Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

Timothy Loher

doi:10.1186/s40317-022-00288-w

Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

Loher, Timothy 2022-06-02 00:00:00 Background: Understanding connectivity is critical to the management of exploited fish stocks, but migratory dynamics of Pacific halibut (Hippoglossus stenolepis) in the Bering Sea and Aleutian Islands region are not well-under - stood. In the current study, 145 Pacific halibut ≥ 82 cm fork length were tagged with Pop-up Archival Transmitting (PAT ) tags to evaluate interannual dispersal, seasonal migration, and depth-specific habitat use. Results: Endpoint locations obtained after 1 year at liberty (n = 79), fishery recoveries after 2–3 years at liberty (n = 5), and at-liberty geopositions based on light data (n = 5313 estimates from 109 fish) indicated geographically distinct movement patterns: Pacific halibut tagged in the Western and Central Aleutian Islands remained within the island groups in which the fish had been tagged; fish in the eastern Bering Sea remained in that ocean basin, moving among International Pacific Halibut Commission (IPHC) regulatory areas and into Russian waters; those tagged south of Unimak Pass in IPHC Regulatory Area 4A displayed the greatest amount of emigration, dispersing eastward both seasonally and interannually to as far south as Washington State. Analysis of daily maximum depth and temperature data from 113 individuals demonstrated group-level variation in summer temperatures experienced by the fish and in the timing, duration, and synchrony of movement to deep-water wintering grounds. Conclusions: Depth-specific habitat use was suggestive of regionally explicit migratory contingents, while interan- nual dispersal patterns were consistent with the existence of multiple functional spawning units. The results may guide future research to examine cross-basin connectivity in the Northern Bering Sea and provide inputs for numeri- cal modelling of individual movements, larval advection, and recruitment analyses. Keywords: Hippoglossus stenolepis, Migration, Spawning, Satellite tagging, Bering Sea Okhotsk, and southward into the northern Sea of Japan. Background These fish are highly migratory, with complex and widely Pacific halibut (Hippoglossus stenolepis) is a wide-rang - dispersing life-history stages [1]. The species supports ing Pleuronectid flatfish species distributed from north - substantial subsistence, recreational, and commercial ern California, USA, northward throughout the Gulf of fisheries [2] and although managed as a single unit stock Alaska and Bering Sea, westward into the Russian Sea of throughout US and Canadian waters [3] research con- ducted in North American waters suggests internal pop- *Correspondence: tim.at.martingale@gmail.com ulation structure at roughly ocean-basin scales [4]. Data collected using Pop-up Archival Transmitting (PAT) tags Martingale Marine Ecological Research, 7019 14th Ave NE, Seattle, WA 98115, USA are consistent with the hypothesis that mature fish have © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Loher Animal Biotelemetry (2022) 10:18 Page 2 of 21 a higher probability of remaining within the basins of the known to conduct seasonal spawning migrations that are eastern Pacific Ocean (defined as the Gulf of Alaska, east - composed of both onshore–offshore movement [5, 12, ern Bering Sea, and the Aleutian Islands region) to spawn 13] and large-scale (> 1000 km) alongshore redistribution than dispersing among basins to do so [4, 5]. Consist- [5, 13–16]. Without constraining recoveries to a relatively ent with this, although population genetic analyses have short window representing the summer feeding period demonstrated little genetic differentiation throughout [12] or having the ability to evaluate each fish’s individ - North American waters [6, 7], best-available evidence ual’s location within its seasonal migration trajectory, suggests that the western Aleutian Islands may support seasonal movement may be confounded with interannual a population component that is significantly different dispersal, whether ontogenic [10] or due to adult straying than in either the eastern Pacific Ocean [6, 7] or in Asian [17]. Resolving among forms of dispersal can be impor- waters of the southern Sea of Okhotsk [7]. tant for understanding how fish movements interact with However, neither existing satellite-tag data nor genetic target [sensu 18, 19] and non-target [sensu 20, 21] fisher - analyses provide insight into connectivity at the annual ies, and for properly defining regional stock components to multi-annual scales over which Pacific halibut fisher - [sensu 3] intended to reflect effective spawning biomass ies are managed and prosecuted or the decadal-scale units [sensu 4]. processes comprising the ontogeny of a species in which In the current study, interannual connectivity in Pacific 8–16 year-olds are the most-represented demographic halibut was investigated using PAT tags. In brief, PAT in the directed fishery [3] and is known to live in excess tags are electronic tags that contain an automated release of 50 years [8]. Satellite tagging has primarily investi- mechanism, sensor package, and satellite-broadcast capa- gated seasonal dispersal (i.e., movements executed over bilities, allowing for environmental data to be collected the course of 6–7 months), while genetic studies speak while attached to the host fish. Recovery/broadcast dates to reproductive isolation over hundreds to thousands of (and, hence, period at liberty) may be pre-specified, and years. As such, there is a need to conduct adult connec- the tag’s final position is determined by the receiving sat - tivity studies ranging from interannual to generational- ellite [22]. This has the advantage of allowing for determi - scale to bridge this gap. nation of final location even if fish move to areas absent Perhaps the most substantial effort to better under - of fishery effort, such as where fisheries are excluded stand interannual dispersal in Pacific halibut was a large- (e.g., the IPHC Closed Area; Fig. 1), or where reporting scale Passive Integrated Transponder (PIT tag) study of physical recoveries is reduced (e.g., in Asian waters conducted along the North American coast between for fish tagged in North America). In addition, locations 2001 and 2009 [9]. Over the course of that study, 67,000 during time at liberty can be inferred from archived light fish were tagged and > 3000 were recovered via intensive data [sensu 23] and may allow for state-space modelling dockside sampling. Statistical modeling of the recovery of movement [24–26] that cannot be achieved using con- distributions, along with analyses derived from a long his- ventional mark-recapture data. PAT tags have been used tory of conventional tagging [10, 11], provided consider- to investigate seasonal movements in Pacific halibut; in able insight regarding migration among geographic areas, particular, identifying key spawning grounds in the BSAI including relationships between movement rates and fish region and their connectivity to components of the com- size. However, these results were subject to limitations mercial fishery [13]. However, due to their relatively high due to the nature of the tags and the recovery mecha- cost (~ $4000 US), the distribution of such releases has nism [1]. Conventional tag recoveries (including PIT been patchy and has lacked coverage in one region that tags) are dependent primarily upon commercial fisher - may be key to understanding basin-scale stock structure ies and, thus, relatively little information can be obtained in Pacific halibut: adjacent to Unimak Pass (Fig. 1), which in regions, where fishing effort is low or within which represents a primary connection between the Eastern the mixing of tagged fish into the fishable population is Bering Sea (EBS) and Gulf of Alaska (GOA), from both incomplete [9]. As a result, limited insight was obtained an oceanographic [27] and biological [28, 29] perspective. regarding dispersal within the Bering Sea and Aleutian Here, we use satellite-generated final positions, fish - Islands (BSAI) region or between the BSAI region and the ery-recovery locations, at-liberty longitudes estimated Gulf of Alaska (GOA). Also, the PIT tags were implanted from ambient light data, and depth trajectories during subcutaneously with no external markings placed on the time-at-liberty [sensu 12] to infer regional variation in fish; the tags could only be detected electronically. Thus, interannual dispersal, the timing and duration of sea- the location data were limited to the scale of entire fish - sonal migrations, and depth-specific regional habitat ing trips: i.e., it was known within which commercial use of Pacific halibut in the BSAI region and in the offload each fish had been recovered, but not the precise western GOA adjacent to Unimak Pass. recapture location or date. In addition, Pacific halibut are L oher Animal Biotelemetry (2022) 10:18 Page 3 of 21 Fig. 1 International Pacific Halibut Commission regulatory areas and geographic features referenced in this manuscript. Note that the western Aleutian Islands (i.e., the Near and Rat Islands) are located in the eastern hemisphere, with IPHC Regulatory Area 4B crossing the Antimeridian at approximately Amchitka Pass Mk10 PAT tags measured 170 mm in length and Methods 40 mm in maximum diameter, with a plastic-coated Tag deployments braided-cable antenna protruding from the distal end. A total of 145 Pacific halibut were tagged with either PSAT Flex tags measured 131 mm in length and 42 mm Wildlife Computers (Redmond, Washington, USA) in maximum diameter, also with a distal antenna. The Mk10 or with Lotek Wireless (St. Johns, Newfound- tags were programmed to record depth (at a resolu- land, Canada) PSAT Flex Pop-up Archival Transmitting tion of 4 m) every 30 s and ambient light levels every (PAT) tags during the boreal summers (June–August) minute and, upon surfacing, to transmitting their data of 2008 and 2009 (Mk10), and during June 2016 to the US National Oceanic and Atmospheric Admin- (PSAT Flex) (Table 1). All fish were captured during istration’s polar-orbiting satellites, administered by the the IPHC’s Fishery-Independent Setline Survey (FISS) Advanced Research and Global Observation System [30] using benthic longline gear rigged with 16/0 cir- (Argos). For tags that were not physically recovered cle hooks at 5.5 m spacing, baited with chum salmon prior to their programmed broadcast dates, archived (Oncorhynchus keta), and soaked for approximately 6 h environmental data were transmitted as aggregated prior to retrieval. Fish that were in excellent condition (“binned”) depth data, depth profiles, and light-based and of commercially legal size [i.e., ≥ 82 cm fork length twilight curves that could subsequently be used to (FL)] were tagged at pre-selected and regularly spaced define habitat use and estimate location during time- stations throughout the BSAI and far-western GOA at-liberty. Depth data were binned into consecutive (Fig. 2). Individual fish were randomly selected at each 8-h blocks that included the minimum and maximum station to achieve a tagged demographic that was rep- resentative of the surveyed population. Loher Animal Biotelemetry (2022) 10:18 Page 4 of 21 Table 1 Tag deployment data Tagging region Date range(s) Sample sizes FL range (mean ± SD) Total Displ. Light Depth Usable Western Aleutian 06 Jun–31 Aug, 2008 22 18 20 20 20 84–171 (106.9 ± 20.8) Central Aleutian 28 May–03 Jun, 2008 18 15 16 16 16 83–129 (95.2 ± 13.0) Transition Zone 18 Jun–13 Jul, 2008 24 17 23 23 23 82–122 (94.4 ± 10.87) EBS Shelf Edge 10 Jun–01 Aug, 2008; 49 19 32 32 34 82–135 (97.7 ± 17.7) 11–26 Jun, 2016 30 Jun–20 Jul, 2008; 32 9 18 22 23 82–139 (119.9 ± 28.6) EBS Shelf Islands 06 Jun–21 Aug, 2009 Dates over which fish were tagged, total number of tags deployed, number of tags that either reported remotely after 1 year at liberty or were recaptured multiple years after deployment (Displ.), produced light data suitable for estimation of at-liberty longitude or daily maximum depth data and, therefore, produced data used in analyses (Usable); and the range and mean forklengths of the fish included, for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the Aleutian Islands and Eastern Bering Sea (EBS), and