Apparent Influences of Host-Plant Distribution on the Structure and The

Biochemical Systematics and Ecology 37 (2009) 6–15

Contents lists available at ScienceDirect

Biochemical Systematics and Ecology

journal homepage: www.elsevier.com/locate/biochemsyseco

Apparent inﬂuences of host-plant distribution on the structure and the genetic variability of local populations of the Purple Clay (Diarsia brunnea)

Carine Luque a,b, Luc Legal a,*, Salima Machkour-M’Rabet c, Peter Winterton d, Charles Gers a, Michael Wink b a ECOLAB, UMR 5245 CNRS/Universite´ Paul Sabatier, Bat IVR3, 118 route de Narbonne, F-31062 Toulouse, France b Institut fu¨r Pharmazie und Molekulare Biotechnologie – abteilung Biologie, Universita¨t Heidelberg, Im Neuenheimer Feld 364, D-69120 Heidelberg, Germany c Ecologı´a y Conservacio´n de Fauna Silvestre, El Colegio de la Frontera Sur, Avenida Centenario Km 5.5, AP 424, 77900 Chetumal, Quintana Roo, Mexico d Universite´ Paul Sabatier, 118, Route de Narbonne, F-31062 Toulouse cedex 4, France article info abstract

Article history: Diarsia brunnea (Lepidoptera, Heterocera, Noctuidae, Denis and Schiffermuller, 1775) is an Received 3 October 2008 abundant oligophagous moth occurring in the French Pyrenees. No or little influence of the Accepted 18 January 2009 forest type was found on population densities. In order to study the genetic structure of two separate moth populations in a natural forest and in a plantation, genomic finger- Keywords: printing with ISSR markers (Inter Simple Sequence Repeats) was used. The goal was to Noctuidae search for potential spatial structuring which could be influenced by differences in forest ISSR type. Population structure Host-plant dependence No detectable genetic differences were observed between the populations of the two forest sites. But, although it was not possible to separate on the simple basis of the sampling site, a non-spatial structuring of three sub-populations became apparent. Three of the host plants known for this moth are present in the sampled locations. Three genetically distinct sub-populations were discovered which correlated with the abundance of the three host plants in the two forest plots. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Lepidoptera are sensitive to habitat and climate change (Harper et al., 2000) and studies have shown that the structure of Lepidoptera communities is influenced by habitat quality (Clarke et al., 1997; Luque et al., 2007). Forestry management especially has consequences in various communities linked by a trophic chain (Hammond and Miller, 1998). However, past studies of the structure of some Lepidoptera populations have mostly dealt with endangered butterflies (Rhopalocera), such as Lycaeides melissa samuelis, Nabokov (Packer et al., 1998)orParnassius sp. (Keyghobadi et al., 1999; Megle´ cz et al., 1999; Matter and Roland, 2002), with pests like the moths Helicoverpa armigera,Hu¨ bner (Zhou et al., 2000), Ostrinia nubilalis, Hu¨ bner (Bourguet et al., 2000; Martel et al., 2003), Plutella xylostella L. (Roux et al., 2007) or with species of economic utility such as Bombyx mori L. (Nagaraju et al., 2001). Nagaraju et al. (2001) described that for such studies genomic fingerprinting with ISSRs (Inter Simple Sequence Repeats, Zietkiewicz et al., 1994) is a good analytical technique, has a better cost-benefit ratio than RFLP (Restriction Fragment Length Polymorphism), is more reliable and repeatable than RAPD (Randomly Amplified Polymorphic DNA) and does not

* Corresponding author. Tel.: þ33 (0)5 61 55 61 36; fax: þ33 (0)5 61 55 61 96. E-mail address: [email protected] (L. Legal).

0305-1978/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.bse.2009.01.008 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 7 involve coding regions (which can be under selection pressure) such as AFLP. ISSR markers are directly linked to the existence of microsatellites of variable lengths as the PCR primers bind directly to microsatellites. In addition, we showed that, for some noctuids (Luque et al., 2002), ISSR is sufficiently precise to ensure specific determinations and even to detect individual polymorphism. Absence of a band is interpreted as the loss of a locus/allele through either the deletion of the SSR site or a chromosomal rearrangement (Wolfe and Liston, 1998). ISSRs are thus considered and treated as dominant markers (Casu et al., 2005). Information provided by such markers is phenotypic (Zietkiewicz et al., 1994). Some recent studies using ISSRs on very geographically mobile moths (Hyles euphorbiae L. complex, Sphingidae or P. xylostella L., Plu- tellidae) demonstrated inter- and intra-population relationships with this genetic marker (Hundsdoerfer et al., 2005; Roux et al., 2007). A species of noctuid, the Purple Clay (Diarsia brunnea, Lepidoptera, Heterocera, Noctuidae, Denis and Schiffermuller, 1775) was caught in large numbers during a programme on forestry management impact on moth communities in the Pyrenees, allowing broad genetic analysis. This moth is an oligophagous noctuid. Like many specialised moths, it is a relatively mobile species, but over a limited area. A crossing distance of about 1 km (the distance between the two forests in this study) is quite frequent for such specialised butterflies, which live in small local populations (Baguette, 2003). However, in the present study a discontinuity between the two forests occurred in the form of a sub-alpine grassland and a rocky cliff 400 m higher than the two sampled forests (Luque et al., 2007 for more details). In this way, dispersal is possible but limited and to a multi-generation scale. Floristic composition and relative abundance of the plant species present are known for the two forests (Cassagne, 2003; Luque, 2004). Only three potential host plants for the Purple Clay (Skinner, 1984; Santin, 1998) were observed: bilberries (Vaccinium sp.), ferns (Dryopteridaceae) and brambles (Rubus sp.), with different abundances. The aim was to find out how the population of Purple Clay is structured.

