Sarsina violascens spatial and temporal distributions affected by native vegetation strips in eucalyptus plantations

The objective of this work was to evaluate spatial and temporal distributions of Sarsina violascens (Lepidoptera: Noctuidae: Lymantriinae) in two Eucalyptus cloeziana plantations, one with native vegetation strips (WNVS) and another without them (ONVS). Adults were collected with light traps, which were installed: inside an area of native vegetation (Cerrado), 100 m from the edge; at the contact zone between the native vegetation area and the E. cloeziana plantation; inside the E. cloeziana plantation, 250 m from the edge; at the central part of the native vegetation strip, around 500 m from the edge (WNVS) or in the contact zone between two E. cloeziana compartments (ONVS); and inside the E. cloeziana plantation, 500 m from the edge. The number of S. violascens adults collected was 240 in the system WNVS and 1,378 in the system ONVS. The lower number of individuals in the system WNVS was probably due to favored biological control provided by higher species richness with the use of native vegetation strips. These strips, intermingled with E. cloeziana plantations, allow a higher proportion of native forest in the landscape and can help to reduce S. violascens infestations.


Introduction
The spatial distribution of insect outbreaks in Eucalyptus spp.(Myrtales: Myrtaceae) plantations can be aggregated, random, or uniform (Silva et al., 2014), and rainfall and temperature are the main factors affecting the temporal distribution of lepidopteran pests (Zanuncio et al., 2000).Knowledge of the spatial and temporal distributions of forest insect outbreaks is fundamental for using the proper control methods, determining damage level, incorporating spatial dynamics within the population model, and optimizing sampling techniques (Zanuncio et al., 2000;Silva et al., 2014).
Sarsina (Lepidoptera: Lymantriinae) is a Neotropical genus of Noctuidae, with 8-10 species superficially described, native to Mexico and Central and South Americas.Sarsina violascens (Herrich-Schäeffer, 1856) (Lepidoptera: Noctuidae: Lymantriinae) occurs from the south of Mexico to Argentina.This insect is an important pest in central Brazil and in portions of Mexico where Eucalyptus spp.plantations are established (Ciesla, 2011).In Brazil, this species is widely distributed in the states of: Bahia, in the Northeast Region; Espírito Santo, Minas Gerais, and São Paulo, in the Southeast Region; Pará, in the North Region; and Paraná, Rio Grande do Sul, and Santa Catarina, in the South Region (Zanuncio et al., 2000;Bernardi et al., 2011).In Mexico, S. violascens has been reported in the states of Tabasco, in the Southeast Region, and of Veracruz, in the South-Central Region (Ciesla, 2011).
Richer environments in plant species generally have a higher number of Lepidoptera species with a lower number of individuals of each species (Zanuncio et al., 2003;Freitas et al., 2005;Zanuncio et al., 2006).The reduction of species homogeneity, with the preservation of the remnants of the native vegetation in Eucalyptus spp.plantations (Cunningham et al., 2005;Hawes et al., 2009;Macedo-Reis et al., 2013), represents an important ecological principle for pest management (Pereira et al., 2001;Murta et al., 2008;Ribeiro et al., 2013).
The objective of this work was to evaluate spatial and temporal distributions of Sarsina violascens (Lepidoptera: Noctuidae: Lymantriinae) in two Eucalyptus cloeziana plantations, one with native vegetation strips (WNVS) and another without them (ONVS).

