Herpetofauna of Paranapiacaba : expanding our knowledge on a historical region in the Atlantic forest of southeastern Brazil

The largest area of preserved Atlantic forest is located in the southern portion of Brazil. The region of Paranapiacaba is depicted in Brazilian zoological studies as one of the first and most intensely sampled areas of the state of Sao Paulo.We provide a concise list of reptiles and amphibians from the Paranapiacaba Municipal Park. It represents the first comprehensive survey of the group in the area. We recorded 136 species of reptiles and amphibians from field surveys, museum collections and the literature. The anuran diversity of Paranapiacaba is greater than that of Estacao Ecologica de Boraceia, which has been considered the most distinctive areas in Sao Paulo in terms of amphibian diversity. The rich history of herpetological research in the region, including the occurrence of the two most threatened species in Brazil, converts the area to an important conservation landmark for the Brazilian herpetofauna.

The Atlantic Forest Domain (sensu AB 'SABER 2003) is formed by a mosaic of environments within several biogeographic subunits (CÂMARA 2003, SILVA & CASTELETI 2005, CARNAVAL et al. 2009).It exhibits great latitudinal, altitudinal and longitudinal variation, featuring a unique diversity of flora and fauna, and high levels of endemism (MYERS et al. 2000, LAURANCE 2009, RIBEIRO et al. 2011).The Atlantic Forest now is 11.7% of its original extension, and more than half of its remnants are restricted to the Serra do Mar formation, in the states of São Paulo and Paraná (GALINDO-LEAL & CÂMARA 2005, RIBEIRO et al. 2009).
The herpetofauna of the state of São Paulo is extremely diverse, representing approximately 30% of all Brazilian diversity (ROSSA-FERES et al. 2011, ZAHER et al. 2011).The amphibian fauna of the Serra do Mar, Serra da Mantiqueira and Serra de Paranapiacaba is clearly the richest of the state (GARCIA et al. 2009a).Moreover, most of the Squamata reptiles of São Paulo are endemic to elevated areas within the Serra do Mar range (MARQUES et al. 2004, RODRIGUES 2005, ROSSA-FERES et al. 2008).
The geomorphological formation knwon as "Serra de Paranapiacaba" encompasses an interiorized portion of the Serra do Mar, and part of the Atlantic plateau along the southern portion of the state of São Paulo (GUIX et al. 2000, CRUZ & FEIO 2007).The district known as "Alto da Serra", "Alto da Serra de Paranapiacaba", "Alto da Serra de Cubatão", or simply as "Paranapiacaba," is located at the edge of the Atlantic Plateau between the metropolitan area of São Paulo and its coast,and it is part of the Serra de Paranapiacaba (BOKERMANN 1966).This district was founded in the nineteenth century, along with the construction of the first railroad of the state (Santos-Jundiaí railway) (LOPES & KIRIZAWA 2009).The region was also the setting point of one of the first conservation units of Brazil, the Reserva Biológica do Alto da Serra de Paranapiacaba (REBIO) (LOPES & KIRIZAWA 2009).
Although Paranapiacaba still represents one of the most intensively surveyed localities of the state of São Paulo, being especially relevant for anurans (DIXO & VERDADE 2006), its herpetofauna remains poorly known.Additionally, only a small fraction of the information available in scientific collections, and which deals with specific aspects of the anuran community, has been published (e.g., BOKERMANN 1968, SAZIMA & BOKERMANN 1978, POMBAL & CRUZ 1999, PATTO & PIE 2001, OLIVEIRA et al. 2008, PMSA 2008, VERDADE et al. 2009).
Here, we provide a concise list of the reptiles and amphibians from the Parque Natural Municipal Nascentes de Paranapiacaba.Our study represents the first comprehensive herpetofaunal survey for the region.

