MATERIALS AND METHODS

The survey of the parasitoids of M. persicae in Virginia tobacco was carried out during 2010, 2011, 2012, and 2013, where seasons lasted from October to December of each year. The main tobacco producing regions in Rio Grande do Sul were visited, totaling 42 cities (Table 1). The visited crops were managed conventionally using synthetic products such as fertilizers, herbicides, fungicides and insecticides throughout the growing process.

Table 1 Cities of Rio Grande do Sul where tobacco leaf collections with infestations of M. persicae were carried out in each season to verify the occurrence of its parasitoids. 

The methodology outlined by ^{Kavallieratos et al. (2005)} was adapted to this study. There was not set an experimental design, and tobacco leaves attacked by M. persicae, with different levels of infestation, were collected randomly. The leaves were then stored in plastic bags and sent to the Laboratory of Entomology of the University of Santa Cruz do Sul (UNISC), where they were cut into squares (3×3 cm), without accounting for the density of aphids in them. The material was conditioned in plastic containers (9.5 cm long × 7 cm wide × 5 cm deep), acclimatized at 26±2 °C for 10 d for further screening and verification of parasitoids emergence.

The emerged parasitoids were identified at a species level according to ^{Wharton et al. (1997)} and ^{Kavallieratos et al. (2001)}. Dr. Marcus Vinicius Sampaio, professor of the Federal University of Uberlândia, confirmed the identification of the specimens. Subsequently, the material was collected and stored in alcohol (70%) at the Entomological Collection of Santa Cruz do Sul (SESC).

In addition to the survey of the parasitoid occurrence, a correlation was made between the parasitoid species found and the temperature (°C) of the cities. For this purpose, the average temperature of spring was considered according to data obtained from ^{Climate-Data.Org (2017)}. This temperature was used because it is the period of planting, flowering, and the emergence of tobacco in the South region of Brazil, and consequently it represents a higher incidence of aphids.

For the correlation analysis between the occurrence of parasitoids and the temperature, the data of the cities only were used when ten or more parasitoids emerged from the collected aphids during all the crop seasons. As the sample number was different in each city and season, the total proportion of individuals in each site was considered.

The distribution map of the cities visited was plotted using the CorelDRAW® X7. The regression models were constructed using SigmaPlot 11.0 software (^{SigmaPlot, 2008}).

RESULTS AND DISCUSSION

During the four crop seasons studied, 2963 parasitoids emerged from M. persicae, of which 2305 (78%) were Aphidius colemani Viereck (Hymenoptera: Braconidae) and 659 (22%) corresponded to Praon volucre (Haliday) (Hymenoptera: Braconidae). Among the 42 cities visited, there was the occurrence of parasitoids in 25 of them (Figure 1). Summing up the data of the four seasons, P. volucre was the most abundant species in 13 cities and A. colemani in 12 (Table 2). The low occurrence of parasitoids or their absence in some cities can be justified by the small sample in some of them, either by not locating crops infested with aphids or due to the excessive use of chemical agents on the crops.

Figure 1 Cities in Rio Grande do Sul state where the surveys were carried out, highlighting the sites where there were occurrences of the parasitoids of M. persicae in tobacco crops. 

Table 2 Number of parasitoids sampled in the cities of Rio Grande do Sul during four tobacco crop seasons. 

^{Silva et al. (2012)} had already reported the occurrence of these parasitoid species in tobacco in Rio Grande do Sul. According to a bibliographical survey, A. colemani came from the Mediterranean and Central Asian regions. Since 1992, it has been marketed in several countries for the control of aphids in protected crops (^{van Lenteren, 1997}).

In the past, it was successfully used in southern Brazil to control wheat aphids (^{Gassen and Tambasco, 1983}), adapting to the climatic conditions. Aphidius colemani is considered a dominant species among those found in aphids in South America and presents a high potential as a biological control agent (^{Sampaio et al., 2007}), corroborating the results of this study.