a transition zone between those ecosystems and the western Gulf of Alaska depths experienced by the fish during each binning subunits at fishery-relevant scales. In particular, prior period. migratory [4] and genetic [7] research has suggested that During tagging, the fork length of each tagged fish was east–west movement of Pacific halibut may be limited recorded and tags were secured to the fish via 15–18 cm by deep-water passes through the Aleutian Island Chain leaders of 130-kg test nylon monofilament coated with (i.e., at Amchitka and Amukta Passes; Fig. 1). Tagging adhesive-lined polyolefin, anchored to the fish using studies have demonstrated different rates and distances surgical-grade titanium darts that were inserted through of dispersal for Pacific halibut found north versus south the pterygiophores on the eyed side of the fish, roughly of Unimak Pass [9]. Analyses also suggest that fisheries 2.5 cm medial to the dorsal fin. Tag assemblies weighed of shallow island ecosystems in the Eastern Bering Sea approximately 120 g in air, representing 0.15–1.5% of (EBS) may display different dynamics than farther off - the initial body weights of the tagged fish as estimated shore [32, 33]. Thus, for the subsequent analyses, data from their lengths [31]; however, the tags were slightly were grouped by tag-deployment location within geo- (i.e., < 5 g) buoyant in water. The tags were programmed graphically distinct regions (Fig. 2): (1) the Western Aleu- to detach and report during the non-spawning (feed- tian Islands, representing all waters south of 55.5° N lat. ing) period, 365 days after deployment, thereby produc- and west of a line extending from 51° N lat. by 180° W ing endpoint data representing interannual dispersal lon. (i.e., Amchitka Pass) to 54° N by 178° W; (2) the Cen- that was absent of seasonal migration to the greatest tral Aleutian Islands, representing waters south of 53° N degree possible. Mk10 PAT tags were programmed to and between 171.7° W (i.e., Amukta Pass) and a line report within 2 days of the occurrence of any premature extending from 51° N by 180° W to 54° N by 178° W; (3) release event (i.e., upon floating to the surface prior to the BSAI-GOA Transition Zone, representing southern the programmed date of detachment). PSAT Flex were IPHC Regulatory Area 4A and spanning 164–171.7° W so not equipped with surface-detection capabilities and, as to include all waters south, and within 40 km to the therefore, only initiated satellite broadcasts on the pro- north, of the Fox Islands (i.e., representing a region of grammed tag-detachment date. For both tag types, the mixing among the eastern Aleutian Islands, far-western dart-and-tether assemblies remained embedded in the GOA, and around Unimak Pass); (4) the Eastern Bering fish following tag detachment to serve as a conventional Sea Shelf Edge, representing waters of the continental tag that allowed for a third location to be obtained if the shelf edge from 135 to 500 m depth, north of 54.6° N, and fish were subsequently recaptured. Both the tag bodies spanning 165.4–178.5° W, and; (5) the Eastern Bering Sea and tethers were printed with tag numbers and contact Shelf Islands, representing shallow (< 105 m) water within information and unique identification numbers were also 40 km of St. Matthew Island and the Pribilof Islands. In engraved on the tagging darts. a Pacific halibut management context, the Western and Central Aleutian Regions comprise the western and east- Tagging regions ern halves, respectively, of IPHC Regulatory Area 4B; the The analyses contained herein are intended to describe Transition Zone comprises southern Regulatory Area dispersal patterns during the summer feeding period, 4A; the EBS Shelf Edge spans northern Area 4A and the within geographic regions that represent biological offshore extent of Regulatory Area 4D; and the EBS Shelf L oher Animal Biotelemetry (2022) 10:18 Page 5 of 21 Fig. 2 Deployment and endpoint locations of Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during the boreal summers of 2008, 2009, and 2016. Colors represent five geographic regions within which individuals were pooled for analysis. Only fish whose tags produced data are plotted. Circles indicate tagging locations, where closed symbols represent final locations that were amenable to analyses of interannual dispersal; open symbols are fish whose final locations were not included but which produced depth or light data. Open diamonds connected by solid lines indicate final positions after 1 year at liberty; dashed lines indicate fish that were physically recaptured after 2–3 years at liberty. The precise location of one fish that was recaptured in coastal Kamchatkan waters was unknown, and is denoted with a question mark. Note that Kamchatka and the Western Aleutians are located in the eastern hemisphere, with Antimeridian located at approximately Amchitka Pass in the south and the western Gulf of Anadyr in the north Islands are contained within Regulatory Areas 4C (Pri- successfully reported via satellite, positions were deter- bilof Islands) and 4D (St. Matthew Island) (Figs. 1 and 2). mined by the receiving satellite(s) from the Doppler shift of the transmitted radio frequency in successive uplinks Final locations, linear displacement, and interannual received during each satellite pass [22]. Each satellite- dispersal generated position estimate was assigned a “location Final fish locations were obtained for fish that were either class” by Service Argos that indicated the positional accu- recaptured in fisheries or whose tags reported through racy of the estimate; locations reported herein represent the Argos satellite system. Tag rewards were offered to the first broadcast for each tag for which positional accu - individuals who captured tagged fish, as well as for tags racy was reported to be < 1000 m. Mk10s that detached found adrift or awash following their detachment. Final earlier than their scheduled date were programmed to locations for the recaptured fish reported herein repre - broadcast upon surfacing if they remained at zero depth sent the coordinates that were provided to the IPHC by for a 2-day period; locations reported herein represent the individuals who returned those tags. For tags that the first position obtained thereafter. Tag drift that may Loher Animal Biotelemetry (2022) 10:18 Page 6 of 21 have occurred between tag surfacing and reporting was of individual twilight events can be biased relative to true not estimated. PSATs were not equipped with zero-depth twilight due to water clarity or cloud cover. In contrast, premature-release detection and were, therefore, pro- the mean of two similarly biased events during a single grammed to report only after 365 days at liberty. As such, day may still result in an accurate estimate of solar noon, reliable final locations for fish whose PSATs detached and so all days in which both sunrise and sunset could be prematurely could not be obtained and these locations characterized were included in the analysis. were omitted from analysis. Longitude was calculated from twilight data by: (1) For analyses of interannual dispersal, only fish whose computing local solar noon as the mean of the mid-sun- final positions were obtained after 1 full year at liberty rise and mid-sunset times; (2) adjusting local solar noon were included; herein defined as occurring within a win - to the nearest minute (i.e., at the same resolution as the dow of 360–370 days after tagging. Positions obtained light data) in accordance with the Equation of Time [34] after shorter or longer periods were omitted, because for the given date; (3) comparing the adjusted local noon they had the potential to be biased due to seasonal migra- to 1200 UTC (Coordinated Universal Time), wherein tion, which may be substantial in magnitude [4, 5]. Such each hour of offset between those two times indicated 15° endpoints may be more representative of spawning/win- of longitude relative to the Prime Meridian. tering locations than the locations to which those fish When longitude is known, latitude may be deter- would have migrated the following summer, whether mined based on day length on any given date. However, displaying homing behavior [14, 17], straying [17], or light-based latitudes were not estimated herein, because engaged in ontogenic redistribution [9]. The dispersal results for Pacific halibut have been found to be highly of individuals whose locations were obtained during the variable [35] due to reductions in estimated day length winter spawning period has been reported elsewhere that arise from cloud cover and water turbidity in the [4]. Herein, interannual displacement between mark and North Pacific Ocean. In addition, changes in the vertical reporting was calculated for each fish included in the distribution of the tagged fish from day-to-day and sea - analysis as the linear distance between its tagging loca- sonally can introduce intractable variance and bias into tion and either its physical recapture or first accepted twilight-derived estimates of local day length. satellite-derived reporting position, ignoring land masses Plots of light-based longitude estimates according to that might lie between those endpoints. date (i.e., demonstrating potential east–west disper- sal during time at liberty) were constructed for each At‑liberty position estimation fish and using regionally aggregated data. Inferring the Daily longitude estimates of tagged Pacific halibut dur - displacement of Pacific halibut from light-based longi - ing their times at liberty were estimated using archived tude is somewhat subjective, because factors that skew ambient light data. MK10 PAT tags were programmed light readings taken at depth can introduce variability to identify and broadcast “twilight events” in which light into those estimates, even for relatively stationary tags levels either increased or decreased at rates that were [35]. Isolated position estimates must be considered consistent with sunrise and sunset, respectively. Wildlife cautiously, whereas series of positions located consist- Computers’ proprietary software Global Position Esti- ently away from a point of reference are more likely to mator version 3 (GPE3) was used to extract the twilight indicate displacement of the fish. Individual outliers are data from each tag’s broadcast file, and each twilight often easy to identify within individual trajectories (e.g., event (i.e., a series of nine sequential light readings) was Fig. 3a); series of positions estimates were evaluated for visually inspected. Putative sunrise/sunset events that plausibility considering the magnitude of displacement did not exhibit smoothly sloping light levels or contained that they nominally suggested with respect to their dis- null values were rejected. In addition, twilight events tance from known release-recovery locations and other were rejected if the fish displayed a change in depth dur - light-based longitude estimates over the time elapsed ing the putative event such that the changing light levels between consecutive position estimates. Herein, series might have resulted from vertical movement as opposed of estimates that suggested smoothly trended east–west to true sunrise or sunset; e.g., light levels declining as a redistribution were considered plausible. Series of five fish descended. Curves were not rejected if a fish’s depth or more “clumped” estimates were evaluated for plausi- change was inverse to the slope of the twilight event; bility on the basis of their distance from prior and sub- e.g., light levels increasing as a fish descended. The lat - sequent positions. Implied displacement was considered ter would be consistent with crepuscular vertical migra- implausible if it could only have been reached from the tions and provide additional evidence that the event was nearest series of prior plausible position estimates by the truly sunrise/sunset. Days on which only one twilight fish having swum continuously at ≥ 2 m per second for event was identifiable were omitted, because the timing consecutive 24-h periods without resting. L oher Animal Biotelemetry (2022) 10:18 Page 7 of 21 Fig. 3 Light-based longitude estimates (upper panels) and daily maximum depth profiles (lower panels) during periods at liberty (2008 and 2009) of: a one Pacific halibut tagged in the