Phenetic structure of two Bactrocera tau cryptic species (Diptera: Tephritidae) infesting Momordica cochinchinensis (Cucurbitaceae) in Thailand and Laos

Zoology 116 (2013) 129–138

Contents lists available at SciVerse ScienceDirect

Zoology journal homepage: www.elsevier.com/locate/zool

Phenetic structure of two Bactrocera tau cryptic species (Diptera: Tephritidae) infesting Momordica cochinchinensis (Cucurbitaceae) in Thailand and Laos Jean-Pierre Dujardin a , Sangvorn Kitthawee b,∗ a b

Laboratoire Génétique and Évolution des Maladies Infectieuses (GEMI), UMR 2724 CNRS/IRD, 911 Avenue Agropolis, 34394 Montpellier, France Department of Biology, Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400, Thailand

a r t i c l e

i n f o

Article history: Received 1 March 2012 Received in revised form 23 June 2012 Accepted 10 July 2012 Available online 16 January 2013 Keywords: Bactrocera tau Population structure Sexual shape dimorphism Species identification

a b s t r a c t Morphometric variation with respect to wing venation patterns was explored for 777 specimens of the Bactrocera tau complex collected in Thailand (nine provinces) and Laos (one locality). Cryptic species B. tau A and C were identified based on their wing shape similarity to published reference images. In Thailand, the B. tau A species was identified in four provinces and the B. tau C species in seven provinces, and both species in one locality of Laos. The objective of the study was to explain the geographic variation of size and shape in two cryptic species collected from the same host (Momordica cochinchinensis). Although collected from the same host, the two species did not show the same morphological variance: it was higher in the B. tau A species, which currently infests a wide range of different fruit species, than in the B. tau C species, which is specific to only one fruit (M. cochinchinensis). Moreover, the two species showed a different population structure. An isolation by distance model was apparent in both sexes of species C, while it was not detected in species A. Thus, the metric differences were in apparent accordance with the known behavior of these species, either as a generalist (species A) or as a specialist (species C), and for each species our data suggested different sources of shape diversity: genetic drift for species C, variety of host plants (and probably also pest–host-relationship) for species A. In addition to these distinctions, the larger species, B. tau C, showed less sexual size and shape dimorphism. The data presented here confirm the previously established wing shape differences between the two cryptic species. Character displacement has been discussed as a possible origin of this interspecific variation. The addition of previously published data on species A from other hosts allowed the testing of the character displacement hypothesis. The hypothesis was rejected for interspecific shape differences, but was maintained for size differences. © 2012 Elsevier GmbH. All rights reserved.

1. Introduction Bactrocera (Zeugodacus) tau (Walker) is an important pest of Cucurbitaceae fruits and vegetables. It is widespread throughout Southeast Asia (Drew and Roming, 1997; White and Elson-Harris, 1992). Typically, it damages and/or reduces the yield of products similar to those infested by the melon fly Bactrocera (Zeugodacus) cucurbitae (Coquillet), one of the most detrimental fruit flies (Yang et al., 1994). In several areas, especially in Taiwan and China, B. tau is the dominant pest in comparison to B. cucurbitae (Yang et al., 1994). Due to morphological variation and subsequent taxonomic confusion, B. tau has been recorded under several different names such as Dasyneura tau Walker, Dacus hageni de Meijere, Dacus caudatus var. nubilus Hendel, Dacus nubilus spp. femoralis Hendel, and Dacus (Zeugodacus) tau (Walker). Seven cryptic species were

∗ Corresponding author. E-mail address: [email protected] (S. Kitthawee). 0944-2006/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.zool.2012.07.004

recognized by cytogenetic techniques and tentatively named B. tau A–G (Baimai et al., 2000). An additional species (B. tau I) was identified after multi-locus enzyme electrophoresis and cytochrome oxidase subunit I (COI) gene sequencing (Jamnongluk et al., 2003a; Saelee et al., 2006). Using the modern approach of morphometrics (Rohlf and Marcus, 1993), our study explores the morphological variation of cryptic species A and C in adults obtained from the same host species. Both species may be recognized using genetic techniques (Baimai et al., 2000; Saelee et al., 2006; Jamnongluk et al., 2003b; Thanaphum and Thaenkham, 2003), but the geometric morphometrics approach also applies. Between these morphologically close species, subtle but consistent wing shape differences were disclosed (Kitthawee and Dujardin, 2010). Cryptic species of insects showed distinct wing shapes in other groups as well, such as in kissing bugs (Villegas et al., 2002; Matias et al., 2001; Dujardin et al., 2009; Márquez et al., 2011), sandflies (De la Riva et al., 2001), mosquitoes (Ruangsittichai et al., 2011), scythridids (Roggero and Passerin d’Entrèves, 2005), parasitoid hymenoptera (Baylac et al., 2003; Villemant et al., 2007; Kitthawee and Dujardin, 2009),

130

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

syrphids (Francuski et al., 2009), fruit flies (Kitthawee and Dujardin, 2010) and screwworm flies (Lyra et al., 2009). Furthermore, as in the present study, wing geometry may be useful for detecting intraspecific variation (Gumiel et al., 2003; Broide et al., 2004; Aytekin et al., 2007; Ayala et al., 2011; Ceballos et al., 2011; Hernández et al., 2011). In the latter application, the main advantage of geometric morphometry is its ability to consider the variation of size and shape separately (Dujardin, 2008, 2011). The present study compared the geometric patterns of wing venation between and within the B. tau A and the B. tau C cryptic species on specimens having developed in the same host species. B. tau A infests a wide range of diverse cucurbits (e.g., Cucurbita sp., Luffa sp., Coccinia sp., etc.), whereas B. tau C occurs specifically in the gac fruit, Momordica cochinchinensis (Lour.) Spreng (Baimai et al., 2000; Jamnongluk et al., 2003a; Saelee et al., 2006). In Thailand, the two species have been shown to present clear-cut differences in their wing shape, allowing unambiguous morphometric identification (Kitthawee and Dujardin, 2010). Moreover, based on the same geometric approach, it was also possible to distinguish specimens of B. tau A according to the fruit where the immature stages developed. Thus, the shape variation of B. tau specimens collected in Thailand could be subdivided according to species and host. Since both species may develop sympatrically in the same fruit, M. cochinchinensis, it is possible to exclude the influence of the host environment on wing shape to further explore its variation and focus on possible isolation effects due to geography. Thus, the present study performs a population structure analysis based on morphometric traits (size and shape) of B. tau A and C infesting the same fruit species (M. cochinchinensis). Samples were collected from 15 localities throughout Thailand and Laos, and their corresponding morphometric variation was described separately for each species. A previous morphometric study on these two cryptic species (Kitthawee and Dujardin, 2010) did not explore geographic variation but described the size and shape differences that could help differentiate the two cryptic species. It observed non-overlapping interspecific size variation when larvae of both species developed in the same fruit, suggesting a possible character displacement, as already observed in Tephritidae (Marsteller et al., 2009). To further explore this hypothesis, we combined the present data with the previously published data obtained for species A collected from other host plants in Thailand. Thus, both size and shape variation could be explored in relation to the locality and/or the host.

