Evaluating the influence of environmental and spatial variables on diatom species distributions from Melville Island (Canadian High Arctic)

Diatom species assemblages were identified and enumerated from the surface sediments of 45 lakes and ponds across a wide spectrum of spatial and environmental gradients on Melville Island, Nunavut/N.W.T, Arctic Canada. Whereas the most common taxa were similar to those recorded elsewhere in the Cana...

Full description

Bibliographic Details
Published in:Botany
Main Authors: Keatley, Bronwyn E., Douglas, Marianne S.V., Smol, John P.
Format: Article in Journal/Newspaper
Language:English
Published: Canadian Science Publishing 2008
Subjects:
Online Access:http://dx.doi.org/10.1139/b07-118
http://www.nrcresearchpress.com/doi/full-xml/10.1139/B07-118
http://www.nrcresearchpress.com/doi/pdf/10.1139/B07-118
Description
Summary:Diatom species assemblages were identified and enumerated from the surface sediments of 45 lakes and ponds across a wide spectrum of spatial and environmental gradients on Melville Island, Nunavut/N.W.T, Arctic Canada. Whereas the most common taxa were similar to those recorded elsewhere in the Canadian High Arctic, significant differences in assemblages existed between sites located in the different bioclimatic zones of Melville Island. For example, taxa recorded in the most lushly vegetated bioclimatic zone were similar to those found in lushly vegetated regions elsewhere in the Canadian Arctic Archipelago, and generally different from diatoms in the poorly vegetated regions on Melville Island. Of the measured environmental variables, pH, specific conductivity, surface area, elevation, and chlorophyll a explained significant portions of the variance in diatom assemblage composition at the scale of the entire island. However, only total dissolved nitrogen was an important explanatory variable within the most lushly vegetated bioclimatic zone. The strongest ecological relationship was between diatoms and pH, and regression and calibration by weighted averaging produced predictive models with r 2 boot of 0.432 to 0.746 and RMSEP of 0.341 to 0.242. Spatial factors were of little importance, confirming that diatoms are not likely to be dispersal limited, at least at the landscape scale explored in this study.