Western Aleutian Islands, from date of tagging (top of upper panels) to date of final tag reporting (bottom of upper panels), and; b, c two Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the far-western Gulf of Alaska. All data were obtained via satellite transmission (i.e., the tags were not recovered). The approximate longitudinal span of International Pacific Halibut Commission regulatory areas is indicated by shading; note that the Western Aleutians are located in the eastern hemisphere, with the Antimeridian bisecting IPHC Regulatory Area 4B at approximately Amchitka Pass. The fish depicted in a remained in shallow water throughout the year and the light data provide no evidence that it departed its tagging location. However, note the single outlier position in early June of 2009: this likely represents a biased daylight curve, as the position is farther eastward than a Pacific halibut could likely swim during the time elapsed relative to the remainder of the position estimates. The fish depicted in b and c emigrated from their tagging region and had final tag-reporting locations in the Gulf of Alaska and coastal Washington State, respectively; the timing of their migrations was coincident with their movement to deep water in autumn In addition to plotting estimates over time, group-level histograms. Those distributions were inspected to evalu - at-liberty redistribution was evaluated using longitude ate whether they were described by a single mode or estimates that were standardized relative to each indi- multiple modes. For each mode that was identified, vidual’s known initial location, as: (Lon − Lon ), where mean, median, and skewness were calculated. Median i t Lon = the longitude at the initial (tagging) location and and skewness values near 0 would indicate little net Lon = light-based longitude estimate at time . That is, directional shift in at-liberty fish distributions relative to t t each light-based longitude estimate was expressed as known tagging locations; negative values would indicate being X° either to the west (negative values) or east (posi- tive values) of the fish’s known initial location. The data were then pooled by region and plotted as frequency Loher Animal Biotelemetry (2022) 10:18 Page 8 of 21 net westward redistribution; and positive values eastward annual mean depth occupied; and the shallow-water redistribution. phase defined as the remainder of the year. Maximum depth data for each fish were then inspected to determine Seasonal depth distributions individual deep-water phases, similarly defined as the Depth data were analysed to characterize depth-spe- longest temporally consistent period (i.e., occurring dur- cific habitat use and the extent and timing of migra - ing boreal winter, as per the regional mean profile) dur - tion to deep-water spawning grounds. Herein, it was ing which that individual persistently occupied depths assumed that Pacific halibut spend considerable time at deeper than the pooled regional mean. These data were or near the seafloor each day, such that the maximum then used to identify the deep- and shallow-water phases depth recorded during each 8-h data-binning period was for each individual and compute mean depth within each likely to represent seafloor depth, and multiple analyses phase (i.e., deep-water and shallow-water) for those indi- were conducted using all tags within each geographic viduals. Individual mean depths were then averaged to region for which maximum-depth data were obtained derive regionally explicit mean depths occupied during (Table 1). A primary objective was to evaluate the degree each depth phase; this eliminated, to the greatest degree to which seasonal migration may vary by region. Such practicable, biasing the regional means in favor of fish movements are typically composed of offshore migra - whose tags transmitted more data than others. tion in the autumn with return to shallow waters in the To investigate whether water temperature might spring. Prior research [4] has compared geographic vari- influence the timing of the autumn offshore migra - ance in seasonal habitat use by defining a fixed spawning tion, conditions experienced by migratory individuals period among all regions of interest that was equivalent were compared among geographic regions. Migratory to the period over which active spawning (i.e., spawn- individuals were defined as those that moved to depths rise behavior) [36] has been reported from archival- below the regional mean annual depth in autumn and tagging data [12, 37]. However, the high-resolution data remained in deep water for at least 4 weeks thereafter. required to confidently conduct spawn-rise analyses [38] For each such individual, mean temperature experienced for Pacific halibut are primarily limited to the GOA and during the 6-week period spanning 28 July and 7 Sep- applying any single spawning period to all regions within tember was calculated; this period was chosen, because an analysis may fail to capture important geographic dif- it represented the period during which the tagged Pacific ferences in spawn timing; and, in the current context, off - halibut in all regions persistently occupied, on average, shore–onshore migrations that occur prior to, and after, their shallowest depths. For each migratory individual, the active spawning are of considerable interest. Herein, an autumn departure date was then identified that rep - the available depth data were used to define regionally resented the date on which the fish descended below the explicit summer shallow-water and winter deep-water regional average depth and persistently occupied deeper phases after standardizing data resolution among all water thereafter. Based on this date, the average tempera- tags analyzed by converting high-resolution (i.e., 30 s) ture experienced by the individual was then computed data from physically recovered tags to 8-h-binned mini- for the 1-week period (i.e., 168 h) immediately prior to its mum and maximum values. Each region’s pooled, aver- departure to deep water, as well as for each of the three age annual depth profile was divided into two periods 1-week periods prior to that (i.e., producing 4 weekly representing the occupation of habitat that was either values spanning the 4 weeks prior to autumn depar- shallower than, or deeper than, the region’s population- ture). Mean summer temperatures experienced among level annual mean. First, mean depth within each region fish within each region were compared to determine was computed among all individuals for all available bin- whether fish inhabiting different regions had experienced ning periods throughout deployment. Next, a region- significantly different pre-migration conditions. Within ally explicit annual mean depth occupied was calculated regions, the five sequential temperatures were compared using the longest duration for which data were contigu- to determine whether conditions changed significantly in ously available among regions: i.e., 21 June of the year of the weeks prior to offshore migration, and to potentially tagging through 3 June of the year of tag-reporting. Then, identify threshold temperature(s), or net change in tem- mean maximum daily depth (MMDD) profiles [sensu 12] perature, that might have initiated migration. were constructed for each region as 3-day moving aver- ages of the binned data (i.e., averaged across twelve con- Statistical analyses secutive 8-h bins), to smooth short-period fluctuations. For all fish for which a date of tag detachment was Each region’s deep-water (spawning) phase was then obtained, the relationship between mean fish length and defined as the longest period over which the smoothed tag-retention period (i.e., to test whether subsequent mean depths were continuously deeper than the region’s analyses might be size-biased) was examined via linear L oher Animal Biotelemetry (2022) 10:18 Page 9 of 21 regression. Regional differences in mean fish lengths, Table 2 Tukey HSD comparisons: fish lengths and tagging depths tagging depths, seasonal depth distributions (i.e., mean shallow- and deep-water phase depths), and linear dis- Comparison WA CA TZ SE SI placement were investigated via multiple analysis of vari- WA – 0.235 0.207 0.228 0.480 ance (MANOVA) in which region was specified as the CA 0.978 – 1.000 1.000 0.000 independent (categorical) variable and the other factors TZ 0.976 1.000 – 1.000 0.000 as dependent. Tukey HSD post hoc multiple comparisons SE 0.754 0.412 0.427 – 0.000 were conducted for factors that were determined to be SI 0.190 0.421 0.464 0.021 – significant. Regional variance in the proportion of satel - lite transmissions received per tag, mean temperatures Results of Tukey HSD pairwise post hoc comparisons of the mean length of fish contributing to analyses in the current study (above the diagonal) and the mean experienced by fish, and the mean number of light-based depths at which fish were captured and released (below the diagonal), for Pacific longitude estimates generated per tag were examine via halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Western Aleutian (WA) and Central Aleutian (CA) one-way ANOVA with Tukey HSD post hoc multiple Islands, the Southeast Bering Sea continental Shelf Edge (SE) and Shelf Islands comparisons conducted where significance was detected. (SI), and a Transition Zone (TZ) between those regions and the western Gulf of Statistical tests were conducted using TIBCO Statistica Alaska Bold values indicate comparisons that were significantly different at p ≤ 0.05 version 13.5.0.17 (Palo Alto, California, USA) and errors reported in the text of this manuscript will represent one standard deviation unless otherwise specified. had been tagged at St. Matthew Island. Tags from 119 fish generated endpoint locations via satellite transmission: Results 76 of these reported within the specified annual-displace - Tag deployments ment window of 360–370 days at liberty; 43 reported A total of 145 Pacific halibut were tagged: 115 from 28 after shorter periods, ranging from 14 to 343 days. Fork May to 31 August 2008; 17 from 6 June to 21 July 2009; length was not found to be a significant determinant of and 13 from 11 to 26 June 2016 (Table 1; Fig. 2). Deploy- tag retention periods; i.e., smaller fish did not shed their ments occurred throughout the study area in 2008; but, tags at a significantly higher rate than larger fish [linear only at the EBS Shelf Islands in 2009 and only on the regression: df (1, 112), F = 0.028, p = 0.870, R = 0.0002]. northern EBS Shelf Edge in 2016. Tags were deployed in The proportion of binned data that were received via each tagging region roughly in proportion to the abun- satellite broadcast from tags that transmitted data sum- dance of Pacific halibut ≥ 82 cm FL and ranged from a maries was highly variable, ranging from 7.0 to 96.7%. minimum of 18 individuals tagged in the Central Aleu- Mean proportion of data received varied significantly tian Islands region to a maximum of 49 individuals [ANOVA: df (4, 112), F = 3.893, p = 0.005] by region, tagged along the EBS Shelf Edge (Table 1). Fish ranged as follows: Western Aleutian Islands = 59.6 ± 13.1%; from 82 to 171 cm FL. For fish whose tags subsequently Central Aleutian Islands = 61.5 ± 13.1%; Transi- produced at least one category of useable data (i.e., final tion Zone = 57.8 ± 19.6%; Eastern Bering Sea Shelf location, light, or depth data; n = 116), significant differ - Edge = 73.2 ± 24.3%; Shelf Islands = 77.9 ± 24.3%. Tukey ences in mean fork length were detected among tagging HSD post hoc comparisons indicated that reception rates regions [MANOVA: df (4, 62), F = 7.052, p < 0.001]. Mean were similar among the regions located along the Aleu- fish lengths were not significantly different within the tian Ridge and significantly higher, and similar to each three regions occurring along the Aleutian Ridge or on other, for the Eastern Bering Sea Shelf Edge and Shelf the EBS Shelf Edge; fish tagged at the EBS Shelf Islands Islands. were on average larger than in all other regions (Tables 1 None of the tags that were deployed in 2016 (i.e., PSAT and 2). Flex) were recaptured within their first year at liberty, successfully communicated via satellite, or produced Tag and data recoveries: via fishery and satellite data downloads. However, three tags were recovered in Twenty-one