Influence of monsoon-related riparian phenology on yellow-billed cuckoo habitat selection in Arizona

Aim: The western yellow-billed cuckoo (Coccyzus americanus occidentalis), a Neotropical migrant bird, is facing steep population declines in its western breeding grounds owing primarily to loss of native habitat. The favoured esting habitat for the cuckoo in the south-western United States is low-elevation riparian forests and woodlands. Our aim was to explore relationships between vegetation phenology patterns captured by satellite phenometrics and the distribution of the yellow-billed cuckoo, and to use this information to map cuckoo habitat. Location: Arizona, USA. Methods: Land surface phenometrics were derived from satellite Advanced Very High-Resolution Radiometer (AVHRR), bi-weekly time-composite, ormalized difference vegetation index (NDVI) data for 1998 and 1999 at a resolution of 1 km. Fourier harmonics were used to analyse the waveform of the annual NDVI profile in each pixel. To create the models, we coupled 1998 satellite phenometrics with 1998 field survey data of cuckoo presence or absence and with point data that sampled riparian and cottonwood–willow vegetation types. Our models were verified and refined using field and satellite data collected in 1999. Results: The models reveal that cuckoos prefer areas that experience peak greenness 29 days later, are 36% more dynamic and slightly (< 1%) more productive than their average cottonwood–willow habitat. The results support a scenario in which cuckoos migrate northwards, following the greening of riparian corridors and surrounding landscapes in response to monsoon precipitation, but then select a nesting site based on optimizing the near-term foraging potential of the neighbourhood. Main conclusions: The identification of preferred phenotypes within recognized habitat can be used to refine future habitat models, inform habitat response to climate change, and suggest adaptation strategies. For example, models of phenotype preferences can guide management actions by identifying and prioritizing for conservation those landscapes that reliably exhibit highly preferred phenometrics on a consistent basis.