2. Materials and methods 2.1. Sample collection and species identification A total of 777 adult B. tau, 384 females and 393 males, were obtained from infested M. cochinchinensis collected in Thailand (from 2007 to 2009) and Laos (in 2009). The infested M. cochinchinensis were placed in a plastic box containing sawdust at the bottom and brought to the laboratory at Mahidol University for rearing until adults emerged. Adult flies were maintained in a plastic cage (23.5 cm × 34.5 cm × 23 cm) at 25 ± 2 ◦ C and 70 ± 5% relative humidity. They were fed with a mixture of sugar and yeast hydrolysate (3:1); 10% honey in distilled water was also provided for 2 weeks to allow the development of wing patterns suitable for identification. Both left and right wings of female and male fruit flies were mounted on glass slides and photographed using a digital camera connected to a stereomicroscope at 40× magnification. To avoid possible optical distortion at the periphery of the optical lens, each wing was positioned at the center of the visual field. Right wings only were used for morphometric analyses. Each wing was digitized

Fig. 1. Wing of Bactrocera sp. showing 12 landmarks used in geometric morphometric analysis. All landmarks are junctions of two different veins (type I landmarks according to Bookstein, 1991).

by the same person at 12 “type I” landmarks (venation intersections; see Fig. 1) (Bookstein, 1991). The wings were classified separately for males and females into species A or C according to their similarity to the shape of each reference species. The reference images for B. tau species A and C (Kitthawee and Dujardin, 2010) were obtained from the CLIC bank at http://www.mpl.ird.fr/morphometrics/clic/index.html (Dujardin et al., 2010). These images were of 9 females and 18 males for B. tau A, and 20 males and 14 females for B. tau C. The identification technique used the shortest distance between each specimen and the mean reference shape; both the Procrustes and the Mahalanobis distances were used, as described in Dujardin et al. (2010). 2.2. Size comparison To compare overall wing size among different populations, we used the isometric estimator known as “centroid size” (CS) derived from coordinate data. It is defined as the square root of the sum of the squared distances between the center of the configuration of landmarks and each individual landmark (Bookstein, 1991). The CSs were compared among groups by both parametric (ANOVA) and non-parametric analyses. Using an ANOVA, and separately for each species, mean CSs were tested for the effects of sex, locality, and their interaction. The non˜ et al., parametric analyses were based on permutations (Caro-Riano 2009), making it possible to compare both means and variances (CSV). The results were illustrated with quantile boxes. Sexual size dimorphism (SSD) was also tested by non-parametric methods. It was estimated as the absolute difference between centroid sizes. 2.3. Shape variation and population structure analysis The shape variables were obtained using the generalized Procrustes analysis (GPA) superimposition algorithm, and subsequent projection of the residues into an Euclidean space to produce the “partial warps” (PW) (Rohlf, 1990). For the identification of specimens, because of the small sample size of female reference images relative to the number of shape variables (PW), the set of their first seven principal components (first seven “relative warps”) were used as input for the computation of Mahalanobis distances. For each sex, shape divergence in both species and localities was explored by the Mahalanobis distances and illustrated by an UPGMA tree. The shape variance of each species and sex was

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

calculated using metric disparity (MD) as in Zelditch et al. (2004). For the sake of comparisons, the same estimators were computed from previously published data (Kitthawee and Dujardin, 2010). The statistical significance of shape-derived Mahalanobis distances, as well as of differences in shape variance (MD), was estimated using non-parametric tests based on permutations (1000 runs). The difference in shape between sexes (SShD, for sexual shape dimorphism) was estimated by the Procrustes distance between sexes. To test for the existence of a non-allometric component, we applied a simple linear regression analysis of SShD on SSD. If significant, it would mean that the shape difference between sexes depends on the size difference. Using an ANCOVA, and separately for each species, mean relative warps scores were tested for the effects of sex, locality, centroid size and their interaction (Kitano et al., 2012). Because there was no significant interaction between centroid size and other factors (sex and locality), interactions with centroid size were excluded from the final model for the analysis of RW1, 2 and 3. Discriminant shape analyses were performed for each sex separately, combining the two species, and an UPGMA tree was derived from the Mahalanobis distances. These trees were expected to illustrate the separation of the two species and also their respective geographic arrangement. The correlation between Mahalanobis distances and geographic distances was estimated separately by sex and species using a Mantel test.

2.4. Allometry The above-mentioned ANCOVA provided a first answer to the question of size influence on shape. To estimate a possible allometric effect in geographic structuring as apparent in shape variation, we computed the linear regression of the first and second discriminant factors on size (CS). This analysis was performed separately for each sex and species, allowing estimation of the contribution of size to shape variation (coefficient of determination). The possible allometric effect on sexual shape dimorphism (SShD) was explored in a different way: first, female–male mean size differences (SSD) and corresponding mean shape distances (SShD) were computed in each locality, separately for each species; then a linear regression test was performed between SShD and SSD across the localities (n = 18). This analysis was performed to detect a possible contribution of SSD to SShD, i.e., an allometric component of SShD.

