The density, species composition, and possible change in the status of pack ice seals within the Weddell Sea were investigated during the 1997/1998 summer cruise of the RV „Polarstern“ (ANT-XV/3, PS48). Comparisons were made with previous surveys in the Weddell Sea where it was assumed that all seals were counted in a narrow strip on either side oft he ship or aircraft. A total of 15 aerial censuses were flown during the period 23 January – 7 March 1998 in the area bounded by 07°08' and 45°33' West longitude. The censused area in the eastern Weddell Sea was largely devoid of pack ice while a well circumscribed pack ice field remained in the western Weddell Sea. A total of 3,636 (95.4 %) crabeater seals, 21 (0.5 %) Ross seals, 45 (1.2 %) leopard seals and 111 (2.9 %) Weddell seals were observed on the pack ice during a total of 1,356.57 linear nautical miles (244.2 nm) of transect line censused. At a mean density of 21.16 1/nm2 over an area of 244.2 nm, it is the highest densities on record for crabeater seals, density of up to 411.7 1/nm2 being found in small areas. The overall high densities of seals (30.18 1/nm2) recorded for the eastern Weddell Sea (27.46 1/nm2, 0.27 1/nm2, and 0.66 1/nm2 for crabeater, leopard and Weddell seals respectively) is a consequence of the drastically reduced ice cover and the inverse relationship that exists between cover and seal densities. Ross seal densities (0.08 1/nm**2) were the lowest on record fort the area. It is suggested that seals largely remain within the confines of the pack ice despite seasonal and annual changes in its distribution. Indications are that in 1998 the El Niño has manifested itself in the Weddell Sea, markedly influencing the density and distribution of pack ice seals.

Densities were calculated by dividing (1) the area surveyed during each census = width of trip delineated by sighting bars (1-5 on both sides combined = 1023 ft or 0.1682 nm) multiplied by the toral length of census line for each census, by (2) numbers of each species encountered in during each census. Furthermore, at some points densities for individual frames were calculated to illustrate some points, but not for overall densities per census (i.e. Bester and Odendaal did not calculate densities for each frame, but the then the mean +- SD (SE for all the frames of a single species).