Abstract
Abundance estimates based on adequate survey design and count methodology are needed for population monitoring and modelling, and for assessing the results of conservation actions taken to boost or maintain population size at desired target levels. We monitored Bonaire's population of yellow‐shouldered parrot Amazona barbadensis rothschildi using systematic distance sampling surveys in 2009–2017, and developed a Bayesian state‐space logistic model to predict changes in abundance resulting from increased human‐induced mortality in 2018–2066. Survey‐based abundance estimates (mean ± bootstrapped SE) were 0.172 ± 0.020 parrots ha‐1 and 2924 ± 340 parrots at a survey region covering 17 000 ha. Model‐based posterior distribution estimates (mean ± MCMC SD) of maximum population growth rate, maximum sustainable mortality rate, maximum sustainable mortality, population carrying capacity and equilibrium population size were 0.179 ± 0.129, 0.090 ± 0.064, 219 ± 135, 5623 ± 2043 and 2811 ± 1022 parrots. With low to moderate mortality rates (0.001– 0.100, 0.101–0.250), predicted population sizes (mean ± MCMC SD) were 2963 ± 668 and 2703 ± 1660 parrots in 2018, and 2754 ± 690 and 2297 ± 1301 parrots in 2066. With high mortality rates (0.251–0.500), predicted population sizes were 1780 ± 1160 parrots in 2018 and 26 ± 139 parrots in 2066. Because the relative importance and magnitude of human–parrot conflicts are unknown but may be unsustainable, we consider the parrot population vulnerable to the risk of extinction during the modelled time horizon. Therefore, we recommend long‐term monitoring and modelling for assessing changes in abundance and the results of conservation actions taken to keep the population above 2800 parrots in the survey region (i.e. population size N > 2.5% percentile of the posterior distribution of population carrying capacity K).
Highlights
How many parrots live in a defined survey region? Can parrot numbers in the survey region remain stable above a desired target level? Answering these basic questions can be challenging when survey region coverage and parrot counts are incomplete (Buckland et al 2001, 2008, 2015, Marques et al 2007, 2010, Nichols et al 2009)
Abundance estimates based on adequate survey design and count methodology are needed for population monitoring and modelling, and for assessing the results of conservation actions taken to boost or maintain population size above the desired target level
Conventional and multiplecovariate detection models generated similar mean density estimates, suggesting that model selection was of secondary importance for abundance inferences in the survey region (Buckland et al 2001)
Summary
How many parrots live in a defined survey region? Can parrot numbers in the survey region remain stable above a desired target level? Answering these basic questions can be challenging when survey region coverage and parrot counts are incomplete (Buckland et al 2001, 2008, 2015, Marques et al 2007, 2010, Nichols et al 2009). We present a population assessment of the yellowshouldered parrot Amazona barbadensis rothschildi on the island of Bonaire, Caribbean Netherlands (Fig. 1) For this population assessment, we used abundance estimates from systematic distance sampling surveys in 2009–2017, and model-based simulations of changes in abundance resulting from increased human-induced mortality in 2018–2066. Parrot roost counts have been conducted by volunteers on Bonaire since the 1980s (for additional information, see < www.dcbd.nl/monitoring/yellow-shouldered-parrotcounts >) These roost counts have been treated as population censuses (i.e. total counts), assuming complete coverage of roosting sites (i.e. all sampling units within the sampling frame) and complete detection of roosting parrots (i.e. all elements within the sampling units), while not accounting for differences in data quality and count effort (e.g. the experience and number of volunteers and roosts year–1), nor quantifying count precision (see e.g. the variance estimator developed by Casagrande and Beissinger 1997). (Roberts et al 2014: 39), while not accounting for observation and process error variances in trend modelling and estimation (Kéry et al 2009, Kéry and Schaub 2012, Hostetler and Chandler 2015, Rivera-Milán et al 2016)
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