131

2.5. Character displacement hypothesis Various conditions must be satisfied to assess character displacement (Losos, 2000). We verified the most important one: is the level of differences in sympatry greater than expected by chance? This had to be verified by comparing sympatric and allopatric configurations: does the difference observed in sympatry significantly exceed the difference observed in allopatry? Sympatry was tested at two levels, fruit species (F) and locality (L). Fruits were Coccinia grandis, Cucumis sativus, Cucurbita moschata, Trichosanthes tricuspidata and M. cochinchinensis (Kitthawee and Dujardin, 2010); localities were as reported in Table 1, also including Kanchanaburi and Nan in Thailand (Kitthawee and Dujardin, 2010). The following configurations were processed: (i) development in the same kind of fruit and in the same locality (1F1L) or (ii) in different localities (1F2L); (iii) development in different fruit species but in the same locality (2F1L) or (iv) in different localities (2F2L); (v) development in one fruit species whatever the number of localities (1F, 1L or 2L), or (vi) in two fruits species whatever the number of localities (2F, 1L or 2L); (vii) development in one locality, from the same fruit species or not (1L, 1F or 2F) and (viii) in different localities from the same fruit species or not (2L, 1F or 2F).

2.6. Software Data collection, specimen allocation to species A and C, analyses and graphical outputs were performed using various modules of the CLIC package version 24 (http://bioinfo-prod. mpl.ird.fr/morphometrics/clic/index.html), the PHYLIP package (J. Felsenstein, http://evolution.gs.washington.edu/phylip.html) and NJPLOT software (Perriere and Gouy, 1996). The Mantel test used the software created by Cavalcanti (http://life.bio. sunysb.edu/morph/soft-mult.html).

3. Results 3.1. Sample collection and species identification A total of 393 males (362 from Thailand and 31 from Laos) and 384 females (351 from Thailand and 33 from Laos) were retained as convenient material for morphometric analyses (Table 1 and Fig. 2).

Table 1 Number of Bactrocera tau specimens collected from different localities in Thailand and Laos. Localities

Latitude (N)

Longitude (E)

Code

A M

C F

M

F

Thailand Chiang Mai Chiang Mai Chiang Mai Chiang Mai Chainat Khon Kaen Kamphaeng Phet Loei Mae Hong Son Phetchabun Phetchabun Satun Ubon Ratchathani

18 40 38.88 18 40 38.88 18 40 38.88 18 40 38.88 15 10 41.21 16 38 10.90 16 27 59.41 17 29 23.66 19 18 22.85 16 47 37.81 16 47 37.81 06 51 34.60 15 08 03.98

099 02 53.16 099 02 53.16 099 02 53.16 099 02 53.16 100 07 41.50 101 54 58.22 099 19 28.33 101 43 52.38 097 58 03.07 101 14 08.77 101 14 08.77 100 09 01.40 104 53 33.56

CM(MC)-2 CM(MC)-4 CM(MC)-5 CM(MC)-6 CN(MC)-1 KK(MC)-2 KP(MC)-1 LO(MC)-5 MS(MC)-1 PE(MC)-1 PE(MC)-3 ST(MC)-3 UB(MC)-1

22 – – 21 – – 21 – – 10 20 – 15

41 – – 21 – – 20 – – 5 20 – 18

13 37 28 39 20 18 – 21 41 – 18 18 –

7 35 28 40 18 6 – 20 42 – 12 18 –

Laos Laos Laos

15 06 55.45 15 06 55.45

105 48 10.01 105 48 10.01

LA(MC)-1 LA(MC)-2

12 –

5 –

– 19

– 28

121

130

272

254

Total A, Bactrocera tau A; C, B. tau C; E, east; N, north; F, females; M, males.

132

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

Fig. 2. Geographic origin of the infested fruits in Thailand (CM, Chiang Mai; CN, Chainat; KK, Khon Kaen; KP, Kamphaeng Phet; LO, Loei; MS, Mae Hong Son; PE, Phetchabun; ST, Satun; UB, Ubon Ratchathani) and Laos (LA).

From the sample of 384 females, 130 females were unequivocally closer to reference images of B. tau A, and 254 females to species B. tau C (Table 1). From the 393 males, 121 specimens were similar to the reference images of B. tau A, and 272 to B. tau C (Table 1). In the total sample, including reference images, in males as well as in females, two groups were clearly apparent (Fig. 3, see also Figs. 4 and 5) which corresponded to the specimens attributed above to either species B. tau A or C. The remaining analyses were thus performed considering these two groups as two different species, namely species B. tau A and B. tau C. Between the Procrustes- and Mahalanobis-based classifications, the disagreement was 1% in males and 3% in females.

3.2. Size variation between and within species Disregarding the geographic origin of the samples, 59% of the interspecific comparisons were significant in males and 51% in females (after Bonferroni correction). This size difference pattern was not consistent across localities; in some locations, such as C6 and P3 for males, and C6 and LA for females, there was no detectable difference between species. Considering the total sample of each

Fig. 3. First (horizontal axis) and second (vertical axis) relative warps (RW) of female (top) and male (bottom) insects when the specimens collected in Thailand and Laos are combined with the reference specimens for B. tau A and C (Kitthawee and Dujardin, 2010). In each sex, the external polygon represents the limits of the total sample, the internal polygons (arrows) represent the limits of the reference samples (9 females of B. tau A and 14 females of B. tau C; 18 males of B. tau A and 20 males of B. tau C).

species, species C tended to be larger than species A (Figs. 4 and 5); this was significant in males (P < 0.0010) but not in females. In males as well as in females, size variances differed between species only when comparing species collected from different localities (15% of significant comparisons in males, 5% in females, details not shown). Comparing the total of all individuals in each species allowed us to detect a significantly larger variance of size (CSV) in species A in both sexes (Table 2, vertical reading). In both species, centroid size significantly differed according to locality, sex, and their interactions (see the ANOVA output, Table 3). This was confirmed by the non-parametric analyses: considering the total sample in each species, sexual size dimorphism (SSD, Table 4) was significant (P < 0.0001), with females larger than males. Mean size was significantly different between localities in 62% of species A males, 51% of species C males, 52% of species A females and 47% of species C females. Within species A, the size variance (CSV) was significantly greater in females (P < 0.001), while no such sexual difference could be detected in species C (Table 2, horizontal reading). CSV was significantly different between localities in 38% of species A males, 7% of species C males, 0% of species A females and 2% of species C females.

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

133

Fig. 4. Males of B. tau. Left: the UPGMA tree according to the Mahalanobis distances. Right: quantile plots of centroid size variation. Each box shows the median as a line across the middle and the quartiles (25th and 75th percentiles) at its ends. Units are pixels. Abbreviations: C2, C4, C5, C6, four Chiang Mai localities; CN, Chainat; KK, Khon Kaen; KP, Kamphaeng Phet; LO, Loei; MS, Mae Hong Son; P1, P3, two Phetchabun localities; SL, Satun; UB, Ubon Ratchathani; LA, Laos.

