Siwicke et al.: Spawning behavior of Reinhardtius hippoglossoides in the Bering Sea and Aleutian Islands 57 
following D’yakov (1982). Results of later work indicate a 
larger L;9, but the study had limited sample sizes and its 
Ls9 has not been adopted for stock assessments (Cooper 
et al., 2007). Following the acknowledgment that there are 
2 separate cohorts of oocytes simultaneously developing 
(Kennedy et al., 2011), the large cohort to be spawned this 
year and the small cohort to be spawned in the following 
year, maturity ogives of 2 different stocks of Greenland hal- 
ibut have been reanalyzed with the small cohort of oocytes 
considered to be immature. In analysis of archival tag data 
from work in eastern Greenland, reclassifying females as 
having only developing oocytes (i.e., only the first cohort 
of oocytes that will not be mature in 12 months) led to a 
decrease in the estimate of spawning stock biomass of an 
average of 56%, with associated declines in total egg pro- 
duction; notably, estimates of L;,. increased from 63.8 to 
80.2 cm FL for one stock and from 61.2 to 74.1 cm FL for 
the other (Kennedy et al., 2014). Lengths of tagged female 
Greenland halibut that exhibited spawning behavior could 
be used to inform decisions regarding maintaining or reex- 
amining the L;,. used in stock assessments. 
Our objective was to assess whether we could identify 
spawning behavior in data from archival tags implanted 
in Greenland halibut, and if so, to characterize it. If 
spawning behavior can be detected in tag data, we could 
infer and clarify several aspects of the reproduction of 
Greenland halibut in the BSAI: 1) the timing of spawning, 
2) the depth at which eggs are released, 3) the frequency 
of spawning in a year (i.e., batch or total), and 4) whether 
individual females spawn in consecutive years. 
Materials and methods 
Data collection 
Archival tags have been implanted in Greenland hali- 
but since 2003 during the longline surveys conducted by 
the NOAA Alaska Fisheries Science Center in the BSAI 
region. Briefly, longlines are set along the continental 
slope in the BSAI, and Greenland halibut captured on 
predetermined hooks that are in good condition and lon- 
ger than 50 cm FL are placed in a tank with flowing sea- 
water prior to tagging. The minimum length threshold of 
50 cm FL meant that tagged fish were likely to be mature 
(Cooper et al., 2007). 
A total of 297 Greenland halibut had Lotek’ archival 
tags (Lotek Wireless Inc., Newmarket, Canada) inter- 
nally implanted, with the specific tag model varying by 
year (2003: model LTD-1250; 2005-2008: model LTD-1300; 
2011: model LAT-2800). Tags recorded temperature and 
pressure (converted to depth) at regular sampling fre- 
quencies between 1 and 15 min, depending on the year 
of deployment. The tag, surgical site, and tools were dis- 
infected by using 3% Betadine, tags were inserted into 
' Mention of trade names or commercial companies is for identi- 
fication purposes only and does not imply endorsement by the 
National Marine Fisheries Service, NOAA. 
the body cavity through a 2-cm-long incision made in the 
abdominal wall of the right side of the fish, and the inci- 
sion was closed with surgical staples. Fish were released 
in close proximity to their site of capture, and tags were 
recovered by commercial fishing vessels (trawl or long- 
line). Size (FL) of Greenland halibut was recorded at 
release, and FL, weight, and sex (determined with visual 
inspection of gonads) were recorded for some but not all 
tagged fish that were recaptured. 
Identified spawning behavior 
Archival tag data, specifically depths converted from 
pressure records, from tags implanted in Greenland hal- 
ibut were used to identify and characterize the timing, 
depths, and frequency of spawning behavior for this spe- 
cies. Depth data were analyzed to find spawning behavior 
defined as “abrupt spikes,” which consist of rapid ascents 
to an apex followed by rapid descents, as outlined in Seitz 
et al. (2005), Loher and Seitz (2008), and Murphy et al. 
(2017). We first filtered each fish’s depth record to the 
most extreme 0.1% of ascents and descents to focus on 
a smaller more manageable subset (Suppl. Fig. 1) (online 
only). Larger percentages (>1% and >5%) were explored 
but did not add any insight into spawning behavior. In 
addition, spawning rises are expected to occur over peri- 
ods of only hours to days in an individual’s year. When 
an extreme ascent closely preceded an extreme descent, 
the entire depth record for that day was further ana- 
lyzed. Greenland halibut may use more of the water col- 
umn than Atlantic and Pacific halibut (e.g., Albert et al., 
2012), and rapid ascents and descents could be related to 
other behaviors that are not related to spawning. Often, 
extreme ascents or descents were identified in isolation, 
or a descent was followed by an ascent; however, given our 
assumption that a spawning rise is a rapid ascent from 
the seafloor to spawn midwater followed by a rapid return 
to the seafloor, these other events were not further ana- 
lyzed (Suppl. Fig. 2) (online only). Among flagged events, 
sex-specific similarities aided in deciding which events to 
classify as spawning behavior (e.g., Murphy et al., 2017). 
The start date, time of day, and initial depth were iden- 
tified for the time directly before the initiation of the first 
spawning rise, and the end date, time of day, and final 
depth were recorded at the completion of the last spawn- 
ing rise, such that multiple rises would be encompassed 
in this period. The exact times when a rise was initiated 
and completed were determined on a case-by-case basis, 
by qualitatively and visually assessing the depth profile 
preceding and following rises to best capture the indi- 
vidual behavior. The time of day at which an individual 
reached a minimum depth during spawning rises (..e., 
the apex) was noted to assess whether there was a daily 
preferred time of spawning. Additionally, the maximum 
ascent and descent rate (meters per minute) occurring 
between the start and end times of a spawning event 
were also compared. 
Spawning rises of female Greenland halibut were 
further investigated to discern whether they occurred 
