An ornithological inventory in a reforested woodlot in western Paraná state, southern Brazil

This paper presents an ornithological inventory taken between March and December of 2017 in the Brazilian state of Paraná. Although the surroundings contain one relatively well-known location in regard to ornithology, Iguaçu National Park, several other areas merit exploration, among them the Santa Helena Relevant Ecological Interest Area (ARIE-SH). The 1,479 ha ARIE-SH is essentially a large remnant of the Atlantic Forest located adjacent to the city of Santa Helena, Paraná, which commencing in the 1980s has undergone considerable reforestation with both native and exotic species, including fruit species. Prior to censusing bird species and to better characterize the avian community, we conducted a bibliographic review of earlier ornithological studies carried out at ARIE-SH. In addition, we conducted opportunistic and unsystematic observations in nearby locations. We recorded 311 species of birds. During the current inventory, and others conducted in the area since 2016, we recorded six Atlantic Forest endemic species, five species threatened in Paraná, and two globally threatened species. Common species which showed high Index of Frequency in Lists include (in descending order) Basileuterus culicivorus, Leptotila verreauxi, Cnemotriccus fuscatus, Corythopis delalandi, Turdus leucomelas and Arremon flavirostris. Nine species observed while conducting this census (Laterallus exilis, Amazona vinacea, Herpsilochmus longirostris, Campylorhamphus trochilirostris, Casiornis rufus, Campylorhynchus turdinus, Myiothlypis flaveola, Eucometis penicillata and Sporophila palustris) are the first records for this region.


INTRODUCTION
divided the development of ornithology in the state of Paraná into four phases: The "Natterer period" (encompassing all regional ornithological works from the 19 th century), the "Chrostowski period" (beginning of the 20 th century to the 1930s), the "Mayer Period" (between the 1940s and 1960s) and the "Current Period" (commencing in the 1970s). The most outstanding investigations of the state's wild birds occurred during the first three chronological periods.
The first ornithological investigations of Paraná were conducted by the Austrian naturalist Johann Natterer. During his stay in Paraná between September 1820 and May 1821 he visited numerous localities along its coastal plain (Serra do Mar). He also passed through the state's interior lands west of the Serra do Mar collecting zoological material (Straube, 1993). Between 1921 and1924, Polish naturalists Tadeusz Chrostowski and Tadeusz Jaczewski, sent by the Polish Museum of Natural History, searched a wide region from the state's center to its far west, as well as along the Ivaí, Piquiri, and Paraná rivers. During this expedition, they obtained approximately 260 bird taxa, which resulted in the first great ornithological collection for Paraná in this century (Scherer-Neto & Straube, 1995). In 1930, Emil Kaempfer carried out an extensive ornithological expedition in Paraná. He crossed the state from east to west, collecting material from the coastal plain to the western seasonal forests surrounding Foz do Iguaçu (Straube, 2015(Straube, , 2016(Straube, , 2017. Until the mid-20 th century, western Paraná was richly endowed with a variety of bird species, among the most diverse in the Brazilian south and southeast regions (Scherer-Neto et al., 2011). This was largely due to the 180,000 ha of continuous vegetation currently constituting Iguaçu National Park (PNI). This region has been a center of ornithological studies since the 20 th century (Scherer-Neto & Straube, 1995), initiated primarily by naturalists and later by ornithologists investigating the local avifauna community (Straube et al., 2004).
Only 28% of the original Atlantic Forest vegetation cover still stands (Rezende et al., 2018) and, in Brazil, the remaining forest is now fragmented gins. There is a forest opposite ARIE-SH to the west and a reforested strip opposite it to the other compass points, with the water, forest, and reforested strip acting as environmental buffers. Six areas adjacent to ARIE-SH are also non-systematically inventoried (Fig. 1).
Santa Helena County in an area that had experienced rather extreme topographic alteration. Inserted in the hydrographic basins of the Paraná and Iguaçu rivers (western Paraná) located on the left bank (east) of the Paraná river, Santa Helena County (centered at 24°51′37″S, 54°19′58″W) had about one third (260 km²) of its territory flooded in 1982 after damming of the Paraná river with construction of the Itaipu hydroelectric facility (Cavarzere et al., 2020).
In order to shelter and protect regional flora and fauna, as well as to rescue fauna displaced from its habitat due to the formation of this reservoir, the two nation consortia responsible for Itaipu dam (Brazil and Paraguay) officially created seven ecological reserves (Biological Refuges) in the 1980s, two in Brazil (Bela Vista and Santa Helena) and five in Paraguay (Itabó, Limoy, Carapá, Tati Yupi and Yui Rupá); another one, encompassing both countries (Maracaju Binational Reserve) was also created. One of these reserves was initially named Refúgio Biológico Santa Helena (RBSH) and is entirely within Santa Helena County (Fig. 1). The Refuge had an area of 1,479 ha and is isolated from adjacent areas by an excavated channel and a lake, making it an artificial island (Cavarzere et al., 2020).
RBSH was later reclassified as a Relevant Ecological Interest Area, thereby becoming the Santa Helena Relevant Ecological Interest Area (ARIE-SH), a protected area designated for sustainable use within the legal scope of the National System for Protected Areas (SNUC). As such, it is also a government defined Integral Protected Area designed to preserve water and mineral resources, fauna and flora, with sustainable tourism and scientific research being the only permitted human activities (Brasil, 2002). ARIE-SH is also integrated into the Paraná Biodiversity Program, which covers more than 2 million ha and connects semideciduous forests to the Araucaria forests along the Iguaçu and Paraná rivers.
According to Koeppen's classification system (Koeppen, 1948;Alvares et al., 2013), the region that includes ARIE-SH is climatically Subtropical Humid Mesothermal (Cfa), with rainfall distributed uniformly throughout the year (average of 1,650 mm). December and January have the highest average rainfall (150 and 175 mm, respectively), and July and August have the lowest (75 and 100 mm, respectively). The region's average temperature is 22℃, with highs reaching 40℃ in the summer and lows down to -2℃ in the winter; frosts are infrequent (Kliver, 2010).

