Pathogenicity, Virulence and the Interaction of Metarhizium anisopliae and Beauveria bassiana against Phyllophaga vetula (Coleoptera: Melolonthidae)

1Research Center in Biotechnology. Universidad Autónoma del Estado de Morelos, Campus Chamilpa. Avenida Universidad 1001. 62209, Chamilpa, Cuernavaca, Morelos, Mexico. 2School of Higher Studies of Xalostoc. Universidad Autónoma del Estado de Morelos, Av. Nicolás Bravo S/N Parque Industrial Cuautla, Ayala, Morelos, México. C.P. 62715. 3Research Center for Biological. Campus Chamilpa. 4Faculty of Agricultural Sciences. Campus Chamilpa. 5Faculty of Biological Sciences. Campus Chamilpa.

Melolonthidae larvae have slightly to strongly curved, C-shaped bodies, distinctive legs, and hardened head capsules, and they are referred to as white grubs.Mexico is a centre of diversity for the Melolonthidae 1 , and a large number of species of the genus Phyllophaga have been recorded (386 in Mexico).Only relatively few species cause economic damage 2 ; these include Phyllophaga obsoleta (Blanchard), P. ravida (Blanchard), and P. vetula Horn, which are distributed throughout the Mexican highlands 3 .
Phyllophaga spp.cause severe damage to maize, sorghum, wheat, sugarcane, bean, amaranth and peanut in Mexico, Central America and the USA.Historically, control measures for white grubs have depended mainly on persistent chemicals, but because of concerns regarding safety and environmental contamination, other forms of control such as biological control have been proposed 3 .
Due to their underground habitat during development, the grubs are susceptible to infection by microorganisms such as viruses, bacteria, protists, fungi and nematodes 3,4 , with the latter having a high potential for use in the control of microbial growth 5 .An ecologically safe strategy in combating soil pests, the use of entomopathogens, is based on the identification of a complex of pest species and their native pathogens and to subsequently select the microorganism with the greatest potential for this purpose, taking as benchmarks the virulence, mobility, persistence, specificity and production costs of the pathogen 6 .
Moist conditions, a relatively stable temperature, and protection against ultraviolet light from the soil favour the infection of larval melolonthids by entomopathogenic fungi 7 , providing them with a high potential to act as control agents against rhizophagous larvae.Previous work involving bioassays with larval melolonthids has been inconsistent, largely because these bioassays have been performed using larvae collected in the field because breeding these species is difficult as a result of their annual cycles and underground habit.The objective of this study was to determine the pathogenicity, virulence and interactions between native Morelos state strains of Metarhizium anisopliae and Beauveria bassiana against P. vetula.

Fungal isolates
Seven isolates of M. anisopliae and 20 of B. bassiana that were previously obtained in a survey conducted in Morelos State to collect native isolates of entomopathogenic fungi from infected white grubs or an insect tramp (Galleria mellonella) in a maize field were used 8 .The fungi were grown on Sabouraud dextrose agar (SDA) that included 5 g/l of mild peptone, 5 g/l casein peptone, 40 g/l dextrose, and 1.5 g/l agar.The culture was adjusted to pH 5.6 ± 0.2, was incubated in a dark room at 27 ± 1 °C for 15 d to induce sporulation, and was then preserved at 4 ± 2 °C. .Conidia were recovered from the Petri dishes using distilled water (with 0.05% Tween 80) in a laminar flux chamber (CFLV-80; Aparatos de Laboratorio BG, Mexico).The conidia were counted in a Neubauer chamber.

P. vetula larvae
A large number of insects were required to perform the bioassays.For the pathogenicity and interaction experiment, a total of 1000 thirdinstar larvae of P. vetula were collected in the field in Villa de Ayala, Morelos, Mexico, placed individually in 30-ml plastic cups covered with plastic lids, transported to the laboratory, and maintained at 25±1°C for 7d before the bioassay.The collected larvae were separated based on the presence of palidia that were almost parallel in the last abdominal segment (raster) and 23-30 pali 9 .A small piece of carrot was added to each cup for food.
For the LC 50 bioassays, a total of 600 firstinstar larvae of P. vetula were field-collected at the same site but were maintained at 25±1°C up to the third instar.