Materials and Methods
The study was carried out in the following municipalities of the state of Minas Gerais, Brazil: Paineiras (18º57'S, 45°26'W, at a 600-m altitude), in a plantation system WNVS; and Paraopeba (19º17'S, 44°29'W, at a 700-m altitude), in a system ONVS.In the system WNVS, vegetation strips of Cerrado vegetation (25-m width each) were kept every 500 m of reforestation.These strips were connected to each other, to vegetation reserves, or to vegetation islands (400-m width, every 10 ha).Areas near dams, lakes, paths, and waterways were considered of permanent preservation.In a landscape context, the area in Paineiras has a higher proportion of native forest than in Paraopeba.
Paineiras belongs to the central mesoregion of the state of Minas Gerais and to the microregion of the municipality of Três Marias, whereas Paraopeba belongs to the metropolitan mesoregion of the municipality of Belo Horizonte and to the microregion of the municipality of Sete Lagoas, also in the state of Minas Gerais.Both studied areas have tropical climate, classified as Aw according to Köppen, with a dry season.Their native vegetation belongs to the Cerrado biome, and the evaluated areas were located on the soil type Latossolo Vermelho-Amarelo (Oxisol) (Durães et al., 2011).
The botanical families of the species found in the native vegetation strips studied were: Lepidopteran individuals were sampled with light traps in the two E. cloeziana compartments (WNVS and ONVS).In the beginning of both studies, the trees were two years old and 2.6-4.0-mtall, and were spaced at 3.0×2.5 m.Leaf-cutting ants were controlled with sulfluramid baits, which have no effect on other insects (Zanetti et al., 2008).Vegetation under E. cloeziana trees (mostly weeds) was not removed.
Five light traps, model Intral 012 (Intral, Caxias do Sul, RS, Brazil) were installed per system with the help of a global positioning system (GPS): inside an area of native vegetation (Cerrado), 100 m from the edge; at the contact zone between the native vegetation area and the E. cloeziana plantation; inside the E. cloeziana plantation, 250 m from the edge; at the central part of the native vegetation strip, around 500 m from the edge (WNVS) or in the contact zone between two E. cloeziana compartments (ONVS); and inside the E. cloeziana plantation, 500 m from the edge.The contact zones between the Cerrado biome and the E. cloeziana plantation (WNVS), as well as among E. cloeziana compartments (ONVS), corresponded to land roads, approximately 7 m wide.The distance between the traps was at least 250 m, in order to minimize possible interferences between them.
Sarsina violascens adults were collected every 15 days, from October to April, in the five light traps, in both plantation systems.The traps were equipped with a fluorescent black-light bulb, model F15 T12 LN (Havells Sylvania Brasil Iluminação Ltda., São Paulo, SP, Brazil), emitting wavelength from 290 to 450 mm, powered by a 12-V and 55-A battery.They were installed at a 2-m height, considering the midpoint of the lamp length (Zanuncio et al., 2014).The collection period included pre-and post-rain seasons in the studied regions (Durães et al., 2011).The traps were turned on between 18:00 and 19:00 p.m. and turned off between 07:00 and 08:00 a.m.(13-hour operation), since S. violascens adults have a nocturnal habit (Dall'Oglio et al., 2013).The collection of the insects was postponed one or two days in case of rain (Dall'Oglio et al., 2013).Captured insects were removed from 20-L plastic bags coupled to a trap funnel.These bags had newspaper strips and an ethyl acetate pot with a wick, to hasten the death and prevent excessive flaking of the insects (Bernardi et al., 2011).These procedures were based on traits of S. violascens.The species can have three generations per year in Mexico, with active life stages during eight months (Ciesla, 2011).However, the number of individuals in the third generation is reduced to low levels due to the control by natural enemies in the first two.In Brazil, adult specimens are active from March to December, depending on local climate.In addition, both sexes fly and are active at night (Zanuncio et al., 2000;Bernardi et al., 2011), and the female deposits up to 40 eggs, singly or in strips, on foliage.The duration of the egg stage is around 11 days, whereas the duration of the larva stage is around 37 days.Larvae have nocturnal habit and congregate on the lower third of boles of host trees during the day, entering a pre-pupa stage for two days when feeding is completed.Pupation lasts around nine days and occurs in foliage, tree trunks, and understory vegetation, i.e., low height vegetation beneath the canopy (Carlos et al., 2010).
Collected S. violascens adults were sorted, cataloged, and mounted in the Laboratory of Biological Control of Insects (LCBI) at Universidade Federal de Viçosa (UFV), in the state of Minas Gerais, Brazil.The males and females were deposited in the Regional Museum of Entomology of the institution.The confirmation of species identification was performed by the specialist Dr. Olaf Hermann Hendrik Mielke, of the Department of Zoology of Universidade Federal do Paraná (UFPR), in the state of Paraná, Brazil, after comparisons with keys, taxonomic descriptions (Lafontaine & Fibiger, 2006), and with the material previously deposited at that institution.
The number of S. violascens adults captured were recorded fortnightly in the five traps, monthly in each trap, and per collection site in each studied area (WNVS and ONVS), from October 2012 to April 2013.The average temperature and monthly rainfall were obtained monthly from the nearest weather station.