Flora
ARIE-SH was originally covered by semideciduous seasonal forests (IBGE, 2012). By the mid-1970s, a large part of this area had been clear-cut to accommodate agriculture. Since then, more than 6,680 ha of forests have naturally regenerated (Kliver, 2010;Fundação SOS Mata Atlântica, 2020). In 1974 construction began at the Itaipu Binacional hydroelectric dam, 120 km to the south of Santa Helena. To permit construction of the dam and filling of the lake behind it, the area that includes ARIE-SH was expropriated.
In 1981, reforestation began at what would become ARIE-SH. The reforestation project introduced 18 exotic and 24 native species (eight fruit trees) on 183 approximately 100 × 50 m plots. The most used exotic species were Coffee bush Leucaena leucocephala (Lam.) de Wit, Malabar plum Syzygium cumini (L.) Skeels and Rosewood Tipuana tipu (Benth.) Kuntze. The main native species plots were guabiroba Campomanesia xanthocarpa Mart. Ex O. Berg and Brazilian cherry Eugenia uniflora L. More currently, the areas of vegetation have undergone natural regeneration and are in various stages of succession (Kliver, 2010;Tambarussi et al., 2019).
In her study of ARIE-SH, Kliver (2010) noted the predominance of vegetation in the secondary stage of regeneration and that the original native vegetation could be found in small lowland areas with a predominance of herbaceous substrate in soaked soil. More recently, Tambarussi et al. (2019) determined that approximately 74% of ARIE-SH's area could be classified as an Extensive Use Zone where reforestation with exotics resulted in very poor forest regrowth, 22.4% classified as being a Natural Zone, 3.3% classified as being a Recovery Zone, and only 0.1% classified as an Intangible Zone. When contrasted with areas planted with exotic species, areas that were replanted with native species showed more advanced regeneration, with herbaceous and shrub species occupying the understory, a marked presence of terrestrial ferns, and the richest number of tree species -thus presenting an intermediate successional stage (Tambarussi et al., 2019).
Contrarily, the stands of exotic species show low species diversity in the understory and little natural regener-ation, even after decades of reforestation. Exceptions are the plots with Coffee bush, the seeds of which were used for germination and dispersion studies (Dalmolin et al., 2011). This species is small and leads to an open canopy that allows for the greatest entry of light and the understory's consequent development. With the senescence of the first planted specimens in these plots, space opened for large-scale natural regeneration. Exotics plots with less diversity are those planted with Malabar plum and mango Mangifera indica L., where practically nothing grows in the shaded understory. In these plots, the soil is covered only by thick litter, making it difficult to establish other forms of life. Additionally, allelopathy, systemic with the cultivation of Coffee bush (Scherer et al., 2005) and, especially, Malabar plum, cannot be disregarded as a cause of low regeneration in these areas (Cavarzere et al., 2020). and December of 2017 (Table 1). We used 5-species lists (Mackinnon & Phillips, 1993) when the observer takes notes of all seen or heard species. Each list contains five species, and one species cannot be repeated in the same list but can appear in subsequent ones. We generated Index of Frequency in Lists (IFL) by dividing the number of lists in which a given species occurred by the total number of accumulated lists (Ribon, 2010).