3.3. Shape variation between and within species and population structure analysis Shape variation was illustrated with UPGMA trees obtained from Mahalanobis distances (Figs. 4 and 5). The Procrustes distance between species A and C was 0.033 in females, and 0.042 in males. In each species, the ANCOVA analysis (Table 3) focused on the first three relative warps (RW1, 2 and 3) representing a total of 70% and 65% of the total shape variation for species A and C, respectively. Regarding RW1, the results were similar for both species, revealing a non-significant effect of size on shape, but significant effects of sex and geography and their interaction. RW2 produced similar results, but was significantly affected by size in species C (P = 0.0006). RW3

was weakly influenced by sex in species A (P = 0.0473), with no significant interaction of sex differences and localities (P = 0.1419). The Procrustes distance between sexes in each locality, as listed in Table 4 (see column SShD), was generally higher in species A (d = 0.050) than in species C (d = 0.032); this difference was tested by non-parametric tests (permutations, 1000 cycles) and found to be significant (P < 0.002). It was generally commensurate with or larger than the interspecific Procrustes distance (0.033–0.042, see above). The variance of shape was estimated by the metric disparity (MD). In species A, MD was 0.00079 for males and 0.00066 for females. This was in agreement with previous data of species A infesting other fruits species, where MD ranged from 0.00062

Table 2 Variance estimators for shape (MD) and centroid size (CSV) in each sex of Bactrocera tau species A and species C. Statistical comparisons between sexes (horizontal reading) and between species (vertical reading). MD

B. tau A B. tau C P

CSV

M

F

0.00079 0.00052 <0.001

0.00066 0.00049 <0.001

P <0.001 0.140 –

M

F

11,671 4541 <0.001

61,145 3721 <0.001

CSV, centroid size variance; MD, metric disparity; F, females; M, males; P, significance after non-parametric tests.

P <0.001 0.080 –

134

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

Fig. 5. Females of B. tau. Left: the UPGMA tree according to the Mahalanobis distances. Right: quantile plots of centroid size variation. Each box shows the median as a line across the middle and the quartiles (25th and 75th percentiles) at its ends. Units are pixels. Abbreviations: C2, C4, C5, C6, four Chiang Mai localities; CN, Chainat; KK, Khon Kaen; KP, Kamphaeng Phet; LO, Loei; MS, Mae Hong Son; P1, P3, two Phetchabun localities; SL, Satun; UB, Ubon Ratchathani; LA, Laos.

to 0.00088 in males and from 0.00062 to 0.00073 in females (Kitthawee and Dujardin, 2010). MD was significantly higher in species A (Table 2). Within species C, and only there, no significant differences of MD could be detected between sexes. Species C males and females showed a significant and positive correlation (P < 0.0228 and P < 0.0104, respectively) between geographic distances (Table 5) and Mahalanobis distances (derived

from PW), while species A males and females did not (P > 0.37 and P > 0.38, respectively) (Table 6). 3.4. Allometry The allometric study was conducted within species. As shown by the ANCOVA (Table 3), size had no significant effect at all on RW1–3

Table 3 ANOVA (CS) and ANCOVA (RW1–3): effect of sex and locality (Loc) on size (ANOVA); effect of size, sex and locality on shape (ANCOVA). CS F Species A ANOVA CS ANCOVA RW1 (50%) RW2 (11%) RW3 (09%) Species C ANOVA CS ANCOVA RW1 (40%) RW2 (14%) RW3 (11%)

Sex P

3.40 1.02 3.46

1.15 12.05 2.24

0.0666 0.3134 0.0643

0.2835 0.0006 0.1346

F

Sex × Loc

Loc P

F

P

F

P

82.25

<0.0001

90.17

<0.0001

18.67

<0.0001

1036.14 154.40 3.98

<0.0001 <0.0001 0.0473

4.15 4.94 3.76

0.0013 0.0003 0.0028

4.75 3.38 1.67

0.0004 0.0059 0.1419

496.22

<0.0001

38.92

<0.0001

26.67

<0.0001

1481.37 475.08 8.87

<0.0001 <0.0001 0.0030

4.57 19.92 6.8

<0.0001 <0.0001 <0.0001

3.84 17.99 10.10

0.0002 <0.0001 <0.0001

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138 Table 4 The data shown here test for a correlation between sexual shape dimorphism (SShD) and sexual size dimorphism (SSD) in Bactrocera tau. For each sample, SSD is the absolute value of the size difference between sexes, and SShD is the corresponding shape (Procrustes) distance. To test for an allometric influence, the regression coefficient of SShD on SSD was computed and was not significant (n = 18, F(1, 16) = 0.68, P = 0.4204, R-squared = 0.0410). Locality

Species

SShD

SSD

P(SSD)

CM CM KP PE PE UB LA CM CM CM CM CN KK LO MS PE ST LA

A A A A A A A C C C C C C C C C C C

0.050 0.049 0.055 0.064 0.055 0.042 0.055 0.029 0.032 0.032 0.032 0.032 0.028 0.033 0.028 0.029 0.041 0.041

59 176 2 136 106 43 14 17 82 85 61 63 23 61 89 90 109 87

0.000 0.000 0.930 0.007 0.000 0.000 0.639 0.337 0.000 0.000 0.000 0.004 0.277 0.000 0.000 0.000 0.000 0.000

135

Table 6 Population structure of Bactrocera tau. Non-parametric Mantel test estimating the statistical significance of the correlation between geographic and shape distances separating the provinces. Species

Sex

N

R

Z ≥ obs Z

B. tau A

Male Female

5 5

0.0811 0.0797

0.3737 0.3792

B. tau C

Male Female

8 8

0.7193 0.7031

0.0228* 0.0104*

* Statistically significant at P = 0.05. N, Number of provinces; R, correlation coefficient; Z ≥ obs Z, number of times pseudo Z was equal to, or larger than, observed Z after 1000 runs, with Z the standardized correlation coefficient.

Table 7 The 1F1L configuration refers to interspecific comparisons of specimens developed in the same fruit species in the same locality; 2F1L in two different fruit species but in the same locality, etc. N is the number of interspecific comparisons corresponding to a given configuration. For shape comparisons, values are mean Procrustes distances between species. For size, values are the absolute differences of centroid size between species.