tags were neither physically recaptured longline fisheries on dates that were later than the tags nor communicated via satellite and were, therefore, had been programmed to detach and report, their batter- lost entirely. An additional three tags produced satellite ies having failed prematurely. These tags were recovered uplinks that were of insufficient strength to determine after periods at liberty of 701, 777, and 1155 days at lib- their final positions or upload any data. Of the remain - erty. Although no environmental data were obtained for ing 121 fish, two that had been tagged with Mk10s were these fish, final coordinates were obtained for two that recaptured in commercial fisheries prior to scheduled tag were recaptured in US waters. The third was recaptured reporting, after periods at liberty of 11 and 343 days; both in Russian waters in the Western Bering Sea (i.e., coastal Loher Animal Biotelemetry (2022) 10:18 Page 10 of 21 Kamchatka) and a precise location was not obtained. and Rat Islands group, bounded between Near Strait to These fish were recovered in late spring and summer the west and Amchitka Pass to the east; for the Central (i.e., late May, July, and August, respectively) and so their Aleutians, interannual displacement was confined to recoveries will be plotted; however, their final locations the Andreanof Islands between Amchitka and Amukta will not be included in calculations of regional dispersal, Passes (Fig. 2). From a fishery-management perspec - having not satisfied the criterion of occurring within a tive, this represented retention within IPHC Regulatory multiple of 360–370-day post-release. In addition to tags Area 4B (Fig. 1). Pacific halibut that were tagged in the recaptured in fisheries, seven tags (all Mk10s) were found EBS were more dispersive than observed in the Aleu- awash after having broadcast their data and producing tian Island regions. Those that departed the EBS Shelf final positions. These tags were downloaded to recover Islands dispersed either eastward on the EBS shelf (n = 1) their full scheduled-broadcast records. or northwestward into Russia waters (n = 1). Those that departed the EBS Shelf Edge displayed similar behav- Displacement and interannual dispersal ior, moving into the shallow waters of Bristol Bay (n = 2) Endpoint-derived mean interannual linear displacements and into Russian waters (n = 2) from just westward of (Table 3) were greatest for the Transition Zone; some- the Russian EEZ boundary to the Gulf of Anadyr. The what lower for the EBS Shelf Edge and Shelf Islands; and fish that was physically recaptured in coastal Kamchatka lowest in the Western and Central Aleutian Islands. How- after roughly 3 years at liberty brought the total number ever, variance in displacements was high in all regions of Pacific halibut that had dispersed to Russian waters to and none of the observed differences were statistically four, resulting in ~ 14% interannual emigration to Russia significant [MANOVA: df (4, 62), F = 1.677, p = 0.167]. for fish tagged in the EBS. Interannual dispersal was characterized by the following Within the Transition Zone, Pacific halibut that were general patterns (Fig. 2): (1) for Pacific halibut tagged in tagged north of Unimak Pass produced final endpoints the Western and Central Aleutian Islands, all final tag- that were exclusively within the Bering Sea (n = 6), within reporting locations were within the region in which the IPHC Regulatory Area 4A. One of these fish was recap - fish had been tagged; (2) Pacific halibut tagged in the tured in August of 2010, after 2 years at liberty (771 days), EBS moved among tagging regions and among IPHC to provide a third known location: it was captured in regulatory areas with final locations that were exclusively northern Area 4A, approximately 2 km from its initial within the Bering Sea, ranging from the Gulf of Anadyr, release location. In contrast, fish that were tagged south Russia, to Bristol Bay, Alaska; (3) Pacific halibut tagged of Unimak Pass (i.e., in the far-western GOA; southern in the Transition Zone displayed the greatest amount of Regulatory Area 4A) emigrated at a relatively high rate emigration from their tagging region, with migrants dis- (~ 55%), exclusively eastward, dispersing among IPHC persing across the GOA as far south as coastal Washing- Regulatory Areas: 4A (n = 5), 3B (n = 4), 3A (n = 1), and ton State. 2A (n = 1). One fish that emigrated to Regulatory Area 3B For Pacific halibut tagged in the Aleutian Islands, after 1 year at liberty was subsequently recaptured in its interannual displacement was confined to the Island third summer at liberty (1117 days) and had moved far- group that defined each tagging region: i.e., for the West - ther eastward into Regulatory Area 3A. ern Aleutians, fish remained resident within the Near At‑liberty dispersal inferred from light data A total of 109 tags generated twilight data that allowed Table 3 Dispersal results for the estimation of daily local noon (Table 1), producing 5313 daily longitudes estimates, while those fish were at Tagging region Mean Light‑lon range Light‑lon N liberty (Table 3). The number of at-liberty position esti - displ. ± SD (mean ± SD) (km) mates per fish ranged from 1 to 152 and averaged from 41 to 66, depending upon region (Table 3). However, no Western Aleutian 79 ± 37 3–97 (41 ± 7) 819 significant differences were detected in the average num - Central Aleutian 39 ± 68 15–81 (46 ± 8) 742 ber of longitude estimates per fish among regions [one- Transition Zone 313 ± 70 1–113 (51 ± 7) 1175 factor ANOVA: df (4, 104), F = 1.541, p = 0.196]. EBS Shelf Edge 189 ± 19 1–152 (45–6) 1417 For the Western and Central Aleutian Islands, inspec- EBS Shelf Islands 100 ± 211 1–141 (66 ± 8) 1160 tion of individual light-based longitude trajectories (36 Mean interannual linear displacement, range in and mean number of light- of 40 tags; Table 1) did not yield sufficient evidence to based daily longitude estimates obtained for individual fish, and total number conclude that any individual fish had departed its tag - of longitude estimates produced for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the ging region during its time at liberty. That is, despite Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those those individuals having generated numerous light-based ecosystems and the western Gulf of Alaska L oher Animal Biotelemetry (2022) 10:18 Page 11 of 21 longitudes to both the east and west of their regions’ late September and early December, spent 4–5 months in boundaries, no individual generated five consecutive IPHC Regulatory Areas 3B and 3A, and returned to Reg- estimates that would have plausibly placed that individ- ulatory Area 4A in late April. Their longitudinal migra - ual outside of its tagging region. Similarly, out-of-region tions were coincident with movement to deep water in movement could not be identified for any fish tagged on autumn and return to the depths at which they had been the EBS Shelf Edge or at the Shelf Islands, including for tagged in spring (Fig. 4a, b, lower panels). Light-based individuals whose final positions unequivocally demon - longitudes also allowed for the timing of regional depar- strated that they had departed the region. In contrast, ture to be identified for two migratory individuals. The seasonal departure followed by return to the tagging first of these fish (Fig. 3b) departed the Transition Zone region was evident for two fish tagged in the Transition at the end of December. The second (Fig. 3c) produced Zone south of Unimak Pass (Fig. 4a, b, upper panels). considerably fewer position estimates, but the available These individuals departed the Transition Zone between data suggest November departure. As with the seasonally Fig. 4 Light-based latitude estimates (upper panels) and daily maximum depth profiles (lower panels) during 2008 and 2009 for: a, b two Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the far-western Gulf of Alaska, and; c one Pacific halibut tagged at the Pribilof Islands in the eastern Bering Sea, from date of tagging (top of upper panels) to date of final tag reporting (bottom of upper panels). All data were obtained via satellite transmission (i.e., the tags were not recovered). The approximate longitudinal span of International Pacific Halibut Commission regulatory areas is indicated by shading. All three fish moved to deep water during the winter and returned to shallow water in spring. Eastward seasonal emigration is clearly evident in a and b. The fish depicted in c is known to have emigrated from its tagging area on the basis of its mid-winter depths (i.e., no such depths occur at the Pribilof Islands), but neither the timing or magnitude of its emigration can be determined. Longitude estimates are absent during its deep-water phase; the remainder of the profile is characterized by apparent measurement error of 5–7° relative to the fish’s known endpoint positions Loher Animal Biotelemetry (2022) 10:18 Page 12 of 21 migratory individuals, the longitudinal migrations of eastward redistribution was apparent after October. As these fish were associated with movement to deep water with the individual trajectories, pooled unstandardized during winter. longitudinal data for Pacific halibut tagged in the EBS For Pacific halibut that were tagged in the EBS (on both regions were relatively uninformative with respect to sea- the Shelf Edge and at the Shelf Islands), light-based longi- sonal and interannual redistribution. tude estimates provided little information regarding their In all regions other than the Transition Zone, pooled movements. Many of their interannual displacements standardized at-liberty longitude estimates were were relatively short distance and, for fish whose migra - described by a single frequency mode (Fig. 6) that was tions were considerable, their movements were executed centered within one degree of longitude relative to the to a considerable degree along north–south vectors fishes’ tagging locations: displacement in the mean (Fig. 2). Such movements are not well-resolved by longi- and median was small (< 0.25°) in the Western Aleu- tudinal position estimates. tians and on the EBS Shelf Edge; and somewhat larger Pooled among all individuals tagged within each region, (0.75–0.91°) for the Shelf Islands and Central Aleutians unstandardized light-based at-liberty longitude estimates (Table 4). These frequency distributions displayed posi - for Pacific halibut tagged along the Aleutian Ridge (Fig. 5) tive (eastward) skew, which was relatively minor in the displayed some clear seasonal trends. A decrease during EBS (< 0.6) and moderate in the Aleutian Island regions winter in the frequency with which estimates were gener- (~ 1.4). Few longitude estimates occurred beyond ~ 5° of ated was apparent in all regions, and most evident in the the median and those observations were symmetrically Western Aleutian Islands (Fig. 5a). This was coincident distributed, consistent with the occurrence and mag- with movement to deeper water (see next section), with nitude of the apparent random measurement error that fish reaching depths at which ambient light levels are was evident in many of the individual longitude trajecto- expected to be below the tags’ detection thresholds. For ries (e.g., Fig. 4c). In contrast, fish tagged in the Transi - fish tagged in the Transition Zone (Fig. 5c), progressive tion Zone produced a distinctly bimodal distribution of Fig. 5 Light-based longitude estimates during periods at liberty for all Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the a Western and b Central Aleutian Islands, and c Transition Zone between the eastern Aleutians, Southeastern Bering Sea, and western Gulf of Alaska, from dates of tagging (top of panels) through final tag reporting (bottom of panels). Dashed lines indicate ocean passes separating regions located along the Aleutian Ridge and the approximate extent of International Pacific Halibut Commission Regulatory areas in the Gulf of Alaska. Note that the Western Aleutians and the Komandorskiye Ostrova are located in the eastern hemisphere, with the Antimeridian bisecting IPHC Regulatory Area 4B at approximately Amchitka Pass L oher Animal Biotelemetry (2022) 10:18 Page 13 of 21 Fig. 6 Frequency distributions of light-based at-liberty location (longitude) estimates