INTRODUCTION

The oat (Avena sativa L.) is a multipurpose cereal that can be used for consumption by humans as grain or by animals as forage, hay, pasture, or silage, as well as for ground coverage for no-till systems (^{DAL MOLIN, 2011}; ^{CASTRO et al., 2012}).

In recent years, this cash crop has become an important alternative during Winter, especially in southern Brazil. Therefore, more productive plants of high quality are needed in response to the improvement of environmental stimuli (^{SILVA et al., 2012}; ^{HAWERROTH et al., 2013}).

Currenty, 10% of the oats produced in Brazil are grown in the state of Mato Grosso do Sul. However, the cultivated area is likely to increase, as A. sativa is an alternative to corn in crop rotation and succession systems in the Fall/Winter (^{RICHETTI; CECCON, 2008}; ^{CONAB, 2019}).

The Dourados region has the potential to grow oats, as this crop has low water requirements, wide tolerance to soil pH (4.5 to 8.5), and resistance to the frosts that occur in autumn and Winter in this area. In addition, large quantities of straw are produced, representing a better use of the soil, decreasing erosion, while weed growth is inhibited by the allelopathic effects of oats (^{CASTRO et al., 2012}). On the other hand, some factors may negatively affect yields, especially pests such as the aphid Rhopalosiphum padi (^{Linnaeus, 1758}) (Hemiptera: Aphididae) and the caterpillar of Spodoptera frugiperda (Smith, 1797) (Lepidoptera: Noctuidae), which has a high incidence in oats in traditional growing areas. The effect of the attack of R. padi can be also be exacerbated by the transmission of viruses such as the barley yellow dwarf virus (BYDV) in Winter cereals (^{ROZA-GOMES et al., 2008}).

Given the few scientific studies on pests of oats, most of the information on pest management has been generally based on other similar crops, such as wheat (^{ROZA-GOMES et al., 2008}; ^{PEREIRA; SALVADORI, 2011}; ^{PANIZZI et al., 2015}). Therefore, information is needed on the diversity of insect species associated with oats in order provide a framework for pest control based on Integrated Pest Management (IPM) (^{GALLO et al., 2002}).

This study was aimed at assessing the entomofauna and fluctuations in their populations in oats in Dourados, Mato Grosso do Sul, Brazil.

MATERIALS AND METHODS

The study was conducted at the Experimental Farm of Agricultural Sciences (FAECA) of the Federal University of Grande Dourados (UFGD), Dourados (22° 23 S, 54° 98 W), Mato Grosso do Sul, Brazil, in the 2014 and 2015 harvests, in an area of three hectares planted with oats.

In 2014, pre-planting desiccation was carried out on April 11^th with the herbicides glyphosate (480 SL) and 2,4-D + amine (720 SL) at the doses of three and one L/ha, respectively. Sowing occurred on April 28^th, at a density of 80 seeds per linear meter and basic fertilizing of N-P-K of 400 kg/ha of 8-20-20.

The pesticides used during oat growth were picoxystrobin (200 SC) + cyproconazole (80 SC) at a dose of 0.3 L/ha on July 10^th and trifoxystrobin (150 SC) + protioconazole (175 SC) 0.4 L/ha on July 25^th. No post-emergence insecticides and herbicides were used.

In the 2015 harvest, pre-planting desiccation was carried out on April 24^th with the herbicides glyphosate (480 SL) and 2,4-D + amine (720 SL) in the dosages of four and one L/ha, respectively. Sowing was performed on May 18^th, at a density of 80 seeds per linear meter and basic fertilization of N-P-K of 400 kg/ha of 8-20-20.

The crop practices used consisted of propiconazole (250 EC) + diphenoconazole (250 EC) applications at a dosage of 0.150 L/ha on June 22^nd, picoxystrobin (200 SC) + cyproconazole (80 SC) at a dose of 0.3 L/ha on July 15^th, and trifloxystrobin (150 SC) + protioconazole (175 SC) 0.4 L/ha on July 31^st. No post-emergence insecticides and herbicides were used.

