Susceptibility of Plutella xylostella (Lepidoptera: Plutellidae; Linnaeus 1758) to Beauveria bassiana Bb9205, Metarhizium anisopliae Ma9236 and Heterorhabditis bacteriophora HNI0100

The diamondback moth ( Plutella xylostella ) is a major pest of broccoli worldwide. It mainly causes leaf defoliation and generates annual losses of 80%. In this study we evaluated the susceptibility of P. xylostella to entomopathogens Heterorhabditis bacteriophora HNI0100, Beauveria bassiana Bb9205 and Metarhizium anisopliae Ma9236. The methodology was based on the inoculation of third instar larvae of P. xylostella with 5x10 1 , 1x10 2 , 3x10 2 , 6x10 2 and 1,2x10 3 IJs/cm 2 of H. bacteriophora HNI0100 and evaluated them after 24, 48 and 72 h and 1x10 4 , 1x10 5 , 1x10 6 , 1x10 7 and 1x10 8 con/cm 2 of B. bassiana Bb9205 and M. anisopliae Ma9236, which were evaluated during two weeks. At a dose of 1,2 x10 3 JIs/cm 2 , P. xylostella had a susceptibility to H. bacteriophora HNI0100 of 91,66%. Similarly, B. bassiana Bb9205 and M. anisopliae Ma9236 had a mortality of 95,33 and 99,67% at 1x10 5 con/cm 2 . The results suggest that the use of strains of entomopathogenic nematodes and fungi is an innovative alternative for the control of P. xylostella . However, studies on the interaction of nematodes and fungi and Plutella xylostella are necessary.

The diamondback moth has a four-stage life cycle (egg, larva, pupa and adult) of 15 to 40 days depending on regional weather conditions; the entire life cycle takes place on the leaves abaxial surface of the host plant (Sarfraz et al. 2005).The egg-stage lasts 3.2 days at 20 ºC.During the 15 days of its larval instars (four instars), the pest causes the greatest damage to broccoli crops (Somvanshi & Ganguly 2007).White spots on the leaves evidence surperficial mines that alter the photosynthesis process, and cause a decrease in the size and quality of the product intended for consumption (Franco 2001, Schroer et al. 2005, Chavez & Hurtado 2010).Eighteen days before the adult hatches, the pupa is covered with a silk thread on the abaxial surface of the leaf (Talekar & Shelton 1993, Chavez & Hurtado 2010).Adults are nocturnal; males emerge seeking a partner for copulation and are found on the leaves during the day.The average adult female lays 160 to 360 eggs (Furlong et al. 2013).
This insect pest is controlled through the continued use of agrochemicals.A 99-day crop can be treated three to five times with insecticides such as organophosphates, pyrethroids and carbamates (Sarfraz et al. 2005).The indiscriminate use of these products by farmers has led to excessive application of the same and an increase in resistance issues with this lepidopteran pest (Monzón 2001, Sáenz 2012).For this reason, other alternatives of control have been evaluated such as physical barriers, as light and resistant covers to prevent the pest from accessing the plants (Chavez & Hurtado 2010) or biological control using parasitoids such as Diadegma insulare or Cotesia plutellae or entomopathogenic fungi and nematodes, which have shown significant advance in managing the diamondback moth (Sarfraz et al. 2005, Schroer et al. 2005, Bertolaccini et al. 2010).
With a 85% mortality, the entomopathogenic nematode, Heterorhabditis bacteriophora (Poinar 1976) (Rhabditida: Heterorhabditidae) is one of the species with the greatest potential to control P. xylostella (Shinde & Singh 2000, Somvanshi & Ganguly 2007).This nematode species is a cruise-type forager that can enter through the anus, spiracles or cuticle (Sánchez 2002).It is hermaphrodite during the first generation of its life cycle and has a tooth that allows it to perforate any insect cuticle, which gives it a competitive advantage over other biocontrol agents (Sáenz & Lopez 2010).
The success of using entomopathogenic nematodes and fungi in biological control strategies depends on the strains of natural enemies and their relationship with the pest (Sarfraz et al. 2005).The populations of M. anisopliae, B. bassiana and H. bacteriophora differ genetically and biologically depending on the region in which they are isolated; therefore, it is necessary to assess their virulence and pathogenicity to the pest to be controlled (Furlong et al. 2013).
The objective of this study was to assess the susceptibility of P. xylostella to H. bacteriophora HNI0100, B. bassiana Bb9205 and M. anisopliae Ma9236 for future studies on the interaction of fungi and nematodes to control diamondback moth.

