Beetle Population Structure at the Crossroads of Biogeographic Regions in Eastern Europe: the Case of Tatianaerhynchites Aequatus (Coleoptera: Rhynchitidae)

NORTH-WESTERN JOURNAL OF ZOOLOGY 12 (1): 166-177 ©NwjZ, Oradea, Romania, 2016 Article No.: e151203 http://biozoojournals.ro/nwjz/index.html

Beetle population structure at the crossroads of biogeographic regions in Eastern Europe: The case of Tatianaerhynchites aequatus (Coleoptera: Rhynchitidae)

Natalia MUNTEANU MOLOTIEVSKIY1,*, Anna MOLDOVAN1, Svetlana BACAL2 and Ion TODERAS1

1. Laboratory of Systematic and Molecular Phylogeny, Institute of Zoology, Academy of Sciences of Moldova, 1 Academiei Str., Chisinau MD-2028, Republic of Moldova. 2. Laboratory of Entomology, Institute of Zoology, Academy of Sciences of Moldova, 1 Academiei Str., Chisinau MD-2028, Republic of Moldova *Corresponding author, N. Munteanu Molotievskiy, Tel: +373 22 73 98 09; Fax +373 22 73 12 55, E-mail: [email protected]

Received: 16. May 2014 / Accepted: 12. August 2015 / Available online: 30. May 2016 / Printed: June 2016

Abstract. Tatianaerhynchites aequatus (L.) (Coleoptera: Rhynchitidae), is an economically important pest species of fruit crops (Rosaceae). This study investigates population genetic structure of T. aequatus at the crossroads of biogeographic regions in Eastern Europe. To estimate gene flow and genetic variation, we sequenced a 794-bp fragment of the mitochondrial cytochrome oxidase I (cox1) gene in 52 specimens from 13 populations across the Republic of Moldova, Romania and Ukraine. Percent divergence among populations ranged from 0.09% to 2.14%. Nested clade analysis of 15 haplotypes showed evidence of genetic structuring that is inferred to be restricted gene flow/dispersal but with some long distance dispersal. Analysis of molecular variance showed genetic structuring and some level of restricted gene flow among populations. Haplotypes formation and distribution are related to biogeographic regionalization. We revealed that haplotype B is predominant, being the only one spread across all three biogeographic regions, indicating that this haplotype has a high adaptive potential. Knowledge on population structure and haplotypes with different adaptive potential may be important for pest management and biodiversity conservation programs.

Key words: Tatianaerhynchites aequatus, weevil, mtDNA, population structure, Eastern Europe, biogeographic regions, crossroads.

Introduction 1999). Thus, at large scales the intersections of biogeographic zones may be evolutionarily active The distribution of species on Earth is not ho- zones, the origination sites of new taxa or adapta- mogenous but follows multifactorial patterns, tions in existing taxa (Spector 2002). Naturally, called ecoregions, determined by climate, geology zones where two or more biogeographic assem- and their changes in time. Ecoregions are not blages come into contact have high genetic diver- separated by fixed boundaries; these rather, repre- sity. Concentrating genetic material and driving sent zones within which complex ecological and up evolutionary processes, biogeographic cross- evolutionary processes most strongly interact. roads present great interest for both purposes of Ecological heterogeneity is the essential feature of biodiversity conservation and management of biogeographic crossroad territories and deter- economically important species. mines the variety of selective regimes to which The Republic of Moldova is located in south- species adapt. But classical evolutionary theory eastern Europe at the crossroads of three bio- maintains that gene flow between populations ef- geographic regions: I) Region of European De- fectively counters the effects of divergent selection ciduous Forests (54.13% of total area), II) Region of on connected populations (Mayr 1963). Recent Mediterranean Forests (15.59%) and III) Region of evidence suggests, however, that ecological het- Euro-Asian Steppe (30.28%) (Rodríguez 2009) (Fig. erogeneity and strongly differing selection pres- 1A). The Region of European Deciduous Forests sures across ecotones may drive rapid evolution- covers a large bioclimatic and geologic range from ary changes in populations even in the face of oceanic to continental forests, from floodplain to gene flow, perhaps driving evolution and speci- mountain forests up to the alpine timberline. The ation at edges of habitats and enriching biotas (En- ecoregion consists of lowland to submontane aci- serink 1997, Smith et al. 1997, Schneider et al. dophilus oak and mixed oak forests, mixed oak-

Population structure of Tatianaerhynchites aequatus in Eastern Europe 167

A B Figure 1. Maps of the Republic of Moldova indicating: A – Biogeographic regions; B – Natural districts.

hornbeam forests, as well as lowland to submon- steppe ecosystems are characterized by xerophyte tane hemiboreal and nemoral pine forests on the species, including perennial grasses, ephemeral North European glaciated plain. It is bordered by species, and sub-shrubs. The steppe flora is rich the lowland-colline subcontinental meadow and diverse, most of species belonging to four steppes and dry grassland vegetation on the east- families: Asteraceae, Fabaceae, Poaceae and La- ern side, hemiboreal spruce and pine-spruce for- miaceae. Four districts are recognized within Re- ests to the north, beech and mixed beech forests of gion of Euro-Asian Steppe: Beltsy Steppe District the Carpathians to the south, and the beech and (3), Chiuluc-Solonets Steppe District (4), Bugeac mixed beech forests of the Baltic and Western Steppe District (10) and the Lower Dniester Steppe Europe to the west (Spiecker 2003). Seven districts District (12) (Gheideman 1966, Cazantev 2002). are recognized within Region of European De- The apple fruit weevil, Tatianaerhynchites ciduous Forests across Moldova: Northern- aequatus (Linnaeus, 1767) (Coleoptera: Rhynchiti- Moldovan Forest-Steppe District (see nr. 1 on dae) is known as an important pest of fruit crops map), the Middle Prut Forest-Steppe District (2), in Europe (Ter-Minasyan 1950, Kryzhanovskii Dniester-Reut Forest-Steppe District (5), the Mid- 1974, Savkovskiy 1976, Zayanchkauskas 1984, Ke- dle Dniester Forest-Steppe District (6), Codry For- hrli & Pasquier 2013). The weevil inhabits most of est District (7) and the Lower Prut Forest-Steppe Europe (except north), Caucasus, Eastern Kazakh- District (8) (Fig. 1B). The Region of Mediterranean stan, Turkmenistan, Asia Minor, Syria, Iran and Forests is generally under semi-arid conditions, Jordan (Angelov 1980, Legalov 2003). characterized by low forest cover. The forest vege- In Moldova the apple fruit weevil was first re- tation is generally composed of open woodlands corded in 1912 by Vitcovskii (1913). Additional in- with scattered trees and xerophytic shrubs; the formation on the damage caused by T. aequatus in main tree vegetation is represented by Quercus pu- fruit tree orchards and measures for its control are bescens. Two districts are recognized within this presented in the report of the bio-entomological biogeographic region: South-Moldovan Forest- station conducted over ten years (Veresceaghin Steppe District (9) and the Lower Dniester Forest- 1922). Information on the species distribution and Steppe District (11). Region of Euro-Asian Steppe abundance within the country can be found in is a vast ecoregion of Eurasia in the temperate some agricultural and faunistic studies (Medvedev grasslands, savannahs and shrub lands biome. The & Shapiro 1957, Veresciaghin 1964, Vinnichenko

