Anuran species composition of Cancão Municipal Natural Park, Municipality of Serra do Navio, Amapá state, Brazil

Abstract In this study, the first survey of anuran species in the Cancão Municipal Natural Park is presented, a protected area of approximately 370 hectares of Amazonian forest located in the northwest center region of the state of Amapá, Brazil. The work was performed during the dry and rainy season, through active visual and auditory survey, totaling 216 man hours of sampling effort. Forty-nine species of anuran amphibians were recorded in the Cancão Municipal Natural Park, including three new records: Hyalinobatrachium iaspidiense, Pristimantis cf. ockendeni, and Scinax garbei. Three species, Hyalinobatrachium iaspidiense, Ameerega pulchripecta, and Anomaloglossus baeobatrachus, are listed as Data Deficient and one is listed as Vulnerable (Atelopus hoogmoedi) according red lists of IUCN. The rarefaction curve cumulative species did not reach an asymptote, indicating that site has potential for species that have not yet been recorded. Nine species were represented by only one individual and were considered rare in the studied environments, eight species were defined as common, and the 32 remaining species were classified as having intermediary abundance. Our data indicated that Cancão Municipal Natural Park contains a considerable portion of the anurans species richness of Amapá state, turn the area into a place of great importance for the conservation of the anurans of the Eastern Amazon.


Introduction
Most of the currently documented amphibian species in Brazil have been discovered during the last forty years ). These new species descriptions, which have occurred at regular rates, are a strong indication that the Brazilian amphibian fauna is poorly known (Peloso 2010). Brazil has the highest diversity of amphibian species on the planet with 1080 species, 1039 of which are anurans, 36 caecilians, and five salamanders (Segalla et al. 2016). According to a recent publication of species list, 308 species of anurans (29.6 % of the species known in Brazil), 18 gymnophionans and five caudates (Hoogmoed and Galatti 2016) are known in the Brazilian Amazon, representing approximately one-third of the total of amphibians recorded for the country (Ávila-Pires et al. 2010, Neckel-Oliveira et al. 2013. This amphibian species richness can be considered underestimated in number and complexity when considering enormous areas of Brazil which have yet to be inventoried, and there are many localities were surveys have been insufficient (Silvano and Segalla 2005). Aditionaly, the political limits and geographic distributions, the existence of cryptic species (Fouquet et al. 2007), sampling gaps due to the concentration of researches in a few areas (Azevedo-Ramos and Galatti 2002), sampling effort used appropriate methods for inventories of amphibians (Miranda et al. 2015) and problems in various taxonomic groups, frustrate attempts to obtain a comprehensive understanding of Brazil amphibians (Silvano and Segalla 2005).
In the Amazonian biome, studies on anurans are concentrated in states of Amazonas, Pará, Rondônia and Acre, other localities in the Brazilian Amazon lack inventories (Azevedo-Ramos and Galatti 2002), a fact observed for the Amapá state. Although be inserted in a region of great interest for conservation and presents gaps knowledge on anurans (Azevedo-Ramos and Galatti 2002), Amapá state is little studied in relation to its anurans fauna (Lima 2008, Queiroz et al. 2011, Pereira-Júnior et al. 2013, Araújo and Costa-Campos 2014, Costa-Campos 2015, Benício and Lima 2017. In this context, the present study aims to provide the list of species of anuran amphibians that occur in the area of the Cancão Municipal Natural Park, municipality of Serra do Navio, state of Amapá, eastern Amazon.