MATERIAL AND METHODS
The present work was conducted at the Parque Natural Municipal Nascentes de Paranapiacaba (PNMNP), municipality of Santo André.The park occupies 426 ha of mountainous Atlantic Forest formation, between 23°47'4.9"S-23°45'27.9"S and 46°18'19.4"W-46°17'7.8"W,at an altitude range of 850 to approximately 1174 m (PMSA 2008).The park surrounds the Paranapiacaba district, adjoining the REBIO, and also the locality known as Campo Grande da Serra, which corresponds to the deactivated Campo Grande train station (BOKERMANN 1966).It is also contiguous with the Parque Estadual da Serra do Mar, and is located at the end of the Mogi river valley (Fig. 1).Historical factors, for instance extraction of natural resources for railroad maintenance and for the construction of the village of Paranapiacaba, led to a predominantly altered vegetational landscape.
The climate of the PNMNP is classified as altitudinal tropical and mesothermic super humid.Mean annual temperatures vary little, ranging from 14° or 15°C to 21° or 22°C.The region is within the greatest cell of precipitation in Brazil (3,300 mm mean average).There is no hydric deficit in the PNMNP, and humidity is high throughout the year.The high humidity, combined with the orographic effect coming from the Serra do Mar, produce a fog that is distinctive (GUTJAHR & TAVARES 2009).
Three sets of pitfall traps were used.Each was composed of two stations with ten 100-liter plastic recipients, totaling 60 recipients installed.Every recipient was connected to another by eight m of drift-fences.The fences were 1 m high, and were buried approximately 20 cm into the ground, passing through the center of each recipient.The pitfall sites were located at least 300 m apart from each other, in order to maintain the spatial independence of the sample units.The sites were selected as to encompass different variables, such as arboreal physiognomy and the influence of water bodies.Pitfall traps were inspected daily totaling 111 days of opened traps (6,600 recipients-days).
In addition to pitfall traps, visual and audio surveys were performed at eight different sites within the PNMNP and also in adjacent areas.Sampling was carried out preferentially during the evening and at night, by a team of two to four researchers.Anuran vocalization was also registered with a portable digital recorder (Sony ICD-P630F) to assist species identification.
Data collected by other means than the methods mentioned above were classified as occasional encounters, and pertain exclusively to random species records.Hence, this sample effort was not considered for data analysis.We collected the following information for each specimen registered by us: location, method of collection, date and time of collection, type of environment, activity patterns, and approximate climate conditions.All specimens collected were deposited in the Coleção Célio F.B. Haddad, Universidade Estadual Paulista "Júlio de Mesquita Filho", Rio Claro, São Paulo (amphibians) and Coleção Herpetológica do Museu de Zoologia da Universidade de São Paulo, São Paulo (amphibians and reptiles).
We used the sampling data to analyze species composition and richness (number of species).Relative abundance was estimated as a percentage of the number of individuals from each species over the total number of individuals registered.Relative abundance was calculated exclusively based on data collected by pitfall traps.Efficiency of the sampling method was evaluated using species rarefaction curves (COLWELL & CODDINGTON 1994, GOTELLI & COLWELL 2001, THOMPSON et al. 2003), with 95% confidence interval and 1000 aleatorizations, using the software EstimateS 8.20 (COLWELL 2009).Richness was assessed through a non-parametric first order Jackknife index (HELTSCHE & FORRESTER 1983, HELLMANN & FOWLER 1999).Sample effort was considered as the number of open pitfall traps per sampling day.Specimens collected outside the sampling period were not considered in the analysis.
In order to define patterns of species distribution in Paranapiacaba, we allocated each taxon into one of the following five categories: 1) species distribution restricted to the Paranapiacaba region, also encompassing the Estação Ecológica de Boracéia (23°38'S and 45°52'W) (HEYER et al. 1990) (Par); 2) species endemic to the Serra do Mar range within the state of São Paulo (S); 3) species distributed in the entire Serra do Mar ZOOLOGIA 31 (2): 126-146, April, 2014 formation, from south to southeastern Brazil (Se); 4) species broadly distributed within the Atlantic forest (southeastern, northeastern and/or southern Brazil) (Af); 5) species broadly distributed in Brazil, with records outside the Atlantic Forest Domain (Br).