168 N. Munteanu Molotievskiy et al.

1988, Poiras 1998, Munteanu 2006). aequatus and to provide quantitative estimates of Adults of apple fruit weevil feed on the young maternal gene flow, we assessed sequence varia- buds from mid spring, and later on the still green tion among populations throughout all three bio- fruits. From early to mid-summer the eggs are laid geographic regions across Moldova and in seeds of the young fruits, and the fruits stalk is neighbouring countries, Romania and Ukraine. gnawed so the fruit dries up and falls off. Larval We amplified and sequenced a 794 bp fragment of development and maturation occur inside the mtDNA from the cytochrome oxidase I (cox1) re- dried out or already fallen fruit. Given the ecologi- gion because this segment is known to have un- cal conditions in the researched area, the apple usually high sequence variation in other curcu- fruit weevil is a bivoltine species (Gontarenko lionoid species (Langor & Sperling 1997; Laffin et 1978). Adults and larvae overwinter in the soil at a al. 2004, Kim & Sappington 2004, Ruiz et al. 2009, depth of about 10-18 cm (Poiras 1998). Riedel et al. 2010), and has been used to elucidate Damages caused by pest insects significantly population structure, taxonomy, and phylogenetic affect fruit yield and its quality. In order to de- relationships of many insect species (Boyce et al. velop sustainable management plans and avoid 1994; Langor & Sperling 1995, 1997; Sakai et al. depletion of insect genetic resources, it is essential 2001; Laffin et al. 2005a, b; Sanchez-Sanchez et al. to accumulate knowledge on pest species popula- 2012; Sequeira et al. 2012). Rapid evolution rate of tion structure (Avise et al. 1987). Genetic analyses mtDNA allows detection of differences among have thus become an important tool in many stud- populations that diverged a relatively short time ies of pest species dispersal and demography, ago. Such sequences provide phylogenetic infor- monitoring of current infestations and the preven- mation that may be analysed to identify bio- tion of further ones, and aid in the planning and geographic regions with genetically distinct popu- implementation of control measures (Estoup & lations and to uncover historical patterns respon- Guillemaud 2010, Ridley et al. 2011, Kirk et al. sible for current genetic structure. 2013). Using genetic markers the evolutionary history of a group can be investigated to determine whether smaller management units may exist be- Materials and Methods low the species level (Moritz 1994, Knapen et al. Live adults of T. aequatus were obtained from 13 localities 2003). Recent studies suggest that species as a unit across Moldova, Romania and Ukraine (Fig. 2). All speci- may not be the appropriate scale at which diver- mens were collected between April 2007 and July 2008 by sity should be measured (Hughes et al. 1997, Luck sweep netting from leaves and flowers of Rosaceae trees, et al. 2003). In case of restricted gene flow among using a standard sweep net, 25 cm in diameter, at the rate populations, important species traits may become of 50 sweeps per plot. Insects were killed either by ex- geographically structured, and local populations tended freezing at − 20°C or by placing them in 70 - 100% ethanol and storing them at − 70°C before DNA extrac- may exhibit only a part of the ecological functions tion. Specimens were identified based on morphological of entire species (Hughes et al. 1997). In order to characters (Legalov 2003). avoid failure of managing all functions provided A total of 52 individuals were selected for genomic by the pest, management programs with regard to DNA extraction, four per population (Table 1). The head population structure are needed. and thorax of each individual were used for the extraction Only few studies on population genetic struc- with a Promega Wizard® Genomic DNA Purification Kit. ture of curculionoid species in Europe are known Corresponding vouchers were placed in the collections of Entomology Museum at the Institute of Zoology, Acad- (Kajtoch et al. 2009, Kajtoch 2011, Kajtoch et al. emy of Sciences of Moldova. Extracted DNA was stored 2012, Avtzis & Cognato 2013, Bertheau et al. 2013). at − 20°C before amplification by polymerase chain reac- This is the first study regarding population genetic tion (PCR). A 794-bp segment of the 3' end of the cyto- structure in T. aequatus populations. The results chrome oxidase I (cox1) gene was amplified. Amplifica- will contribute to the understanding of species tion reactions were performed in a 25 μL reaction solu- colonization and establishment ability, and further tion. Each PCR mixture contained 12.5 μL of double- elaboration of management practices. Population distilled H2O, 2 μL of 25 mmol/L MgCl2, 2.5 μL of 10x PCR buffer, 0.75 μL of 10 mmol/L dNTPs, 3 μL of DNA genetic structure studies at the crossroads of bio- template, and 2 μL of 10 μmol/L primers (Simon et al. geographical regions can serve as a model and 1994): forward TL2-N-3014 (5′ - help us reveal genetic material dispersion poten- TCCAATGCACTAATCTGCCATATTA-3′) and reverse tial. C1-J-2183 (5′-CAACATTTATTTTGATTTTTTGG-3′). Taq To understand the population structure of T. DNA polymerase 0.2 μL (1 Unit) was added as final step

Population structure of Tatianaerhynchites aequatus in Eastern Europe 169

Figure 2. Map indicating the sampling localities and haplotype distribution of Tatianaerhynchites aequatus in the Repub- lic of Moldova and neighbouring countries Romania and Ukraine. Symbols for populations (sampling sites) Bac, But, Cal, Cap, Chi, Erj, Ior, Mic, Pan, Per, Vol, Sta and Tar, were derived from the name of cities, towns or villages situated nearest to that population (see Table 1 for full names). Legend refers to most common haplotypes, with unique or sin- gleton haplotypes designated by letters on pie charts. The size of each pie chart reflects relative sample size of that sample.

