Our study area has 63,000ha and encompasses the most densely populated portion of the metropolitan region of João Pessoa, including the municipalities of Cabedelo, Santa Rita, Bayeux, Lucena and João Pessoa (34° 58′87″W 7° 15′46″S and 34° 47′19″W 6° 54′43″S) (Fig. 1). A recent study indicates that 57% of the original vegetation cover were lost between 1974 and 2010, and the most affected areas were the rain forests, sandbank forests (restingas) and tabuleiro forests (Stevens, 2014). Currently, the 14,847ha of native vegetation left are scattered over 236 forest remnants, most smaller than 50ha and surrounded by urban agglomerations or agricultural fields (mainly sugarcane plantations) under peri-urban conditions. About 20% of the remaining vegetation is protected by law within 18 protected areas, either within the federal law 9985/2000 or other legal instruments at the state and municipal spheres. Nine protected areas showed minimum level of security to permit access and fieldwork inside. They vary in size, internal level of fragmentation, surrounding matrix, predominant vegetation, presence of waterbodies, and administration regime (Table 1). Altogether, they conserve a mosaic of ecosystems composed of distinct types of forests, from rain to drier forests (restinga and tabuleiro forests), salt marshes, mangrove swamps and estuaries. In this study we consider urbanization as a multidimensional process that manifests itself through rapidly changing human population and changing land cover (sensuSeto et al., 2013) and, consequently, assume that all remaining vegetation inside the study area are facing urbanization.

Fig. 1.

Study area in the metropolitan region of João Pessoa, Northeast Brazil. The nine protected areas are shown in black.

(0.49MB).

Between June 2014 and May 2015, we used 50 Mackinnon lists of 10 species to sample birds in each of the nine protected areas, starting early in the morning (around 6 am) to maximize species detection (Bibby et al., 1998). In the Mackinnon method (Ribon, 2010), each species is recorded only once in each list until the maximum number of species in the list is reached (10 in our study). After that, a new list is started. We produced 10 lists per day per area during five consecutive days, totaling 450 lists. An advantage of this method is that the list may be generated throughout a day without application time, or on subsequent days. Furthermore, it allows the calculation of species frequency by list and area, uncovering common and rare species (Manhães and Loures-Ribeiro, 2011). Nonetheless, we were conservative and excluded any doubtful record from our lists.

We recorded the birds visually with binoculars Nikon 7×50, aurally through species vocalization and confirmed species identification with specialized literature and sound recordings of private collection (Sick, 1997; Erize et al., 2006; Sigrist, 2009). We alternated consecutive field surveys between smaller and larger protected areas to control for possible seasonal effects on bird detection, activity and occurrence along the gradient of protected area size. Furthermore, we restricted the surveys to the forest formations of each protected area (i.e., rain forest, tabuleiro forest, restinga forest; Table 1) to minimize compositional differences attributed to habitat type. The taxonomic classification of the species followed the Brazilian Ornithological Records Committee (Piacentini et al., 2015). We searched for the conservation status of identified species in the red lists of IUCN and the Brazilian government. Introduced species were also identified.

To quantify the relative contribution of each protected area for the regional bird species richness, we divided the species richness of each area by the regional species richness. Our expectation was that smaller areas showed smaller proportion of the regional richness than larger areas. A simple linear regression was performed in JPM 8 (SAS Institute Inc., Cary, NC, USA) to test this relationship. We used the Jackknife method of first order to estimate the regional species richness following the equation: Sjack1=Sobs+Q1 (m−1/m), where Sjack1 is the estimated species richness; Sobs is the observed number of species in all lists; Q1 is the number of species recorded in a single list; and m is the number of lists. We obtained the estimates and species accumulation curves with the software Estimates 9.1 with 1000 randomizations (Colwell, 1997).

To assess how differences in protected area size and isolation affected species composition, we constructed a dissimilarity matrix of taxonomic composition based on the Jaccard distance (1−J) and used Mantel tests to correlated it with matrices of geographic distance and difference in size. We expected that less isolated and similar-sized protected areas showed an avifauna more similar to each other than more isolated, different-sized areas. We produced matrices and ran Mantel tests in Primer 6 (Clarke and Gorley, 2006). To identify the most common and rarest species in the region, we calculated the frequency of occurrence of each species by dividing their number of occurrence by the total number of occurrence of all species in the lists.

We recorded a total of 126 species (7 endemics), 17 orders and 41 bird families (Table S1). The richness estimated with Jackknife method of first order was approximately 136 species (Fig. S1), indicating the presence of 10 species more in our study region. Thraupidae and Tyrannidae, with 16 species each, were the most representative families, followed by Trochilidae (8 spp.), Thamnophilidae (6 spp.), Rhynchocyclidae (5 spp.), Picidae (5 spp.) and Columbidae (5 spp.). Eighteen of the 41 families were represented by only one species (Fig. S2).

As expected, the richest area was the largest reserve (Gargaú Private Reserve of Natural Heritage, 1058ha), with 95 species, accounting for 75.4% of the regional species richness. The Buraquinho Refuge of Wildlife (518ha) and the Xem-Xem State Park Forest (182ha), second and third largest areas, followed with 80 and 62 species, respectively (63.5% and 49.2% of the regional richness). Five areas ranging from 22ha to 155ha protected between 40.5% and 43.7% of the regional richness: Cuiá Natural Municipal Park (56 species), Arruda Câmara Zoo Botanical Park (55 species), Cabedelo Sandbank National Forest (54 species), Cabo Branco Park (51 species) and Paraíba Federal University Forest (51 species). The smallest area, the Cidade Verde Municipal Forest Nursery (14.8ha), protected as much as 35.7% of the regional richness (45 species). When we modeled the proportion of regional richness in each protected area as a function of its size, a clear positive relationship arose (R2=0.916; F1,7=77.2; P<0.0001; Fig. 2), showing the remarkable contribution of larger areas for bird conservation.