Materials and methods
Study area. Fieldwork was conducted at the Cancão Municipal Natural Park (Figure 1), municipality of Serra do Navio, Amapá state (0.90263°N, 52.00505°W and 0.90858°N, 52.00422°W). The study area covers approximately 370 hectares of primary forest, including terra-firme rainforests, streams, open areas, and treefall gaps. The climate of the region is Equatorial (Am) according Köppen-Geiger classification and the average temperature is 27.6 °C, varying seasonally between 25.8 to 29.0 °C, with annual rainfall approximately 2,850 mm with monsoon period between February and May, when the monthly rainfall is nearly 400 mm (Alvares et al. 2013).
Sampling. Animals were registered during diurnal and nocturnal active visual search and auditory census in different microhabitats used by frogs (Heyer et al. 1994). These methods were conducted by three researchers for three consecutive days from January to December 2013, resulting in a sampling effort of 216 hours/man. A wide variety of environments were surveyed including ponds, brooks, forest interior, temporary ponds, and other water bodies. These environments were sampled in four mainly sites in the Cancão Municipal Natural Park ( Figure 2): Terra firme trail at Cancão forest (0.90275°N, 52.00497°W); River Amapari trail (0.90083°N, 52.01347°W), Treefall gap at stream Cancão 01 (0.91183°N, 52.00205°W) and Treefall gap at Cancão forest 02 (0.91388°N, 51.99977°W).
Data analysis. To analyze the anurans species richness, rarefaction curves of species were constructed based on the number of individuals and number of samples (Gotelli and Colwell 2001) with 1000 randomizations. Species richness estimators Jacknife1 and Bootstrap were used for determine the expected richness of anurans (Colwell 2013). To determine similarities of species compositions amongst habitats sampled, cluster analyses were performed by the UPGMA method, using the modified index of similarity of Jaccard (Clarke 2003). This analysis was performed using ESTIMATES 9.1 (Gotelli and Colwell 2001).
The dominances were represented by Whittaker Diagram, obtained by ranking species, starting with the most abundant, along the x-axis and the logarithm abundances on the y-axis. Rare species were those represented by a single individual (singletons). The other species were classified as having intermediate abundance. The pattern of the species abundance distribution was fitted to the geometric, logarithmic, lognormal, and broken-stick models. Model fit was assessed by the chi square adherence test (Magurran 2011) using the software PAST version 2.17c (Hammer et al. 2001).  The Spearman correlation coefficient analysis was performed to compare climatic conditions (avail able from the NHMET database) during the sampling period with abundance. To check the influence of environmental data on amphibian abundance, multiple re gression analyses were conducted, including data on rain fall, temperature, and humidity as independent variables. The normality of the data was tested with the KOLMOGOROV-SMIRNOV analysis (Zar 1999). Statistical analyses were performed with BIOESTAT 5.3 software (Ayres et al. 2007), using a significance in dex of P < 0.05 for all analyses.
Hyalinobatrachium iaspidiense is known from Brazil, Ecuador, French Guiana, Guyana, Peru, Suriname, Venezuela, and is expected to occur in the Amazonian areas between the Ecuadorian and Peruvian localities and the Guiana region (Castroviejo-Fisher et al. 2011). This record is the first for Amapa and extends the known distribution of the species 1,020 km east from the type locality Quebrada de Jaspe, San Ignacio de Yuraní, Bolívar state, Venezuela (Silva e Silva and Costa-Campos 2016).
Pristimantis cf. ockendeni is distributed throughout the Amazonian basin of Peru, Ecuador, southern Colombia, and Brazil in the states of Acre and Amazonas (Rodrigues et al. 2004). This is the first state record for Amapa, extending the range 986 km NW from the Manaus, Amazonas state (Silva e Silva et al. 2015).  . calcarata 20 B. cinerascens 21 B. dentei 22 B. fasciata 23 B. geographica 24 B. multifasciata. Scinax garbei is known from Ecuador, adjacent Peru, Bolivia, Colombia, and Venezuela (Frost 2018). In Brazilian Amazonia, it has been recorded from Amazonas (França and Venâncio 2010) and Pará states (Ávila-Pires et al. 2010). In this study, we present the first record of the species in the state of Amapá, extending the species distribution in the Brazilian Amazonia by 525 km northward from the two localities in the state of Pará: Rio Xingu and Rio Curuá-Una (Silva e Silva and Costa-Campos 2014).
The frog species richness estimated for the area by Bootstrap and Jack-knife was 54 and 63 species, respectively, and the rarefaction curve cumulative species did not reach an asymptote. We believe that site has potential for species that have not yet been recorded ( Figure 4). The Hylidae was the most species-rich family (17 species), followed by the Leptodactylidae (10) and Craugastoridae (4).
Nine species (Allophryne ruthveni, Rhinella martyi, Pristimantis zimmermanae, Boana calcarata, B. dentei, Osteocephalus oophagus, Scinax garbei, Leptodactylus lineatus and Pipa pipa) were represented by only one individual and were considered rare in  The dendrogram obtained from cluster analysis evidences three major groups: (A) sites located in the treefall gaps, (B) sites in the Terra firme forest, and (C) sites belonging to the Amapari River with temporary ponds. The group (A) is characterized by the higher Boana multifasciata, B. cinerascens and Pithecopus hypochondrialis, species occurring in open areas. For the group (B), the most frequent species were Rhinella margaritifera complex species and Pristimantis chiastonotus. The last group, (C), is characterized by the high frequency of occurrence of Allobates femoralis and Adenomera andreae. The coefficient of cophenetic correlation for the cluster analysis was 0.997 (Figure 7).