On the other hand, endoparasitoid P. volucre, of Palearctic origin, was also introduced in Brazil for the control of wheat aphids, establishing itself and becoming part of the group of parasitoids with potential use as control agents of different aphid species in different crops (^{de Conti et al., 2008}). Nowadays, Praon volucre is a cosmopolitan species of great importance for several crops, both in field conditions and in protected environments in Brazil (^{Silva et al., 2008}). It may be related to the adaptation of the species to the different climatic conditions of each region.

In Greece, ^{Kavallieratos et al. (2005)} support that A. colemani and Diaeretiella rapae (M'Intosh) (Hymenoptera: Braconidae) are the principal parasitoid species of M. persicae in tobacco. In contrast, ^{Kavallieratos et al. (2004)} found that P. volucre was the dominant parasitoid species of M. persicae in a different tobacco growing area of Greece, whereas D. rapae was not recorded in that area. According to ^{Starý (1970)}, interspecific relations are influenced by the geographical distribution of parasitoids which also affects their occurrence.

With respect to the proportion of individuals in each municipality and the average temperature in spring, a correlation for temperature with respect to the proportion of P. volucre was verified (r²=0.92), that is, with the increase in temperature, there was a proportional decrease in the number of individuals of this species (Figure 2). For the occurrence of A. colemani, there was a correlation in which the proportional incidence of individuals of this species increased at higher temperatures (r²=0.84)(Figure 3).

Figure 2 Correlation between the proportion of P. volucre emerged from M. persicae according to the different spring average temperatures of each city. 

Figure 3 Correlation between the proportion of A. colemani emerged from M. persicae according to the different spring average temperatures of each city. 

The results demonstrate that P. volucre presents a greater predominance in cities with average temperatures in spring equal to or lower than 22 °C (Figure 2). These results were also observed in the evaluation of the potential of P. volucre as an agent for the control of the aphids Uroleucon ambrosiae (Thomas) and Macrosiphum euphorbiae (Thomas) (Hemiptera: Aphididae) (^{de Conti et al., 2008}; ^{de Conti et al., 2010}). High parasitism rates were observed at temperatures between 18 °C and 22 °C considering these climatic conditions favored mummification, emergence of parasitoids, and increasing in the longevity.

The parasitoid A. colemani presented predominance in cities with temperatures above 22 °C (Figure 3). This had also been observed by ^{Zanini et al. (2006)} in their study on aphids of the species Sitobion avenae (Fabricius) (Hemiptera: Aphididae), and by ^{Sampaio et al. (2007)} and ^{Sampaio et al. (2005)} in their work on the development of A. colemani at different temperatures and different climatic regions. The species presented a high emergence of individuals at temperatures above 22 °C, being possible to report emergence at even higher temperatures in some warmer regions.

The fact that A. colemani has a higher tolerance at high temperatures may explain its predominance in agricultural environments in the southern region of Brazil. Based on the literature, A. colemani is formed by a species group, which are important biological control agents: A. colemani, Aphidius transcaspicus Telenga, and Aphidius platensis Brethes. This diversity can have an impact on the plasticity of the species in different environmental conditions (^{Tomanovic´ et al., 2014}). This type of study is important to know the thermal limits of each species of parasitoids in order to infer the species most adapted to each climatic situation that in the future could be more effective as a tool in integrated pest management.

CONCLUSIONS

Two species of parasitoids A. colemani and P. volucre were surveyed on tobacco farms in Rio Grande do Sul, Brazil parasitizing M. persicae, being possible to infer that there is a variation in the occurrence of these natural enemies according to temperature.

Under the conditions of this study, temperature exerted a direct influence on the populations of parasitoids of M. persicae. Individuals of P. volucre occurred preferably at temperatures below 22 °C, unlike to A. colemani, which had a clear preference for higher temperatures, above 22 °C, showing a different thermal tolerance between both parasitoid species.

Therefore, the results demonstrate that there is a possibility of using the natural enemies found for the control of M. persicae in tobacco growing in Rio Grande do Sul state.