General models of the spatial distribution of porpoises require representative data and parsimony: Comment on Skov & Thomsen (2008)
Tougaard, J.; Wisz, M. (2010). General models of the spatial distribution of porpoises require representative data and parsimony: Comment on Skov & Thomsen (2008). Mar. Ecol. Prog. Ser. 399: 295-297. http://dx.doi.org/10.3354/meps08471
In: Marine Ecology Progress Series. Inter-Research: Oldendorf/Luhe. ISSN 0171-8630; e-ISSN 1616-1599, more
|   |
| Keyword |
|
| Author keywords |
Line transect survey; Phocoena phocoena; Spatial modelling; Offshorewind farms; Dynamic habitat modelling |
| Abstract |
The small-scale distribution of wide-ranging marine top predators has important implications for environmental management. Skov & Thomsen (2008; Mar Ecol Prog Ser 373:173-186) concluded from a study on Horns Reef, North Sea, that tidally driven upwelling is the most important factor governing distribution of harbour porpoises. For 3 principal reasons such a conclusion cannot be reached through their analysis: (1) the data used are unlikely to be representative of natural and undisturbed conditions of the habitat; (2) the selection of explanatory variables probably led to spurious correlations with no general explanatory value; and (3) there was no assessment of statistical uncertainty, model fit, or parsimony, nor any model validation, and thus no indication of the robustness of the results. Therefore, there is no statistical support for any general conclusions regarding the factors explaining small-scale distribution of harbour porpoises. |
|