P(SSD) = significance level of SSD after 1000 permutations.

in species A, but influenced RW2 of species C. The contribution of size variation to the first two discriminant functions among geographic locations ranged from 0% to 2% in species C females, from 0% to 9% in C males, from 2 to 19% in species A females, and from 7% to 12% in A males (detailed analysis not shown). From the values of shape (SShD) and size (SSD) sexual dimorphism as reported in Table 4 for each locality (n = 18), we computed the regression coefficient of SShD on SSD and could not detect any significance (F[1,16] = 0.68, P = 0.4204, R-squared = 0.0410). We also computed the correlation coefficient of corresponding shape (MD) and size (CSV) variances (detailed analyses not shown), and did not detect any statistical significance.

Configuration

N

1F1L 2F1L 1F2L 2F2L 1L (1 or 2F) 2L (1 or 2F) 1F (1 or 2L) 2F (1 or 2L)

4 4 55 32 8 87 59 36

a

Shape

Size

Females

Males

Females

Males

0.0325 0.0410 0.0358a 0.0394a 0.0367 0.0371 0.0356 0.0395

0.0405 0.0473 0.0444 0.0452 0.0439 0.0447 0.0441 0.0454

59.79 71.49 71.06 47.73 65.64 62.48 70.30 50.37

101.99 84.31 78.93 59.21 93.15 71.67 80.49 61.99

Denotes statistically different values (P < 0.0003). F, fruit(s); L, locality(ies).

4. Discussion In previous studies, B. tau A and C were investigated using different techniques, including modern morphometrics, with the objective of clarifying the group’s systematics (Baimai et al., 2000; Jamnongluk et al., 2003a; Thanaphum and Thaenkham, 2003; Saelee et al., 2006; Kitthawee and Dujardin, 2010). Here we used the morphometric approach to perform a comparative population structure study within the two species. We explored their morphometric variability in relation to sex and geography when the same host plant was shared during preimaginal development. By selecting samples of B. tau A and C infesting the same fruit species, we excluded the influence of the host and focused on geography. Since no reliable climatic data were available for our samples, the geographic effects on shape were quantified using distances between provinces. In addition to these data, we also made use of previously published data to further explore the idea of possible character displacement as suggested by Kitthawee and Dujardin (2010).

3.5. Character displacement hypothesis Table 7 presents the complete set of comparisons testing for a significant increase of morphological divergence in sympatry. The results showed opposite trends for shape and size. Interspecific size differences increased when in sympatry, be it microsympatry (development in the same kind of fruit) or geographic sympatry (development in the same locality). This trend was consistent and particularly clear for males. Interspecific shape differences did not show any significant changes whether or not the specimens were found in sympatry, but interspecific Procrustes distances were found to be systematically (but only slightly) lower in sympatry (see Table 7).

Table 5 Distance in kilometers between the provinces in Thailand (CM, Chiang Mai; CN, Chainat; KK, Khon Kaen; KP, Kamphaeng Phet; LO, Loei; MS, Mae Hong Son; PE, Phetchabun; ST, Satun; UB, Ubon Ratchathani) and Laos (LA) where Bactrocera tau specimens were collected.

CN KK KP LO MS PE ST UB LA

CM

CN

KK

KP

LO

MS

PE

ST

UB

405.7 379.3 247.6 312.8 133.4 312.3 1319 736 820.4

250.7 167.1 308.8 513.2 215.2 925 511.4 609.1

276.9 96.92 512.4 74.55 1104 359.5 448.8

280.1 346.9 206.9 1072 613.9 709.2

445.5 93.61 1194 427.1 508.4

444.3 1404 869.4 953.6

1111 432.3 522.7

1055 1106

97.72

LA

136

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

4.1. Shape, species and geography In spite of significant variation of size between localities in around 50% of the comparisons, the contribution of size difference to shape divergence among localities was low (species A) or not significant (species C). The geographic shape heterogeneity therefore must have causes other than variation in size. For each species and separately for each sex, we explored the isolation-by-distance model hypothesis by performing Mantel tests between metric and geographic distances. A positive and significant correlation with geographic separation was disclosed for species C only. In this species, genetic drift, probably enhanced by geographic isolation, may appear as a general explanation for shape divergence among localities. The situation was different in species A, for which no correlation was found between metric and geographic distances. Diversifying selection as opposed to genetic drift might be a valid explanation for species A, but much less for species C because there is no reason to expect the diversifying effect of selection to increase with geographic distance. Since Bactrocera species are known to be able to fly long distances (50–100 km) (Fletcher, 1989; Chen et al., 2006), the different population structures of species A and C might be related to different dispersive capacities. However, we do not know the difference between the flying capacities of both species. The different population structure of species A and C regarding shape could also have other causes than geographic isolation. Actually, the two species also showed different size and shape variance (Table 2), and different size and shape sexual dimorphism (Table 4). Species A is easy to rear in the laboratory, since it is adapted to at least five different fruits (Kitthawee and Dujardin, 2010). For shape as well as for size, species A showed greater variability than species C (Table 2). This observation was in accordance with the host influence on shape (Kitthawee and Dujardin, 2010), an influence liable to explain a more variable morphology in a more generalist species. In addition, infesting different fruits may also mean more possibilities to be passively transported due to commercial activities. For this species, we therefore expect both lower geographic structuring and greater variation, including greater sexual dimorphism. In contrast, species C is difficult to rear in the laboratory and it is known to be a much more specialized species: its morphology is adapted to a single host (M. cochinchinensis), and its possibility of long-distance migration depends on a single host only. Since more host species may be related to more passive transportation opportunities, gene flow is expected to be greater in species A than in species C, resulting in more spatial structuring for species C. 4.2. Shape, species and sex Sexual dimorphism is generally explained by selection on reproductive traits (Fairbain, 1997). To maximize female encounter rates, selection might favor a faster developmental rate in males, a mechanism that may explain the smaller size of males, or selection might favor larger females to increase fecundity, or it might favor larger size, hence larger wings, in the more dispersing sex. None of these mechanisms could be clearly verified in Tephritidae (Sivinski and Dodson, 1992). An alternative hypothesis, not excluding the effect of selection on reproductive traits, proposes that sexual dimorphism evolves to reduce intraspecific competition for food (Fairbain, 1997). Higher values of sexual size and shape dimorphism were found in species A. In addition, and in contrast to species C, species A also showed statistically different size variance and different shape variance between sexes. The possible causes of variance dimorphism are unclear, but this observation is intuitively expected in a more generalist species. Since species A is able to develop in

