Evaluation of Indigenous Entomopathogenic Nematodes as Potential Biocontrol Agents against Popillia japonica (Coleoptera: Scarabaeidae) in Northern Italy

Simple Summary The Japanese beetle Popillia japonica is considered one of the most harmful organisms in the world for crops and the urban landscape. Popillia japonica spends most of its life cycle in the soil as a larva. At this stage, this beetle is more susceptible to biological control agents like entomopathogenic nematodes, which are obligate parasites of main soil-inhabiting insects, killing their host in just a few days. In 2014, P. japonica was detected in Northern Italy between the Piedmont and Lombardy regions. This research aims to investigate the natural occurrence of indigenous and locally adapted entomopathogenic nematodes along the Piedmont part of the Ticino river and test the most performing of them via bioassays. Natural isolates were recovered from 39 out of the 155 soil samples collected. The virulence of all entomopathogenic nematodes assessed by laboratory and semi-field assays highlighted that two isolates resulted in more efficiency in controlling P. japonica grubs. This result is very encouraging, and the use of these natural biocontrol agents against this pest is a fundamental component of eco-friendly management. Abstract The natural presence of entomopathogenic nematodes (EPNs) has been investigated in the Piedmont region (Northern Italy) in areas infested by the Japanese beetle Popillia japonica. Thirty-nine out of 155 soil samples (25.2%) were positive for EPNs. Most of the samples contained only steinermatids (92.3%), 5.1% contained heterorhabditids, and one sample (2.6%) contained both genera. All the recovered isolates were identified at species level both morphologically and molecularly. Steinernema carpocapsae was the most abundant and it was mainly distributed in open habitats, such as perennial meadows, uncultivated soils, and cropland, characterized by sandy loam soil texture and acidic pH. Steinernema feltiae has been found associated mainly with closed habitats such as coniferous and deciduous woodland, characterized by sandy loam-texture and extremely acidic soil. The three isolates of Heterorhabditis bacteriophora were collected only in open habitats (perennial meadows and uncultivated fields) characterized by strongly acidic soils with sandy loam texture. The virulence of all EPN natural strains was evaluated by laboratory assays against P. japonica third-instar larvae collected during two different periods of the year (spring, autumn). The results showed that larval mortality was higher for pre-wintering larvae than post-wintering ones. The five more promising EPN isolates were tested in the semi-field assay in which H. bacteriophora natural strains have been shown to be more efficient in controlling P. japonica grubs. All of these results are finally discussed considering the use of these natural EPNs as biological control agents against P. japonica, within an eco-friendly perspective of management.

The identification of particularly aggressive natural EPN strains would contrast this phytosanitary problem by proposing the use of these natural enemies probably performing better in those specific edaphic conditions.

Field Sampling and Isolation of Nematodes
During the spring of 2017, 155 soil samples were collected from different habitats (meadows, croplands, uncultivated fields, and woodlands) never treated, in previous years, with biocontrol agents. All sampling areas were distributed within an area of approximately 116 km 2 along the western side of the Ticino river, in Novara province, in the Piedmont region (Northern Italy) ( Figure 1).
Insects 2020, 11, x 3 of 16 The identification of particularly aggressive natural EPN strains would contrast this phytosanitary problem by proposing the use of these natural enemies probably performing better in those specific edaphic conditions.

