Genetica

, Volume 139, Issue 10, pp 1241–1249 | Cite as

Genetic variation in a female genital trait evolved by sexual coevolution

  • Georgina Jiménez Ambriz
  • Diana Mota
  • Carlos Cordero
Article

Abstract

Understanding the patterns of genetic variation of traits subject to sexual selection is fundamental for explaining its evolutionary dynamics and potential for sexual coevolution. The signa of female Lepidoptera are sclerotized structures located on the inner surface of the genital receptacle that receives the spermatophore during copulation (the corpus bursae), whose main function is tearing the spermatophore envelope. Comparative data indicate that the evolution of signa has been influenced by sexually antagonistic coevolution with spermatophore envelopes. We looked for additive genetic variation in the size and shape of signa in females of the butterfly Callophrys xami (Lycaenidae) from two localities (BG and FC) in Mexico City. We also looked for genetic variation in female body size and in the size of corpus bursae. There were significant between-population differences in female body size, signa width and three signa shape traits. We found significant extranuclear maternal effects in one component of signa shape in the BG population, and in body weight, signa length and in one uniform component of signa shape in the FC population. Extranuclear maternal contributions could permit the evolution of female adaptations even if these reduce male fitness. We found additive genetic variation in signa length and width only in one population (BG); heritability estimates were high: 0.96 and 0.8, respectively. The existence of additive genetic variation in signa size could be, at least in part, a result of relaxed sexually antagonistic selection pressures due to the low level of polyandry exhibited by this species. Our results imply that there is currently potential for further sexual coevolution in this trait.

Keywords

Female genitalia Heritability Sexual selection Lepidoptera 

Introduction

In insects, genitalia are especially useful for species identification because their degree of differentiation between related species is generally greater than that of non genital characters, indicating that they have evolved in a particularly rapid and divergent way (Eberhard 1985, 2010a, b). On the other hand, genitalia generally show a degree of complexity that seems unnecessary to accomplish the sole function of ejaculate transfer to the female (Waage 1979; Eberhard 1985, 2010a, b; Hosken and Stockley 2004; Jagadeeshan and Singh 2006). Genitalia are usually composed of multiple structures that appear to play different roles during sexual interactions (Eberhard 1985; Tuxen 1970; Jagadeeshan and Singh 2006). A number of studies have shown some of the functions of different genital structures (e.g. Waage 1979; Jagadeeshan and Singh 2006; Moreno-García and Cordero 2008; several papers in Joly and Schmitt 2010 and Leonard and Córdoba-Aguilar 2010). For example, Waage (1979) found that the distal segment of the ‘penis’ of a damselfly is used during copulation for removing the sperm from previous mates, and Eberhard (1996, 2010a) discussed compelling evidence that several structures of the male genitalia are used to stimulate females internally in a number of taxa. These studies, together with the morphological complexity observed in many animal taxa (Eberhard 1985; Tuxen 1970), suggest that selection, probably acting in multifarious ways (Joly and Schmitt 2010; Leonard and Córdoba-Aguilar 2010), has been the main force behind the rapid evolutionary divergence of genitalia. In particular, many authors consider that the process of sexual coevolution, resulting from the action of reciprocal selective pressures exerted by each sex on the other, has played a crucial role in genital evolution (Eberhard 1985; Hosken and Stockley 2004; Cordero 2005, 2010; Minder et al. 2005; Brennan et al. 2007; Joly and Schmitt 2010; Leonard and Córdoba-Aguilar 2010; Wedell and Hosken 2010; Sánchez et al. 2011).

Sexually selected traits are commonly subject to strong and continuous selection and, therefore, it is very important to know the amount and nature of genetic variation for these traits (Simmons and Moore 2009; Walsh and Blows 2009; Chenoweth and McGuigan 2010). This is relevant for different reasons. In the case of intrasexual selection, such as male-male combat (Darwin 1871) or sperm competition (Parker 1970; Simmons 2001), it is important to estimate the potential for escalated intrasexual evolution (Simmons 2001), whereas in the case of mate choice and sexual conflict is important to estimate the potential for sexual coevolution (Parker 1979; Alexander et al. 1997; Holland and Rice 1998; Hosken and Stockley 2004; Arnqvist and Rowe 2005; Minder et al. 2005; Brennan et al. 2007; Joly and Schmitt 2010; Leonard and Córdoba-Aguilar 2010; Wedell and Hosken 2010; Sánchez et al. 2011). In general, although the number of studies has increased in recent times (general reviews of sexually selected traits: Snook et al. 2010; Simmons and Moore 2009; Walsh and Blows 2009; Chenoweth and McGuigan 2010; insect genitalia: Arnqvist and Thornhill 1998; House and Simmons 2005; Higgins et al. 2009; Simmons et al. 2009; Kamimura and Iwase 2010; Schäfer et al. 2011), more information on genetic variation of phenotypic traits that are, or are believed to be, subject to strong sexual selection in a variety of model species is necessary to understand the dynamics of genetic variation in the face of different types of sexual selection (Simmons and Moore 2009; Walsh and Blows 2009; Chenoweth and McGuigan 2010).

