Forest bird monitoring on a Caribbean Island where distance sampling is complicated
Speaker: Dr Steffen Oppel, RSPB Centre for Conservation Science, (Cambridge)
Many forest birds in the Caribbean region are threatened by habitat loss, invasive species, and potentially climate change. To better understand the status and population trends of forest birds, regular population monitoring is necessary. On Montserrat, a small British Overseas Territory, the eruption of a volcano caused massive forest loss in 1997, and most forest birds are now confined to a 1100 ha forest patch in the centre of the island. Annual forest bird monitoring has been conducted since 1999, and initially indicated that an endemic bird (Montserrat Oriole, Icterus oberi) was rapidly heading towards extinction. Intensive research into demographic processes of this species revealed that the oriole was highly susceptible to volcanic activity, but that the population could rebound after years with heavy rain early in the breeding season. Since 2010 volcanic activity has slowed down, and based on an integrated population model the extinction probability of the Montserrat Oriole was estimated to be <1%, and the population appears to be stable. The monitoring established in 1999 was based on distance sampling to account for imperfect detection, but because estimating distances to vocalising birds was unreliable in a lush tropical forest, the design was changed in 2011 to include repeated visits in each breeding season. Long-term trend estimates are now based on binomial mixture models, which account for imperfect detection, but may not be powerful enough to detect small changes in population size. While the status of most forest birds is reasonably secure, robust long-term monitoring of the forest birds is essential to detect any population declines. This talk provides an overview of this monitoring programme which records data for 15 different species, but so far only data for two globally threatened species have been used and the remainder is available for collaborative projects.