Water birds
We inventoried the water birds linked with ARIE-SH following the Scott & Carbonell (1986) directory who used the Ramsar Convention, which suggests that water birds are those that depend on wetlands, to determine the families and, consequently, the species of aquatic avifauna. Except for time of beginning censuses, we deliberately used Lara's (1994) methodology so that water bird communities could be compared over an interval of approximately 33 years. We determined transects around the ARIE-SH perimeter (32 km) and navigated by motorized boat at a constant speed of about 5 km/h, covering a total of 256 km in all campaigns (Table 1). Inventorying took place between 13:00 and 17:00, starting on the eastern margin, going around ARIE-SH's perimeter, and ending on its western margin. On a rainy day (September 30), we were forced to interrupt the inventory at the halfway point. The same observers (ISQ and VC) conducted censuses using 8 × 20 and 8 × 42 binoculars. Whenever pos-sible, we photographed species and taped their vocalizations with a digital recorder and directional microphone.

Qualitative census
We additionally took occasional, non-systematic observations at six locations close to ARIE-SH (Fig. 1, Table 2).

Analysis
We used the Jaccard Similarity Index defined according to Krebs (1989) to compare the aquatic bird community over an interval of approximately 33 years, where S = Jaccard Similarity Index between campaigns i and j; a = number of species that occur in both campaign i and campaign j (co-occurrence); b = number of species that occur in campaign j but are absent in campaign i; c = number of species that occur in campaign i but are absent in campaign j. We did not compare the terrestrial community between periods due to the heterogeneity of methodologies used in relevant studies and the fact that the present censuses did not sample the entire ARIE-SH area.
We used the Vegan package (Oksanen et al., 2007) within the R environment (R Core Team, 2019) to build species accumulation curves and calculate the similarity index. We further used the warbleR package (Araya-Salas & Smith-Vidaurre, 2017) to present the spectrogram.
We used the proposal by Piacentini et al. (2015) for bird taxonomy. Atlantic Forest endemic bird species are in agreement with Vale et al. (2018) and threat status follows global (IUCN, 2019), national (Brasil, 2014) and state (Paraná, 2018) references.

Reviews
Literature Scherer-Neto's (1997) data provided the foundation for our ARIE-SH avifauna information. The author ob-  tained a total of 143 species over four campaigns carried out every three months. During each campaign, the author covered ARIE-SH trails over three consecutive days (but also in adjacent areas, therefore, outside the ARIE-SH), alternating the starting points between the beginnings and the ends of the trails. He recorded bird species using transect counts and mist-netting and carried out transects in different environments.
In a study restricted to water species, Lara (1994) also inventoried locations near ARIE-SH, observing 28 bird species specifically at ARIE-SH. The author censused every three months over one year using the method we chose to use to census water birds.
Censusing at both ARIE-SH and at the Bela Vista Biological Refuge (RBBV) in the county of Foz do Iguaçu, Paraná, Seger et al. (1993), including Scherer-Neto's (1997) records, logged a total of 280 bird species but did not distinguish between the two locations in their records.
The most recent available ARIE-SH avifauna data are from Kliver's (2010) inventoried carried out in 1998 and in 2010. The author recorded 269 (24 orders and 61 families) species in 1998 and 197 (20 orders and 51 families) species in 2010, for a total of 271 species in 24 orders and 62 families. Between October 2009 and January 2010, the author conducted 80 visits to ARIE-SH at different times of the day, with a mean duration of 1.5 hours each. Observers carried out transects covering 53 km of trails with a constant speed of about 1.5 km/h. Kliver (2010) followed the same sequence on transects, alternating days of drought and rain and censusing during the mornings and afternoons.

Online database
As of March 27, 2020, only Wiki Aves showed bird records (97 species) from Santa Helena County. However, the data supplied by the citizen scientists did not accurately report observation locations; we therefore did not incorporate Wiki Aves records in our compilation.