The signa (plural of signum) are sclerotized structures located on the inner wall of the corpus bursae, a sac-like organ in which males deposit a spermatophore during copulation, and their main function is to tear or pierce the external wall of the spermatophore, thus allowing females access to the resources contained in it (Hinton 1964; Drummond 1984; Rogers and Welles 1984; Tschudi-Rein and Benz 1990; Cordero 2005; Galicia et al. 2008). Recently, Sánchez et al. (2011) found evidence in support of the hypotheses that the evolution of signa has been influenced by sexual coevolution. Briefly, the hypothesis is as follows: (1) since in polyandrous Lepidoptera female sexual receptivity is inversely related to the amount of spermatophore remaining in the corpus bursae, sperm competition could select for spermatophore envelopes difficult to break that increase the length of the period of diminished sexual receptivity; (2) signa evolved as female counter-adaptations to these hard-to-break spermatophore envelopes evolved by males in response to sperm competition. This hypothesis predicts that signa evolved in polyandrous taxa and were lost when selection favored the evolution of monandry in females. Sánchez et al. (2011) performed a comparative test of this prediction in which they found, as expected, a statistically significant association between evolution of monandry and loss of signa in a sample of 37 taxa. Since, according to the hypothesis, a continuous process of sexually antagonistic coevolution is the main force behind the evolution of signa, a low level of genetic variation in this trait is expected. Here, we provide estimates of the amount of genetic variation underlying differences in signa size and shape in the butterfly Callophrys xami (Reakirt) (Lycaenidae), as well as in corpus bursae length. Females of this species have a pair of spine-shaped signa located in opposite sides of the distal third of the corpus bursae (Fig. 1a). In this species, females are slightly polyandrous (mean ± SD of spermatophore number = 1.37 ± 0.6; Cordero 1999), while males are polygamous (Cordero and Soberón 1990; Cordero et al. 2000).
Fig. 1

a Corpus bursae of female Callophrys xami, showing one of the paired spine-shaped signa (S) through its transparent wall; our measure of corpus bursae length is indicated with the white arrow. DB ductus bursae. b Drawing of the C. xami consensus signum shape (obtained from the General Superimposition Procrustes Analysis) illustrating the measures of length (l) and width (w), and the landmarks (1, 4, 7) and semilandmarks (2, 3, 5, 6, 8–10) configuration, used in the relative warps analysis

Materials and methods

Species studied and laboratory rearing

Callophrys xami (Lycaenidae) is a multivoltine species that inhabits dry, rocky habitats of North America and whose larvae feed on several species of Crassulaceae; its main host plant in the study area is Echeveria gibbiflora DC. Fifteen isofemale lines were established by using females collected in the Pedregal de San Ángel nature preserve located in the Ciudad Universitaria campus of the Universidad Nacional Autónoma de México (CU-UNAM), in the south of Mexico City. Ten females were collected near the Jardín Botánico, five in January 2008 and five in February 2008 (there were not matings between the offspring of these two groups because the offspring of females collected in January were too old to mate when the offspring of females collected in February emerged), and are called here the BG “population”; other five females were collected in March 2008 near the Facultad de Ciencias (CU-UNAM) and are called here the FC “population”. We used a small number of isofemale lines because C. xami, although it is present throughout the year, it is usually rare and finding several females within a reasonably short period of time was not possible. Although both “populations” are within CU-UNAM, we conservatively decided to consider them as two populations because they are located in opposite sides of the large (eight lines) Insurgentes Avenue, one of the busiest roads in Mexico City; the distance between collecting sites was 1.6 km. Females were fed ad libitum with a 10% glucose solution every morning. To obtain eggs, females were introduced in white, translucent, one liter plastic containers with a leaf of E. gibbiflora, located under a 125 W light bulb during 1.5 h per day. Each egg was carefully transferred to a plastic petri dish (8 cm diameter × 1.1 cm height) that contained one 9 cm2 fragment of E. gibbiflora and a piece of paper towel on the bottom. Once eggs hatched, larvae were raised individually until pupation; leaf fragments were changed every three days to prevent food limitation. All E. gibbiflora leaves used in the study were collected in CU-UNAM. Pupae were kept individually in plastic petri dishes (8 cm diameter × 2.2 cm height) until eclosion. Petri dishes were kept at ambient temperature and a 12 light—12 dark photoperiod; the petri dishes were stored in two metal shelves in an insectarium. Development time (from oviposition to adult emergence) was about 2 months.

Experimental crosses

To estimate the genetic basis of variation in signa size and shape, we used an experimental protocol approaching a diallelic design; since we did not have inbred parental lines for all diallelic crosses (see below), we made a pedigreed population analysis to estimate genetic variance components (Fry 2004).

Crosses were performed in cylindrical mating cages (58 cm high × 26 cm diameter) made of mesh cloth and a framework of metal wire (Jiménez and Soberón 1988/1989); one female and two males were introduced in a cage and the cage was hung outdoors in a garden located in the CU-UNAM during sunny days (as is usual in this species, not all mating essays were successful). In the case of the BG population, offspring of each of the two groups of five families were first crossed with members of its own family—considering the relatively low level of polyandry observed in this species (Cordero 1999), most of these crosses probably were between full siblings—to produce five inbred lines. Diallelic crosses were obtained by crossing four females of each inbred line with males from each of the other four inbred lines. In the case of the FC population we did not have a first generation of inbred crosses but only five maternal families and, therefore, we were unable to perform a real diallelic design. Instead, we planned crosses approaching a diallelic design with the progeny of the five maternal families, and accounted for this in the mixed model used to estimate genetic variation (see below).

