Intraspecific Mitochondrial DNA Variation and Historical Biogeography ...

ECOLOGY AND POPULATION BIOLOGY

Intraspecific Mitochondrial DNA Variation and Historical Biogeography of the Eastern Lubber Grasshopper, Romalea microptera SERAP MUTUN1

AND

DAVID W. BORST

Department of Biological Sciences, Illinois State University Normal, IL 61790

Ann. Entomol. Soc. Am. 97(4): 681Ð696 (2004)

ABSTRACT The genetic and geographical variation, gene ßow, and the historical biogeography of the eastern lubber grasshopper, Romalea microptera (⫽guttata) (Houttuyn), were examined by sequencing a 420-bp region of the mitochondrial cytochrome b gene. Individuals (168) were collected from 12 sites in the southern United States that covered most of the range reported for this species. These populations contained 49 mitochondrial DNA haplotypes. A high level of genetic diversity was observed in these populations (3.8%), most of which was due to variation within populations. The highest genetic variation was found in a northern Florida population (collected at St. Marks, FL), and the lowest was found in a southern Florida population (Copeland, FL). Estimates of historical and current population sizes suggest that most of the lubber populations drastically declined in size at some point in the past. In contrast to previous studies on several other species in this region, phylogenetic analyses (PAUP) and haplotype age phylogenies (PHYLIP) showed no major geographic structure. These observations suggest that the distribution of this species in the past may have been homogeneous, rather than the patchy distribution that is currently observed. Alternatively, this result may reßect the absence of long-term barriers for the dispersal of this species. Either of these might have contributed to the lack of a genetic structure divided geographically into east-west groupings that is seen in other species from this region. KEY WORDS eastern lubber grasshopper, mtDNA, intraspeciÞc variation, genetic variation, gene ßow

THE GEOGRAPHIC RANGE AND the genetic population structure of a species are determined by a combination of current and historical factors (Cox and Moore 1980). If these factors can be described and their effects distinguished, it may be possible to construct a detailed evolutionary history of a species. However, it is often difÞcult to distinguish the individual importance of each factor that has played a role in determining the current distribution of a species, because these factors have considerable overlap in their effects on species distribution (Myers and Giller 1990). In addition, the relative importance of historical versus current factors is inßuenced by the dispersal ability of the species. The population structure of a species with restricted dispersal ability is determined mostly by historical factors, because there is limited gene ßow between populations (Bermingham and Avise 1986). These species tend to show a more heterogeneous genetic structure that contains population subdivisions across their range (Rhodes et al. 1996). In contrast, species with high dispersal ability have higher rates of gene ßow between populations, resulting in a more homogeneous genetic structure across their 1 Current address: Abant Izzet Baysal University, Fen-Ed. Fak. Biyoloji Bol., 14280 Bolu, Turkey. E-mail: [email protected].

range. Therefore, species with lower mobility are considered to be more suitable as candidates for historical biogeographic studies. The Pleistocene glacial cycles were a major historical factor that determined the contemporary distribution of North American organisms (Reeb and Avise 1990). Although the glaciers never advanced beyond the middle latitudes of the United States, the climatic ßuctuations associated with these events had considerable effects upon the biota throughout the southern United States (Williams et al. 1993). This seems to have been particularly true for temperate species. During the periods of glaciation, the range of these species would have been pushed into the southern region of the United States and onto areas of the continental shelf exposed by the lower sea levels. During interglacial periods, temperate species would have expanded their range to the north and retreated from the continental shelf as the sea level rose (Nilsson 1983). Changes in temperature and precipitation that accompanied the glacial cycles would have further inßuenced the genetic structure of populations by changing the distribution of suitable habitats. One way to investigate the effect of historical events on the genetic structure of species is to examine the geographic distribution of mitochondrial DNA

0013-8746/04/0681Ð0696$04.00/0 䉷 2004 Entomological Society of America

682

ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA

Vol. 97, no. 4

Fig. 1. Distribution range of R. microptera and the sampling sites used in this study. Shaded area shows the states in which this species has been observed (Rehn and Grant 1959). As indicated in the Materials and Methods, we failed to Þnd specimens in North and South Carolina (NC and SC). Numbers on the map indicate the collection sites. The haplotypes found at each collection site and their frequencies in parentheses are indicated in the table at the bottom. Collection site locations and their abbreviation used here and elsewhere are as follows: 1) HNT, Huntsville, TX; 2) ARK, Bluff City, AR; 3) NAT, Natchitoches, LA; 4) LAF, Lafayette, LA; 5) MIS, Pascagoula, MS; 6) PEN, Pensacola, FL; 7) STM, St. Marks, FL; 8) COP, Copeland, FL; 9) SHK, Shark Valley, FL; 10) FLC, Florida City, FL; 11) ATH, Athens, GA; and 12) ALA, Gadsten, AL.

(mtDNA) lineages. Mitochondrial DNA has proven to be particularly useful for population studies because it undergoes no recombination, is maternally inherited, and has a simple sequence organization (Harrison 1989). In some studies, it has been shown that mtDNA haplotypes (unique mtDNA sequences) are often geographically localized within the range of species (Avise et al. 1979, Lansman et al. 1983, Phillips 1994). This makes it possible to relate the geographic position in a gene tree (Avise et al. 1987), providing information on the evolutionary history of a species and the dispersal of its lineages (Avise and Nelson 1989). Such intraspeciÞc phylogeographic studies have been done in a variety of animals, including many insect species (Zehnder et al. 1992, Vogler and DeSalle 1993, Costa and Ross 1994, Nurnberger and Harrison 1995, Juan et al. 1996). The eastern lubber grasshopper, Romalea microptera (Houttuyn), can be found from east Texas to the tip of the Florida peninsula, and as far north as central North Carolina and Arkansas (Fig. 1; Rehn and Grant 1959, 1961; however, see our comments about the current distribution of this species below). This ßightless species is very limited in its dispersal ability and shows a patchy distribution across its range. The distribution range of this temperate species suggests that it would have been severely affected by the climate changes during the Pleistocene (Johannesen and Loeschke 1996). Hence, R. microptera would seem to be an ideal species for studying the effects of these events on the genetic structure of a low-mobility species.

In this study, we determined the degree of genetic differentiation across the range of R. microptera and compared these results with the patterns observed in several other species in the same region. In addition, the results were analyzed to determine whether R. microptera used northern Florida (between St. Marks and Pensacola) as a refuge area during glacial maxima. It has been suggested that other animal species, including some insects, used this area as a refuge during these periods (Neill 1957). This is the Þrst study in which these questions have been investigated at the molecular level for this species. Several species, including mammals, Þsh, and insects exhibited congruent patterns of genetic differentiation in the southern United States (i.e., eastern versus western clade separation; Bermingham and Avise 1986, Vogler and DeSalle 1993, Ellsworth et al. 1994). Because of its low mobility and current distribution, we anticipated that R. microptera would have a pattern of genetic subdivision similar to that observed in other species located in the same region. We failed to detect such a distribution, suggesting either that barriers for the dispersal of this species were not present or that the past distribution of this species was more homogeneous than currently observed. Materials and Methods Eastern lubber grasshoppers were collected from 12 different sites across its range during the summers of 1995Ð1998 (Fig. 1). As an attempt to avoid collecting

