Confidence in ecological indicators: A framework for quantifyinguncertainty components from monitoring data

Abstract The value of an ecological indicator is no better than the uncertainty associated with its estimate. Nev-ertheless, indicator uncertainty is seldom estimated, even though legislative frameworks such as theEuropean Water Framework Directive stress that the confidence of an assessment should be quantified.We introduce a general framework for quantifying uncertainties associated with indicators employed toassess ecological status in waterbodies. The framework is illustrated with two examples: eelgrass shootdensity and chlorophyll a in coastal ecosystems. Aquatic monitoring data vary over time and space; vari-ations that can only partially be described using fixed parameters, and remaining variations are deemedrandom. These spatial and temporal variations can be partitioned into uncertainty components operatingat different scales. Furthermore, different methods of sampling and analysis as well as people involved inthe monitoring introduce additional uncertainty. We have outlined 18 different sources of variation thataffect monitoring data to a varying degree and are relevant to consider when quantifying the uncertaintyof an indicator calculated from monitoring data. However, in most cases it is not possible to estimateall relevant sources of uncertainty from monitoring data from a single ecosystem, and those uncertaintycomponents that can be quantified will not be well determined due to the lack of replication at differ-ent levels of the random variations (e.g. number of stations, number of years, and number of people).For example, spatial variations cannot be determined from datasets with just one station. Therefore, werecommend that random variations are estimated from a larger dataset, by pooling observations frommultiple ecosystems with similar characteristics. We also recommend accounting for predictable patternsin time and space using parametric approaches in order to reduce the magnitude of the unpredictablerandom components and reduce potential bias introduced by heterogeneous monitoring across time. Wepropose to use robust parameter estimates for both fixed and random variations, determined from a largepooled dataset and assumed common across the range of ecosystems, and estimate a limited subset ofparameters from ecosystem-specific data. Partitioning the random variation onto multiple uncertaintycomponents is important to obtain correct estimates of the ecological indicator variance, and the magni-tude of the different components provide useful information for improving methods applied and designof monitoring programs. The proposed framework allows comparing different indicators based on theirprecision relative to the cost of monitoring