Kerala Bird Atlas 2015-2020: features, outcomes and implications of a citizen-science project

Citizen-science driven exercises (e.g., bird surveys) and online platforms (e.g., eBird) provide voluminous data on bird occurrence but the data quality is often compromised. Bird atlases are based on standardized surveys and describe the distribution of bird species over a predefined region and have less biases, and thus are better suited for use in research. The recently concluded Kerala Bird Atlas (henceforth KBA) is Asia’s largest bird atlas in terms of geographical extent, sampling effort and species coverage. The entire state of Kerala was systematically surveyed twice a year during 2015–2020 and over 0.3 million records of 380 species from 25,000 checklists were aggregated. The dataset was filtered and various metrics were estimated. A few cells could not be surveyed due to logistical constraints or when they fell outside the state boundary. This variation in sampling effort had a minimal effect on survey completeness; the slope of the species accumulation curve suggests near-complete species sampling in over 70% of the cells. After eliminating nocturnal and pelagic species, in total 361 species were analyzed. Species count was higher in the dry season (Jan.–Mar.) than in the wet season (Jul.–Sep.). Species richness (count) and evenness were higher in the northern and central districts than in the southern districts. High elevation regions of the southern Western Ghats were the largest contiguous areas lacking sufficient sampling. Although most of the endemics were concentrated in the Western Ghats, threatened species were as likely to occur along the coasts as in the Ghats. The KBA dataset is a valuable resource for testing various ecological hypotheses and suggesting science-backed conservation measures. The Kerala Bird Atlas model could be replicated for similar atlases in other states or biogeographic regions of India.

Methods
The state of Kerala was divided into cells of size 3.75 minutes x 3.75 minutes (equivalent to 6.6 km x 6.6 km) aligned to Survey of India maps. A total of 1,096 cells were laid out covering the entire state. Each cell was further divided into four quadrants of size 3.3 km x 3.3 km. Each quadrant was then sub-divided into 9 sub-cells of size 1.1 km x 1.1 km. A single, randomly selected sub-cell in every quadrant was chosen for the survey. Field surveys were conducted from 2015 to 2020 twice a year, during dry (mid-Jan. to mid-Mar.) and wet (mid-July to mid-Sep.) seasons, exactly for 60 days in each season per year. Survey teams consisted of 2–5 volunteers, every bird species seen or heard during the 15-minute sampling session was noted. Volunteers uploaded checklists from the surveyed sub-cells to eBird, using the website or the app, and this data was later downloaded for analysis.