Introduction

Aphids (Hemiptera: Aphididae) are among the most successful families of insects, and many are serious agricultural and forest pests (FAO 1991; Sorensen 2003; van Edhem and Harrington 2007). The cypress aphid, Cinara cupressi (Buckton), was first described as Lachnus cupressi in England (Buckton 1881), and is a native species of Europe and the Near East (from eastern Greece to Iran). This aphid, which is considered to be one of the hundred most important invasive pests in the world (Watson et al. 1999; FAO 2009; Montalva et al. 2010a), is now present in Southwest Asia, North America, Africa and South America (Blackman and Eastop 1994; Ciesla 2011). In Chile, the aphid was first detected in the Tarapacá Region during 2003 on trees of Cupressus sp., and the occurrence of this pest was studied between 2005 and 2007 (INFOR 2008). It has been recorded on 23 different tree species in Chile, where the most important concern for their ecological and biodiversity value are the native tree species Austrocedrus chilensis (D. Don.) Pic. Serm. & Biz and Fitzroya cupressoides (Mol.) Johnston, which are in ecologically vulnerable and endangered categories (Baldini and Alvarado 2008; Montalva et al. 2010a). To mitigate the damage from this pest in Chile, chemical control or releases of the parasitoid Pauesia juniperorum Stary (Hymenoptera: Braconidae) have been used but both of these approaches yielded erratic results (Baldini et al. 2008).

During a survey of natural enemies on C. cupressi between 2008 and 2012 in the South of Chile, Syrphidae larvae and Coccinellidae were found feeding on C. cupressi and in the southern city of Puerto Natales. Hemerobiidae was found regulating the population of this aphid in some colonies. Two entomopathogenic fungi were also found in the surveys on Cinara species: Lecanicillium sp. and Neozygites osornensis Montalva & Barta (Montalva et al. 2013a, b). Worldwide, mainly parasitoids associated with Cinara sp. have been described as well as some species of fungi and insect predators. However, the most used control agents under field conditions are parasitoids, most commonly species of Pauesia (Montalva et al. 2010a). The study of such natural enemies of the cypress aphid as entomopathogenic fungi is highly important in Chile, since applications of chemical insecticides to control Cinara species associated with the native tree A. chilensis are forbidden in national parks without the permission of the national health authorities (CONAF 2006).

In general, aphids feed on phloem sap via extremely fine maxillary stylets that penetrate phloem sieve tubes. This mode of feeding greatly reduces the possibility that aphids will ingest any viruses, bacteria, or protozoa present on plant surfaces. Fungi are the most important microbial pathogens of aphids because these pathogens can infect hosts through the integument (Barta and Cagáň 2006; Steinkraus 2006). More than 750 species of fungi are pathogens of arthropods, and many of them offer great potential for managing insect pests such as aphids (Lacey et al. 2015). Different species of entomopathogenic fungi from the phyla Ascomycota and Entomophthoromycota are associated with aphids (Humber 1991). Entomopathogenic ascomycetes include hundreds of species, but only a few of these are specific for aphids. Lecanicillium lecanii (Zimmerman) Zare & W. Gams (Hypocreales: Clavicipitaceae) is one of the most important hypocrealean of aphids (Humber 1991).

The use of entomopathogenic fungi to control pests has low environmental impact in comparison to chemical insecticides. Moreover the fungi are both target-specific and often economically feasible to use (Shah and Pell 2003; Rodríguez and Arredondo 2007; Ansari et al. 2011). Laboratory bioassays with Lecanicillium attenuatum Zare & W. Gams, L. lecanii, and L. longisporum (Petch) Zare & W. Gams have shown high efficacy for controlling different aphid species in many parts of the world (Cortez 2007; Kim 2007; Vu et al. 2007; Kim et al. 2008; Kim and Roberts 2012). L. longisporum and L. muscarium (Petch) Zare & W. Gams have been formulated and commercialized in Europe as the mycoinsecticides Vertalec® and Mycotal®, respectively, to control aphids and whiteflies in crops (Diaz et al. 2009). The objective of the present work was to identify three fungal isolates of Lecanicillium obtained from naturally infected C. cupressi in southern Chile, and to determine their pathogenicity in comparison to Lecanicillium isolates from commercial mycoinsecticides against C. cupressi nymphs under laboratory conditions.

