Haplotype characterization of the COI mitochondrial gene in Chrysoperla externa ( Neuroptera : Chrysopidae ) from different environments in Jaboticabal , state of São Paulo , southeastern

The green lacewings (Chrysopidae) belong to the Order Neuroptera and are described as voracious predators in the larval stage and sometimes also in their adulthood. They are an important group used in integrated biological control in field and horticultural crops. Individuals of Chrysoperla externa were collected during 2007 until March 2008 in five different locations in Jaboticabal, SP, with all the seasons sampled. Thirty six sequences with 805 pairs of bases for the gene mitochondrial Citochrome Oxidase I (COI) were analysed. The genetic parameters revealed 24 haplotypes for this population, a total of 36 mutations and haplotype diversity of 0.956. The data of genetic distance and population structure calculated for this population considering the different areas and seasons, revealed a great genetic similarity and high degree of genetic sharing between individuals sampled. It showed that the species Chrysoperla externa from Jaboticabal, SP, is a single population, without genetic structure neither due to the area of origin nor to the seasons of the year.


Introduction
Among the so-called green lacewings (Neuroptera: Chrysopidae), the species of the genus Chrysoperla are well known as beneficial species in the biological control of agricultural pests (Senior and McEwen, 2001).The family Chrysopidae contains over 1,200 species and subspecies divided into at least 80 genera and subgenera distributed around the world on all continents, except Antarctica (Brooks and Barnard, 1990;Brooks, 1997).The genus Chrysoperla contains 36 species and four species can be found in Brazil: C. defreitasi (Brooks, 1994) , C.
population will only be established if the environmental pressure factors are matched by their corresponding support capacities.Therefore, knowing their genetic structure is of paramount importance in the process of choosing pest controlling populations to be field released.

Biological material
The thirty six specimens used in this study were collected in five places throughout 2007 until March 2008.All the seasons were sampled in an area of approximately 504 ha (Table 1), containing both wild and cultivated vegetation, located in the municipality Jaboticabal, in the state of São Paulo, Brazil (21º 15' 22" S and 48º 18' 58" W).The different collection places were chosen according to their different vegetation culture characteristics plus geographical location, so that the areas chosen were not neighbouring areas.
Thus, the selected places for sampling were: P1experimental nursery, area of native vegetation type, inside a stream riparian vegetation; P2 -green house, place of cultivation of various fruit trees, such as some citrus, cashew, passion fruit, coconut, and also corn and sunflower seed; P3 -mixed orchard, an area which cultivates only citrus species, such as oranges, lemons and tangerines; P4monoculture orchard, an area which cultivates only guava and; P5 -an area of eucalyptus, a eucalyptus cultivation (Figure 1).The identification of individuals was performed performed by Lourenço et al. (2006) using two populations of the Brazilian species Chrysoperla externa, the results showed that the two populations were genetically distinct, suggesting the existence of a genetic structure in populations of the species in question.
However, until now, very little has been done in terms of describing and understanding the intraspecific and interspecific variation among these distinct populations of Chrysoperla externa, despite these parameters being very important to reach a correct specific taxonomic identification, which guarantees success in programs of biological control, in addition to allowing the proper understanding of gene flow process and population structure among sub-populations.In molecular methods, mitochondrial DNA is generally considered an excellent marker for genealogical inferences at the intraspecific level and particularly so, the gene cytochrome oxidase I (COI), which is one of the genes more widely utilised in systematic and populational studies (Avise, 2004).
Thus, the purpose of this research is the investigation of Chrysoperla externa's genetic diversity, genetic and population structure in the study area, utilising mitochondrial gene markers.The importance of the present study is derived from the well known fact that the adaptability of any given population is dependent on its phenotypic and ecological plasticity as well, and that the pest control field success of a biological agent -in this case, green lacewings -, is directly dependent on the choice of the population chosen to be released in the field.Obviously, the introduced pest controlling belonged.Among-haplotype comparisons were based on Φ-analogues of Wright's (1965) F-statistics.Significance was determined using 1,000 permutations with statistical significance of 5%.

