Estimation of Median Lethal Concentration of Three Isolates of Beauveria bassiana for Control of Megacopta cribraria (Heteroptera: Plataspidae) Bioassayed on Solid Lygus spp. Diet

The kudzu bug, Megacopta cribraria (F.), is an urban nuisance and significant agricultural pest. The median lethal concentrations of three strains of Beauveria bassiana (Balsamo), including the Mississippi Delta native strain (NI8) isolated from Lygus lineolaris (Palisot de Beauvois), the commercial strain BotaniGard® (GHA) (Victor, NY, USA), and the B. bassiana strain isolated from M. cribraria (KUDSC), were estimated on kudzu bug adults. A technique developed to evaluate B. bassiana against L. lineolaris was used. Younger adults (eight days after collection) were treated with NI8 and GHA and older adult (50 days after collection) were treated with NI8, GHA and KUDSC. Higher concentrations (n × 106, n × 107) of NI8 and GHA caused kudzu bug mortality two days after treatment in younger adults and similar concentrations of NI8, GHA, and KUDSC caused mortality one day after treatment in older adults. Lower concentrations (n × 104, n × 105) were not significantly different in mortality between strains. LS50 values of the KUDSC were significantly lower than NI8 and GHA values in older adults. This is the first available information on median lethal concentration of B. bassiana on kudzu bug adults bioassayed on artificial diet. It was determined that B. bassiana (KUDSC and NI8) are highly effective for young adults at very low doses (LC50 1.98–4.98 viable spores per mm2).

Nucleotide sequences of eukaryotic translation elongation factor 1 alpha (EF-1α) and internal transcribed spacer 1 and 2 of ribosomal RNA (ITS) were obtained using [24]. Nucleotide sequence reads were used to search databases to identify similarity to EF-1α and ITS sequences from other B. bassiana isolates.

Production of KUDSC Spore Powder
The Mississippi Delta native strain NI8 isolated from L. lineolaris and the commercial strain GHA formulated as BotaniGard ® were obtained from stored sources of spore powder maintained at the USDA-ARS Southern Insect Research Unit (SIMRU). NI8 and GHA are produced at SIMRU in a regular basis for the L. lineolaris research program, stored sources are kept at´80˝C. To produce KUDSC spore powder, a small-scale of biphasic culture system for solid-substrate fermentation was used according to the method used at SIMRU as follow: a singe germinated spore was selected from a plate (plate used for identification above) manually with a toothpick and suspended in 500 mL of CSYE broth (40 g /L Glucose, 10 g/L KNO 3 , 5 g/L KH 2 PO 4 , 2 g/L MGSO 4 , 0.05 g/L CACL 2 , 2.5 g/L Yeast Extract). The suspension was agitated in a shaker (Incubator Shaker Series Excella E25, New Brunswick Scientific Co., Inc., Edison, NJ, USA) for three days at 25˝C. Ten plastic bags (PPB75SEH6/V35-53 580 mmˆ385 mmˆ360 mm) that contained 1000 mL of barley (Minnesota Grain Inc., SASO, East Grand Fork, MN, USA) and 600 mL of water were sealed, autoclaved and prepared as production medium. Aliquots of 50 mL of the inoculum (suspended B. bassiana spores-CSYE broth) were injected into each bag and placed in an environmental control room (27˝C, 85% RH, 12 D:12 L Photoperiod) for 10 days and allowed to ferment. The injected bags were routinely shaken to avoid coagulation or clusters allowing for optimal dispersion of the inoculum throughout the fermented substrate. The sporulated substrate was transferred to paper sacks (30.48 cmˆ17.78 cmˆ43.18 cm Barrel, Kraft) for ten more days or until the moisture content of the resulting conidia dried to a w ď 0.3. Moisture endpoint was measured using a water activity meter (AquaLab-Decagon Devices, Serie 3-0105641113; Decagon Devices Inc., Pullman, WA, USA). Conidia were separated from the dried substrate using graded sieves (Grainger, Sieve SS Frame 8, SS Mesh # 30 and 100) on a vibratory shaker (Advantech, 8 Test Sieve Shaker, Model Mainll-16Y908; Endecotts Limited, London, UK). Harvested spore powder from the KUDSC strain and samples of NI8 and GHA were examined for spore germination and spore quantification (spores mm 2 ). Amounts of 0.5 g of harvested spore powder that contained 1.20ˆ10 11 , 1.18ˆ10 11 , and 1.19ˆ10 11 spores per gram for NI8, KUDSC, and GHA, respectively were suspended in 50 mL of 0.04% Tween-80 (Sigma-Aldrich P8074) and diluted to obtain final concentrations of 7.02ˆ10 7 , 6.95ˆ10 7 , and 6.90ˆ10 7 spores per mL, respectively. Suspensions of conidia were sprayed on five disposable microscope cover slips using a spray tower modified from a Burgerjon tower [19] that covered an area of 38.5 cm in diameter. Concentrations (spores per mm 2 ) were quantified by counting spores deposited [19]. The process was replicated five times using the final concentrations (nˆ10 7 ) (n = diluted harvested spore powder/strain: 7.02, 6.95, 6.90 for NI8, KUDSC and GHA, respectively) prepared above. Resulting data were analyzed by analysis of variance [25]. Lower test concentrations (nˆ10 6 , nˆ10 5 , nˆ10 4 ) were extrapolated based on dilution of nˆ10 7 concentrations, and the number of spores applied was corrected for viability (germination) [26] for all concentrations ( Table 1). Aliquots of 6 mL of suspension (nˆ10 7 ) of each strain provided concentrations of about 350 viable spores per mm 2 ( Table 1).

