Entomopathogenic nematodes in agricultural areas in Brazil

Entomopathogenic nematodes (EPNs) (Steinernematidae and Heterorhabditidae) can control pests due to the mutualistic association with bacteria that kill the host by septicemia and make the environment favorable for EPNs development and reproduction. The diversity of EPNs in Brazilian soils requires further study. The identification of EPNs, adapted to environmental and climatic conditions of cultivated areas is important for sustainable pest suppression in integrated management programs in agricultural areas of Brazil. The objective was to identify EPNs isolated from agricultural soils with annual, fruit and forest crops in Brazil. Soil samples were collected and stored in 250 ml glass vials. The nematodes were isolated from these samples with live bait traps ([Galleria mellonella L. (Lepidoptera: Pyralidae) larvae]. Infective juveniles were collected with White traps and identified by DNA barcoding procedures by sequencing the D2/D3 expansion of the 28S rDNA region by PCR. EPNs identified in agricultural areas in Brazil were Heterorhabditis amazonensis, Metarhabditis rainai, Oscheios tipulae and Steinernema rarum. These species should be considered pest biocontrol agents in Brazilian agricultural areas.

Entomopathogenic nematodes (EPNs) Steinernematidae and Heterorhabditidae can control pests due to mutualistic association with bacteria of the genus Xenorhabdus (Thomas & Poinar) and Photorhabdus (Boemare, Louis & Kuhl), respectively 1,2 . These nematodes penetrate the host through natural openings or through the cuticle transporting bacteria into the hemocele 3,4 where they reproduce and kill the host from septicemia within 24 to 48 hours 5,6 , making the environment favorable for nematode development and reproduction 7 . Infective juveniles seek another host in the soil when the insect host resources run out 8 . Interest in these biological control agents is increasing 9 due to the reduced efficiency of conventional chemical and cultural methods for insect soil management and the broad spectrum of EPN hosts 10 .
EPNs are globally distributed, with different species and groups according to geographic regions 11,12 . Information on EPNs and their symbiotic bacteria is scarce in many countries, including in Brazil. Heterorhabditis amazonensis Andaló and Steinernema brasiliensis Nguyen were reported as native species in Brazil 13,14 and Heterorhabditids indica and H. baujardi were reported in Rondônia state, Brazil 15 . Isolates of H. amazonensis, H. baujardi, H. indica and H. mexicana were found in Minas Gerais state, Brazil 16,17 .
EPN identification, adapted to environmental and climatic conditions of cultivated areas is important for sustainable pest suppression in integrated management programs in agricultural areas of Brazil. The objective was to identify EPNs from agricultural soils with annual, fruit and perennial crops in Brazil. A zero to 25 cm deep soil sample was taken per sampling point and placed in 2L labeled plastic bags, stored in a Styrofoam box and transferred to the laboratory. The geographical coordinates of each sample were obtained with Garmin GPS device Etrex Vista H 2.8. Nematodes were isolated in the laboratory using fifth instar Galleria mellonella Linnaeus (Lepidoptera: Pyralidae) larvae. Briefly, each soil sample was packed into a 250 mL a glass vial with five G. mellonella larvae. These vials were covered and stored without light at 25 ± 2 °C. After three to seven days, dead G. mellonella showing nematode infection symptoms were removed, rinsed in distilled water and transferred to White traps 34 . The infective juveniles (IJs) were again inoculated in G. mellonella larvae for multiplication. Five G. mellonella larvae were used in a Petri dish (9 cm diameter) containing two moistened paper filters with 1.5 mL of solution with 100 JIs/larvae, for each nematode species. The samples were covered with PVC plastic and stored at 25 ± 2 °C and RH > 80% 35 . After three days, dead larvae were transferred to a White trap 34 at 25 ± 2 °C for seven and 15 days. The IJs that left the G. mellonella carcasses were collected with distilled water every two days and stored at 18 °C. The nematode samples were named FCA 01 to 14 and stored in the Entomopathogenic Nematode Bank of the Nematology Laboratory at FCA/UNESP. Vials contained 1 M NaCl solution and were frozen at − 80 °C until DNA extraction and EPN identification.