July 2004

MUTUN AND BORST: PHYLOGEOGRAPHY OF EASTERN LUBBER GRASSHOPPER

specimens from the same egg pod, the distribution site was divided into grids, and the samples were collected randomly from the area. We failed to obtain specimens from either North or South Carolina, despite several collecting trips to locations mentioned in the literature and appeals to colleagues in the area. Thus, the current range of the eastern lubber grasshopper does not seem to include many if any of the regions in these two states. The number of specimens from each population and their locations are given in Fig. 1. Some specimens were maintained in separate cages in the laboratory and fed with lettuce and oatmeal until analyzed. Other individuals were preserved at ⫺80⬚C until analysis. Two grasshopper species, Taeniopoda eques (Burmeister) from Tucson, AZ, and Taeniopoda tamaulipensis (Rehn) from Victoria, Tamaulipas, Mexico, which are closely related to R. microptera in their morphology (Rehn and Grant 1959), were used as outgroups. mtDNA was isolated using the alkaline-lysis method described in Tamura and Aotsuka (1988). The use of tissue rich in mtDNA relative to nuclear DNA is preferred for the extraction of mtDNA. Therefore, hind leg muscle from adults was homogenized with buffer A (0.25 M sucrose, 10 mM EDTA, and 30 mM Tris-HCl, pH 7.5) and centrifuged (2,000, 5,000, and 12,000 rpm). The pellet from the last centrifugation step was resuspended in buffer B (0.15 M NaCl, 10 mM EDTA, and 10 mM Tris, pH 8.0) and the mtDNA puriÞed by extracting twice with phenol-chloroform and precipitating with ethanol. After puriÞcation, samples were checked using a 1% agarose gel in Tris borate-EDTA buffer (0.045 M Tris-borate, pH 8.0, and 0.001 M EDTA) to ensure that we successfully isolated mtDNA. A 420-bp region of the cytochrome b (cty b) gene was ampliÞed by polymerase chain reaction (PCR) (each 100-␮l reaction volume contained 10 ␮l of 10⫻ PCR buffer, 200 ␮M dNTP, 10 ␮M of each primer, 2.5 mM MgCl2, and 2 U of Taq polymerase [catalog no. D-1806, Sigma, St. Louis, MO]), and 2 ␮l of mtDNA extract for 30 cycles (denaturation, 92⬚C for 45 s; annealing, 52⬚C for 1 min; and primer extension, 72⬚C for 1 min) with a thermal cycler (Gene Amp PCR System 2400, Applied Biosciences, Foster City, CA). The following primer set was used for the PCR ampliÞcations in our study: CBJ (5⬘ primer) TAT GTA CTA CCA TGA GGA ACA AAT AT and CBN (3⬘ primer) AAT ACA CCT CCT AAT TTA TTA GGA AT (Operon, Alameda, CA). This primer set has been used previously to amplify a region of the cty b gene in a variety of insect species (Simon et al. 1994). The PCR samples were run on agarose gels (1%) to determine whether the desired part (a 420-bp region ßanked by the above-mentioned primers) of the cyt b gene had been ampliÞed, and whether there were multiple bands due to either heteroplasmy or possible contamination with pseudogenes. The PCR products were then cleaned (Gene Clean II, Bio 101, La Jolla, CA) and sequenced using dye-termination cycle sequencing (Sanger et al. 1977). Brießy, each sample was ampliÞed by PCR (25 cycles of 96⬚C for 10 s, 50⬚C

683

for 5 s, and 60⬚C for 4 min) by using ddNTPs labeled with different ßuorescent dyes (Applied Biosystems). The products were analyzed with an ABI 310 DNA sequencer (Applied Biosystems) by using the protocol suggested by the manufacturer. Both strands of each sample were sequenced to minimize PCR artifacts, ambiguities, and base-calling errors. Sequences were aligned using a pairwise similarity measurement (Mac Vector, version 4.0, IBI, Symantec Corp., CT). Distances for pairwise comparisons of haplotypes were calculated using the Kimura twoparameter model (Kimura 1980). An AMOVA analysis (Arlequin program, version 1.1; Schneider et al. 1997) was used to determine the distribution of genetic variation at different hierarchical levels over the range of the species. The Fst values generated by the AMOVA analysis allowed the genetic distances between pairs of populations to be determined. These distances were then used in a hierarchical analysis of population differentiation (ExcofÞer et al. 1992). The ⌽ statistics yielded by AMOVA were tested for signiÞcance using 1000 random permutations. The Holsinger and Mason-Gamer (1996) method (H&M-G) was used to estimate the geographic distribution pattern of genetic variation as well as the genetic diversity within each population. The results of this analysis were used to create a dendrogram depicting the pattern of genetic differentiation among populations. This method provided a detailed description of the hierarchical relationship between the populations and an estimate of the genetic differentiation (gˆ st) for each node on the dendrogram. The sequences were then transferred to PAUP, version 3.01 (Swofford 1990) to determine the percentage of sequence divergence both within and between populations by using pairwise comparisons of the haplotypes. The evolutionary relationships of the haplotypes were estimated using maximum parsimony analysis. The homologous sequences of T. eques and T. tamaulipensis were used as outgroups to detect any geographic polarity. In addition to the unweighted parsimonious analysis, we performed different weighting trials (transversion versus transition; 2:1, 5:1, and 10:1). The bootstrap method was applied with 500 replications. Computational limitations restricted the use of bootstrapping option to unweighted parsimony searches. We also constructed a median network analysis to estimate the evolutionary relationship among haplotypes (Bandelt et al. 1995, 1999, 2000). The data were then transferred to PHYLIP (PHYLIP 3.5c; Felsenstein 1993) to obtain a maximum likelihood-based unrooted haplotype tree by using DNAMLK (DNA Maximum Likelihood Method). This search was repeated three times (conÞdence trials; 1, 101, 1001) with 20 randomly entered haplotypes. After each run with different seed numbers, the logs of likelihood values for each trial were compared as suggested by Felsenstein (1993). The genealogy obtained from DNAMLK was compared with the PAUP tree.

Table 1. Haplotype sequences grouped by population and collection site

684 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 97, no. 4


Table 1. continued

July 2004

685

For the location of each collection site, see Figure 1. All haplotypes found in each population are shown. Common (shared) haplotypes are shown in bold.

Table 1 continued

686 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 97, no. 4

July 2004 Table 2.


687

Hierarchical analysis of variance by using different grouping schemes for the mtDNA cytochrome b gene Test 1

Group 1 (HNT, ARK, LAF, NAT, MIS, ATH, ALA), Group 2 (PEN, STM, COP, SHK, FLC) Source of variation Between groups Among populations in groups Within populations

df 1 10 156

% 3.59 34.65 62.06

⌽ 0.035 0.383 0.405

p NS *** ***

2 9 156

5.39 33.32 59.28

0.053 0.373 0.407

NS *** ***

3 8 156

0.22 39.43 60.35

0.002 0.395 0.395

NS *** ***

Test 2 Group 1 (HNT, LAF, NAT, MIS, ARK), Group 2 (ATH, ALA), and Group 3 (PEN, STM, COP, SHK, FLC) Between groups Among populations in groups Within populations Group 1 (HNT, ARK), Group 2 (LAF, NAT), Group 3 (MIS, ATH, ALA), and Group 4 (PEN, STM, COP, SHK, FLC) Between groups Among populations in groups Within populations

Test 3

%, percentage of variation; ⌽, Þxation indices; p, signiÞcance of percentage variation and Þxation indices (estimated from 1,023 random permutation tests). See Fig. 1 for abbreviations of haplotype locations. NS, nonsigniÞcant. * p ⬍ 0.05; ** p ⬍ 0.01; *** p ⬍ 0.001.