mtDNA haplotype diversity
A 805-bp fragment was used in all the analyses of the 36 individuals, including 10 individuals from P1 locality, five individuals from P2, five individuals from P3, 10 individuals from P4 and six individuals from P5.No insertions, deletions or stop codons were detected in the analyses.The average percentage base composition was 42.7% thymine, 29.4% adenine, 13.5% cytosine and 14.4% guanine.Overall, 34 polymorphic sites were observed, defining 24 haplotypes, with an overall diversity (h) of 0.956 ± 0.023.Nucleotide diversity was lower (Pi) than 0.00462 ± 0.00065, the total number of mutations (η) was 36 and the average number of nucleotide differences (k) was 3.722 (Table 2).

Data analysis
The genetic distance was calculated for the total population distributed among distinct localities and seasons of the year, using the model of Kimura 2-parameter (Kimura, 1980).For the population as a whole, the highest value of genetic distance was 0.043, found between individuals from the place P1 and P5 (experimental nursery and area of eucalyptus, respectively).In both groups, individuals collected in the same locality and season of the year presented the same value of genetic distance, which was 0.005 (Tables 3 and 4).A dendrogram was produced using the Neighbor-Joining (Saitou and Nei, 1987), and it was tested by 1,000 random permutations using bootstrap, with the values on the basal nodes.The analysis of molecular variance calculated for the sample partitioned in five localities and the four seasons showed values of F ST = 0.000 and 100% of genetic variation is intrapopulational, these values were significant (p = 0.000) only when the five collecting localities were jointly analysed (Table 5).
by genitalia (Freitas and Penny, 2001) and all specimens were stored in absolute alcohol for molecular analysis and numerically coded -voucher number.All the biological material used in the analysis is deposited in the Laboratório de Biologia Molecular de Crisopídeos do Departamento de Fitossanidade da FCAV-UNESP.

Total DNA extraction, polymerase chain reaction (PCR) and sequencing
The total DNA was extracted from a single individual preserved in ethanol, using the set of reagents Wizard ™ Genomic DNA Purification Kit (Promega), supplemented with 20 mg/ml Proteinase K, following the manufacturer's protocols with some small changes.The homogenate individuals were incubated at 55 ° C for 3 hours.A fragment of the mitochondrial gene encoding the first subunit of COI (cytochromo oxidase) was amplified via PCR (Mullins and Faloona, 1987) using the primers C1-J-2183 (5' CAACATTTATTTTGATTTTTTGG 3') and TL2-N-3014 (5' TCCATTGCACTAATCTGCCATATTA 3') (Simon et al.,1994).PCR was performed in 25 µL reactions containing 2 µL 10X buffer, 1.5 mm MgCl 2 , 0.4 µm dNTPs, 0.4 µm each primer, 1.5 unit Taq polymerase.The thermal cycling conditions were: initial denaturation at 94° C (2 minutes), 35 cycles of 94° C (40 seconds), 55° C (50 sec.), 72° C (1 second) and 72° C (10 minutes) for final annealing.Parts of the PCR products were run on 1% agar electrophoresis gel to check the amplification of DNA.All reaction products were purified by using the set of reagents Wizard ® SV Gel and PCR Clean-Up System (Promega), following the manufacturer's protocols.Purified PCR products were sequenced using the same primers used in PCR directly using BigDye™ Terminator version 3.1 (Perkin-Elmer Applied Biosystems).Sequencing was performed on an ABI Prism 3100 Genetic Analyzer, with the same conditions on PCR.After the sequencing, the fragment was submitted to several washes with isopropanol 75% followed by centrifugation.