Field Sampling and Isolation of Nematodes
During the spring of 2017, 155 soil samples were collected from different habitats (meadows, croplands, uncultivated fields, and woodlands) never treated, in previous years, with biocontrol agents. All sampling areas were distributed within an area of approximately 116 km 2 along the western side of the Ticino river, in Novara province, in the Piedmont region (Northern Italy) ( Figure  1). Each soil sample weighed about 1 kg and resulted from the mixture of five subsamples of about 200 cm 3 dug with the shovel within an area of 5 × 5 m. All samples were georeferenced in WGS84 (World Geodetic System 1984), labeled as POP 1-155, and before being placed in the plastic bags, Popillia japonica larvae were removed from the soil samples and counted. Samples were transferred to the CREA-(Research Centre of Plant Protection and Certification) laboratory of Florence to be processed. Once in the lab, soil samples were placed into plastic containers covered with a perforated lid, and EPNs were recovered by the insect bait method [20]. For each sample, five last-instar Galleria mellonella L. (Lepidoptera: Pyralidae) larvae were placed onto the soil and incubated at 23 °C for one week. Soil samples were stirred every day to foster the contact of the bait larvae with the entomopathogenic nematodes present in the sampled soil. Mortality was checked every two days and dead specimens showing signs of EPNs were rinsed in sterile distilled water and placed Each soil sample weighed about 1 kg and resulted from the mixture of five subsamples of about 200 cm 3 dug with the shovel within an area of 5 × 5 m. All samples were georeferenced in WGS84 (World Geodetic System 1984), labeled as POP 1-155, and before being placed in the plastic bags, Popillia japonica larvae were removed from the soil samples and counted. Samples were transferred to the CREA-(Research Centre of Plant Protection and Certification) laboratory of Florence to be processed. Once in the lab, soil samples were placed into plastic containers covered with a perforated lid, and EPNs were recovered by the insect bait method [20]. For each sample, five last-instar Galleria mellonella L. (Lepidoptera: Pyralidae) larvae were placed onto the soil and incubated at 23 • C for one week. Soil samples were stirred every day to foster the contact of the bait larvae with the entomopathogenic nematodes present in the sampled soil. Mortality was checked every two days and dead specimens showing signs of EPNs were rinsed in sterile distilled water and placed individually into modified white traps [21] to collect the emerged infective juveniles (IJs). Emerging nematodes were used to infect five G. mellonella separate larvae (100 IJs/larva) in Petri dishes (diam. 3.5 cm) to verify pathogenicity, complete Koch's postulates, and establish the culture.
Nematodes capable of infesting G. mellonella larvae for the second time were subsequently collected, stored in single flasks in the darkroom, and incubated at 12 • C. Each flask containing EPNs acquired the same label as the soil from which the contained nematode was isolated (see above). The pH of the soil samples with EPNs was analyzed and correlated with site location, habitat, and soil type, as extracted from Brenna et al. [22] and IPLA [23]

Nematode Identification
Both molecular and morphologic examinations were used for EPN species identification. For morphobiometrical studies, IJs and first-generation males of each isolate were heat-killed in warm water at 65 • C. Then, they were fixed in triethanolamine formalin (TAF) solution (2 cc triethanolamine, 7 cc of 40% commercial formaldehyde solution, and 91 cc distilled water), processed to glycerin by a modified glycerine-ethanol series of the Seinhorst rapid method, and then permanently mounted in anhydrous glycerin on microscope glass slides. Specimens were observed under a light microscope equipped with a Leica MC 170 HD digital camera (Leica Microsystems, Heerbrugg, Switzerland) and the specimens were measured using LEICA Application Suite (LAS) Version 4.9.0 (Leica Microsystems, Heerbrugg, Switzerland).
The features measured for the IJs (n = 10 for each EPN) were body length (L), distance from the anterior end to excretory pore (EP), distance from the anterior end to the base of the pharynx (ES), and tail length (T). Moreover, the following ratios were calculated: a (body length/maximum body diameter), b (body length/distance from the anterior end to the base of the pharynx), c (body length/tail length), D% (distance from the anterior end to the excretory pore/distance from the anterior end to the base of the pharynx), and E% (distance from the anterior end to excretory pore/tail length). The IJs' body length of each species was matched with nematodes that have a similar length in the polytomous key of Nguyen [24], and then other specific characteristics were used to identify the species. Finally, the spicule shape of the males was compared with the original description to confirm the identification.
For the molecular analysis, three specimens of nematodes were collected for each isolate and individually put in a 0.2 mL tube containing 50.0 µL InstaGene Matrix (Bio-Rad, Hercules, CA, USA), 1.5% Sodium Dodecyl Sulfate (SDS), and were frozen at −80 • C overnight. Afterwards, samples were rapidly thawed at 55.0 • C, added to with 2.5 µL Proteinase K (Qiagen, Hilden, Germany) 20.0 µg/µL, and incubated at the same temperature for 3 h. Proteinase K was inactivated by heating at 96.0 • C for 10 min and DNA was recovered by alcoholic precipitation adding 100.0 µL of cold absolute ethanol for each sample. Pellets were air-dried and re-suspended in 20.0 µL of double-distilled water. Amplifications of the internal transcribed spacer locus (ITS) were performed using primers TW81 (5 -GTTTCCGTAGGTGAACCTGC-3 ) and AB28 (5 -ATATGCTTAAGTTCAGCGGGT-3 ) following a touchdown PCR protocol on 2720 Thermal Cycler (Applied Biosystem, Foster City, CA, USA). Briefly, 5.0 µL of DNA was used in a 50.0 µL amplification reaction containing 1X DreamTaq Hot Start PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) and 0.6 mM of each primer. The thermal protocol adopted was as follows: an initial denaturation/hot start activation step for 3 min at 95 • C followed by 10 cycles of denaturation at 94 • C for 60 s, primers annealing for 60 s from 50 to 45 • C, decrease of −0.5 • C/cycle, elongation at 72 • C for 1.5 min, followed by 35 traditional amplification cycles with denaturation at 94 • C for 60 s, primers annealing for 60 s at 45 • C, elongation at 72 • C for 1.5 min, followed by a final elongation step of 7 min at 72 • C. The PCR products were sequenced at the Centro di Servizi per le Biotecnologie di Interesse Agrario Chimico ed Industriale (CIBIACI), University of Florence, Italy. Sequences were submitted in GenBank. Species attribution was defined with a BLAST similarity search. To solve ambiguities in species attribution inside the Steinernema genus, sequences belonging to this species were further characterized by performing a maximum likelihood (ML) phylogenetical reconstruction based on the ITS locus. The ML tree was computed starting from an alignment 1156 positions long, choosing GTR + G + I as the nucleotide substitution model, and validated with 1000 bootstrap pseudo-replicates.