Entomopathogens: H. bacteriophora HNI0100, M. anisopliae Ma9236 and B. bassiana Bb9205.
After screening different strains of Steinernematidae and Heterorhabditidae, entomopathogenic nematode H. bacteriophora HNI0100 was used for the laboratory tests (Delgado-Ochica & Sáenz 2012).The Biological Control Laboratory at Pontificia Universidad Javeriana provided the infective juveniles, which were multiplied in vivo by infecting last instar larvae of Galleria mellonella (Linnaeus 1756) (Lepidoptera: Pyralidae) following the methodology of Kaya & Stock (1997).The entomopathogenic fungi B. bassiana Bb9205 and M. anisopliae Ma9236 were obtained from the Laboratorio de Control de Calidad de Bioinsumos Agrícolas -Control de Bioinsumos Disciplina de Entomología, Cenicafé, Chinchiná -Caldas -Colombia.The fungi were activated by infecting P. xylostella.For the optimal growth and production of conidia, the reactivated strains were grown in PDA (potato dextrose agar) and oatmeal agar for 15 days at 25 °C.

Entomopathogenic Evaluation:
The susceptibility of P. xylostella to H. bacteriophora HNI0100, B. bassiana Bb9205 and M. anisopliae Ma9236 was evaluated by transferring one larva of third instar to 2 oz plastic containers, each one containing 5 g of broccoli leaf and 60 g of sterile river sand (Table 2).The infective juveniles (IJs) were suspended in distilled water with Tween 80 (0.1 % v/v) and inoculated over the sand and broccoli leaf.Larvae mortality by entomopathogenic nematodes was evaluated during 24, 48 and 72 hours.Subsequently, once dead, the larvae were transferred to White traps to recover infective juveniles.Fungal conidia were suspended in saline solution with Tween 80 (0.1 % v/v) and inoculated over the sand and broccoli leaf.Larvae mortality by entomopathogenic fungi was evaluated during two weeks.
Statistical Analysis: P. xylostella mortality percentages were tested for variance homogeneity (Levene) and normality (Kolmogorov-Smirnov); the assumptions for parametric data were met, demonstrating that treatment variances were the same and their errors had a normal 2 1 x 10 2 1 x 10 5 1 x 10 5 3 3 x 10 2 1 x 10 6 1 x 10 6 4 6 x 10 2 1 x 10 7 1 x 10 7 5 1,2 x 10 3 1 x 10 8 1 x 10 8 6 0 0 0 distribution.A univariate ANOVA with a factorial arrangement was used to determine whether dose and time means were the same or different in relation to the mortality percentage of each test (entomopathogenic nematodes and fungi) and to determine the interaction of both factors.Subsequently, multiple comparisons using Tukey and Scheffe tests were performed, employing a 95% probability, to identify the dose with the highest mortality of P. xylostella larvae, as well as the time during which the highest mortality occurred.The tests were performed using Statistix 10 software.
Similarly, the highest larvae mortality occurred at 48 and 72 hours post-inoculation.This experiment replicated the interaction between dose and time (F=205.71,p=0.0000) mentioned by Goettel et al. (1993).Larvae cavaders displayed a yellow-ocher coloration, which is an indicative of a IJs infection (Figure 2a).Additionally, the IJs migrated from cadavers (Figure 2b).
The highest mortality produced by B. bassiana Bb9205 (95,33%) was at a dose of 1x10 5 con/cm 2 15 days after inoculation, and by M. anisopliae Ma9236 (99,67%) was at a dose of 1x10 5 con/cm 2 15 days after inoculation.The interaction between dose and time for B. bassiana (F=32.15,p=0.0000) and M. anisopliae (F=171.94,p=0.0000) suggests a trend of increasing mortality over time.2007) demonstrated that the percentage of mortality increased with time, establishing that the mortality generated in the first 48 hours is due to a mutualistic relationship with Photorhabdus luminescens (Poinar 1976), which gives the nematode a competitive advantage allowing it to kill its host quickly (Mason & Wright 2000, Sáenz 2012).
According to Mason & Wright (1997) and Baur et al. (1995), the yellow-ocher coloration displayed by P. xylostella after infection by H. bacteriophora HNI0100 is not the characteristic symptomatology for Heterorhabditidae, which usually display a reddish hue.This divergence is attributed to the insect's original color (green), the concentration of bacterial cells of P. luminescens in each infective juvenile that enters the larva (30-200 IJs), the concentration of bacteria in the insect's hemolymph and the refraction of light on the cuticle (Silva et al. 2002, Waterfield et al. 2009).
The larvae had a symptomatology characteristic of infection by nematodes, displaying flaccidity and little mobility, as reported by Nyasani et al. (2008).
Third instar larvae were susceptible to B. bassiana Bb9205 and M. anisopliae Ma9236.Both fungi had the highest mortality at a dose of 1x10 5 con/cm 2 on day 15 after inoculation; this is comparable to results obtained by Franco (2001), who establish that the percentage of mortality increases with time, and is corroborated by studies by Loc & Chi (2007), Thuy (2001) Butt & Goettel (2000), that conclude that fungal pathogenicity is characterized by a gradual increase in mortality with extended exposure time.According to Quesada-Moraga & Vey (2004), Hui Wu et al. (2010) and Anaisie et al. (2011), concentrations of 1x10 5 con/cm 2 are low to cause a mortality percentage exceeding 60% in P. xylostella; these authors recommend the use of concentrations in excess of 1x10 9 con/cm 2 to produce a larval mortality of 80%.This study demonstrates that a high mortality can be obtained at low-doses, depending on the species and strain of the entomopathogenic fungus (Roy & Pell 2000, Furlong & Pell 2001, Sun et al. 2002, Furlong 2004).Sarfraz et al. (2005) indicate that the success of biological control strategies depends on the accurate identification of natural enemies and association with the strain; misidentifications can prompt the failure of the pest management program.Ibrahim & Low (1993) and Shelton et al. (1998) state that strains of entomopathogenic fungi produce different responses in mortality of individuals because their virulence and pathogenicity are not the same.By conducting tests on P. xylostella with six isolates of B. bassiana and two of M. anisopliae, Godonou et al. (2009) found that the strain with the highest mortality (94%) was Bba5653, establishing that the percentage and death rate of the pest can be variable.A study by Vandenberg et al. (1998) determined that the survival times for P. xylostella larvae, inoculated at different doses, were variable for two strains of B. bassiana, pathogenicity decreased with an increased dose, which agrees with our results.They also established that survival times and mortality vary between the two isolates, this is related to strains and their virulence.
The relationship between doses, mortality, biological controllers and pest generally responds to a positive correlation between the dose and the pathogen ability to kill the pest (Pena et al. 1991, Ferron 1981, Butt & Goettel 2000).In this study, the nematodes dose with the highest mortality was the most concentrated (1,2x10 3 JIs/cm 2 ), which may suggest a positive correlation between dose and mortality.Contrary, B. bassiana Bb9205 and M. anisopliae Ma9236 showed a possible negative correlation, thus 1x10 5 con/cm 2 , the dose with de highest mortality, is one of the lowest doses tested.This result may be explained by Butt & Goettel (2000), who mentioned that for entomopathogenic fungi there is a threshold dose to kill a pest, nevertheless the exact nature of this relationship has not been defined yet.Studies by Goettel et al. (1993) and Vandenberg et al. (1998) report a negative correlation between doses and mortality at the highest concentrations, occurring an autoinhibition; which may explain our mortality results for the fungal strains.
This study is of high importance as it gives way to establish interactions between fungi and nematodes that can generate possible synergy or additivity to pest mortality.

