Landlocked Arctic charr (Salvelinus alpinus)
populations in sub-Arctic and Arctic Greenland lakes
were sampled with multi-mesh-sized survey gillnets. The
study covered a range of small shallow lakes (0.01 km2,
maximum depth <3.3 m) to large deep lakes (43 km2,
maximum depth >200 m). Arctic charr were found in
one to three di€erent forms in lakes with maximum
depths >3 m. A dwarf form occurred in all lakes in-
habited by Arctic charr and was the only form in lakes
with maximum depths <8 m. In deeper lakes with
maximum depths >20 m and a surface area <0.5 km2,
larger charr were found, although in low numbers, the
length-frequency distribution being unimodal with a tail
towards large sizes. In lakes with a maximum depth
>20 m, large-sized charr were more abundant, and the
length-frequency distribution of the population was bi-
modal, with a ®rst mode around 10±12 cm and a second
mode around 26±37 cm. In a single large and deep lake, a
distinct medium-sized pelagic zooplankton-eating charr
form occurred. Maximum size of individual charr was
signi®cantly positively correlated with lake maximum
depth and volume, and the mean size of large-sized charr
was signi®cantly positively correlated with lake volume.
Our study indicates that the charr population structure
became more complex with increasing lake size. More-
over, the population structure seemed to be in¯uenced
by lake-water transparency and the presence or absence
of three-spined stickleback (Gasterosteus aculeatus).