Table 1. Sampling localities of Tatianaerhynchites aequatus, geographical coordinates, number of individuals and haplotypes identified in each sampling locality.

Locality Country Latitude Longitude Sample Haplotypes Biogeographic (N) (E) size (n) (no. individuals) regions Bacioi (Bac) Moldova 46°54′44″ 28°53′02″ 4 B(2), I(1), O(1) II Butuceni (But) Moldova 47°32′01″ 29°01′33″ 4 C(2), L(2) I Calfa (Cal) Moldova 46°54′15″ 29°22′31″ 4 I(1), J(3) II Capriana (Cap) Moldova 47°07′03″ 28°30′06″ 4 G(2), F(2) I Chisinau (Chi) Moldova 47°00′20″ 28°51′27″ 4 A(2), B(2) I Erjovo (Erj) Moldova 47°48′13″ 29°00′24″ 4 C(3), L(1) I Iordanovca (Ior) Moldova 46°23′38″ 28°54′33″ 4 B(1), H(3) III Micauti (Mic) Moldova 47°10′20″ 28°44′56″ 4 G(2), A(1), B(1) I Panasesti (Pan) Moldova 47°09′11″ 28°31′07″ 4 E(3), F(1) I Peresecina (Per) Moldova 47°15′08″ 28°46′08″ 4 K(1), M(2), L(1) I Volintiri (Vol) Moldova 46°25′36″ 29°36′21″ 4 D(4) III Starokozache (Sta) Ukraine 46°20′14" 29°59′07″ 4 D(4) III Tartaria (Tar) Romania 45°56′02″ 23°24′50″ 4 N(3), K(1) I

before samples were placed into GeneAmp PCR System min. Each experiment was associated with negative 9700 thermal cycler (Applied Biosystems, California, (without DNA template) control. PCR products were USA). Amplification was performed with 39 cycle pro- verified by agarose gel electrophoresis, and cleaned using gram (each cycle consisting of denaturation at 94°C for 2 the QIAquik PCR Purification Kit (Qiagen). Sequencing min, annealing at 54°C for 30 s and extension at 72°C for 1 was from both strands with the same primers used for min), followed by a final extension step at 72°C for 10 PCR amplification, using BigDye 2.1 and an ABI PRISM

170 N. Munteanu Molotievskiy et al.

3730 automated sequencer (Applied Biosystems, Foster Results City, CA, USA). Sequence chromatograms were assem- bled and edited with Sequencher 4.6 (Gene Codes Corp., A total of 52 specimens were analysed. The 794 bp Ann Arbor, Michigan). fragment of mitochondrial cox1 gene was ampli- For each population, haplotype diversity (h) and nucleotide diversity (π) was calculated using Nei (1987) fied and successfully sequenced. Tatianaerhynchites equations. The average number of nucleotide differences aequatus was represented by 13 populations. But, (k) was calculated using equation of Tajima (1983). Fu’s Fs Cap, Chi, Erj, Mic, Pan, Per, and Tar belong to the test statistic (Fu 1997) was calculated by using total segre- Region of European Deciduous Forests (Region I); gating sites. Kimura two-parameter genetic distances be- Bac and Cal to the Region of Mediterranean For- tween haplotypes were calculated. Sequence alignments ests (Region II); and Ior, Vol and Sta to the Region were performed by means of the CLUSTAL W function of Euro-Asian Steppe (Region III) (Figs. 1 and 2). (Thompson et al. 1994) of the MEGA 4.0 program (Ta- mura et al. 2007) using an IUB DNA. Alignment statistics, Nucleotide substitutions among all individu- as e.g. the identification of phylogenetically informative als of T. aequatus were found at 34 sites spread sites, were assessed in MEGA 4.0. TREE PUZZLE 5.2 throughout the entire segment, accounting for software (Schmidt et al. 2002) was used to estimate data 4.3% of all bases (Table 2). Of the variable sites, set specific parameters as transition/transversion ratios four occurred in 3rd base position, 23 in 2nd base and the -parameter for the Γ-distribution based correc- position and five occurred in the 1st base position tion of rate heterogeneity among sites. For phylogenetic of codons. From revealed substitutions, eight were reconstruction from nucleotide sequence alignments, the transversions (A↔T=3.93; A↔C=1.83; T↔A=3.18; most appropriate models of DNA sequence evolution were chosen according to the rationale outlined by T↔G=1.63; C↔A=3.18; C↔G=1.63; G↔T=3.93; Posada & Crandall (1998). Phylogenetic trees were con- G↔C=1.83), and four were transitions (A↔G= structed using a subset of sequences. We retained only 16.5; T↔C=9.16; C↔T=19.64; G↔A=33.11). The one sequence for each haplotype. Organism phylogenies nucleotide frequencies were estimated as follows: were constructed with the Neighbour Joining (NJ) A = 30.05%, T/U = 37.22%, C = 17.35%, and G = method in MEGA 4.0 software tool (Tamura el al. 2007) 15.38%. using the Kimura two-parameter (K2P; Kimura 1980) Among the 15 haplotypes of T. aequatus (Table models of nucleotide substitution and γ-distribution based model of rate heterogeneity (Yang 1993). Addi- 1), haplotype B was most frequently observed (six tional Maximum Parsimony (MP) phylogenies were con- individuals from four populations). Haplotype L structed in MEGA 4.0 using the close-neighbour- consisted of four individuals from three popula- interchange (CNI) for heuristic search option with initial tions. Haplotypes I and K were each shared by tree by Random addition (100 replicates). Tree topology two individuals from two populations, haplotypes confidence limits were explored in non-parametric boot- A and F by three individuals from two popula- strap analyses over 1,000 pseudo-replicates. Neocoe- tions, G by four individuals from two populations, norhinidius pauxillus Germar and Byctiscus betulae L. (Col- eoptera: Rhynchitidae) were used as outgroups. C by five individuals from two populations and D Nested clade analysis was conducted using two pro- by eight individuals from two populations. The grams. First, TCS 1.18 (Clement et al. 2004) was used to other six haplotypes were each unique to one construct a network of haplotypes with 95% confidence population, one of them O being unique to one in- limit for parsimony (Templeton et al. 1995) and a clado- dividual, M containing two individuals and the gram showing nesting structure of the haplotypes. Sec- rest E, H, J and N by three individuals (Table 1, ond, GeoDis 2.6 (Posada et al. 2000) was employed to cal- Fig. 3). Haplotypes A, C, E, F, G, K, L, M, and N culate distances between populations where haplotypes or clades were found, based on the cladogram generated were spread through Region I, haplotypes I, J, O by TCS 1.18 and coordinates of sampling localities (Table through Region II, and Haplotypes D and H 1). Templeton's inference key (Templeton 2011) was ap- through Region III. Only haplotype B was found plied to determine the likely mechanism of the observed in all 3 biogeographical regions (Fig. 2). genetic structure. Sequence divergence among haplotypes Nucleotide sequence diversity was calculated using ranged from 0.09% to 2.14%. Within thirteen DNAsp v.5 (Librado & Rozas 2009). Arlequin 3.5 (Excof- populations of T. aequatus, the lowest genetic di- fier & Lischer 2010) was used to estimate F statistics indices for genetic variation between populations of T. aequa- versity parameter was revealed in the Cal popula- tus. tion and the highest in the Bac population. Fu’s statistics showed a negative value of FS for But population (Table 3). Parsimony analysis provided adequate resolu-

