|Comparison of methods for determining key marine areas from tracking data|Tancell, C.; Philips, R.A.; Xavier, J.C.; Tarling, G.A.; Sutherland, W.J. (2013). Comparison of methods for determining key marine areas from tracking data. Mar. Biol. (Berl.) 160(1): 15-26. hdl.handle.net/10.1007/s00227-012-2050-2
In: Marine Biology. Springer: Heidelberg; Berlin. ISSN 0025-3162, more
|Authors|| || Top |
- Tancell, C.
- Philips, R.A.
- Xavier, J.C.
- Tarling, G.A.
- Sutherland, W.J.
There is an urgent need to identify key marine areas for conservation, particularly in the high seas. A range of techniques have been applied to tracking data from higher predators, particularly seabirds and pinnipeds, to determine the areas of greatest use. This study compared three commonly used methods—kernel, first-passage time and state-space modelling—and a new approach, minimum displacement rate, for the analysis of data from the wandering albatross Diomedea exulans of Bird Island, South Georgia, tracked during the chick-rearing period. Applied to a single track, these four models identified similar marine areas as important. The greatest similarity in areas identified occurred when model assumptions were shared (such as slow speed indicating spatial preference) even when methods modelled these assumptions differently (e.g. Bayesian inference versus cumulative density surface). A gridded overlap approach applied to all tracks revealed core areas not apparent from results of any single analysis. The gridded approach also revealed spatial overlap between methods based on different assumptions (e.g. minimum displacement rate and kernel analysis) and between individuals. Although areas identified as important by kernel and first-passage time analysis of a single track were biased towards resting locations during darkness, this does not negate the requirement for their protection. Using the gridded overlap approach, two distinct core regions were identified for the wandering albatross; one close to the breeding colony and another 800 km to the North–West in the high seas. This convenient and pragmatic approach could be applied to large data sets and across species for the identification of a network of candidate marine protected areas in coastal and pelagic waters.