Influence of pastoral systems on Mahanarva spectabilis (Distant) (Hemiptera: Cercopidae) and the entomopathogen Metarhizium anisopliae (Metsch.) Sorokin

The influence of Urochloa brizantha (variety Marandu) grazing systems on Mahanarva spectabilis (Distant) and the entomopathogen Metarhizium anisopliae (Metsch.) was studied to understand the benefits of integrated systems in pest management. The pastoral systems studied were: (M) monoculture, (SP) silvopastoral and (ICLF). We assessed the number, per square meter, of alive spittlebug nymphs or infected by M. anisopliae as well as the demanded number of entomopathogen sprays in each pasture system to control the pest. Throughout the experiment period, M. spectabilis was the unique species found. Silvopastoral had a higher number of alive nymphs and a lower percentage of the infected nymphs compared to pasture in monoculture; however, in both systems, only one spray of M. anisopliae was enough to keep the pest below its threshold. In agrosilvopastoral system, there was no spittlebugs infestation. Thus, intensified production systems such ICLF may be more sustainable, considering pest aspects.


Introduction
Brazilian Govern has many strategies to reduce rates of Brazilian greenhouse emissions. Among them, there is the Plan ABC -Low Carbon Emission Agriculture (Brasil, 2012) where production systems like Crop-Livestock-Forest integration (ICLF) are encouraged. This strategy reduces the emission of gases that produce the greenhouse effect. Thus, facilitating more sustainable agricultural methods. These systems are developed to raise both the productivity and financial gain for the rural population and reduce the risks of environmental deterioration, enhancing the chemical, physical and biological soil richness (Santos et al. 2008) as well as communicating scientific know-how on the physiological ecology of the various species of plants and govern their interactions with the flora and fauna of their locale (Santos et al., 2010).
In light of these complex production systems, multidisciplinary studies are required not only to prove their viability but also to technically subsidize growers who decide to adopt those systems to produce food, timber, beef and milk.
Studies in soil nutrient distribution and their effect on the soybean yield (Diel et al. 2014); carbon stocks and soil nitrogen (Sacramento et al. 2013), pasture yield (Paciullo et al., 2011) under shade trees reveal that ICLF bring yield advantages. However, little is known about the influence of these systems on the community of insect pests and their natural.
Application of the entomopathogenic fungus Metarhizium anisopliae (Metsch.) (Hypocreales: Clavicipitaceae) to control spittlebugs in pasture and sugarcane are usual in Brazil (Hernándéz-Dominguez et al., 2016). However, the pathogenicity of the fungus depends on abiotic environmental characteristics, including soil moisture, which plays a crucial role in the fungal maintenance and multiplication (Alves et al., 2011).
Some studies have shown that shadow promoted by trees in silvopastoral systems reduce solar radiation, avoid soil moisture loss and increase air humidity (Paciullo et al., 2008;Law et al., 2011;Pezzopane et al., 2015) which contribute to entomophathogen establishment (Meyling et al., 2009).
In this study, we evaluated the possible influence of integrated systems on M. spectabilis infestation as well as the efficiency of M. anisopliae to control that spittlebug.

Methods
The experiment were carried out from November 2013 to July 2014 in Sinop-MT-Brazil (11° 51'S, 55° 35' W and 384 m altitude), a transition region between Cerrado and Amazon Rainforest (Araújo et al., 2009). The climate of the region is type Aw according to Köppen climate classification, owing a tropical winter and annual average temperature of 25ºC, Relative Humidity of 82.5% and a precipitation of 2,550 mm (National Institute of Meteorology 2018).
The three systems evaluated were: (M) monoculture pasture, (SP) silvopastoral system and (ICLF) agrosilvopastoral. The silvopastoral system had triple rows every 30 meters of the hybrid H13 Eucalyptus, (Eucalyptus grandis vs Eucalyptus urophylla), with 3m x 3.5m between the individual trees ( Fig.1). Palisade grass (U. brizantha) cv Marandu was sown between Eucalyptus rows at a density of 12 Kg.ha -1 (80% germination). The monoculture pasture was also sown as silvopastoral system. In agrosilvopastoral system, soybean was cultivated between eucalyptus rows during the summer followed by maize in autumn intercropped with palisade grass in order to offer pasture to cattle during the crops off-season (june -october). All production system had 2 ha.
Four plots (4 x 38m) per treatment were installed as perpendicular transects to Eucalyptus rows in a way that 16m length faced to south side of the rows (receiving direct sunlight) and 16m faced to north side (under the shade trees) ( Figure 1). In the monoculture pasture, four plots of 25m² were randomly arranged and installed within the area. To establish forage uniformity in silvopastoral systems and monoculture pasture, plants were cut every 28 days to mimic the recommended grazing height of 10 cm (Jayme et al., 2009). All the harvested biomass was transferred away from the site to remove any type of physical barrier that could offer solar radiation protection to spittlebugs nymphs (Chiaradia et al., 2014).

Quantification of the live and infected nymphs
Every week the evaluations were done on 1m² per sample point. The number of alive M. spectabilis nymphs was quantified at the plant base as well as the number of nymphs infected by fungus.
The sampling points in agrosilvopastoral and silvopastoral systems were chosen at 3 and 15m, respectively, from north and south faces of the central Eucalyptus hedgerow in a total of four samples per plot. The evaluation ranges were selected prior based on the projection of the tree shade to evaluate the influence of shading on the spittlebug population. In the monoculture pasture, the four samples were randomly selected in each plot.
Samples of adults were collected and sent to Dr Gervásio Silva Carvalho (Pontifica Universidade Católica do Rio Grande do Sul), a specialist in Cercopidae, in order to identify the species occurring in the experimental area.