tion of the relationships among haplotypes (Fig.

Population structure of Tatianaerhynchites aequatus in Eastern Europe 171

Table 2. Nucleotide sequence differences between observed haplotypes mtDNA cytochrome oxidase I (cox1) gene of Tatianaerhynchites aequatus collected at locations in the Republic of Moldova, Romania and Ukraine. Dots (.) refer to invariant sites, hyphens (-) denote an alignment gap/missing data.

Nucleotide position

Haplo-

type 7 8 9 16 68 74 92 98 110 143 173 230 266 311 313 323 395 416 425 452 527 566 578 707 710 722 726 772 778 780 786 792 794 107 A - A T T C A T T A G A T G T G G T C A T T A G C T A A A A G T A A T B - ...... A . . . . C - . . . . G ...... A . . . . D - ...... A . . . A E - . . . . . C ...... A . . . . F - ...... G ...... T . . . . . A . . . . G - ...... A . . . T . . . . C ...... A . . . . H G ...... A ...... A . . . G I - . C A T ...... A ...... T . T G A C G G G J - . C A T . . C ...... T . T G A C G G G K - . . . . G . . G A T . A C . . A T G . . G A . C . . . . A . . . . L - . . . . G . . G A T . A C . . A T G . . G A . C . T . . A . . . . M - . . . . G . . G A T . A C . . A T G C . G A . C . . . . A . . . . N G . . . . G . . G A T . A C . . A T G . . G A . C . . . . A . . . . O G . . . . G . . G A T . A C A . A T G . . G A . C . . . . A . . . .

NJ MP Figure 3. Phylogenetic trees constructed from 794 bp nucleotide sequences of the mitochondrial cox1 gene of Tatianaerhynchites aequatus. Right: Maximum-parsimony tree (MP). Left: Neighbour-Joining tree with Kimura two-parameter distances. Bootstrap values over 50 are given for both MP and NJ. Letters in parentheses denote codes of the populations (see Table 1) and the number following the code letter represents the number of individuals in the population with the identical haplotype. I and II and III indicate the biogeographical regions.

3). Fifteen haplotypes were grouped into two clus- haplotype network had three main clades contain- ters in both NJ and MP trees. The first group in- ing, eight one-step clades, and three two-step cludes 10 haplotypes of 11 populations from all clades. three biogeographic regions. The second group in- Nested contingency analysis (Table 4) showed cludes five haplotypes from five populations associations between geographic locations and within boundaries of Region I and II. Bootstrap clades at both first and second clade levels. From support for first group was 75% in NJ and 53% in nested clade analysis we inferred one pattern at MP, while for second it was 100% and 99%, re- first clade level and one pattern at second level. spectively. The two inferred patterns were restricted gene Haplotype network construction yielded a flow/dispersal but with some long distance dis- single network with a maximum of 26 mutational persal (clade 1-2), and restricted gene flow with steps between any two haplotypes (Fig. 4). The isolation by distance (restricted dispersal by dis-

172 N. Munteanu Molotievskiy et al.

Table 3. Genetic diversity parameters and neutrality test Fu’s (FS) in the mtDNA cytochrome oxidase I (cox1) sequences within 13 populations of Tatianaerhynchites aequatus.

Population Sample Haplotype Average number Nucleotide Fu’s size (n) diversity (h) of nucleotide differences (k) diversity (π) statistic (FS)* Bacioi (Bac) 4 0.833 11.6667 0.01471 2.885 Butuceni (But) 4 0.667 8.0000 0.01009 -5.100 Calfa (Cal) 4 0.500 1.0000 0.00126 1.099 Capriana (Cap) 4 0.667 3.3333 0.00420 3.153 Chisinau (Chi) 4 0.667 0.6667 0.00084 0.540 Erjovo (Erj) 4 0.500 6.0000 0.00757 4.419 Iordanovca (Ior) 4 0.500 1.0000 0.00216 1.099 Micauti (Mic) 4 0.833 2.5000 0.00316 0.461 Panasesti (Pan) 4 0.500 1.5000 0.00189 1.716 Peresecina (Per) 4 0.833 1.1667 0.00147 0.650 Volintiri (Vol) 4 0.500 5.5000 0.00694 4.220 Starocazacie (Sta) 4 0.667 0.6667 0.00084 0.540 Tartaria (Tar) 4 1.000 7.0000 0.00885 0.461

*Fu’s (FS): A negative value of FS is evidence of an excess number of alleles, as expected from recent population expansion or

from genetic hitchhiking. A positive value of FS is evidence for a deficiency of alleles, as expected from a recent population bottleneck. Statistical significance: Not significant P ˃ 0.02.

Figure 4. Haplotype network and nested design for 15 haplotypes of Tatianaerhynchites aequatus. Letters in circles represent haplotypes (see Table 1) and lines between haplotypes represent a one-step muta- tion changes. Dotted boxes represent one-step clades, dashed boxes represent two-step clades, and solid boxes represent three-step clades. Black circles represent a hypothetical missing haplotype in the network.

Table 4. Nested contingency analysis of geographical associations for mtDNA sequence data from Tatianaerhynchites aequatus.