obtained for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those ecosystems and the western Gulf of Alaska. Values have been standardized to reflect degrees westward (negative values) or eastward (positive values) of each fish’s known initial tagging location. The vertical dashed line in each panel indicates a value of 0, such that distributions not centered at that location indicate substantial residency of the tagged halibut at some distance away from their initial longitudes Loher Animal Biotelemetry (2022) 10:18 Page 14 of 21 Table 4 Characteristics of frequency distributions of in some regions, the tagged population’s deep-water standardized geolocations phase extended beyond that date. Mean tagging (fish-release) depth varied significantly Tagging region Mean ± se Median ± se Skewness ± se among regions [MANOVA: df (4, 62), F = 2.64, p < 0.042]. Western Aleutian 0.14 ± 0.11 0.23 ± 0.11 1.37 ± 0.09 This was expected, given that the regions in the EBS Central Aleutian − 0.87 ± 0.14 − 0.91 ± 0.14 1.38 ± 0.09 were chosen to represent depth-specific habitat. As such, TZ Mode 1 − 1.22 ± 0.09 − 1.23 ± 0.09 − 0.38 ± 0.08 mean tagging depth (Table 5) was significantly (Tukey TZ Mode 2 13.0 ± 0.23 13.1 ± 0.24 0.08 ± 0.25 HSD, p = 0.021) greater for fish released along the EBS EBS Shelf Edge 0.17 ± 0.10 0.15 ± 0.10 0.06 ± 0.07 Shelf Edge than at the Shelf Islands; no other pairwise EBS Shelf Islands − 0.61 ± 0.10 − 0.75 ± 0.10 0.55 ± 0.07 comparisons were significant (Table 2). Characteristics of frequency distributions (see Fig. 6) of light-based at-liberty Mean depths during the tagged individuals’ summer- location (longitude) estimates, standardized relative to initial locations, of Pacific time shallow-water phases (Table 5; Fig. 7) varied sig- halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags nificantly among regions. Mean shallow-water phase during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone (TZ) between those ecosystems and the western Gulf of Alaska depths for fish tagged on the EBS Shelf Edge were signifi - cantly greater than for fish in all other regions (all com - parisons: Tukey HSD, p < 0.001) and shallow-water phase standardized longitude estimates composed of a primary depths for fish tagged in the Western Aleutian Islands mode positioned ~ 1.2° westward of the fishes’ tagging were greater than at the EBS Shelf Islands (Tukey HSD, locations and a smaller secondary mode centered ~ 13° p = 0.038). No other pairwise comparisons were signifi - eastward, in the Gulf of Alaska. The secondary mode was cant at p ≤ 0.05. For the most part, relative differences consistent with winter locations of fish that were deter - in depth distribution persisted throughout the fishes’ mined to be seasonally migratory (Fig. 4). movements to deeper water during the winter (Table 5; Fig. 7), with one exception: Pacific halibut tagged at the Seasonal depth distributions EBS Shelf Islands exhibited the deepest mean deep- The timing of tagging and, therefore, the precise span water phase (563 ± 156 m), while in no other region did over which depth data were available varied by tagging the mean deep-water phase exceed 500 m. Fish tagged in region. For the purposes of plotting and analysis, the the Transition Zone displayed the shallowest deep-water “year” over which depth data were included was defined phase, averaging 369 ± 69 m. Differences were signifi - as the 365-day period over which the greatest amount cant between the Shelf Islands and the Central Aleutian of data was available among all regions: from 19 June of Islands (Tukey HSD, p = 0.028), between the Shelf Islands the tagging year through 18 June of the subsequent year. and the Transition Zone (Tukey HSD, p = 0.002), and This provided contiguous data for all regions except the between the EBS Shelf Edge and the Transition Zone Western Aleutian Islands. Fish were tagged earlier in the (Tukey HSD, p = 0.034). Western Aleutians than in other regions, resulting in data In addition to the observed differences in phase- that terminated approximately 2 weeks earlier than else- specific mean depths (Table 5), the timing of offshore– where: i.e., on 3 June of the year after tagging. The analy - onshore redistribution, length of deep-water residency, ses presented subsequently demonstrate that terminating and general form of the mean maximum daily depth all regions’ depth profiles on 3 June to maintain a unified (MMDD) trajectories differed among tagging regions. period among all regions would have been inappropriate: Table 5 Depth results Tagging region Mean depths (meters) Deep‑ water phase Tagging (± SD) Shallow (± SD) Deep (± SD) Arrive Depart Days Western Aleutian 144 ± 90 158 ± 40 453 ± 65 09 Nov 07 May 180 Central Aleutian 133 ± 103 139 ± 29 411 ± 85 12 Nov 20 May 190 Transition Zone 123 ± 62 141 ± 49 369 ± 69 25 Nov 04 May 161 EBS Shelf Edge 187 ± 80 214 ± 72 468 ± 93 12 Dec 04 Jun 175 EBS Shelf Islands 74 ± 35 107 ± 19 534 ± 156 27 Dec 11 Jun 194 Mean depths at which fish were captured and released (tagging), during the fishes’ summer shallow-water and winter deep-water phases, the mean dates of arrival and departure from and resulting population-level average duration occupying deep-water habitat for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those ecosystems and the western Gulf of Alaska L oher Animal Biotelemetry (2022) 10:18 Page 15 of 21 Fig. 7 Three-day moving averages of maximum daily depth for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea, and a Transition Zone between those ecosystems and the western Gulf of Alaska. Note that these profiles include all individuals for which depth data were generated, even if they did not conduct an offshore migration (e.g., see Fig. 3a, lower panel) and, therefore, tend to result in group-level mid-winter depths that are shallower than the group’s average deep-water phase depth (i.e., Table 5) Pacific halibut tagged in both Aleutian regions behaved form of the MMDD trajectories was quite similar for all similarly, arriving on average in relatively deep water in three regions (Fig. 7a). In contrast, fish tagged in the EBS mid-November and returning to shallower water in May. arrived at deep-water grounds approximately 1 month Deep-water phase duration and depth were shorter and later than along the Aleutian Ridge and did not return shallower, respectively, in the Transition Zone than far- to shallow water until June (Table 5; Fig. 7). Unlike in ther westward in the Aleutians (Table 5); however, the all other regions, Pacific halibut tagged at the EBS Shelf Loher Animal Biotelemetry (2022) 10:18 Page 16 of 21 Table 6 Mean temperatures (degrees Celsius) experienced by region, Pacific halibut tagged in the Central and Western Pacific halibut (Hippoglossus stenolepis) tagged with pop-up Aleutian Islands remained within their island groups (i.e., archival transmitting tags during 2008 and 2009 in the Aleutian within the Andreanof and Near-Rat Islands, respectively), Islands, Eastern Bering Sea (EBS), and a Transition Zone between while those tagged on and along the EBS continental those ecosystems and the western Gulf of Alaska, that undertook shelf moved among Bering Sea regulatory areas and from seasonal offshore migrations during the autumn of the year in Alaskan to Russian waters. These dispersal patterns are which they were tagged consistent with summer-to-winter PAT-tagging studies Tagging Midsummer D‑4 D‑3 D‑2 D‑1 that have indicated basin-scale reproductive segrega- region tion with considerable mixing within those basins [4]; research that has shown Samalga Pass to be an oceano- Western 4.8 ± 0.6 4.5 ± 0.5 4.5 ± 0.5 4.5 ± 0.5 4.5 ± 0.5 Aleutian graphic [39] and ecological [40–42] boundary within the Central Aleu- 5.3 ± 0.4 5.5 ± 0.8 5.2 ± 0.7 5.2 ± 0.7 5.0 ± 0.6 Aleutian Island ecosystem; and population-genetic analy- tian ses that suggest relative isolation of Pacific halibut in the Transition Zone 5.2 ± 0.6 5.4 ± 0.5 5.2 ± 0.7 5.1 ± 0.6 5.0 ± 0.6 Aleutian Islands westward of Amchitka Pass [7]. In addi- EBS Shelf Edge 2.9 ± 0.4 3.0 ± 0.6 3.0 ± 0.6 3.0 ± 0.6 3.0 ± 0.6 tion, light-based geolocation demonstrates that connec- EBS Shelf 3.3 ± 0.9 2.6 ± 1.3 2.6 ± 1.3 2.7 ± 1.2 2.8 ± 1.1 tivity among regulatory areas may be seasonally cyclic, Islands with fish emigrating from their tagging region for the “Midsummer” was defined as 28 July through 7 September, when fish were winter and returning the following spring. This was evi - resident on shallow grounds. “D-1” represents mean temperature during the week prior to the fishes’ departure to deep water; D-2 is the week prior to D-1, dent at both the individual and population level for fish and so forth. The superscripts after the midsummer means indicate two groups tagged south of Unimak Pass. determined to be significantly different (p < 0.05) from one another. In no region In contrast to the large degree of out-of-area movement were mean temperatures experienced during the 4 weeks prior to departure found to be significantly different from the region’s midsummer temperature demonstrated by Pacific halibut tagged south of Unimak Pass, there was insufficient evidence to suggest move - ment of tagged fish out of either the Central or Western Islands displayed a distinctly bi-phasic offshore migra - Aleutian Islands, at either interannual or seasonal scales. tion, initially moving from December until late February Although numerous tags produced light-based longitude to depths of approximately 300 m, then descending to estimates that might have placed them outside of these > 400 m during March and April (Fig. 7b). regions, inspection of the individual trajectories failed to Mean temperatures experienced during midsummer reveal series of consecutive estimates indicative of east– by seasonally migratory Pacific halibut (Table 6) varied west redistribution beyond the regions’ boundaries, and significantly [ANOVA: df (4, 51), F = 35.264, p < 0.001] the distribution of pooled, standardized values in both according to tagging region. Tukey HSD post hoc com- Aleutian regions showed little skew and no secondary parisons indicated that the temperatures experienced peaks consistent with cross-pass movement. Rather, the were similar among regions located along the Aleutian nature of the longitude estimates to the east and west of Ridge and significantly lower for fish tagged on the East - the passes that define the Aleutian tagging regions was ern Bering Sea Shelf Edge and at the Shelf Islands. In no most consistent with estimation error induced by local region was a significant change in the experienced tem - environmental conditions biasing the perception of local peratures detected between midsummer and any of the noon relative to its true value [35]. four 1-week periods prior to the initiation of departure to Although light-based geolocation failed to resolve sea- deeper water in autumn. sonal movement among areas in the EBS—perhaps due to the ability of individuals to migrate from summer feeding sites to deeper-water spawning grounds without Discussion moving considerably along east–west axes—the end- The current study builds upon prior analyses of relative point data clearly demonstrated interannual migration spawning segregation [4] and genetic population struc- from the USA into Russian waters (n = 4). Ultimately, ture [6, 7] to enhance our understanding of connectiv- movement between North American and Asian waters ity within the Pacific halibut stock of the eastern Bering on the northern Bering Sea continental shelf is not sur- Sea and Aleutian Islands. The Pacific halibut that were prising: there are no known oceanographic or geologi- tagged herein exhibited basin-specific dispersal in which cal features that might impede such connectivity. Rather, fish tagged in the Bering Sea had summer distributions larval transport modelling predicts that some proportion that were within the Bering Sea, while fish tagged south of larvae spawned in the EBS are likely to be delivered of Unimak Pass were more dispersive and either occu- to Russian coastal habitat [43], and our understanding pied or transited all IPHC regulatory areas in the Gulf of ontogenic [9, 10] and seasonal [4, 5, 44] cross-basin of Alaska and US Pacific Northwest. Within the BSAI L oher Animal Biotelemetry (2022) 10:18 Page 17 of 21 movements in the GOA should lead to