Morphological measurements

One day after emergence, each adult butterfly was carefully introduced, with the wings folded, in a glassine envelope where it was weighed in an analytical balance (Scientech™, model SA120; resolution 0.0001 g) and its left forewing length (from the axilar sclerites to the apex) measured with a caliper (Mitutoyo™; accuracy of 0.02 mm). Each butterfly was individually numbered on the ventral side of one of its posterior wings with a nontoxic marker (Sharpie™). Female progeny of all crosses were stored at −20°C until dissection. Length of left wing and weight at emergence were included in the genetic analysis as estimates of body size.

To measure female genital traits, the abdomens of the female progeny of the experimental crosses were dissected under a dissection microscope (Olympus™, model SZH10). The corpus bursae was placed fresh on a scaled slide (resolution 0.01 mm) and photographed with a digital camera (Canon™, model PowerShot A640) mounted on the dissection microscope. Then, the corpus bursae was opened, the inner area where the signa are located was held flat on a glass slide, and the two signa were photographed with the same camera mounted on a microscope (Carl Zeiss™, model Primo Star). On the digital photographs we measured: (1) corpus bursae length (CBL) as the distance between the two most proximal (the “bottom” of the bursae) and distal (the union with the ductus bursae) points of the corpus bursae (Fig. 1a); (b) signum length as the distance between the apical point and the middle point of a line joining the two most basal points (Fig. 1b); and (3) signum width as the length of the longest line perpendicular to signum length (Fig. 1b). Digital photographs were processed with the tpsDig software v.2.12 (Rohlf 2008b). Data from both signa were included in the mixed model as repeated measures of each individual butterfly.

Geometric morphometrics was used to obtain quantitative estimates of signa shape. Three 2- and 3-type landmarks and seven semilandmarks (Bookstein 1991) were digitized on signa images with the tpsDig software (Rohlf 2008b; Fig. 1). Using the tpsRelw v.1.46 software (Rohlf 2008a), we got the consensus configuration of landmarks by a General Procrustes Superimposition Analysis (GPA; Fig. 1b). We also specified a slider file for sliding semilandmarks. After computation of principal and partial warps analysis from the bending energy matrix, we got the u1 and u2 uniform components of shape scores for each specimen; we used these scores for genetic analysis. We calculated centroid size as a form-independent measure of signa size. Since the relative warp analysis showed that 92.39% of the non-uniform component of shape was explained by the first four relative warps (RW1 to RW4), we used these warps in the genetic analysis (Table 1; Rohlf and Marcus 1993).
Table 1

Percentage of variation in the shape of signa explained by each consecutive relative warp

Relative Warp (RW)