The cultivated area used for the study was 13,520 m², subdivided into 80 plots of 169 m² each. In 2014, the first assessment was carried out at 12 days after emergency (DAE), followed by weekly evaluations, totaling 10 assessments in the year. In 2015, the first assessment was performed at 10 DAE, with nine evaluations.

Twenty insects were sampled in the field and identified at species level in the laboratory with the aid of identification keys (^{GALLO et al., 2002}; ^{PEREIRA; SALVADORI, 2011}).

For each sampling point, a 0.5 × 0.5 (0.25 m²) metal frame was used, and the number of insects found within the area delimited by the frame was counted. One sampling point was examined per plot. Aphids were examined by removing ten plants within the frame and counting the insects found in their aerial and root parts. For the other species, direct counting of insects presents in the plants and on the soil, and surface was carried out within the area delimited by the frame.

The entomofauna was analyzed using the ANAFAU software (^{MORAES et al., 2003}). Following ^{SILVEIRA NETO et al. (1995)}, the species with the highest indices, based on abundance, constancy, dominance, and frequency, were considered predominant.

Abundance was determined using the mean and standard error of the mean number of individuals sampled per species, thus determining a confidence interval (CI) at 5 and 1% probability and establishing the following abundance classes: sa = superabundant (discrepant values of insect number, identified by the residue analysis); va = very abundant (number of individuals greater than the upper limit of 1% CI); a = abundant (number of individuals within the upper limits of the CI at 5 and 1%); c = common (number of individuals within 5% CI); d = dispersed (number of individuals between the lower limit of CI at 5 and 1%); and r = rare (number of individuals lower than the lower limit of CI at 1%) (^{THOMAZINI; THOMAZINI, 2002}; ^{MORAES et al., 2003}).

The confidence interval of the mean frequency (percentage of individuals of a species in relation to the total number of individuals sampled) was determined with a probability of 5%, and the following classification was adopted: sf = super frequent (discrepant values of the number of insects, identified with the residue analysis); vf = very common (frequency greater than the upper limit of 5% CI); f = frequent (frequency within 5% CI); and lf = uncommon (frequency lower than the lower limit of 5% CI) (^{THOMAZINI; THOMAZINI, 2002}; ^{MORAES et al., 2003}).

Constancy was determined for each species by assessing the percentage of samples with a given species, calculated as follows: C = (number of species x / total number of samples) × 100. According to the values obtained, species were grouped into w = constant (C > 50%); y = accessory (C between 25 and 50%); and z = incidental (C < 25%). Species with frequencies exceeding the limit calculated by the formula were considered dominant: D = 1/total number of species × 100 (^{THOMAZINI; THOMAZINI, 2002}).

The relationship between the population of R. padi and syrphids was examined with the chi-square test (x²), which compares observed and expected frequencies (^{RODRIGUES et al., 2010}).

Temperature and rainfall data were obtained from the weather station of the Experimental Farm of Agricultural Sciences (FAECA) of the Federal University of Grande Dourados (UFGD).

RESULTS AND DISCUSSION

During the sampling period, eight species divided into four orders and six families were collected in 2014, totaling 26,262 specimens (Table 1). In 2015, eight species representing four orders, seven families and 5,099 specimens were sampled (Table 2).

In 2014, the order Hemiptera had the highest indices, with Rhopalosiphum padi and Schizaphis as dominant and constant species. In 2015, R. padi was superdominant, superabundant, super frequent, and constant.

Aphids were the main pest insects attacking oats in the present study, especially R. padi, thus confirming its status as the most important pest in cereals worldwide, and among the most important aphids in agriculture (^{LOECK et al., 2006}; ^{BLACKMAN; EASTOP, 2007}).

Our results are in accordance with those by ^{LAU et al. (2009)}, that reported R. padi and S. graminum as the main aphids found in wheat and oats Southern Brazil and the state Mato Grosso do Sul. According to these authors, R. padi has adapted well to Brazilian conditions, increasing the risk of transmission of the barley (BYDV)/cereal yellow dwarf viruses (CYDV) to which this insect is a vector.