Bioassay Procedure
Serial dilutions of four test concentrations of NI8, GHA, and KUDSC strains (nˆ10 7 , nˆ10 6 , nˆ10 5 , nˆ10 4 spores/mL) were prepared to treat kudzu bug adults and evaluate mortality and infection. Kudzu bug adults (unknown age) from the South Carolina collection site were separated into two groups. The first group (called young adults) was sprayed with NI8 and GHA 8 days after field collection on 8 October 2012. The second group (called older adults) were held for 50 days and sprayed with NI8, GHA, and KUDSC strains on 20 November 2012. The KUDSC strain was not applied to the young adults, as that strain was received and isolated after the first group was treated. Both bioassays were conducted at room temperature (23˘2˝C and 50%˘5% RH). Each assay treatment (concentration of individual strain) was replicated four times with 10 adults per treated replicate (380 and 520 individual per young and old adult groups, respectively). Control insects were sprayed with 6 mL water (water control). Treatments of NI8, GHA, and KUDSC (nˆ10 7 , nˆ10 6 , nˆ10 5 , nˆ10 4 spores/mL) concentrations (Table 1) were similarly delivered in a 6 mL spray volume. After application, insects were placed individually into a 29.7 mL SOLO cups with a solid diet developed for use in L. lineolaris bioassays [19]. RH of 80% within a diet solo cup (bioassay arena) has been reported [19]. Adults were examined daily for 10 days for mortality and sporulation. Dead insects were retained in the same cup until completion of the 10 days trial to observe sporulation. Sporulation percentage was measured for cadavers on which sporulation occurred (presence of mycelial growth) out of total dead.

Statistical Analysis
Computations for all experiments were performed using SAS system software [25]. A randomized complete block design with factorial arrangements was used for each group of insects as follow: 2ˆ5ˆ3 (young adults) and 3ˆ5ˆ3 (old adults) for mortality (strains: NI8 and GHA; concentrations: nˆ10 7 , nˆ10 6 , nˆ10 5 , nˆ10 4 spores/mL; and evaluation times: 3, 5, and 10 days after sprayed) and 2ˆ5 (young adults) and 3ˆ5 (old adults) for sporulation (strains: NI8, GHA, and KUDSC; and concentrations: same as above). Each treatment combination was repeated four times. Nonparametric estimates of the survival function of kudzu bugs were compared between treatments using PROC LIFETEST [25]. Statistical differences in the survival of M. cribraria were declared based on the log-rank statistic. Mortality and infection were analyzed by using PROC GLM to detect differences between treatments for each group of insects. Mortality and sporulation data for each group of insects and each strain were analyzed by PROBIT [25] using common logarithm (log to the base 10) of the concentration value.