Results Nucleotide Diversity. A 420-bp region of the mitochondrial cytochrome b gene was analyzed in 168 individuals from 12 collection sites. Forty-nine unique mitochondrial DNA sequences (haplotypes) were identiÞed (Table 1). These contained 51 substitution sites, of which 13 sites (⬇25%) showed transversion substitutions, 33 sites (⬇65%) showed transition substitutions, and Þve sites (⬇10%) showed parallel mutations. No additions or deletions were observed in this region when it was compared with the homologous region in other insect species (Crozier and Crozier 1993, Jermiin and Crozier 1994, Flook et al. 1995). In T. eques, there were three haplotypes out of 11 individuals sequenced and in T. tamaluipensis, there were two haplotypes out of Þve individuals. The most common mtDNA sequences of R. microptera (H-1) and of the two outgroup species used in this study, T. eques (TEQ-A) and T. tamaulipensis (TAM-1), have been submitted to GenBank (accession numbers AF 138855, AF 138856, and AF 138857, respectively). Nucleotide diversity for each population was calculated from the average pairwise differences between each sequence (Nei 1987). Among 1,126 pairwise comparisons, 33 cases showed the least possible divergence estimate (0.2% or one base pair difference). The largest sequence divergence was 3.8% between haplotype 9 (H-9, from STM) and H-15 (from Pensacola [PEN]). When the sequence divergence was analyzed by population, the lowest intrapopulation sequence divergence was 0.5% and the highest was 2.9% for the Copeland (COP) and the St. MarkÕs (STM) populations, respectively. Although seven (14%) of the 49 haplotypes were shared among populations, most (42 or 86%) were singletons (found only in one population). The most common shared haplotype (H-1) was found in 16 individuals in seven populations (Fig. 1). The second

most common was H-28, found in six individuals in four populations. The third most common (H-42) was found in eight individuals in three populations. Most of these individuals (six of eight) in H-42 were clustered in one location (COP). Analysis of Geographic Structure of R. microptera. The data were analyzed using the AMOVA analysis (Arlequin, version 1.1; Schneider et al. 1997) to investigate population subdivisions. For this analysis, the lubber populations were grouped in several alternative geographic groups to reveal geographic clustering of genetic differentiation (Table 2). In each trial, the total genetic variation was separated into three components. These components were the genetic variation found 1) between geographical groups, 2) among populations in each group, and 3) within each population. These trials showed that estimates of genetic variation between geographical groups were not signiÞcant (P ⱕ 0.106) (Table 2). This could reßect the small number of populations sampled, or the high rates of historical and/or ongoing gene ßow. In contrast, estimates of genetic variation found in the other two components (among populations in each group, and within each population) were signiÞcant in all of the geographical groupings tested. Indeed, the variance calculated for the second component (among populations in each group) was signiÞcantly higher (P ⱕ 0.001) than those for Þrst component (between geographical groups), indicating that the populations within each group are not homogeneous. Overall, these trials revealed that most of the genetic variation observed in lubber grasshoppers was due to the variance component within each population. Hence, it seems that each population developed its own genetic variation. The geographic structure and genetic differentiation of the R. microptera populations were also determined using the H&M-G (Holsinger and Mason-

688


Vol. 97, no. 4

Fig. 2. Dendrogram depicting the hierarchical relationship between populations of R. microptera. The number following the population abbreviation (see Fig. 1) is the estimated amount of within population diversity for that population. At each node, the genetic differentiation between that population and the group of populations on the hierarchy, gˆ st, is given with the probability of the genetic differentiation at that node.

Gamer 1996) method. The H&M-G analysis did not show any major geographic structure for these populations. However, considerable genetic differentiation was observed across the range of the species, and each population was signiÞcantly different from each other (Fig. 2). For example, a hierarchical analysis of these populations revealed that the Florida City (FLC) population was most closely related to the PEN population and was signiÞcantly different (P ⱕ 0.001) from the other 11 populations grouped together. Likewise, the STM population had the largest genetic differentiation value of the 12 populations (gˆ st ⫽ 0.100) and was also signiÞcantly (P ⱕ 0.001) different from all of the other populations grouped together. The H&M-G method was also used to estimate the local genetic diversity of each population. The STM population had the largest local genetic diversity (0.093) followed by the PEN population (0.0084). The COP population had the lowest local genetic diversity (0.0039). Overall, the results from the H&M-G analysis indicate that each of the populations is signiÞcantly different from the other populations, suggesting that each has been isolated for sufÞcient time to become genetically distinct. Phylogeographic Analysis. Haplotype trees were created using maximum parsimony (with PAUP) and maximum likelihood (with PHYLIP) analyses. The genealogies were then used to examine the phylogeographic relationships of the haplotypes. The haplotype trees obtained with PAUP were rooted with two outgroups, T. eques and T. tamaulipensis, whereas the tree obtained with PHYLIP was unrooted. Because the unweighted and the weighted haplotype trees obtained from PAUP had the same topologies, only the unweighted tree is shown (Fig. 3). The PAUP genealogy did not resolve the relationships between most R. microptera haplotypes due to

Fig. 3. Haplotype tree derived from PAUP analysis. Outgroup haplotypes are abbreviated as TAM1 and TAM2 (haplotypes of T. tamaulipensis) and TEQ-A, TEQ-C and TEQ-AB (haplotypes of T. eques). Haplotypes of R. microptera are designated by their location (see Fig. 1). The numbers on the branches indicate the percentage of the bootstrap replicates in which that node was supported.

polytomies (Fig. 3). Furthermore, most of the small haplotype clusters within the large polytomic clade included haplotypes from geographically distant populations. Thus, there was no evidence for strong geographical clustering in the PAUP trees. The evolutionary relationship of the haplotypes was further analyzed using their estimated ages that were calculated by DNAMLK (a subprogram in PHYLIP) (Fig. 4). H-9 was in the most basal position in this genealogy, suggesting that it was the most ancient haplotype. H-43 and H-44 were the youngest haplo-

July 2004


Fig. 4. Haplotype genealogy of eastern lubber grasshopper haplotypes generated using their estimated ages. The age of each haplotype is indicated by the branch length to the Þrst coalescing node. The haplotype with an asterisk is the oldest haplotype. Estimates of age were made with the PHYLIP program using the DNAMLK subprogram. See Fig. 1 for the location of each haplotype.

types, based on branch length and their rank order within the hierarchy. The haplotypes seem to be divided into two distinct clades based on their estimated ages. The smaller clade contained H-5 and H-6 from the STM population. Likewise, H-7 and H-48 from the STM and ALA populations, respectively were monophyletic within this clade. In the larger clade, two distinct subclades were detected with small amounts of internal structure. Although the PHYLIP analysis, like the PAUP analysis, did place some haplotypes from the same geographic locations together, the PHYLIP analysis showed little correlation between geographic location and haplotype clusters according to their ages.