% of variation

Cumulative % of variation

1

44.17

44.17

2

32.87

77.04

3

11.59

88.64

4

3.75

92.39

5

2.81

95.19

6

1.94

97.14

7

1.28

98.42

8

0.69

99.12

9

0.59

99.70

10

0.15

99.85

11

0.08

99.93

12

0.03

99.96

13

0.02

99.98

14

0.01

99.99

15

0.01

100.00

16

0.00

100.00

Statistical analysis and genetic variance estimation

As mentioned above, we used an experimental design approaching the diallelic design to estimate the additive and dominance genetic variance components, as well as paternal and maternal effects, on the measured traits. However, as not all our lines were inbred, we made a Pedigreed Population Analysis to estimate genetic variance components (Fry 2004). First, we used the INBREED procedure of SAS statistical software (SAS Institute 2002) to estimate the coefficients matrix of additive genetic variance () for the progeny of every possible pair of families resulting from the crosses. This matrix was used to estimate the fraternity coefficients matrix to calculate dominance genetic variance. The additive genetic coefficients and fraternity coefficients were used to analyze genetic variance components for the measured traits according to the Cockerman and Weir model for reciprocal crosses (1977). In this model, individual phenotypes of progeny from father i and mother j are calculated as:
$$ Y_{ijk} = \mu + N_{i} + N_{j} + T_{ij} + P_{i} + M_{j} + K_{ij} + W_{k(ij)} $$
Where i and j are the paternal and maternal families, respectively; Ni and Nj are the nuclear contribution from the paternal and maternal families, respectively; T is the interaction between both nuclear contributions; Pi and Mj are the extranuclear paternal and maternal contributions, respectively; Kij is the interacting extranuclear contribution from both families, and Wk(ij) is the specific effect of the K individual (Fry 2004). The variances of these effects are, respectively: σN2, σT2, σP2, σM2, σK2 and σW2. These variances were estimated using a mixed model with the MIXED procedure from SAS (SAS Institute 2002). We modeled the covariance between the different types of relatives as linear combinations of the variances corresponding to the variance components of the Cockerman and Weir model in this way:
$$ COV_{ij} = a_{ij} \sigma_{N}^{2} + b_{ij} \sigma_{T}^{2} + c_{ij} \sigma_{P}^{2} + d_{ij} \sigma_{M}^{2} + f_{ij} \sigma_{K}^{2} + e_{ij} \sigma w^{2} , $$
where COVij corresponds to covariance between progeny from i and j families, aij are the additive genetic coefficients; bij are the fraternity coefficients for every possible pair of families; and cij, dij, fij and eij values depend on the shared parents of the pair of families (Fry 2004). We specified this covariance structure in the MIXED procedure, as described by Fry (2004), to estimate σN2, σT2, σP2, σM2, σK2 and σW2 variances. As the relative family structure from this experiment was of first cousins and double first cousins, we estimated the genetic additive variance (VA) and the genetic dominance variance (VD) from σN2 = 1/8VA and σT2 = 1/16VD respectively, assuming that there are not significant epistatic interactions and that populations from which mothers were obtained were panmictic. All variance estimations were made with the Restricted Maximum Likelihood method (REML) which is adequate for unbalanced designs like ours. We tested if variance estimates were significantly different from zero with a likelihood ratio test (Shaw 1991). To do this, we analyzed the data with a model constraining the variance component to be tested to a value of zero, then we obtained the difference in the restricted log-likelihood between the constrained and the unconstrained model; twice this difference was then contrasted with the asymptotic Χ2 distribution (df = 1). We also incorporated in the model the original populations of maternal families. Since the lights used for inducing oviposition where in the top level of the metal shelves where the petri dishes containing eggs, larvae and pupae were stored, when the lights were on there was a gradient of light and temperature from top to bottom of the shelves. For this reason, we included a blocking factor describing the spatial distribution of petri dishes on the shelves as fixed factors. This procedure allowed us to estimate genetic effects by statistically eliminating potential “position effects” (Grafen and Hails 2002). The significance of fixed factors was tested with the “Type 3 Tests of Fixed Effects” performed by the MIXED procedure. We calculated the heritability of traits in which we found significant additive genetic variation by using the formula:
$$ h^{2} = \frac{{V_{A} }}{{V_{A} + V_{D} + V_{P} + V_{M} + V_{K} + V_{E} }} $$
Where VA is the additive genetic variance, VD is the dominance genetic variance, VP and VM are the variance attributable to paternal and maternal extranuclear contributions respectively, VK is the variance due to extranuclear genetic interactions and VE is the environmental variance. Standard errors for heritability estimates were calculated with the Delta method (Lynch and Walsh 1998).

Results

Differences between populations and possible effect of microenvironment

Adult body weight, left wing length, signa width, and three of the seven components of signa shape (RW2, RW3 and u2) were significantly larger in the BG population (Table 2). Since rearing conditions were uniform, these results suggest the existence of genetic differences between populations. Since genetic variance estimations are specific for the population and time in which they are estimated, we performed genetic variance computations separately for each population.
Table 2

Means and standard errors for each phenotypic trait measured in females from two populations (BG and FC) of the butterfly Callophrys xami

 

FC

BG

Source of Variation

 

Mean

SE

Mean

SE

Population

Block

Adult body weight (g)

0.039

0.001

0.050

0.001

51.25(1,371)d

3.02(4,371)b

Left wing length (mm)

15.68

0.079

16.59

0.060

37.78(1,369)d

1.9(4,369)

Corpus bursae length (mm)

2.121

0.018

2.055

0.014

3.78(1,322)a

1.61(4,369)

Signa length (mm)

0.260

0.004

0.262

0.003

0.21(1,634)

2.07(4,634)a

Signa width (mm)

0.084

0.001

0.089

0.001

6.99(1,633)c

2.03(4,633)b

Relative warp 1 (RW1)

−0.002

0.008

0.000

0.006

0.04(1,647)

1.03(4,647)

Relative warp 2 (RW2)

0.044

0.006

−0.027

0.005

73.61(1,647)d

1.49(4,647)

Relative warp 3 (RW3)

0.009

0.004

−0.006

0.003

7.18(1,647)c

7.21(4,647)d

Relative warp 4 (RW4)

−0.002

0.002

0.000

0.002

0.24(1,647)

0.9(4,647)

Centroid

244.029

4.928

243.063

3.625

0.02(1,648)

1.27(4,648)

Uniform component (u1)

−0.009

0.006

0.004

0.005

2.26(1,648)

1.25(4,648)

Uniform component (u2)

−0.013

0.005

0.010

0.003

15.03(1,648)d

4.65(4,648)d

The effect of population and position of the developing females within the metal shelves are given in the last two columns, respectively. F values (gl) were obtained from the type 3 test of fixed effects calculated with the MIXED procedure. Sample sizes: FC = 147, BG = 231. Key: aP < 0.1, bP < 0.05, cP < 0.01 and dP < 0.001

The spatial distribution of petri dishes with individual larvae on the shelves in the insectarium (“block” in Table 3) affected adult body weight, signa width and two components of signa shape (RW3 and u2). These results suggest plasticity in the expression of these traits to microenvironmental variation; however, our experiment was not designed to estimate genotype-environment interactions.
Table 3

Estimates of genetic variance for body size, corpus bursae size, and size and shape of signa for Callophrys xami butterflies belonging to two populations (Pop), BG and FC

Trait

Pop

VA

VD

VP

VM

VK

VE

CVA

CVD

CVP

CVM

CVK

CVE

h2

Adult body weight

BG

0

1.09E−20

0

0

0

0.0002c

0

2.136E−09

0

0

0

0.2497

0

 