diverse host plants, there is a greater chance of finding a differential effect of food on sexes or differential resource utilization by the sexes or intersexual dietary divergence. Moreover, different studies in other insects have shown that males and females can exhibit different responses to food quality/quantity (Mackauer, 1996; Blanckenhorn, 1997; Morin et al., 1999) and/or temperature (Blanckenhorn, 1997; Morin et al., 1999; Fischer and Fiedler, 2000, 2001). SShD has generally been attributed to allometry in Drosophila (Gidaszewski et al., 2009; Baylac and Penin, 1998; Gilchrist et al., 2000; Debat et al., 2003). However, SShD could not be satisfactorily interpreted herein as an allometric effect. Since there was no significant correlation between observed SSD and SShD values across localities (Table 4), i.e., no detectable influence of size variation on shape variation, non-allometric components probably played a role in the SShD of Bactrocera. Studies in other organisms – including another Tephritidae genus (Marsteller et al., 2009) – also recognized non-allometric components for SShD (O’Higgins and Collard, 2002; Schaefer et al., 2004; Leigh, 2006; Gidaszewski et al., 2009). Another sex-related observation was that the mean shape divergence between species was higher among males (Procrustes distance = 0.043) than among females (Procrustes distance = 0.033; P < 0.001). Greater interspecific variance in one sex than in the other is predicted by the “correlational selection” argument which proposes that the selective forces acting on traits of one sex are not independent of the traits of the other sex (Fairbain, 1997). But the reason why it is the male sex that is affected seems clearly related to sexual selection (Marsteller et al., 2009). When resources are relatively common so that females can easily escape male harassment by changing from one place to another, males have to develop attractive strategies (Lloyd and Sivinski, 2004). Males of some Tephritidae species use the wings as vibratory and/or visual displays during courtship (Sivinski et al., 2000). For instance, Ceratitis capitata (Tephritidae) males produce a buzzing sound with their wings as they attempt to mate, and females select as mates those with louder sounds (Lloyd and Sivinski, 2004). If wing shape affects courtship displays, it should differ between species, and especially between males. In addition to explaining the previous observation, this hypothesis would also account for inducing sexual dimorphism (Andersson, 1994). 4.3. Sympatry and allopatry The initial definition of the character displacement concept (Wilson and Brown, 1955; Brown and Wilson, 1956) did not anticipate the actual complexity of its demonstration: “. . . the situation in which, when two species overlap geographically, the differences between them are accentuated in the zone of sympatry.” Specifically, in sympatry, selection was assumed to minimize attempts at hybridization (by mistaken identity) as well as competition between the two species. The problems of obtaining unambiguous evidence from natural observations have been discussed by Grant (1972), and more recently by Losos (2000). Typically, character displacement was suspected when a greater difference was observed between species developing in sympatry than in allopatry; “soon after the theory was promulgated, ecologists and evolutionary biologists were seeing evidence for character displacement everywhere” (Losos, 2000). Herein, we focused on the very first condition for diagnosing character displacement: a greater difference in sympatry than in allopatry, more than would be expected by chance. We had suggested the hypothesis of character displacement as an explanation of interspecific size variation because of nonoverlapping sizes between species A and C infesting the same fruit species (Kitthawee and Dujardin, 2010). It can be seen from Figs. 4 and 5 that a non-overlapping size variation within the same fruit species was not confirmed here. However, when comparing

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

different sympatry and allopatry configurations by combining previously published and present data (see Table 7), we observed that interspecific size differences were systematically greater in sympatry (with a single exception in females). Even though the statistical significance of these differences was > 5% (around P = 10%), the trend was consistent. Thus, size differences between these two species could still be explained as a character displacement (Losos, 2000). There was no such evidence for interspecific shape differences. Furthermore, Procrustes distances between species were systematically lower when infesting the same fruit species or when occupying the same locality. This greater shape homogeneity in the same fruit species was highly significant in females (P < 0.0003). Thus, as already observed (Kitthawee and Dujardin, 2010), the host is shaping the insects. For shape, the very first condition necessary for assuming character displacement, greater difference in sympatry, was not verified here. 5. Conclusions In sum, the present study provided three main observations related to population structure, sexual dimorphism and the character displacement hypothesis: (i) the metric properties described here for species A and C of B. tau in Thailand and Laos fit the known biology of both species well, one a specialist species whose large-scale migration possibilities depend on a single host species, the other a generalist species with a greater probability of passive transportation due to various hosts; (ii) the sex-related observations were coherent with the model of sexual selection on courtship accelerating lineage shape diversification; (iii) the comparison of the present data with previously published data involving more host plants supported the hypothesis of character displacement for interspecific size differences, but not for shape divergence. Acknowledgments We thank Mr. Wimol Kijbamrung for providing specimens from Laos, and C. Apiwhatnasorn (Mahidol University, Bangkok, Thailand) for encouraging this work. This study was supported by the TRF/BIOTEC Special Program for Biodiversity Research and Training Grant BRT R 352052, the Commission on Higher Education, RMU5080060 (Thailand) and the Faculty of Science, Mahidol University, as well as by IRD grants number HC3165-3R165-GABIENT2 and HC3165-3R165-NV00-THA1. References Andersson, M., 1994. Sexual Selection. Princeton University Press, Princeton. ˜ H., Dujardin, J.P., Nil, R., Simard, F., Fontenille, D., 2011. ChroAyala, D., Caro-Riano, mosomal and environmental determinants of morphometric variation in natural populations of the malaria vector Anopheles funestus in Cameroon. Infect. Genet. Evol. 11, 940–947. Aytekin, A.M., Alten, B., Caglar, S., Ozbel, Y., Kaynas, S., Simsek, F.M., Kasap, O.E., Belen, A., 2007. Phenotypic variation among local populations of phlebotomine sand flies (Diptera: Psychodidae) in southern Turkey. J. Vector Ecol. 32, 226–234. Baimai, V., Phinchongsakuldit, J., Sumrandee, C., Tigvattananont, S., 2000. Cytological evidence for a complex of species within the taxon Bactrocera tau (Diptera: Tephritidae) in Thailand. Biol. J. Linn. Soc. 69, 399–409. Baylac, M., Penin, X., 1998. Wing static allometry in Drosophila simulans males (Diptera Drosophilidae) and its relationships with developmental compartments. Acta Zool. Acad. Sci. Hung. 44, 97–112. Baylac, M., Villemant, C., Simbolotti, G., 2003. Combining geometric morphometrics with pattern recognition for the investigation of species complexes. Biol. J. Linn. Soc. 80, 89–98. Blanckenhorn, W.U., 1997. Altitudinal life history variation in the dung flies Scathophaga stercoraria and Sepsis cynipsea. Oecologia 109, 342–352. Bookstein, F.L., 1991. Morphometric Tools for Landmark Data: Geometry and Biology. Cambridge University Press, New York. Broide, J.S., Dujardin, J.P., Kitron, U., Gürtler, R.E., 2004. Spatial structuring of Triatoma infestans (Hemiptera: Reduviidae) populations from northwestern Argentina using wing geometric morphometry. J. Med. Entomol. 41, 643–649. Brown, W.L., Wilson, E.O., 1956. Character displacement. Syst. Zool. 5, 49–64.