689

Finally, we used a median network analysis to investigate possible relationships between the R. microptera haplotypes. The result of this analysis is shown in Fig. 5. The resulting network was complicated, reßecting the total number of haplotypes being analyzed. In addition, parallel and back mutations may have increased the complexity of the network. The network analysis indicates that the two most widespread haplotypes (H-1 and H-28) were directly or indirectly the founders of most other haplotypes (Fig. 5). However, the results of the network analysis were similar to those obtained from PAUP. No significant clade structure was detected based on the geographic origins of the haplotypes. Estimates of Historical and Current Population Sizes. Estimates of ␪s (the current population size) and ␪k (the historical population size) were calculated using Arlequin (version 1.30) for the 12 populations (Table 3). These values can be used to infer changes in the size of each population over time (Ewens 1972, Waterson 1975). In the analysis shown, the populations were divided into two groups based on their geographic location. Other groupings gave similar results. The estimates of ␪s for populations in group 1 (located in Florida) ranged from 1.611 to 2.731. For populations in group 2 (not located in Florida), the estimates of ␪s ranged from 1.489 to 4.361. The ␪k estimates for group 1 ranged from 1.956 to 5.407 and for group 2 from 1.956 to 4.627. These results indicated that some populations (e.g., Gadsten, AL [ALA]; Huntsville, TX [HNT]; and Bluff City, AR [ARK]) declined drastically, whereas others (e.g., Athens, GA [ATH] and COP) showed a smaller decrease in size. Only two populations (STM and FLC) showed an increase in their population size (Table 3). Gene Flow and Dispersal. Because gene ßow is an important force shaping the genetic structure of populations, estimates of gene ßow are important for understanding the dispersal of a species over its range (Chenoweth et al. 1998). The gene ßow between the R. microptera populations was estimated using Arlequin (Table 4). The lowest estimated gene ßow (Nm ⫽ 0.355) was obtained between the populations in FLC and ATH. The highest estimated gene ßow (Nm ⫽ 3.022) was obtained between the populations in ALA and ARK. For some populations, there was a clear correlation between geographic distance and gene ßow. For example, the three populations in southern Florida form an approximate straight line, with the maximum distance (COP to FLC) ⬇100 miles. The Shark Valley, FL (SHK) population is near the midpoint between the other two populations. The estimated rate of gene ßow parallels this relationship. The gene ßow between FLC and COP was estimated as Nm ⫽ 0.550, whereas the gene ßow between COP and SHK was Nm ⫽ 0.854 and between FLC and SHK was Nm ⫽ 0.913. However, the relationship between distance and gene ßow was less clear for populations that are more geographically distant. For example, the estimated gene ßow between the HNT and the MIS populations was 1.889. However, the estimated gene ßow between the HNT and the COP was Nm ⫽ 2.053,

690


Vol. 97, no. 4

Fig. 5. Network analysis of the 49 mtDNA haplotypes found in R. microptera. Filled circles represent hypothetical intermediate haplotypes. The numbers within the open circles indicate the haplotypes that were found in this study. Each slash and corner represents one mutational step between haplotypes.

although these two populations are geographically much more distant (Table 4). We also estimated the genetic differentiation (Fst) and the coefÞcients of coancestry (D) between speciÞc pairs of populations by using the Arlequin program. As shown in Table 4, estimates of Fst ranged from 0.141 to 0.584, and estimates of D ranged from 0.153 to 0.877 (where D ⫽ 0 is identical shared anTable 3. Current and historical population size estimates (␪s and ␪k, respectively) for each R. microptera population Group 1 populations (PEN, STM, COP, SHK, FLC) 95% conÞdence Population ␪s SD for ␪s ␪k interval for ␪k PEN STM COP SHK FLC

3.072 4.361 1.489 1.767 2.460

1.527 1.835 0.732 1.013 1.201

4.627 3.720 1.905 5.407 2.202

1.581Ð13.507 1.472Ð9.089 0.761Ð4.452 1.779Ð16.627 0.770Ð5.975

Group 2 populations (HNT, ARK, LAF, NAT, MIS, ATH, ALA) ␪s SD for ␪s ␪k 95% conÞdence Population interval for ␪k HNT ARK LAF NAT MIS ATH ALA

1.839 2.731 1.611 1.767 1.808 2.649 2.731

1.068 1.389 0.899 1.013 0.941 1.331 1.389

3.828 4.627 2.486 1.956 2.097 2.684 4.627

1.196Ð12.229 1.581Ð13.507 0.851Ð6.961 0.598Ð6.127 0.739Ð5.620 0.906Ð7.676 1.581Ð13.507

Estimates were obtained from Arlequin (version 1.03). See Fig. 1 for abbreviations of haplotype locations.

cestry). The results indicated that all pairwise estimates were signiÞcant. The genetic distances between these population pairs did not show any relationship to geographic location. For example, the Fst estimate between the Lafayette, LA (LAF), Natchitoches, LA (NAT), and the HNT populations was 0.442, whereas the Fst estimate between the FLC and the HNT populations was 0.439. In summary, the results of this analysis indicated a signiÞcant and high amount of genetic differentiation in most of the R. microptera populations but no strong correlation between genetic differentiation and geographic distance. Discussion In the past decade, there has been a rapid development in new molecular techniques and approaches for analyzing the population structure of a species. These can be used to directly measure the genetic divergence between populations of a species. Genetic diversity has been previously estimated in a variety of insect species, with values ranging from 2.0 to 11.1% for Glossina morsitans (Westwood) (Trick and Dover 1984), from 0.17 to 1.2% for Schzaphis graminum (Rondani) (Powers et al. 1989), and from 0.18 to 0.92% for Melanoplus sanguinipes (F.) (Chapco et al. 1992). In this study, we found that the genetic diversity of R. microptera ranged from 0.2 to 2.9% within individual populations, and from 0.2 to 3.8% over the entire range of the species. This level of genetic diversity is similar

Fst ⫽ 0.264* D ⫽ 0.306 Nm ⫽ 1.393 Fst ⫽ 0.442** D ⫽ 0.584 Nm ⫽ 0.630 Fst ⫽ 0.466* D ⫽ 0.628 Nm ⫽ 0.571 Fst ⫽ 0.209** D ⫽ 0.234 Nm ⫽ 1.889 Fst ⫽ 0.507** D ⫽ 0.709 Nm ⫽ 0.484 Fst ⫽ 0.393* D ⫽ 0.500 Nm ⫽ 0.770 Fst ⫽ 0.427** D ⫽ 0.557 Nm ⫽ 0.669 Fst ⫽ 0.326** D ⫽ 0.395 Nm ⫽ 1.029 Fst ⫽ 0.195* D ⫽ 0.217 Nm ⫽ 2.053 Fst ⫽ 0.352* D ⫽ 0.434 Nm ⫽ 0.919 Fst ⫽ 0.439** D ⫽ 0.578 Nm ⫽ 0.638

HNT

Fst ⫽ 0.183* D ⫽ 0.202 Nm ⫽ 2.230 Fst ⫽ 0.253** D ⫽ 0.292 Nm ⫽ 1.470 Fst ⫽ 0.269** D ⫽ 0.313 Nm ⫽ 1.355 Fst ⫽ 0.319** D ⫽ 0.384 Nm ⫽ 1.066 Fst ⫽ 0.141* D ⫽ 0.153 Nm ⫽ 3.022 Fst ⫽ 0.268** D ⫽ 0.312 Nm ⫽ 1.364 Fst ⫽ 0.175** D ⫽ 0.192 Nm ⫽ 2.354 Fst ⫽ 0.250** D ⫽ 0288 Nm ⫽ 1.496 Fst ⫽ 0.308* D ⫽ 0.369 Nm ⫽ 1.118 Fst ⫽ 0.459** D ⫽ 0.614 Nm ⫽ 0.588

ARK

Fst ⫽ 0.319** D ⫽ 0.384 Nm ⫽ 1.066 Fst ⫽ 0.379** D ⫽ 0.477 Nm ⫽ 0.816 Fst ⫽ 0.413** D ⫽ 0.533 Nm ⫽ 0.709 Fst ⫽ 0.252* D ⫽ 0.290 Nm ⫽ 1.482 Fst ⫽ 0.359** D ⫽ 0.445 Nm ⫽ 0.890 Fst ⫽ 0.274** D ⫽ 0.321 Nm ⫽ 1.320 Fst ⫽ 0.375** D ⫽ 0.470 Nm ⫽ 0.832 Fst ⫽ 0.452** D ⫽ 0.602 Nm ⫽ 0.604 Fst ⫽ 0.562** D ⫽ 0.827 Nm ⫽ 0.388