FC

0

0

0

6.34E−05c

0

0.0002c

0

0

0

0.2083

0

0.324

0

Left wing length

BG

0

0

0.0289

4.17E−18

0

1.375c

0

0

0.0103

1.239E−10

0

0.0711

0

 

FC

0

0

0

0.0739

0

1.1268c

0

0

0

0.0173

0

0.0675

0

Corpus bursae length

BG

0

0

0

0.002

2.27E−19

0.076c

0

0

0

0.0221

2.328E−10

0.1347

0

 

FC

0

0

0

0

0

0.0727c

0

0

0

0

0

0.1274

0

Signa length

BG

0.0078c

0

2.51E−37

0.0003

0

0

0.3404

0

1.929E−18

0.0633

0

0

0.9666c

 

FC

0

0

0

0.0006a

0

0.0031c

0

0

0

0.0961

0

0.2111

0

Signa width

BG

0.0005b

0

0

0

0

0.0001c

0.2505

0

0

0

0

0.1214

0.8099c

 

FC

0

0

0

9.09E−06

0

0.0004c

0

0

0

0.0347

0

0.2288

0

Relative warp 1 (RW1)

FB

0

0

0

7.75E−05

0

0.0138c

0

0

0

−0.332

0

−4.4346

0

 

FC

0.0123

1.581E−05

4.02E−05

0.0002

0.0002

0.0079c

−7.8922

−0.2824

−0.4503

−1.1054

−0.895

−6.3229

0.5957

Relative warp 2 (RW2)

BG

0

0

0

0.0005

9.54E−06

0.0078c

0

0

0

−0.8652

−0.1165

−3.3289

0

 

FC

0

4.198E−06

9.91E−06

0

0.0002

0.0089c

0

0.0446

0.0685

0

0.3146

2.0627

0

Relative warp 3 (RW3)

BG

0

4.709E−19

0.0002

0.0006b

0

0.0033c

0

−6.2833

−1.3929

−2.2423

0

−5.2357

0

 

FC

0.0032

1.635E−06

0

0.0022

6.12E−06

0.0019c

−18.608

−0.4206

0

−15.4885

−0.8138

−14.4167

0.4356

Relative warp 4 (RW4)

BG

0

1.091E−19

0

0

4.98E−05

0.0012c

0

1.104E−06

0

0

−23.5804

−113.9935

0

 

FC

0

0

0

0

0

0.0011c

0

0

0

0

0

585.9739

0

Centroid

BG

17498.03

182362.36

0

1111.222

0

3693.5c

0.5493

1.7733

0

0.1384

0

0

0.08707

 

FC

0

0

0

194.785

0

4535.55c

0

0

0

0.0552

0

0.2666

0

Uniform component (u1)

BG

3.83E−21

0

2.15E−06

2.73E−06

0

0.0078c

9.52E − 09

0

0.2255

0.804

0

13.6343

4.9E − 19

 

FC

0

0

0

0.0016

0

0.01C

0

0

0

−1.9033

0

−4.7897

0

Uniform component (u2)

BG

0

0.013

1.91E−20

0.0003

0

0.0038

0

8.0842

9.806E−09

1.1415

0

0

0

 

FC

0

0

0

0.0013b

0

0.0035c

0

0

0

−13.1462

0

−21.5018

0

VA additive genetic variance, VD dominance genetic variance, VP and VM paternal and maternal extranuclear contributions, respectibly, VK interaction effect of extranuclear contributions, VE environmental effects. Coefficients of variation (CVi) for each of the i variance component and narrow sense heritabilities (h2) are also shown. Significant components are given in bold letter (key: aP < 0.05, bP < 0.01, cP < 0.001)

Genetic variation and heritability estimates

We found significant additive genetic variance for signa length and width in the BG population but not in the FC population (Table 3). We did not find significant additive genetic variance for body size traits, corpus bursae traits or signa shape traits in any population (Table 3). Heritability estimates for signa length and width in the BG population were substantial: 0.97 (SE = 0.09) and 0.81 (SE = 0.003), respectively.

We did not find significant dominance variance in any trait in any population (Table 3). Virtually all traits exhibited significant environmental variance (Table 3). Significant extranuclear maternal effects were detected in both populations. In the BG population extranuclear maternal effects were observed in one of the components of signa shape (relative warp RW3), whereas in the FC population they were observed in weight at emergence, signa length and in one uniform component of signa shape (u2). In both populations, sample size prevented the estimation of genetic covariances. When we pooled both populations and calculated covariances by using a simple model of pooled reciprocals without self crosses (Lynch and Walsh, 1998), none of the genetic covariances were significant (data not shown). However, the meaning of these estimates is questionable considering that the populations differed in three out of four signa and body size variables, and in three out of seven signa shape traits.