PCR.
Genomic DNA was obtained for each population (FCA 01 to 14) from three individuals of each EPN, extracted using Lysis Buffer Holterman [(HLB) (800 μ g proteinase K/ml, β -mercaptoethanol 1% (v/v), 0.2 M NaCl and 0.2 M Tris HCl pH 8)] 36 . A total of 25 μ L of HLB was diluted in 25 μ L of ultrapure water totaling 50 μ L in a 0.2 mL Eppendorff tube. A drop of this solution (5 μ L) was placed on a glass slide, where the nematodes were individually cut into three parts and placed in the same 0.2 mL tube. The 45 μ L of remaining solution was used to wash the slide and added to the tube with the sectioned nematode. Samples were submitted to PCR at 65 °C for 2 h, 99 °C for five minutes and stored at − 20 °C 37 . The universal primers D2A (5′ -CAAGTACCGTGAGGGAAAGTTG-3′ ) and D3B (5′ TCGGAAGGAACCAGCTACT A-3′ ) were used to amplify the D2/D3 expansion segment of 28S rDNA by PCR 38 . A total of 12.5 μ L of Gotaq Hot Start (Promega, São Paulo State, Brazil), with the reagents necessary for reaction: 5 U/μ L Taq, 100 mM of each NTP and 25 mM MgCl 2 , 9.5 μ L of nuclease free water (Promega), and 1 μ L of each primer [10 mM] and 1 μ L of cDNA from each representative population of target and non-target species, totaling 25 μ L per reaction was submitted to pCR at 94 °C for seven minutes; followed by 35 cycles at 94 °C for 60 seconds, 55 °C for 60 seconds, 72 °C for 60 seconds; and 72 °C for 10 minutes 39 . Five μ L of PCR product was used for electrophoresis in TAE buffer 40 on 1% agarose gel, stained with ethidium bromide (0.02 mg/mL), visualized and photographed under UV light. The result of the amplification was compared to the molecular weight marker VIII.
The amplified fragments of D2/D3 expansion 28S rDNA were sequenced with the Big Dye Terminator kit (Applied Biosystems) 41 . A reagent mix containing 2 μ L Big Dye, 3.2 mmol sense primers, 3.0 μ L of amplified product containing 400 ng DNA and 2.0 mL of water was prepared for the product end of the PCR reaction. The reaction for sequencing was carried out according to manufacturer's instructions (Applied Biosystems) with further purification of the amplified product by precipitation with isopropanol. Samples were denatured at 95 °C for three minutes and electrophoresis performed in an ABI Prism 377 DNA Sequencer unit (Applied Biosystems).
The sequences were aligned and compared to nucleotide polymorphism identification with the aid of BioEdit Aligment Sequence Editor Program. The EPN population sequences were compared with other nematode species in the database (GenBank, http://www.ncbi.nlm.nih.gov) for identification based on genetic similarity.
For phylogenetic analysis, the multiple alignments between the sequences of the region D2/D3 of the different isolates were edited manually with BioEdit Sequence Aligment program when phylogenetically uninformative columns were excluded from the analyses. Phylogenetic analyses were inferred using the Maximum Likelihood method based on the Kimura 2-parameter model 42 , considered the best fitting model for sequence evolution determined using the BIC scores (Bayesian Information Criterion) implemented in 6 MEGA program 42 . Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. A discrete Gamma distribution was used to model evolutionary rate differences among sites (4 categories (+ G, parameter = 2.3432)). The model variation rate allowed for some sites to be evolutionarily invariable ([+ I], 19.1558% sites). The tree was drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 48 nucleotide sequences, including 16 obtained in the present study (FCA01 to FCA16) and 32 from the GenBank. All positions containing gaps and missing data were eliminated. A total of 292 positions was obtained in the final dataset. Trees were sampled at intervals of 1000 generations and Caenorhabditis elegans (Brenner) was selected as outgroup. Evolutionary analyses were conducted in 6 MEGA program 42 .

Results
The nematodes obtained from White traps were inoculated in new G. mellonela larvae, which demonstrates parasitism by the isolated entomopathogen. EPNs were found in 16 soil samples, corresponding to 8% of 201 samples. Seven (35%) of 20 samples from forest plantation areas had nematodes (FCA 04, FCA 05, FCA 06, FCA 07, FCA 08, FCA 10 and FCA 15). In annual crops, three (8.1%) out of 37 samples had nematodes: isolated FCA 11, detected in sandy soil in irrigated rice in Botucatu, São Paulo State, Brazil and isolates FCA 16 and FCA 03 in soybean crops in clay soils in Palotina, Paraná State, Brazil. Among 97 samples taken from orchards, six (6.2%) were positive for EPNs, FCA 12 isolates grown in sandy soils with wild raspberry, FCA 13 in citrus soil and FCA 14 in mango soil in São Manuel, São Paulo, Brazil. FCA 01 and FCA 02 isolates were also found in clay soil with citrus in Botucatu, São Paulo State, Brazil (Fig. 1). EPN samples were not found in plowed soil, native forest and pasture areas.
Amplification of D2/D3 expansion 28S rDNA gene of EPN isolates produced 590 bp fragments, whose sequences were deposited in the GenBank under the accession codes KRO11843 and KRO11858 (Fig. 2) (AF331905). However, these isolates formed two groups that had three polymorphisms between them. One group with FCA 09, FCA 11 and FCA 12 isolates were similar to each other, but different in 3 bp from the group comprising FCA 10, FCA 13 and FCA 14 isolates. The phylogenetic tree (Fig. 2) (Table 1).

