Research Article |
Corresponding author: Massimiliano Virgilio ( massimiliano.virgilio@africamuseum.be ) Academic editor: Teresa Vera
© 2015 Joannes Van Cann, Massimiliano Virgilio, Kurt Jordaens, Marc De Meyer.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Van Cann J, Virgilio M, Jordaens K, De Meyer M (2015) Wing morphometrics as a possible tool for the diagnosis of the Ceratitis fasciventris, C. anonae, C. rosa complex (Diptera, Tephritidae). In: De Meyer M, Clarke AR, Vera MT, Hendrichs J (Eds) Resolution of Cryptic Species Complexes of Tephritid Pests to Enhance SIT Application and Facilitate International Trade. ZooKeys 540: 489-506. https://doi.org/10.3897/zookeys.540.9724
|
Previous attempts to resolve the Ceratitis FAR complex (C. fasciventris, C. anonae, C. rosa, Diptera, Tephritidae) showed contrasting results and revealed the occurrence of five microsatellite genotypic clusters (A, F1, F2, R1, R2). In this paper we explore the potential of wing morphometrics for the diagnosis of FAR morphospecies and genotypic clusters. We considered a set of 227 specimens previously morphologically identified and genotyped at 16 microsatellite loci. Seventeen wing landmarks and 6 wing band areas were used for morphometric analyses. Permutational multivariate analysis of variance detected significant differences both across morphospecies and genotypic clusters (for both males and females). Unconstrained and constrained ordinations did not properly resolve groups corresponding to morphospecies or genotypic clusters. However, posterior group membership probabilities (PGMPs) of the Discriminant Analysis of Principal Components (DAPC) allowed the consistent identification of a relevant proportion of specimens (but with performances differing across morphospecies and genotypic clusters). This study suggests that wing morphometrics and PGMPs might represent a possible tool for the diagnosis of species within the FAR complex. Here, we propose a tentative diagnostic method and provide a first reference library of morphometric measures that might be used for the identification of additional and unidentified FAR specimens.
Ceratitis anonae , Ceratitis fasciventris , Ceratitis rosa , fruit flies, cryptic species, integrative taxonomy, wing morphometrics, Posterior Group Membership Probability
“True” fruit flies (Diptera, Tephritidae) are represented by more than 4,000 phytophagous species of which 25–30% feed on fruits. Many of them include major agricultural pests affecting crop production (
Previous attempts to separate the three morphospecies using alternative diagnostic methods showed contrasting results. Both PCR - RFLP (
This paper aims at exploring morphometric differentiation between FAR morphospecies and genotypic clusters in an integrative taxonomic framework (see
We considered a set of 227 specimens genotyped at 16 polymorphic microsatellites (
A preliminary methodological experiment aimed at quantifying morphometric differences between sexes (see
PERMANOVAs revealed differences in multivariate patterns of 7 male and 7 female C. rosa specimens for both wing landmarks and wing band areas (Tab.
Preliminary methodological control. PERMANOVA testing differences in multivariate patterns of wing landmarks and wing band areas of 14 C. rosa specimens in response to: sex (male / female), wing (right / left), image (1, 2) and measure (A, B). d.f.: degrees of freedom; MS: mean square estimates; F: pseudo-F. Probability of Monte Carlo simulations: n.s.: not significant a P<0.05; ***: P<0.001.
wing landmarks | wing band areas | ||||||
---|---|---|---|---|---|---|---|
d.f. | MS | F | MS | F | |||
Sex = S | 1 | 0.0220 | 68.64 | *** | 0.1390 | 66.23 | *** |
Wing = W | 1 | 0.0001 | 0.30 | n.s. | 0.0006 | 0.28 | n.s. |
Image = I(W) | 2 | 0.0000 | 0.10 | n.s. | 0.0011 | 0.52 | n.s. |
Measure = M(IxW) | 4 | 0.0000 | 0.03 | n.s. | 0.0001 | 0.05 | n.s. |
S x W | 1 | 0.0001 | 0.45 | n.s. | 0.0006 | 0.29 | n.s. |
S x I(W) | 2 | 0.0000 | 0.06 | n.s. | 0.0001 | 0.03 | n.s. |
S x M(WxI) | 4 | 0.0000 | 0.04 | n.s. | 0.0001 | 0.04 | n.s. |
Residual | 96 | 0.0003 | 0.0021 |
The first two PC axes of PCAs including all specimens accounted for 57.9% and 47.6% of variation in males and females, respectively. PCAs of both wing landmarks (Suppl. material
The average individual PGMP within morphospecies (Figure
Individual assignments to morphospecies. Posterior group membership probabilities (PGMPs) of male and female specimens as resulting from Discriminant Analysis of Principal Coordinates of wing landmarks (upper) or wing band areas (lower). Prior groups: C. anonae (white), C. fasciventris (light blue), C. rosa (dark blue).