i) Does spatial genetic structuring occur due to the physical separation between the two sites? If this is the case, distribution into two populations, including speciﬁc alleles, corresponding to the two sampling sites could be expected. ii) Could non-spatial structuring be expected due to relationships with host-plant distribution (Michalakis et al., 1993)? For other oligophagous Lepidoptera, it has been demonstrated that an oviposition host-plant preference exists (Thomas and Singer, 1987; Hanski and Singer, 2001). More recently, Bossart (2003) showed that a polyphagous species can present covariance of preference and performance, even if the population is not subjected to obvious internal host-associated structuring.

In this paper, we explore these hypotheses by using genomic ﬁngerprinting by ISSR and we propose the most probable explanation regarding our analysis of the population structure.

2. Materials and methods

2.1. Population sampling

Samples were collected as part of a larger programme studying communities of moths in a forestry context (Luque, 2004; Luque et al., 2007). Two forests differing only in forestry treatment were chosen (Luque et al., 2007). They are both situated in the mountainous level of vegetation (1400–1450 m) (Ozenda, 1994), in the Bethmale valley of Ariege (central Pyrenees, France). The sites are approximately 1 km apart and are separated by a ridge and a stretch of sub-alpine grassland. They have the same exposure (N–NE), topography, climate and pedology (pH 4–5, relative humidity 55–75% and C/N ¼ 20) (Fig. 1; Luque et al., 2007). The two sites are: i) an unmanaged forest (U) with Abies alba Mill. and Fagus sylvatica L., which has not been exploited for at least 150 years; ii) a plantation (P) of A. alba about 40 years old, planted on a cut natural forest of F. sylvatica (Cassagne, 2003; Luque, 2004). Sampling sessions were carried out during the period of the year when moths are active (from late June to beginning of September) once per month over three years (1999–2001) to avoid biases due to annual climatic variations, especially harsh winters that inﬂuence most species’ diapause during this season. Samples were taken using light traps constructed with two 250 W mercury-vapour bulbs. The traps were intentionally placed deep inside the two forests in order to cover a radius of attraction of only around 50 m. However, as discussed in Luque et al. (2007), some corridor effects are possible (up to 75 m of radius) in the plantation. Light traps are known to favour the capture of triﬁne Noctuidae such as our model species (Luque et al., 2007). This can be a problem for obtaining absolute population estimates and for studying moth population dynamics (Raimondo et al., 2004), but it is of less concern if the aim is to compare populations over time or between sites (Yela and Holyoak, 1997; Luque et al., 2007). A few caterpillars were collected in May 2005 on brambles and ferns, but after pupation, it was found that only one feeding on fern was actually D. brunnea. 8 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15

Fig. 1. Location of the two sites in the Pyrenees (France). Source: French National Geographic Institute. P: Plantation, U: Unmanaged forest.

A single sampling point was used for each site. To avoid biases due to humidity, temperature and moonlight (moonless nights or small new moons) differences (Yela and Holyoak, 1997; Luque et al., 2007), collections were done on the same day (pair-wise) in the two studied sites. A preliminary sampling effort for an entire night was carried out at each site. It can be assumed that in such habitats more than 90% of the moths are active early in the evening during almost an hour. This is why the sampling duration used was an hour and a half. This short sampling duration essentially excluded crepuscular species such as certain Sphingidae, or small Geometridae which are active late at night. Therefore, biases due to the short sampling period did not influence the sampling of Purple Clays. In another study, a similar relatively short time of sampling ensured only samples belonging to the sampled sites were collected (Luque et al., 2007). Moths were killed with KCN and stored at 20 C giving negligible effects on yield and quality of DNA (Schneider et al., 1999). Species determination of the moths was carried out using standard identification handbooks (Skinner, 1984; Leraut, 1992). These morphological determinations of Diarsia, more precisely differentiation with close species, were confirmed by ISSR (Luque et al., 2002). Only D. brunnea males were used to avoid genetic pollution from spermatophores that could remain in female abdomens. A total of 121 individuals were analysed, from both the plantation (P: 75 males) and the unmanaged forest (U: 46 males). C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 9

2.2. DNA extraction

Half an abdomen was removed from frozen moths (thoracic part of the caterpillar to avoid fat) and incubated overnight at 50 Cin700ml of lysis buffer (10 mM Tris, pH 7.5, 25 mM EDTA, 75 mM NaCl) with 1 mg of proteinase K (Boehringer, Mannheim) and 35 ml of 20% SDS, or in guanidinium thiocyanate buffer (4 M guanidinium thiocyanate, 0.1 M Tris–HCl, pH 7.5), 1%b-mercaptoethanol, followed by a standard phenol–chloroform protein extraction. DNA was precipitated with 800 ml of cold isopropanol, centrifuged, washed, dried and re-suspended in TE buffer (Sambrook et al., 1989; Swatschek et al., 1994).

