We present the results of an alternative method of estimating krill abundance, and producing maps of krill distribution, from the data collected during the 2000 CCAMLR survey of the Scotia Sea. Using a probabilistic Bayesian Maximum Entropy (MaxEnt) technique, density values for the unsurveyed off-transect portions of the survey area were found by interpolation, and thus values for total biomass across the survey area, and within individual small-scale management units (SSMUs), can be inferred. The MaxEnt formalism allows an objective choice of the parameters of the interpolation, and hence an objective choice of the most probable reconstruction of krill distribution, given the data. The Bayesian framework also allows an intrinsic calculation of the error in any resulting density estimate. During the survey, data were integrated at approximately 1 nm intervals, resulting in 9,586 observed data points. We therefore inferred the krill density on a 1 nm × 1 nm grid of pixels. Density values (g m-2) were calculated for all 1,726,332 pixels in the reconstructed area. The MaxEnt maps of the density distribution of krill across the survey area are presented. The resulting total biomass inferred for the survey area was 207.98 million tonnes, with a standard deviation of 10.08 million tonnes. The inferred biomass of krill in each of the SSMUs was: Antarctic Peninsula Pelagic Area (APPA) 65.192 Mt; Antarctic Peninsula West (APW) 0.753 Mt; Drake Passage West (APDPW) 35.874 Mt; Drake Passage East (APDPE) 0.397 Mt; Bransfield Strait West (APBSW) 0.167 Mt; Bransfield Strait East (APBSE) 1.381 Mt; Elephant Island (APEI) 10.496 Mt; Antarctic Peninsula East (APE) 0.003 Mt; South Orkney Pelagic Area (SOPA) 30.799 Mt; South Orkney West (SOW) 0.988 Mt; South Orkney North East (SONE) 0.005 Mt; South Orkney South East (SOSE) 0.243 Mt; South Georgia Pelagic Area (SGPA) 57.194 Mt; South Georgia West (SGW) 1.742 Mt; South Georgia East (SGE) 1.742 Mt.
Número de documento:
Punto(s) de la agenda