an expectation of With respect to seasonal migration and depth-specific considerable cross-basin mixing in the Bering Sea, par- habitat usage, regional differences were apparent. In the ticularly at young ages. Both systems are characterized EBS, Shelf Island fish moved to deeper winter spawn - by spawning that is concentrated in submarine canyons ing habitat than fish tagged along the Shelf Edge, despite along their eastern and central margins, and westward- having summered in relatively shallower water that was flowing shelf-edge currents [45, 46]. Basin-scale con- considerably farther from their shelf-edge winter destina- nectivity over the course of Pacific halibut life history is tions. In addition, the annual depth trajectories of Shelf likely to display approximately the same spatial structure Island fish were on average considerably more biphasic in both systems. Within the Bering Sea, individuals that (i.e., dual-stage) than observed on the Shelf Edge. This are derived from EBS spawning and which settle in nurs- dual-stage movement pattern is common in individu- eries along the Asian coast should be expected to return als and, in cases in which the data allow for a detailed to the North American spawning stock if the population evaluation of active spawning (i.e., putative egg release) is to maintain long-term stationarity. It is perhaps only [12, 36, 38], appears to represent an initial period of pre- surprising that the current study detected as much inter- spawn staging that is followed by active spawning at the annual migration to Russian waters as it did, given that deeper stratum. The depths associated with each stratum those movements are counter to the expected direction vary according to individual and examples of dual-stage of mean population-level dispersal of benthic-stage indi- movement occurred in all tagging regions in the current viduals: i.e., from west to east. This would suggest that study. Ultimately, the dual-stage nature of the group-level overall migration rates across the Russia–USA maritime MMDD profile for Shelf Island fish derives from these border are non-trivial, which is an important observation fish having exhibited more-synchronous movements from the perspective that stock assessment modelling, than were observed elsewhere. In other regions, indi- policy analyses, and management decisions have tradi- vidual dual-stage migrations were obscured by averag- tionally assumed that Pacific halibut in waters of Canada ing among individuals that did so with variable timing. In and the USA exist in a closed system in which there is addition, differences in sex ratio among tagged individu - no exchange with population(s) in the western half of als within each region could obscure this pattern, if one the species’ geographic range. The omission of this con - sex more consistently undertakes dual-stage migration nectivity when characterizing the function of EBS Pacific than the other. We are unable to address that hypothesis halibut stocks to-date [e.g., 3, 4, 9] is likely due to an here, because reliable techniques for evaluating sex with- absence of data regarding its magnitude and dynamics. out sacrificing the individuals had not yet been devel - Future work should seek to generate migration-rate data oped and sex was, therefore, unknown. However, taken along this axis. together, the displacement, at-liberty longitude estimates, The observation that out-of-area dispersal was highest and depth profiles suggest that the Pacific halibut tagged for Pacific halibut that were tagged in the far-Western along the Shelf Edge and at the Shelf Islands likely occu- GOA is consistent with prior PIT-tagging research [9] in pied the same regional slope spawning grounds (i.e., which approximately 90% of the fish that emigrated east - in Pribilof, Pervenets, and Zhemchug Canyons; Fig. 1) ward from IPHC Regulatory Area 4A into GOA regula- [12], but with somewhat different timing and depth tory areas were individuals that had been tagged south of preferences. the Aleutian Ridge. The spatial discontinuity in connec - Pacific halibut recruitment is believed to be environ - tivity within this region (i.e., very different mixing pat - mentally driven, via favorable plankton productivity terns on opposite sides of the Aleutian Ridge) highlights or larval transport [47] and the maintenance of broad the challenges associated with defining management spawning and migratory periods in marine fish popula - subregions that are intended to represent stock structure tions likely represents ecological bet-hedging [48] that and function [3] when constrained to the boundaries of ensures long-term recruitment success and stock pro- historical management units that may not have been ide- ductivity. Evidence of migratory contingency has been ally crafted to do so. For assessment and policy purposes, observed in Pacific halibut in the Gulf of Alaska [14] data obtained from Regulatory Area 4A are assigned to and US Pacific Northwest [16], and analyses suggest that Bioregion 4 [3], which largely describes stock status and Pacific halibut in the EBS have exhibited shifts in their dynamics of the eastern Bering Sea. However, the current distribution and habitat use over the last three decades results demonstrate that dispersal from IPHC Regulatory in response to changing ocean temperatures [49, 50]. Area 4A is more complex: 4A North represents the EBS, Maintaining a diversity of life-history strategies may be while 4A South represents the western GOA and the critical to the successful management of exploited stocks. individuals tagged within it displayed considerable move- Parameterizing the timing and duration of larval release, ment across the GOA and into the US Pacific Northwest. and the depths (i.e., current regimes) into which larvae Loher Animal Biotelemetry (2022) 10:18 Page 18 of 21 are released, is required for the construction of larval of onshore–offshore migration. For example, an early transport models and for evaluating relative recruitment trawl survey might sample inshore waters prior to the potential among population components. arrival of seasonal migrants and the setline survey sub- Pacific halibut tagged along the Aleutian Ridge, in both sequently occupy the slope after fish have departed, such the Aleutian Islands and the far-western GOA, were that neither survey fully indexes the population. A bet- observed to move to deep water and return to shallow ter understanding of the timing of these movement and habitat roughly 1 month earlier than fish tagged in the their interannual variability would be required to under- EBS and there was evidence of a gradual advancement stand to what extent such mismatch is likely in any given in the timing of deep-water occupancy moving anticy- year, but the ultimate result may be time-varying selec- clonically from the Western Aleutians to the EBS Shelf. tivity within each survey that should be accounted for in Although the available data did not provide any evidence the assessment models. The IPHC stock assessment does of a threshold temperature, or change in local tempera- have a structure that can account for time-varying selec- ture, that might initiate the autumn offshore migra - tivities [53] and the effects of seasonal migration on fish - tion, the observation that fish residing in cooler waters ery selectivity [54] and the incorporation of time-varying departed later in the year is consistent with findings from functions in assessment models [e.g., 55, 56] has received research conducted both to the south and north of the attention in the literature. current study. Pacific halibut tagged in the eastern Gulf of Inspection of Pacific halibut depth profiles also high - Alaska have been shown to initiate offshore migrations as lights the difficulty of evaluating at-liberty movements early as September and largely arrive at their wintertime for any given individual in the absence of highly resolved depths by the end of October [12]. Emerging data sug- location data. For example, fish tagged along the EBS gest that Pacific halibut that summer in Norton Sound Shelf Edge were found at shallower average depths in July (Fig. 2), in the northeastern Bering Sea, may move to of 2009 than they had occupied when they were tagged. deep water as late as March and spawn in April and May This suggests that their preferred summer habitat was (A Flanigan, University of Alaska Fairbanks, USA, per- likely shallower in 2009. However, from the available data sonal communication). Similarly, latitudinal gradients in is it difficult to identify the precise nature of the habitat spawn timing have been reported for North Pacific starry shift nor identify its driver(s). In particular, light-based flounder (Platichthys stellatus) and rex sole (Glyptoceph - longitudinal estimates provided little information regard- alus zachirus). Spawning in these species occurs approxi- ing their movements (e.g., relative to fish located along mately 4 months earlier off the coast of California than in the Aleutian Ridge), because large-scale redistribution the southeast Bering Sea [51]. may occur on the EBS continental shelf along north– In addition to its ramifications on interannual recruit - south axes. The associated changes in stock distribution ment potential and vulnerability to seasonal fisheries may be quantifiable through refined methodology, such [12], regional variance in seasonal migration timing can as advanced mobile acoustics [57, 58], ongoing develop- affect our understanding of the distribution and demo - ments in geomagnetic-sensing electronic tags [59, 60], graphic structure of populations via their interactions and the adaptation of statistical models for tracking ben- with survey design. For example, the depth trajectories in thic and epibenthic marine species [24–26]. In addition, the EBS remained upward-sloping throughout June and the use of mark-report PAT tags, which generate pop-up into July of 2009 (i.e., the year after tagging), indicating locations in the absence of archived environmental data that these fish were still in the process of returning to and, therefore, cost considerably less than standard PAT their summer habitat when the assessment surveys com- tags, could be considered. Large-bodied species such as menced. In the EBS, two platforms are used to index the Pacific halibut can be tagged with multiple such tags that abundance of Pacific halibut: the IPHC FISS that inten - are programmed to release at any desired interval, pro- sively surveys the continental shelf edge and US National viding more-detailed migration histories than can any Marine Fisheries Service trawl surveys [52] within shal- single pop-up tag [sensu 61] as well as providing vali- lower continental shelf habitat. Because catchability var- dated locations to increase the accuracy of tracking mod- ies between the two survey techniques and they can be els [62]. conducted with slightly different timing in any given year, Ultimately, to fully understand population ecology in our perception of fine-scale distribution and habitat use Pacific halibut we will need to fill the scaling-gaps that exist may be affected. Differences in relative catchability can be within the existing data: in particular, a lack of informa- interpreted as differences in underlying abundance even tion describing generational-scale processes and informa- in a homogeneous population. Relative survey timing can tion on sex-specific behavior. With respect to the latter, either oversample or undersample the population if the ultrasonic [63] and genetic techniques [64] have now been timing of the surveys does not correspond with timing developed that allow for determination of the sex of Pacific L oher Animal Biotelemetry (2022) 10:18 Page 19 of 21 Consent for publication halibut during tagging and can be employed in future work. The author has given his consent to the manuscript being published. With respect to the former, the interannual time scales addressed herein are still far from generational in this Competing interests The author declares that he has no competing interests. species. Although it is rare for satellite-tagging studies to exceed 1 year due to hydraulic drag effects [65], tag shed - Received: 21 February 2022 Accepted: 20 April 2022 ding [66], and biofouling [67], the general lack of fouling and wear on PAT tags that have been physically recovered on Pacific halibut, after 2–3 years at liberty in the current study, suggests that PAT-tagging can likely be conducted References 1. Carpi P, Loher T, Sadorus LL, Forsberg JE, Webster RA, Planas J, Jasonowicz at multi-year time scales for this, and similar, species. In A, Stewart IJ, Hicks AC. Ontogenetic and spawning migration of Pacific addition, the tags used in the current study (i.e., Mark 10 halibut: a review. Rev Fish Biol Fish. 2021;31:879–908. https:// doi. org/ 10. PAT tags) were considerably larger than current-generation 1007/ s11160- 021- 09672-w. 2. Kong T, Tran H, Prem C. Fisheries data overview (2021). Meeting Docu- satellite tags and tended to suffer higher failure rates than ment IPHC-2022-AM098–06 Rev_1. Seattle: International Pacific Halibut more-recent studies on Pacific [e.g., 68] and Atlantic [38, Commission. 