Measuring occupancy for an iconic bird species in urban parks

Urbanisation is a significant and increasing threat to biodiversity at the global scale. To maintain and restore urban biodiversity, local communities and organisations need information about how to modify green spaces to enhance species populations. 'Citizen science' initiatives monitoring the success of restoration activities also require simple and robust tools to collect meaningful data. Using an urban monitoring study of the bellbird (Anthornis melanura), we offer advice and guidance on best practice for such monitoring schemes. Three independent surveys were undertaken across 140 locations in Christchurch's urban parks. Detection probabilities (estimated from six repeat five-minute point counts at each location per survey) were used to calculate unbiased occupancy estimates for the second and third surveys. A single five-minute count had c. 60% chance of detecting bellbirds at a location where they were present, while the cumulative detection probability increased to almost one after five repeat counts per survey. Bellbird detection probabilities varied between surveys (albeit weakly) but not with environmental conditions. Occupancy, which declined slightly over the study period, was highest in parks with more native woody vegetation, less paved areas and close to the Port Hills (which were mostly Riverbank/Conservation parks). Robust estimates of bellbird occupancy require at least three repeat counts per location per survey within a short time frame, with multiple locations ideally surveyed concurrently.