Search

There is an urgent need for reliable data on the impacts of deforestation on tropical biodiversity. The city-state of Singapore has one of the most detailed biodiversity records in the tropics, dating back to the turn of the 19th century. In 1819, Singapore was almost entirely covered in primary forest, but this has since been largely cleared. We compiled more than 200 y of records for 10 major taxonomic groups in Singapore (>50,000 individual records; >3,000 species), and we estimated extinction rates using recently developed and novel statistical models that account for “dark extinctions,” i.e., extinctions of undiscovered species. The estimated overall extinction rate was 37% (95% CI [31 to 42%]). Extrapolating our Singapore observations to a future business-as- usual deforestation scenario for Southeast Asia suggests that 18% (95% CI [16 to 22%]) of species will be lost regionally by 2100. Our extinction estimates for Singapore and Southeast Asia are a factor of two lower than previous estimates that also attempted to account for dark extinctions. However, we caution that particular groups such as large mammals, forest-dependent birds, orchids, and butterflies are disproportionately vulnerable. [ABSTRACT FROM AUTHOR]

Aim: Research in island biogeography has long focused mainly on present‐day island configurations. Recently, there has been an increasing focus on islands' past histories of land connection, shape and size. Moreover, continental islands (=shelf islands) have received less attention than oceanic islands, and species inventories from extremely small islands are lacking in many datasets. We examine the effects of sea‐level rise since the Last Glacial Maximum (LGM) on bird species diversity and composition of tropical shelf islands in Southeast Asia. Location: Sundaland. Taxon: Birds. Methods: We compiled avifaunal island inventories for 94 islands using an exhaustive literature review of historic surveys of larger islands combined with our own comprehensive island surveys from both small and large islands. Using generalised least‐squares models with spatial autocorrelation, we assessed the importance of traditional biogeographical parameters including area, maximum elevation, distance from mainland and geographical isolation, along with post‐LGM effects of change in island area and duration since isolation. We also compared the species composition on similar‐sized shelf islands from two categories—recently submerged and unsubmerged—using non‐metric multidimensional scaling. Results: Post‐LGM effects on species diversity are minimal and insular diversity is instead well explained by present‐day island characteristics, such as area, distance to mainland and proportion of land surrounding an island within a 10 km radius (Cox and Snell Pseudo‐R2 = 0.803). Avifaunal diversity is similar across recently submerged and unsubmerged small shelf islands. Main conclusion: Avifaunal diversity on tropical shelf islands equilibrates rapidly after isolation, indicating that both extinction and immigration rates are high. In particular, a high immigration rate of dispersive species maintains diversity, especially on small islands. Over‐water dispersal is generally restricted to short distances among Sundaic birds. Consequently, the diversity of an island can be maintained by the presence of large or stepping‐stone islands near it. [ABSTRACT FROM AUTHOR]

Insects as a group are suffering rapid declines in many parts of the world but are also poorly studied relative to vertebrate taxa. Comprehensive assessments of insect declines must account for both detected and undetected species. We studied extirpations among butterflies, a particularly well-known insect group, in the highly developed and biologically well-surveyed island city-state of Singapore. Building on existing butterfly species lists, we collated museum and naturalist records over the last two centuries and used statistical models to estimate the total extirpation rate since the first major collections in 1854. In addition, we compiled a set of traits for each butterfly species and explored how they relate to species discovery and extirpation. With a database of 413 native species, 132 (32%) of which are recorded as extirpated in Singapore, we used a statistical model to infer that, in addition, 104 unknown species (95% CI 60–162) were likely extirpated before they were ever discovered, suggesting a total extirpation rate of 46% (41–51%). In the trait analyses, we found that butterfly species that were discovered later were weakly associated with rarer larval host plants and smaller wingspans, while species that persisted for longer were weakly associated with higher larval host plant abundance and lower forest-dependence. This exercise is one of the first to offer a holistic estimate of extirpations for a group of insects by accounting for undetected extirpations. It suggests that extirpations among insects, specifically in the tropics, may be higher than naïve estimates based only on known records. [ABSTRACT FROM AUTHOR]