Genetic diversity, phylogenetic and phylogeographic analyses of Oncideres impluviata (Germar, 1823) (Coleoptera: Cerambycidae) in Rio Grande do Sul state, Brazil

ABSTRACT The Cerambycidae Oncideres impluviata (Germar, 1823) is an important insect pest for Acacia mearnsii De Wild in Southern Brazil. The damage caused by their girdling behavior reduces tree productivity, specially in the early years of plant establishment, when girdling is performed on the main trunk of trees. Here, we used a fragment of the mtDNA COI gene to analyze the genetic diversity, population structure and demography of O. impluviata in Southern Brazil, as well as to present the first hypothesis of phylogenetic relationships among species of the genus Oncideres. Our results identified five distinct haplotypes among the populations of O. impluviata, with the most common haplotype identified as O.imp_COI_01. The phylogenetic inferences corroborated the monophyly of O. impluviata with maximum statistical support. In addition, the phylogeny recovered three main population strains that are largely congruent with the haplotype network, which includes two lineages that are found in different edaphic regions of Rio Grande do Sul (Serra do Sudeste and Encosta Inferior do Nordeste). This is the first molecular phylogenetic assessment of O. impluviata. Our findings provide insights into the evolution of a significant species for the Brazilian forestry sector, as well as new resources for planning of pest management strategies.


Introduction
The Brazilian forestry sector has significantly contributed to the country's economy, with eucalyptus and pine plantations occupying an estimated area of 7.53 and 1.93 million hectares, respectively (IBÁ, 2022).The forestry sector also includes species from genus Acacia Martius, 1829 (Fabaceae), whose contribution has been noted mainly in Southern Brazil, with 51.000 hectares of Acacia mearnsii De Wild (2021) in the State of Rio Grande do Sul (AGEFLOR, 2022).The cultivation of A. mearnsii, also known as black wattle, plays an important social and economic role, as it involves more than 35.000 family forest owners who work in the generation of products.These include plant extracts rich in tannins agents and phenols that originate tannins, which are used for leather tanning and animal nutrition; flocculants for water treatment processes (vegetable origin); dispersants used as additives where spraying of liquids is necessary; adhesives, resins and phenolicbased for industries; natural flavoring soluble in water for stimulating food consumption at all stages in the diet of pigs, chickens, sheep, cattle, fish and shrimp (AGEFLOR, 2022).
The increase in acreage and homogeneous plantations of black wattle in Rio Grande do Sul has also promoted an upsurge of insect pests Genetic diversity, phylogenetic and phylogeographic analyses of Oncideres impluviata (Germar, 1823)

A B S T R A C T
The Cerambycidae Oncideres impluviata (Germar, 1823) is an important insect pest for Acacia mearnsii De Wild in Southern Brazil.The damage caused by their girdling behavior reduces tree productivity, specially in the early years of plant establishment, when girdling is performed on the main trunk of trees.Here, we used a fragment of the mtDNA COI gene to analyze the genetic diversity, population structure and demography of O. impluviata in Southern Brazil, as well as to present the first hypothesis of phylogenetic relationships among species of the genus Oncideres.Our results identified five distinct haplotypes among the populations of O. impluviata, with the most common haplotype identified as O.imp_COI_01.The phylogenetic inferences corroborated the monophyly of O. impluviata with maximum statistical support.In addition, the phylogeny recovered three main population strains that are largely congruent with the haplotype network, which includes two lineages that are found in different edaphic regions of Rio Grande do Sul (Serra do Sudeste and Encosta Inferior do Nordeste).This is the first molecular phylogenetic assessment of O. impluviata.Our findings provide insights into the evolution of a significant species for the Brazilian forestry sector, as well as new resources for planning of pest management strategies.
Currently, the distribution of O. impluviata is restricted to southern South America, being reported in Argentina (Misiones), Brazil (South and Southeast regions), Paraguay, and Uruguay (Monné, 2022).This insect has specific morphological characters for identification (Fig. 1), measuring 13 mm -20 mm in length and 4 mm -6 mm in width.Dorsal region has a yellowish-brown color with grayish hairs and yellowish spots on the entire surface of the elytra and glossy black dots near the pronotum (Bondar, 1953;Amante et al., 1976;Pedrozo, 1980).Adults of O. impluviata girdle the branches that are going to be used by females to lay eggs (Baucke, 1962).This injury knocks down the branches, which become dry and suitable for feeding and development of larvae (Amante et al., 1976;Pedrozo, 1980).Girdled trees tend to bifurcate their trunk, which drastically reduces wood production (Costa et al., 2014).A high infestation of O. impluviata in the early years of forest establishment can result in the death of injured trees (Baucke, 1962).Despite its economic importance, no study has been carried out with the aim of characterizing the population genetic structure of O. impluviata in Southern Brazil.Genetic structure studies of this economically important insect pest can contribute to understand biological and evolutionary aspects that can be applied to pest management.
In general, mitochondrial genes have an advantage of high rates of evolution, compared to the nuclear DNA, and have relatively conserved regions (Meyer, 1994;Sosa-Gómez et al., 2012); therefore, are remarkably useful for organisms succinctly investigated (Guo et al., 2020).In the present study, we used a fragment of the mtDNA COI gene to analyze the genetic diversity, population structure, and demography of O. impluviata in the State of Rio Grande do Sul, Brazil.In addition, the phylogenetic relationships among the populations of O. impluviata, as well as among the species of the genus Oncideres Lacordaire, 1830 were explored.