LAF

Fst ⫽ 0.399** D ⫽ 0.509 Nm ⫽ 0.752 Fst ⫽ 0.403** D ⫽ 0.516 Nm ⫽ 0.739 Fst ⫽ 0.259** D ⫽ 0.301 Nm ⫽ 1.423 Fst ⫽ 0.363** D ⫽ 0.451 Nm ⫽ 0.876 Fst ⫽ 0.292** D ⫽ 0.345 Nm ⫽ 1.211 Fst ⫽ 0.477** D ⫽ 0.649 Nm ⫽ 0.546 Fst ⫽ 0.455* D ⫽ 0.607 Nm ⫽ 0.598 Fst ⫽ 0.558** D ⫽ 0.816 Nm ⫽ 0.395

NAT

Fst ⫽ 0.462** D ⫽ 0.620 Nm ⫽ 0.582 Fst ⫽ 0.405** D ⫽ 0.411 Nm ⫽ 0.982 Fst ⫽ 0.405** D ⫽ 0.519 Nm ⫽ 0.734 Fst ⫽ 0.340** D ⫽ 0.416 Nm ⫽ 0.966 Fst ⫽ 0.429** D ⫽ 0.561 Nm ⫽ 0.663 Fst ⫽ 0.404** D ⫽ 0.517 Nm ⫽ 0.737 Fst ⫽ 0.526** D ⫽ 0.748 Nm ⫽ 0.449

MIS

Fst ⫽ 0.201* D ⫽ 0.225 Nm ⫽ 1.976 Fst ⫽ 0.402* D ⫽ 0.515 Nm ⫽ 0.742 Fst ⫽ 0.380** D ⫽ 0.478 Nm ⫽ 0.815 Fst ⫽ 0.533** D ⫽ 0.763 Nm ⫽ 0.436 Fst ⫽ 0.498** D ⫽ 0.691 Nm ⫽ 0.502 Fst ⫽ 0.584** D ⫽ 0.877 Nm ⫽ 0.355

ATH

Fst ⫽ 0.304** D ⫽ 0.363 Nm ⫽ 1.142 Fst ⫽ 0.206** D ⫽ 0.230 Nm ⫽ 1.926 Fst ⫽ 0.420** D ⫽ 0.545 Nm ⫽ 0.689 Fst ⫽ 0.394* D ⫽ 0.501 Nm ⫽ 0.767 Fst ⫽ 0.519** D ⫽ 0.733 Nm ⫽ 0.462

ALA

Fst ⫽ 0.339** D ⫽ 0.414 Nm ⫽ 0.973 Fst ⫽ 0.481** D ⫽ 0.656 Nm ⫽ 0.538 Fst ⫽ 0.434* D ⫽ 0.569 Nm ⫽ 0.651 Fst ⫽ 0.539** D ⫽ 0.774 Nm ⫽ 0.427

PEN

Fst ⫽ 0.384** D ⫽ 0.485 Nm ⫽ 0.799 Fst ⫽ 0.339** D ⫽ 0.414 Nm ⫽ 0.972 Fst ⫽ 0.467** D ⫽ 0.631 Nm ⫽ 0.568

STM

Estimate of the fixation indices (Fst), coancestry coeficients (D), and migration rates (Nm) between population pairs of the eastern lubber grasshopper

Fst ⫽ 0.369** D ⫽ 0.460 Nm ⫽ 0.854 Fst ⫽ 0.475** D ⫽ 0.550 Nm ⫽ 0.550

COP

SHK

Fst ⫽ 0.353* D ⫽ 0.436 Nm ⫽ 0.913

Estimates are based on the analysis of mitochondrial cytochrome b gene sequences using Arlequin (version 1.03) (* p ⬍ 0.01; ** p ⬍ 0.001). See Fig. 1 for abbreviations of haplotype locations.

FLC

SHK

COP

STM

PEN

ALA

ATH

MIS

NAT

LAF

ARK

Table 4.

July 2004 MUTUN AND BORST: PHYLOGEOGRAPHY OF EASTERN LUBBER GRASSHOPPER 691

692


that observed in the cytochrome b gene of other insects. In the oak gallwasp, Andricus quercustozae (Bose), the genetic diversity of the cytochrome b gene between different populations ranged from 0.2 to 4.2% (Rokas et al. 2003). Likewise, the marble gallwasp, Andricus kollari (Hartig), had sequence diversity between 2.8 and 3.8% in the cytochrome b gene between its lineages (Stone et al. 2001). Thus, the genetic diversity of the eastern lubber grasshopper seems to be in line with that observed in other insect species. Most populations of R. microptera showed high haplotype diversity, and the number of private haplotypes was higher than the number of shared haplotypes in most of the populations. It has been previously observed that more alleles are maintained in species with multiple populations than in species with a single large population (Maruyama 1970). Therefore, the high haplotype diversity of the lubber grasshopper may result from its patchy distribution, a characteristic of this species that we encountered while collecting the specimens. Our analysis suggests that most of the populations we studied were larger in the past (see below). The decrease in population size may have been in response to changing environmental conditions. This, combined with the limited dispersal ability of this species, would probably have caused populations to become more fragmented, resulting in the patchy distribution seen today across most of the range of this species. Another result of this isolation and the limited dispersal ability of the lubber grasshopper would have been the development of private haplotypes as new mutations arose in the isolated populations. Indeed, the estimates of the current and historical population sizes indicate that nine of the 12 populations that we studied have had population declines (Table 3). In most cases, the current population size seems to be less than one-half of the maximum size in the past (e.g., the HNT and SHK population). However, some populations seem to have had less drastic declines, and the STM population may have had a small increase in size (␪s ⫽ 4.361 and ␪k ⫽ 3.720). These estimates probably do not adequately reßect the magnitude of changes in population size. Indeed, Þeld observations indicate that lubber grasshopper populations can show drastic declines in size (⬎99.9%) over short periods of time (Lamb et al. 1999). Populations that repeatedly ßuctuate in size will have close estimates of ␪s and ␪k because the effective size of such populations would be equal to the harmonic mean of the population size (Nei 1987). In some previous studies, the number of mtDNA haplotypes found in a population or a species has been shown to depend on the sample size and the amount of sequence surveyed (Neigel and Avise 1993). However, we failed to detect such a relationship in this study. The number of individuals from each population that we analyzed in this study varied from a maximum of 32 (COP) to a minimum of nine (HNT). The sample sizes of the remaining populations varied from 10 and 18. The number of haplotypes in these populations ranged from four to seven, with six hap-