Natural History Museums
We found two specimens (MZUSP 75893 and MZUSP 75894) of Great Dusky Swift Cypseloides senex collected on April 28, 1982 at the mouth of the São Francisco Falso River in Santa Helena County, as well as another eight species from Santa Helena housed at MHNCI, of which four were collected at ARIE-SH between 1987-1991 (Appendix).

Bird census
We recorded a total of 125 bird species in at ARIE-SH using two methodologies. Three species are Atlantic Forest endemics (Aramides saracura, Picumnus temminckii, and Tachyphonus coronatus) and one is endangered at state level (Busarellus nigricollis). We recorded no nationally or globally threatened species at ARIE-SH.

Land birds
We accumulated 65 five-species lists, recording 78 species while hiking trails at ARIE-SH. Two Atlantic Forest endemic forest species were recorded only during the land bird census (P. temminckii and T. coronatus). The species accumulation curve shows a tendency towards stabilization (Fig. 2). IFL values varied from one (0.016) to 23 (Basileuterus culicivorus, 0.359) records in lists (Appendix).

Water birds
During the inventories taken while traveling around the shore of ARIE-SH, we recorded 74 species in 15 orders and 32 families. Twenty-two of the birds recorded were "water birds" as defined following Scott & Carbonell's (1986) classification of aquatic birds. We recorded the regionally endangered Busarellus nigricollis exclusively on ARIE-SH's margins. The species accumulation curve for aquatic species did not show stabilization (Fig. 3).

Qualitative census
Our records (including occasional, non-systematic observations) and compilations resulted in the identification of 311 bird species for ARIE-SH and surroundings. The number of endemic species rises to six (A. saracura, P. temminckii, Amazona vinacea, Automolus leucophthalmus, Myiornis auricularis and T. coronatus). Regarding state level threatened species, we compiled one critically endangered species (Sporophila palustris), one endangered species (B. nigricollis) and four vulnerable species (A. vinacea, Campylorhamphus trochilirostris, Casiornis rufus and Sporophila leucoptera). At the national level, no threatened species were encountered. At the global level, two species are listed as endangered (A. vinacea and S. palustris). We recorded 17 species in the area that had not been registered (Appendix). Lara (1994) recorded 28 aquatic bird species (eight exclusive) while we detected 22, of which two were only recorded by us (Table 3). The similarity index between current and previous water bird censuses was Sji = 0.67.

DISCUSSION
There are a few exceptional preterit records which merit mention as they are no longer present regionally, such as the Scaled Pigeon Patagioenas speciosa and the Crested Eagle Morphnus guianensis, both obtained in Marechal Cândido Rondon, some 25 km east of Santa Helena (Straube & Bornschein, 1989, 1995. Another important record is that of the Large-tailed Antshrike Mackenziaena leachii, obtained in Foz do Iguaçu, Pato Bragado, Marechal Cândido Rondon and Guaíra by Kaempfer (Naumburg, 1937). The Plumbeous Pigeon P. plumbea is also worth mentioning, deposited at MHNCI and currently quite rare in interior Paraná. Some 100 km further east, Pinto & Camargo (1956) presented several species collected in Porto Camargo (Icaraíma municipality), such as the Undulated Tinamou Crypturellus undulatus, the Three-toed Jacamar Jacamaralcyon tridactyla, the Helmeted Woodpecker Celeus galeatus, the Black-capped Antwren Herpsilochmus atricapillus, the Russet-mantled Foliage-gleaner Syndactyla dimidiata and the Crested Oropendola Psarocolius decumanus. Scherer-Neto et al. (1997) reported three important (undocumented) singletons: White-browed Woodpecker Piculus aurulentus, Rufous-headed Tanager Hemithraupis ruficapilla and Green-throated Euphonia Euphonia chalybea, Atlantic Forest endemics which currently do not have records in this region of the state. Such species are of paramount importance for characterizing the avifauna once present. Therefore, they indicate community composition changes which resulted from habitat modification.