Materials and methods

Origin and preparation of the fungi

During 2007–2013, C. cupressi populations were observed for entomopathogenic fungi in 25 localities of southern Chile (36°49′41″–53°10′00″S, 70°56′00″-73°03′04″W) (Montalva et al. 2010b, 2013a, b). Three Lecanicillium strains were isolated from naturally infected C. cupressi cadavers in different localities (Table 1) and were studied further in the bioassays presented here.

Table 1 Source of fungal isolates used in the characterization and pathogenicity test against nymphs of Cinara cupressi

The isolates are preserved at the Laboratory of Servicio Agrícola y Ganadero from Osorno, Chile, with the method of water stasis, as described by López-Lastra et al. (2002). To preserve these fungi in water stasis, sporulating cultures grown on potato dextrose agar (PDA; Difco®, Becton, Dickinson & Company, USA) were cut into small blocks (approximately 3 mm2) and placed into sterile cryotubes (40 × 10 mm with screw caps) containing 2 ml of sterile deionized water that were then kept at 4 °C. Cultures were recovered from the storage by aseptically transferring a block of inoculum from the water onto a fresh PDA plate. The isolates were rejuvenated once from the material preserved by passage through a host before the bioassays (Montalva et al. 2010b, 2014b).

These isolates are also preserved in the USDA-ARS Collection of Entomopathogenic Fungal Cultures (ARSEF; Ithaca, NY, USA) as L-1, L-2, and L-3 (accessioned as ARSEF 13278, ARSEF 13279, and ARSEF 13280, respectively). Two other Lecanicillium strains, L. longisporum ARSEF 5126 and L. muscarium ARSEF 5128, used for the formulation of the mycoinsecticides Vertalec® (Koppert B.V., The Netherlands) and Mycotal® (Koppert B.V., The Netherlands), respectively, were included in the pathogenicity bioassays as reference strains (Table 1). The Chilean isolates were compared morphologically (size and shape of the main structures) and physiologically (rate of growth and fungal sporulation) with the ex-neotype isolate of Lecanicillium lecanii CBS 102067 provided by CBS-KNAW Fungal Biodiversity Centre (Utrecht, Netherlands) on in vitro material (Table 1) and with other aphid-pathogenic Lecanicillium species noted by Zare and Gams (2001).

Morphological and physiological evaluation of the fungi

Conidial inocula from the isolates were grown at 24 ± 1 °C and in darkness on PDA for ten days (Zare and Gams 2001). Semi-permanent slide mounts were prepared in lactophenol-cotton blue according to Humber (2012). Fungal microstructures were examined with an optical microscope (Nikon Eclipse E600), measured using Motic Images Plus 2.0 at 1500 magnifications, and documented with a digital camera (Nikon DS-Fi1). Measurements were based on 200 objects per microstructure from which mean values and SE were calculated. To evaluate the radial growth of the isolates, small blocks from seven-day-old cultures (about 15 mm3) were placed individually in the center of a new Petri dish (90 mm diam) containing PDA, which was marked with two perpendicular axes. The radial growth of culture was measured for both axes with ruler daily during ten days (24 ± 1 °C and in darkness) and for each day a mean for the radial growth obtained from the axes was calculated. The conidial production of each fungal isolate was calculated on day 21. Conidia were scraped from the culture surface with a spatula, suspended in 10 ml sterile 0.1% Tween 80®. The suspension was vortexed for 3 min, and then filtered through hydrophobic cotton. The numbers of conidia in the suspension were quantified with a Neubauer improved haemocytometer. The single temperature used for radial growth and conidial production was chosen for ready comparison with other specific studies with these species in order to complement the taxonomic identification. Four independent replicates and four repetitions were done for each isolate.