2022. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- 69] halibut. Furthermore, the IPHC has developed both 2022- am098- 06. pdf. Accessed 20 Feb 2022. 3. Stewart I, Hicks A, Webster R, Wilson D. Summary of the data, stock surgical [70] and external [71] tagging techniques resulting assessment, and harvest decision table for Pacific halibut (Hippoglossus in in situ retention periods in excess of 6 years and elec- stenolepis) at the end of 2021. Meeting document IPHC-2022-AM098-10. tronic archival tag manufacturers currently provide multi- Seattle: International Pacific Halibut Commission. 2021. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- 2022- am098- 10. pdf. Accessed 20 Feb sensor tags with operational life and logging capacities on the order of 7–9 years [e.g., Star Oddi (Reykjavik, Iceland) 4. Seitz AC, Loher T, Farrugia TJ, Norcross BL, Nielsen JL. Basin-scale DST-centi (www. star- oddi. com/ produ cts/ data- logge rs/ reproductive segregation of Pacific halibut (Hippoglossus stenolepis). Fish Manag Ecol. 2017;24:339–46. https:// doi. org/ 10. 1111/ fme. 12233. minia ture- depth- logger); Lotek Wireless (St. Johns, Can- 5. Loher T, Seitz AC. Seasonal migration and environmental conditions ada) LAT-2000 Series (https:// www. lotek. com/ produ cts/ experienced by Pacific halibut (Hippoglossus stenolepis), elucidated from lat20 00- series/) archival tags]. Ultimately, there is consider- pop-up archival transmitting tags. Mar Ecol Prog Ser. 2006;317:259–71. https:// doi. org/ 10. 3354/ meps3 17259. able potential to continue expanding the scope and scale of 6. Nielsen JL, Graziano SL, Seitz AC. Fine-scale population genetic structure research on individual behavior and population-level con- in Alaskan Pacific halibut (Hippoglossus stenolepis). Conserv Genet. nectivity using advanced electronic-tagging technology 2010;11(3):999–1012. https:// doi. org/ 10. 1007/ s10592- 009- 9943-8. 7. Drinan DP, Galindo HM, Loher T, Hauser L. Subtle genetic population and a variety of multidisciplinary approaches. structure in Pacific halibut. J Fish Biol. 2016;89:2571–94. https:// doi. org/ 10. 1111/ jfb. 13148. Acknowledgements 8. IPHC (International Pacific Halibut Commission). The Pacific halibut: biol- Kathy Bareza, Dan Rafla, Tucker Soltau Andy Vatter and Tom Wilson conducted ogy, fishery, management. Technical report no. 59. Seattle: International the tagging described in this study, aboard the fishing vessels Free to Wander, Pacific Halibut Commission. https:// www. iphc. int/ uploa ds/ pdf/ tr/ IPHC- Kema Sue, Pacific Sun, and Saint Peter. Erica Anderson-Chau, Claude Dykstra, 2014- TR059. pdf. Accessed 20 Feb 2022. Tracee Geernaert, Ed Henry, Eric Soderlund, and Evangeline White provided 9. Webster RA, Clark WG, Leaman BM, Forsberg JE. Pacific halibut on the invaluable logistical assistance through the IPHC Fishery-Independent Setline move: a renewed understanding of adult migration from a coastwide Survey. Ian Stewart, David Wilson, Josep Planas, and two anonymous reviewers tagging study. Can J Fish Aquat Sci. 2013;70:642–53. https:// doi. org/ 10. provided feedback and suggestions that improved its quality. 1139/ cjfas- 2012- 1371. 10. Hilborn R, Skalski J, Anganuzzi A, Hoffman, A. Movements of juvenile Author contributions halibut in IPHC regulatory areas 2 and 3. Technical report no. 31. Seattle: This study was conceived and planned by TL, who supervised all tag deploy- International Pacific Halibut Commission. 1995. https:// www. iphc. int/ ments, conducted all analyses, and wrote this manuscript. The author read uploa ds/ pdf/ tr/ IPHC- 1995- TR031. pdf. Accessed 20 Feb 2022. and approved the final manuscript. 11. Kaimmer SM. Pacific halibut tag release programs and tag release and recovery data, 1925 through 1998. IPHC technical report no. 41. Seattle: Funding International Pacific Halibut Commission. 2000. https:// www. iphc. int/ The sole Funding body for this work was the International Pacific Halibut uploa ds/ pdf/ tr/ IPHC- 2000- TR041. pdf. Accessed 20 Feb 2022. Commission (IPHC), and all tags were deployed from IPHC-chartered stock 12. Loher T. Analysis of match-mismatch between commercial fishing assessment survey vessels. periods and spawning ecology of Pacific halibut (Hippoglossus stenolepis), based on winter surveys and behavioural data from electronic tags. ICES J Availability of data and materials Mar Sci. 2011;68:2240–51. https:// doi. org/ 10. 1093/ icesj ms/ fsr152. The data sets generated and/or analyzed during this study are Public Domain, 13. Seitz AC, Loher T, Norcross BL, Nielsen JL. Dispersal and behavior of Pacific as per the International Pacific Halibut Commission’s (IPHC) current Data Use halibut Hippoglossus stenolepis in the Bering Sea and Aleutian Islands and Confidentially Policy (dated 2 August 2019), and may be obtained from region. Aquat Biol. 2011;12:225–39. https:// doi. org/ 10. 3354/ ab003 33. the IPHC upon request. 14. Nielsen JK, Seitz AC. Interannual site fidelity of Pacific halibut: potential utility of protected areas for management of a migratory demersal fish. Declarations ICES J Mar Sci. 2017;74(8):2120–34. https:// doi. org/ 10. 1093/ icesj ms/ fsx040. Ethics approval and consent to participate 15. Loher T, Blood CA. Seasonal dispersion of Pacific halibut (Hippoglos - All tagging was conducted under the auspices of the International Pacific sus stenolepis) summering off British Columbia and the US Pacific Halibut Commission. Northwest evaluated via satellite archival tagging. Can J Fish Aquat Sci. 2009;66:1409–22. https:// doi. org/ 10. 1139/ F09- 093. Loher Animal Biotelemetry (2022) 10:18 Page 20 of 21 16. Loher T, Soderlund E. Connectivity between Pacific halibut Hippoglossus 36. Seitz AC, Norcross BL, Wilson D, Nielsen JL. Identifying spawning behavior stenolepis residing in the Salish Sea and the offshore population, demon- in Pacific halibut, Hippoglossus stenolepis, using electronic tags. Environ strated by pop-up archival tagging. J Sea Res. 2018;142:13–124. https:// Biol Fish. 2005;73:445–51. https:// doi. org/ 10. 1007/ s10641- 005- 3216-2. doi. org/ 10. 1016/ jsear es. 2018. 09. 007. 37. Loher T, Seitz AC. Characterization of active spawning season and depth 17. Loher T. Homing and summer feeding site fidelity of Pacific halibut for eastern Pacific halibut (Hippoglossus stenolepis), and evidence of prob - (Hippoglossus stenolepis) in the Gulf of Alaska, established using satellite- able skipped spawning. J Northwest Atl Fish Sci. 2008;41:23–36. https:// transmitting archival tags. Fish Res. 2008;92:63–9. https:// doi. org/ 10. doi. org/ 10. 2960/j. v41. m617. 1016/j. fishr es. 2007. 12. 013. 38. Fisher JAD, Robert D, Le Bris A, Loher T. Pop-up satellite archival tag (PSAT ) 18. Rose GA, Leggett WC. Eec ff ts of biomass-range interactions on the temporal data resolution affects interpretations of spawning behavior catchability of migratory demersal fish by mobile fisheries: an example of and vertical habitat use. Anim Biotelem. 2017;2017(5):27. https:// doi. org/ Atlantic cod (Gadus morhua). Can J Fish Aquat Sci. 1991;48:843–8.10. 1186/ s40317- 017- 0137-8. 19. Rose GA. Cod spawning on a migration highway in the north-west Atlan- 39. Ladd C, Hunt GL Jr, Mordy CW, Salo SA, Stabeno PJ. Marine environ- tic. Nature. 1993;366:458–61. https:// doi. org/ 10. 1038/ 36645 8a0. ment of the eastern and central Aleutian Islands. Fish Oceanogr. 20. Gordoa A, Lesch H, Rodergas S. Bycatch: complementary information for 2005;14(s1):22–38. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00373.x. understanding fish behaviour. Namibian Cape hake (M. capensis and M. 40. Coyle KO. Zooplankton distribution, abundance and biomass relative to paradoxus) as a case study. ICES J Mar Sci. 2006;63:1513–9. https:// doi. water masses in eastern and central Aleutian Island passes. Fish Ocean- org/ 10. 1016/j. icesj ms. 2006. 05. 007. ogr. 2005;14:77–92. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00367.x. 21. Hobday AJ, Hartog JR, Timmiss T, Fielding J. Dynamic spatial zoning to 41. Jahncke J, Coyle KO, Hunt GJ. Seabird distribution, abundance and diets manage southern bluefin tuna (Thunnus maccoyii) capture in a multi- in the central and eastern Aleutian Islands. Fish Oceanogr. 2005;14:160– species longline fishery. Fish Oceanogr. 2010;19:243–53. https:// doi. org/ 77. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00372.x. 10. 1111/j. 1365- 2419- 2010. 00540.x. 42. Konar BH, Edwards MS, Bland A, Metzger J, Ravelo A, Traiger S, Weitzman 22. Keating KA. Mitigating elevation-induced errors in satellite telemetry BP. A swath across the great divide: kelp forests across the Samalga Pass locations. J Wildl Manag. 1995;59:801–8. https:// doi. org/ 10. 2307/ 38019 60. biogeographic break. Cont Shelf Res. 2017;143:78–88. https:// doi. org/ 10. 23. Block AB, Dewar H, Williams T, Prince ED, Farwell C, Fudge D. Archival tag-1016/j. csr. 2017. 06. 007. ging of Atlantic Bluefin tuna (Thunnus thynnus thynnus). Mar Technol Soc 43. Sadorus LL, Goldstein ED, Webster RA, Stockhausen WT, Planas JV, Duffy- J. 1998;32:37–46. https:// doi. org/ 10. 1007/ 978- 94- 017- 1402-0_3. Anderson JT. Multiple life stage connectivity of Pacific halibut (Hippoglos - 24. Pederson MW, Righton D, Thygesen UH, Andersen KH, Madsen H. Geolo- sus stenolepis) across the Bering Sea and Gulf of Alaska. Fish Oceanogr. cation of North Sea cod (Gadus morhua) using hidden Markov models 2021;30(2):174–93. https:// doi. org/ 10. 1111/ fog. 12512. and behavioural switching. Can J Fish Aquat Sci. 2018;65(11):2367–77. 44. Leaman BM, Geerneart TO, Loher T, Clark WG. Further examination of https:// doi. org/ 10. 1139/ F08- 144. biological issues concerning an extended commercial fishing season. In: 25. Le Bris A, Fisher JAD, Murphy HM, Galbraith PS, Castonguay M, Loher Sadorus L, editor. Report of assessment and research activities 2001. Seat- T, Robert D. Migration patterns and putative spawning habitats of tle: International Pacific Halibut Commission; 2002. p. 53–73. Atlantic halibut (Hippoglossus hippoglossus) in the Gulf of St. Lawrence 45. Stabeno P, Schumacher JD, Ohtani K. The physical oceanography of the revealed by geolocation of pop-up satellite archival tags. ICES J Mar Sci. Bering Sea. In: Loughlin TR, Ohtani K, editors. Dynamics of the Bering Sea. 2018;75:135–47. https:// doi. org/ 10. 1093/ icesj ms/ fsx098. Fairbanks: Alaska Sea Grant; 1999. p. 1–28. https:// doi. org/ 10. 4027/ dbs. 26. Nielsen JK, Mueter FJ, Adkinson MD, Loher T, McDermott SF, Seitz AC. 1999. Eec ff t of study area bathymetric heterogeneity on parameterization and 46. Stabeno PJ, Bond NA, Hermann AJ, Kachel NB, Mordy CW, Overland JE. performance of a depth-based geolocation model for demersal fishes. Meteorology and oceanography of the Northern Gulf of Alaska. Fish Ecol Model. 2019;24:18–34. https:// doi. org/ 10. 1093/ icesj ms/ fsx040. Oceanogr. 2004;10:81–98. https:// doi. org/ 10. 1016/J. CSR. 2004. 02. 007. 27. Stabeno P, Ladd C, Napp JM. Transport through Unimak Pass, Alaska. 47. Clark WG, Hare SR. Eec ff ts of climate and stock size on recruitment and Deep-Sea Res. 2002;II(49):5919–30. https:// doi. org/ 10. 1016/ S0967- growth of Pacific halibut. N Am J Fish Manag. 2002;22:852–62. 0645(02) 00326-0. 48. Tamario C, Sunde J, Petersson E, Tibblin P, Forsman A. Ecological and 28. Lanksbury JA, Duffy-Anderson JT, Mier KL, Bisby MS, Stabeno PJ. Distribu- evolutionary consequences of environmental change and management tion and transport patterns of northern rock sole, Lepidopsetta polyxystra, actions for migrating fish. Front Ecol Evol. 2019;7:271. https:// doi. org/ 10. larvae in the southeastern Bering Sea. Prog Oceanogr. 