Clade Permutational 2 statistic Probability 1-2 66.00 0.000* 1-6 7.50 0.445 1-7 4.00 0.233 2-1 67.75 0.000* 2-2 8.63 0.061 2-3 1.88 0.387 Total 82.50 0.000*

* Significant at the 0.05 level.

tance in non-sexual species) (clade 2-2). At the lations within groups (45.87%) was higher than highest clade level we could not discriminate be- variation among groups (21.37%). Overall, 32.76% tween fragmentation and isolation by distance be- of the detected variation was within populations. cause of sampling constraints (Table 5). Significant F statistic indices (FSC=0.58, p < 0.001

AMOVA detected structuring between differ- and FST=0.67, p < 0.001) also supported structuring ent populations (Table 6). Variation among popu- of T. aequatus populations.

Population structure of Tatianaerhynchites aequatus in Eastern Europe 173

Table 5. Demographic inferences from nested clade distance analysis (Templeton et al. 1995; Templeton 2011) of mtDNA in Tatianaerhynchites aequatus.

Clade Inference chain Inferred pattern Haplotypes in 1-2 1-2-3-5-6-7-Yes Restricted Gene Flow/Dispersal but with some Long Dis- tance Dispersal Haplotypes in 1-6 1-2-11-17-No Inconclusive Outcome Haplotypes in 1-7 1-19-20-2-11-17-No Inconclusive Outcome One-step clades in 2-1 1-2-3-5-6-13-14-Yes Sampling design inadequate to discriminate between Con- tiguous Range Expansion, Long Distance Colonization, and Past Fragmentation One-step clades in 2-2 1-2-3-4-No Restricted Gene Flow with Isolation by Distance (Restricted Dispersal by Distance in Non-sexual species) One-step clades in 2-3 1-2-11-17-No Inconclusive Outcome Total 1-2-3-5-15-16-18-Yes Geographical Sampling scheme inadequate to discriminate between Fragmentation and Isolation by Distance.

Table 6. AMOVA results for tests of genetic divisions between populations of Tatianaerhynchites aequatus.

Source of variation d.f. Variance components Percent of variation F statistics** Significance

Among groups* 2 1.00 21.37 FCT=0.21 P < 0.039

Among populations within groups 10 2.15 45.87 FSC=0.58 P < 0.001

Within populations 39 1.54 32.76 FST=0.67 P < 0.001 Total 51 4.7

*Groups represent regions: 1) European Deciduous Forests, 2) Mediterranean Forests, 3) Euro-Asian Steppe.

**F statistics (FCT, FSC and FST) range from 0 to 1. A zero value implies complete panmixis (populations are freely interbreeding). A value of one implies that all genetic variation is explained by the population structure, and that populations do not share any genetic diversity.

Discussion Ceutorhynchus obstrictus (Marshan), 11 haplotypes in 176 individuals from 16 localities from North This was the first study to investigate the genetic America and Europe (Laffin et al. 2005b); Pissodes diversity and spatial structure of pest species T. strobi (Peck), 36 haplotypes in 130 individuals aequatus at a crossroads of three biogeographical from 13 localities across Canada (Laffin et al. regions (Region of European Deciduous Forests, 2004); P. yunnanensis Langor et Zang, 21 haplo- Region of Mediterranean Forests and Region of types in 60 individuals from seven localities in Euro-Asian Steppe) using mtDNA sequences of southwestern China (Zhang et al. 2007); Polydrusus cytochrome oxidase I (cox1) gene. The 0.09% to inustus Germar, 15 haplotypes in 92 individuals 2.14% sequence divergence among haplotypes of from 19 localities in Central Europe (Kajtoch et al. T. aequatus was similar to divergence among hap- 2009); Centricnemus leucogramus Germar, 41 haplo- lotypes of Pissodes strobi (Laffin et al. 2004) and P. types in 65 individuals from 13 localities in Cen- yunnanensis as it was later noticed (Zhang et al. tral Europe (Kajtoch et al. 2009); Copturus aguacatae 2007) over the same portion of mtDNA. High in- Kissinger, 14 haplotypes in 27 individuals from traspecific sequence diversity for T. aequatus spe- four localities in Mexico (Engstrand et al. 2010); cies may indicate that the mtDNA genome evolves Dendroctonus approximatus Dietz, 29 haplotypes in at a much faster rate in this species than in other 71 individuals from 15 localities in Mexico (San- groups of insects studied to date (Boyce et al. 1994, chez-Sanchez et al. 2012); Curculio elephas L., 31 Langor & Sperling 1997). haplotypes in 160 individuals from 10 localities in The number of haplotypes of T. aequatus (15 Greece (Avtzis & Cognato 2013). This suggests haplotypes in 52 individuals from 13 localities) is that T. aequatus exhibits a high degree of genetic quite high, taking into account that specimens differentiation among populations. The high ge- were sampled in a much smaller geographic area, netic diversity in T. aequatus can be explained by if compared with previous studies on curculionoid the influence of different ecological conditions species: Ceutorhynchus neglectus Blatchley, two within biogeographic regions and transition areas haplotypes in 138 individuals from 23 localities in between regions, which may enhance evolution- Canada and the United States (Laffin et al. 2005a); ary changes. The combination of high haplotype