2.3. ISSR–PCR ampliﬁcation

0 For amplifications, 300 ng of total DNA was used as a template, plus 10 pmol primer 5 -(CA)10, 0.1 mM of dGTP, dCTP, and dTTP, 0.075 mM dATP, 1 mCi [a-33P]-dATP, 2.5 mlof10 amplification buffer (100 mM Tris–HCl, pH 8.5, 500 mM KCl, 5% Triton X-100, 15 mM MgCl2) and 0.15 units Taq polymerase (Pharmacia Biotech, Freiburg) in a total volume of 25 ml. After an initial denaturation (4 min at 94 C), 40 cycles of 45 s at 94 C, 45 s at 40 C, and 45 s at 72 C were performed on a Biometra thermocycler; then at 72 C for 20 min, followed by 4 C for storage. After mixing with a specific blue marker (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, 0.05% xylene cyanol FF) and denaturation at 90 C for 5 min, PCR products were electrophoretically separated on a high resolution acrylamide gel (Sequagel matrix, National Diagnostics, Atlanta, USA) at 65 W for 4 h (size 45 30 cm). The separation of ISSR fragments on such gel can enhance the resolution of ISSR bands (Arcade et al., 2000). After drying, the gel was exposed to an X-ray film (Kodak Biomax MR) overnight at room temperature and developed (Kodak).

2.4. Data analysis

As ISSR is a dominant marker, ampliﬁed fragments were considered as di-allelic and bands were scored as the two alleles of a locus. A band present (coded 1) represents the dominant allele (Aa and AA individuals have the (1) phenotype) whereas the (0) phenotype is noted for aa individuals, when the band is absent (coded 0). To determine the number of potential sub-populations for each location, pooling all samples, two different approaches were used:

2.4.1. Self-organising map (SOM) The SOM, an unsupervised neural network, referred to as a Kohonen neural network, approximates the probability density function of the input variables and performs a non-linear projection of the multivariate data into two-dimensional space (Kohonen, 2001): the SOM map. The final SOM map is an (X,Y) lattice composed of hexagons. The map is separated into similarity groups. Samples are finally set in these groups in different cells (hexagons) depending on their SOM classification. Some cells can be empty due to the absence of samples corresponding to these characteristics (Gevrey et al., 2004; Roux et al., 2007; Bouzid et al., 2008).

2.4.2. Bayesian clustering method We used the Bayesian genetics model, implemented in STRUCTURE 2.2.2 (basic algorithm: Pritchard et al., 2000;and extension to the method: Falush et al., 2007), to infer the potential number of populations in our study. This method was designed to identify K (unknown) populations and to assign each individual, with a probability (qi), to one population (cluster) or more than one if their genotypes indicate that they are admixed. Population structure is detected by departures from Hardy–Weinberg and linkage equilibrium which could result by recent admixture, migration and/or hybridisation. To determine the most probable number K of population units, the program was run five times for different values of K (from 1 to 6) with our data sets using the no-admixture and the correlated-frequencies models. For each value of K, the Markov Chain Monte Carlo (MCMC) algorithm was run with a burn-in period of 1,000,000 steps followed by 500,000 steps. When we obtained the ‘‘true value’’ of K, we assessed the average coefficient of membership (Q) of each of the sampled populations to the inferred clusters. Then, each genotype was assigned to these clusters, based on values of the individuals’ proportion of membership (qi). Referring to Barilani et al. (2007), we decided to assign each individual to only one cluster if qi > 0.90, otherwise it was associated to two or more clusters (admixture individuals). To determine the genetic parameters of (sub)population structure, the presence/absence matrix was first separated into sub-populations determined by SOM/Structure common groups. Discriminant analysis with a jack-knifed classification matrix (which computed the proportion of correct classifications) was performed with SYSTAT 8.0 to validate the sub-populations (tolerance 0.001), before they were analysed using POPGENE 1.32 (Yeh et al., 1997). Some parameters of genetic diversity within sub-populations were calculated: observed number of alleles per locus (Na), effective number of alleles per locus (Ne), Shannon’s Information index (I; Lewontin, 1972), percentage of polymorphic loci (P). Nei’s (1973) coefficient of gene differentiation (Gst) and level of gene flow (Nm) were also computed using this software. Analysis of Molecular Variance (AMOVA) was also carried out using this matrix in the form of input files for WINAMOVA 1.05 (Excoffier, 1992; Excoffier et al., 1992)byAMOVA-PREP 1.01 (Miller, 1998). Three-level genetic variance partitioning was used: 10 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 among sub-populations, among sites within sub-populations and among individuals within sub-populations. Seven thousand permutations were employed to ensure significant testing.

2.5. Relative abundance of host plants

A procedure to estimate plant species abundance/covering was performed by following the Braun-Blanquet cover abundance scale (Mueller-Dombois and Ellenberg, 1974). Three nested quadrats following Barbour et al. (1987) were performed in each of the two forests. A discussion on uses and misuses of these measurements is detailed for the same study sites in Cassagne (2003).