Bioassays
The procedure to inoculate each isolate was the same for all of the experiments.For the analysis of pathogenicity, 27 treatments (seven isolates of M. anisopliae and 20 of B. bassiana) with 15 larvae each were tested.Field-collected larvae were surface-sterilized with a 2% NaCl solution, washed with distilled water, and placed on paper towels to eliminate excess water.The bioassay followed a modified "maximum challenge test" methodology, which is useful for separating virulent from non-virulent isolates at the early stages of entomopathogen screening programmes 10 .Because of the great variability in isolates 11 , various conidial densities were used in the bioassay rather than sporulating strains.
For the interaction bioassay, the conidia of M. anisopliae (HI-019) and B. bassiana (HI-113) were evaluated alone and in combined doses of 1 × 10 8 con/ml, and distilled water (with 0.05% Tween 80) was used as a control.A completely randomized design was used, and each experimental unit was composed of 15 third-instar larvae of P. vetula.Each treatment was applied in the same way as in the pathogenicity bioassays.All bioassays were carried out in triplicate.The PEREZ  After treatment, the larvae were placed individually into 30-ml cups with a piece of carrot as a food source.The larvae were maintained at 25±1°C, and mortality was evaluates by touching the grub on the thoracic segments with a probe.

Statistical analysis
The percentage mortality data were arcsine transformed for statistical analysis.After processing the data, we performed analysis of variance (ANOVA) and Tukey's multiple comparisons of means at a significance level of 0.05 using the statistical package SAS 9. 1 (2003).Probit analysis was performed to estimate the mean lethal concentration 50 (LC 50 ), and confidence intervals (CIs) were generated using the statistical package Polo Plus 12 .

RESULTS AND DISCUSSION
Pathogenicity is a qualitative measure of the ability of a pathogen or parasite to cause disease in a host (5).The fungal isolates of M. anisopliae and B. bassiana showed different pathogenicity against P. vetula.In general, the M. anisopliae isolates (Table 1) were more pathogenic than those of B. bassiana (Table 2), corroborating other studies with Phyllophaga spp 13,14 .M. anisopliae isolates from a Phyllophaga sp.host were more pathogenic (46.66 to 73.33%) than those from an insect tramp, G. mellonella (00.00 to 20%).In this way, differential susceptibility of Phyllophaga spp. to fungal infection has been reported elsewhere 15,16 , and in P. polyphylla, larval infection never exceeded 30% for B. bassiana or M. anisopliae 17 .
The mortality caused by the highly pathogenic isolate of M. anisopliae HI-019 (86.06%) decreased significantly (P Â 0.05) when the inoculation was simultaneous with B. bassiana HI-113 (61.06%), but the mortality was statistically the same as when a grub was inoculated with only B. bassiana (52.73%) (Table 3).In the biocontrol of insect pests, the efficacy of treatment with multiple pathogens has not been frequently investigated but may have some potential in effective management efforts.Co-infection in the field is not commonly reported; however, coinfection by Entomophthora aulicae and Paecilomyces canadensis was reported for Lymantria dispar in field observations of epizootic disease in a gypsy moth population in Japan 18 .In the rhizosphere, Phyllophaga spp.are frequently subject to co-infection by pathogens of distinct species 5 .From the experiments presented here, no beneficial effect was apparent in using the two fungi together.Similar results have been reported for other insect hosts 19 .Recent information about the antimicrobial activity of secondary metabolites isolated from B. bassiana and M. anisopliae has identified potentially bioactive substances with antimicrobial activity 20 , which can cause one fungal infection to outcompete another.
The estimated LC 50 values of M. anisopliae isolates Ma17 and Ma19 against thirdinstar larvae of P. vetula were 4.749 × 10 7 conidia/ mL and 7.684 × 10 8 conidia/mL, respectively, which are statistically equivalent.Thus, additional studies must be conducted to further evaluate these isolates against white grubs under greenhouse and/or field conditions 10,14 .Similarly, the more virulent strains can be considered as candidates for sustainable agriculture based on a strategy of conservation biological control 21 .

Table 1 .
et al.: STUDY OF Metarhizium anisopliae & Beauveria bassiana Mortality of third-instar larvae of P. vetula caused by the conidia from seven isolates of Metarhizium anisopliae up to 30 d after inoculation.The conidial concentration was 1 × 10 8 c/mL (n=15).All isolates obtained for this study were from locations within Morelos, Mexico.

Table 2 .
Mortality of third-instar larvae of P. vetula caused by the conidia from seven isolates of Beauveria bassiana up to 30 d after inoculation.The conidial concentration was 1 × 10 8 c/mL (n=15).All isolates obtained for this study were from locations within Morelos, Mexico.

Table 3 .
Mortality caused by isolates of M. anisopliae (HI-019) and B. bassiana (HI-113) alone and in combination at conidial densities of conidia/mL against third-instar larvae of P. vetula