Aggregated retention, in which patches of trees (aggregates) remain unlogged within larger harvested units, was first applied commercially in 1986. A primary goal was to maintain greater diversity of forest-dependent species through harvest, relative to conventional clearcutting. Despite its global application, the long-term benefits for biodiversity and the comparative responses of disparate taxonomic groups to aggregated retention are largely unknown. A critical knowledge gap relates to the role of 'forest influence' - whether and to what extent aggregates affect biodiversity in neighboring harvested areas. We sampled plants, beetles, and spiders/harvestmen in the world's three oldest aggregated-retention sites (21-26. years old), matched with three recently harvested sites (5-8. years old). For each taxonomic group, we compared species composition between undisturbed aggregates and regenerating forests to assess the 'lifeboating' function of aggregates. For each group, we also modeled changes in species composition, and in the numbers of aggregate- and regeneration-affiliated species, with distance from the aggregate edge into the regenerating forest along transects at north-facing edges. For all three taxa, species composition differed between aggregates and regenerating forests in both older and recent sites, confirming the long-term effectiveness of aggregates for lifeboating. The compositional difference between habitats was significantly greater at recent than at older sites for plants, but not for invertebrates. Plants and spiders/harvestmen responded to forest influence, with a marginal response for beetles. Responses for plants and spiders generally manifested as increased numbers of aggregate-affiliated species and decreased numbers of regeneration-affiliated species in regenerating areas closer to edges. Our results indicate that aggregated retention has short- and long-term benefits for biodiversity reflecting both the lifeboating and forest-influence functions of aggregates. However, variation in the responses of plants, beetles and spiders suggests that these benefits cannot be generalized among taxa. We advocate broader application of aggregated retention in forests managed for timber production and encourage managers to incorporate the benefits of forest influence in harvest designs by arranging aggregates to reduce average distances from harvested areas to unlogged habitats.