3. Results

Among the total bands produced (n ¼ 60), only polymorphic bands belonging solely to D. brunnea were selected (variable bands common with other close species were discarded from analysis; Luque et al., 2002). Ultimately, a total of 16 variable ‘‘diagnostic’’ bands (Luque et al., 2002) belonging only to D. brunnea were studied for 121 individuals.

3.1. Structure of the population

Neither of the analyses, SOM or STRUCTURE, allowed the separation of individuals according to the sampled sites. However, after pooling all the samples, the SOM treatment discriminated three sub-populations called A (n ¼ 45), B (n ¼ 25), and C (n ¼ 51) (Fig. 2). Using probability assignments estimated by STRUCTURE, all individuals were partitioned into three sub-populations or clusters (K ¼ 3) showing a gathering of initial populations evaluated by the Q probability value (Table 1): sub-population A (n ¼ 43) is totally homogenous while sub-populations B (n ¼ 25) and C (n ¼ 53) show few genetic flux/introgression events (Fig. 3). Discriminant analysis with a Jack-knifed classification matrix showed that these three sub-populations differed signifi- cantly (Table 1).

3.2. Genetic diversity

The percentage of polymorphic loci varied from 81% (sub-population A) to 62.5 % (sub-population B) (Table 2). The average number of alleles observed per locus (Na) was similar among three sub-populations (Table 2). The number of effective alleles per locus (Ne) ranged from 1.000 to 2.000 with a mean of 1.326 for sub-population A, from 1.000 to 1.77 with a mean of 1.184 for sub-population B, and from 1.000 to 1.999 with a mean of 1.3016 for sub-population C. For the overall population, Ne ranged from 1.025 to 1.945 with an average of 1.455, Shannon’s Information index I ranged from 0.234 (sub-population B) to 0.351 (sub-population C), with an average of 0.412 for the population. For the overall population and between collecting sites, Gst ¼ 0.0075 (Nm ¼ 66.0793). Among sub-populations Gst varied: GstBC ¼ 0.1385 (Nm ¼ 3.1091), GstAC ¼ 0.2163 (Nm ¼ 1.8116) and GstAB ¼ 0.3121 (Nm ¼ 1.1020). Between collecting sites for the three groups, Gst were: GstA ¼ 0.0282 (Nm ¼ 17.2551), GstB ¼ 0.1575 (Nm ¼ 2.6738) and GstC ¼ 0.0661 (Nm ¼ 7.0687). The AMOVA analysis showed, with a high degree of signiﬁcance (p < 0.0001), that individual variability represents 51.45% of the total variance with only 8.58% coming from special variability, e.g., between components of each sub-population differing by the collection site (Table 3). Sub-population diversity represents 39.96% of genetic diversity. Such a high signiﬁcance of variance components was obtained from 7000 permutations. One thousand permutations were enough to obtain p < 0.001.

4. Discussion

ISSR–PCR is already known as a very effective method to understand intra-specific and genetic structures of populations (Fang and Roose, 1997; Ge and Sun, 1999; Nagaraju et al., 2001; Ge et al., 2003; Casu et al., 2005; Hundsdoerfer et al., 2005; Roux et al., 2007; Bouzid et al., 2008). As expected, this molecular marker is effective as the present study found genetic variability in both individuals and sub-populations. In this study, we intentionally kept a low number of genetic characters (16 among a possible 60). This harsh selection of variable genetic characters is justified considering that at least 4 other close species of Diarsia/Xestia were flying at the same time as D. brunnea (Luque et al., 2002). Among our samples some were identified as hybrids or at least results of introgression events between species (data not shown). Therefore, only specific genetic characters belonging to our studied species were taken into account. In addition, this selection meant that the number of variable and informative genetic characters was the same as that used in similar population genetic studies. Numbers were: (i) this study: 16 informative genetic differences; (ii) using mitochondrial DNA, among close species, average value: 10 (Aubert et al., 1999; Albre et al., 2008); (iii) with ISSR, inside populations but using more primers, values on moths: 5–25 (Hundsdoerfer et al., 2005; Roux et al., 2007); in Ligula (Cestodes) average value (removing Chinese samples which may belong to another species): 18 (Bouzid et al., 2008). C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 11

B C

Clusters, SOM map, Ward algorithm 6

4 BC A

Fig. 2. SOM discrimination of the three sub-populations of Purple Clay. Sub-population A, Sub-population B, Sub-population C. Only discriminated groups (putative sub-populations) are indicated (not individuals). The area covered by the three groups indicates overall genetic variability and is independent of the number of individuals. The resulting tree (Ward algorithm; see Roux et al., 2007 for more details) indicates the total number of different genotypes in each sub-population (note that the number of branches differs from the total number of individuals sampled as some individuals were identical).

4.1. Structure of the population

Both methods of analysis (SOM and STRUCTURE), and the very low value of Gst (0.0075), confirmed the single population hypothesis as being the most plausible. This hypothesis will have to be subsequently proved by subjecting individuals from other valleys in the Pyrenees to ISSR. At our local scale, this allowed us to discard the presence of specific alleles in each of our sampled locations (absence of two genetically distinct populations), despite the apparently efficient physical barrier between

Table 1 Membership probability (Q) of individuals at each of the three inferred clusters as determined by STRUCTURE software. The highest value of a sample assigned to one cluster (for K ¼ 3) is indicated in bold. n: number of individuals.