Analysis of the sequences
The raw sequence chromatographs were edited using the program Chromas version 1.55 © (Technelysium Pty Ltd, 1998), and aligned using the program BioEdit v. 5.0.9 (Hall, 1999).Nucleotide diversity (Pi, Nei 1987, Eq. 10.5 or 10.6), haplotype diversity (Hd, Nei 1987, Eq. 8.4 and 8.12), total number of mutations Eta (η, Nei 1987, Eq. 10.3), average number of nucleotide differences (k) and number of polymorphic (segregating) sites (S) were calculated with DnaSP v. 4.00 for the specimens as a whole (Rozas et al. 2003).The average percentage of bases in all sequences of DNA samples was estimated using the program MEGA v.4 (Tamura et al., 2007).In order to estimate the levels of genetic distance, a distance matrix was calculated through the model of Kimura 2-parameter (Kimura, 1980), and the dendrogram was obtained using Neighbor-Joining (Saitou and Nei, 1987).It was tested by 1,000 random permutations using the MEGA v.4 (Tamura et al., 2007) program.The Arlequin v. 2.000 (Schneider et al., 2000) program estimated the hierarchical analysis of molecular variance (AMOVA, Excoffier et al., 1992), between the distinct places and seasons in which the individuals

Discussion and Conclusions
As has been said before, correct specific identification is essential for success in programs of biological control and therefore the knowledge of intraspecific variation is fundamental in this identification.Green lacewings (Chrysopidae) are an important group used in integrated biological control in field and horticultural crops (Horne et al., 2001;Senior and McEwen, 2001), but little is known about the genetic structure of this group.The species Chrysoperla externa is largely distributed, and therefore has an enormous potential for biological control, according to the  considerations of Mackauer (1976).Thus, the populational genetic structure of the species was examined, from five different places of collection and in all seasons of the year, utilising 36 individuals, which were sequenced and had a total of 805 nucleotide sites examined.

Fixation indices
During this study, it was possible to infer that the genetic distance is extremely low, with values that approximate 0.005 independently of the type of analysis -including either the total population, distinct collecting localities or season of the year.This indicates a great genetic similarity in this population of Chrysoperla externa in terms of sharing of haplotypes.Low values of genetic distance were also found for the group of cryptic species of the group Chrysoperla carnea (Lourenço et al., 2006), studied using four different gene markers, as was also the case in a study, using isozymes, performed utilising seven populations of green lacewings of the genus Chrysoperla (Wells, 1994).
This great genetic similarity found for the species Chrysoperla externa from Jaboticabal, SP highlights its genetic mitochondrial integrity, which independent of the collection locations and the season of the year sampled, still generated results indicating that it is the same population.This can be viewed in the dendrogram generated by the Neighbor-Joining method, which has three major branches, supported by the values of "bootstrap" (Figure 2).Each of the three branches contains results derived from individuals collected in all five localities and four seasons of the year uniformly distributed.Thus, it is not possible to identify a structure in this population, neither geographical nor seasonal.The absence of structure in this population may be due to the high degree of gene flow (indicated by the value of F ST = 0,000) among the individuals from distinct places, leading to the genetic homogeneity.
Therefore, is possible to conclude that the species Chrysoperla externa from Jaboticabal, SP, to this genetic marker -COI -forms a single population and that in the use of samples of this population for biological control, neither the collection location, within the general study area, nor the seasons of the year when the samples were collected, will interfere in the establishment of mass rearing of individuals for posterior field release.However, to understand the species as a whole, more studies are necessary, sampling additional populations from different localities and also utilising other, genetic and morphological markers, not only to understand each in situ population, but also to understand the evolutionary and systematic pattern of each one of them based on their respective genetic variability.

Table 1 .
Sample areas, identification codes (Voucher) and number of Chrysoperla externa individuals (N) analysed in Jaboticabal, SP.

Table 3 .
Nei's genetic distance calculated for five places of origin of Chrysoperla externa from Jaboticabal, SP.

Table 4 .
Nei's genetic distance calculated for seasons of years of Chrysoperla externa from Jaboticabal, SP.

Table 2 .
Polymorphic sites for mitochondrial gene COI of Chrysoperla externa.

Table 5 .
ANOVA analysis of Chrysoperla externa in five places.