Discussion

In this study of signa, a female genital trait whose evolution has been influenced by sexual coevolution (Galicia et al. 2008; Sánchez et al. 2011), we found significant additive genetic variation and large heritability values in signa length and width in one population (BG) of the butterfly C. xami but not in another nearby population (FC; Table 3). The existence of additive genetic variation in the BC population seems unexpected since sexual selection is generally considered a strong directional selection force (Tomkins et al. 2004; Kotiaho et al. 2008). One possible explanation for this apparent paradox is that selection acting on signa in C. xami is very weak or null and thus the erosion of additive genetic variation has been very slow. The fact that females of this species tend to be monogamous (mean ± SD number of copulations per female is = 1.37 ± 0.6; Cordero 1999), which could decrease the strength of sexual selection for thick spermatophore envelopes is in agreement with this explanation. The observation of relatively thin spermatophore envelopes in C. xami (Sánchez and Cordero in preparation) is also consistent with this explanation. However, if this hypothesis is true, we do not have an explanation of why in the second population (FC) we failed to detect additive genetic variation and significant heritabilities for signa characteristics, considering that this population is contiguous and appears to have a mating system similar to that of the BG population (C. Cordero personal observation). A second hypothesis (suggested by an anonymous reviewer) for the existence of significant additive genetic variance in signa size in the BG population is that sexually antagonistic selection pressures (potentially the main type of selective force responsible for the sexual coevolution of signa; Cordero 2005; Sánchez et al. 2011) may help maintain additive genetic variation, if the “optimal” values of the coevolving traits of males (i.e. spermatophore envelope characteristics) and females (signa size and shape) continually shift. Once again, if this hypothesis is true, we do not have an explanation for the absence of additive genetic variation in signa characteristics in the FC population. Of course, it is possible that the lack of additive genetic variation in signa size in the FC population is an artifact produced by the small sample of field collected females (N = 5) used to estimate the genetic parameters. For this reason, the results obtained from population FC should be considered preliminary.

Zeh and Zeh (2005) proposed that extranuclear maternal contributions could promote female adaptations even if these reduce male fitness, therefore, the significant extranuclear maternal effects on signa size and shape that we detected in C. xami (BG: RW3; FC: signa length and u2) are interesting because they could give females an advantage in the sexually antagonistic coevolutionary race between spermatophore envelopes and signa (Cordero 2005; Sánchez et al. 2011).

We did not detect significant additive genetic variation in components of signa shape (Table 3). The function of signa in this species is piercing the spermatophore envelope—thus allowing the female to open the spermatophore and having access to the resources contained in it—and the spine-shaped signa seems ideal for this end. Thus, the lack of additive genetic variance for signa shape could result from directional selection. The lack of additive genetic variation in corpus bursae length and body size measurements also could be due to the erosion of genetic variation in these characteristics as a result of directional selection. For example, the size of the corpus bursae affects the maximum amount of nutritious ejaculate the female can receive, and increased female body size is generally correlated with female fecundity, therefore these two variables could have a strong influence on female fitness. However, we cannot discard the possibility that the significant environmental variance observed in most traits (Table 3) contributed to undercover additive genetic variance in corpus bursae and body size traits.

If the significant differences between populations in body size and in the size and shape of genital traits are real, they are somewhat puzzling because the two populations are very close to each other (both of them are in the small Reserva Ecológica del Pedregal de San Ángel, within the campus CU-UNAM). Since experimental conditions were uniform, these results suggest the existence of genetic differences between populations (although we should remember that the data from the FC population are preliminary due to the small number of females sampled). One hypothesis is that the very busy Insurgentes Avenue separating both populations (this avenue is probably the longest in Latin America, crossing virtually all Mexico City from North to South) is a barrier unexpectedly difficult to cross for C. xami. Research on the population genetics of C. xami in the Reserva Ecológica del Pedregal de San Ángel, and studies of the effect of streets and traffic on flight behavior will help us to test these ideas.

Notes

Acknowledgments

Georgina Jiménez was supported by a grant from PROFIP-DGAPA, UNAM. This project was possible thanks to the financial support to C. Cordero from the Instituto de Ecología, UNAM, and PAPIIT-DGAPA, UNAM (IN223508). Our sincere thanks to Alejandra Parga, Saraí Cruz and Raúl Martínez for technical help; Dr. Raul Rueda and Patricia Romero (IIMAS, UNAM) for statistical advice; and to Dr. Rosario Rodríguez for her support. Two anonymous reviewers provided insightful and very helpful suggestions on a previous version of our manuscript.

