Phytoseiidae (Acari: Mesostigmata) on plants of the central region of the Brazilian Cerrado

Four surveys of Phytoseiidae species associated with native plants were conducted in ten fragments of Cerrado sensu stricto to determine the species associated with the plants of this region. Twelve of the most common plant species were sampled in each fragment. Twenty six phytoseiid species were recorded during the surveys. Stryphnodendron adstringens was the host plant that had the highest number of species (eight species). Amblyseius neochiapensis Lofego, Moraes and McMurtry, 2000 and Neoseiulus tunus (De Leon), 1967 were the most common phytoseiid species. These species occurred on 24 and 20 host species, respectively. Approximately 76 % of the phytoseiids found belonged to the subfamily Amblyseiinae Muma, 1961. Almost 27 % of phytoseiid species recorded were associated with only one plant species. In addition, approximately two thirds of species found in this study also occur in the Atlantic Forest. The comparison of our results from the central areas with previous surveys in peripheral areas of the Cerrado showed that only 12 phytoseiid species were common to both areas.


INTRODUCTION
Many species of Phytoseiidae Berlese, 1913 are predators, some of which proved to be highly effective in controlling pest mites in various crops (McMurtry and Croft 1997;Moraes and Flechtmann 2008). Therefore, there is a strong need for more and comprehensive studies about phytoseiid species.
Surveys on Phytoseiidae in specific regions, biomes or crops constitute the first step to further studies about the use of predatory mites in the biological control (Gerson et al. 2003). Many surveys like these have been conducted in South America, being the bases for important programs to control of pest mites such as Aceria guerreronis Keifer, 1965 (Gondim Jr. and Moraes 2001;Lawson-Balagbo et al. 2008), Mononychellus tanajoa (Bondar, 1938) (Farias et al. 1981; Moraes and McMurtry 1983;Moraes et al. 1991) and Tetranychus evansi Baker and Pritchard, 1960(Furtado et al. 2004Fiaboe et al. 2007;Guanilo et al. 2008c;2010;Rosa et al. 2005).
In Brazil, surveys on phytoseiid species from natural areas are mostly concentrated in areas of Atlantic Forest of state of São Paulo (Buosi et al. 2006;Castro and Moraes 2010;Daud et al. 2007;Demite et al. 2011;Feres and Moraes 1998;Feres and Nunes 2001;Feres et al. 2005;Zacarias and Moraes 2002). Though the Cerrado is the second largest Brazilian biome, surveys were made only on peripheral areas (Lofego et al. 2004;Lofego and Moraes 2006;Demite et al. 2009) and so far no information is available on these predators in its central areas.
Moreover, the Cerrado is currently threatened by the rapid increase in agricultural activities and constant loss of natural areas (Ministério do Meio Ambiente 2007), which may consequently threat its characteristic mite fauna. So, the objective of this study was to do the inventory of phytoseiid species associated with native plants of this biome, specifically in fragments of cerrado sensu stricto.

Study areas
This study was conducted in ten fragments of Cerrado, located in southeastern and central Brazil ( Figure 1, Table 1). The climate in the region is characterized as Aw of Köppen (Ribeiro and Walter 1998) with two distinct seasons: a rainy season (October to March) and a dry season (April to September). The chosen areas are situated in the core region of biome, considered the original distribution of the Cerrado (Ministério do Meio Ambiente 2007).

Sampling
Samples were taken in fragments between October 2009 and June 2010. Twelve species of more abundant native plants were sampled in each area, belonging to different strata and plant families. Abundance of plants was evaluated visually. At least two individuals of each plant species selected were sampled, in total 57 plant species, in all fragments (Table  1).
The material collected from each plant species consist of leaves and flowers (when present). We used the following procedure for mite extraction: firstly, the material was washed separately in buckets containing eight liters of 30 % ethanol. After each wash, the ethanol was filtered through sieves with 25 µm nylon mesh. With a wash-bottle, the debris on the mesh was washed with 67 % ethanol into vials filled with the same alcohol for preservation. These procedures were done in the fragments, immediately after the samplings. So, the vials were placed in plastic boxes and brought to laboratory by car.
The material was screened under a stereoscopic microscope in laboratory. Phytoseiids found were mounted in Hoyer's medium (Moraes and Flechtmann 2008) and the slides labelled. The collection data is indicated in the following order for   Origin of the material examined -Chapadão do Céu-GO (B): Myrtaceae sp. 1: I-10 (1); Myrtaceae sp. 3: X-09 (2).

DISCUSSION
Amblyseius neochiapensis and N. tunus were found on a large number of hosts (24 and 20 host species, respectively). However, Lofego and Moraes (2006) classified these mites as infrequent due to the small number of hosts on which they were found, in the periphery of the Cerrado, in São Paulo. Demite et al. (2009) also found N. tunus in peripheral areas of Cerrado at Mato Grosso, but associated with only one host. The differences between the results of this study and the above cited may be related to geographical position, climatic differences or floristic composition among fragments, which may alter species composition from one region to another (Felfili et al. 2002;Reis et al. 2000;Bridgewater et al. 2004).
According to some authors, leaf structures are important factors determining the composition of phytoseiid species in plants (Beard and Walter 2001;Hadamar et al. 1986;Moraes et al. 1993). Our data partially agree with this statement, since 27 % of mite species did not occur in more than one host, suggesting a very specific mite-host relationship. Furthermore, some species groups had preference for some kind of leaf structure. For example, species of the genus Phytoseius almost always occurred in plants with trichomes, such as C. brasiliense and X. aromatica.
However, other species did not show any preference at all for specific host. Amblyseius neochiapensis, E. citrifolius, and N. tunus occurred in plants with diverse leaf structures, such as smooth leaves of M. albicans, leaves with trichomes of Q. grandiflora and leaves with domatia of S. adstringens. Lofego and Moraes (2006) found similar results in Cerrado-Atlantic forest transition area, in which Amblyseius acalyphus Denmark and Muma, 1973 occurred in host with different foliar structures, whereas the majority of Transeius belotti (Moraes and Mesa), 1988 occurred in only one host. Therefore, we conclude that the occurrence of some phytoseiid in the Cerrado may be restricted to some plant species. This restricted distribution is likely related to foliar structures or feeding resources available in the hosts. Nonetheless, other species apparentely are not affected by these characteristics, and occur in a wide variety of plants in the region. Taken together, these results point to the highly complex distribution of phytoseiid mites in the Cerrado and reinforce the importance of the conservation of this biome. Most of species found in this study (17 in total) were previously recorded in Atlantic Forest remnants (Castro and Moraes 2010;Demite et al. 2011;Feres and Moraes 1998;Feres et al. 2005;Zacarias and Moraes 2002). Comparing our results with previous records from peripheral areas of Cerrado (Lofego et al. 2004;Lofego and Moraes 2006;Demite et al. 2009), only 12 species are common between these two areas.
In conclusion, the results suggest that few Phytoseiidae species are endemic from Cerrado, a fact also pointed out by Tixier and Kreiter (2009). However, due to the small number of surveys and the high rate of endemic plants in this biome (Myers et al. 2000), these results show probably only a fraction of the real diversity of phytoseiid in the region, since the richness of animal groups tend to increase as plant diversity increases (Brose 2003;Dennis et al. 1998;MacArthur and MacArthur 1961). Therefore, more surveys should be conducted in the Cerrado, including physiognomies and hosts not considered in this study. This would improve the knowledge about phytoseiid species richness from this Neotropical savannah.