Kari E Norman, Timothée Poisot 

Comprehensive monitoring of biodiversity to direct conservation action is foundational to addressing the ongoing biodiversity crisis. As integrative monitoring programs increasingly come online in response to multilateral biodiversity agreements, establishing best practices for optimal design is critical. Selecting the appropriate algorithm for identifying sample sites is both necessary for robust inference from biodiversity data and largely neglected from broader monitoring network discussions. Here, we benchmark the performance of four common selection algorithms, outline the characteristics of the suite of algorithms suitable for ecological monitoring design, and offer recommendations for their best use. While all algorithms outperformed simple random samples, performance differences were negligible between algorithms. We recommend instead that practitioners choose algorithms based on feature availability, which varies greatly between algorithms.