137

˜ H., Jaramillo, N., Dujardin, J.P., 2009. Growth changes in Rhodnius Caro-Riano, pallescens under simulated domestic and sylvatic conditions. Infect. Genet. Evol. 9, 162–168. Ceballos, L.A., Piccinali, R.V., Marcet, P.L., Vazquez-Prokopec, G.M., Cardinal, M.V., Schachter-Broide, J., Dujardin, J.P., Dotson, E.M., Kitron, U., Gürtler, R.E., 2011. Hidden sylvatic foci of the main vector of chagas disease Triatoma infestans: threats to the vector elimination campaign? PLoS Negl. Trop. Dis. 5 (10), e1365, http://dx.doi.org/10.1371/journal.pntd.0001365. Chen, C.-C, Dong, Y.-J., Li, C.-T., Liu, K.-Y., 2006. Movement of the oriental fruit fly, Bactrocera dorsalis (Hendel) (Diptera: Tephritidae), in a guava orchard with special reference to its population changes. Formosan Entomol. 26, 143–159. De la Riva, J., Le Pont, F., Ali, V., Matias, R., Mollinedo, S., Dujardin, J.P., 2001. Wing geometry as a tool for studying the Lutzomyia longipalpis (Diptera: Psychodidae) complex. Mem. Inst. Oswaldo Cruz 96, 1089–1094. Debat, V., Begin, M., Legout, H., David, J.R., 2003. Allometric and nonallometric components of Drosophila wing shape respond differently to developmental temperature. Evolution 57, 2773–2784. Drew, R.A.I., Roming, M.C., 1997. Overview – Tephritidae in the Pacific and SouthEast Asia. In: Allwood, A.J., Drew, R.A.I. (Eds.), Management of Fruit Flies in the Pacific. ACIAR Proceedings No. 76. Australian Center for International Agricultural Research, Canberra, pp. 43–46. Dujardin, J.P., 2008. Morphometrics applied to medical entomology. Infect. Genet. Evol. 8, 875–890. Dujardin, J.P., 2011. Modern morphometrics of medically important insects. In: Tibayrenc, M. (Ed.), Genetics and Evolution of Infectious Diseases. Elsevier, pp. 473–501. Dujardin, J.P., Costa, J., Bustamante, D., Jaramillo, N., Catalá, S., 2009. Deciphering morphology in Triatominae: the evolutionary signals. Acta Trop. 110, 101–111. Dujardin, J.P., Kaba, D., Henry, A.B., 2010. The exchangeability of shape. BMC Res. Notes 3, 266. Fairbain, D.J., 1997. Allometry for sexual size dimorphism: pattern and process in the coevolution of body size in males and females. Annu. Rev. Ecol. Syst. 28, 659–687. Fischer, K., Fiedler, K., 2000. Sex-related differences in reaction norms in the butterfly Lycaena tityrus (Lepidoptera: Lycaenidae). Oikos 90, 372–380. Fischer, K., Fiedler, K., 2001. Dimorphic growth patterns and sex-specific reaction norms in the butterfly Lycaena hippothoe sumadiensis. J. Evol. Biol. 14, 210–218. Fletcher, B.S., 1989. Life history strategies of tephritid fruit flies. In: Robinson, A.S., Hooper, G. (Eds.), Fruit Flies: Their Biology, Natural Enemies and Control. Elsevier, Amsterdam, pp. 195–208. ´ A., Milankov, V., 2009. Wing geometric morphometFrancuski, L., Ludoˇski, J., Vujic, ric inferences on species delimitation and intraspecific divergent units in the Merodon ruficornis group (Diptera Syrphidae) from the Balkan peninsula. Zool. Sci. 26, 301–308. Gidaszewski, N.A., Baylac, M., Klingenberg, C.P., 2009. Evolution of sexual dimorphism of wing shape in the Drosophila melanogaster subgroup. BMC Evol. Biol. 9, 110. Gilchrist, A.S., Azevedo, R.B., Patridge, L., O‘Higgins, P., 2000. Adaptation and constraint in the evolution of Drosophila melanogaster. Evol. Dev. 2, 114–124. Grant, P.R., 1972. Convergent and divergent character displacement. Biol. J. Linn. Soc. 4, 39–68. Gumiel, M., Catalá, S., Noireau, F., de Arias, A.R., Garcia, A., Dujardin, J.P., 2003. Wing geometry in Triatoma infestans (Klug) and T. melanosoma Martinez Olmedo and Carcavallo (Hemiptera: Reduviidae). Syst. Entomol. 28, 173–179. Hernández, M.L., Abrahan, L.B., Dujardin, J.P., Gorla, D.E., Catalá, S.S., 2011. Phenotypic variability and population structure of peridomestic Triatoma infestans in rural areas of the Arid Chaco (Western Argentina): spatial influence of macroand microhabitats. Vector-Borne Zoon. Dis. 11, 503–513. Jamnongluk, W., Baimai, V., Kittayapong, P., 2003a. Molecular phylogeny of tephritid fruit flies in the Bactrocera tau complex using the mitochondrial COI sequences. Genome 46, 112–118. Jamnongluk, W., Baimai, V., Kittayapong, P., 2003b. Molecular evolution of tephritid fruit flies in the genus Bactrocera based on the cytochrome oxidase I gene. Genetica 119, 19–25. Kitano, J., Mori, S., Peichel, C.L., 2012. Reduction of sexual dimorphism in streamresident forms of threespine stickleback Gasterosteus aculeatus. J. Fish Biol. 80, 131–146. Kitthawee, S., Dujardin, J.P., 2009. Diachasmimorpha longicaudata: reproductive isolation and geometric morphometrics of the wings. Biol. Control 51, 191–197. Kitthawee, S., Dujardin, J.P., 2010. The geometric approach to explore the Bactrocera tau complex (Diptera: Tephritidae) in Thailand. Zoology 113, 243–249. Leigh, S.R., 2006. Cranial ontogeny of Papio baboons (Papio hamadryas). Am. J. Phys. Anthropol. 130, 71–84. Lloyd, J.E., Sivinski, J., 2004. Sexual selection. In: Capinera, J.L. (Ed.), Encyclopedia of Entomology. Kluwer Academic Publishers, Dordrecht, Boston, London, pp. 1992–1993. Losos, J.B., 2000. Ecological character displacement and the study of adaptation. Proc. Natl. Acad. Sci. U. S. A. 97, 5693–5695. Lyra, M.L., Hatadani, L.M., de Azeredo-Espin, A.M., Klaczko, L.B., 2009. Wing morphometry as a tool for correct identification of primary and secondary new world screwworm fly. Bull. Entomol. Res. 23, 1–8. Mackauer, M., 1996. Sexual size dimorphism in solitary parasitoid wasps: influence of host quality. Oikos 76, 265–272. Márquez, E.J., Jaramillo-O, N., Gómez-Palacio, A., Dujardin, J.P., 2011. Morphometric and molecular differentiation of a Rhodnius robustus-like form from R. robustus