Results and Discussion
The population dynamics of S. violascens differed between Paineiras (WNVS) and Paraopeba (ONVS) in samples from October 2012 to April 2013.This dynamics was similar to the one reported for Oxydia vesulia (Cramer, 1779) (Lepidoptera: Geometridae) in a similar study in the same regions, from October 1993 to April 1994 (Santos et al., 2002).Native vegetation strips, intermingled with E. cloeziana plantations, provided a higher proportion of native forest in the landscape and this probably contributed to the reduction of the number of S. violascens adults collected in the light traps: 240 in the system WNVS and 1,378 in the system ONVS (Table 1).The lower number of S. violascens adults in the system WNVS suggests that higher plant species richness in the landscape, compared to that of the system ONVS, can maintain a low frequency of this pest species due to biological factors.This agrees with the fact that several small shelters and higher proportions of native forest provide better conditions of biological balance than a few or no native vegetation compartments (Silva et al., 2010;Dall'Oglio et al., 2013;Macedo-Reis et al., 2013).Therefore, the lower occurrence of S. violascens individuals in the system WNVS might be explained by the higher heterogeneity in plant species richness and by the higher quantities of sources of pollen, nectar, and alternative hosts in this system, which favors natural enemies (Zanuncio et al., 2009;Macedo-Reis et al., 2013).Additionally, the higher number of S. violascens in the system ONVS and the abrupt rise in this number at the end of the sampling period indicate that a more homogeneous plant species richness increases the frequency of organisms.This occurs mainly when S. violascens reaches eucalyptus plantations without native vegetation strips, where its number increases rapidly, because of the lower number of biological control agents (Bernardi et al., 2011).
Sarsina violascens showed a more uniform frequency in the system WNVS, with a lower number of individuals from January to April 2013, whereas a higher number of Lepidoptera individuals was observed in the system ONVS in March and April 2013 (Figure 1 A).On the one hand, the reduced number of S. violascens individuals collected in the system WNVS throughout the evaluation months indicated that the higher plant species richness and greater proportion of native forest in the landscape in this system contributed to avoid the population explosion of the pest.Furthermore, it showed that this species, despite being able to establish itself in native hosts, is under control in richer plant species environments (Zanuncio et al., 2000;Bernardi et al., 2011).On the other hand, the smaller area with native vegetation in the system ONVS hinders the maintenance of residual populations of S. violascens; however, the lower plant species richness and native forest proportion in the landscape favored a population increase of the pest in February 2013, with higher populations in April 2013 (Figure 1 A).
The number of S. violascens individuals collected was higher in the third collection site, inside the E. cloeziana plantation, in both the systems WNVS and ONVS (Figure 1 B).However, this number was (1) Trap 1, inside a native vegetation (Cerrado) area, 100 m from the edge; trap 2, at the contact zone between the native vegetation area and the E. cloeziana plantation; trap 3, inside the E. cloeziana plantation, 250 m from the edge; trap 4, at the central part of the native vegetation strip, around 500 m from the edge (WNVS) or in the contact zone between two E. cloeziana compartments (ONVS); and trap 5, inside the E. cloeziana plantation, 500 m from the edge.
higher in the collection sites of the system ONVS, and the value and distribution of S. violascens individuals per site showed a lower variation in the system WNVS.High populations and outbreaks of pest species, such as S. violascens, were linked to factors including low vegetal species richness (Zanuncio et al., 2001(Zanuncio et al., , 2003(Zanuncio et al., , 2006)).The expansion of Eucalyptus spp.plantations in areas of the Brazilian Cerrado contributes to this outcome, because the poorer understory favors the unbalance of insect pest populations (Pereira et al., 2009;Silva et al., 2010).The studied species was reported as a dominant one with wide distribution and abundant, floating populations (Zanuncio et al., 2001;Lopes et al., 2007;Bernardi et al., 2011).In addition, its adaptation to Eucalyptus spp. is favored, because it is a polyphagous species with many native hosts of the family Myrtaceae (Zanuncio et al., 2000).
Precipitation and temperature, in general, did not affect the number of S. violascens individuals in the system WNVS (Figure 2 A), but in the system ONVS, there was an increase at the end of the rainy season (Figure 2 B).Adults were not recorded on some dates in the system ONVS, but their number increased from February 2013 onwards, reaching a population peak (with 800 collected individuals) after two months, in the second half of April 2013, which suggests a lower population stability in more simplified systems.
Intermingling native vegetation with E. cloeziana plantations in the system WNVS helped in the reduction of the number of S. violascens adults.Moreover, the interconnection of preservation areas by strips and the higher proportion of native vegetation in the landscape can contribute to decrease insulation, which facilitates the dispersion of natural enemies in the environment  Pesq.agropec.bras., Brasília, v.51, n.6, p.703-709, jun.2016 DOI: 10.1590/S0100-204X2016000600001 (Murta et al., 2008;Zanuncio et al., 2009;Tavares et al., 2013).This apparently is the major mechanism for maintaining the insect pests of Eucalyptus spp.below the damage level (Soares et al., 2007;Zanuncio et al., 2009), showing the importance of the technique in the integrated management of pests in Eucalyptus spp.

Conclusion
Native vegetation strips in forest plantations represent an ecological principle for integrated management of forest pest, and higher proportions in the landscape can help to reduce Sarsina violascens infestations in Eucalyptus spp.plantations.

Figure 1 .
Figure 1.Number of Sarsina violascens (Lepidoptera: Noctuidae: Lymantriinae) adults collected, monthly (A) and per collection site (B), with light traps installed in Eucalyptus cloeziana plantations with and without native vegetation strips in the state of Minas Gerais, Brazil.

Figure 2 .
Figure 2. Number of Sarsina violascens (Lepidoptera: Noctuidae: Lymantriinae) adults collected with light traps, as well as mean temperature and monthly precipitation, in Eucalyptus cloeziana plantations with (A) and without (B) native vegetation strips in the state of Minas Gerais, Brazil.

Table 1 .
Number of Sarsina violascens (Lepidoptera: Noctuidae: Lymantriinae) adults collected with light traps in the first and second fortnights, from October 2012 to April 2013, in Eucalyptus cloeziana plantations in systems with (WNVS) and without (ONVS) native vegetation strips.