relative to that within habitat types. This is exactly what we observed when comparing thrips populations from cool/moist vs. hot/dry habitats (Fig. 5).

Habitat choice is generally regarded as an important factor in speciation in insects (Via, 1999). Nevertheless, given that thrips are only weak active flyers, it is safe to assume that individuals are randomly wind-dispersed and cannot actively select a preferred habitat over long distances. Therefore, we are in favour of an alternative explanation and hypothesize that barriers to gene flow other than active habitat choice, such as selection against migrants or hybrid sterility exist. Molecular methods have previously detected cryptic differentiation in Thrips tabaci that has been associated with host-plant adaptation (Brunner et al., 2004). Indeed, reciprocal laboratory host-plant experiments (Chatzivassiliou, 2002; Chatzivassiliou et al., 2002) suggested strong host-specific (physiological) adaptations in this species. While T. tabaci populations from both host plants apparently thrived on leek, those originally collected from leek failed to survive on tobacco. So far, similar studies that could further clarify the genetic patterns observed in the native range of the western flower thrips are lacking.

The divergence in habitat use of western flower thrips can be viewed as a case of incipient cryptic adaptive radiation, as we found no correlation between morphology and genetics. There is indeed growing evidence that temperature rather than genetics is a key factor in colour variation (L.A. Mound, pers. comm.) as has been shown in T. tabaci which produces darker forms in cooler conditions (Murai & Toda, 2002).

However, the topic remains puzzling as the role of different colour forms responsible for the global invasion success of the western flower thrips is still discussed. For example, the virtual absence of dark forms in F. occidentalis in European greenhouses is, as yet, unexplained. It is also worth mentioning, that the colour forms of Frankliniella schultzei differ in their transmission of tospoviruses (Wijkamp et al., 1995).

In conclusion, molecular evidence presented in this study strongly suggests F. occidentalis forming two distinct, well-supported clades within its native range. Do these two clades represent cryptic species between which gene flow has ceased? Or are they ‘ecotypes’ (e.g., lineages with partial reproductive isolation as a consequence of local adaptation)? Genetic differentiation between the two clades is FST = 0.24. Although estimates of FST alone cannot be used to conclude whether all gene flow between two groups has ceased (Whitlock & McCauley, 1999), it is useful to put them into the context of similar estimates for other groups. For example, FST values for F. occidentalis clades equate with estimates for host races of the gall-moth Gnorimoschema gallaesolidaginis (0.16; Nason et al., 2002) and the pea aphid Acyrthosiphon pisum (0.21; Via, 1999), respectively. In contrast, estimates obtained for sympatric species of Gryllus crickets (Harrison, 1979), or cryptic species in Thrips (Brunner et al., 2004) are 0.92, and 0.91, respectively. Similarly, average nucleotide sequence divergence between the two habitat-associated clades is 3.8 ± 0.8%. This is substantially lower than those detected among morphologically distinguishable thrips species (range 16–27.5% (Brunner et al., 2002), or host-specific, morphologically indistinguishable sibling species of Australian gall-forming thrips (range 8–16% (Crespi et al., 1998). Thus, differentiation between F. occidentalis clades clearly falls within the range of typical ecotypes within a given taxon.

This study provided novel insight into the population structure of a notorious invasive species. The ecological niche adaptation revealed in this study may be among the key factors determining the astonishing invasion potential of western flower thrips. We speculate that under the novel conditions of the introduced range (i.e. mainly greenhouse conditions), invaders originating from a restricted source area and/or possessing pre-adapted genotypes have a selective advantage. Nevertheless, future analysis will have to contrast the genetic structure of native and introduced F. occidentalis populations on a global scale.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