Sites Sub-populations n Clusters

123 PA 251.000 0.000 0.000 UA 181.000 0.000 0.000 P B 19 0.000 0.993 0.007 U B 6 0.000 0.981 0.019 P C 31 0.002 0.067 0.931 U C 22 0.001 0.118 0.881 12 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 PPPUU U

A B C

Fig. 3. Analysis of the population and discrimination of the Purple Clay under a Bayesian approach computed by STRUCTURE 2.2.2 software with K ¼ 3. Each individual is represented by a single vertical line broken into K segments whose length is proportional to the estimated membership (probability qi) in the K clusters. The three sub-populations are presented with the same grey scale as in Fig. 2. P: Plantation, U: Unmanaged forest and ‘‘A’’, ‘‘B’’, ‘‘C’’ for the three sub- populations/ecotypes discriminated. them. In another way, as the passage of D. brunnea from one forest to the other is reduced (total absence of individuals of this species sampled in the alpine prairie separating U and P), this supports the hypothesis that the local genetic variations observed are due to local adaptations. Before seeking genetic diversity, and more especially relationships among sub-populations inside a population, an understanding of how this population is structured was essential. This situation, with non-spatial structuring represented by a mixture of individuals from the two sites (plantation and unmanaged forest), led us to seek a different population structure organisation, one not generated by spatial separation. The distribution of three sub-populations was demonstrated by ‘‘SOM’’ and ‘‘STRUCTURE’’ and conﬁrmed by discriminant analysis. A very low variability for each of these sub-populations between capture sites was shown. It can be concluded that within collected samples, a single population (or part of a population) was found to be composed of three putative sub-populations of D. brunnea. We collected more individuals in the plantation than in unmanaged forest (Tables 4 and 5). This could be due to the lower biodiversity in this site leading to lower competition between species which may have the same ecological exigencies (Albre et al., 2008), many other species of Diarsia (or closely related Noctuidae) are present in unmanaged forest while they are absent from plantations (Luque et al., 2007). This was especially noticeable for sub-population B (Table 4).

4.2. Genetic diversity

The AMOVA results showed very little variability within the three sub-populations (A, B and C). Genetic variability in each sub-population appears to be lower than individual genetic variability. Furthermore, Gst within groups were smaller than Gst between groups. This conﬁrms the separation of the population into three sub-populations made by SOM/STRUCTURE analysis in which individuals seem to be grouped according to genetic similarities. Results from ‘‘among sites within sub-populations’’ demonstrate that these three sub-populations are genetically more homogeneous than the population. Comparison of the different indices (Na, Ne and P) also shows that there were few differences among the three sub-populations. Sub-population A, compared to the two others, had a higher average number of observed alleles per locus and percentage of polymorphic loci, but the smallest average number of effective alleles per locus. The main difference among the three sub-populations is the number of individuals captured at each site. For B, fewer individuals (n ¼ 8 versus 17) were captured in unmanaged forest. For this last sub-population, the average number of alleles observed (Na) was lower than for A. But the same number calculated with information from the entire population (Ne) shows that the number of alleles could have been as high as for A.

Table 2 Jack-knifed classiﬁcation matrix for the three sub-populations A, B and C (Tolerance 0.001).

Jack-knife estimates % Correct

ABC Input A4500100 B 0 23 2 93 C084381

Total 45 33 43 91 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 13

Table 3 Genetic variability of Diarsia brunnea in the Pyrenees (population) detected by ISSR analyses and for each sub-population (A, B, and C).

Sub-population Na (SD) Ne (SD) I (SD) P (%) A(n¼45) 1.812 (0.403) 1.326 (0.384) 0.305 (0.260) 81.25 B(n¼25) 1.625 (0.500) 1.184 (0.234) 0.216 (0.219) 62.50 C(n¼51) 1.750 (0.447) 1.302 (0.378) 0.291 (0.263) 75.00

Population (n¼121) 2.000 (0.000) 1.456 (0.381) 0.412 (0.236) 100.00

Na, observed number of alleles per locus; Ne, effective number of alleles per locus; I, Shannon’s Information index; P, percentage of polymorphic loci; n, number of individuals.