Diagnostic performance at different identification thresholds. Proportions of male and female specimens consistently assigned (a) to morphospecies (i.e. of specimens for which the highest posterior group membership probabilities (PGMP) corresponds to the prior morphospecies grouping) and (b) to genotypic cluster (i.e. of specimens for which the highest PGMP corresponds to the prior A, F1, F2, R1, R2 genotypic grouping) when considering wing landmarks (left) or wing band areas (right) and by using different assignment thresholds (no threshold, PGMP = 0.95, PGMP = 0.99).
PERMANOVAs evidenced significant interspecific differences in both wing landmarks and wing band areas (for both males and females, Table
PERMANOVA and a posteriori comparisons (t-statistic) testing differences in multivariate patterns of wing landmarks among morphospecies (C. anonae, C. fasciventris, C. rosa). d.f.: degrees of freedom; MS: mean square estimates; F: pseudo-F. Probability of Monte Carlo simulations: n.s.: not significant a P<0.05; ***: P<0.001, **: P<0.01; *: P<0.05 (after False Discovery Rate Correction for repeated a posteriori comparisons). | ||||||||
males | females | |||||||
d.f. | MS | F | d.f. | MS | F | |||
Morphospecies | 2 | 0.0021 | 2.34 | * | 2 | 0.0039 | 6.63 | *** |
Residual | 90 | 0.0009 | 54 | 0.0006 | ||||
Pair-wise a posteriori comparisons (males and females: lower and upper diagonal matrix, respectively). | |||
C. anonae | C. fasciventris | C. rosa | |
---|---|---|---|
C. anonae | - | *** | * |
C. fasciventris | n.s. | - | *** |
C. rosa | * | n.s. | - |
PERMANOVA and a posteriori comparisons (t-statistic) testing differences in multivariate patterns of wing landmarks among genotypic clusters (A, F1, F2, R1, R2). d.f.: degrees of freedom; MS: mean square estimates; F: pseudo-F. Probability of Monte Carlo simulations: n.s.: not significant a P<0.05; ***: P<0.001, **: P<0.01; *: P<0.05 (after False Discovery Rate Correction for repeated a posteriori comparisons). | ||||||||
males | females | |||||||
d.f. | MS | F | d.f. | MS | F | |||
Genotypic clusters | 4 | 0.0037 | 6.10 | *** | 4 | 0.0013 | 2.30 | ** |
Residual | 45 | 0.0006 | 10 | 0.0006 |
Pair-wise a posteriori comparisons (males and females: lower and upper diagonal matrix, respectively) | |||||
A | F1 | F2 | R1 | R2 | |
---|---|---|---|---|---|
A | - | n.s. | n.s. | n.s. | n.s. |
F1 | *** | - | * | n.s. | n.s. |
F2 | *** | n.s. | - | * | n.s. |
R1 | n.s. | ** | ** | - | n.s. |
R2 | ** | *** | ** | ** | - |
The preliminary methodological experiment showed significant differences between male and female C. rosa wing morphometrics and suggests that these differences are consistent across species. Wing size and shape is supposed to have a main role in visual and vibrational courtship displays of tephritid fruit flies and sexual dimorphism in wing morphometrics has been already shown in tephritids (e.g.
PERMANOVAs showed that C. fasciventris, C. anonae, C. rosa as well as their five genotypic clusters (
Distance thresholds are currently implemented in DNA barcoding identification of species (
This study suggests that wing morphometrics might represent a possible tool for the diagnosis of species within the Ceratitis FAR complex. In this respect, PGMP of individuals might be calculated and used to quantify the proximity of individuals to each morphospecies (see
The authors would like to acknowledge all colleagues and collaborators who contributed to the sample collection. The study was supported by FAO/IAEA via the Coordinated Research Project ‘Resolution of cryptic species complexes of tephritid pests to overcome constraints to SIT application and international trade’.
