How are we choosing species for direct seeding restoration in megadiverse neotropical landscapes? Recommendations from projects implemented in Brazil

Direct seeding for ecological restoration is a cost‐effective and scalable method, but its effectiveness largely relies on choosing the appropriate species. Understanding the criteria used to select species is critical for optimizing restoration outcomes, especially in the face of limited resources and ongoing environmental changes. This study evaluated the criteria used for species selection in direct seeding restoration projects implemented in Brazil. A systematic literature review was conducted, analyzing studies published up to August 2024. We assessed species selection criteria, seed source traceability, and seed quality across various biomes. Species selection is predominantly based on logistical criteria, with seed availability (20%) and natural occurrence (19%) being the most frequently cited. Ecological and functional traits, such as drought or thermal stress tolerance, were rarely considered (<10%). Seed source traceability was absent in over 71% of the studies, raising concerns about reduced genetic diversity and maladaptation of restored populations to future conditions. Additionally, pre‐sowing seed quality testing, such as germination and vigor assessments, was conducted in fewer than 40% of studies. The overrepresentation of tree and shrub species, particularly legumes, highlights a bias toward forested ecosystems, neglecting non‐forested biomes such as savannas and grasslands. Our findings emphasize integrating functional traits into species selection and prioritizing seed source traceability to enhance restoration success. Establishing traceability programs and centralized databases would ensure ecological suitability and genetic diversity, aligning restoration practices with international environmental targets and climate adaptation goals.