Laboratory Virulence Assays
The experiment was repeated twice, using Popillia japonica larvae collected from infested soil of the Piedmont region in May (post-wintering 3rd-instar larvae) and September 2018 (pre-wintering 3rd-instar larvae) following results obtained in Paoli et al. [25]. The grubs were transferred to the CREA-DC laboratory in Florence (Italy) and maintained at 4 • C in native soil. Each test was carried out within 10 days after larval collection. Larvae were acclimated individually in Petri dishes at 20 • C for 4 days before nematode inoculation, to evaluate the absence of preventive infestation by natural EPNs.
Each EPN strain of POP (see above) collected from the positive soil samples (see Results section) was reared at 24 • C in greater wax moth G. mellonella late-instar larvae, and the infective juveniles (IJs) were recovered using modified white traps. After storage at 12 • C for a maximum of 2 weeks, they were kept at 20 • C for 24 h before the tests.
Each POP isolate was tested against post-wintering 3rd-instar larvae (n = 6) and pre-wintering 3rd-instar larvae of P. japonica (n = 6). The experimental unit consisted of a Petri dish (3.5 cm diameter) with two layers of filter paper (Whatman No. 1) and one larva. A distilled water suspension of 250 µL containing about 300 IJs (about 31 IJs/cm 2 ) was inoculated into each Petri dish. Concerning the mixed isolates, the proportions of the two EPNs were as follows: 45% S. carpocapsae/55% S. feltiae for POP 138 and 61% H. bacteriophora/39% S. carpocapsae for POP 80. In the control (n = 6), only distilled water was added. Petri dishes were stored at 23 • C in the dark. Larval mortality was checked daily for 14 days. Dead larvae were transferred individually to modified white traps to determine the percentage of emergence (percentage of cadavers from which the progeny emerged) and the time of emergence (i.e., number of days from the death of P. japonica larvae to the moment when IJs started to emerge from the infested host cadaver).

Semi-Field Virulence Assay
Five of the most performing EPN strains (POP 139: Steinernema carpocapsae, POP 48: S. feltiae, POP 16: Heterorhabditis bacteriophora, POP 80: mix of S. carpocapsae and H. bacteriophora, and POP 138: mix of S. carpocapsae and S. feltiae) were tested against P. japonica 3rd instars in semi-field experiments. Based on the data collected in September 2018, the criteria used for the selection of the EPNs were, in order of importance, percentage of mortality, and median lethal times to 50% (LT50). Larvae of P.·japonica were collected from the infested cornfield (pre-wintering 3rd instars) and maintained at 4 • C in native soil at the CREA-DC laboratory in Florence (Italy) until the time of the assay.
Larvae were acclimated individually in plastic cups, containing about 30 g of sterile soil and ryegrass seeds. Cups were incubated at 20 • C for 4 days before nematode inoculation. The selected EPN strains were reared and stored as above.
The experimental unit (n = 12 for each EPN) consisted of a plastic box (14.5 × 14.5 cm) with 400 g of sterile soil with about 13% humidity, determined by the gravimetric water content method [26], mixed with 2 g of ryegrass seeds and 4 P. japonica larvae at a 2 cm depth. A distilled water suspension of 1 mL containing about 5000 IJs (about 24 IJs/cm 2 ) was inoculated into every box. Concerning the mixed isolates, the proportions of the two EPNs were as follows: 21% S. carpocapsae/79% S. feltiae for POP 138 and 87% H. bacteriophora/13% S. carpocapsae for POP 80. Only distilled water was added into the control (n = 12 boxes with 4 larvae each). Plastic boxes were stored at 23 • C in darkness for one week. Larval mortality was evaluated at the end of the experiment. Dead larvae were placed individually on modified white traps and observed daily to record the IJs emergence. Infective juveniles were collected and counted after 10 days to determine the progeny, which represents the reproductive potential.