174 N. Munteanu Molotievskiy et al. diversity, as observed in our data, can be a signa- types. Concerning haplotype K, it was sampled ture of a rapid demographic growth from a small from two localities situated at long distance one effective population size (Avise 2000). from another, has also few connections with other According to Neighbour Joining (NJ) and low-frequency haplotypes, but sampling design Maximum Parsimony (MP) phylogenetic trees has not permitted to detect haplotype K at high constructed for the T. aequatus populations; there frequency. Clade 2-3 which contains haplotypes I is a separation between two groups of haplotypes and J, connected to clade 2-1 (first subnetwork) (Fig. 3). The first comprises eleven populations through 12 mutational steps could be examined as from all three biogeographic regions, and the sec- a potential subnetwork at a deeper investigation of ond, five populations from two biogeographic re- the territory populated by T. aequatus. gions (I and II). Therefore, within the first group, it The haplotype network data suggests a high can be observed the separation of haplotypes into interconnectivity between populations in geo- smaller subclades according to regions they be- graphic vicinity. Consequently, only few popula- long. The results suggest that haplotypes are tions are isolated from each other. Nested clade grouping into a subclade that comprises haplo- distance analysis revealed Restricted Gene types from all three biogeographic regions, a sub- Flow/Dispersal but with some Long Distance clade that contains only haplotypes from Region II Dispersal in one step clade 1-2 that groups haplo- (haplotypes I and J) and a subclade that contains types from all three biogeographic regions, from only haplotypes from Region III (haplotypes D populations situated in central (Bac, Chi, Mic and and H). Seven of ten haplotypes were shared Pan), north-eastern (Erj and But) and southern among populations from first group, while a (Ior, Vol and Sta) part of the country, at some dis- lower proportion of haplotypes (two of five) were tance one from another. shared among the populations from second group. The cladogram of haplotype distribution re- This suggests that haplotypes from second group veals the relationship between population struc- are more differentiated than those from first group tures and the biogeographic regions of Moldova. of populations. Linkage between populations is due to the loca- Genetic differentiation among populations of tion of the investigated territory at the intersection T. aequatus may have pest management implica- of biogeographic regions, haplotypes formation tions if intraspecific genetic variability is an indi- being determined by the climate features and spe- cation of significant variability in biological pa- cific plant associations of the biogeographic re- rameters such as fecundity, development, and gions and districts. A better structuring of haplo- survival (Zhang et al. 2007). Although the basic types can be observed at natural districts level. biology of T. aequatus is known (Gontarenko 1978) Thus, haplotype C is characteristic to the Middle most of the research has been conducted in few lo- Dniester Forest-Steppe District (6) and haplotype cations (Poiras 1998). This biological knowledge M to Codry Forest District (7). Transition between cannot be extrapolated over the entire range of T. populations of these two districts is achieved by aequatus, thus further investigations are needed. haplotype L that is found in both districts. An- Two subnetworks were obtained from mito- other specific haplotype is N, identified for Oak chondrial sequence haplotype network (Fig. 4). A Forest ecosystems and Forest-Steppe of low hills star-like pattern in which one very common haplo- and plateaus from Romania (Florea 2004). Transi- type, presumably ancestral, lies at the centre of a tion between populations from Codry Forest Dis- network and is connected by independent muta- trict (7) (Moldova) and Oak Forest and Forest- tion steps to many rarer haplotypes, is usually re- Steppe of low hills and plateaus (Romania) is garded as indicative of a population having re- achieved through haplotype K, identified for both cently expanded in size from one or a small num- countries. Specific ecological conditions of the dis- ber of founders (Avise 2000; Mardulyn 2001). In tricts contribute to development of new haplo- present study, the largest subnetwork consisting types, although migration of individuals and gene of clade 2-1 with haplotype B in the centre and flow may still occur. second subnetwork consisting of clade 2-2 with The Codry Forest District covers over 15% of haplotype K conform to this description, subnet- the republic territory, the landscape being pre- works are linked to each other by 11 mutational dominantly natural, with rounded mountain-tops steps. Haplotype B is a high-frequency interior and ancient landslides. Forest is comprised mainly haplotype connected to lower-frequency haplo- of beech (Fagus sylvatica) and oak (Quercus petraea Population structure of Tatianaerhynchites aequatus in Eastern Europe 175 and Quercus robur), in associations with cherry Acknowledgements. The authors wish to thank the two (Cerasum avium), chequer tree (Sorbus torminalis), anonymous reviewers for their constructive comments wild pear (Pyrus pyraster) and crab apple (Malus which helped improve the paper. This work was supported by the Centre for Graduate, Post-Graduate sylvestris). More humid climatic conditions, eleva- Education and Advanced Training of the Academy of tions, topography and specific vegetation assem- Sciences of Moldova. Studies were carried out in the blages drove up the formation of three specific Laboratory of Systematic and Molecular Phylogeny, haplotypes, namely E, F and G, which are distinc- Institute of Zoology, Academy of Sciences of Moldova. tive only for Codry Forest District. Haplotype B is widespread in all three biogeographic regions, evident of regional intersection. Formation of hap- References lotype J is due to environmental variation at the boundary between the Lower Dniester Forest- Angelov, P.A. (1980): Fauna Bulgarica, 10, Coleoptera, Steppe District (11) and the Lower Dniester Steppe Curculionidae, IV Part, Calandrinae. 2. Sofia, Bulgaria. Avise, J.C. (2000): Phylogeography: the history and formation of District (12). species. Harvard University Press, Cambridge, Massachusetts, The Euro-Asian Steppe is represented by two USA. haplotypes, D and H, both distributed in the Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb, C.A., Saunders, N.C. (1987): Intraspecific Bugeac Steppe District (10). The relationship be- phylogeography: the mitochondrial DNA bridge between tween these two haplotypes is ambiguous because population genetics and systematics. Annual Review of Ecology there are two possible connections of the haplo- and Systematics 18: 489-522. Avtzis, D.N., Cognato, A.I. (2013): Genetic structure of Curculio type D to the rest of the network, each with a 95% elephas Gyll. (Coleoptera, Curculionidae) in Greece: an probability (Fig. 4). important pest of sweet chestnut. Journal of Pest Science 86: 491- Our results suggest that gene flow and hence 497. Bertheau, C., Schuler, H., Arthofer, W., Avtzis, D.N., Mayer, F., some level of dispersal occur among bio- Krumbock, S., Moodley, Y., Stauffer, C. (2013): Divergent geographic regions within state, with even higher evolutionary histories of two sympatric spruce bark beetle levels likely within districts. Using nested clade species. Molecular Ecology 22: 3318-3332. Boyce, T.M., Zwick, M.E., Aquadro, C.F. (1994): Mitochondrial analysis, we were able to highlight specific demo- DNA in the bark weevils: phylogeny and evolution in the graphic events comparing detected haplotypes Pissodes strobi species group (Coleoptera: Curculionidae). with biogeographical regions. Molecular Biology and Evolution 11: 183-194. Cazantev, M., Mucilo, M., Sirodoev, Gh., Andreev, A., Gorbunenko, AMOVA and nested contingency analysis also P. (2002): The Republic of Moldova. Ecological Network (map). showed structuring between haplotypes and spe- Publishing house “Iulian”, Chisinau, Moldova. cific populations, and detected restricted gene Clement, M., Derington, J., Posada, D. (2004): TCS: estimating gene genealogies. Version 1.18 (computer program). Brigham Young flow among populations. In AMOVA, only 21.37% University, Provo, Utah, USA. of total variation was among groups. Variation Engstrand, R., Cibrian-Tovar, J.C., Cibrian-Jaramillo, A., among populations within groups was higher and Kolokotronis, S.-O. (2010): Genetic variation in avocado stem weevils Copturus aguacatae (Coleoptera: Curculionidae) in accounted for 45.87% of total variation. This, along Mexico. Mitochondrial DNA 21 (S1): 38-43. with significant F statistics indices, suggests that Enserink, M. (1997): Life on the edge: rainforest margins may there are genetically differentiated populations spawn species. Science 276: 1791-1792. Estoup, A., Guillemaud, T. (2010): Reconstructing routes of with restricted levels of gene flow between them. invasion using genetic data: why, how and so what? Molecular The results suggest that haplotype formation Ecology 19: 4113-4130. and distribution is related to biogeographic re- Excoffier, L., Lischer, H.E.L. (2010): Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses gions. Our study revealed that haplotype B of the under Linux and Windows. Molecular Ecology Resources 10: species T. aequatus is the predominant haplotype, 564-567. being the only one spread across all three bio- Florea, N. (2004): Map of ecological units in Romania. Proceedings of the Romanian Academy, Series B, Vol. VI, Issue 1: 109-116. geographic regions in Moldova. This indicates that Fu, Y.X. (1997): Statistical tests of neutrality of mutations against this haplotype has very high adaptive potential. population growth, hitch-hiking, and background selection. Study of population structure and knowledge of Genetics 147: 915-925. Gheideman, T.S. (1966): The vegetation covers of the Moldavian different haplotypes’ adaptive potential is essen- SSR (Summary of the PhD Thesis). Chisinau. tial for developing management programs suitable Gontarenko, M. (1978): Fruit protection from apple fruit rhynchites. for protecting crops without loss of insect genetic Horticulture, Viticulture and Oenology in Moldova 1: 33-35. Hughes, J., Daily, B.G.C., Ehrlich, P.R. (1997): Population diversity: resources. its extent and extinction. Science 278: 689-692. Kajtoch, L. (2011): Conservation genetics of xerothermic beetles in Europe: the case of Centricnemus leucogrammus. Journal of Insect Conservation 15: 787-797.

