Optimal allocation of sampling effort in lake sediment studies

A sampling algorithm requiring only bathymetric and catchment parameters has been derived from several recent empirical developments and should drastically reduce the number of sites and replicates necessary to characterize the distribution, trace element concentrations, and sediment accumulation rates of lake sediments. A test of this algorithm with sediments from 96 sites in three basins of morphometrically complex Jack Lake, Ontario, demonstrates the utility of this approach. A priori application of the algorithm would have reduced the number of sample sites on our sampling grid from 96 to 22 (a reduction of 77%), without losing precision. This reduction in sample sites (and in time and money expended) would allow for sampling additional lakes or sites of local significance (e.g., river or creek mouths, pollution point sources), or measuring additional parameters. The relationships between lake morphometry, sediment distribution, and sediment texture, along with the relationships between sediment texture and inter- and intra-site variability, can also be used to extrapolate sediment characteristics from a few samples to the whole lake or to predict the zone to which sediment is focused.