Multiple Choice, Multiple Answers

For more than a century, amateur birdwatchers have contributed to scientific knowledge. Each winter since 1900, volunteers in North America have counted birds in fixed circles of land as part of an annual survey, and similar schemes now operate in many countries. Smartphone apps have greatly expanded this kind of citizen science: millions of checklists are now submitted every year, recording which species were seen, where and when. These records have allowed researchers to map changes that no professional team could track alone. Long-term counts have shown that the wintering ranges of many species have moved northward over recent decades, a shift that matches rising winter temperatures. Data from volunteers have also helped to identify steep declines in some common farmland birds. Volunteer data are not without problems. Observers are concentrated near cities and roads, so remote areas are under-recorded, and experienced birders detect more species than beginners do. Rare or colourful birds may also be reported more eagerly than common ones. Rather than discarding such data, statisticians build these biases into their models, for example by recording how long each observer searched and how skilled they are. When this is done, studies comparing volunteer records with professional surveys have generally found that the two give similar estimates of population trends.