Discussion
Our data indicate that Cancão Municipal Natural Park contains a considerable portion of the anurans species richness of Amapá state. The anuran fauna corresponded to 65.7 % of the recorded species for the Tumucumaque Mountains National Park (Lima 2008), 86.9 % of the species found in the River Cajari Extractive Reserve (Queiroz et al. 2011), 60.9 % species recorded during the surveys conducted of the Rio Curiaú Environmental Protection Area , and 90.6 % species of anurans recorded in the Amapá National Forest (Bemício and Lima 2017). This high anurans richness for the Amazonian biome is highly underestimated considering taxonomic problems, recent descriptions of species and taxonomic revisions (Peloso et al. 2014;Vaz-Silva et al. 2015).
The results obtained from the rarefaction curve and the Jack-knife1 and Bootstrap estimators suggest that the species composition is still underestimated, and more longterm studies may reveal the presence of additional species in the area. Future studies should be complemented with combined and/or different approaches in fieldwork, such as the use of pitfall traps for leaf-litter species, increased visual search times (Freitas et al. 2017).
The finding of a large number of species of the families Hylidae and Leptodactylidae was similar to the results of other studies and follows the pattern found in neotropical environments (Segalla et al. 2016  The record of Atelopus hoogmoedi and Ameerega pulchripecta in the area studied is relatively important. Atelopus hoogmoedi is a terrestrial and diurnal species, and is most commonly found at small streams in primary forest (Ouboter and Jairam 2012). The species occurs in the Amazonian lowlands of Colombia, Ecuador, and eastern Peru, to Amazonas, Pará, Amapá (Brazil), and the Guianas (Frost 2018). Ameerega pulchripecta was hard to find, and it has been heard only during less than an hour around dawn and again around twilight (Costa- . Additionally, its distribution appears restricted to Serra do Navio, in the state of Amapá, northeastern Brazilian Amazon. These species are classified as vulnerable and data deficient by the IUCN due to their areas of occurrence, status and little known ecological requirements (IUCN 2017).
The cluster analysis of the anuran assemblages generated three groups. The group A and B showed a more differentiated assemblage. Group C are located on the right bank of the Amapari River, and presented high values of abundance and richness. The results can be attributed to the similar characteristics between the sites. The main hypotheses proposed to explain barrier formation separating populations and causing the differentiation of species in Amazonia during the course of geological history are based on different factors (Haffer 2008). According the river hypothesis, rivers may play a major role in creating and maintaining high levels of spatial separation of populations (Vaz-Silva et al. 2015).

Conclusions
The results of the present study thus provide new data on geographic distribution of species showed three new records of the Brazilian Amazonian and important insights into the diversity of amphibians in the northern Brazil. The high amphibian richness recorded in this study for the eastern Amazon, combined with the presence of populations of Data Deficient or Vulnerable species, contributes to the knowledge on species, reinforcing the importance of the Cancão Municipal Natural Park for the conservation of anurans species.