Dose-Mortality Response of M. cribraria to B. bassiana
All isolates tested were pathogenic to kudzu bug adults. However, the mortality among isolates was highly variable (Figures 1 and 2). The GHA commercial strain had the lowest measured performance at all concentrations at all evaluation times. No significant differences in mortality were found between the KUDSC and NI8 strains at the highest concentrations tested (nˆ10 6 and nˆ10 7 ) for studies with old adults at 10 days after treatment ( Figure 2). Mortality in the first group with younger adults exposed to NI8 was three-to eight-fold greater than that of GHA, but no significant differences were found in sporulation between strains at all concentrations except the nˆ10 6 (F = 6.37; df = 3, 119; p = 0.0024). Mortality at three and five days for young adults (F = 0.50; df = 3, 119; p = 0.6105 and F = 0.91; df = 3, 119; p = 0.4042) and old adults (F = 0.91; df = 3,159; p = 0.4382 and F = 1.11; df = 3, 159; p = 0.3453) exposed to the lowest concentrations (nˆ10 4 ) of all strains was not significantly different from those insects exposed to water. Results from analyses of dose-mortality response by insect age and dose ratio for the three isolates indicated high variability in dose-mortality responses between adult age groups ( Table 2). Dose-ratios were higher for young adults than those for older adults. Isolate GHA was less virulent than NI8 in young adults. At 7.02ˆ10 7 , NI8 strain eventually killed about 60% while GHA eventually killed only about 40% of the insects. Comparison of LD 50 values among the three isolates in older adults did not show any significant differences among strains.

Time-Mortality Response of M. cribraria to B. bassiana
Time to mortality was measured through routine post-treatment observations. Higher rates plateaued faster than lower rates (Figures 3 and 4). The earliest mortality recorded was with the highest concentration of both B. bassiana strains on two and three days after treatment in young adults ( Figure 3) while older insects treated with NI8 and KUDSC strains began to exhibit mortality at one and two days after treatment (Figure 4). Mortality analyzed by the test of equality with the strata statement in´log (survival probability) PROC LIFETEST indicated no significant differences between concentrations for all isolates, with the exception of younger adults treated with NI8 (Log-Rank X 2 = 11.67, df = 4, p = 0.019) (Figures 3 and 4).   plateaued faster than lower rates (Figures 3 and 4). The earliest mortality recorded was with the highest concentration of both B. bassiana strains on two and three days after treatment in young adults ( Figure 3) while older insects treated with NI8 and KUDSC strains began to exhibit mortality at one and two days after treatment (Figure 4). Mortality analyzed by the test of equality with the strata statement in −log (survival probability) PROC LIFETEST indicated no significant differences between concentrations for all isolates, with the exception of younger adults treated with NI8 (Log-Rank X 2 = 11.67, df = 4, p = 0.019) (Figures 3 and 4).

B. bassiana Dose-Sporulation response (LS50) on M. cribraria: Effects of Strain and Concentration
The regression analysis performed (cubic trend model) with sporulation percentage per concentration, showed significant correlation between sporulation and spore concentration ( Figure  5). Sporulation increased when spore concentrations increased, ranging the r 2 values from 0.94 to Figure 4. Survival probability of old adults (treated 50 days after collection) of Megacopta cribraria exposed to Beauveria bassiana at different concentrations under laboratory conditions. Insects were fed with artificial diet after spray. p = 0.05, LIFETEST of Equality Over Strata.