Reforestation
ARIE-SH lies approximately 80 km northeast from PNI, an area containing 335 bird species (Straube et al., 2004). This great species richness is a reflection of PNI's large size, abundant native vegetation and diversity of habitats, which include mixed and seasonal forests. However, because ARIE-SH and PNI are close to one another and share a seasonal forest biome, their avifauna compositions should have been similar before ARIE-SH was deforested (Willis, 1979). In addition, small (≥ 20 ha) reforested seasonal semideciduous Atlantic Forest remnants in northern Paraná do have equal or higher bird richness and similar functional group structures when compared to contiguous native vegetation. Thus, they can maintain levels of biodiversity and reduce species extinction debt (Santos-Júnior et al., 2016). Furthermore, as long as contiguous to native fragments, interior Atlantic Forest reforested areas in São Paulo state with native and exotic plant species proved to be ecologically more functional for birds and to contribute to the maintenance of resident and visiting species by providing food resources (Athiê & Dias, 2012). Since we recently recorded only 125 species within ARIE-SH, we present a few arguments why such differences currently occur.
Some trophic guilds-specialized forest species, such as large frugivores and understory insectivores, tend to disappear from degraded environments (Willis, 1979;Ribon et al., 2003). Some of these species were only recorded in past ARIE-SH censuses, such as the large frugivores Crypturellus obsoletus and Penelope superciliaris. The former, a terrestrial species having low sensitivity to forest fragmentation (Anjos, 2006), uses small forest remnants (10 ha) as long as there is continuous connec- Table 3. List of aquatic bird species recorded within ARIE-SH in previous (Lara, 1994) and recent censuses. tion among forest patches (Barbosa et al., 2017). Because there is continuous reforested vegetation bordering the eastern margins of the lake connecting ARIE-SH to PNI, this species' absence strongly suggests habitat modification. Although Penelope superciliaris avoids habitat modification (Pereira- Ribeiro et al., 2018), it can be found in small forest remnants (Anjos, 2001) and even in disturbed ones (pers. obs.). Both C. obsoletus and P. superciliaris are game birds and, especially in the case of P. superciliaris, invite intense poaching activity, which may play a major role in their dearth (Pereira- Ribeiro et al., 2018). This conclusion is corroborated by the fact that P. superciliaris was quickly sighted during our visits to nearby locations (Appendix) that do not suffer the hunting pressure. The ecological consequences of this species absence are of paramount importance as both species are able to swallow fruits with large seeds that smaller bird species cannot disperse, greatly contributing to the fruit's trees spreading throughout its environment (Pizo, 2004).
Dendrocolaptids and thamnophilids are examples of understory insectivorous. Species belonging to this tropic guild are among the first to disappear from forest fragments at the earliest sign of habitat modification (Stouffer et al., 2011). Therefore, a more permeable matrix facilitates their dispersal within landscapes. Not only matrices, but passive restoration, is a cheap and effective solution to recover some taxocenes in the Atlantic Forest (Guerrero & Rocha, 2010). We did not record any dendrocolaptid species and only three thamnophilids, with T. caerulescens being the most common. This species is less sensitive to forest fragmentation (Ribon et al., 2003;Anjos, 2006), occurring in several small fragments (Barbosa et al., 2017) and, apparently, is not affected by changes in the vegetation's structure at ARIE-SH.
In an area of similar size to ARIE-SH in the municipality of Rio Claro, São Paulo state, a Eucalyptus sp. woodlot with native understory in interior São Paulo state, the bird community is impoverished when compared with a 10-times smaller native remnant (Willis, 2003), a trend in Neotropical forests (Iezzi et al., 2018). The fact that non-deciduous tree species (permanent canopy cover precluding the passage of sunlight) have been used for the reforestation of ARIE-SH resulted in the absence of a native understory. A clear understory contributes to the loss of bird species in such environments, as already reported for the Brazilian state of Espírito Santo (Marsden et al., 2001) and for Argentina (Iezzi et al., 2018). Although Eucalyptus sp. forests can be important within a matrix context, meaning bird species can use them to move among fragments (Barbosa et al., 2017), this is not the case in ARIE-SH, which is surrounded by water and in extremely close proximity to monocultures on the east, such as soybean and maze. Because large-scale (at least 10-year-old) tree planting in corridors adjacent to mature forests results in rapid increased abundance and expanded distribution of forest birds (Pejchar et al., 2018), we draw attention to the need to investigate whether the plant species used for reforestation are adequate for fauna usage of the protection strip bordering Itaipu dam and whether it is being used for travel among National Parks.
Although other empirical studies have observed that habitat modification can lead to an absence of specialist species, it is possible that the noted absence of a species in ARIE-SH is due to recording error or observational failure and that the species was present. Observational failure can be caused by (1) imperfect detection, especially evident when the chance of detecting a species is less than 1 (Mackenzie et al., 2017), or (2) because the species are in ARIE-SH microhabitats that we did not visit.