138

J.-P. Dujardin, S. Kitthawee / Zoology 116 (2013) 129–138

Larousse 1927 and R. prolixus Stal, 1859 (Hemiptera, Reduviidae). Acta Trop. 120, 103–109. Marsteller, S., Adams, D.C., Collyer, M.L., Condon, M., 2009. Six cryptic species on a single species of host plant: morphometric evidence for possible reproductive character displacement. Ecol. Entomol. 34, 66–73. Matias, A., De la Riva, J.X., Torrez, M., Dujardin, J.P., 2001. Rhodnius robustus in Bolivia identified by its wings. Mem. Inst. Oswaldo Cruz 96, 947–950. Morin, J.P., Moreteau, B., Petavy, G., David, J.R., 1999. Divergence of reaction norms of size characters between tropical and temperate populations of Drosophila melanogaster and D. simulans. J. Evol. Biol. 12, 329–339. O’Higgins, P., Collard, M., 2002. Sexual dimorphism and facial growth in papionin monkeys. J. Zool. 257, 255–272. Perriere, G., Gouy, M., 1996. WWW-Query: an on-line retrieval system for biological sequence banks. Biochimie 78, 364–369. Roggero, A., Passerin d’Entrèves, P., 2005. Geometric morphometric analysis of wings variation between two populations of the Scythris obscurella species-group: geographic or interspecific differences? (Lepidoptera: Scythrididae). SHILAP Revista de Lepidopterologia 33, 101–112. Rohlf, F.J., 1990. Rotational fit (Procrustes) methods. In: Rohlf, F., Bookstein, F. (Eds.), Proceedings of the Michigan Morphometrics Workshop. University of Michigan Press, Ann Arbor, pp. 227–236. Rohlf, F.J., Marcus, L.F., 1993. A revolution in morphometrics. TREE 8, 129–132. Ruangsittichai, J., Apiwathnasorn, C., Dujardin, J.P., 2011. Interspecific and sexual shape variation in the filariasis vectors Mansonia dives and Ma. bonneae. Infect. Genet. Evol. 11, 2089–2094. Saelee, A., Tigvattananont, S., Baimai, V., 2006. Allozyme electrophoretic evidence for a complex of species within the Bactrocera tau group (Diptera Tephritidae) in Thailand. Songklanakarin J. Sci. Technol. 28, 249–257.

Schaefer, K., Mitteroecker, P., Gunz, P., Bernhard, M., Bookstein, F.L., 2004. Craniofacial sexual dimorphism patterns and allometry among extant hominids. Ann. Anat. 186, 471–478. Sivinski, J., Dodson, G., 1992. Sexual dimorphism in Anatrepha suspensa (Loew) and other tephritid fruit flies (Diptera: Tephritidae): possible roles of developmental rate, fecundity, and dispersal. J. Insect Behav. 5, 491–506. Sivinski, J., Aluja, M., Dodson, G., Freidberg, A., Headrick, D., Kaneshiro, K., Landolt, P., 2000. Topics in the evolution of sexual behavior in the Tephritidae. In: Aluja, M., Norbom, A.L. (Eds.), Fruit Flies (Tephritidae): Phylogeny and Evolution of Behavior. CRC Press LLC, Boca Raton, FL, pp. 751–792. Thanaphum, S., Thaenkham, U., 2003. Relationships of forms within the Bactrocera tau (Walker) (Diptera: Tephritidae) taxon based on heat shock protein 70 cognate sequences. Ann. Entomol. Soc. Am. 96, 44–53. Villegas, J., Feliciangeli, M.D., Dujardin, J.P., 2002. Wing shape divergence between Rhodnius prolixus from Cojedes (Venezuela) and R. robustus from Mérida (Venezuela). Infect. Genet. Evol. 2, 121–128. Villemant, C., Simbolotti, G., Kenis, M., 2007. Discrimination of Eubazus (Hymenoptera Braconidae) sibling species using geometric morphometrics analysis of wing venation. Syst. Entomol. 32, 625–634. White, I., Elson-Harris, M., 1992. Fruit Flies of Economic Significance: Their Identification and Bionomics. CAB International, Wallingford, UK. Wilson, E.O., Brown, W.L., 1955. Revisionary notes on the sanguinea and neogagates groups of the ant genus Formica. Psyche (Stuttgart) 62, 108–129. Yang, P., Carey, J., Dowell, R., 1994. Tephritid fruit flies in China: historical background and current status. Pan-Pacific Entomol. 70, 159–167. Zelditch, M.L., Swiderski, D.L., Sheets, H.D., Fink, W.L., 2004. Geometric Morphometrics for Biologists: A Primer. Elsevier, Academic Press, New York.