*2.4. Data Analysis*

To analyze the effect of treatment on developmental time of *C. perspectalis* larvae and pupae, we fitted generalized linear models (GLM) with a Poisson error distribution and log link function. Mortality as well as germination data were analyzed using GLM with a binomial distribution and logit link. Treatment and replication were set as fixed effects. The analyses were performed in SAS® Studio for Linux [25] using the GLM procedure (PROC GENMOD) of SAS/STAT® module [26]. Means were separated by the least-square means (LSMEANS) statement of SAS with Tukey–Kramer adjustment for multiple comparisons. *p*-values <0.05 were considered statistically significant. Lethal concentrations (LC50 and LC90) were estimated using Probit analysis (PROC PROBIT). Differences in area not covered by *I. fumosorosea* mycelium were compared by an exact Wilcoxon two-sample test (PROC NPAR1WAY) of the SAS/STAT® module.

#### **3. Results and Discussion**

Most BTM larvae successfully passed to pupa regardless of treatment (Table 1), and no statistically significant effect of treatment on the duration of the larval stage was observed (χ<sup>2</sup> = 10.08, df = 5, *p* = 0.0730). Similarly, treatments had no significant effect on the duration of the pupal stage (χ<sup>2</sup> = 0.19, df = 5, *p* = 0.9992), but higher mortality was observed in all treatments; the maximum mortality of 46.4% pupae was found with the highest concentration of fungal treatment.


**Table 1.** The effects of *Isaria fumosorosea* on the development of *Cydalima perspectalis*.

1 Concentration of conidia per milliliter of suspension.

Mycosis was observed only in treatments of 1 × 10<sup>7</sup> and 1 × 10<sup>8</sup> conidia per 1 mL when 10% and 28.6% of pupae cadavers, respectively, were obviously infected by the fungus (Table 1). Infection by *I. fumosorosea* was later confirmed when fungus sporulated (Figure 1).

Interestingly, several adults that emerged from larvae treated by the highest conidia concentration were malformed (Table 1, Figure 2) and died in 1–2 days. A similar effect was observed when *I. fumosorosea* was applied to *Spodoptera littoralis* (Boisd.) [27].

**Figure 1.** (**a**) Early mycosis of*Isaria fumosorosea* on*Cydalima perspectalis* pupa; (**b**) Cadaver of*C. perspectalis* covered with sporulating fungus.

**Figure 2.** (**<sup>a</sup>**,**b**) Malformed adults of *Cydalima perspectalis* emerged in a group of larvae treated by *Isaria fumosorosea* at a concentration 1 × 10<sup>8</sup> conidia/mL.

The cumulative mean mortality, including mortality in malformed adults, varied among treatments and reached a maximum value of 60% when larvae were treated by a suspension of 1 × 10<sup>8</sup> conidia per 1 mL (Figure 3a). Thus, the highest mortality corrected for mortality in the control group [28] was only 42.9%. Although the effect of treatment on mortality was significant (χ<sup>2</sup> = 18.67, df = 5, *p* = 0.0022) and no significant differences were found between replications (χ<sup>2</sup> = 0.45, df = 1, *p* = 0.5004), the results indicate the very low susceptibility of *C. perspectalis* to *I. fumosorosea*.

The low efficacy is rather surprising, because the strain CCM 8367 of *I. fumosorosea* used in this study was previously found to be highly virulent against several pest species. For example, the mortality of pupae of *C. ohridella*, an invasive pest of *Aesculus hippocastanum* in Europe [29], treated by blastospores or conidia suspension of concentration 5 × 10<sup>7</sup> spores per 1 mL reached 100% over a few days [20]. Later, the high efficacy of this strain was confirmed against *Spodoptera littoralis* (Boisd.) in which an application of CCM 8367 blastospores at a concentration of 5 × 10<sup>7</sup> per 1 mL caused larval mortality >90% [27]. The high efficacy of CCM 8367 under laboratory conditions similar to that used in the present study was reported also against Colorado potato beetle, *Leptinotarsa decemlineata*, (Say) (Coleoptera: Chrysomelidae) larvae [30]. This indicated that the strain could be a prospective biocontrol agent, although some side effects against non-target natural enemies were also reported [31].

**Figure 3.** (**a**) Mean cumulative mortality (±SE) of *Cydalima perspectalis*(including mortality of malformed adults) treated with various concentrations of *Isaria fumosorosea* conidia. A generalized linear model was fitted and pairwise between treatment differences were tested using the least-square means. Different letters indicate significant differences between columns (*p* < 0.05); (**b**) Log-probit regression line of concentration-mortality response of *C. perspectalis* to *I. fumosorosea*.

The log-probit regression line describing the relationship between concentration and mortality has a form *y* = −1.264 + 0.135*x* (Figure 3b), but the slope was not statistically significant (χ<sup>2</sup> = 3.27, df = 1, *p* = 0.071). Thus, the extrapolated values of LC50 = 2.42 × 10<sup>9</sup> and LC90 = 7.88 × 10<sup>18</sup> were very high. For example, this contrasts with the LC50 and LC90 values of 1.03 × 10<sup>6</sup> and 8.67 × 107, respectively, reported for *L. decemlineata* [30].

