Don't bet against the natal homing abilities of marine fishes.

Whether marine fishes are capable of homing to their natal areas has long been something of an enigma. For some estuarine species or sharks (which have extended nondispersal juvenile stages or are born as relatively large, fully formed juveniles), the answer is clearly 'yes' (Thorrold et al. ; Feldheim et al. ), but for most marine fishes, the issue is much more mysterious. Many species have free-floating eggs, and most have pelagic, passively dispersing larvae. It is challenging to imagine how adult fish might navigate to a region of the ocean they experienced only as eggs or larvae, and easier to assume that such dispersal leads inexorably to high gene flow, and even panmixia. One way to resolve the conundrum would be to track fish from hatching to reproduction, but for marine fishes with tiny eggs and drifting larvae, this is notoriously difficult to do (Bradbury & Laurel ). In this issue of Molecular Ecology, Bonanomi et al. () use a creative approach to solve this challenge for Atlantic cod (Gadus morhua) populations that mingle in the vicinity of Greenland. They show that cod that disperse more than a 1000 km away from Iceland as eggs and larvae, then spend years growing on the far side of Greenland, while mixing with two local populations, return as adults to spawning areas near Iceland - and further, that this behaviour has remained stable over more than six decades. They manage this feat with a clever use of historical cod tracking data, modern genomic data and genetic analysis of decades-old DNA obtained from archived materials. Their results have important implications for our view of the biocomplexity of marine fish populations, and how we should manage them.