Molecular characterization of the fungi

The six fungal isolates were grown on PDA for 21 days at 24 ± 1 °C, 75 ± 5% RH in darkness. The protocol by Montalva et al. (2014a) was used for DNA extraction. DNA pellet was suspended in 50 µl of buffer TE pH 8.0. The DNA was quantified (Nanodrop ND-1000, DE, USA) and maintained at −80 °C. PCR reactions were done with universal fungal primer pairs NS1/NS2 (NS1: 5′-GTAGTCATATGCTTGTCTC-3′ and NS2: 5′-GGCTGCTGGCACCAGACTTGC-3′; White et al. 1990) and Bt1a/Bt1b (Bt1a: 5′-TTCCCCCGTCTCCACTTCTTCATG-3′, Bt1b: 5′-ACGAGATCGTTCATGTTGAACTC-3′; Glass and Donaldson 1995), that amplify the small subunit (SSU) rRNA (rDNA) and β-tubulin genes, respectively. PCR reactions with the primer pairs NS1/NS2 and Bt1a/Bt1b were performed in a volume of 25 µl per sample containing 10 µl of EconoTaq™ PLUS GREEN 2X Master Mix (Lucigen, Middleton, WI, USA), 0.4 µM of each primer, and 25 ng µl−1 of DNA. The thermal profile of the PCR (MultiGene™ Gradient Thermal Cycler, Labnet International Inc., Woodbridge, NJ, USA) condition for primer pair NS1/NS2 included an initial denaturation at 94 °C for 3 min, followed by 35 cycles with a new denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 40 s, and with a final extension of 10 min at 72 °C. The PCR protocol for the primer pair Bt1a/B-t1b was similar to the above except that the annealing was at 60 °C for 30 s, and the extension at 72 °C for 30 s. PCR products were run in 1.2% (w/v) agarose gels with 0.04 µl ml−1 of GelRed™ nucleid acid stain (Biotium Inc., Hayward, CA, USA) in a TBE buffer at 75 V for 60 min. The PCR amplicons were observed and photographed in a UV light transilluminator. The DNA amplicons for both genes were purified and directly sequenced in both directions (Macrogen, Seoul, South Korea). An analysis of sequence similarities was done with the Basic Local Alignment Search Tool (BLAST, NCBI, USA) and compared online for sequence homologies (T-coffee) with other sequences in the European Bioinformatics Institute database (EBI, Cambridge, UK). For the phylogenetic relationships between Lecanicillium sp. isolated from C. cupressi in Chile (ARSEF 13278, ARSEF 13279 and ARSEF 13280) and with Lecanicillium species reported on other hosts, a multiple and concatenated alignment of partial sequences of SSU rDNA and β-tubulin genes was performed with MUSCLE (Edgar 2004), implemented in Seaview 4.0 software (Gouy et al. 2010). Sequences of CBS 102067, ARSEF 5126, ARSEF 5128, ARSEF 13278, ARSEF 13279 and ARSEF 13280 for SSU rDNA and β-tubulin genes were deposited in the GenBank database (National Centre for Biotechnology Information, Bethesda, Maryland, USA) under accession numbers KX818212, KX806649, KX818213, KX806650, KX818214, KX806651, KX818215, KX806652, KX818216, KX806653, KX818217 and KX806654, respectively. A phylogenetic tree was constructed (Seaview 4.0) by the maximum-likelihood (ML) method using the GTR model in PhyML 3.0 (Guindon and Gascuel 2003). Branch support on ML tree was calculated from 1000 bootstrap replicates. DNA sequences of SSU rDNA and β-tubulin genes from L. lecanii CBS 102067 (ex-neotype isolate), L. longisporum ARSEF 5126, L. muscarium ARSEF 5128 were used in this study, as well as DNA sequences from GenBank such as for L. attenuatum (GenBank accessions AF339614 and AH015608 obtained from the ex-neotype isolate of this fungus, CBS 170.76) and L. muscarium (GenBank Accessions EF641854 and EU000248). Verticillium dahliae Klebahn (GenBank accessions U33637 and DQ266153) was used as outgroup in the phylogenetic analysis.