Statistical Analysis
Data were first checked for normality and homogeneity of variance using Shapiro and Levene tests, respectively. When these assumptions were not met, non-parametric tests were used.
The habitat preference between the two Steinernema species was compared via contingency table analysis and the χ 2 test. The preference between S. carpocapsae and S. feltiae isolates for the soil pH of the sampling sites was compared using the Welch t-test.
In laboratory assays, lethal times to 50% (LT50) were estimated by linear interpolation according to Marcus and Eaves [27]. The non-parametric Mann-Whitney test was used to compare the numbers of dead P. japonica larvae in May and September, and the pre-wintering larval mortality caused by the two Steinernema species, the LT50 between the two Steinernema species, and their emergence from cadavers. A t-test was used to compare the day of IJs emergence between S. carpocapsae and S. feltiae isolates. Larval mortality and progeny of the semi-field assay were analyzed using the Kruskal-Wallis test, followed by pairwise comparisons using the Wilcoxon rank-sum test with Benjamini and Hochberg's correction.
Data were analyzed using R version 3.6.3 [28].

Field Sampling and Isolation of Nematodes
Soil samples were collected in 155 selected areas trying to maintain a balance of representativeness between the two main habitats: woodlands and meadows with 77 samples and 72 samples, respectively. Six more areas were sampled in other habitats poorly represented in the ecosystem.
Most of the positive samples were recovered from perennial meadows and woodlands with a percentage of 51.3% and 38.5%, respectively. In particular, H. bacteriophora was present only in strongly acidic soils (pH 5.01-5.16) with sandy loam texture, collected only in open habitats (perennial meadows and uncultivated fields). Steinernema carpocapsae was mainly distributed in sites with extremely acidic-moderately acidic soil (pH 4. 16-5.65) and sandy loam soil texture. It was more frequently recorded in open habitats, such as perennial meadows (78.9%), uncultivated soils with wild weeds and shrubs (10.5%), and in one cropland with alfalfa (5.3%). Only one S. carpocapsae strain (POP 139) was isolated from a woodland composed mainly of elm and black locust.
Steinernema feltiae was found associated with soils with sandy loam texture, pH ranging from extremely acidic to moderately acidic (3.71-5.78), and mainly closed habitat such as coniferous and deciduous woodland (81.3%). Most S. feltiae isolates (62.5%) were extracted from soils sampled in lowland mixed woods mainly composed of oak, black locust, hornbeam, hazel, and chestnut. One positive sample (POP 102) was isolated from soil collected from Norway spruce woodland and two S. feltiae strains were extracted from mixed coniferous/deciduous forests, in particular Scots pine/English oak (POP 91) and Scots pine/chestnut (POP 127). Instead, 18% of S. feltiae strains were isolated from perennial meadows.
Concerning the habitat preferences between the two Steinernema, S. carpocapsae was mainly related to open habitats, while S. feltiae to closed ones (χ 2 = 17.85; df = 1; p < 0.0001). Further, the pH values of the habitats play a decisive role in the distribution of these two species: S. feltiae was found in more acidic soils (4.50 ± 0.17) in comparison to the number of S. carpocapsae (5.25 ± 0.06) (F = 18.658; df = 1, 18.322; p = 0.0003981).

