Abstract
Juvenile Atlantic salmon Salmo salar from three allopatric populations ( LaHave , Sebago and Saint‐Jean) were placed into artificial streams with combinations of four non‐native salmonids: brown trout Salmo trutta , rainbow trout Oncorhynchus mykiss , Chinook salmon Oncorhynchus tshawytscha and coho salmon Oncorhynchus kisutch . Non‐additive effects, as evidenced by lower performance than predicted from weighted summed two‐species competition trials, were detected for S. salar fork length ( L F ) and mass, but not for survival, condition factor or riffle use. These data support emerging theory on niche overlap and species richness as factors that can lead to non‐additive competition effects.