Vol. 97, no. 4

lotypes in the COP population and Þve haplotypes in the HNT population. Overall, the correlation between sample size and number of haplotypes (R2 ⫽ 0.120) was poor. Of the 12 locations examined, the COP population had the lowest number of private haplotypes. The relatively low number of haplotypes found in the COP population could reßect the inadvertent collection of multiple individuals from a few egg pods. However, the sampling area at this site extended over ⬎2 miles. Given the limited dispersal ability of this species, it seems unlikely that this result reßects a sampling artifact. Alternatively, this population may have had frequent population crash(es) in the recent past. A catastrophic drop in population size could result from habitat ßooding, wildÞres, or parasitic infections (Lamb et al. 1999) and cause the random loss of haplotypes. Such bottleneck effects (caused by disease, parasitism, predation, or natural disasters) have been used to explain the low levels of mtDNA haplotypes in other insect species (Wilson et al. 1985). Was Northern Florida a Refugium for R. guttata. The northwestern area of Florida is well known for the diversity of its endemic plant and animal species. It contains isolated populations of many temperate species that moved to northern Florida during the ice ages to escape the cooling effects of this event. Among the groups of animals that seem to have used northern Florida as a refugium during this period are many insects (Neill 1957). Furthermore, the Apalachicola region seems to be an important zone separating the eastern and western clades of many unrelated taxa that show a continuous range in the southern United States (Avise et al. 1987). It has been suggested that this pattern resulted from the presence of a barrier in the Apalachicola area that may have affected the eastwest movement of species, thereby allowing the genetic differentiation of these populations. This barrier may have been due the broadening of the Apalachicola River due to the elevated sea level during the interglacial cycles (Neill 1957). Similar to these previous studies, we observed the highest levels of local genetic diversity in the STM and PEN populations (Fig. 2), two populations in northern Florida that are separated by the Apalachicola River. These observations are consistent with the hypothesis that the geologic history of the Apalachicola region may have played an important role in maintaining the high genetic diversity of these populations. Furthermore, the presence of the most divergent (presumably the oldest) haplotype (H-9) in the STM population suggests that northern Florida may have been a refugium for the eastern lubber grasshopper as well. However, the distribution of the genetic diversity of R. microptera does not Þt the refugium model very well. In other species it has been observed that the highest genetic diversity is observed near the refugium and less genetic diversity is observed in populations more distant from this center (Chapco 1997). The position of H-9 in the haplotype network (Fig. 5), and in the PAUP tree (Fig. 3), does not demonstrate such a relationship. In all trials of the PAUP analysis H-9 was placed between the two outgroup haplotypes.

July 2004


Thus, it seems more likely that H-9 is a retained ancestral polymorphism in the mtDNA lineage and that it is older than the separation of this species from its ancestral stock. In the haplotype network, a haplotype with an internal position gives rise to more intermediately aged or younger haplotypes (Templeton 1992). Therefore, neither the haplotype network nor the PAUP tree for this species supports the refuge hypothesis. Gene Flow and Dispersal. The genetic structure of most species seems to be governed by several components, including the dispersal ability of the species (Korman et al. 1993). R. microptera is a ßightless, aggregative species and seems to move only 50 Ð100 m during its lifetime (D. Whitman, personal communication). The low mobility of this species would strongly limit the rate of gene ßow between populations and affect its genetic structure. Due to this low mobility, we anticipated that gene ßow and genetic differentiation among populations would be partitioned as a function of geographic distance. Indeed, the analysis of our data by AMOVA indicates that the genetic diversity of this species was mostly attributable to variation within each collection site (Table 2). This is consistent with the eastern lubber grasshopper being a low-mobility species. Nevertheless, the large-scale distribution of some haplotypes suggests that there has been a recent, longdistance dispersal event(s), contrary to the expectations for a low-mobility species (Phillips 1994). Furthermore, the estimates of gene ßow rates between some geographically distant populations were unusually high. It seems likely that some haplotypes were missed in locations between the populations that we sampled. Likewise, the observed haplotypes might be extinct in the intermediate populations. Thus, estimates of gene ßow between geographically distant populations that cannot be explained by the mobility of this species may reßect the error inherent in any large-scale sampling study. Alternatively, the high levels of gene ßow observed in this study may reßect historical gene ßow rather than ongoing gene ßow between the current populations of eastern lubber grasshoppers, because historical and current gene ßow cannot be distinguished with available methods. Extinction and recolonization events may also be responsible for some of the high gene ßow estimates. In an extinction-colonization study of a milkweed beetle species, Tetraopes tetraophthalmus (Fo¨ rster), the extinction rate for haplotypes was estimated to be 0.02 per generation. This rate could either double or halve the Fst estimate, depending upon the migration pattern (McCauley and Eanes 1987). For example, if migration rates and local extinction rates are high, the apparent change in genetic variance within each population would increase (Slatkin 1987). This could also create phylogenetic tree topologies unrelated to geography. Thus, the absence of a strong correlation between the topology of the PAUP tree and the geographical distribution of R. microptera may reßect the extinction and recolonization events in the history of this species. Finally, it has been argued that mobility

693

and rate of gene ßow are not always correlated (Daly 1989). Phillips (1994) showed that the high levels of gene ßow occurring between populations of a sedentary species, the spotted salamander, Ambystoma maculatum, could not be explained by the dispersal ability of this species. He attributed these results to stochastic sorting of ancestral polymorphism. Contemporary and Historical Phylogeographic Structure. Previous studies of species in southeastern United States have revealed a high level of genetic differentiation with major geographic clustering of mtDNA haplotypes into eastern and western clades. These congruent phylogenetic patterns among diverse animal groups have been interpreted as the result of common historical factors that affected the distribution of these species (Avise et al. 1979, Vogler and DeSalle 1993). However, this general pattern has not been observed in all species examined in this region. For example, there was no geographic structure for the haplotypes of the tobacco budworm, Heliothis virescens (F.), which inhabits nearly the same area as these other species (Roehrdanz et al. 1994). The lack of geographic structure for the haplotypes of this insect might be anticipated given the migratory nature of this species. Because the eastern lubber grasshopper currently has limited vagility and a patchy distribution, we anticipated that this species would show low diversity within patches, and larger geographic patterns of genetic differentiation, similar to the patterns observed in many other species from this area. Although we found a high level of mtDNA variability in this species, most of this variability was attributable to the individual populations (Table 4; Fig. 2). Furthermore, no major geographic clustering of haplotypes was observed in this species over its range in the southern United States. In this study, haplotype genealogy obtained from the PAUP analyses did not resolve the evolutionary relationships between most R. microptera haplotypes. However, the polytomic structure of the PAUP tree that includes most of the R. microptera haplotypes may indicate the historical events that have left their imprint on the genetic structure of populations. The application of a mitochondrial DNA molecular clock is controversial, and its calibration requires a known vicariant event in the region or a fossil record (Rand 1994), neither of which is available for the Orthoptera of North America. However, a conventional mtDNA clock calibration (2% sequence divergence per million years between pairs of lineages) has been used for other insect species, including Hawaiian Drosophila, Heliconius erato, L., Salganea species group, and A. kollari (DeSalle et al. 1987, Brower 1994, Maekawa et al. 1999, Stone et al. 2001). Using the 2% sequence divergence/MY yields an estimated divergence time of 2 MYA for the separation of haplotype 9 from T. eques. Because mtDNA lineage separation can predate species or population separation (Avise 1992), it is likely that haplotype 9 diverged before the speciation

694


of R. microptera. The same estimation procedure would give a divergence time of 1.5Ð1.9 MYA for the polytomous node of the haplotype tree. The glacial and interglacial cycles of the Pleistocene and associated climatic and geologic changes greatly affected the distribution of many animal species in the southern United States. As a result, many species were subdivided over their range. The exact causes (habitat fragmentation, changes in temperature and the relative humidity, or the other related factors) of population subdivision in R. microptera are not known. However, these factors have been shown to affect other species during the Pleistocene (Templeton et al. 1990), so it seems likely that these factors also affected the genetic population structure of R. microptera. Nevertheless, the mtDNA haplotypes of the eastern lubber grasshopper showed an unstructured geographic distribution. One possible explanation for this situation is that the distribution of this species was homogeneous in the past, rather than patchy. This could have contributed to the absence of a genetic structure divided geographically into the east-west groupings seen in other species from this region. Thus, R. microptera may have dispersed into the southern United States before or during the early Pleistocene. After dispersing across much of its range, this homogeneous population began to subdivide in response to the changing environmental and climatic conditions. The remaining fragments of the population developed their own private haplotypes in isolation. Alternatively, our results could reßect the absence of longterm barriers to the dispersal of this species. Clearly, human intervention can affect dispersal of organisms, and this has the potential to drastically alter the genetic structure of natural populations. Although R. microptera is used widely for student dissections, these are typically alcohol-Þxed specimens. Dispersal might also involve Þshermen, who use this species as live bait. Nevertheless, the existence of private haplotypes in every population studied argues against human-assisted dispersal as the cause of our observations. A more likely mechanism for the occasional long-distance dispersal of this species might involve strong winds such as hurricanes. During historical times (and presumably at times in the past) the hurricane season has peaked in the fall, at a time when adult eastern lubber females have mated and are reproductively active. Thus, the dispersal of a single female would be sufÞcient to found a new, distant population. Furthermore, hurricanes move across the southeastern United States going either westerly (entering from the Atlantic Ocean) or easterly (entering from the Gulf of Mexico), so over time the pattern of these events would be random. Over time, these new populations would develop their own private haplotypes due to their isolation. Such events, even if rare, could result in an unstructured geographic distribution of genetic diversity unlike that observed in most other species in the same region.