Sampling of adult specimens of O. impluviata
The sampling procedure was performed in five sites of A. mearnsii plantations in Rio Grande do Sul, Brazil.Adults of O. impluviata were randomly collected from recently girdled branches which were lying on the ground (Fig. 2a, 2b; Table 1).Five to 15 adults were collected per site.Adult specimens were chosen to be captured instead of larvae because the branches girdled by O. impluviata often host other cerambycid opportunistic species, which could lead to inaccurate species identification (Paulino-Neto, 2004).
The samples were placed into 50 mL Falcon type plastic tubes containing 96% ethanol.The species was identified by comparison with the photo of the lectotype (Nearns and Tavakilian, 2015;Nearns and Nascimento, 2019), as well morphological characteristics (Bondar, 1953;Amante et al., 1976;Pedrozo, 1980).After identification, O. impluviata adults, both males and females, were stored at -20 °C until extraction of genomic DNA.

DNA extraction, amplification and sequencing
Genomic DNA extraction was performed using parts of adult specimens (head, legs or thoracic muscles).We used the Qiagen DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany), following manufacturer's instructions.Specimens were carefully removed from the Falcon tubes, placed on paper towel and air dried at room temperature.Specimens used for DNA extraction were vouchered and deposited in the Entomological Collection 'Padre Jesus Santiago Moure' (Universidade Federal do Parana, Curitiba, Brazil; DZUP), the Entomological Collection of the National Museum (Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; MNRJ) and Integrated Pest Management Laboratory (Universidade Federal de Santa Maria, Santa Maria, Rio Grande do Sul, Brazil).

Sequence editing and alignments
Sequence analysis and editing were performed using Pregap and Gap4 programs within the Staden package (Staden et al., 2000).Partial mtDNA COI sequences were aligned using the software BioEdit version 7.2.6 (Hall, 1999).These sequences were then translated into protein sequences to check for any translation problems and confirm the absence of premature stop codons.

Statistical analyses and genetic differentiation
The number of polymorphic sites (S), number of haplotypes (h), haplotype diversity (Hd ± SE), nucleotide diversity (π ± SE) and

Total
neutrality tests (Tajima's D and Fu FS) were calculated individually for each population and for the total populations using the DnaSP v. 5.10.01 (Librado and Rozas, 2009).The nucleotide composition was calculated using MEGA version 7.0.26(Kumar et al. 2016) and statistical analysis of parsimony was performed using TCS 1.21 (Clement et al., 2000), within the PopART program to generate a haplotype network (Leigh and Bryant, 2015).Paired population differentiation (F ST ) and analysis of molecular variance (AMOVA) were performed on Arlequin 3.5.2.2 (Excoffier and Lischer, 2010), and the significant values were estimated with 10,000 permutations.F ST values were used to estimate the number of female migrants per generation (1-F ST / 2F ST ) (Wright, 1943).For the AMOVA, the data were grouped according to the physiographic regions of the State of Rio Grande do Sul (http://coralx.ufsm.br/ifcrs/fisiografia.htm).The populations of Capela de Santana and Montenegro are in the physiographic region Encosta Inferior do Nordeste; General Câmara in the region known as Depressão Central; Canguçu and Encruzilhada do Sul in the Serra do Sudeste (Fig. 3).
The relationship between genetic distance (F ST ) and geographic distance (km) was used to test isolation by distance (IBD).The geographical distance between each population pair was estimated using the Google Maps Distance Calculator (https://www.daftlogic.com/projects-advancedgoogle-maps-distance-calculator.htm#), and analysis was performed using IBD version 1.52 with 10.000 randomizations (Bohonak, 2002).