Most species are distributed on mountainous areas of the Atlantic forest from south and southeastern Brazil (60.5%).Eight of those are strictly endemic to the Paranapiacaba region (10.5%) and 14 (18.4%) are endemic to the Serra do Mar formation of the state of São Paulo.The remaining species are broadly distributed throughout the Atlantic Forest (23.7%), and 15.8% of the registred anurans occur in more than one biome within Brazil (Table I).
Species rarefaction curves from pitfall traps data stabilized for sampled amphibians (Fig. 62).The jackknife index calculated for anurans (Jack1 = 17 ± 0) corresponded fairly to the number of observed species.

Reptiles
We recorded fifty six species of reptiles for the PNMNP and the Paranapiacaba region (Table II).Of these, 39 were snakes classified in four families: Colubridae (six species), Dipsadidae (30), Elapidae (one), Viperidae (one), and Tropidophiidae (one); and 13 were lizards from five families: Anguidae (two species), Gekkonidae (one species), Gymnophtalmidae (five), Leiosauridae Table I.Species composition of anurans from Parque Municipal Nascentes de Paranapicaba (PNMNP), Santo André, São Paulo.Type of data collected: 1.Primary data from field surveys during the study period; 2. secondary data gathered from herpetological collection catalogs, 3. secondary data from Verdade et al. (2009)  A significant proportion of reptile species were broadly distributed throughout the Atlantic Forest Domain (41.1%).One snake is endemic to São Paulo (Atractus serranus) and 10.7% of the species are restricted to the Serra do Mar formation of southeastern Brazil.Most reptile species recorded in this study, (four), and Teiidae (one).There were also two amphisbaenids; and two chelonians .Secondary data was responsible for 29 records.Field surveys encountered 18 snakes, 10 lizards and two chelonians, seven of which represent new records for the area according to collection records (see MARQUES 2009) (Table II).however, also occur outside the Atlantic forest (46.4%), in the Cerrado or the Amazon (Table II).
The species accumulation curve for the reptiles sampled by pitfall traps could not be stabilized, even though it showed a slight tendency for stabilization (Fig. 62).The calculated richness index (Jack1 = 14.9 ± 1.9) suggests that the absolute reptile diversity has not yet been sampled by this method.
Simultaneously, 27 species of reptiles were added to previously published data for the area: 16 snakes, seven lizards, two amphisbeanids, and two chelonians (MARQUES 2009).Contrary to MARQUES (2009), we did not consider the following species: Leposternon microcephalum, Kentropyx paulensis, Chironius fuscus, C. foveatus, Pseudoboa serrana, Siphlophis pulcher, and Bothrops jararacussu.Even though their distribution range may include Paranapiacaba, there were no collection records corroborating their presence in the region.However, this may be the result of life habitat, altitudinal range or collection bias, and it is possible that these species will be recorded for the region in the near future.The only exception is the lizard Kentropyx paulensis, listed in MARQUES ( 2009), which is an endemic of Cerrado formations (VALDUJO et al. 2009).Some records were not included in our list, such as Erythrolamprus typhlus and Thamnodynastes strigatus.Those snakes have been registered for nearby localities.However, as we were not able to match specimens to a precise geographic coordinate, their presence in the PNMNP could not be confirmed.On the other hand, species such as Amphisbaena alba, A. dubia, Atractus reticulatus, Erythrolamprus jaegeri, Oxyrhopus guibei, and Sibynomorphus mikanii were included as probable occurrences.Although they are not believed to occur within the PNMNP boundaries, their distribution is expected for the macro region where Paranapiacaba is inserted (MARQUES et al. 2009), with records for Santo André and adjacent municipalities.One single collection record was catalogued for the snake Echinanthera melanostigma (IB 1640).However, as the authors could not confirm the identification of the specimen, this species was not included herein.