Discussion
Nematology surveys with G. mellonella baiting technique are useful to detect Steinernematidae and Heterorhabditidae species as well as other rhabditids. The molecular technique used was adequate to identify nematode isolates, enabling knowledge of its biodiversity and contributing to the detection of new isolates that may be used in biological control programs of insect pests.
The sequence of D2/D3 expansion 28S rDNA gene analysis by DNA barcode technique was useful for the diagnosis of H. amazonensis, S. rarum, M. rainai and O. tipulae. The phylogenetic tree obtained from the 48 aligned sequences of the expansion D2/D3 EPNs with four distinct groups support the molecular identification of these nematodes isolated from soil samples. This technique has been used to diagnose plant and animal parasites and entomopathogenic nematodes with accurate and reliable results, such as one Pratylenchus penetrans (Cobb) specimen in potato 41 , Bursaphelenchus fungivorus (Franklin & Hooper) in coconut fiber 43 , M. rainai (Carta & Osbrink) in soil cultivated with soybean 44 and Metarhabditis blumi (Sudhaus) parasitizing the ear canal of cattle 45 .
The finding of H. amazonensis and S. rarum in seven (3.5%) of the 201 soil samples, shows the reduced occurrence of these organisms in Brazil compared to surveys in Western Canada (20%) 46 , Argentina (13.2%) 47 and Spain (23.3%) 48 . However, the prevalence of 3.5%, of these species, is similar to surveys in Turkey (2%, 9.1%) 25,49 , Azores Archipelago, Portugal (3.9%) 19 and Minas Gerais state, Brazil (9%) 29 . The absence of EPNs in plowed soil areas suggests inadequate conditions for nematode survival, but zero EPN detection in the natural forest was unexpected and may be due the low soil samples collected in this area. This result may also indicate the need for a higher number of samples taken at different soil depths. The species habitat and soil type affected EPNs recovery 25,50 . Our samples with nematodes were obtained from, clay (75%) and sandy (25%) soils, indicating the mobility and survival of EPNs in soils rich in sand, but S. rarum occurred in clay (FCA 09 and FCA 10) and sandy (FCA,11,12,13 and 14) soils. Many EPNs positive samples (89.65%) were obtained in acid soils in Nepal 51 . Six EPNs were found in soils with PH < 4, which is uncommon, but this has also been reported in Belgium 52 . Steinernema spp. are widespread in different regions and niches, whereas Heterorhabditis spp. are common in forest and river banks 53 . Heterorhabditis and Steinernema species were found in Sandy soils with pH < 6, whereas representatives of Heterorhabditis, mainly in Sandy soils with pH > 6 were found on nine islands of Açores 19 .
Steinernema rarum detection in sandy soil with Rubus edaeus, Mangifera indica and Oryza sativa crops and clay soils with Citrus sp. and Eucalyptus sp. demonstrates the habitat diversity of this nematode, which has been detected in Olea europea L. crops in Argentina 54 , Carya illinoinensis orchards (Wangenh.) in Mississippi and Louisiana State, USA 28 and in sandy soils with annual and forest cultures in China 55 . In Brazil, this nematode was detected in soils with Nicotiana tabacum L., Triticum aestivum L., Glycine max and in native vegetation 56 . Steinernema rarum showed desiccation tolerance and freezing intolerance 57 , features that may be related to its adaptation to Brazilian soils. This species has shown promise in biological control programs against Antonomus grandis Boheman (Coleoptera: Curculionidae) and Spodoptera frugiperda J.E. Smith (Lepidoptera: Noctuidae) 27 . Metarhabditis rainai detection in fruit orchards, annual crops and forest areas shows the broad habitat diversity of this nematode. This species was reported in Glycine max crop in Araras, São Paulo State, Brazil 44 . Other species, such as Metarhabditis blumi, are considered entomopathogenic and associated with the bacteria Alcaligenes faecalis Castellani & Chalmers, Flavobacterium sp. (Bernardet & Grimont) and Providencia vermicola (Somvanshi), which killed G. mellonella larvae 58 . Other genera, besides Heterorhabditis and Steinernema may feed on and kill insects 44 . Rhabditidae nematodes are usually bacteriophages, found in insect carrions 59 . However, M. rainai was described from specimens isolated from the Oscheios tipulae termite gut 60 detected in soil samples and may parasitize insects and was also found to be associated with Tipula paludosa Meigen (Diptera: Tipulidae) larvae 61 . This nematode has been frequently isolated in soil samples from around the world, enabling species population 62 . Oscheius genus (Rhabditidae) includes free-living, vertebrate and invertebrate parasitic species 63 , such as Oscheios chongmingensis (Zhang), considered a facultative entomopathogenic nematode 64 . The bacteria symbiosis with this genus species is similar to that of Steinernematidae and Heterorhabditidae 65 .
EPNs identified in agricultural areas in Brazil were Heterorhabditis amazonensis, Metarhabditis rainai, Oscheios tipulae and Steinernema rarum. These species should be considered a pest biocontrol agent.