The Fundão tailings dam held back approximately 50 million cubic meters of tailings from an Fe ore mine in the state of Minas Gerais, Brazil. The dam failed on 5 November 2015, releasing tailings into the Rio Doce watershed and causing an environmental disaster in a biodiversity hotspot. To guide conservation action following the event, Brazilian authorities demanded an impact assessment focused on terrestrial species threatened by extinction. Postdisaster impact assessment is substantially more challenging than predevelopment impact assessment. Predisaster baseline data were sparse, and much was unknown about how the Fundão dam failure impacted terrestrial species threatened with extinction. Baseline reconstruction and impact pathway validation and characterization was critical. Ecosystem reconstruction revealed that 1580 ha of terrestrial ecosystems were destroyed, including approximately 480 ha of Atlantic rainforest. Collaboration with local experts identified 346 species of vertebrates, invertebrates, and vascular plants threatened with extinction that were impacted or potentially impacted. Species composition and impacts varied within 4 distinct segments of the Rio Doce watershed. Nine potential impact pathways were identified using data and literature review combined with information provided by regulators, experts, and media. Not all were valid. Validating and characterizing each pathway within each watershed segment involved data evaluation, an ecological risk screening, and field assessments. Impact pathway evaluation proved critical to dispelling misconception, accurately understanding postdisaster impacts, and directing conservation action; the importance of this step cannot be overemphasized. Integr Environ Assess Manag 2020;16:676–680. © 2020 SETAC

ABSTRACT. We analyzed the genetic diversity of populations of two sympatric species of Lymania (Bromeliaceae), both endemic to the Atlantic rainforest of southern Bahia (Brazil). Lymania azurea has a restricted occurrence, while Lymania smithii has a wider distribution. Our aim was to provide genetic data to contribute to the design of more efficient conservation strategies for these bromeliads, possibly justifying inclusion in the official Brazilian list of Endangered Species. Up to now, L. azurea has been classified by the Brazilian Ministry of the Environment as "data deficient". We sampled four populations of L. azurea throughout its distribution area in southern Bahia and two populations of L. smithii in the same region. Genotyping was performed with 48 random amplified polymorphic DNA markers. Based on the Jaccard genetic similarity index, L. smithii has greater diversity than L. azurea. An analysis of molecular variation showed greater genetic variance within than between populations for both species. L. azurea was found to have 20% inbreeding, probably due to population fragmentation, with L. smithii showing only 10%. When we analyzed pairs of populations of L. azurea within a conservation unit, we found low population structure (Ф ST = 0.098), apparently due to a large degree of gene flow between them. In disturbed areas, we found a higher Ф ST (0.372). We found low genetic variability for L. azurea, probably as a consequence of habitat fragmentation, supporting the need for its inclusion in the Brazilian list of endangered flora.