Phylogenetic analyses
Phylogenetic analyses were performed to infer the relationships among the populations of O. impluviata and included 17 species of Oncideres and representatives of five genera of Onciderini (Supplementary material, Table S1), which were used as outgroups.The trees were rooted in Hypsioma hezia Dillon & Dillon, 1945 in all analyses.The data matrix used for the phylogenetic analyses comprised a total of 95 terminals and 612 aligned positions of COI.The alignment included both sequences newly generated for this study and sequences obtained from Barcode of Life Data System (BOLD) (http://www.boldsystems.org/).Sequences obtained from BOLD were only reduced to suit them to the length of our alignments.Phylogenetic analyses were performed under Bayesian and Maximum Likelihood criteria.
Prior the phylogenetic analyses, the dataset was partitioned considering codon positions separately and the best-fitting model of nucleotide substitution for each partition was selected under Akaike Information Criterion (AIC) in PartitionFinder 2.1.1 (Lanfear et al., 2017).Bayesian inference (BI) was carried out using MrBayes 3.2.6 (Ronquist et al., 2012), using partitioned data with their corresponding best-fitting evolutionary model (SYM+G, HKY+I, TRN+G for 1st, 2nd and 3rd codon position, respectively) with eight chains of 50 million generations each and sampling trees every 1000 generations.The analysis convergence was assessed in Tracer 1.7.1 (Drummond and Rambaut, 2007) and the maximum clade credibility tree was generated discarding the output of the 10% initial generations in TreeAnnotator 1.8.4 (Heled and Bouckaert, 2013).Posterior probability values (PP) were used to infer branch support in the BI tree.Maximum likelihood (ML) analysis was performed in RAxML 7.2.6 (Stamatakis, 2006) using unlinked GTR + I + G nucleotide substitution models for each partition with 1.000 bootstrap (BS) pseudo-replicates.Both BI and ML analyses were implemented in CIPRES (Miller et al., 2010).

Genetic diversity and phylogenetic analyses
A total of 48 sequences of COI (with 612 bp) from five populations of O. impluviata from Rio Grande do Sul, Brazil, were analyzed in the genetic diversity study.Twenty-six polymorphic sites (S), haplotype diversity (Hd) of 0.520 ± 0.073 and nucleotide diversity (π) of 0.01479 ± 0.00206 were observed when combining all sequences (Table 2).Hd in the municipalities of CA and MN varied from 0.564 to 0.800, and π ranged from 0.01014 to 0.01827.In addition, the neutrality tests (Tajima´s D: -0.34857, P = 0.70; and Fu´s Fs: 2.06, P = NA) were not significant, suggesting a pattern of constant size in populations (Table 2).The average nucleotide composition among the 48 sequences of O. impluviata was 38.3% thymine (T), 17.2% cytosine (C), 28.5% adenine (A) and 15.9% guanine (G).
Both BI and ML phylogenetic inferences generated congruent results, supporting the monophyly of O. impluviata with strong statistical support and recovering three well-supported clades within the species: two closely related including the O.imp_COI_02 and O.imp_COI_03, both comprising samples from CA and MN; and one containing O.imp_COI_01, O.imp_COI_04 and O.imp_COI_05, comprising all samples from CS, ES, GC and one sample from each CA and MN (Fig. 4).In addition, our analyses recovered a clade containing all samples of a close related species, Oncideres saga (Dalman, 1823) (PP = 0.83; BO = 71).Oncideres was recovered as paraphyletic in relation to Psyllotoxus griseocinctus Thomson, 1868 (PP = 1; BO = 95), which was recovered as sister group of Oncideres cervina Thomson, 1868.Except for Oncideres digna Bates, 1865 and Oncideres putator brevifasciata Dillon & Dillon, 1946 all other Oncideres species sampled by more than one specimen in the analyses were strongly supported as monophyletic (Fig. 4).