Nematode Identification
Morphological and molecular analysis revealed that the EPN isolates matched with three described species: 21 S. carpocapsae, 17 S. feltiae, and 3 H. bacteriophora. Thirty-seven samples contained one strain, while two samples were composed of two EPNs associated with the following: POP 80 with H. bacteriophora and S. carpocapsae, and POP 138 with S. carpocapsae and S. feltiae. The morphological and morphometric data of nematodes were generally in agreement with the respective original description (Table S1).
Molecular identification based on the ITS sequence was performed on 117 individual nematodes coming from 39 POP isolates (three specimens/isolate). The amplification with TW81-AB28 primer pairs generated (at least) an amplicon in 37 out of the 39 isolations (94.9-96.0% successful amplifications). Comparison between the 96 obtained sequences led to identifying only three unique sequences. The BLAST homology search assigned the three sequences to as many different species: H. bacteriophora (MW226884), S. carpocapse (MW226882), and S. feltiae (MW226883).
Even if the analysis of large numbers of samples usually highlights the characteristic hypervariability of the ITS locus, the absence of nucleotide polymorphisms in the three species could be due to a very limited geographic provenance of the samples.
The ML tree based on the ITS locus clearly distinguished each Steinernema species, attributing them to a monophyletic origin. Moreover, the tree was able to univocally differentiate S. carpocapse from S. feltiae with a high bootstrap value ( Figure S1).
There was a significant difference among the mortality caused by the five different EPNs and the control (Kruskal-Wallis χ 2 = 52.596; df = 5; p < 0.0001) (Figure 2).  In particular, the isolates with H. bacteriophora, POP 80 and POP 16, caused the highest mortality, at 97.9% and 91.7%, respectively. Meanwhile, the isolates with S. carpocapsae and S. feltiae alone and mixed with each other killed a similar number of P. japonica larvae (54.2% POP 138 and POP 139, 39.6% POP 48) (Table 3). Regarding progeny, there was a significant difference between the number of IJs produced by the five different EPNs (Kruskal-Wallis χ 2 = 51.57; df = 4; p < 0.0001). In particular, the H. bacteriophora POP 16 reproduced more than the others, while no statistical differences were detected among POP 48, POP 80, POP 138, and POP 139 (Figure 3).

Discussion
In this study, the susceptibility of Popillia japonica larvae to indigenous EPN isolates, recovered from soil collected in the infested areas, was evaluated. The recovery frequency of 25.2% is higher than that reported in previous surveys on EPNs carried out in Italy, e.g., 5% [29], 6.5% [30], 14% [31],