References

  1. Alexander RD, Marshall D, Cooley J (1997) Evolutionary perspectives on insect mating. In: Choe JC, Crespi BJ (eds) The evolution of mating systems in insects and arachnids. Cambridge University Press, Cambridge, pp 4–31CrossRefGoogle Scholar
  2. Arnqvist G, Rowe L (2005) Sexual conflict. Princeton University Press, PrincetonGoogle Scholar
  3. Arnqvist G, Thornhill R (1998) Evolution of animal genitalia: patterns of phenotypic and genotypic variation and condition dependence of genital and non-genital morphology in water strider (Insecta: Heteroptera: Gerridae). Genet Res Camb 71:192–212CrossRefGoogle Scholar
  4. Bookstein FL (1991) Morphometric tools for landmark data: geometry and biology. Cambridge University Press, CambridgeGoogle Scholar
  5. Brennan PLR, Prum RO, MacCracken KG, Sorenson MD, Birkhead TR (2007) Coevolution of male and female genital morphology in waterfowl. PLoS ONE 2(5):e418PubMedCrossRefGoogle Scholar
  6. Chenoweth SF, McGuigan K (2010) The genetic basis of sexually selected variation. Ann Rev Ecol Evol Syst 41:81–101CrossRefGoogle Scholar
  7. Cockerham CC, Weir BS (1977) Quadratic analysis of reciprocal crosses. Biometrics 33:187–203PubMedCrossRefGoogle Scholar
  8. Cordero C (1999) Is spermatophore number a good measure of mating frequency in female Callophrys xami (Lycaenidae)? J Lep Soc 53:170–171Google Scholar
  9. Cordero C (2005) The evolution of signa in female Lepidoptera: natural and sexual selection hypotheses. J Theor Biol 232:443–449PubMedCrossRefGoogle Scholar
  10. Cordero C (2010) On the function of cornuti, sclerotized structures of the endophallus of Lepidoptera. Genetica 138:27–35PubMedCrossRefGoogle Scholar
  11. Cordero C, Soberón J (1990) Non-resource based territoriality in the butterfly Xamia xami (Lepidoptera: Lycaenidae). J Insect Behav 3:719–732CrossRefGoogle Scholar
  12. Cordero C, Macías R, Jiménez G (2000) The number of copulations of territorial males of the butterfly Callophrys xami (Lycaenidae). J Res Lepid 35:78–89Google Scholar
  13. Darwin CR (1871) The descent of man and selection in relation to sex. John Murray, LondonCrossRefGoogle Scholar
  14. Drummond BA III (1984) Multiple mating and sperm competition in the Lepidoptera. In: Smith RL (ed) Sperm competition and the evolution of animal mating systems. Academic Press, Orlando, pp 291–371Google Scholar
  15. Eberhard WG (1985) Sexual selection and animal genitalia. Harvard University Press, CambridgeGoogle Scholar
  16. Eberhard WG (1996) Female control: sexual selection by cryptic female choice. Princeton University Press, PrincetonGoogle Scholar
  17. Eberhard WG (2010a) Evolution of genitalia: theories, evidence, and new directions. Genetica 138:5–18PubMedCrossRefGoogle Scholar
  18. Eberhard WG (2010b) Rapid divergent evolution of genitalia: theory and data updated. In: Leonard JL, Córdoba-Aguilar A (eds) The evolution of primary sexual characters in animals. Oxford University Press, NY, pp 40–78Google Scholar
  19. Fry JD (2004) Estimation of genetic variances and covariances by restricted maximum likelihood using PROC MIXED. In: Saxton A (ed) Genetic analysis of complex traits using SAS. SAS Institute Publication 59554, Cary, NC, pp 12–34Google Scholar
  20. Galicia I, Sánchez V, Cordero C (2008) On the function of signa, a genital trait of female Lepidoptera. Ann Entomol Soc Am 101:786–793CrossRefGoogle Scholar
  21. Grafen A, Hails R (2002) Modern statistics for the life sciences. Oxford University Press, NYGoogle Scholar
  22. Higgins SL, Hosken DJ, Wedell N (2009) Phenotypic and genetic variation in male genitalia in the seedbug, Lygaeus equestris (Heteroptera). Biol J Linn Soc 98:400–405CrossRefGoogle Scholar
  23. Hinton HE (1964) Sperm transfer in insects and the evolution of haemocelic insemination. In: Highnam KC (ed) Insect reproduction. Symposium of the royal entomological society of London, London, pp 95–107Google Scholar
  24. Holland B, Rice WR (1998) Chase-away sexual selection: antagonistic seduction versus resistance. Evolution 52:1–7CrossRefGoogle Scholar
  25. Hosken DJ, Stockley P (2004) Sexual selection and genital evolution. Trends Ecol Evol 19:87–93PubMedCrossRefGoogle Scholar
  26. House CM, Simmons LW (2005) The evolution of male genitalia: patterns of genetic variation and covariation in the genital sclerites of the dung beetle Onthophagus taurus. J Evol Biol 18:1281–1292PubMedCrossRefGoogle Scholar
  27. Jagadeeshan S, Singh RS (2006) A time-sequence functional analysis of mating behaviour and genital coupling in Drosophila: role of cryptic female choice and male sex-drive in the evolution of male genitalia. J Evol Biol 19:1058–1070PubMedCrossRefGoogle Scholar
  28. Jiménez G, Soberón J (1988–1989) Laboratory rearing of Sandia xami xami (Lycaenidae, Eumaeini). J Res Lepid 27: 259–271Google Scholar
  29. Joly D, Schmitt M (eds) (2010) Animal genitalia in evolution. Genetica (Special Issue) 138:1–152Google Scholar