Vol. 97, no. 4

Acknowledgments We thank D. W. Whitman who graciously volunteered time, comments, and information about the studied species. We also thank L. Barientos (Instituto Tecnolo´ gico, Victoria, Mexico) who kindly provided T. tamaulipensis specimens from Mexico. We are grateful to A. P. Capparella, S. S. Loew, E. L. Mockford, and C. J. Phillips (Department of Biological Sciences, Illinois State University) for helpful comments and suggestions. We also express our appreciation to D. W. Whitman for insight into the natural history of lubber grasshoppers. These studies were supported in part by a graduate student fellowship from Abant Izzet Baysal University, Turkey, and two research grants from ISU Chapter of Phi Sigma (to S.M.) and National Science Foundation grants BIR-9510979 and DBI-9978810 (to D.W.B.).

References Cited Avise, J. C. 1992. Molecular population structure and the biogeographic history of a regional fauna: a case history with lessons for conservation biology. Oikos 63: 62Ð76. Avise, J. C., and W. S. Nelson. 1989. Molecular genetic relationships of the extinct seaside sparrow. Science (Wash DC) 243: 646 Ð 648. Avise, J. C., C. Gibling, J. Davidson, J. Laerm, J. C. Patton, and R. A. Lansman. 1979. Mitochondrial DNA clones and matriarchal phylogeny within and among geographic populations of the pocket gopher, Geomys pinetis. Proc. Natl. Acad. Sci. U.S.A. 76: 6694 Ð 6698. Avise, J. C., J. Arnold, R. M. Ball, E. Bermingham, T. Lamb, J. E. Neigel, C. A. Reeb, and N. C. Saunders. 1987. IntraspeciÞc phylogeography: the mitochondrial DNA bridge between population genetics and systematics. Annu. Rev. Ecol. Syst. 18: 489 Ð522. Bandelt, H.-J., P. B. Forster, B. C. Sykes, and M. B. Richards. 1995. Mitochondrial portraits of human populations using median networks. Genetics 141: 743Ð753. Bandelt, H.-J., P. Forster, and A. Rohl. 1999. Median-joining networks for inferring intraspeciÞc phylogenies. Mol. Biol. Evol. 16: 37Ð 48. Bandelt, H.-J., V. Macaulay, and M. Richards. 2000. Median networks: speedy construction and greedy reduction, one simulation, and two case studies from human mtDNA. Mol. Phyl. Evol. 16: 8 Ð28. Bermingham, E., and J. C. Avise. 1986. Molecular zoogeography of freshwater Þshes in the southeastern United States. Genetics 113: 939 Ð965. Brower, A.V.Z. 1994. Rapid morphological radiation and convergence among races of the butterßy Heliconius erato inferred from patterns of mitochondrial DNA evolution. Proc. Natl. Acad. Sci. USA 91: 6491Ð 6495. Chapco, W. 1997. Molecular evolutionary genetics in orthopteroid insects, pp. 337Ð354. In S. K. Gangware, M. C. Muralirangan, and M. Muralirangan [eds], The bionomics of grasshoppers, katydids and their kin. CAB, Wallingford, United Kingdom. Chapco, W., R. A. Kelln, and D. A. McFadyen. 1992. IntraspeciÞc mitochondrial DNA variation in the migratory grasshopper, Melanoplus sanguinipes. Heredity 69: 547Ð 557. Chenoweth, S. F., J. M. Hughes, C. P. Keenan, and S. Lavery. 1998. Concordance between dispersal and mitochondrial gene ßow: isolation by distance in a tropical teleost, Lates calcarifer (Australian barramundi). Heredity 80: 187Ð197. Costa III, J. T., and K. G. Ross. 1994. Hierarchical genetic structure and gene ßow in macrogeographic populations

July 2004


of the eastern tent caterpillar (Malacosoma americanum). Evolution 48: 1158 Ð1167. Cox, C. B., and P. D. Moore. 1980. Biogeography: an ecological and evolutionary approach. Wiley, New York. Crozier, R. H., and Y. C. Crozier. 1993. The mitochondrial genome of the honeybee Apis mellifera: complete sequence and genome organization. Genetics 133: 97Ð117. Daly, J. C. 1989. The use of electrophoretic data in a study of gene ßow in the pest species Heliothis armigera (Hu¨ bner) and H. punctigera Wallengren (Lepidoptera: Noctuidae). In H. D. Loxdale and J. den Hollandere [eds.], Electrophoretic studies on agricultural pests. Syst. Assoc. Spec. Vol. 39: 115Ð141. DeSalle, R., T. Freedman, E. M. Prager, and A. C. Wilson. 1987. Tempo and mode of sequence evolution in mitochondrial DNA of Hawaiian Drosophila. J. Mol. Evol. 26: 157Ð164. Ellsworth, D. L., R. L. Honeycutt, N. J. Silvy, J. W. Bickham, and W. D. Klimstra. 1994. Historical biogeography and contemporary patterns of mitochondrial DNA variation in white-tailed deer from the southeastern United States. Evolution 48: 122Ð136. Ewens, W. J. 1972. The sampling theory of selectively neutral alleles. Theor. Pop. Biol. 3: 87Ð112. Excoffier, L., P. E. Smouse, and J. M. Quattro. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131: 479 Ð 491. Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package), Version 3.5c. Dep. of Genetics, University of Washington, Seatle. Flook, P. K., C.H.F. Rowell, and G. Gellissen. 1995. The sequence, organization, and evolution of the Locusta migratoria mitochondrial genome. J. Mol. Evol. 41: 928 Ð941. Harrison, R. G. 1989. Animal mitochondrial DNA as a genetic marker in population and evolutionary biology. Trends Ecol. Evol. 4: 6 Ð11. Holsinger, K. E., and K. J. Mason-Gamer. 1996. Hierarchical analysis of nucleotide diversity in geographically structured populations. Genetics 142: 629 Ð 639. Jermiin, L. S., and R. H. Crozier. 1994. The cytochrome b region in the mitochondrial DNA of the ant Tetraponera rufoniger: sequence divergence in Hymenoptera may be associated with nucleotide content. J. Mol. Evol. 38: 282Ð 294. Johannesen, J., and V. Loeschke. 1996. A hierarchical analysis of genetic structure and variability in patchily distributed coexisting Chiastocheta species (Diptera: Anthomyiidae). Heredity 76: 437Ð 448. Juan, C., K. M. Ibrahim, P. Oromi, and G. M. Hewitt. 1996. Mitochondrial DNA sequence variation and phylogeography of Pimelia darkling beetles on the Island of Tenerife (Canary Islands). Heredity 77: 589 Ð598. Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitution through comparative studies of nucleotide sequences. J. Mol. Evol. 16: 111Ð120. Korman, A. K., J. Mallet, J. L. Goodenough, J. B. Graves, J. L. Hayes, D. E., Hendricks, R. Luttrell, S. D. Pair, and M. Wall. 1993. Population structure in Heliothis virescens (Lepidoptera: Noctuidae): an estimate of gene ßow. Ann. Entomol. Soc. Am. 86: 182Ð188. Lamb, M. A., D. J. Otto., and D. W. Whitman. 1999. Parasitism in eastern lubber grasshopper by Anisia serotina (Diptera: Tachinidae) in Florida. Fla. Entomol. 82: 365Ð 371. Lansman, R. A., J. C. Avise, C. F. Aquadro, J. F. Shapira, and S. W. Daniel. 1983. Extensive genetic variation in mito-