Discussion
In this study, the susceptibility of Popillia japonica larvae to indigenous EPN isolates, recovered from soil collected in the infested areas, was evaluated. The recovery frequency of 25.2% is higher than that reported in previous surveys on EPNs carried out in Italy, e.g., 5% [29], 6.5% [30], 14% [31], and 6% or 15.5% from pinewood or holm-oak wood habitats, respectively [32]. The EPNs' occurrence in soil samples is highly variable in different surveys, with a recovery frequency ranging from less than 1% to more than 50% [33][34][35]. Our recovery frequency was similar to that reported in the Catalonia region of Spain (23.3%) by Garcia del Pino and Palomo [36] and in the USA (California, 26.3%, and Oregon, 23.7%) [37,38].
Steinernematids are usually recovered more often than heterorhabditids during non-targeted surveys [39]. This proportion is confirmed by our research, as well as by Tarasco et al. [30] in a previous study in Italy; however, it differs from the first survey on the biodiversity of EPNs in Italy, carried out in 1988 in the Emilia-Romagna region, where heterorhabditids formed 68% of the entomopathogenic nematodes [40].
Moreover, two different EPN species were simultaneously extracted with the Galleria mellonella bait method from single soil samples. Several field surveys reported that steinernematid species often occur sympatrically (e.g., Stuart and Gaugler [43]; Campbell et al. [44]; Půža and Mráček [45]), and the co-occurrence of S. carpocapsae and S. feltiae had already been found by Garcia del Pino and Palomo [36]. On the contrary, concomitant steinernematid and heterorhabditid occurrences are less frequently reported [46,47]. Some laboratory assays showed the inter-specific interactions between different EPN species and their capability to produce mixed progeny inside the same cadaver [48][49][50][51]. Lewis et al. [52] asserted that species of Steinernema can penetrate and reproduce in the same insect host, though one species may suffer the effects of the competition more than the other one. This work confirmed this aspect of the two species of Steinernema, and also the co-existence of S. carpocapsae and H. bacteriophora. The proportion of progeny that emerged from Galleria larvae in the Kock's postulates assay was similar in both cases (POP 138: 45% S. carpocapsae/55% S. feltiae and POP 80: 61% H. bacteriophora/39% S. carpocapsae); however, after storage and the multiplication of EPNs in G. mellonella larvae to obtain fresh progeny for the semi-field assay, the presence of S. carpocapsae IJs decreased in both cases, at 21% for POP 138 and 13% for POP 80. Moreover, these percentages reached very low values in the progeny that emerged from P. japonica larvae that died in the semi-field assay (3.7% in POP 80; 8.8% in POP 138).
The EPN occurrence is influenced by the soil characteristics, especially by the texture [53]. Most studies report that EPNs are more prevalent in soils with a high sand content, which promotes the mobility and survival of nematodes [37,54,55]. Indeed, in this study, 97.4% of the sampled soils had a sandy loam texture, typical of the areas along the Ticino river, and only one S. carpocapsae isolate (POP 44) was extracted from an area characterized by silty soil texture.
Another important soil parameter that influenced the EPN occurrence is the pH [38]. It is commonly reported that EPNs were found in soils with a wide spectrum of pH, from acidic (pH 4) to alkaline (pH 8). Mwaniki et al. [56] suggested that steinernematids are more suited to pH < 6 and heterorhabditids to pH > 6. Since all sampled soils in the study area were characterized by low pH [22,23], this can explain the low occurrence of Heterorhabditis in our sampling area where only three H. bacteriophora isolates were collected from soils with a strongly acidic pH. All the Steinernema isolates were extracted from soils with a pH < 6, and four S. feltiae isolates (POP 101, POP 127, POP 152, and POP 153) from the soil with a pH < 4. Only a few works reported EPNs found in soil with a pH < 4 [57,58].
In this study, different EPN species showed a distinct habitat preference. Open habitats, such as perennial meadows, cropland, and uncultivated fields, were suitable for S. carpocapsae, compared to S. feltiae which has been found mainly in closed habitats (woodland). The literature on habitat preference for EPNs is various and contradictory. In some cases, soil samples collected from agricultural lands had more EPNs than those from natural habitats [36,[59][60][61]. However, opposing results have also been found with EPNs more prevalent in natural habitats than agricultural ones [62][63][64]. In this survey, EPNs were found in both habitats, even if perennial meadows, generally considered agricultural habitats, are not subject to intense management which can alter the edaphic conditions through soil tillage, exposing EPNs to unfavorable abiotic and biotic conditions. Akhurst and Bedding [65] suggested that the abundance of suitable insect hosts seems to be crucial for EPN occurrence and distribution. Therefore, both forests and open habitats can host different EPN species, and they are generally more frequent in insect-rich habitats. In this survey, perennial meadows and uncultivated fields presented a high number of P. japonica larvae (45 ± 14 larvae/m 2 ). In woodlands, however, no larvae of this pest were found, but it is known that coniferous and deciduous forests are rich in insects which pupate in the soil, creating an ideal environment for the persistence of EPNs [66,67].
The laboratory assays carried out with EPN strains isolated from our study area showed that there was a significant difference in the mortality of larvae collected in the two different periods. Paoli et al. [25] reported that P. japonica 3rd-instar larvae were more susceptible to H. bacteriophora in the pre-wintering than in the post-wintering period. This work not only confirmed the different susceptibility of P. japonica larvae to H. bacteriophora, but also to steinernematids.
The 39 isolates were all highly virulent to G. mellonella larvae, producing 83.3-100% larval mortality in 24-72 h after 31 JIs/cm 2 exposition (data not shown), but high intraspecific variability in the percentage mortality of P. japonica larvae was shown for both steinernematids recovered. These differences among strains of S. carpocapsae and S. feltiae had also been previously reported by Simões et al. [68] and Tarasco [69], against G. mellonella. The isolates containing H. bacteriophora alone or mixed with steinernematids were the most virulent ones and confirmed the results shown by Marianelli et al. [13].

Conclusions
This study of the occurrence and the pathogenicity of different EPN strains in this geographical area infested by P. japonica indicated that soils in the Ticino valley are rich in EPNs. Moreover, the nematodes isolated during this survey could be used in biological control programs against this dangerous alien pest, since native isolates of EPNs possess physiological traits that are adapted to local ecological conditions.
The idea of supporting the ecosystem by amplifying the entomopathogenic fauna that thrives in its native soil could be considered the basis for an eco-friendly approach against alien pests.