Estimating ecological thresholds for good status using natural variation in reference conditions

Abstract Thresholds to distinguish between good and degraded ecosystem status are required to assess and manage the condition of marine systems under international conventions (e.g. Marine Strategy Framework Directive, MSFD), but many frameworks do not yet have such thresholds and consistent, transparent and robust methodologies to fill this gap are currently lacking. Here we estimated thresholds for good status, as the proportion of state that must remain relative to a minimally impacted reference condition (scale 0–1; reference condition mean = 1), using two different methods (range of natural variation RNV, which quantifies the natural variability of the ecological state in time series of reference conditions, and statistically detectable change SDC, an alternate approach, which uses the lower confidence interval for the same monitoring data). These methods were applied to 92 pre‐existing minimally impacted time series of biomass, density and abundance indicators from seagrass, plankton lifeforms, benthic invertebrates and tropical reef fish. The RNV method estimated a range of average good‐state thresholds between 0.40 and 0.85 depending on indicator species group with the majority between 0.58 and 0.78. The SDC method estimated higher thresholds (0.65–0.94) with the majority ranging between 0.82 and 0.90. We found that environmental characteristics and species life history traits can be used to predict thresholds for five of the indicator species groups (e.g. benthic invertebrate longevity predicted thresholds). Synthesis and applications. The framework outlined here presents a transparent, robust, scalable and pressure‐independent approach to estimate the probability of an ecosystem component being in good status for a range of marine systems. This straightforward and easily understood approach can be used to estimate thresholds for good status where they are missing for ecological condition assessments under legislation (e.g. Good Environmental Status under the MSFD) and make progress towards achieving international biodiversity objectives. In practical terms, this enables managers to take defensible decisions on whether action is needed to manage human activities and improve ecosystem health.