According to the species accumulation curve, the pitfall trap method was efficient in sampling leaf litter anurans.Reptile diversity was not fully sampled by this method, and we expect that an increased sampling effort would have collected a number of additional species.
Altitudinal factors play an important role in the species composition of the Atlantic Forest Domain (HEYER et al. 1990, VASCONCELOS et al. 2010), and altitude seems to be a limiting factor for some species at Paranapiacaba, for instance the snakes Bothrops jararacussu and Chironius laevicollis (MARQUES 2009).Alternatively, some species are restricted to elevated areas, for example Heterodactylus imbricatus and Atractus serranus.On the other hand, species normally abundant within the Serra do Mar range were scarce in the PNMNP.That appeared to be the case for Hylodes asper, otherwise very common in nearby lo-

Species composition
The amphibian diversity recorded for Paranapiacaba (80 ssp.) in this work is higher than the diversity known for the Estação Ecológica de Boracéia (67 ssp.), which has been considered the most distinctive area in São Paulo in terms of amphibian diversity (HEYER et al. 1990, ARAÚJO et al. 2009a).The diversity of reptiles (56 ssp) was also very representative for the state (CONDEZ et al. 2009, FORLANI et al. 2010).
We recorded 15 anuran species for Paranapiacaba that were not present in previous records: Ischnocnema cf.spanios, Table II.Species composition of reptiles from Parque Municipal Nascentes de Paranapicaba, Santo André, São Paulo (PNMNP).Type of data collected: 1.Primary data from field surveys during the study period, 2. secondary data gathered from herpetological collection catalogs, 3. secondary data from Marques (2009)  Variables such as high humidity and the almost continuous rainy season certainly play an important role in species assemblages in Paranapiacaba, especially for anurans.For instance, the substantial diversity (11 species) of anurans that have direct development (Brachycephalidae) could account for these favorable climatic conditions.The dominance of brachycephalids appears to be a recurrent pattern among Neotropical litter frog faunas (SCOTT 1976, DUELLMAN 1988, HEINEN 1992, GIARETTA et al. 1997, 1999, ROCHA et al. 2001, 2007, VAN SLUYS et al. 2007, ALMEIDA-GOMES et al. 2008, SANTOS-PEREIRA et al. 2011, SIQUEIRA et al. 2011), and it is possibly associated with higher humidity values in cloud forests of higher altitudes (GIARETTA et al. 1999).
The relationship between altitudinal gradients and richness and equitativity of reptile species in the Neotropical region has been debated (SCOTT 1976, FAUTH 1989, HOFER & BERSIER 2001, DIXO & VERDADE 2006, VASCONCELOS et al. 2010) and not fully understood for the Atlantic forest.However, there is an apparent pattern for Atlantic forest localities: frog density seems to be higher in leaf litter assemblages collected in higher altitudes.This pattern has been corroborated for a few localities (GIARETTA et al. 1999, VASCONCELOS et al. 2010, SIQUEIRA et al. 2011).In the present work, the overall abundance of leaf litter anurans was higher than in lower altitudes (ROCHA et al. 2001, 2007, VAN SLUYS et al. 2007).
Paranapiacaba is the type locality of 23 species of anurans.Historical exploration of the area and the considerable sampling effort made by zoologists during decades could account for such a diverse species catalog.Nevertheless, as one examines anuran diversity in adjacent areas (Estação Ecológica de Boracéia and Parque das Neblinas) (HEYER et al. 1990, BERNECK et al. 2008, 2013, GARCIA et al. 2009b), a gradient of high diversity and similar species composition can be observed throughout this portion of the Serra do Mar.