4.3. Three sub-populations versus three host plants?

The question is now to find a hypothesis for this three sub-population distribution. Our former work allowed us to discard the hypothesis of the presence of a complex of sibling species (Luque et al., 2002). The frequency of the 3 genotypes was not time dependent as we collected all 3 in similar proportions from late June to September (data not shown). Therefore, we finally favoured an ‘‘environmental’’ hypothesis. Considering our sampling protocol, we were able to reject any hypotheses related with elevation, temperature, precipitations. As D. brunnea is an oligophagous species, a link between population structure and host plants was then, not the first hypothesis chosen. However, only three of the described host plants for this species (Santin, 1998) are present in the two sites in different proportions: two of them share a similar abundance at the two sites (bramble and fern), the third (bilberry) is more abundant in the plantation (Table 5). Even though only a single caterpillar clearly belonging to D. brunnea was collected (on ferns), it was classified with the individuals of sub-population A, suggesting that sub-population A may correspond to individuals growing on this plant. Furthermore ferns are phylogenetically notably distinct from brambles and bilberries (Hilu et al., 2003) and may be in accordance with the strict genetic separation of sub-population A from the others. Another clue is the correspondence between per site/covering of bilberries in the two forests and number of samples belonging to sub-population B (Table 5). Therefore, samples belonging to sub-population C are putatively assigned to brambles. In this way, a plausible hypothesis could be a distribution into ecotypes determined by host plants. Bossart (2003) demonstrated, for polyphagous Lepidoptera, and in independent analyses of preference-performance covariance that links between female choice and progeny performance ‘‘exist at the level of variation among individuals of a single population’’. Moreover, different authors have shown that, for oligophagous Lepidoptera, an oviposition host-plant preference exists (Thomas and Singer, 1987; Hanski and Singer, 2001). It seems to be influenced by the relative abundance of the host plants. These authors assume a genetic basis for this preference, as individuals from a place with an abundant host plant still prefer that plant even when they are introduced to another site where the plant is not so available, but an alternative host species predominates. In this study, the question of genetic determination was approached in another way, separating different sub-populations and evaluating comparisons with host-plant abundances. This hypothesis should be confirmed, but it apparently favours the theory of a genetic component in host-plant choice. The gap which separates host-plant choice studies from genetic population structure studies is therefore in the process of being bridged. When a species is faced with ecosystem changes, genetic variation within populations can confer greater resilience (Luck et al., 2003). Host-plant distribution depends on the forest structure, which, for two similar sites, is the result of forestry management practices. In this case, if ecotypes are confirmed to be dependent on a genetically controlled host- plant preference, this structure of the Purple Clay population might confer an advantage in resistance to forest changes without obvious damage to population size, or representation relative to other species. The potential link between forestry impact and the structure and genetic diversity of moth populations should therefore now be investigated more in detail.

Table 4 Analysis of molecular variance (AMOVA) for 121 individuals in 3 sub-populations of Diarsia brunnea (with 2 collecting sites for each) using 16 ISSR markers.

Source of variance df SSD MSD Variance components %Total p-Value Among sub-populations 2 106.546 53.273 1.161 39.96 <0.0001 Among sites 3 18.120 6.040 0.249 8.58 <0.0001 within sub-populations Among individuals 115 171.946 1.495 1.495 51.45 <0.0001 within sub-populations

Statistics include degrees of freedom (df), sum of squares (SSD), mean squared deviations (MSDs), variance component estimates, the percentage of the total variance (% total) contributed by each component. The probability (p) of obtaining a more extreme component estimated by chance alone (signiﬁcance tests after 7000 permutations). 14 C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15

Table 5 Abundance of individuals of Diarsia brunnea per ecotype and their potential host plants in the two forests.

n Cover (%)

UP UP A 20 25 Ferns 8 5 B 8 17 Bilberries 2 15 C 18 33 Brambles 10 7 n, number of individuals of Diarsia brunnea; Cover, percentage of soil surface area covered by the plant; P, Plantation; U, Unmanaged forest.

Acknowledgements

We would like to thank Heidi Staudter for technical help and advice on ISSR, Herve´ Gryta for his advice on data processing, Young-Seuk Park and Sovan Lek for SOM results. Thanks are also due to University Paul Sabatier (UPS), Toulouse, France and to the Alexander von Humboldt Foundation (A.v.H) for ﬁnancial support (UPS grant for Carine Luque and A.v.H. grant for Luc Legal).