We thank B. Frey and F. Schwaller for providing invaluable help in the lab. Personal comments of L.A. Mound and B. Duffy during the preparation of the manuscript were highly appreciated. Constructive comments of two anonymous reviewers improved the manuscript. We are also grateful to the American Horticultural Society for permission to use their climatic data. This study was supported by Swiss National Science Foundation Grant 3100-064045.00.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information
  • Brunner, P.C. & Frey, J.E. 2004. Isolation and characterization of six polymorphic microsatellite loci in the western flower thrips Frankliniella occidentalis (Insecta, Thysanoptera). Mol. Ecol. Notes 4: 599601.
  • Brunner, P.C., Fleming, C. & Frey, J.E. 2002. A molecular identification key for economically important thrips species (Thysanoptera: Thripidae) using direct sequencing and a PCR-RFLP-based approach. Agr. Forest Entomol. 4: 127136.
  • Brunner, P.C., Chatzivassiliou, E.K., Katis, N.I. & Frey, J.E. 2004. Host-associated genetic differentiation in Thrips tabaci (Insecta; Thysanoptera), as determined from mtDNA sequence data. Heredity 93: 364370.
  • Bryan, D.E. & Smith, R.F. 1956. The Frankliniella occidentalis (Pergande) complex in California. Univ. Calif. Public Entomol. 10: 359410.
  • Chatzivassiliou, E.K. 2002. Thrips tabaci: an ambiguous vector of TSWV in perspective. Thrips and Tospoviruses. In: 7th International Symposium on Thysanop (R.Marullo & L.A.Mound, eds), pp. 6975. Australian National Insect Collection, Canberra.
  • Chatzivassiliou, E.K., Peters, D. & Katis, N.I. 2002. The efficiency by which Thrips tabaci populations transmit tomato spotted wilt virus depends on their host preference and reproductive strategy. Phytopathology 92: 603609.
  • Crespi, B.J., Carmean, D.A., Mound, L.A., Worobey, M. & Morris, D. 1998. Phylogenetics of social behavior in Australian gall-forming thrips: evidence from mitochondrial DNA sequences, adult morphology and behavior, and gall morphology. Mol. Phylogenet. Evol. 9: 163180.
  • DeSalle, R. 1995. Molecular approaches to biogeographic analysis of Hawaiian Drosphilidae. In: Hawaiian Biogeography: Evolution on a Hot Spot Archipelago (L.Wagner & V.A.Funk, eds), pp. 7289. Smithonian Institute Press, Washington, DC.
  • Douglas, M.R. & Endler, J.A. 1982. Quantitative matrix comparisons in ecological and evolutionary investigations. J. Theor. Biol. 99: 777795.
  • Evanno, G., Regnaut, S. & Goudet, J. 2005. Detecting the number of clusters of individuals using the software structure: a simulation study. Mol. Ecol. 14: 26112620.
  • Everett, R.A. 2000. Patterns of pathways of biological invasions. Trends Ecol. Evol. 15: 177178.
  • Excoffier, L., Laval, G. & Schneider, S. 2005. Arlequin ver. 3.0: an integrated software package for population genetics data analysis. Evol. Bioinform. Online, 1: 4750.
  • Finston, T.L. & Peck, S.B. 1995. Population structure and gene flow in Stomion: a species swarm of flightless beetles of the Galàpagos Islands. Heredity 75: 390397.
  • Futuyma, D.J. & Peterson, S.C. 1985. Genetic variation in the use of resources by insects. Annu. Rev. Entomol. 30: 217238.
  • Goudet, J. 1995. FSTAT (version 1.2): a computer program to calculate F-statistics. J. Hered. 86: 485486.
  • Harrison, R.G. 1979. Speciation in North American field crickets: evidence from electrophoretic comparisons. Evolution 33: 10091023.
  • Hasegawa, M., Kishino, H. & Yano, T. 1985. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol. 22: 160174.
  • Haughland, D.L. & Larsen, K.W. 2004. Ecology of North American red squirrels across contrasting habitats: relating natal dispersal to habitat. J. Mammal. 85: 225236.
  • Lewis, P.O. 1998. A genetic algorithm for maximum-likelihood phylogeny inference using nucleotide sequence data. Mol. Biol. Evol. 15: 277283.
  • Morse, J.G. & Hoddle, M.S. 2006. Invasion biology of thrips. Annu. Rev. Entomol. 51: 6789.
  • Mound, L.A. 1983. Natural and disrupted patterns of geographical distribution in Thysanoptera (Insecta). J. Biogeogr. 10: 119133.
