Census and effective population sizes of white‐spotted charr (Salvelinus leucomaenis) in a fragmented landscape

Several habitat models have been proposed to predict population size for stream fishes and to guide habitat assessment and monitoring techniques. However, most models do not incorporate the potential advantage of molecular genetic markers. We conducted a field survey and microsatellite DNA analyses to quantify the relationships among genetic diversity, census/effective population size and habitat variables in fragmented populations of white‐spotted charr (Salvelinus leucomaenis). The census population size significantly increased with the stream length, the number of pools and a pool‐riffle sequence index, a proxy for channel‐unit habitat type complexity within reaches. Population density was correlated with the pool‐riffle sequence index only. Genetic diversity and effective population size were not correlated with the habitat variables or census population size. There was a lack of isolation‐by‐distance population structure in the studied populations. Our results suggest that stream length and the number of pools within reaches associated with habitat complexity are the habitat variables that explain the majority of variation in population size of white‐spotted charr. Our findings provide further evidence that census population size per se is a poor indicator of the inclusive genetic diversity within populations in a fragmented landscape.