Skip to main content

Avian survival on islands

Cite this dataset

Beauchamp, Guy (2020). Avian survival on islands [Dataset]. Dryad.


Island species are often predictably different from their mainland counterparts. Milder climates and reduced predation risk on islands have been involved to explain shifts in body size and a suite of life history traits such as clutch size and offspring growth rate. Despite the key role of adult survival on risk taking and reproduction, the prediction that living on islands increases adult survival has yet to be tested systematically. I gathered data on adult annual apparent survival from island and mainland year-round resident species of birds from around the world. With this large dataset (708 species), I found that species of birds living on islands showed higher apparent survival than their mainland counterparts in the two hemispheres and at all latitudes controlling for several known predictors of adult survival including body size, clutch size, and breeding system. These results shed light on the ecological factors that influence survival on islands and extend the life history island syndrome to adult survival.


I searched the literature for avian studies that reported survival rates. I focused on year-round resident species, as most island species are not migratory. I first consulted earlier extensive reviews of survival rates in birds [15, 18, 20-22]. I complemented this set by performing online database searches using key words such as ‘adult’, ‘survival’, ‘annual’ and ‘birds’. Studies were included in the analysis if they met the following criteria: (1) the study provided estimates of adult annual survival for year-round resident species, (2) survival was not based on age ratios or DNA fingerprinting, (3) the study population was not subject to extensive harvest or artificial conditions that can influence survival such as predator control, (4), the study did not include urban populations or recently introduced species. Most studies reported survival estimates for two sexes combined but if this was not the case I took the arithmetic mean of the two sexes.


No funding

No funding