Genetic distance between populations
Pairwise genetic distances among the populations of O. impluviata evidenced significant distances (Table 3).However, the analysis of female migrants per generation showed a weak and low evidence for this hypothesis.Furthermore, O. impluviata populations showed no statistically significant correlation of genetic distance (F ST ) vs geographical distance (P = 0.572).

Molecular analysis of variance (AMOVA)
The analysis of molecular variance (AMOVA) was conducted grouping populations according to the physiographic regions of Rio Grande do Sul: Serra do Sudeste, Depressão Central and Encosta Inferior do Nordeste.The results showed significant genetic differentiation among populations within groups (Φ SC = 0.85304; P < 0.0001) and within populations (Φ ST = 0.72641; P < 0.0001).No statistical significance differences were observed by grouping within physiographic regions (Φ CT = -0.8617;P = 1.0000) (Table 4).

Discussion
Genetic diversity, genetic distance, and molecular analysis of variance Specimens of O. impluviata, an important insect pest of A. mearnsii in Southern Brazil, were collected in five Acacia plantation sites in Rio Grande do Sul.Analyses of COI resulted in five haplotypes, representing the first molecular characterization and genetic diversity study of O. impluviata.In addition, our data are limited to an analysis of a partial mtDNA cytochrome Oxidase I fragment.
Although the specimens of O. impluviata were collected in its native range, the genetic diversity among the sampled populations was relatively low, contrary to our expectations of higher genetic diversity which is usually expected from native populations (Puillandre et al., 2008).We also found that O.imp_COI_01 was the most common haplotype in the localities with the larger areas of cultivation of A. mearnsii, such as General Câmara, Capela de Santana and Encruzilhada do Sul.The two lineages (O.imp_COI_04 and O.imp_COI_05) found in different edaphic regions, Serra do Sudeste and Encosta Inferior do Nordeste, should be monitored in further landscape adaptation studies.The force of gene flow in each landscape can balance the genetic diversity and form frequency of specific lineages (Dong et al., 2021).The reduction of native areas, the expansion of acaciculture at Rio Grande do Sul, along with the proximity of plantations A. mearnsii associated with unappropriated management strategies, have contributed significantly to the increase in population density and dispersion of O. impluviata.Older plantations of black wattle can be important bridges of migration of O. impluviata to new plantations, especially in areas where legislative measures are not taken for this insect pest management (Costa et al., 2014).This same pattern was observed in old plantations of Mimosa scabrella Bentham (Baucke, 1958) which is a tree species from the same botanical family as A. mearnsii.
The preference of O. impluviata for A. mearnsii may be associated to the volatile compounds present in this plant species (Berkov et al., 2000).These are fundamental to stimulate feeding, copulation, and oviposition (Paulino-Neto, 2004).Studies should be carried out to know which compounds emitted by A. mearnsii, particularly in the initial establishment in the field, are attractive to this insect.
Low diversity of O. impluviata can be associated with biotic and biotic factors, in addition to its limited samples.Oncideres impluviata is a univoltine insects (Costa et al., 2014;Ono, 2015), as well polyphagous, and reproduces through sexual reproduction, it is worth to note that females of this genus usually oviposit one egg per incision, but there are cases of multiple eggs per branch (Rogers, 1977;Solomon, 1995;Paulino-Neto, 2016).They seek to lay eggs on plants that, due to their characteristics (branch diameter, age, height, vigor, nutritional quality, and others), provides greater fitness to their offspring (Paulino-Neto, 2004).Therefore, it is possible that more than one specimen emerged from the same branch or other branches of the same tree, all of which are maternally related (Machado et al., 2020b), as possibly is the case of insects found in Capela de Santana, Encruzilhada do Sul and General Câmara.
The correlation between genetic distance (FST) and geographical distance didn't show any statistically significance, which suggests absence of distance isolation.However, pairwise genetic distance shows significant population structure when comparing the collection sites (GC x MN, GC x CA, MN x CS, MN x ES, CS x CA, and CA x ES), regardless of its geographic distance.This suggests that most populations are not interbreeding freely, excepting some low FST value (GC x CS, GC x ES, CS x ES, and MN x CA).These pairwise comparisons agree with the AMOVA results, which showed that the populations presented considerable degree of differentiation, within populations lacking haplotype sharing.The number of female migrants of O. impluviata between the collection sites was not relevant in this study.
Still, new specimen collections should be carried out in order to expand the possibilities of finding more haplotypes in another areas in Rio Grande do Sul state and Brazil.