Fig. 2.

Relationship between the size of protected areas (ha) and the proportion of regional bird species richness (126 spp.) in the metropolitan region of João Pessoa, Brazil.

(0.07MB).

Pairwise similarity averaged 52.6% across the protected areas, varying from 36.5% (Gargaú Private Reserve of Natural Heritage vs. Cidade Verde Municipal Forest Nursery) to 71.4% (Arruda Câmara Zoo Botanical Park vs. Paraíba Federal University Forest). There was a trend for distance-decay in species composition (Rho=0.379; P=0.055) supported by lower similarity between more distant areas (Fig. 3). Differences in size also influenced species composition (Rho=0.421; P=0.04), with similar-sized areas being more taxonomically similar than different-sized areas (Fig. 3). Nonetheless, in both cases the strength of the relationship (Rho values) was low, suggesting that the size and isolation of the protected areas are not the main drivers of species composition.

Fig. 3.

Relationship of species dissimilarity (1−J) with geographic distance and difference in size for nine protected areas in the metropolitan region of João Pessoa, Brazil.

(0.13MB).

The most frequent species was Pitangus sulphuratus (Great kiskadee; Tyrannidae), taking place in all areas and in more than half of the 450 lists. Other common species included Coereba flaveola (Thraupidae), Vireo olivaceus (Vireonidae), Fluvicola nengeta (Tyrannidae), and Columbine talpacoti (Columbidae), occurring in 30%-37% of the lists (Table S1). In addition to P. sulphuratus, other 18 species occurred in all protected areas, but at lower frequencies. Ten species were recorded only once in the entire study.

Only one species, Picumnus fulvescens (Picidae), is recognized by the IUCN as near threatened (NT). However, six species appear in the Brazilian Red List of Endangered Species in worrying categories: Penelope superciliaris (Cracidae; CR – Critically endangered), Momotus momota (Momotidae; EN – Endangered), Conopophaga melanops (Conopophagidae; VU – Vulnerable), Xenops minutus (Xenopidae; VU), Platyrinchus mystaceus (Platyrinchidae; VU) and Xiphorhynchus atlanticus (Dendrocolaptidae; VU). Only two introduced species were detected: Estrilda astrild (Estrildidae) and Passer domesticus (Passeridae), but P. domesticus occurred in 25.8% of the lists.

In addition to more investment in basic infrastructure and full-time staff (Silva-Junior and Santos, 2017), we strongly recommend the responsible agencies to follow the federal law 9.985/2000 and create the management council and the management plan of the four areas officially recognized as ‘conservation units’. Lawfully, these actions should have been taken in the five years after creation, but to date only the Cabedelo Sandbank National Forest is in conformity with legislation. This legal noncompliance threatens the protected areas with downgrading, downsizing and degazettement (Bernard et al., 2014) and make public use of their ecosystem services almost unfeasible. Three municipal protected areas, which are not yet recognized as conservation units, face even more legal insecurity as their creation act may be revoked more easily. In this sense, we also recommend the official recognition of these areas as conservation units under federal law 9.985/2000.

Despite these challenges, the remarkable bird diversity of small and large urban reserves suggest that birdwatching for contemplative and educational purposes may help to change the course of conservation in the region, provided that it is well planned and based on the current legislation. Birdwatching is the act of observing and identifying birds in their native habitats; it is becoming the most rapidly growing and most environmentally conscious segment of ecotourism and provides economic hope for many threatened natural areas around the world (Sekercioglu, 2002). Positive impacts of birdwatching include link between avian diversity and local income, financial incentive to conserve wildlife, less impact and more income than typical tourism, increased local control due to unique bird species, visitation of areas outside traditional tourist itineraries, protection of unprotected areas with desired species, valuation of local natural history knowledge, education and employment of local guides, generation of funds for bird conservation and contribution to ornithological knowledge (Sekercioglu, 2002). Negative impacts involve disturbing birds by playing tapes and by approaching, increased nest predation and nest abandonment, increased disturbance of rare and/or threatened birds, visitor-related pollution and habitat destruction, cash leaks from local communities, resentment by excluded locals and cultural degradation associated with tourism (Sekercioglu, 2002).

Following Sekercioglu's recommendations for optimal birdwatching, we suggest that birdwatching initiatives in ours and other urbanizing regions of the Brazilian Atlantic Forest adhere to and insist on ethical birding conduct, avoid nests and young birds as much as possible, show particular care with threatened and rare species, minimize tape use and try to minimize being seen, do not approach further once a bird notices the watcher, stick to established roads/trails/walkways, use scopes for observation and photography, educate locals about birds and their financial benefits, support local and low-impact establishments and contribute to non-governmental organizations active in bird conservation. Providing a comprehensive birdwatching planning extrapolates the limits of this study, but its principles are presented here. We hope conservationists and decision makers find this information useful and give more attention to the urban reserves under their responsibility.