Similarity
The Similarity Index between the composition of the aquatic avifauna found by Lara (1994) suggests that almost 70% of the aquatic avifauna remained the same at ARIE-SH. This index is relatively high when compared to those found by da Da Silva-Jr. (2007), who monitored bird composition within the Caiapó River Valley in the state of Goiás, pre-and post-filling of small lakes behind two dams. The author obtained Sji = 0.23, an expressive, but expected, dissimilar result since the author censused bird communities during completely different times: before and after damming. Lara (1994) visited ARIE-SH some five years after Itaipu dam began filling up. Then, microhabitats which are not currently found, may have been present, such as sand beaches suitable for finding several piper species. We only recorded groups of the White-backed Stilt Himantopus melanurus (not on beaches, but afloat on macrophytes) and lone individuals of Solitary Sandpiper Tringa solitaria, while Lara (1994) could find another three of these migratory species. In addition, the author detected the Cattle Egret Bubulcus ibis, a very common species in pastures which we did not see on aquatic environments. Apart from the Gray-necked Wood-Rail Aramides cajaneus, commoner to the northeast of the state (Scherer-Neto et al., 2011), the remaining species we did not record within ARIE-SH were actually seen by us in similar habitats on other locations in the municipality. Similarly, the two species Lara (1994) failed to detect during her censuses probably result from randomness.

Species accounts
Busarellus nigricollis (Black-collared Hawk): Considered endangered in Paraná (Paraná, 2018). This species occurs in almost all of Brazil (Bierregaard et al., 2020), but there are few records of its presence in Paraná. We saw a lone individual on the shores of Lake Itaipu in a bay with exposed stumps surrounded by riparian vegetation (24°47′46″S, 54°21′56″W) in June, August, September and November (Fig. 4). There was an abundance of aquatic macrophytes observed, such as Eicchornia spp. and Salvinia sp. First recorded in Paraná in Altônia in 1989 (Straube & Bornschein, 1995), some five years after Itaipu lake damming. No collector found this large and conspicuous species in this region before, which suggests a Quagliato, I.S. & Cavarzere, V.: Bird records for western Paraná Pap. Avulsos Zool., 2021;v.61: e20216130 7/23 recent distribution expansion in the state due to water habitats (absent before damming) created by Itaipu dam. Currently, the species seem to occur almost exclusively on the banks of the Paraná (Scherer-Neto & Straube, 1995) and Paranapanema rivers.
Laterallus exilis (Gray-breasted Crake): Data deficient in Paraná (Paraná, 2018). It has been observed in the southern region of the Brazilian Amazonia, eastern Brazil (Taylor, 2020) and in northeastern Argentina (Pearman et al., 2000). We skinned an individual who collided with a wall of a UTFPR -Santa Helena faculty building (24°50′43″S, 54°20′38″W) on August 29, 2019 (UTFPR-SH-010). This observation took place about 600 km from the nearest prior sighting of this species, Serra do Mar, Paraná (Batista, 2015), and is the first documented record of the species outside Paraná's coastal region (Scherer-Neto et al., 2011;Fig. 5). This species inhabits predictable habitats such as flooded grasslands, rice fields, wetlands, and can be easily detected through the playback technique. Its few records in the state denotes a probable underrepresentation and more individuals may be found in specific searches for the species.

Amazona vinacea (Vinaceous-breasted Parrot):
Vulnerable in Paraná (Paraná, 2018). An Atlantic Forest endemic occurring in southeastern Brazil, eastern Paraguay and the province of Misiones in Argentina (Collar et al., 2020). The species has disappeared from most areas in Paraguay and Argentina where they had been historical-ly recorded. Most notably the Itaipu Reserves Complex in Paraguay is of paramount importance for the species (Cockle et al., 2007). The species has a distribution that is quite coincident with that of Araucaria angustifolia (Bertol.) Kuntze. Our observation is one of the few sightings of A. vinacea in a Seasonal Semideciduous Forest (Urben-Filho et al., 2008). We saw six individuals flying over the UTFPR-SH campus (24°50′42″, 54°20′36″W) on December 14, 2018 (Fig. 6). They were silent and appeared to be coming from either the Limoy Reserve on the Paraguayan side of the Paraná River or from ARIE-SH.

