Horning and Mellish: Predation on Eumetop/as /ubatus by Somniosus pacificus in the Gulf of Alaska 
303 
Figure 3 
Antemortem and postmortem temperatures recorded by life history transmitters 
(LHX tags) for 11 mortality events detected in juvenile Steller sea lions ( Eumetopias 
jubatus) with rapid onset of transmissions. Temperatures were recorded by LHX 
tags in 30-min increments and subsequently transmitted through the Argos satel- 
lite system. All events occurred in the Gulf of Alaska between 2008 and 2013 (for 
further details, see Table 1 and the Materials and methods section). Data from only 
one tag are shown for each animal. If data from both tags were available, then the 
data patterns from one tag that are not shown here did match the patterns of the 
other tag shown here. Because these 11 events occurred on different dates, data 
from one tag from each event were arbitrarily assigned to one of 3 sections of this 
graph solely for clarity. Gray diamonds indicate the regional estimates of sea-surface 
temperatures (for sources, see Table 1). All tags sensed light or air by or before they 
began to transmit at the end of each temperature record; arrows indicate the end 
of records and onset of transmissions for each tag. The gray shaded area at the top 
of this graph indicates the normal range of abdominal temperatures for healthy sea 
lions (Fig. 1). 
decline in peak temperature is consistent with the rate 
observed in laboratory simulations (n = 15) of tag trans- 
fers to water: 0.88°C min -1 (SD 0.14; Fig. 2C; Horn- 
ing and Mellish, 2009). In all 11 cases, tags uplinked 
at the first opportunity after each had sensed ambient 
light, air, or both, indicating immediate tag extrusion 
at sea. This notion was confirmed by the locations of 
initial uplinks at sea. In all 11 events, the recorded 
postmortem temperatures coincided with regional SST 
estimates (Fig. 3; Table 1). 
One type-III event was observed. This event yielded 
only a single complete data set with temperatures that 
gradually declined (peak AT'=-0.16°C min -1 ) over 18+ 
h to 4.51°C, close to the regional SST of 3.98°C (Fig. 4; 
Table 1). Transmissions began 16 d post mortem. The 
algor mortis model yielded a mass estimate of only 24 
kg, significantly below the mass of the animal at re- 
lease (146 kg; model output shown in Fig. 4) and the 
estimated mass at the age of death (158 kg), and indi- 
cating partial dismemberment (Horning and Mellish, 
2009). 
Of the remaining 3 events, 2 events fell into a type- 
IV scenario. In the first event (TJ52), only a single 
transmitter uplinked. This tag recorded a precipitous 
temperature drop (peak A7 1 = -0.98°C min -1 ) to 5.53°C, 
but the tag did not sense light or air and did not trans- 
mit until 11 d post mortem (Table 1; Fig. 5). Multiple 
uplink temperatures of 11.98°C corresponded to the re- 
gional SST estimate of 11.90°C, more than double the 
recorded postmortem temperatures. 
In the second event (TJ64), both tags sensed precipi- 
tous drops (peak A7 , =-0.93°C min -1 ) to a temperature 