Morphometric data
Data type: morphometric data
Explanation note: Specimen list and details and raw morphometric data (for both wing landmarks and wing band areas).
Map of sampling locations
Data type: occurence
Explanation note: Number of sampled specimens for each morphospecies are indicated in parentheses.
Wing landmarks and wing band areas
Data type: species data
Explanation note: List of wing landmarks and wing band areas considered in this study.
Wing landmarks and wing band areas
Data type: species data
Explanation note: Position of wing landmarks and wing band areas (numbers according to Suppl. material
Preliminary methodological experiment: unconstrained ordination of wing landmarks
Data type: species data
Explanation note: Principal component analysis (PCA) showing morphometric differences in wing landmarks of 14 C. rosa specimens across sexes (M, F), wings (LW: left wing, RW: right wing), repeated images of the same wing (1, 2), repeated measures of the same image (A, B).
Preliminary methodological experiment: unconstrained ordination of wing band areas
Data type: species data
Explanation note: Principal component analysis (PCA) showing morphometric differences in wing band areas of 14 C. rosa specimens across sexes, wings (LW: left wing, RW: right wing), repeated images of the same wing (1, 2), repeated measures of the same image (A, B).
Unconstrained ordination of wing landmarks across sexes of each morphospecies
Data type: species data
Explanation note: Principal component analysis (PCA) showing morphometric differences in wing landmarks between sexes of each morphospecies (C. anonae, C. fasciventris, C. rosa). (all 227 specimens included).
Unconstrained ordination of wing landmarks
Data type: species data
Explanation note: Principal component analysis (PCA) showing morphometric differences in wing landmarks between males and females (a) C. anonae, C. fasciventris and C. rosa and (b) genotypic clusters A, F1, F2, R1, R2.
Unconstrained ordination of wing band areas
Data type: species data
Explanation note: Principal component analysis (PCA) showing morphometric differences in wing band areas between males and females (a) C. anonae, C. fasciventris and C. rosa and (b) genotypic clusters A, F1, F2, R1, R2.
Constrained ordination of wing landmarks
Data type: species data
Explanation note: Discriminant analysis of principal coordinates (DAPC) maximising morphometric differences in wing landmarks between males and females (a) C. anonae, C. fasciventris and C. rosa and (b) genotypic clusters A, F1, F2, R1, R2.
Constrained ordination of wing band areas
Data type: species data
Explanation note: Discriminant analysis of principal coordinates (DAPC) maximising morphometric differences in wing band areas between males and females (a) C. anonae, C. fasciventris and C. rosa and (b) genotypic clusters A, F1, F2, R1, R2.
Average individual assignments
Data type: species data
Explanation note: Average individual posterior group membership probabilities (PGMPs) of male and female specimens of three morphospecies (C. anonae, C. fasciventris and C. rosa) and five genotypic clusters (A, F1, F2, R1, R2) as resulting from the analysis of wing landmarks (light grey) and wing band areas (dark grey).
Individual assignments to genotypic clusters A, F1, F2, R1, R2
Data type: molecular data
Explanation note: Posterior group membership probabilities (PGMPs) of male and female specimens as resulting from Discriminant Analysis of Principal Coordinates of wing landmarks (upper) or wing band areas (lower). Prior groups: A, F1, F2, R1, R2 (from white to dark blue).
Morphometric differences across morphospecies (wing band areas)
Data type: morphometric data
Explanation note: PERMANOVA and a posteriori comparisons (t-statistic) testing differences in multivariate patterns of wing band areas among morphospecies (C. anonae, C. fasciventris, C. rosa). d.f.: degrees of freedom; MS: mean square estimates; F: pseudo-F. Probability of Monte Carlo simulations: n.s.: not significant a P<0.05; ***: P<0.001, **: P<0.01; *: P<0.05 (after False Discovery Rate Correction for repeated a posteriori comparisons).
Morphometric differences across genotypic clusters (wing band areas)
Data type: morphometric data
Explanation note: PERMANOVA and a posteriori comparisons (t-statistic) testing differences in multivariate patterns of wing band areas among morphospecies (C. anonae, C. fasciventris, C. rosa). d.f.: degrees of freedom; MS: mean square estimates; F: pseudo-F. Probability of Monte Carlo simulations: n.s.: not significant a P<0.05; ***: P<0.001, **: P<0.01; *: P<0.05 (after False Discovery Rate Correction for repeated a posteriori comparisons).