Herpsilochmus longirostris (Large-billed Antwren):
Its distribution in Brazil extends from the states of Mato Grosso, Tocantins and Goiás, extending to the south of Ceará and south of Piauí, Paraná and São Paulo (Zimmer & Isler, 2020). Considered endemic to the Brazilian Cerrado (Silva, 1995), it has a characteristic distributions pattern typical of Cerrado species, which arrive in the interior of Paraná traveling down the Paraná River. Known to be found along the banks of the Ivaí and Paraná rivers (Straube & Urben-Filho, 2005), sightings of this species have been logged approximately 70 km to the north in Palotina (Osaki, 2016) and 100 km to the north in Guaíra (Freitas, 2011), evidencing their presence in western Paraná. While traversing ARIE-SH's western marginal trail (24°49′59″S, 54°22′04″W) on June 9, 2017, we saw, photographed (Fig. 7), and recorded their vocalizations. Our record represents the southernmost record of the species in Brazil.   western Paraná Pap. Avulsos Zool., 2021;v.61: e20216130 9/23 Campylorhamphus trochilirostris (Red-billed Scythebill): Vulnerable in Paraná (Paraná, 2018), it has the widest distribution of its genus, appearing in all five regions of the country and in 22 Brazilian states (Marantz et al., 2020). Although it has been logged in Paraná's Tibagi River Basin (Anjos et al., 1997), its general distribution pattern indicates that it occurs across Paraná along the Paraná River. Our record is of an individual that collided with a window at the Federal University of Paraná's campus in the city of Palotina and is the southernmost record of this species in Brazil (Fig. 8). (Paraná, 2018), it is found in Bolivia, central Brazil (Mato Grosso, Goiás, Minas Gerais, Mato Grosso do Sul, Paraná and São Paulo, and recently reported in Rio Grande do Sul), Paraguay and northern Argentina. During the austral winter, it can be found in small numbers in Peru and north and northeastern Brazil (Scholes, 2020). There are few records of the species having been observed in Paraná (Scherer-Neto et al., 2011). We observed two individuals and taped their vocalizations while walking a trail on the east bank of ARIE-SH (24°50′43″S, 54°21′40″W) on October 20, 2019 (Fig. 9). Our observation is about 250 km south of the previous most southern logged observation in the state (Souza, 2019).

Casiornis rufus (Rufous Casiornis): Vulnerable in Paraná
Campylorhynchus turdinus (Thrush-like Wren): Found in Brazil's Amazon region, both to the north and south of the Amazon River, also in the states of Mato Grosso (Pantanal), Bahia and northern Espírito Santo (Kroodsma et al., 2020). It has been suggested that the geographic expansion of the species towards the south is a result not only of deforestation, but also of climate change (Hayes et al., 2018). The first record of the species in Paraná was from the city of Foz do Iguaçu (Bencke et al., 2008), which is an urban environment. We first observed it (Fig. 10) western Paraná Pap. Avulsos Zool., 2021;v.61: e20216130 11/23 ests found in the country's interior, which includes northeastern Paraná, but has not been previously recorded in western Paraná (Fitzpatrick et al., 2004;Scherer-Neto et al., 2011). Model et al. (2014) reported its presence in an urban fragment of the city of Cascavel (24°57′20″S, 53°27′19″W), but this report is not documented. Our record from ARIE-SH (24°50′03″S, 54°21′02″W) represents the southernmost record of the species in Brazil, approximately 150 km from the previous southernmost documented sighting (Stencel & Caxambu, 2018). It probably reached western Paraná via the reforested margins of Itaipu Lake. We taped the vocalizations (Fig. 11) of at least two individuals on the 25 th of May and 15 th of November in 2017 and on the 20 th of October in 2019.
Eucometis penicillata (Gray-headed Tanager): This species ranges from western portion of the Brazilian state of Maranhão, to the states of Mato Grosso, Minas Gerais and São Paulo (Hilty, 2020). There are few records of the species from Paraná, and all are centered to the northeast bordering São Paulo state (Scherer-Neto et al., 2011). The previous most southern logged observation was at the Caiuá Ecological Station (Scherer-Neto et al., 2008, 2011

