The Muscidae (Diptera) from the Atlantic Forest of Serra de Paranapiacaba, southestern Brazil

A first survey of Muscidae in the State of São Paulo (Southeastern Brazil) is presented here with a one‐year of Malaise trap collecting from August 2010 to July 2011 at the Biological Reserve Alto da Serra de Paranapiacaba (23°46′00′′‐23°47′10′′S, 46°18′20′′‐46°20′40′′W, 750‐891 m of altitude). A total of 1,284 individuals of muscids were collected, and 15 genera and 39 species of Muscidae were identified. So far, only one muscid species had been recorded to the Reserve, which now has its Muscidae diversity increased to 40 species. Thirteen species are new records for the State of São Paulo. With this, the number of species of Muscidae species known to occur in the State of São Paulo is increased to 169. The interval between November and February was higher in number of individuals and number of species. Muscidae presented a seasonal pattern, with more abundance and diversity in that interval. The study area is covered by secondary forest and very close to São Paulo metropolitan area, and the composition of the fauna of Muscidae signalizes this environment changing and anthropic stress with nine species with synanthropic habits, two of them are typically synanthropic species.


Study area and collection
This survey took place in the Biological Reserve Alto da Serra de Paranapiacaba ( Fig. 1)  in the municipality of Santo André, State of São Paulo, Brazil. The reserve is situated slightly under the Capricorn Tropic (23°27′S), in a transition area between the Tropical and the Subtropical climates zones (Gutjahr & Tavares, 2009). According to Köppen's classification, the climate type is Cwa, temperate and humid, with dry winter and hot summer. The mean temperature at summer (December-March) is 20℃, and 14℃ on winter (June-September), with temperature range of 6℃ (Gutjahr & Tavares, 2009). The area is predominantly covered with secondary Atlantic Forest under different successional stages, and that altered flora is due to historical anthropic intervention (Sugiyama et al., 2009), as this area is very close to São Paulo metropolitan area.

Identification
The specimens were recognized until the genera and species level, using identification keys in . When necessary, taxonomic revisions were consulted, containing updated identification keys and descriptions (e.g., Pereira-Colavite & Carvalho, 2012;Nihei & Carvalho, 2007; among others). All studied material was pinned and, in most cases, males and/or females were dissected to study their terminalia, which were compared to drawings and descriptions in the literature. The dissected terminalia were cleared with 10% KOH solution, then neutralized with acetic acid solution and washed with water; after examination, they were stored in microvials with glycerine and pinned with the respective specimen. The material was deposited at the Museu de Zoologia, Universidade de São Paulo, São Paulo (MZSP).

Data analyses
Quantitative data were analyzed by calculating Shannon diversity and Berger-Parker dominance indexes (Magurran, 1988) to compare the diversity and proportional abundance of species among the four stations of the year. These calculations were performed in the software DivEs v2.0 (Rodrigues, 2005). To analyze the seasonality abundance in the four seasons of the year, we conducted the Kruskal-Wallis statistical test for non-parametric samples using the software Past version 2.17c (Hammer et al., 2001). We have applied a Kruskal-Wallis post-test called Bonferroni Correction, used to reduce the chances of obtaining false-positive results (type I errors) when several pairwise tests are made for one dataset (Nakagawa, 2004).
The estimation of the number of species and extrapolation of this variable and their respective confidence in-tervals (95%) as a function of abundance was performed in R using the statistical package iNEXT (Hsieh et al., 2016) based on the methodologies proposed by Colwell et al. (2012) and Chao et al. (2014).
Collector curves were obtained for the Biological Reserve Alto da Serra de Paranapiacaba and for each seasonal season within the same area.
Quantitative data analysis reveals that the interval between November and February had both more individuals ( Fig. 2) and more species (Fig. 3) collected. Those three-month period have a sequence of 60-55-56 individuals, while other months reached up to 36 individuals. For species number, we had a variation of 14 to 18 species. Shannon diversity index confirms summer with the highest diversity, followed by spring, autumn and winter (Table 2). On the other hand, Berger-Parker dominance shows summer with the lowest rate, followed by spring and autumn, and with winter with the highest rate. In Berger-Parker dominance, the lower the index the higher the dominance of such taxon ( Table 2).
The estimated values of Chao for the richness of muscid species in the Biological Reserve was 65.009 ± 12.198. The Chao estimator is accurate in its lower limit and based on it, we can infer that about 52 species are expected in the area, seven more than the observed richness. This value can reach 105 species, which is an overestimate due to the 14 singletons (species with only one individual) (31.11%) ( Table 3). There is also no trend towards stabilization of the collector curve (Fig. 4a), corroborating Chao's analysis for the Biological Reserve.
Estimates of Muscidae richness for each season of the year, as well as their extrapolation, reveal interesting patterns of sampling effort for this conservation unit. In Table 3 we can see that there are higher values in the summer than in any other season, without a trend of sta-bilization in the collector curve for summer, spring and autumn, while in the winter the collections were sufficient (Fig. 4b).
The seasonality analysis using Kruskal-Wallis statistical test (Table 4) showed a seasonal pattern for the abundance of Muscidae throughout the year, with significative difference between the number of individuals distributed in the stations. The difference among the stations found at Kruskal-Wallis test was due to the abundance of individuals in the summer when compared to the winter and fall seasons. Furthermore, the circular analysis (Fig. 5) supported the seasonal variation of Muscidae indicating a clustered distribution (r = 0,913).

