Variable retention mitigates negative effects of traditional harvesting on biodiversity, maintaining legacies in harvested forests as habitat for sensitive species, and generating other habitats for early-seral species. Their assessment, including drivers, has a high concern for society. Likewise, species association with ecosystems in the landscape (as different forest types, grasslands and peatlands) are unknown for some taxa and regions, and could explain species flow into harvested areas. Therefore, we evaluated the forest harvest impact on ground-dwelling spiders and vascular plants, their diversity and association to vegetation types in the landscape matrix, and richness and abundance correlations with forest structure. Six natural common habitats were studied in Tierra del Fuego (Argentina), where variable retention has been implemented: aggregated (AR) and dispersed (DR) retention in harvested Nothofagus pumilio forests, unharvested primary forests (NPF) as reference, grasslands (G), peatlands (P), and unharvested N. antarctica forests (NAF). We surveyed spiders (N = 432, by 6 habitats × 6 replicates × 6 collections × 2 years), and vascular plants (N = 36, by 6 habitats × 6 replicates), and characterized forest structure in wooded ecosystems. We determined for both taxa richness, relative abundance (total captures/ground cover) and Shannon-Wiener and Pielou indices, compared assemblage composition and indicator species among harvesting treatments and vegetation types, and analyzed correlations. We sampled 752 individuals of spiders from 33 species (six families), and 79 vascular plant species (28 families) that averaged 52% total vegetation cover. Total spider captures and overall richness were higher in DR > NPF > AR, although differences were not detected at plot level for any variable. For understory vascular plants, richness and cover were the highest on DR, followed by AR and NPF, as well as at plot level, with no affectation in indices. Likewise, highest richness and abundance occurred in NAF and NPF for spiders, and in NAF and G for vascular plants, with unique assemblages in each vegetation type despite the shared species. Assemblages also differed in harvested areas, including species of other vegetation types mainly in DR, where new conditions were generated by reduction of forest structure variables (evidenced by negative correlations), while AR maintained species composition similar to NPF, contributing to the conservation of forest specialists. Indicators occurred for both taxa in several habitats. This study highlights the importance of different vegetation types for spider and plant conservation at landscape level, while provides tools for developing monitoring strategies and conservation policies.