Sporophila palustris (Marsh Seedeater):
Critically endangered in Paraná (Paraná, 2018). Sporophila palustris Marsh Seedeater appears in the center and center east and south of Brazil, northern Argentina, and possibly northern Paraguay (Lowen et al., 1996;Jaramillo, 2020;   Vizentin- Bugoni et al., 2013). It breeds in its southern range: Rio Grande do Sul, Brazil, Uruguay, and northern Argentina (BirdLife International, 2017). A sighting of the species in northwestern Paraná was confirmed over two decades ago (Scherer-Neto & Straube, 1995) and again, much more recently, in eastern Paraná. On November 2, 2016, Paludo (2016) observed two S. palustris individuals in Campina Grande do Sul near Curitiba in eastern Paraná; two days later, we logged and photographed (Fig. 13) one male S. palustris in a small open field on the UTFPR-SH campus. The brief interval and the relatively close proximity of these two S. palustris sightings indicates that flocks of the species were likely migrating south at the beginning of their reproductive season. The individual was among hundreds of S. caerulescens and dozens of S. lineola feeding on native the sourgrass Digitaria insularis (L.) Fedde in an area where the exotic 2 m-tall Guinea grass Megathyrsus maximus (Jacq.) predominated (24°50′49″S, 54°20′07″W). We did not spot S. palustris on following days. These hundreds of other individuals stayed at this location for one week, after which their numbers were significantly lower. We suspect the site served as a resting point while flying over during a migration route. These grasses were cleared in 2017, and new construction has taken their place.
On the same date we also saw two male S. collaris and, on the next day, one male S. leucoptera. This S. leucoptera observation was 70 km south of its previous most southern sighting in Paraná, but our sighting is undocumented.

CONCLUSION
There is a relatively great amount of information about the birds of ARIE-SH, but most of it was generated more than 30 years ago, rendering this location as extremely under sampled. The terrestrial bird community at ARIE-SH seems to have undergone a major change over the intervening years, especially with respect to species sensitive to habitat modifications. Some terrestrial species were no longer in evidence (large frugivores and understory insectivores), and others were recorded at ARIE-SH for the first time. From our observations, ARIE-SH's aquatic bird community remains relatively unchanged. Although the site had been reasonably well studied, it is of utmost importance to continuously monitor its bird populations to properly document how their community is being affected by habitat modification in the form of flora fragmentation and composition. The changes in the native forest avian species' populations at ARIE-SH reported in our inventories reinforces the importance of avian surveys taking to gauge the effects of severe habitat modification, such as occurred at ARIE-SH, on bird populations, especially endangered species' populations, if the goal of maintaining avian diversity is to be met.

ACKNOWLEDGMENTS
To Universidade Tecnológica Federal do Paraná campus Santa Helena for financial and structural support. To Fundação Araucária for a scholarship for ISQ. To Edson Poier for making water censuses possible. To Itaipu Binacional for allowing this study to be conducted. To Pedro Scherer-Neto who contributed to the difficult-to-obtain literature used here. To Luís F. Silveira for kindly allowing access to MZUSP bird collection. Antenor Silva-Júnior kindly contributed with specimens' data deposited under his care. For accompanying us traversing Santa Helena and region: Adriana B. Bussler, Ana P. Zingler, Jonas P. Grigolo, Júlio C.S.M. de Souza, Liamara C. Zagonel, Lucas P. de Oliveira, Nathália L.E. Januário and Sara O.G. Maicrovicz. Data on ARIE-SH were exhaustively collected thanks to the fundamental collaboration of Denise Lange, Edicléia Bonini, Heleno Brandão, Leonardo Biral (who also identified the sourgrass species), Rejane B. de Oliveira and Tatiane Tambarussi. To André Rüdiger who alerted us to the record of C. trochilirostris. We thank João Maicrovicz and Jaqueline R. dos Santos for contributing with information during the 2020 coronavirus quarantine period. We extend our gratitude to all citizen scientists who, even unintentionally, greatly contribute to our understanding of Brazilian birds. Dione Seripierri kindly provided difficult-to-find references. We conducted our research according to a SISBio permit (56171 -1). Anonymous reviewers greatly contributed to the first draft of this manuscript. Christopher C. Fields reviewed the English.

AUTHORS' CONTRIBUTIONS
I.S.Q.: Conceptualization, Writing -original draft, methodology, data analyses, investigation. Writing -proofreading and editing. V.C.: Supervision, Conceptualization, methodology, data analyses. Writing -proofreading and editing. All authors actively participated in the discussion of the results, reviewed and approved the final version.