References

Arcade, A., Anselin, F., Rampant, P., Lesage, M., Paques, L., Prat, D., 2000. Application of AFLP, RAPD and ISSR markers to genetic mapping of European and Japanese larch. Theor. Appl. Genet. 100, 299–307. Albre, J., Gers, C., Legal, L., 2008. Molecular Phylogeny of the Erebia Tyndarus (Lepidoptera, Rhopalocera, Nymphalidae, Satyrinae) complex of species combining CoxII and ND5 mitochondrial genes: a case study of a recent radiation. Mol. Phylogenet. Evol. 47, 196–210. Aubert, J., Legal, L., Descimon, H., Michel, F., July 1999. Molecular phylogeny of swallowtail butterflies of the tribe Papilionini (Papilionidae, Lepidoptera). Mol. Phylogenet. Evol. 12 (2), 156–167. Baguette, M., 2003. Long distance dispersal and landscape occupancy in a metapopulation of the cranberry fritillary butterfly. Ecography 26, 153–160. Barbour, M.G., Burk, J.H., Pitts, W.D., 1987. Terrestrial Plant Ecology, second ed. Benjamin/Cummings Publishing Company, Menlo Park, California. Barilani, M., Sfougaris, A., Giannakopoulos, A., Mucci, N., Tabarroni, C., Randi, E., 2007. Detecting introgressive hybridisation in rock partridge populations (Alectoris graeca) in Greece through Bayesian admixture analyses of multilocus genotypes. Conserv. Genet. 8, 343–354. Bossart, J.L., 2003. Covariance of preference and performance on normal and novel hosts in a locally monophagous and locally polyphagous butterfly population. Oecologia 135, 477–486. Bourguet, D., Bethenod, M.T., Pasteur, N., Viard, F., 2000. Gene flow in the European corn borer Ostrinia nubilalis: implications for the sustainability of transgenic insecticidal maize. Proc. R. Soc. Lond., B, Biol. Sci. 267, 117–122. Bouzid, W., Lek, S., Mace´, M., Ben Hassine, O., Etienne, R., Legal, L., Loot, G., 2008. Genetic diversity of Ligula intestinalis (L.) (Cestoda: Diphyllobothriidea) based on analysis of inter-simple sequence repeat (ISSR) markers. J. Zoolog. Syst. Evol. Res. 46, 289–296. Cassagne, N., 2003. Evolution de la flore, des humus et de la pe´dofaune associeé sous l’effet de perturbations sylvicoles. University Paul Sabatier, Toulouse, France, pp. 155. Casu, M., Maltagliati, F., Cossu, P., Lai, T., Galletti, M.C., Castelli, A., Commito, J.A., 2005. Fine-grained spatial genetic structure in the bivalve Gemma gemma from Maine and Virginia (USA), as revealed by Inter-Simple Sequence Repeat markers. J. Exp. Mar. Biol. Ecol. 325, 46–54. Clarke, R.T., Thomas, J.A., Elmesm, G.W., Hochberg, M.E., 1997. The effects of spatial patterns in habitat quality on community dynamics within a site. Proc. R. Soc. Lond., B, Biol. Sci. 264, 347–354. Excoffier, L., 1992. WINAMOVA, Analysis of Molecular Variance for Windows. Version 1.04. Genetics and Biometry Laboratory, Geneva, Switzerland. Excoffier, L., Smouse, P.E., Quattro, J.M., 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131, 479–491. Falush, D., Stephens, M., Pritchard, J.K., 2007. Inference of population structure using multilocus genotype data: dominant markers and null alleles. Mol. Ecol. Notes 7, 574–578. Fang, D., Roose, M., 1997. Identification of closely related citrus cultivars with inter-simple sequence repeat markers. Theor. Appl. Genet. 95, 408–417. Ge, X.J., Sun, M., 1999. Reproductive biology and genetic diversity of a cryptoviviparous mangrove Aegiceras corniculatum (Myrsinaceae) using allozyme and intersimple sequence repeat (ISSR) analysis. Mol. Ecol. 8, 2061–2069. Ge, X.J., Yu, Y., Zhao, N.X., Chen, H.S., Qi, W.Q., 2003. Genetic variation in the endangered Inner Mongolia endemic shrub Tetraena mongolica Maxim. (Zygophyllaceae). Biol. Conserv. 111, 427–434. Gevrey, M., Rimet, F., Park, Y.S., Giraudel, J.L., Ector, L., Lek, S., 2004. Water quality assessment using diatom assemblages and advanced modelling techniques. Freshw. Biol. 49, 208–220. Hammond, P.C., Miller, J.C., 1998. Comparison of the biodiversity of Lepidoptera within three forested ecosystems. Ann. Entomol. Soc. Am. 91, 323–328. Hanski, I., Singer, M.C., 2001. Extinction-colonization dynamics and host-plant choice in butterfly metapopulations. Am. Nat. 158, 341–353. Harper, G.L., Piyapattanakorn, S., Goulson, D., Maclean, N., 2000. Isolation of microsatellite markers from the Adonis blue butterfly (Lysandra bellargus). Mol. Ecol. 9, 1948–1949. Hilu, K.W., Borsch, T., Muller, K., Soltis, D.E., Soltis, P.S., Savolainen, V., Chase, M., Powell, M., Alice, L., Evans, R., Sauquet, H., Neinhuis, C., Slotta, T., Rohwer, J., Chatrou, L., 2003. Inference of angiosperm phylogeny based on matK sequence information. Am. J. Bot. 90, 1758–1776. Hundsdoerfer, A.K., Kitching, I.J., Wink, M., 2005. The phylogeny of the Hyles euphorbiae complex (Lepidoptera: Sphingidae): molecular evidence from sequence data and ISSR–PCR fingerprints. Org. Divers. Evol. 5, 173–198. Keyghobadi, N., Roland, J., Strobeck, C., 1999. Influence of landscape on the population genetic structure of the alpine butterfly Parnassius smintheus (Papilionidae). Mol. Ecol. 8, 1481–1495. Kohonen, T., 2001. Self-organizing Maps, third ed. Springer, Berlin, pp. 501. Leraut, P., 1992. Les papillons dans leur milieu. Bordas, Paris, pp. 256. Lewontin, R.C., 1972. The apportionment of human diversity. In: Dobzhansky, T.H., Hecht, M.K., Steere, W.C. (Eds.), Evolutionary Biology. Appleton-Century- Crofts, New York, pp. 381–398. Luck, G.W., Daily, G.C., Ehrlich, P.R., 2003. Population diversity and ecosystem services. Trends Ecol. Evol. 18, 331–336. Luque, C., 2004. Forestry impacts on forest biodiversity: ecological reactivity of Arthropod communities and molecular analysing of a target species Diarsia brunnea (D & S, Lepidoptera). Paul Sabatier University of Toulouse and Ruprecht-Karls University of Heidelberg, France and Germany, pp. 217. Luque, C., Legal, L., Staudter, H., Gers, C., Wink, M., 2002. ISSR (Inter Simple Sequence Repeats) as genetic markers in Noctuids (Lepidoptera). Hereditas 136, 251–253. Luque, C., Gers, C., Lauga, J., Mariano, N.A., Wink, M., Legal, L., 2007. Analysis of forestry impacts and biodiversity in two Pyrenean forests through a comparison of moth communities (Lepidoptera, Heterocera). Insect Sci. 14, 323–338. C. Luque et al. / Biochemical Systematics and Ecology 37 (2009) 6–15 15