Origin and rearing of Cinara cupressi

The laboratory colony of C. cupressi originated from a single individual collected from Puerto Natales, Chile (51°43′39.39″S; 72°29′45.54″W), and was reared on Cupressus macrocarpa Hartw. plants (30 cm high) in a growth chamber at 18 ± 2 °C with a 16:8 (L:D) photoperiod, which ensured the parthenogenetic reproduction of aphids. Hence, each individual produced a genetically identical colony or clonal lineage. Before experiments, aphids were left to reproduce on their optimal host-plants for at least three generations, and to prevent any maternal effect or grand-maternal effects, a consequence of the telescopic reproduction in aphids (Cabrera-Brandt et al. 2010; Saranya et al. 2010).

Pathogenicity bioassay

Conidia of isolates were obtained from 15-day-old cultures grown on wheat grains (100 g) previously autoclaved at 121 °C for 15 min in a 500 ml Erlenmeyer flask, then incubated at 24 °C ± 1 °C and a 12 h photoperiod. After 15 days, 100 ml of 0.1% (v/v) sterile aqueous solution of Tween® 80 was added to the Erlenmeyer flask and shaken vigorously to obtain a conidial suspension. The conidial suspensions were filtered through a double layer of muslin to remove mycelial mats and wheat residues. The numbers of conidia in the suspensions were quantified with a hemocytometer and adjusted with sterile 0.1% (v/v) Tween® 80 to concentrations of 105, 106, 107 and 108 conidia ml−1 for the assays as a preliminary experiment. At the beginning of each assay, the viability of conidia was verified by spreading 50 µl of a conidial suspension (105 conidia ml−1) on PDA. Germination of 100 conidia was quantified from four separate areas on each plate after 24 h incubation at 24 ± 1 °C and 12 h photoperiod. Conidia were considered to have germinated when the length of the germ tube exceeded the conidial diameter. In all cases, more than 95% of conidia were viable.

The pathogenicity of the isolates was tested against third-instar nymphs (N3) of C. cupressi. Ten N3 were placed on a disinfected twig (8 cm long) of C. macrocarpa (washed with a 1% NaOCl solution for 5 min, rinsed three times with sterile distilled water for 2 min, and dried for 10 min before being placed in a Petri dish) inside a Petri dish (9 cm diam) containing a thin layer of 2% (w/v) water agar. The twigs in Petri dishes were sprayed with 0.5 ml of four concentrations of conidial suspensions (105, 106, 107 or 108 conidia ml−1) of each Lecanicillium isolate by a small calibrated atomizer bottle, applied directly for 3 s and at 10 cm of distance to set-ups where nymphs were exposed with four replicates. In the controls, ten nymphs (N3) were sprayed with 0.5 ml of autoclaved 0.1% Tween® 80 (v/v). The treated nymphs were incubated at 24 ± 1 °C, nearly saturated RH (>98%), with a 16:8 (L:D) photoperiod for up to seven days. All treatments were observed at 24 h intervals to record mortality. Aphids with an unusually darker body color and absence of movement were scored as dead (Lacey and Solter 2012). Aphid cadavers were removed from the Petri dish (5 cm diam) with a dampened filter paper and maintained at 24 °C ± 1 °C for 48 h to induce fungal sporulation (Diaz et al. 2009).

Data analysis

All tests were carried out with four independent repetitions. Radial growth, conidial production, size of fungal microstructure and relative values of mortality were analyzed with an ANOVA to determine differences among isolates and the Student–Newman–Keuls multiple range test was performed to separate and compare means when significant (P < 0.05) differences were detected. Lethal times (LT50 and LT90) and lethal concentrations (LC50 and LC90) to kill 50 and 90% of test populations with their respective 95% confidence intervals were calculated by probit analysis with the Wolfram Mathematica and Probit software, respectively (Throne et al. 1995).