176 N. Munteanu Molotievskiy et al.

Kajtoch, L., Brzyska-Nadachowska, K., Babik, W. (2012): Munteanu, N. (2006): Diversity and economic importance of leaf- Development and characterization of microsatellite loci in the rolling weevils (Coleoptera: Rhynchitidae, Attelabidae) on the Centricnemus leucogrammus weevil. Molecular Biology Reports territory of Moldova Republic. Bulletin UASMV 62: 188-191. 39: 11131-11136. Nei, M. (1987): Molecular evolutionary genetics. Columbia Kajtoch, L., Lachowska-Cierlik, D., Mazur, M. (2009): Genetic University Press, New York. diversity of the xerothermic weevils Polydrusus inustus and Poiras, A. (1998): Catalogue of the weevils (Coleoptera, Centricnemus leucogrammus (Coleoptera: Curculionidae) in Curculionoidea) and their host plants in the Republic of central Europe. European Journal of Entomology 106: 325-334. Moldova. Pensoft Publishers, Sofia-Moscow. Kehrli, P., Pasquier, D. (2013): On the track of insects responsible Posada, D., Crandall, K.A. (1998): Modeltest: Testing the model of for misshaped apple. IOBC-WPRS Bulletin 91: 173-174. DNA substitution. Bioinformatics 14: 817-818. Kim, K.S., Sappington, T.W. (2004): Boll weevil (Anthonomus grandis Posada, D., Crandall, K.A., Templeton, A.R. (2000): GeoDis: A Boheman) (Coleoptera: Curculionidae) dispersal in the southern program for the cladistic nested analysis of the geographical United States: Evidence from mitochondrial DNA variation. distribution of genetic haplotypes. Molecular Ecology 9: 487-488. Environmental Entomology 33: 457-469. Ridley, A.W., Hereward, J.P., Daglish, G.J., Raghu, S., Collins, P.J., Kimura, M. (1980): A simple method for estimating evolutionary Walter, G.H. (2011): The spatiotemporal dynamics of Tribolium rates of base substitutions through comparative studies of castaneum: adult flight and gene flow. Molecular Ecology 20: nucleotide sequences. Journal of Molecular Evolution 16: 111- 1635-1646. 120. Riedel, A., Daawia, D., Balke, M. (2010): Deep cox1 divergence and Kirk, H., Dorn, S., Mazzi, D. (2013): Molecular genetics and hyperdiversity of Trigonopterus weevils in a New Guinea genomics generate new insights into invertebrate pest invasions. mountain range (Coleoptera, Curculionidae). Zoologica Scripta Evolutionary Applications 6: 842-856. 39: 63-74. Knapen, D., Knaepkens, G., Bervoets, L., Taylor, M.I., Eens, M., Rodríguez, D.R. (2009): Protected areas of the Republic of Moldova. Verheyen, E. (2003): Conservation units based on mitochondrial An updated review. Europarc Federation: 2-28. and nuclear DNA variation among European bullhead Ruiz, E.A., Victor, J., Hayes, J., Zuniga1, G.Z. (2009): Molecular and populations (Cottus gabio L., 1758) from Flanders, Belgium. morphological analysis of Dendroctonus pseudotsugae Conservation Genetics 4: 129-140. (Coleoptera: Curculionidae: Scolytinae): An assessment of the Kryzhanovskii, O.L. (1974): Pest insects and mites of agricultural taxonomic status of subspecies. Annals of the Entomological plants. Vol 2. Leningrad-Nauka. Society of America 102: 982-997. Laffin, R.D., Dosdall, L.M., Sperling, F.A.H. (2005a): Population Sakai, A.K., Allendorf, F.W., Holt, J.S., Lodge, D.M., Molofsky, J., structure of the cabbage seedpod weevil, Ceutorhynchus With, K.A., Baughman, S., Cabin, R.J., Cohen, J.E., Ellstrand, obstrictus (Marsham) (Coleoptera: Curculionidae): Origins of N.C., McCauley, D.E., O'Neil, P., Parker, I.M., Thompson, J.N., North American introductions. Environmental Entomology 34: Weller, S.G. (2001): The population biology of invasive species. 504-510. Annual Review of Ecology, Evolution and Systematics 32: 305- Laffin, R.D., Dosdall, L.M., Sperling, F.A.H. (2005b): Population 332. structure and phylogenetic relationship of Ceutorhynchus Sanchez-Sanchez, H., Lopez-Barrera, G., Penaloza-Ramirez, J-M., neglectus (Coleoptera: Curculionidae). The Canadian Rocha-Ramirez, V., Oyama, K. (2012): Phylogeography reveals Entomologist 137: 672-684. routes of colonization of the bark beetle Dendroctonus Laffin, R.D., Langor, D.W., Sperling, F.A.H. (2004): Population approximatus Dietz in Mexico. Journal of Heredity 103 : 638-650. structure and gene flow in the white pine weevil, Pissodes strobi Savkovskiy, P.P. (1976): Atlas of the pests of fruit and berry plants. (Coleoptera, Curculionidae). Annals of the Entomological Kiev-Urozhai. Society of America 97: 949-956. Schmidt, H.A., Strimmer, K., Vingron, M., Haeseler, A. (2002): Langor, D.W., Sperling, F.A.H. (1995): Mitochondrial DNA TREE-PUZZLE: maximum likelihood phylogenetic analysis variation and indication of bark weevils in the Pissodes strobi using quartets and parallel computing. Bioinformatics 18: 502- species group in western Canada (Coleoptera, Curculionidae). 504. The Canadian Entomologist 127: 895-911. Schneider, C.J., Smith, T.B., Larison, B., Moritz, C. (1999): A test of Langor, D.W., Sperling, F.A.H. (1997): Mitochondrial DNA alternative models of diversification in tropical rainforests: sequence divergence in weevils of the Pissodes strobi complex ecological gradients vs. rainforest refugia. Proceedings of the (Coleoptera, Curculionidae). Insect Molecular Biology 6: 255- National Academy of Sciences of the United States of America 265. 96: 13869-13873. Legalov, A.A. (2003): Taxonomy, classification and phylogeny of Sequeira, A.S., Stepien, C.C., Sijapati, M., Roque Albelo, L. (2012): the leaf-rolling weevils (Coleoptera: Rhynchitidae, Attelabidae) Comparative genetic structure and demographic history in of the world fauna. Novosibirsk Press. endemic Galapagos weevils. Journal of Heredity 103: 206-220. Librado, P., Rozas, J. (2009): DnaSP v5: A software for Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., Flook, P. comprehensive analysis of DNA polymorphism data. (1994): Evolution, weighting, and phylogenetic utility of Bioinformatics 25: 1451-1452. mitochondrial gene-sequences and a compilation of conserved Luck, G.W., Daily, G.C., Ehrlich, P.R. (2003): Population diversity polymerase chain-reaction primers. Annals of the Entomological and ecosystem services. Trends in Ecology and Evolution 18: Society of America 87: 651-701. 331-336. Smith, T.B., Wayne, R.K., German, D.J., Bruford, M.W. (1997): A Mardulyn, P. (2001): Phylogeography of the Vosges mountains role for ecotones in generating rainforest biodiversity. Science populations of Gonioctena pallida (Coleoptera: Chrysomelidae): a 276: 1855-1857. nested clade analysis of mitochondrial DNA haplotypes. Spector, S. (2002): Biogeographic crossroads as priority areas for Molecular Ecology 10:1751-1763. biodiversity conservation. Conservation Biology 16: 1480-1487. Mayr, E. (1963): Animal species and evolution. Belknap Press, Spiecker, H. (2003): Silvicultural management in maintaining Cambridge, Massachusetts. biodiversity and resistance of forests in Europe-temperate zone. Medvedev, S.I., Shapiro, D.S. (1957): To the knowledge of beetle Journal of Environmental Management 67: 55-65. fauna (Coleoptera) of the Moldavian Soviet Socialist Republic Tajima, F. (1983): Evolutionary relationship of DNA sequences in and the neighbouring regions of Ukraine. Scientific Works 30: finite populations. Genetics 105: 437-460. 173-206. Tamura, K., Dudley, J., Nei, M., Kumar, S. (2007): MEGA4: Moritz, C. (1994): Defining “Evolutionarily significant units” for Molecular Evolutionary Genetics Analysis (MEGA) software conservation. Trends in Ecology and Evolution 9: 373-375. version 4.0. Molecular Biology and Evolution 24: 1596-1599. Population structure of Tatianaerhynchites aequatus in Eastern Europe 177