Conclusion
The inoculations of entomopathogenic nematodes (1,2x10 3 JIs/cm 2 ) and fungi (1x10 5 con/cm 2 ) cause mortality rates exceeding 90% for P. xylostella, in comparison with the majority of studies using higher concentrations to achieve the same mortality.The use of Colombian strains of H. bacteriophora HNI0100, B. bassiana Bb9205 and M. anisopliae Ma9236 is an innovative alternative for the control of P. xylostella and should be considered a management strategy in broccoli production.

Fig. 1 Fig. 2 .
Fig. 1. a. Susceptibility of P. xylostella to H. bacteriophora HNI0100, B. bassiana Bb9205 and M. anisopliae Ma9236 in time.a. H. bacteriophora HNI0100.b.B. bassiana Bb9205.c. M. anisopliae Ma9236.Ratios were calculated for each treatment and used to determine the standard error.Treatments with the same letter had no significant difference (p<0.05).

Table 1 .
Continents where Brassica oleracea (Broccoli) is grown and annual insecticide application costs (U.S.$).CI: Cost of insecticide application, and CIPM: Cost of insecticide application with IPM.

Table 2 .
Treatments and doses used for experimental design of Plutella xylostella susceptibility to entomopathogens H. bacteriophora HNI0100, B. bassiana Bb9205 and M. anisopliae Ma9236.