咨询与建议

看过本文的还看了

相关文献

该作者的其他文献

文献详情 >Estimating the spatio-temporal... 收藏
Estimating the spatio-temporal variation of bird phenology u...

Estimating the spatio-temporal variation of bird phenology using citizen science data

作     者:Pedro Guilherme Antunes Lopes da Silva Nicolau 

作者单位:Universidade de Lisboa 

学位级别:硕士

导师姓名:Tiago Andre Marques;Alison Johnston

授予年度:2017年

主      题:phenology generalized additive models bootstrap pied flycatcher barn swallow 

摘      要:Migratory bird populations all over Europe are deeply changing. In particular, long-distance migrants are being increasingly affected by climate change and have shown difficulties in adapt- ing to the recent and fast changes. Studying the spatial variation in ecological processes is of fundamental importance when unfolding the mechanisms underlying bird population change. The purpose of this study was understanding the variation in arrival and breeding onset with latitude, estimating the length of time between these two processes and describing their rela- tionship. Using data from two citizen science databases, we estimated the latitudinal variation of arrival date and breeding onset for two migratory bird species, Pied Flycatcher (Ficedula hypoleuca) and Barn Swallow (Hirundo rustica), in Great Britain in the year of 2014. Unlike many of the past site-specific localized studies, here we focused on broad population patterns. Generalized additive models were used to produce estimates of both phenological processes, us- ing a spatio-temporal interaction smooth. Arrival was estimated using data from citizen science database, BirdTrack, by modelling the probability of occurrence of each species on complete birding checklists. This allowed the estimation of the date representing arrival of 50% of the population at different locations. For the breeding onset, we used data from citizen science based Nest Record Scheme to model clutch initiation events, as a measure of breeding onset. From this we extracted the date where the estimated peak of breeding occurred at different locations. A ten-folded cross-validation procedure was used to provide metrics of model validation. Both models were bootstrapped to estimate uncertainty, and bootstrap estimates were used to calcu- late the gap between arrival and breeding. Our results proved inconclusive for Barn Swallow, yet the Pied Flycatcher showed a marked reduction in gap length with an increase in latitude. This is the first

读者评论 与其他读者分享你的观点

用户名:未登录
我的评分