58 
Fishery Bulletin 120(1) 
multiple times or singularly within a year and whether 
they occurred in consecutive years. We anticipated find- 
ing multiple rises over the course of several days if batch 
spawning was occurring, as has been observed for Atlantic 
halibut (Murphy et al., 2017). A singular rise would 
indicate total spawning, or a single batch, as has been 
observed for captive Greenland halibut (Dominguez-Petit 
et al., 2013). Evidence of female Greenland halibut having 
spawning rises in consecutive years (from tagged fish that 
were at liberty for more than 1 winter) would indicate that 
spawning can occur every year despite oocyte development 
requiring a period of 2 years (Kennedy et al., 2011; Rideout 
et al., 2012). Such a reproductive scenario is unusual but 
still considered that of a total spawner. In the first year, a 
female produces oocytes, but no spawning occurs. In the 
second and subsequent years, 2 cohorts 
of oocytes are present in the ovaries (a 
small cohort, with less than 1 year of 
Results 
As of 2020, 22 of the 297 (7.4%) Greenland halibut 
implanted with archival tags have been recovered and 
returned to the Alaska Fisheries Science Center. The tags 
from 13 of these fish (6 males, 5 females, and 2 fish of 
undetermined sex) provided usable data through at least 
one presumed spawning period, October—March, with 
6 fish at liberty for multiple years (Suppl. Table) (online 
only). Spawning behavior was identified and characterized 
for 11 tagged Greenland halibut (6 males, 3 females, and 
2 fish of undetermined sex). Release and recapture loca- 
tions were across the BSAI region (Fig. 1). 
Extreme ascents and descents of tagged fish occurred 
throughout the year, but the top 0.1% of vertical movements 
development, and a large cohort, with 
more than 1 year but less than 2 years 
of development); the large cohort is 
spawned in the second year, and spawn- 
ing can occur each subsequent year. 
The depths of the apexes of spawning 
rises for female Greenland halibut were 
carefully estimated because these depths 
were assumed to be the depths at which 
eggs were released. Similar assumptions 
have been made for the family Pleu- 
ronectidae because the only observed 
flatfish spawning behavior has included 
release of eggs at the apex of spawning 
rises (see review in Seitz et al., 2005). 
Estimation of the apex depth varied with 
the sampling frequency of the tag, given 
that an increase in sampling frequency 
increases the probability of capturing the 
actual apex. When the sampling interval 
was 1 min, the minimum depth achieved 
was the estimate of the apex. For sam- 
pling intervals greater than 1 min (i.e., 
4 and 15 min), depths were interpo- 
lated to a 1-min frequency by using the 
intersection of 2 simple linear models, 
the ascent and the descent. The esti- 
mated apex was determined by setting 
these equations equal to each other (i.e., 
the depth at which the 2 lines cross). 
This interpolation was not intended to 
exactly replicate the trajectory of these 
fish, although it was expected to approx- 
imate the depth of the apex better than 
taking the observed minimum depth 
from a low-frequency sampling inter- 
val (Suppl. Figs. 3 and 4) (online only). 
Additional information related to infer- 
ring and interpolating spawning rises is 
available in Supplementary Materials 
(online only). 
Aleutian Islands 
175°W 
Figure 1 
Map of the study area in the eastern Bering Sea and Aleutian Islands show- 
ing locations where Greenland halibut (Reinhardtius hippoglossoides) were 
tagged and released (gray circles) and later recovered (black circles). The tag 
numbers of recovered fish are provided for each location. Dashed lines indicate 
straight-line movement trajectories, and solid gray lines indicate isobaths at 
200, 500, 1000, 1500, and 2000 m. The inset shows the location of the study 
area in relation to Alaska. 