B. bassiana Dose-Sporulation response (LS 50 ) on M. cribraria: Effects of Strain and Concentration
The regression analysis performed (cubic trend model) with sporulation percentage per concentration, showed significant correlation between sporulation and spore concentration ( Figure 5). Sporulation increased when spore concentrations increased, ranging the r 2 values from 0.94 to 0.99. The sporulation of the KUDSC strain was significantly higher than GHA and NI8 at all concentrations: .5-fold greater than GHA at nˆ10 4 , nˆ10 5 , nˆ10 6 , and nˆ10 7 , respectively. Analyses of dose-sporulation response by insect age and dose-ratio for all three isolates showed high variability in dose-sporulation response among adult age groups (Table 3). Dose-ratios for sporulation response (LS 50 ) were significantly higher for young kudzu adults than those ratios found for older adults. Isolate GHA presented less sporulation that NI8 in young adults and KUDSC showed a significantly higher infectivity in old adults among NI8 and GHA strains. 0.27; df = 3, 3, 159; p < 0.0001, 0.8483). Sporulation percentage of KUDSC was 2.0-, 2.0-, 2.1-, and 2.2fold greater than NI8 and 2.0-, 3.0-, 2.1-, and 3.5-fold greater than GHA at n × 10 4 , n × 10 5 , n × 10 6 , and n × 10 7 , respectively. Analyses of dose-sporulation response by insect age and dose-ratio for all three isolates showed high variability in dose-sporulation response among adult age groups (Table 3). Dose-ratios for sporulation response (LS50) were significantly higher for young kudzu adults than those ratios found for older adults. Isolate GHA presented less sporulation that NI8 in young adults and KUDSC showed a significantly higher infectivity in old adults among NI8 and GHA strains.

Discussion
The original goal for this study was to determine the lethal effects of NI8 and GHA strains on populations of adult of kudzu bug that would soon be moving to overwinter sites. The addition of the KUDSC strain to this investigation provided an opportunity to evaluate the lethal effects of this naturally occurring strain on populations of kudzu bug adults and compared its activity to NI8 and GHA. The significant differences in percent mortality among the strains, indicates that young and old adults of M. cribraria can readily acquire lethal doses of conidia from the direct inoculation method (spray delivery) developed for Lygus spp. [19]. The lack of significant differences in -log survival probability (Figures 3 and 4) shows that the Lygus spp. diet was not the optimal diet for M. cribraria. This likely accounts for the mark increase in mortality noted on Days 6 and 8 in the water-control. However, considering the results, the young and old adult controls survive long enough (>10% mortality) on the solid diet to facilitate the comparison of pathogenesis and sporogenesis phases against strains and concentrations. Infection of B. bassiana spores normally is wrapped with haemocytes, but after four days only the germ tube breaks through its envelope and invades the insect's hemolymph [27]. Other observation also indicated that a five-day incubation period at 25˝C is adequate for detection of most lethal infections [28]. In this study, mortality and sporulation are the main factor of evaluation to determine virulence (level of pathogenicity) of B. bassiana against M. cribraria.
The entomopathogenic fungus B. bassiana is known to infect kudzu bug in Asia [14]. Few reports of pathogenicity and subsequent mortality are available from its recent establishment in the U.S. where B. bassiana infection in caged plots of soybean with high densities of M. cribraria has been observed [8]. Two B. bassiana infected kudzu bug specimens were also reported in Georgia [13]. Recently, high incidences of B. bassiana were observed in collections of kudzu bug nymphs and adults from kudzu growing in Mississippi and Tennessee [29]. Infection rates estimated in these studies ranged from 10% to 95%. However, only limited published data on the lethal effects of B. bassiana on kudzu bug are available and no studies of bioassays on artificial diet has been reported.
The entomopathogenic fungus B. bassiana is a globally distributed, monophyletic taxon that has been isolated from a wide range of insect taxa [24]. Selection of virulent genotypes is an important aspect of enhancing microbial control of insects with entomopathogenic fungi [30]. The selection may increase the ability to sporulate on the host cadaver. The isolations of the KUDSC strain from M. cribraria in South Carolina and NI8 from L. lineolaris in the Mississippi Delta are just examples of virulent genotypes with high pathogenicity against targeted pests.Tthe highest concentration of NI8 used in this study (nˆ10 7 ) has been enough to obtain 100% mortality and 97% sporulation in L. lineolaris under laboratory conditions 10 days after spray [19]; but in this study, although more than 85% and 95% mortality was obtained in young and old adults respectively, only 32.5%˘4.7% (SD) and 35.0%˘4.8% (SD) sporulation was observed on treated young and old adults of M. cribraria, respectively. The NI8 strain of B. bassiana showed high pathogenicity to old adults of M. cribraria, but not significantly more mortality than that of the KUDSC strain at the highest test concentrations (nˆ10 6 and nˆ10 7 ) (Figure 2). Young and old kudzu bug adults treated with the commercial strain GHA showed low mortality and sporulation at all concentrations (Figures 1, 2 and 5). Similar results have been observed in several studies using commercial isolates of B. bassiana where moderately effective control of targeted insects was reported [18,21,[31][32][33][34]. The present results support previous observations that pathogenicity of B. bassiana is not always related to the original host or geographical origin [33,[35][36][37][38][39][40][41][42].
One disadvantage of entomopathogenic fungi is the long time period that elapses between exposure and death of the host [39]. The time period between exposure and mortality in this investigation ranged from 1 to 8 days for the highest and lowest concentration tested, respectively. These time periods of mortality do not differ from studies previously reported. For example a median lethal time of 4.9-8.4 days on adults and nymphs was observed on L. lineolaris [19] (Hemiptera: Pentatomidae) [44]. For standard viability determination, B. bassiana conidia are typically incubated from 16-24 h on agar-based media. Germination on the insect integuments should be slower, due to limited free moisture and the time required for penetration. In this study sporulation was observed 2-6 days after insect mortality for all strains at all concentrations. However, the successful accomplishments of fungal growth, which are pre-requisites for pathogenicity, were primarily observed in old adults sprayed with the KUDSC strain ( Figure 5). Sporulation percentage of cadavers treated with KUDSC was 2.2-fold greater than those for NI8 and 3.5-fold greater than those for GHA at the highest test concentration.
Estimating the media lethal dose of the fungal pathogen B. bassiana for M. cribraria acquired via spray acquisition of conidia is a process that has not been attempted; particularly on artificial diet. Estimating the LD 50 and LS 50 for M. cribaria requires not only a reliable method to evaluate the infection of propagules on its body, but also a determination of when to sample the individuals exposed to the inoculum. In the described bioassay, dose-ratios for mortality and sporulation were significantly higher for old adults of M. cribraria than those ratios found for young adults. The high susceptibility obtained in old adults may be occurred due to the low mobility of those insects at the time of the spray, assuming that its low activity could probably let them acquire more conidia than that on young adults, which effectively contributed to its infection, mortality and sporulation. In this experiment Conidial acquisition was a continued process (spray delivery) for a same period of time of the exposure for old and young adults while treated insects were in movement.

