Is sparse local ecological knowledge accurate enough for policy? A seagrass mapping case study from five Greek islands in the Eastern Aegean Sea

The low cost, high applicability and community-based nature of Local Ecological Knowledge (LEK) data for mapping marine benthic habitats has become increasingly popular in recent years. Nonetheless, due to persisting technocratic distrust of LEK often stemming from the unquantifiable nature of bias, the embedding of empirical data at policy levels often lags. Currently, protocols rely on high participation from stakeholders, however, such quotas are often unfeasible, with common limiting factors being the low numbers of knowledgeable individuals in remote parts of the world, and the lack of community trust in the conservation mechanism. In this study, we test whether data from a limited number of interviews with selected artisanal fishers can provide accurate and policy-valuable information as an alternative to high-participation data collection. Our work focuses on the case of Posidonia oceanica seagrass beds around five Greek islands in the eastern Aegean Sea. In comparison with remotely sensed seagrass distribution maps, our results show that the quality of selected LEK data can be exceptionally high (on average 78% accuracy; maximum 92% accuracy), arguing the point that limited-participation LEK should not be discarded as invalid at policy levels. Furthermore, the way LEK is recorded, as well as the number of years of fishing experience of the participants, were found to significantly affect the quality of data being delivered. Lastly, we show that LEK outperformed the governmentally-issued seagrass map, thus providing further evidence of fishers’ LEK ability to deliver policy-relevant data regarding the distribution of their local marine habitats. • Evaluated low-participation fishers Local Ecological Knowledge for seagrass mapping. • Results indicate a mean accuracy of ∼78% (max 92%) against remotely sensed data. • Empirical data outperformed official governmental seagrass maps against remotely sensed data.