Taxonomic comments
The taxonomic status of several species that occur in Paranapiacaba has been intricate (GIARETTA & CASTANHO 1990, POMBAL & CRUZ 1999).Even though we do not intend to provide a taxonomic review of each group, we do consider necessary to elucidate a few taxonomic aspects of some species.VERDADE et al. (2009) mentioned Brachycephalus ephippium for the REBIO and Paranapiacaba.However, through the analysis of collected specimens and their life habits, we believe that this taxon represents an undescribed species, differentiated from B. ephippium by its darker dorsal coloration and ossification pattern.Brachycephalus sp. is currently being described by Paulo Garcia and collaborators for the Parque das Neblinas, municipality of Mogi das Cruzes.The species seems to be restricted to the most superficial leaf litter layers, and active individuals were always found hidden under leaves, differing from the typical pattern of B. ephippium (POMBAL et al. 1994).
Another new brachycephalid sampled, Ischnocnema sp.(aff.guentheri), can be distinguished from the congeneric I. guentheri by a suite of characters that include vocalization and   However, only a more representative sampling of the population and a broader taxonomic revision including the large complex under the name of I. guentheri, will provide sufficient grounds to diagnose this new species.
The Ischnocnema lactea group represents a complex of 14 species distributed in the Atlantic forest, from Rio de Janeiro to Santa Catarina (HEDGES et al. 2008), andBahia (CANEDO &PIMENTA 2010).During our field surveys, we collected three species from this complex: Ischnocnema sp.(gr.lactea), I. cf.spanios, and I. nigriventris.Ischnocnema lactea is known exclusively from the holotype, and its precarious conservation status is responsible for misleading identifications (CANEDO & PIMENTA 2010).The specimens collected by us match the original description in characteristics such as enhanced apical discs with a bifid pattern, absence of webbing, and reduced first hand disc (MIRANDA-RIBEIRO 1926).However, we would need the species to be re-described to arrive at a reliable identification.Therefore, we maintain the taxonomic status of the population found in PNMNP as Ischnocnema sp., pending further revision of the group.
Ischnocnema cf.spanios has been considered restricted to its type locality (Estação Ecológica de Boracéia, São Paulo), but similar specimens were collected from the Atlantic plateau as well, for instance the Parque Estadual de Carlos Botelho and adjacent regions (FORLANI et al. 2010).Still, it is possible that such records refer to more than one taxonomically distinct entity under one name, and a reassessment of the group is necessary for further considerations.
Ischnocnema nigriventris is a small species described for the locality known as "Alto da Serra" (Paranapiacaba) ZOOLOGIA 31 (2): 126-146, April, 2014(BOKERMANN 1966).The species is underrepresented in scientific collections and, because it had an ill-defined type series there used to be a lot of confusion about its identification until it was recently redescribed (BERNECK et al. 2013).Based on the pattern of dark and light blotches on the venter and inner portions of the thighs, enlarged distal discs on external fingers and toes, dorsal and palpebrous tubercles, and a prominent calcar tubercle (HEYER 1985, HEYER et al. 1990, POMBAL & CRUZ 1999, BERNECK et al. 2013) we recognize the population collected in the present study as Ischnocnema nigriventris.
The bufonid Dendrophryniscus cf.brevipollicatus resembles its sympatric D. brevipollicatus, except for the presence of a white rostral band in D. leucomystax.However, morphological and molecular studies on this genus are required for a better taxonomic resolution of the complex.
Scinax perpusillus identified by us might actually correspond to more than one nominal species (HEYER et al. 1990, FAIVOVICH et al. 2005).Therefore, sampled specimens were listed as S. cf.perpusillus.VERDADE et al. (2009) mentioned the following anurans for the Paranapiacaba region: Leptodactylus gracilis, Rhinella hoogmoedi, Scinax fuscomarginatus, and Gastrotheca microdiscus.No additional record was made for R. hoogmoedi and S. fuscomarginatus, therefore their occurrence there cannot be corroborated.Leptodactylus gracilis is known to occur in southeastern and southern Brazil.However, its distributional range does not include the state of São Paulo (ARAÚJO et al. 2009b, ROSSA-FERES et al. 2011).Specimens deposited in the MZUSP correspond to Leptodactylus jolyi, and therefore, L. jolyi seems to be the taxon that occurs in Paranapiacaba.On the other hand, the only Gastrotheca sampled for the PNMNP corresponds to the description available for G. fulvorufa (CARAMASCHI & RODRIGUES 2007, IZECKSOHN & CARVALHO-E-SILVA 2008).The latter was described for Paranapiacaba, and was recently removed from synonymy with G. microdiscus (CARAMASCHI & RODRIGUES 2007).
