CCAMLR

Diet studies are a key component of the CCAMLR predator monitoring program as they provide direct assessments of the prey types and amounts of import to predators, which, in turn, are hypothesized to influence variability in related parameters such as breeding success, foraging trip durations and chick fledging weights. Given the importance and interdependence of diet studies to monitoring work, we compared data on stomach weights and prey types among three US AMLR land-based predator study sites in the Antarctic Peninsula region where different methodologies were used to select, pump and sort the diet samples. At Palmer Station, Adélie penguins selected for diet sampling were not confirmed to be breeding, while all Adélie and chinstrap penguins selected for diets sampling at Admiralty Bay and Cape Shirreff were confirmed breeders. Adélie penguins at Palmer Station had significantly smaller mean stomach weights (349 vs. 550 g) and a significantly higher proportion of birds with small

This paper uses power analysis and bootstrap functions to examine the importance of sample size in detecting differences between two years of fur seal trip duration data. Female foraging trip duration for the first six trips to sea for 1997/98 (n = 30) and 1998/99 (n = 31) at Cape Shirreff, Livingston Island was used to calculate minimum sample sizes necessary to detect differences from year to year. A bootstrap function was used to examine the effect of sample size on variance. Currently CCAMLR-Ecosystem Monitoring Program Standard Method C 1.2 for measuring fur seal trip duration sets the sample size at 40 females. Based upon my results I propose incorporating some flexibility in sample size and suggest changing the protocol from 40 to 25 – 40 females.

Fatty acid signature analysis (FASA) makes use of specific fatty acids, as well as entire profiles, to study dietary relationships at different trophic levels. Previously, FASA has been used in marine ecosystems in which diet determination by more direct methods is difficult and sometimes misleading. This study examined fatty acid profiles in milk from two species of pinniped from the Southern Ocean, that were expected to have highly contrasting diets. Milk samples were collected from Antarctic fur seals (Arctocephalus gazella) in 3 consecutive years, from 1991- 1993 (n=72) and from Southern elephant seals (Mirounga leonina) in 1988 (n=53) at South Georgia. Lipids were extracted and fatty acid profiles determined by temperature programmed gas chromatography. Possible prey species collected from waters around South Georgia were also analysed. Cluster analysis as well as classification and regression trees (CART) indicated that profiles from fur seals and elephant seals were significantly different. Southern elephant seal data could be distinguished from Antarctic fur seals by lower levels of the fatty acids 16:4 n1, 18:2 n6, 18:4 n3, 18:4 n1 and 20:5 n3 and by higher levels of 18:0, 18:1 n9/ n11 (i.e. 18:1 n9 co- eluting with 18:1 n11) and 20:1 n9. Fatty acid signatures from the milk of Antarctic fur seals were closest to krill and fish species that were also known to feed on krill. Southern elephant seal fatty acid profiles were closest to species that are not known as krill predators such as larger notothenids and myctophids. The fatty acid profiles of Antarctic fur seals showed considerable inter- and intra- annual variability which was congruent with diet variability detected using scat analyses. Southern elephant seals showed little variation in profile through lactation. In contrast to previous diet analyses based on examination of stomach contents, the results from FASA were consistent with a fish based diet for Southern elephant seals.

The necessity of synoptic survey covering the South Sandwich Islands that are an integral part of Area 48 is discussed. It is shown that in this case there shall be obtained for the first time for the recent twenty years the material which allows to have an idea of krill simultaneous distribution over the whole water area of Subareas 48.1-48.4 covering a vast zone of krill drift and that of concentration of fishing grounds between the Antarctic Peninsula and the most southern Sandwich Islands up to 20° W. It is emphasized that in this way there shall be obtained the data that are important both to acquire knowledge in appropriateness of krill horizontal distribution over the whole Area 48, and to reveal interrelations between individual Subareas of the same. It is given a supplement to the most up-to-date survey design, taking into account the operation by Russian R/V ATLANTIDA in Subarea 48.4, as being a part of the forthcoming synoptic survey in Area 48.

Backscattering measurements of 14 live individual Antarctic krill (Euphausia superb a) were made at a frequency of 120 kHz in a chilled insulated tank at the Long Marine Laboratory in Santa Cruz, CA. Individual animals were suspended in front of the transducers, were only loosely constrained, had substantial freedom to move, and showed more or less random orientation. One thousand echoes were collected per animal. Orientation data were recorded on video. The acoustic data were analyzed and target strengths determined from each echo. A method was developed for estimating the three-dimensional orientation of the krill based on the video images and was applied to five of them, giving their target strengths as functions of orientation. Scattering models based on a simplified distorted-wave Born approximation (DWBA) method were developed for five animals and compared with the measurements.
Both measured and modeled scattering patterns showed that 120 kHz acoustic scattering levels are highly dependent on animal orientation. Use of these scattering patterns with orientation data from shipboard studies of E. superba gave mean scattering levels approximately 12 dB lower than peak levels. These results underscore the need for better in situ behavioral data to properly interpret acoustic survey results. A generic E. superb a DWBA scattering model is proposed that is scalable by animal length. With good orientation information, this model could significantly improve the precision and accuracy of krill acoustic surveys.

Acoustic scattering experiments involving simultaneous acquisition of broadband echoes and video footage from several Antarctic krill were carried out to determine the effect of animal orientation on echo spectral structure. A novel video analysis technique, applied to extract krill angle of orientation corresponding to each insonification, revealed that echo spectra from krill near broadside incidence relative to the incident acoustic wave exhibited widely spaced, deep nulls, whereas off-broadside echo spectra had a more erratic structure, with several closely spaced nulls of variable depth. The pattern of changes in echo spectra with orientation for the experimentally measured acoustic returns was very similar to theoretically predicted patterns based on a distorted wave Born approximation (DWBA) model. Information contained in the broadband echo spectra of the krill was exploited to . invert the acoustic returns for angle of orientation by applying a newly developed Co variance Mean Variance Classification (CMVC) approach, using generic and animal-specific theoretical and empirical model spaces. The animal-specific empirical model space was best able to invert for angle of orientation. The CMVC inversion technique can be implemented using a generic empirical model space to determine angle of orientation based on broadband echoes from individual zooplankton in the field.

Refinements have been made to the multiple-frequency method that was recently developed to improve the accuracy and precision of in-situ target strength (TS) measurements using split-beam echosounders (Demer et al., J. Acoust. Soc. Am. 105(4) 2359:2376 (1999)]. The multiple-frequency method improves the rejection of unresolvable and constructively interfering target multiples by combining synchronized signals from two or more adjacent split-beam transducers of different frequencies which are not integer multiples of each other. In this study, the method itself was improved by: 1) optimizing the accuracy and precision of the angular and range measurements of the individual frequency detections; 2) more precisely determining the relative three-dimensional (3-D) locations (x, y, and z) and angular orientations (pan and tilt) of the transducers and thus the positional transformation; and 3) increasing the range resolution of one or more of the frequencies. These improvements are demonstrated through controlled test tank experiments using 38 and 120 kHz split-beam transducers and a 200 kHz single-beam transducer. Tolerances for matching target positions at two or more frequencies were determined to be ±0.5 m in radial range and ±0.7° off-axis (compared to ±l.0 m ±l.5° in the previous study). These experimental results indicate that such careful application of the multiple-frequency TS method can reject 100% of multiple targets while allowing 90% of the resolvable single-targets to be measured