DISCUSSION
In the Biological Reserve Alto da Serra de Paranapiacaba, only one muscid species, Philornis mima (Townsend, 1927), had been recorded so far (Townsend, 1927;Lowenberg-Neto & Carvalho, 2013). With the 39 species found in the present survey, the Muscidae diversity of the Reserve is increased to a total of 40 species.
Malaise traps are efficient in collecting very active flying insects that can be intercepted by its tents, especially Hymenoptera and Diptera (Campos et al., 2000). In tropical rain forests of the Neotropical Region, Diptera is the most collected order when using Malaise traps (Dutra & Marinoni, 1994;Brown, 2005     When considering the main surveys of Muscidae for southern Brazil (Rodríguez-Fernández et al., 2006;Krüger et al., 2010;Zafalon-Silva et al., 2018), we observe that the richness in this family, disregarding the Coenosiinae, is 109 species for Paraná State (Rodríguez-Fernández et al., 2006), 59 species for localities in the grasslands and south coast of Lagoa dos Patos (Krüger et al., 2010) and 63 species for the coastal plain of the Rio Grande do Sul (Zafalon- Silva et al., 2018), values that are 31 to  132% higher than that found for Biological Reserve of Paranapiacaba. It is worth mentioning that the morphospecies richness values are also much higher than those found here, varying from 23% of the fauna in Krüger et al. (2010) to 38% in Rodríguez-Fernández et al. (2006), while in this work only four species were not nominally identified (8.89%). Therefore, when disregarding mor-phospecies, the richness in the Biological Reserve of Paranapiacaba was of 41 species, values close to those found by Krüger et al. (2010) and Zafalon-Silva et al. (2018), 45 and 42 species, respectively. In Paraná, the richness was of 67 species (Rodríguez-Fernández et al., 2006), remaining 63% higher than in Paranapiacaba. The collections carried out at the Biological Reserve of   Nihei, S.S. et al.: Muscidae from Paranapiacaba Pap. Avulsos Zool., 2021;v.61: e20216168 7/12 Paranapiacaba used traps that were placed close to each other (separated by hundreds of meters), and this may explain the much higher values in Paraná, where the collections were carried out in almost the entire State (separated by hundreds of kilometers). Nevertheless, we can estimate a great richness for Muscidae in the Atlantic Forest of São Paulo State, since in a reduced area as the Biological Reserve of Paranapiacaba, the richness found was to similar to that found in areas of much larger collections as Rio Grande do Sul, wherein sampled localities were separated by dozens of kilometers. In fact, the Chao estimate in Table 3 extrapolates more species to be collected at Paranapiacaba and, in this case, we can consider estimates higher than the estimated one. Thus, the curve of the collector is still increasing for this area. Interestingly, the fauna found in the Biological Reserve of Paranapiacaba is more similar to the fauna of Paraná, sharing 31.71% of the species, while only 8% of Paranapiacaba species were also observed in Rio Grande do Sul and in Paraná. However, none of those shared species occurs either in the Coastal Plain of RS or in the Sulrio-grandense Shield. And ten species (24.4%) occur in all these locations and have a wide distribution in Brazil.
The seasonal pattern of Muscidae showed the family individuals more abundant on summer, based on Kruskal-Wallis statistical test (Table 4) and circular analysis (Fig. 5). Previous studies have demonstrated that tropical insect abundance decreases in the dry season (e.g., Pinheiro et al., 2002;Da Silva et al., 2011), but far less than in temperate zones (Wolda, 1988). The pairwise test comparing summer and spring resulted in a statistically relevant difference, but after Bonferroni Correction, this difference was not considered statistically significant (Table 4). On the other hand, Bonferroni tests confirmed seasonal pattern of Muscidae with high abundance in the summer and less in autumn and winter.
Da Silva et al. (2011) andPinheiro et al. (2002) performed seasonal analyses for insect abundance in the Brazilian Cerrado (savannah) and found that most insect orders (Coleoptera, Hymenoptera, Hemiptera, Isoptera, etc.) showed a clustered distribution, being more abundant in the wet season. However, in both studies, the order Diptera was randomly distributed, contrary to the results observed here and focused on Muscidae. This could be related to the groups studied in each case, revealing perhaps the most abundant groups of Diptera might be equally distributed the whole year, while Muscidae might be more vulnerable to environmental and/or biotic conditions that affect their abundance. Moreover, these studies were performed in different biomes other than Atlantic Forest. Traditionally, Cerrado is characterized by a bimodal rainfall distribution, with two well-defined, dry and wet, seasons (Silva et al., 2008), whereas Atlantic Forest is characterized by having high temperature and precipitation during the whole year (IBGE, 1992). However, climatic studies in Paranapiacaba have demonstrated that precipitation suffers a decrease in December and January (Gutjahr & Tavares, 2009), when Muscidae reached its peaks in both abundance and diversity. Unfortunately, there are just a few studies with seasonality of Neotropical muscids. Using fruit-baited traps, Azevedo et al. (2015) found the muscids much less abundant in the rainy season in Barbalha municipality (State of Ceará, northeastern Brazil), characterized by Cerrado (savannah) and Caatinga (semi-arid) vegetations. Results of Costacurta et al. (2003) in three localities of State of Paraná (southern Brazil) also supports the correlation between Muscidae abundance and air humidity: the lower the humidity, the higher the abundance. However, as pointed out by Azevedo & Krüger (2013) for southern Brazilian calliphorids, each species may respond differently (even inversely) to temperature and humidity interactions, therefore, it demands a careful examination on the species other than all the family.
The occurrence of exotic and/or synanthropic species in natural preserved areas has been associated to edge effect by some authors (Lima-Ribeiro, 2008;Ferraz et al., 2010;Barbosa et al., 2014). Furthermore, the complex composition of the Muscidae fauna in the Biological Reserve of Paranapiacaba reveals that while this Reserve still bears most of its fauna with typically forest species, the presence of some typically synanthropic species may signalize that its proximity to São Paulo metropolitan area has imposed an increasing anthropic pressure (Leandro & D'Almeida, 2005