Results

Morphological fungal identification

The Chilean entomopathogenic isolates formed white colonies with a deep yellow reverse on PDA. The conidia were short-ellipsoidal to subcylindrical and usually very homogeneous in the shape (Fig. 1a). The phialides were short, strongly tapering, and present either singly or in characteristic whorls of three individual phialides (Fig. 1b). During autumn, sporulation on infected aphids was observed in some localities with an appearance typical for infections caused by the fungal genus Lecanicillium.

Fig. 1
figure 1

Lecanicillium species isolated from Cinara cupressi in southern Chile. a Conidia from pure cultures on PDA. b Conidiophores and conidia from pure cultures on PDA. Bars 20 µm

Table 2 presents the morphometric and physiological characteristics of the Lecanicillium species isolated in Chile. The conidial measurements of the Chilean isolates were larger than those of CBS 102067, L. attenuatum, L. longisporum and were more similar to the average size of L. muscarium conidia (Zare and Gams 2001). The Chilean isolates produce conidia of a remarkably consistent size whereas those of L. muscarium are routinely found to be highly variable in both size and shape (Table 2). Further, the radial growth of CBS 102067 was significantly lower (F3,12 = 35.9, P < 0.001) at 1.8 ± 0.06 cm day−1 than all other isolates being monitored. Isolate ARSEF 13280 presented the highest (F3,12 = 35.9; P < 0.001) radial growth at 3.33 ± 0.15 cm day−1 among all the isolates studied. The highest value for conidial production (F3,16 = 23.0, P < 0.001) at day 10 was observed for ARSEF 13278 with 9.4 × 107 conidia ml−1 (Table 2).

Table 2 Morphometric analysis of taxonomically important structures and physiological characteristics of Lecanicillium isolates from Chile, ex-neotype isolate of Lecanicillium lecanii (CBS 102067) and other close related aphid-pathogenic Lecanicillium species described by Zare and Gams (2001)

Molecular characterization

Partial DNA sequences of SSU rDNA (537 bp) and β-tubulin (531 bp) genes from L. lecanii CBS 102067, L. longisporum ARSEF 5126, L. muscarium ARSEF 5128, and Chilean Lecanicillium isolates ARSEF 13278, ARSEF 13279, and ARSEF 13280 were deposited in the GenBank database (NCBI). Partial DNA sequences of SSU rDNA gene from Chilean isolates ARSEF 13278, ARSEF 13279, and ARSEF 13280 in BLAST analysis (NCBI) showed 100% homology with L. attenuatum and 98-99% with other Lecanicillium species and from other genera in the family Cordycipitaceae. Similarly, BLAST analysis using partial DNA sequences of β-tubulin gene from ARSEF 13278, ARSEF 13279 and ARSEF 13280 showed that these strains were 100% identical with L. attenuatum and showed 99% identity with strains of L. muscarium. Pairwise comparisons (T-coffee) on 537 bp lengths of aligned sequences of SSU rDNA gene showed that Lecanicillium isolates ARSEF 13278, ARSEF 13279, and ARSEF 13280 shared 100% identity among themselves and 100% with L. attenuatum (GenBank AF339614) and L. muscarium ARSEF 5128.

The identity of ARSEF 13278, ARSEF 13279 and ARSEF 13280 with L. lecanii CBS 102067 and L. longisporum ARSEF 5126 was 99%, respectively. On the other hand, pairwise comparisons on 531 bp from sequences of β-tubulin gene showed that ARSEF 13278, ARSEF 13279, and ARSEF 13280 shared 100% identity among themselves and 100% with L. attenuatum (Accession Nº AH015608). The identity of Chilean Lecanicillium isolates with L. lecanii CBS 102067, L. longisporum ARSEF 5126, and L. muscarium ARSEF 5128 was 95, 93 and 99%, respectively.