  30. Kamimura Y, Iwase R (2010) Evolutionary genetics of genital size and lateral asymmetry in the earwig Euborellia plebeja (Dermaptera: Anisolabididae). Biol J Linn Soc 101:103–112CrossRefGoogle Scholar
  31. Kotiaho JS, LeBas NR, Puurtinen L, Tomkins JL (2008) On the resolution of the lek paradox. TREE 23:1–3PubMedGoogle Scholar
  32. Leonard JL, Córdoba-Aguilar A (eds) (2010) The evolution of primary sexual characters in animals. Oxford University Press, NYGoogle Scholar
  33. Lynch M, Walsh B (1998) Genetics and analysis of quantitative traits. Sinauer Associates, SunderlandGoogle Scholar
  34. Minder AM, Hosken DJ, Ward PI (2005) Co-evolution of male and female reproductive characters across the Scatophagidae (Diptera). J Evol Biol 18:60–69PubMedCrossRefGoogle Scholar
  35. Moreno-García M, Cordero C (2008) On the function of male genital claspers in Stenomacra marginella (Heteroptera: Largidae). J Ethol 26:255–260CrossRefGoogle Scholar
  36. Parker GA (1970) Sperm competition and its evolutionary consequences in the insects. Biol Rev 45:525–567CrossRefGoogle Scholar
  37. Parker GA (1979) Sexual selection and sexual conflict. In: Blum MS, Blum N (eds) Sexual selection and reproductive competition in insects. Academic Press, NY, pp 123–166Google Scholar
  38. Rogers SH, Welles H (1984) The structure and function of the bursa copulatrix of the monarch butterfly (Danaus plexippus). J Morphol 180:213–221CrossRefGoogle Scholar
  39. Rohlf FJ (2008a) tpsRelw, relative warps analysis, version 1.46. Department of Ecology and Evolution, State University of New York, Stony BrookGoogle Scholar
  40. Rohlf FJ (2008b) tpsDig, digitize landmarks and outlines, version 2.12. Department of Ecology and Evolution, State University of New York, Stony BrookGoogle Scholar
  41. Rohlf FJ, Marcus LF (1993) A revolution in morphometrics. Trends Ecol Evol 8:129–132CrossRefGoogle Scholar
  42. Sánchez V, Hernández-Baños BE, Cordero C (2011) The evolution of a female genital trait widely distributed in Lepidoptera: comparative evidence for an effect of sexual selection. PLoS ONE 6(8):e22642Google Scholar
  43. SAS Institute (2002) SAS version 9.0. SAS Institute Inc, NCGoogle Scholar
  44. Schäfer MA, Routtu J, Vieira J, Hoikkala A, Ritchie MG, Schlötterer C (2011) Multiple quantitative trait loci influence intra-specific variation in genital morphology between phylogenetically distinct lines of Drosophila montana. J Evol Biol 24:1879–1886PubMedCrossRefGoogle Scholar
  45. Shaw RG (1991) The comparison of quantitative genetic parameters between populations. Evolution 45:143–155CrossRefGoogle Scholar
  46. Simmons LW (2001) Sperm competition and its evolutionary consequences in the insects. Princeton University Press, PrincetonGoogle Scholar
  47. Simmons LW, Moore AL (2009) Evolutionary quantitative genetics of sperm. In: Birkhead TR, Hosken DJ, Pitnick S (eds) Sperm biology: an evolutionary perspective. Academic Press, San Diego, pp 6111–6585 Kindle e-book locationGoogle Scholar
  48. Simmons LW, House CM, Hunt J, García González F (2009) Evolutionary response to sexual selection in male genital morphology. Curr Biol 19:1442–1446PubMedCrossRefGoogle Scholar
  49. Snook RR, Bacigalupe LD, Moore AJ (2010) The quantitative genetics and coevolution of male and female reproductive traits. Evolution 64:1926–1934PubMedGoogle Scholar
  50. Tomkins JL, Radwan J, Kotiaho JS, Tregenza T (2004) Genic capture and resolving the lek paradox. TREE 19:323–328PubMedGoogle Scholar
  51. Tschudi-Rein K, Benz G (1990) Mechanisms of sperm transfer in female Pieris brassicae. Ann Entomol Soc Am 83:1158–1164Google Scholar
  52. Tuxen SL (ed) (1970) Taxonomist’s glossary of genitalia in insects, 2nd edn. Munksgaard, CopenhagenGoogle Scholar
  53. Waage JK (1979) Dual function of the damselfly penis: sperm removal and transfer. Science 203:916–918PubMedCrossRefGoogle Scholar
  54. Walsh B, Blows MW (2009) Abundant genetic variation + strong selection = multivariate genetic constraints: a geometric view of adaptation. Ann Rev Ecol Evol Syst 40:41–59CrossRefGoogle Scholar
  55. Wedell N, Hosken DJ (2010) The evolution of male and female internal reproductive organs in insects. In: Leonard JL, Córdoba-Aguilar A (eds) The evolution of primary sexual characters in animals. Oxford University Press, NY, pp 307–331Google Scholar
  56. Zeh AZ, Zeh DW (2005) Maternal inheritance, sexual conflict and the maladapted male. Trends Genet 21:281–286PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2012

Authors and Affiliations

  • Georgina Jiménez Ambriz
    • 1
    • 2
  • Diana Mota
    • 3
  • Carlos Cordero
    • 1
  1. 1.Departamento de Ecología Evolutiva, Instituto de EcologíaUniversidad Nacional Autónoma de MéxicoCoyoacánMexico
  2. 2.Departamento de Biología Celular, Facultad de CienciasUniversidad Nacional Autónoma de MéxicoCoyoacánMexico
  3. 3.Carrera de Biología, Facultad de CienciasUniversidad Nacional Autónoma de MéxicoCoyoacánMexico

Personalised recommendations