Among hylodids, Hylodes spp.and Crossodactylus spp are problematic.The taxonomy of Crossodactylus is confusing, and morphological diagnostic characters are misleading (HEYER et al. 1990).One specimen resembling a young Crossodactylus was collected in a neighboring area, on the Serra do Mar slope.We were not able to provide a precise identification for that specimen.Collection records, however, confirm at least two species occurring in the region, C. dispar and C. gaudichaudii (see discussion below).Hence, our records for this genus are based solely on secondary data.Hylodes sp.(aff.phyllodes) is closely related to H. phyllodes.The large size of the male (CRC: 33.7-35.9mm), contrasting with the male of H. phyllodes (CRC: 27.6-30.0mm), as well as the coloration pattern of specimens, might be diagnostic for this new species.However, diagnostic characters still need to be established.
Difficulties identifying cryptic species of the Leptodactylus marmoratus complex (sensu HEYER 1973) are common (POMBAL & GORDO 2004, BERNECK et al. 2008).The epithet "marmoratus" does not seem to apply to the populations from Paranapiacaba and Boracéia (BERNECK et al. 2008), and it is possible that more than one species occur at the PNMNP.
Species boundaries are not very clear within the lizard genus Ophiodes.The specimens collected from the PNNMP correspond to Ophiodes fragilis (MÁRCIO BORGES-MARTINS, pers.com.), and we follow that classification until a proper revision is available.
The snake Dipsas indica registered by us for Paranapiacaba corresponds to D. indica bucephala (PETERS 1960).Recent taxonomic changes in the D. indica group have resulted in uncertainty about the status of this subspecies (HARVEY & EMBERT 2008).Therefore, we refrain from unsing the subspecies name here.
The Thamnodynastes recorded for the Paranapiacaba region corresponds to Thamnodynastes "sp.1" of FRANCO & FERREIRA (2002).According to the authors, this taxon corresponds to the one described by MIKAN (1820) as Coluber nattereri.Menawhile, as there is no formal redescription available to legitimize the name, the taxon is considered as Thamnodynastes "sp.1".

Conservational status and final comments
The importance of Paranapiacaba for the conservation of Brazilian amphibians is already well established.It is the single known locality for the only anuran considered to be extinct from Brazil (Phrynomedusa fimbriata), and for another critically endangered species (Hypsiboas cymbalum) (HADDAD 2008, GARCIA et al. 2009a).This alone would be sufficient to establish the importance of the region as a primordial site for conservation.However, declines in anuran populations make the region even more important.
A great deal of attention was given in the 1980's, especially for the REBIO, when research revealed the effects of pollution emerging from the petrochemical pole of Cubatão (DOMINGOS et al. 2009, LOPES et al. 2009).The original vegetation of the area has been altered as a consequence of pollution (DOMINGOS et al. 2000, MORAES et al. 2002, 2003), and that has impacted several groups of animals, including amphibians (LOPES et al. 2009, DOMINGOS et al. 2009, SUGIYAMA et al. 2009, VERDADE et al. 2011, 2012).According to VERDADE et al. (2009VERDADE et al. ( , 2011)), the anurofauna of the REBIO is depauperate, if not in terms of richness, at least in the number of individuals of otherwise common species, for instance Ischnocnema parva.Additionally, the authors indicated that the vegetation physiognomy and the water pH of the REBIO are altered, contrasting with the unaltered pH values obtained from the PNMNP.It has been postulated that the mountainous portion of the REBIO is under direct influence of the wind current emerging from the petrochemical pole of Cubatão, whereas the PNMNP is apparently more protected from it (FERREIRA et al. 2009, VERDADE et al. 2009).These differences are reflected on the anurofauna: species are clearly more abundant in the PNMNP (VERDADE et al. 2009).Unfortunately, we were not able to sample the REBIO, and further comparisons are hindered until standardized sampling is performed simultaneously in the two conservation units.