Templeton, A.R., Routman, E., Phillips, C.A. (1995): Separating diseases, pests and weeds in the MSSR. CAI Scientific Works population structure from population history: a cladistic Proceeding 31-36. analysis of the geographic distribution of mitochondrial DNA Vitcovskii, N. (1913): A brief overview of the main pests and haplotypes in the tiger salamander, Ambystoma tigrinum. diseases of wild and cultivated plants, observed in 1912 in the Genetics 140: 767-782. province of Bessarabia. Proceedings of the Bessarabian Templeton, A.R. (2011): Inference key for the nested haplotype tree Naturalists Society and Life Science Amateurs 4: 263-277. analysis of geographical distances. http://darwin.uvigo.es/ Yang, Z. (1993): Maximum-Likelihood estimation of phylogeny download/geodisKey_06Jan11.pdf from DNA sequences when substitution rates differ over sites. Ter-Minasyan, M.E. (1950): Roller-weevils (Attelabidae). Fauna of Molecular Biology and Evolution 10: 1396-1401. the USSR. Coleoptera 27 (2), Moscow-Leningrad. Zayanchkauskas, P.A. (1984): Insects injuring orchards in Thompson, J.D., Higgins, D.G., Gibson, T.J. (1994): CLUSTALW: Lithuanian SSR. New and rare species of insects from Improving the sensitivity of progressive multiple sequence Lithuanian SSR. Vilnius. alignment through sequence weighting, positions-specific gap Zhang, H., Langor, D.W., Ye, H., Li, Z., Laffin, R.D. (2007): Genetic penalties and weight matrix choice. Nucleic Acids Research 22: divergence among populations of Pissodes yunnanensis 4673-4680. (Coleoptera: Curculionidae) in southwestern China. The Veresceaghin, B.V. (1922): Report on the activity of the Bio- Canadian Entomologist 139: 308-318. Entomological Station for ten years (1911-1921). Chisinau. Veresciaghin, B.V. (1964): Some features of harmful entomofauna of tree plantations in Moldova. Journal of Zoology 43: 185-192. Vinnichenko, I. (1988): Management methods against weevils in intensive apple orchards. Agricultural crops protection against