6. Reliability of Datasets

Because nests of buzzards are large and conspicuous, and observers were able to survey target habitats multiple times in a breeding season, possibilities of missing nests are likely to be minimal. The rationale is as follows. Firstly, all observers committed a large amount of time to search for active nest sites, i.e., observation effort in each site was 20–50 h/week during the breeding season (from April to late July). Second, in four of the five sites, nest-location surveys have been conducted over periods ranging over 5–13 years. In these cases, we used data from the year with the greatest number of recorded nests. In Aichi, surveys were conducted for only three years, with 160 h/week invested, and no new breeding cells were detected during the last two years (S3 Fig).

During the last three years of our observation, only limited numbers of cells were newly detected, as breeding cells, in four of the five sites (Tocihgi: 1, Chiba: 0, Aichi: 0, Fukuoka: 1). The Iwate site was an exception, i.e., we found three new breeding cells in 2011. Because we confirmed that observers invested a similar amount of survey effort between 2011 and the previous three years, these were designated as newly established nests in these areas.