Choosing an efficient portfolio of sampling strategies for monitoring marine biodiversity

Effective monitoring of marine biodiversity is crucial for understanding ecological dynamics, assessing human impacts, and guiding management decisions. Measuring biodiversity is particularly challenging in nearshore subtidal systems because of their wide variety of organisms. To monitor the variety of life in these habitats, most research programs employ multiple sampling methods, but because of the high cost of underwater sampling, only a few methods can be undertaken by any program and limited resources tightly constrain total sampling effort. Despite the complexity and importance of these resource allocation decisions, they are often made on an ad hoc basis, by copying existing programs or following simple statistical rules of thumb (e.g., minimum replication).

Here we develop analyses to evaluate a portfolio of alternative strategies for sampling biodiversity in order to guide decisions about which methods should be employed and how effort should be allocated across them. We construct efficiency curves for each candidate sampling method, using existing data to predict the biodiversity information expected for a given level of investment. We then combine these individual curves into multidimensional efficiency surfaces describing how any combination of sampling strategies yields information about biodiversity. We demonstrate the method analyzing data from rocky reefs in the Santa Barbara Channel, a location where the Marine Biodiversity Observation Network has been developing new monitoring methods and synthesizing data from existing methods. We find that optimal investment strategies are relatively consistent across a variety of biodiversity metrics, but that they vary more dramatically if information about specific taxonomic groups is valued asymmetrically. This method can guide the design of new monitoring projects and can help existing projects determine when new or improved methods should be added to existing sampling portfolios.