摘要 : Monitoring long-term trends in population size is important for species' conservation assessments. However, it may be unfeasible for rare species, for which past records are typically sparse. Here, I investigate the potential of b... 展开 Monitoring long-term trends in population size is important for species' conservation assessments. However, it may be unfeasible for rare species, for which past records are typically sparse. Here, I investigate the potential of birding trip reports as an underappreciated source of biological information to monitor rare species. For this purpose, I used an uncommon species - the Peruvian thick-knee Burhinus superciliaris - as an example to assess population trends over 2000-2010, using flock size dynamics as a proxy. I collected all observations of thick-knees that could be accessed from birding trip reports across the entire geographic range of the species. In total, I compiled 218 records of 2403 individuals, of which 73.9% were fully useable for this study. Mean flock size declined slightly over the study period in Central Peru, whereas in North Chile it increased at the beginning of the decade and then decreased nearly 90% until the end of the study period. This suggests that, at least at the southernmost part of the distribution range of the thick-knee, the magnitude and speed of the change in population size could be sufficient to qualify for a threatened category. The use of birding trip reports as a data source for monitoring trends of rare species seems promising compared to other sources, since sightings of sought-after species are increasingly posted online as a form of tourist attraction and so made publicly accessible for researchers and managers. This study highlights the great utility that birding trip reports freely available online may have to provide retrospective data on rare species that would otherwise be impossible to collect. Nonetheless, they might not be universally valid due to species-specific differences and possible lack of non-detection records. Potential biases due to varying sampling effort or detectability should also be carefully considered. 收起