^{PARIZOTO et al. (2013)} studying the main vector insects of the BYDV virus, identified R. padi and S. graminum as main pests, representing 68.6 and 15%, respectively, of the total aphids sampled in wheat and oats.

In addition, the temperature during this study was ideal for R. padi, resulting in its highest reproduction and development rates, as shown in (Fig. 1), which was around 18.9 and 19.5°C in 2014 and 2015 respectively (^{AUAD et al., 2009}; ^{PEREIRA et al., 2016}).

Figure 1. Climate parameters during the periods of evaluation in Avena sativa in 2014 and 2015 in the municipality of Dourados, Mato Grosso do Sul. 

The order Lepidoptera was as one of the most abundant with 1,929 individuals in 2014, represented by the species Spodoptera frugiperda and Mocis latipes, while in 2015 these same species accounted for 196 specimens. Larvae of S. frugiperda and M. latipes were the predominant species, very frequent and constant in the first year of study. In the second year, S. frugiperda remained dominant, very abundant, and frequent, however the population of M. latipes decreased, becoming a common and frequent species.

One of the major factors affecting the dynamics of S. frugiperda in the environment is the high availability of food, including invasive and cultivated plants (^{GALLO et al., 2002}). Therefore, polyphagia in S. frugiperda may be reflected in its higher indices in oats in this study. Despite that, not all plants equally used, with species of the Poaceae family as the preferred hosts of this pest in America (^{SÁ et al., 2009}; ^{CASMUZ et al., 2010}).

The insects of the family Shyrphidae accounted for 407 specimens in 2014 and 130 in 2015. They were considered dominant, frequent, and constant in this study. In addition, an interaction between the syrphid population and R. padi (Fig. 2) was observed, where in the years of study the largest syrphid populations occurred close to the largest R. padi populations. Table 3 shows the results of the chi-square test indicating the association of the abundance of these species, based on the p-value below the significance level of 0.05. In addition, when abundance of syrphids was high, oat plants were flowering, thus providing food for adults of this insect.

Figure 2. Population fluctuation of Rhopalosiphum padi and Syrphids during the assessments in 2014 and 2015 in Dourados, Mato Grosso do Sul. 

The presence of aphidophagous syrphids is determined by climatic factors, availability of flowers and preys, habitat, and geographical area, which is positively correlated with aphid population (^{AUAD; TREVIZANI, 2005}; ^{BOKINA, 2012}). Moreover, the ability of these insects to oviposite near aphid colonies and their voracity make them effective aphid predators (^{ALMOHAMAD et al., 2009}).

According to ^{BOKINA (2012)}, the main factor affecting the abundance of syrphids in cereals is the density of aphid population, and in crop rotation system, these insects prefer plants with the highest aphid reproduction rates. This author, studying the aphid population in cereals, also found that the largest populations of R. padi occurred in oats, concomitant with the largest syphid populations.

Coleoptera accounted for 570 insects, represented by the species Diabrotica speciosa, Lagria villosa, and Harmonia axyridis in 2014 and 2015. The species L. villosa, D. speciosa, and H. axyridis were considered dominant, common, and frequent, respectively, in the two years of study. H. axyridis, in 2014, was considered an incidental species.

Comparatively, the predator H. axyridis was less abundant than syrphids, but it also is an important biological aphid control agent and has a positive correlation with the population of this pest (^{KOCH, 2003}).

In this study, diversity, richness, and abundance revealed a community composed of a small number of abundant species, acceptable in monocultures (^{SILVEIRA NETO et al., 1995}). Since this crop is relatively recent in the region, this might reflect a pest-insect complex less adapted to this crop.

Our findings indicate which pest and beneficial insects occur in Southern Mato Grosso do Sul in oats. This information provides a framework for studies on Integrated Pest Management, ecology, and taxonomy, as a scientific basis for those involved in the production chain of this grain.

CONCLUSIONS

The entomofauna associated to oat crops was not diversified during the years this study.

The main pest insect associated to Avena sativa is R. padi.

Syrphids are the most abundant insects along with R. padi.