The phylogenetic tree constructed by the maximum likelihood method using concatenated partial SSU rDNA and β-tubulin sequences of Chilean Lecanicillium isolates ARSEF 13278, ARSEF 13279 and ARSEF 13280 and the L. attenuatum (GenBank AF339614) shared the same clade (Fig. 2), indicating that Chilean isolates are more similar to this species than to other Lecanicillium spp. used in this work. In fact, L. muscarium ARSEF 5128, L. longisporum ARSEF 5126, and L. lecanii CBS 102067 (the ex-neotype culture) were placed in distinct clades (Fig. 2), thus confirming that the Chilean isolates can be identified as L. attenuatum.

Fig. 2
figure 2

Phylogenetic tree constructed by the maximum likelihood method using concatenated SSU rDNA and β-tubulin sequences of Chilean Lecanicillium isolates ARSEF 13278, ARSEF 13279 and ARSEF 13280. Other DNA sequences of Lecanicillium species used in this study as reference material were also included in the analysis. Numbers indicate bootstrap values based on 1000 replicates. The scale bar represents the number of expected substitutions per site

Pathogenicity bioassay

Our results indicated that third-instar individuals of the cypress aphid were sensitive to isolates of the fungi tested in laboratory. In general, the mortality of test insects increased with time post-inoculation and with conidial concentration. The mortality of nymphs treated by specific conidial concentrations varied significantly (F4,15 = 11.2, P < 0.001) among isolates. After a four-day incubation, the cumulative mortality of nymphs treated with 1 × 106 conidia ml−1 varied from 4.3% (ARSEF 5126) to 62.5% (ARSEF 13280), and seven days after treatment all exposed nymphs were killed by the action of the fungi ARSEF 13278, ARSEF 13279 and ARSEF 13280 (Fig. 3). The mean mortality of the control nymphs on the fourth day reached 7.5% and was lower than 18% after seven days. There was a significant effect of the isolate on cumulative mortality after four and seven days of incubation (F4,15 = 6.9, P < 0.002; F4,15 = 32.1, P < 0.001; respectively).

Fig. 3
figure 3

Relative cumulative mean mortality ± SE of N3 Cinara cupressi exposed to 1 × 106 conidia ml−1 of ARSEF 13278, 13279, 13280, 5126 and 5128 for up to seven days

The lowest values of lethal times LT50 (≤3.7 days) and LT90 (≤five days), at conidial concentration of 106 conidia ml−1, were obtained with ARSEF 13279. The highest values of both LT50 (4.3 days) and LT90 (7.2 days) were detected for ARSEF 5128. Because ARSEF 5126 killed fewer than 50% of population it was not possible to calculate LT50 or LT90 at 106 conidia ml−1 (Table 3). LC50 values of the isolates (except ARSEF 5126) four days after application varied from 3 × 105 conidia ml−1 (ARSEF 13279) to 106 conidia ml−1 (ARSEF 13278) and LC90 values of the isolates (except for ARSEF 5126), four days after application varied from 7.8 × 106 conidia ml−1 (ARSEF 13278) to 2.9 × 108 conidia ml−1 (ARSEF 5128; Table 3).

Table 3 Lethal time (days) to kill 50% or 90% (LT50 and LT90) with the slope ± SE and lethal concentrations (conidia ml−1) to kill 50% or 90% (LC50 and LC90) of test population with 95% confidence interval (CI) of Cinara cupressi third-instar nymphs treated with the Lecanicillium isolates and incubated at 24 ± 2 °C for up to seven days

Discussion

Based on the comparison of the morphological data obtained from the Chilean Lecanicillium species with the Lecanicillium species studied by Zare and Gams (2001) with the ex-neotype isolate for L. lecanii we can say confidently that the Chilean isolates belong to the Lecanicillium complex (L. attenuatum, L. longisporum or L. muscarium), but are definitively not L. lecanii. Within the current systematics of hypocrealean entomopathogens, it is recognized that the species can be and often are defined only by their genomic characters, and that morphological characters can no longer be regarded as adequate to identify species within this extremely large and taxonomically complex group of fungi. Nonetheless, differences in arrangement morphology and size of conidiogenous cells as well as in homogeneity or variability of conidia sizes within the L. lecanii species complex (which should now be treated as including L. attenuatum) do seem to be significant and highly correlated with the genomically verified identifications of fungi in this complex.