Phylogenetic analyses
Our results support the monophyly of O. impluviata, recognizing three main population strains that are largely congruent with the haplotype network (clades A, B and C), but the statistical support among the clades was negligible to infer with confidence their relationships.Oncideres impluviata is morphologically close to Oncideres guttulata Thomson, 1868.Both species present rounded yellowish pubescent spots on the elytra and a white pubescent macula on the metaventrite (Dillon and Dillon, 1946).However, our analyses did not recover an immediate relationship between these species.Instead, O. saga was found to be the closest relative of O. impluviata, although with moderate statistical support.This is the first molecular phylogenetic study in the genus Oncideres.Our results reveal two important questions about the evolution of Onciderini: (1) the placement of Psyllotoxus within Oncideres, corroborating the close relationship between these genera, as suggested by some authors in the past (Thomson, 1868;Lacordaire, 1872), and (2) hypotheses about the relationship between the species of Oncideres.However, it is still premature to propose taxonomic changes based on our sampling.These findings suggest evidence to redefine the systematic limits of Oncideres and must be taken into consideration in future phylogenetic and systematic studies.

Implications and insights for pest management
In the agricultural area, insect pests as the Lepidoptera Chrysodeixis includens (Walker, 1858) (Lepidoptera: Noctuidae) can get lineages adaptation not only at the local landscapes (Silva et al., 2020b) but also between countries (Perini et al., 2020).Thus, edaphic regions in Rio Grande do Sul can force and create lineages of O. impluviata adapted to specific conditions that should be kept under investigation in further years.
Understanding the occurrence and ecology of O. impluviata in areas where new A. mearnsii tree are being cultivated is essential for the management of this pest species.Main strategies to manage O. impluviata populations includes a rotation every 6-7 years, without thinning or pruning and population density of 2-2.5 thousand trees per hectare (AGEFLOR, 2022).Besides our finding of five genetic lineages, pest management strategies at a local-scale management of O. impluviata might help to reduce local populations which will not be boosted by long-distance dispersal of adults to another regions.
Furthermore, the destruction of twigs girdled by this species seems not to be efficient, since that O. impluviata populations continues to increase in some regions of Rio Grande do Sul state (2006 to 2013), affecting the black wattle plantations (Ono, 2015).This requires, besides legislative tactics, other integrated pest management strategies, including chemical and biological control, plant resistance, behavior control and cultural activities should be performed to manage this pest.

Figure 2
Figure 2 Injuries caused by Oncideres impluviata to Acacia mearnsii in the State of Rio Grande do Sul, Brazil.Girdled fallen branches in a Acacia plantation in General Câmara.Red arrows show branches girdled by O. impluviata (a).Adults of O. impluviata copulating and girdling the main trunk of a young Acacia tree in Encruzilhada do Sul (b).

Figure 3
Figure 3 Physiographic regions of Rio Grande do Sul state, Brazil.

Figure 4
Figure 4 Phylogenetic tree summarizing the results of Bayesian inference (BI) and Maximum likelihood (ML).Tree shows the relationships among species of Oncideres along with the haplotype network of five populations of Oncideres impluviata from Rio Grande do Sul, Brazil.A, B and C depicts clades within Oncideres impluviata.The circle areas in the haplotype network are proportional to the frequencies of each haplotype and hatch markers represent the number of differences among haplotypes.

Table 1
Sampling sites, dates and the number of Oncideres impluviata specimens sequenced from each location.

Table 2
Results of the polymorphism analysis and neutrality tests based on the mtDNA (COI fragment) of five populations of Oncideres impluviata in Rio Grande do Sul, Brazil.FS neutrality tests and their respective p values were not included in the table because they have the same haplotype.

Table 4
Results of the AMOVA analyses of five populations of Oncideres impluviata.Va, Vb, and Vc are the associated variance components and Φ CT , Φ SC , and Φ ST are the fixation indices.