The anuran populations of some species are likely declining in southeastern Brazil (HEYER et al. 1988, WEYGOLDT 1989, ETEROVICK et al. 2005, VERDADE et al. 2011) Much of the information in the literature is merely speculative.A great diversity of factors might contribute to apparent reductions in population size.Life habitats associated with mountainous streams may be related to environmental sensibility of some of those supposedly declining anurans (HEYER et al. 1988, STUART et al. 2004, LIPS et al. 2005, VERDADE et al. 2011).On the other hand, various factors could influence such alterations, such as low population density, stochastic effects caused by natural selection, distributional patterns within the Serra do Mar range, collection biases, or even natural population fluctuations (HEYER et al. 1988, HEYER & MAXSON 1983, MAGNUSSON et al. 1999).
Nevertheless, even though we strongly recommend caution when evaluating patterns of population decline, we stress that species that now seem to be rarer were frequently collected during campaigns between 1950 and 1980, as documented in collection catalogs from MZUSP and ZUEC (see Appendix 1).Human interference, especially on adjacent areas of the Paranapiacaba village, appears to have played an important role on the current local species composition and loss of biodiversity.As a result from the operation of the railroad, several forested areas have been completely suppressed.Our inability to delimitating the exact path of past collecting expeditions, due to the lack of geographic coordinates for them, makes it difficult to reach more definite conclusions.Regardless, the results presented herein are relevant enough to indicate the need for future studies.

Figure 1 .
Figure 1.Geographic position of the study area within the state of São Paulo and the municipality of Santo André, including adjacent conservation units.PNMNP: Parque Natural Municipal Nascentes de Paranapiacaba, PESM: Parque Estadual da Serra do Mar, REBIO: Reserva Biológica do Alto da Serra de Paranapiacaba (Adapted from PMSA 2008).Red dots, numbered from 1 to 13, represent localities of visual survey samples, black dots represent localities of pitfall trap samples (P1, P2 and P3).
. Localities of occurrence at site areas: a. Paranapiacaba or Alto da Serra, b.PNMNP (study area), c: Reserva Biológica do Alto da Serra de Paranapicaba, d: Campo Grande da Serra, e: Santo André municipality.Geographical distribution: Par: species exclusive to Paranapiacaba region, S: endemic for the Serra do Mar from São Paulo state, Se: Serra do Mar formation, south and southeastern Brazil, Af: broad distribution within the Atlantic forest, Br: broad distribution in Brazil.* Represents "Paranapiacaba" or "Alto da Serra" as type locality; ** Paranapiacaba as type locality of one of the species synonyms; ***Campo Grande as type locality.Thoropa taophora*(Miranda-Ribeiro, 1923)

Figure 62 .
Figure 62.Species accumulation curves (solid lines) and richness index (first order Jackknife) for amphibians (black) and reptiles (grey) sampled by pitfall traps during field surveys in the Parque Natural Municipal Nascentes de Paranapiacaba, Santo André, São Paulo.Dashed lines represent respective standard deviation.
. Localities of occurrence at site areas: a. Paranapiacaba or Alto da Serra, b.PNMNP (study area), c: Reserva Biológica do Alto da Serra de Paranapicaba, d: Campo Grande da Serra, e: Santo André and adjacent municipalities.Geographical distribution: Par: species exclusive to Paranapiacaba region, S: endemic for the Serra do Mar from São Paulo state, Se: Serra do Mar formation, south and southeastern Brazil, Af: broad distribution within the Atlantic forest, Br: broad distribution in Brazil.

Table I .
Continued.