The mystery of nocturnal birds in tropical secondary forests

As tropical logging increases in scope, understanding the capacity of tropical secondary forests for sustaining oldgrowth (OG) forest biodiversity has become a critical question in tropical conservation. There is growing realization that tropical conservation cannot be limited to pristine habitats in protected areas (Sekercioglu et al., 2007; Chazdon et al., 2009). To better integrate secondary forests into conservation schemes, it is essential to understand the proportion of native forest biodiversity that can persist in secondary forests (Chazdon et al., 2009). This is particularly urgent in light of the growing significance of the UNReducing Emissions from Deforestation and Forest Degradation in Developing Countries program (Grainger et al., 2009). For preserving tropical biodiversity, the ideal way to sequester carbon is to preserve OG forests. However, the magnitude of the climate change problem means that preserving OG is not enough. Secondary forests can also sequester large amounts of carbon, even 100 years after the start of regrowth (Silver, Ostertag & Lugo, 2000). The Biological Dynamics of Forest Fragments Project (BDFFP) has pioneered the experimental approach in tropical forest fragmentation studies, by taking advantage of forestclearing operations near Manaus, Brazil (Laurance et al., 2002). BDFFP worked with landowners to make sure that they left forest patches in squares of 1, 10 and 100ha to be compared with extensive OG, resulting in an excellent experimental setup. Now, hundreds of papers later and entering its fourth decade, BDFFP scientists are again skillfully taking advantage of circumstances beyond their control (Sberze, Cohn-Haft & Ferraz, 2010). It has proven logistically challenging to keep cleared much of the land that was initially deforested. Even though this has reduced the isolation of the original forest fragments, it has also led to many new and interesting questions about the biodiversity value of the different types of secondary growth (Antongiovanni & Metzger, 2005; Barlow et al., 2007; Laurance et al., 2007). A large part of the tropical conservation literature focuses on diurnal birds, as they are relatively easy to sample. However, sampling nocturnal species can be challenging (Borges, Henriques & Carvalhaes, 2004; Bhattacharjee, 2008) and these birds have received little attention in tropical conservation studies. Sberze et al. (2010) rectify this by combining state-of-the-art modeling with intensive fieldwork. The densities of nocturnal birds can be notoriously difficult to estimate (Borges et al., 2004), especially considering the problems with estimating distance based on vocalizations (Alldredge, Simons & Pollock, 2007). Therefore, Sberze et al. (2010) used occupancy modeling (MacKenzie, 2006) based on detection data to estimate the likelihood that a site is occupied by a species. Extensive field sampling covered all months of the year and the authors accounted for spatial autocorrelation by quantifying it in their models. Vocalization playback of focal species further increased the chances of their detection. The findings are encouraging, although not surprising for those who have field experience with tropical night birds. OG and 30-year-old secondary forest (SF) did not differ in species richness. Each habitat was estimated to host 10–12 nocturnal bird species. Only one species, the potooNyctibius leucopterus, clearly avoided second growth. However, Nyctibius griseus and the nightjarsCaprimulgus griseus andCaprimulgus nigrescens preferred it, defying the authors’ predictions for C. nigrescens. Neither of the focal owl species differed significantly in its preference for OG versus secondary forest, which was also unexpected for Lophostrix cristata. Of the six other nocturnal species detected, two potoo species did not have enough detections for parameter estimation, but three owl species showed higher point estimates of occupancy for SF than for OG. The preference for SF and embedded forest fragments is likely due to the higher abundances of small mammals, on which many owls feed, and better foraging habitat for nightjars that usually prefer to feed in open areas (del Hoyo, Elliott & Sargatal, 1999). Nevertheless, the location of this study presents an unusual best-case scenario for forest birds, as the SF survey points are within 0–2km of extensive primary forest. Birds of prey can cover large areas while foraging and even non-predatory resident tropical songbirds are capable of covering 6km in 2h (Sekercioglu et al., 2007). As the authors themselves are first to point out, some of the seemingly SF-preferring species may have source populations in OG. The possibility of nocturnal birds nesting inOG and ‘commuting’ to SF, analogous to some tropical bats (Evelyn & Stiles, 2003), also remains to be