Under-sampled KBAs, over-sampled roadsides: integrating historical and contemporary records of Malagasy bees

Madagascar boasts remarkable biodiversity and endemism, yet its indigenous bee population is under-explored. Researchers have compiled the initial extensive list of Malagasy bee sightings, trimming 10,721 initial records down to 4,071 confirmed occurrences of 218 geographically marked species out of the 224 species listed across six families – a 89.3% endemic figure.

Despite nearly covering the complete checklist, the dataset remains severely lacking, with most species represented by a handful of records. The sampling efforts were unevenly distributed among the various taxa, with a majority of genera underrepresented due to their ecological rarity, detectability, and collector specialization.

Spatial concentration within a few specific grid cells further amplified these biases. Temporally, the recordings varied significantly, with historical booms attributed to individual collectors and a post-2010 shift towards Apidae-dominated records. Spatially, 79.9% of the 25x25 km grid cells covering Madagascar had no bee records, while 77.1% of the records were found within a 2.5 km radius of roads – a pattern commonly observed globally.

The sampling was heavily clustered near major cities, with frequent sightings of common species near roads and infrequent sightings of rare species spread over larger distances. Out of the 231 Key Biodiversity Areas (KBAs) identified, 68.4% remained entirely unexplored. Of the sites that were sampled, they only accounted for 26.7% of all records, and even within KBAs, the sampling efforts were uneven, leaving a significant proportion of undetected diversity in most locations.

These findings highlight widespread temporal, spatial, taxonomic, and collector-driven biases, emphasizing the urgent need for focused surveys within KBAs and beyond roadsides to enhance coverage of under-documented species and bolster conservation evaluations.