Conclusions
This study resulted in a reliable bioassay system that determined the median lethal concentration and sporulation of B. bassiana against young and old adults of M. cribraria. The LD 50 of 0.83 spores/mm 2 and LS 50 of 64.38 spores/mm 2 obtained for the KUDSC strain could have great potential for controlling late-season populations of kudzu bug. Compared with NI8 LC 50 s and LS 50 s for KUDSC were 5.2-fold and 238-fold lower, respectively. Although no results are presented in this study for mortality and sporulation caused by the KUDSC strain on early season young adults, it is assumed that LD 50 and LS 50 values (spores/mm 2 ) would be a little greater than those values observed for late-season adults. High infection of kudzu bug has been observed in Mississippi and Tennessee (MS, TN, USA), which is suggesting that the B. bassiana strains NI8 (ARSEF13136) and KUDSC (ARSEF13137) may provide useful levels of suppression and a potential bio-control option for kudzu bug in soybean [29]. These experiments were all conducted in the laboratory under conditions that are favorable to B. bassiana. Further studies are needed that examine the pathogenicity of these two strains of B. bassiana on various insect pests in soybeans field plots. In general the study reported here resulted that the bioassay system developed for L. lineolaris [18] was also successful to bioassayed M. cribraria. The current limitation of M. cribraria survival and development on this diet can be improved by modifying the existing Lygus diet. However, this bioassay system has been used to easily set up and manipulate first to fifth instar nymphs of M. cribraria, where the survival nymphs molted to the next instar producing data that showed good fit to the classical probit model for each instar [29].
Author Contributions: Maribel Portilla isolated the KUDSC strain, concluded the experiment, analyzed the data and drafted the manuscript. Walker Jones maintained insect colonies under quarantine and revised the manuscript. Omaththage Perera cloned fungal isolate KUDSC strain for identification and revised the manuscript. Nicholas Seiter found the KUDSC strain and revised the manuscript. Jeremy Greene provided the field insect colony and revised the manuscript. Randall Luttrell revised the manuscript.

Conflicts of Interest:
The authors declare no conflicts of interest.