LT-SEM imaging of *I. fumosorosea* conidia on the cuticle of BTM larvae revealed a high number of spores immediately after treatment (Figure 4a,b), but after 24 and 48 h of incubation, the number of attached spores seemed to be much lower, as we found them only in some places of larvae, usually as small groups or individual conidia. This indicates low conidial attachment to the larvae cuticle. Examination further showed that the number of spores did not germinate (Figure 4d–f). This finding might explain the low virulence of the fungus against *C. perspectalis* because the successful germination of fungus conidia on the host cuticle has been considered to be necessary for infection [32,33]. Several studies documented that the cuticle of some arthropods repress the germination of EPF spores or further development of germlings and appressoria formation [34,35]. One of the reasons might be the presence of antifungal compounds on the cuticle [36,37], which might be also case of *C. perspectalis*.

Results of in vitro experiments using *B. sempervirens* hydro-alcoholic extract revealed that this extract has a negative effect on the germination of *I. fumosorosea* conidia (Figure 5). In the control treated by solvent, the mean germination was 100% (SE = 0, *n* = 3), while on extract-treated agar, the mean germination was only 92.67% (SE = 0.88, *n* = 3) in average. The difference was statistically highly significant (χ<sup>2</sup> = 31.36, df = 1, *p* < 0.0001).

The inhibition zone assay showed the suppression of *I. fumosorosea* mycelium growth on filter paper discs treated by *B. sempervirens* extract. The mean area not covered by mycelium was 0.02 ± 0.01 mm<sup>2</sup> and 53.83 ± 19.86 mm<sup>2</sup> in control and treated discs, respectively. The differences were statistically significant (Z = −2.184, *p* = 0.028).

Our findings indicate the presence of phytochemicals in box tree leaves having some activity against entomopathogenic fungi. Several secondary plant compounds were found to have a negative effect on the germination of *I. fumosorosea* blastospores, indicating that the presence of allelochemicals on a substrate (e.g., insect cuticle or leaf) may be an additional constraint to the survival of entomopathogenic fungi [38]. The *Buxus* trees contain a lot of alkaloids, some of which are sequestered by *C. perspectalis* larvae, while some are metabolized and/or excreted [39]. The antimicrobial activity of substances extracted from *B. sempervirens* by 65% ethanol were found earlier [40], and similar effects of box tree extracts were confirmed by other authors [41]. Thus, it is thus likely that BTM larvae use

phytochemicals obtained from the host plant for their own defense against the invasion of microbial pathogens. This might explain the low efficacy of two strains of entomopathogenic fungi, *I. fumosorosea* (present study), and *B. bassiana* [14] against BTM.

**Figure 4.** Low-temperature scanning electron microscope (LT-SEM) image of *Isaria fumosorosea* conidia on the cuticle of *Cydalima perspectalis* larva. (**<sup>a</sup>**,**b**) Conidia immediately after the fungus application; (**c**) Conidium with germ tube after 24-h incubation; (**d**) Ungerminated conidium after 24-h incubation; (**e**) Group of ungerminated conidia after 24-h incubation; (**f**) Ungerminated conidia after 48-h incubation.

**Figure 5.** Germination of *Isaria fumosorosea* conidia on: (**a**) a control agar plate treated with solvent only; (**b**) agar plate treated with *Buxus sempervirens* extract. Arrows indicate spores with no or little germination peg. Traces of plant extract are visible on the agar surface.

It may be concluded that the strain CCM 8367 of *I. fumosorosea* is not a potent biocontrol agen<sup>t</sup> against *C. perspectalis* and that the reason for the low efficacy of the fungus might be the accumulation of host plant phytochemicals with antimicrobial activity in the fifth-instar larvae cuticle of the pest.

**Author Contributions:** Conceptualization, R.Z.; methodology, R.Z. and J.K.; conducting experiments, R.Z. and J.K.; data analysis, R.Z.; writing—original draft preparation, R.Z.; writing—review and editing, R.Z., J.K. and Z.U.A.; supervision, R.Z.; project administration, Z.U.A.; funding acquisition, Z.U.A. and R.Z. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research is funded by the Foundation for Science and Technology Development of Ton Duc Thang University (FOSTECT), website: http://fostect.tdt.edu.vn, under Grant FOSTECT.2018.12.

**Acknowledgments:** We acknowledge the core facility Laboratory of Electron Microscopy, Biology Centre CAS supported by the MEYS CR (LM2018129 Czech-BioImaging) and ERDF (No. CZ.02.1.01/0.0/0.0/16\_013/0001775). The authors thank Jiˇrí Vanˇeˇcek and Martina Tesaˇrová for their valuable help with SEM imaging.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