695

chondrial DNAs among geographic populations of the deer mouse, Peromyscus maniculatus. Evolution 37: 1Ð16. Maekawa, K., N. Lo, O. Kitade, T. Miura, and T. Matsumoto. 1999. Molecular phylogeny and geographic distribution of wood-feeding cockroaches in East Asian islands. Mol. Phylogenet. Evol. 13: 360 Ð376. Maruyama, T. 1970. Effective number of alleles in a subdivided population. Theor. Pop. Biol. 1: 273Ð306. McCauley, D. E., and W. F. Eanes. 1987. Hierarchical population structure analysis of the milkweed beetle, Tetraopes tetraophthalmus (Fo¨ rster). Heredity 58: 193Ð201. Myers, A. A., and P. S. Giller. 1990. Analytical Biogeography: An Integrated Approach to the Study of Animal and Plant Distributions. Chapman and Hall, New York. Nei, M. 1987. Molecular evolutionary genetics. Columbia University Press, New York. Neigel, J. E., and J. C. Avise. 1993. Application of a random walk model to geographic distributions of animal mitochondrial DNA variation. Genetics 135: 1209 Ð1220. Neill, W. T. 1957. Historical biogeography of present-day Florida. Bull. Fla. State Mus., Biol. Sci. vol. 2: 173Ð223. Nilsson, T. 1983. The Pleistocene: Geology and Life in the Quaternary Ice Age. D. Reidel Publishing Co., Dordrecht, Holland. Nurnberger, B., and R. G. Harrison. 1995. Spatial population structure in the whirligig beetle Dineutus assimilis: evolutionary inferences based on mitochondrial DNA and Þeld data. Evolution 49: 266 Ð275. Phillips, C. A. 1994. Geographic distribution of mitochondrial DNA variants and the historical biogeography of the spotted salamander, Ambystoma maculatum. Evolution 48: 597Ð 607. Powers, T. O., S. G. Jensen, S. D. Kindler, C. J. Stryker, and L. J. Sandall. 1989. Mitochondrial DNA divergence among greenbug (Homoptera: Apididae) biotypes. Ann. Entomol. Soc. Am. 82: 298 Ð302. Rand, D. M. 1994. Thermal habit, metabolic rate and the evolution of mitochondrial DNA. Trends Ecol. Evol. 9: 125Ð131. Reeb, C. A., and J. C. Avise. 1990. A genetic discontinuity in a continuously distributed species: mitochondrial DNA in the American oyster, Crassostrea virginica. Genetics 124: 397Ð 406. Rehn, J. A., and H. J. Grant. 1959. A review of the Romaleinae (Orthoptera: Acrididae) found in America north of Mexico. Proc. Acad. Nat. Sci. Phil. 111: 109 Ð271. Rehn, J. A., and H. J. Grant. 1961. A monograph of the Orthoptera of North America (north of Mexico). Monogr. Acad. Nat. Sci. Phil. vol. I, 12: 231Ð245. Rhodes, O. E., R. K. Chesser, and M. H. Smith. 1996. Population dynamics in ecological space and time. The University of Chicago Press, Chicago, IL. Roehrdanz, R. L., J. D. Lopez, J. Loera, and D. E. Hendricks. 1994. Limited mitochondrial DNA polymorphism in North American populations of Heliothis virescens (Lepidoptera: Noctuidae). Ann. Entomol. Soc. Am. 87: 856 Ð 866. Rokas, A., R. J. Atkinson, L.M.I. Webster, G. Csokas, and G. N. Stone. 2003. Out of Anatolia: longitudinal gradients in genetic diversity support an eastern origin for a circum-Mediterranean oak gallwasp Andricus quercustozae. Mol. Ecol. 12: 2153Ð2174. Sanger, F. S., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain reaction terminating inhibitors. Proc. Natl. Acad. Sci. U.S.A. 74: 5463Ð5467. Schneider, S., J.-M. Kueffer, D. Roessli and L. Excoffier. 1997. Arlequin: a software package for population genet-

696


ics. Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, Switzerland. Simon, C., F. Frati, A. Beckenbach, B. Crespi, H. Liu, and P. Flook. 1994. Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Ann. Entomol. Soc. of Am. 87: 651Ð701. Slatkin, M. 1987. Gene ßow and the geographic structure of natural populations. Science (Wash DC) 236: 787Ð792. Stone G., R. Atkinson, A. Rokas, G. Csoka, and J.-L. NievesAldrey. 2001. Differential success in northwards range expansion between ecotypes of the marble gallwasp Andricus kollari: a tale of two lifecycles. Mol. Geol. 10: 761Ð778. Swofford, D. L. 1990. PAUP: phylogenetic analysis using parsimony, version 3.0. Illinois Natural History Survey, Champaign. Tamura, K., and T. Aotsuka. 1988. Rapid isolation method of animal mitochondrial DNA by alkaline lysis procedure. Biochem. Genet. 26: 815Ð 819. Templeton, A. R. 1992. Human origins and analysis of mitochondrial DNA sequences. Science 255: 737. Templeton, A. R., K. Shaw, E. Routman, and S. K. Davis. 1990. The genetic consequences of habitat fragmentation. Ann. MO Bot. Gard. 77: 13Ð27. Trick, M., and G. A. Dover. 1984. Genetic relationships between subspecies of the tsetse ßy Glossina morritans in-

Vol. 97, no. 4

ferred from variation in mitochondrial DNA sequences. Can. J. Genet. Cytol. 26: 692Ð 697. Vogler, A. P., and R. DeSalle. 1993. Phylogenetic patterns in coastal North American tiger beetles, Cicindela dorsalis, inferred from mitochondrial DNA sequences. Evolution 47: 1192Ð1202. Waterson, G. A. 1975. On the number of segregating sites in genetical models without recombination. Theor. Pop. Biol. 7: 256 Ð276. Williams, M.A.J., D. L. Dunkerley, P. D. Deckker, A. P. Kershaw, and T. Stokes, T. 1993. Quaternary environments. Edward Arnolds, A Division of Hodder & Stoughton, London, United Kingdom. Wilson, A. C., R. L. Cann, S. M. Carr, M. George, U. B. Gyllensten, and K. M. Helm-Bychowski. 1985. Mitochondrial DNA and two perspectives on evolutionary genetics. Biol. J. Linn. Soc. 26: 375Ð 400. Zehnder, G. W., L. Sandall, A. M. Tisler, and T. O. Powers. 1992. Mitochondrial DNA diversity among 17 geographic populations of Leptinotarsa decemlineata (Coleoptera: Chrysomelidae). Ann. Entomol. Soc. Am. 85: 234 Ð240.

Received 30 May 2003; accepted 30 January 2004.