Size‐selective harvesting in commercial fisheries can induce rapid changes in biological traits. While experimental and wild harvested populations often exhibit clear shifts in body size and maturation associated with fishing pressure, the relative contributions of genetic and environmental factors to these shifts remain uncertain and have been much debated. To date, observations of so‐called fisheries‐induced evolution (FIE) have been based solely on phenotypic measures, such as size data. Genetic data are hitherto lacking. Here, we quantify genetic versus environmental change in response to size‐selective harvesting for small and large body size in guppies (Poecilia reticulata) across three generations of selection. We document for the first time significant changes at individual genetic loci, some of which have previously been associated with body size. In contrast, variation at neutral microsatellite markers was unaffected by selection, providing direct genetic evidence for rapid evolution induced by size‐selective harvesting. These findings demonstrate FIE in an experimental system, with major implications for the sustainability of harvested populations, as well as impacts on size‐structured communities and ecosystem processes. These findings highlight the need for scientists and managers to reconsider the capacity of harvested stocks to adapt to, and recover from, harvesting and predation.

SUMMARY Based on morphological characters, cross‐fertility and molecular systematics, two species are currently recognized in the ubiquitous temperate brown algal genus Ectocarpus: the type species E. siliculosus (Dillwyn) Lyngbye and E. fasciculatus Harvey. We studied diversity, cross‐fertility and ecology of Ectocarpus in megatidal areas in northwest France (Western Europe) and propose to reinstate a third species, E. crouaniorum Thuret in Le Jolis. Genotyping of 67 individuals from five localities, including the type locality of E. crouaniorum, using internal transcribed spacer 1 (ITS1) length as a marker, showed that the three species co‐occurred whenever the habitat was suitable. Our survey also revealed a single putative field hybrid between E. crouaniorum and E. siliculosus, and a single individual of a further Ectocarpus genotype. In laboratory experiments, E. crouaniorum was crossed with E. siliculosus and E. fasciculatus. In 12 of 13 crosses, the zygotes did not develop (postzygotic sterility); in one experiment a viable hybrid was produced after crossing a female E. crouaniorum with a male E. siliculosus, but this hybrid was unable to form meiospores. Phylogenetic analysis of five molecular markers from the nuclear, mitochondrial and plastid genomes (in total 1818 bp) confirmed genetic separation of the three species. Ecologically, E. crouaniorum was confined to high intertidal pools and run‐offs, where the gametophyte was common from spring to summer. Another characteristic was that it usually occurred as an epiphyte of up to 12 cm in length on erect thalli of Scytosiphon lomentaria. Sporophytes of E. crouaniorum were found all year long; they were <3 cm in size or microscopic and were epilithic in the same habitat. The presence of a third species of Ectocarpus in Western Europe suggests that species diversity in this genus is larger than recognized during the last 40 years.