  • Murai, T. & Toda, S. 2002. Variation of Thrips tabaci in colour and size. In: Thrips and Tospoviruses: Proceedings of the 7th International Symposium on Thysanoptera (R.Marullo & L.A.Mound, eds), pp. 377378. Australian National Insect Collection, Canberra.
  • Nakahara, S. 1997. Annotated list of the Frankliniella species of the world (Thysanoptera: Thripidae). Contrib. Entomol. Internat. 2: 355389.
  • Nason, J.D., Heard, S.B. & Williams, F.R. 2002. Host-associated genetic differentiated in the goldenrod elliptical-gall moth, Gnorimoschema gallaesolidaginis (Lepidoptera: Gelechiidae). Evolution 56: 14751488.
  • Peakall, R. & Smouse, P.E. 2006. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol. Ecol. Notes 6: 288295.
  • Posada, D. & Crandall, K.A. 1998. Modeltest: Testing the model of DNA substitution. Bioinformatics 14: 817818.
  • Pritchard, J.K., Stephens, M. & Donnelly, P. 2000. Inference of population structure using multilocus genotype data. Genetics 155: 945959.
  • Rohlf, F.J. 1997. NTSYS-pc 2.10e: Numerical Taxonomy and Multivariate Analysis System. Applied Biostatistics Inc., Setauket, NY.
  • Rousset, F. 1997. Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics 145: 12191228.
  • Rozas, J., Sanchez-DelBarrio, J.C., Messeguer, X. & Rozas, R. 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 24962497.
  • Sacks, B.N., Brown, S.K. & Ernest, H.B. 2004. Population structure of California coyotes corresponds to habitat-specific breaks and illuminates species history. Mol. Ecol. 13: 12651275.
  • Sacks, B.N., Mitchell, B.R., Williams, C.L. & Ernest, H.B. 2005. Movements and social structure along a cryptic population genetic subdivision in the coyote. Mol. Ecol. 14: 12411249.
  • Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Floors, P. 1994. Evolution, weighing, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. Am. 87: 651701.
  • Slatkin, M. 1993. Isolation by distance in equilibrium and non-equilibrium populations. Evolution 47: 264279.
  • Swofford, D.L. 2002. PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods). Version 4.0b10. Sinauer, Sunderland, MA.
  • Thomson, J.D., Gibons, 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: 48764882.
  • Tommasini, M.G. & Maini, S. 1995. Frankliniella occidentalis and other thrips harmful to vegetable and ornamental crops in Europe. In: Biological Control of Thrips Pests (A.J.M.Loomans, J.C.Van Lenteren, M.G.Tommasini, S.Maini & J.Riudavets, eds), pp. 142. Wageningen Agricultural University Papers, Wageningen, The Netherlands.
  • Tsutsui, N.D., Suarez, A.V., Holway, D.A. & Case, T.J. 2000. Reduced genetic diversity and the success of an invasive species. Proc. Natl Acad. Sci. USA 97: 59485953.
  • Via, S. 1999. Reproductive isolation between sympatric races of pea aphids. I. Gene flow restriction and habitat choice. Evolution, 53: 14461457.
  • Vogl, W., Taborsky, M., Teuschl, Y. & Honza, M. 2002. Cuckoo females preferentially use specific habitats when searching for host nests. Anim. Behav. 64: 843850.
  • Wallace, A.R. 1880. Island Life. Macmillan, London.
  • Whitlock, M.C. & McCauley, D.E. 1999. Indirect measures of gene flow and migration: FST ≠ 1/(4Nm). Heredity 82: 117125.
  • Wijkamp, I., Almarza, N., Goldbach, R. & Peters, D. 1995. Distinct levels of specificity in thrips transmission of tospoviruses. Phytopathology 85: 10691074.
  • Williamson, M.H. 1996. Biological Invasions. Chapman & Hall, London.
  • Wilson, G.A. & Rannala, B. 2003. Bayesian inference of recent migration rates using multilocus genotypes. Genetics 163: 11771191.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. References
  9. Supporting Information

Table S1 Summary of habitat parameters and body color information scored for 12 native populations of western flower thrips Frankliniella occidentalis.

Table S2 Variable positions in the 433 bp fragment of the COI gene defining 30 haplotypes and their distribution across 12 native populations of Frankliniella occidentalis.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JEB_1946_sm_TableS1-S2.doc145KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.