Control measure
The spittlebugs were controlled whenever each plot reached an infestation of 25 nymphs per m² (Valério & Koller, 1992). Control was performed using a commercial biopesticide of M. anisopliae (Metarril ® ) formulated as wettable powder in a concentration of 2x10 9 viable conidia. The dosage used was 250 g.ha -1 and a spray volume of 150 L.ha -1 . The number of sprays required in each production system and the total number of insects gathered in each system were used as sustainability parameters to evaluated pastoral system.

Statistical analysis
The data were analyzed about their normality and homoscedasticity in order to decide the use of parametric or nonparametric statistical testing according to the nature of the data sets. All analyses were performed using the R software (R Development Core Team, 2008). Although the initial M. spectabilis infestation in silvopastoral system was almost three fold higher than the infestation in monoculture pasture, a single M. anisopliae spray was enough in both systems to maintain the pest population below its threshold throughout the entire rainfall period; this showed that using M. anisopliae as control was effective in both the systems.

Results and discussion
In terms of total nymph collected per m 2 during the experiment, a significant difference was observed between the pastoral systems (p <0.001), with the infestation being most prominent in silvopastoral system (n = 178.44) than in monoculture (n = 95.38) (Figure 3).
In relation to number of nymphs per m 2 in the distances of 3m and 15m from the North and South faces of the Eucalyptus hedgerow in the silvopastoral system, there were no different infestations among these distances (p = 0.94) (Figure 4).

Quantification of infected nymphs
Although the monoculture pasture had the lowest nymph infestation, the percentage of infected nymphs was greater in this system (15.7%) when compared with silvopastoral system (6%) (p <0.001) ( Figure 5). Spatial distribution analysis of infected nymphs in silvopastoral system, as well as the number of alive nymphs, show that neither the presence nor absence of Eucalyptus shading on pasture has any effect on control efficiency of the M. anisopliae on nymphs of M. spectabilis, inasmuch as no difference in the number of infected nymphs was reported between the distances measured (3m and 15m) from the north and south faces of the Eucalyptus hedgerow (p = 0.41) ( Figure 6).
In the pastoral systems with spittlebug infestation (monoculture and silvopastoral), this infestation occurred in the beginning of the rainy season (November 2013) when the total rainfall measured was 239,25 mm, providing enough soil moisture to end the quiescence process of the eggs present in the dry soil (Sujii et al., 2001); thus, rising the first M. spectabilis generation as described by Valério (2009)  ). Thus, more solar radiation reaches the soil surface, which has a high correlation with spittlebug survivance (Lohmann et al., 2010).
The shadow, necessary to protect nymphs of M. spectabilis, promoted by pasture intercropped with maize occurred in the end of March. At this time, M. spectabilis population naturally reduces due to quiescence sparkled by precipitation decrease. Besides, as soybean had been cultivated early the pasture, there was no historic of spittlebug eggs in the area, showing the significant influence of croppasture rotation to break the pest cycle as suggested by Macedo (2009).
In more complex systems, besides food diversity and shelter, moisture and favorable temperatures are provided to establish and facilitate the development of spittlebugs (Ferreira & Marques, 1998;Wink et al., 2005) due to trees canopies which block the direct sunlight, thus reducing soil moisture loss (Koller, 1988). Although the infestation of the M. spectabilis nymphs was affected by solar radiation, the reduction in radiation provided by the trees canopy was not enough to significantly reduce the spittlebug infestation, as there were no differences in the distribution of the pest among the sites with and without trees shadow.
The availability of the forest component can also positively influence the nutritional value of the forage, by increasing the nitrogen concentration of the dry matter (Ribaski et al., 2003). According to Koller & Valério (1987) and Chiaradia et al. (2014), well-nourished plants provide the best pasture nutrition for spittlebugs.
Considering the evaluations started two years after both pasture systems establishment, a lower infestation of M. spectabilis in pasture monoculture indicates the relation pathogen-host (fungusspittlebug) reached a balance and that system offers biotic and abiotic conditions for both organisms.
Eucalyptus may release metabolites with inhibitory activity to plants, animal and microrganisms (Yang et al., 2017;Rieff et al., 2016;Mendes et al., 2013). Those compounds are released in soil during decomposition of leaves, barks and branches or exudates released by the 17 roots, affecting plants, soil microfaune (Rieff et al., 2016) and microorganisms. Zhang et al. (2010) discovered that Eucalyptus plants ranging from 2 -4 years old express the peak of those inhibitory metabolites. This study reinforces our results since the trees in our experiment were three years old. Mendes et al (2013) confirmed in their review about rhizosphere microbiome, compounds released in soil by plants may affect spore germination and fungal growth. Santos et al. (1998) also reported that Eucalyptus species produce metabolites that inhibit the growth of soil fungi.  Further studies are required on inhibitory effects of these metabolites on M. anisopliae to understand the interactions of these integrated systems with the soil microbiota (Rizvi et al., 1999;Sobrero, 2004;Albach et al., 2010).
Although nymph infestation was higher and the percentage of infected nymphs was lower in silvopastoral system, none additional sprays of M. anisopliae was required to maintain the spittlebug population below its threshold action as well as in monoculture pastures. Crocomo (1990), Alves (1998);Pereira et al. (2008) report that the effectiveness of a biopesticide based on the microorganisms is good if accurately performed during the first infestation peaks as we observed in our study.

Conclusion
According to the results in this study, we conclude that agrosilvopastoral system with soybeans in the summer crop and maize intercropped with U. brizanta (var. marandu) was more sustainable, within the parameters evaluated, due to unfavorable condition for the multiplication of M. spectabilis. Therefore, there was no need for M. anisopliae sprays, subsequently decreasing the production costs. In addition, although the forest component, Eucalyptus negatively influenced M. anisopliae, its control was effective enough to maintain the spittlebug population under its action threshold.