Molecular tools based on SSU rDNA gene have been used before to differentiate between morphologically similar Lecanicillium species (Zare and Gams 2008; Lu et al. 2015). However, the pairwise comparison based on SSU rDNA gene, indicated that the Chilean isolates (ARSEF 13278, ARSEF 13279 and ARSEF 13280) could not be distinguished from L. attenuatum (GenBank AF339614) or L. muscarium ARSEF 5128.

The SSU rRNA gene is considered to be a conserved gene, which varies with constant rate, and that can show relatively highly variable regions that provide important information for distinguishing fungal species (Rodriguez et al. 2004). However, this gene might not allow unambiguous separations of related species within the genus Lecanicillium, since some ambiguity can be observed in the analysis of the partial SSU rDNA (Zare and Gams 2008; Lu et al. 2015). In this case it is recommended to use another gene and DNA sequences of β-tubulin gene would be successful in clarifying ambiguity between related Lecanicillium species (Kope and Leal 2006; Leal et al. 2008; Lu et al. 2015). In this study, the β-tubulin sequences unambiguously separated L. attenuatum from L. muscarium and placed the Chilean isolates in L. attenuatum. Furthermore, the phylogenetic tree also confirmed that the Chilean Lecanicillium isolates are L. attenuatum (Fig. 2).

It is important to remember that genome-based identifications of fungi can be treated as definitive only by demonstrating an overwhelming similarity to sequences from the type specimen or from ex-type culture (regardless of the type of type specimen from which it was derived). Comparisons and identifications based on any sources that are not types or type-derived may be misleading or incorrect (Humber 2016).

In the laboratory, the Chilean isolates ARSEF 13278 and 13279 were the ones that produced the greatest numbers of conidia. This represents important data to consider for future production for inundative sprays and we successfully demonstrated that third-instar nymphs of C. cupressi are susceptible to the Chilean fungal isolates we tested. However the virulence varied among the isolates. Kim et al. (2001) evaluated the virulence of the entomopathogenic fungus L. lecanii against Aphis gossypii Glover and reported 100% mortality with a LT50 of 2.7 days. In the present work the lowest value of 3.7 days was obtained from ARSEF 13279 (Table 3). Vu et al. (2007) showed that the application of L. lecanii at a concentration of 1 × 107 conidia ml−1 to Myzus persicae (Sulzer) resulted in an LT50 of 2.1 days, while the application of this entomopathogen to A. gossypii resulted in an LT50 of 1.5 days. The difference in the values of LT50 compared to those obtained in the present study could be explained by the different aphid species and by the conidial suspensions being evaluated. In general, the most virulent fungal isolates often tend to be those isolated from the host (Papierok et al. 1984). This was the case in our work, where the most virulent isolates we tested were obtained from C. cupressi. From our preliminary results, we suggest that the Chilean isolates might be considered for use in the biological control of C. cupressi with an inundative strategy of large-scale sprays of conidia onto the target trees and their pests. However further studies are needed.

Furthermore, advantage of using the Chilean isolates against N3 of C. cupressi (living and feeding in colonies together with the adults) would result in the entire population being susceptible to a fungal infection as noted by Jandricic et al. (2014) and Shrestha et al. (2015) since in Chile C. cupressi reproduces parthenogenetically and forms aphid colonies of genetically identical individuals. Therefore, all the aphids in the colony may be relatively equally susceptible to a single pathogen.

In conclusion, these native Chilean L. attenuatum isolates proved to be pathogenic against N3 of C. cupressi under laboratory conditions and may very well have significant potential as biocontrol agents against this pest. Further laboratory, formulations and field studies will be needed to confirm the probable value of L. attenuatum as an alternative agent for the control of cypress aphids.