Martel, C., Rejasse, A., Rousset, F., Bethenod, M.T., Bourguet, D., 2003. Host-plant-associated genetic differentiation in Northern French populations of the European corn borer. Heredity 90, 141–149. Matter, S.F., Roland, J., 2002. An experimental examination of the effects of habitat quality on the dispersal and local abundance of the butterfly Parnassius smintheus. Ecol. Entomol. 27, 308–316. Meglećz, E., Ne`ve, G., Pecsenye, K., Varga, Z., 1999. Genetic variations in space and time in Parnassius mnemosyne (L.) (Lepidoptera) populations in north-east Hungary: implications for conservation. Biol. Conserv. 89, 251–259. Michalakis, Y., Sheppard, A.W., Noel, V., Olivieri, I., 1993. Population structure of a herbivorous insect and its host plant on a microgeographic scale. Evolution 47, 1611–1616. Miller, M.P., 1998. AMOVA-PREP 1.01: A Program for the Preparation of AMOVA Input Files from Dominant-markers Raw Data. Computer software distributed by author. Mueller-Dombois, D., Ellenberg, H., 1974. Aims and Methods of Vegetation Ecology. John Wiley & Sons, New York. Nagaraju, J., Reddy, K., Nagaraja, G., Sethuraman, B., 2001. Comparison of multilocus RFLPs and PCR-based marker systems for genetic analysis of the silkworm, Bombyx mori. Heredity 86, 588–597. Nei, M., 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. U.S.A. 70, 3321–3323. Ozenda, P., 1994. Ve´ge´tation du continent europeén. Delachaux and Niesle´, Paris, pp. 269. Packer, L., Taylor, J.S., Savignano, D.A., Bleser, C.A., Lane, C.P., Sommers, L.A., 1998. Population biology of an endangered butterfly, Lycaeides melissa samuelis (Lepidoptera; Lycaenidae): genetic variation, gene flow, and taxonomic status. Can. J. Zool. 76, 320–329. Pritchard, J.K., Stephens, M., Donnelly, P.J., 2000. Inference of population structure using multilocus genotype data. Genetics 155, 945–959. Raimondo, S., Strazanac, J.S., Butler, L., 2004. Comparison of sampling techniques used in studying Lepidoptera population dynamics. Environ. Entomol. 33, 418–425. Roux, O., Gevrey, M., Arvanitakis, L., Gers, C., Bordat, D., Legal, L., 2007. ISSR–PCR: tool for population identification and genetic structure of Plutella xylostella. Mol. Phylogenet. Evol. 43, 240–250. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York. Santin, A., 1998. Re´pertoire des plantes-hoˆtes et de substitution des chenilles de Le´pidopte`res du monde. OPIE, Guyancourt, pp. 1125. Schneider, D., Legal, L., Dierl, W., Wink, M., 1999. Androconial hairbrushes of the Syntomis (Amata) phegea (L.) Group (Lepidoptera, Ctenuchinae): a synapomorphic character supported by sequence data of the mitochondrial 16S rRNA gene. Zeitschrift fur Naturforschung C (J. Biosci.) 54, 1119–1139. Skinner, B., 1984. Colour Identification Guide to Moths of the British Isles. Viking Press, London, pp. 267. Swatschek, I., Ristow, D., Wink, M., 1994. Mate fidelity and parentage in Cory’s shearwater (Calonectris diomedea) - field studies and DNA fingerprinting. Mol. Ecol. 3, 259–262. Thomas, C.D., Singer, M.C., 1987. Variation in host preferences affects movement patterns within a butterfly population. Ecology 68, 1262–1267. Wolfe, A.D., Liston, A., 1998. Contributions of PCR-based methods to plant systematics and evolutionary biology. In: Soltis, D.E., Soltis, P.S., Doyle, J.J. (Eds.), Plant Molecular Systematics II. Kluwer Academic Publishers, Dordrecht, pp. 43–86. Yeh, F.C., Yang, R.C., Boyle, T.B.J., Yeh, Z.H., Mao, J.X., 1997. POPGENE, the User-friendly Shareware for Population Genetic Analysis. Version 1.32. Molecular Biology and Biotechnology Centre, University of Alberta, Edmonton. Yela, J.L., Holyoak, M., 1997. Effects of moonlight and meteorological factors on light and bait trap catches of noctuid moths (Lepidoptera: Noctuidae). Environ. Entomol. 26, 1283–1290. Zhou, X.F., Faktor, O., Applebaum, S.W., Coll, M., 2000. Population structure of the pestiferous moth Helicoverpa armigera in the Eastern Mediterranean using RAPD analysis. Heredity 85, 251–256. Zietkiewicz, E., Rafalski, A., Labuda, D., 1994. Genome fingerprinting by simple sequence repeat (SSR)-anchored polymerase chain reaction amplification. Genomics 20, 176–183.