2007;72(1):39–62. 3389/ fevo. 2019. 00271. https:// doi. org/ 10. 1016/j. pocean. 2006. 09. 001. 49. Vestfals C, Ciannelli L, Hoff GR. Changes in habitat utilization of 29. Wright DL, Castellote M, Berchok CL, Ponirakis D, Crance JL, Clapham PJ. slope-spawning flatfish across a bathymetric gradient. ICES J Mar Sci. Acoustic detection of North Pacific right whales in a high-traffic Aleutian 2016;73(7):1875–89. https:// doi. org/ 10. 1093/ icesj ms/ fsw112. Pass, 2009–2015. Endanger Species Res. 2018;37:77–90. https:// doi. org/ 50. Webster RA, Soderlund E, Dykstra CL, Stewart IJ. Monitoring change in 10. 3354/ esr00 915. a dynamic environment: spatio-temporal modelling of calibrated data 30. Ualesi K, Wilson D, Jones C, Rillera R, Jack T. IPHC fishery-independent from different types of fisheries surveys of Pacific halibut. Can J Fish setline survey (FISS) design and implementation in 2020. Meeting Aquat Sci. 2020;77:1421–32. https:// doi. org/ 10. 1139/ cjfas- 2019- 0240. document IPHC-2022-AM098-07. Seattle: International Pacific Halibut 51. Castillo GC. Latitudinal patterns in reproductive life history traits of North- Commission. 2021. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- east Pacific flatfish. In: Baxter B, editor. Proceedings of the international 2022- am098- 07. pdf. Accessed 20 Feb 2022. symposium on North Pacific Flatfish. Fairbanks: Alaska Sea Grant; 1994. p. 31. Courcelles D. Re-evaluation of the length-weight relationship of Pacific 51–72. https:// repos itory. libra ry. noaa. gov/ view/ noaa/ 38469. Accessed 7 halibut (Hippoglossus stenolepis). In: Sadorus L, editor. Report of assess- Apr 2022. ment and research activities 2011. Seattle: International Pacific Halibut 52. Lauth RR, Acuna A. Results of the 2006 eastern Bering Sea continental Commission; 2012. p. 459–70. shelf bottom trawl survey of groundfish and invertebrate resources. 32. Hare SR. Investigation of the role of fishing in the Area 4C CPUE decline. NOAA technical memorandum NMFS-AFSC-176. Seattle: US National In: Sadorus L, editor. Report of assessment and research activities 2004. Marine Fisheries Service. 2007. https:// apps- afsc. fishe ries. noaa. gov/ Publi Seattle: International Pacific Halibut Commission; 2005. p. 185–97.catio ns/ AFSC- TM/ NOAA- TM- AFSC- 176. pdf. Accessed 20 Feb 2022. 33. Loher T. Investigating variability in catch rates of halibut (Hippoglossus 53. Stewart I, Martell S. Development of the 2015 stock assessment. Meeting stenolepis) in the Pribilof Islands: is temperature important? Deep-Sea Res. document IPHC-2015-srb06-02. Seattle: International Pacific Halibut 2008;II(55):1801–8. https:// doi. org/ 10. 1016/j. dsr2. 2008. 04. 002. Commission. https:// www. iphc. int/ uploa ds/ pdf/ srb/ iphc- 2015- srb06- 02. 34. Lynch P. The equation of time and the analemma. Irish Math Soc Bull. pdf. Accessed 7 Apr 2022. 2012;69:47–56. 54. O’Boyle R, Dean M, Legault CM. The influence of seasonal migration on 35. Seitz AC, Norcross BL, Wilson D, Nielsen JL. Evaluating light-based geolo- fishery selectivity. ICES J Mar Sci. 2016;73(7):1774–87. https:// doi. org/ 10. cation for estimating demersal fish movements in high latitudes. Fish Bull. 1093/ icesj ms/ fsw048. 2006;104:571–8. L oher Animal Biotelemetry (2022) 10:18 Page 21 of 21 55. Wilberg MJ, Thorson JT, Linton BC, Berkson J. Incorporating time-varying catchability into population dynamic stock assessment models. Rev Fish Sci. 2010;18(1):7–24. https:// doi. org/ 10. 1080/ 10641 26090 32946 47. 56. Nielsen A, Berg CW. Estimation of time-varying selectivity in stock assess- ments using state-space models. Fish Res. 2014;158:96–101. https:// doi. org/ 10. 1016/j. fishr es. 2014. 01. 014. 57. Cimino M, Cassen M, Merrifield S, Terrill E. Detection efficiency of acoustic biotelemetry sensors on Wave Gliders. Anim Biotelem. 2018;6:16. https:// doi. org/ 10. 1186/ s40317- 018- 0169-4. 58. Lembke C, Lowerre-Barbieri S, Mann DA, Taylow JC. Using three acoustic technologies on underwater gliders to survey fish. Mar Technol Soc J. 2019;2(6):39–52. https:// doi. org/ 10. 4031/ MTSJ. 52.6.1. 59. Stockhausen H, Guðbjörnsson S. The earth’s geomagnetic field and geolocation of fish: first results of a new approach. ICES CM. 2009/B:19. 60. Nielsen JK, Mueter FJ, Adkinson MD, Loher T, McDermott SF, Seitz AC. Potential utility of geomagnetic data for geolocation of demersal fish in the North Pacific Ocean. Anim Biotelem. 2020;8:17. https:// doi. org/ 10. 1186/ s40317- 020- 00204-0. 61. Hussey NE, Orr J, Fisk AT, Hedges KJ, Ferguson SH, Barkley AN. Mark report satellite tags (mrPATs) to detail large-scale horizontal movements of deep water species: first results for Greenland shark (Somniosus microcephalus). Deep Sea Res. 2018;I(134):32–40. https:// doi. org/ 10. 1016/j. dsr. 2018. 03. 62. Gatti P, Fisher JAD, Cyr F, Galbraith PS, Robert D, Le Bris A. A review and tests of validation and sensitivity of geolocation models for marine fish tracking. Fish Fish. 2021;22:1041–66. https:// doi. org/ 10. 1111/ faf. 12568. 63. Loher T, Stephens SM. Use of veterinary ultrasound to identify sex and assess female maturity of Pacific halibut in nonspawning condition. N Am J Fish Manag. 2011;31:1034–42. https:// doi. org/ 10. 1080/ 02755 947. 2011. 64. Drinan DP, Loher T, Hauser L. Identification of genomic regions associated with sex in Pacific halibut. J Heredity. 2018;109(3):326–32. https:// doi. org/ 10. 1093/ jhered/ esx102. 65. Grusha DS, Patterson MR. Quantification of drag and lift imposed by pop-up satellite archival tags and estimation of the metabolic cost to cownose rays (Rhimnoptera bonasus). Fish Bull. 2005;103(1):63–70. 66. Jepsen N, Thorstad EB, Havn T, Lucas MC. The use of external electronic tags on fish: an evaluation of tag retention and tagging effects. Anim Biotelem. 2015;3:49. https:// doi. org/ 10. 1186/ s40317- 015- 0086-z. 67. Hammerschlag N, Cooke SJ, Gallagher AJ, Godley BJ. Considering the fate of electronic tags: interactions with stakeholders and user responsibil- ity when encountering tagged aquatic animals. Methods Ecol Evol. 2014;5:1147–53. https:// doi. org/ 10. 1111/ 2041- 210X. 12248. 68. Loher T, Dykstra CL, Simeon A, Hicks AC, Stewart IJ, Wolf N, Harris BP, Planas JV. Estimation of post-release mortality using acceleration-logging archival pop-up tags for Pacific halibut (Hippoglossus stenolepis) discarded from longlines. N Am J Fish Manag. 2021;42(1):37–49. https:// doi. org/ 10. 1002/ najfm. 10711. 69. Le Bris A, Fisher JAD, Murphy HM, Galbraith PS, Castonguay M, Loher T, Robert D. Migration patterns and putative spawning habitats of Atlantic halibut (Hippoglossus hippoglossus) in the Gulf of St. Lawrence revealed by geolocation of pop-up satellite archival tags. ICES J Mar Sci. 2018;1(1):135–47. https:// doi. org/ 10. 1093/ icesj ms/ fsx098. 70. Loher T, Rensmeyer R. Physiological responses of Pacific halibut, Hippoglossus stenolepis, to intracoelomic implantation of electronic archival tags, with a review of tag implantation techniques employed in flatfishes. Rev Fish Biol Fish. 2011;21(1):1027–37. https:// doi. org/ 10. 1007/ Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : s11160- 010- 9192-4. 71. Loher T, Geernaert TO. Captive holding to develop long-term archival fast, convenient online submission tagging protocols in Pacific halibut. In: Sadorus L, editor. Report of assess- thorough peer review by experienced researchers in your ﬁeld ment and research activities 2015. Seattle: International Pacific Halibut Commission; 2016. p. 445–63. rapid publication on acceptance support for research data, including large and complex data types Publisher’s Note • gold Open Access which fosters wider collaboration and increased citations Springer Nature remains neutral with regard to jurisdictional claims in pub- maximum visibility for your research: over 100M website views per year lished maps and institutional affiliations. At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Animal Biotelemetry Springer Journals http://www.deepdyve.com/lp/springer-journals/dispersal-and-seasonal-movements-of-pacific-halibut-hippoglossus-cT9y4008ar

Loading next page...

References (59)

JK Nielsen (2017)
10.1093/icesjms/fsx040
ICES J Mar Sci, 74
T Loher (2008)
10.2960/j.v41.m617
J Northwest Atl Fish Sci, 41
GA Rose (1991)
10.1139/f91-100
Can J Fish Aquat Sci, 48
T Loher, TO Geernaert, L Sadorus (2016)
Captive holding to develop long-term archival tagging protocols in Pacific halibut
Report of assessment and research activities 2015
PJ Stabeno (2004)
10.1016/J.CSR.2004.02.007
Fish Oceanogr, 10
D Courcelles, L Sadorus (2012)
Re-evaluation of the length-weight relationship of Pacific halibut (Hippoglossus stenolepis)
Report of assessment and research activities 2011
RA Webster (2013)
10.1139/cjfas-2012-1371
Can J Fish Aquat Sci, 70
T Loher (2006)
10.3354/meps317259
Mar Ecol Prog Ser, 317
NE Hussey (2018)
10.1016/j.dsr.2018.03.002
Deep Sea Res, I
T Loher (2018)
10.1016/jseares.2018.09.007
J Sea Res, 142
DL Wright (2018)
10.3354/esr00915
Endanger Species Res, 37
MW Pederson (2018)
10.1139/F08-144
Can J Fish Aquat Sci, 65
T Loher (2009)
10.1139/F09-093
Can J Fish Aquat Sci, 66
T Loher (2008)
10.1016/j.fishres.2007.12.013
Fish Res, 92
T Loher (2011)
10.1080/02755947.2011.635240
N Am J Fish Manag, 31
P Stabeno (1999)
10.4027/dbs.1999
AB Block (1998)
10.1007/978-94-017-1402-0_3
Mar Technol Soc J, 32
A Gordoa (2006)
10.1016/j.icesjms.2006.05.007
ICES J Mar Sci, 63
SR Hare, L Sadorus (2005)
Investigation of the role of fishing in the Area 4C CPUE decline
Report of assessment and research activities 2004
N Hammerschlag (2014)
10.1111/2041-210X.12248
Methods Ecol Evol, 5
C Lembke (2019)
10.4031/MTSJ.52.6.1
Mar Technol Soc J, 2
T Loher (2011)
10.1093/icesjms/fsr152
ICES J Mar Sci, 68
T Loher (2021)
10.1002/najfm.10711
N Am J Fish Manag, 42
C Ladd (2005)
10.1111/j.1365-2419.2005.00373.x
Fish Oceanogr, 14
N Jepsen (2015)
10.1186/s40317-015-0086-z
Anim Biotelem, 3
KO Coyle (2005)
10.1111/j.1365-2419.2005.00367.x
Fish Oceanogr, 14
DS Grusha, MR Patterson (2005)
Quantification of drag and lift imposed by pop-up satellite archival tags and estimation of the metabolic cost to cownose rays (Rhimnoptera bonasus)
Fish Bull, 103
JL Nielsen (2010)
10.1007/s10592-009-9943-8
Conserv Genet, 11
P Stabeno (2002)
10.1016/S0967-0645(02)00326-0
Deep-Sea Res, II
AC Seitz (2011)
10.3354/ab00333
Aquat Biol, 12
T Loher (2008)
10.1016/j.dsr2.2008.04.002
Deep-Sea Res, II
WG Clark (2002)
10.1577/1548-8675(2002)022<0852:EOCASS>2.0.CO;2
N Am J Fish Manag, 22
T Loher (2011)
10.1007/s11160-010-9192-4
Rev Fish Biol Fish, 21
AC Seitz (2005)
10.1007/s10641-005-3216-2
Environ Biol Fish, 73
RA Webster (2020)
10.1139/cjfas-2019-0240
Can J Fish Aquat Sci, 77
JK Nielsen (2020)
10.1186/s40317-020-00204-0
Anim Biotelem, 8
J Jahncke (2005)
10.1111/j.1365-2419.2005.00372.x
Fish Oceanogr, 14
P Carpi (2021)
10.1007/s11160-021-09672-w
Rev Fish Biol Fish, 31
KA Keating (1995)
10.2307/3801960
J Wildl Manag, 59
DP Drinan (2018)
10.1093/jhered/esx102
J Heredity, 109
A Le Bris (2018)
10.1093/icesjms/fsx098
ICES J Mar Sci, 1
DP Drinan (2016)
10.1111/jfb.13148
J Fish Biol, 89
BH Konar (2017)
10.1016/j.csr.2017.06.007
Cont Shelf Res, 143
MJ Wilberg (2010)
10.1080/10641260903294647
Rev Fish Sci, 18
P Gatti (2021)
10.1111/faf.12568
Fish Fish, 22
GA Rose (1993)
10.1038/366458a0
Nature, 366
JAD Fisher (2017)
10.1186/s40317-017-0137-8
Anim Biotelem, 2017
R O’Boyle (2016)
10.1093/icesjms/fsw048
ICES J Mar Sci, 73
M Cimino (2018)
10.1186/s40317-018-0169-4
Anim Biotelem, 6
P Lynch (2012)
10.33232/BIMS.0069.47.56
Irish Math Soc Bull, 69
AC Seitz, BL Norcross, D Wilson, JL Nielsen (2006)
Evaluating light-based geolocation for estimating demersal fish movements in high latitudes
Fish Bull, 104
JA Lanksbury (2007)
10.1016/j.pocean.2006.09.001
Prog Oceanogr, 72
BM Leaman, TO Geerneart, T Loher, WG Clark, L Sadorus (2002)
Further examination of biological issues concerning an extended commercial fishing season
Report of assessment and research activities 2001
A Nielsen (2014)
10.1016/j.fishres.2014.01.014
Fish Res, 158
C Tamario (2019)
10.3389/fevo.2019.00271
Front Ecol Evol, 7
AC Seitz (2017)
10.1111/fme.12233
Fish Manag Ecol, 24
C Vestfals (2016)
10.1093/icesjms/fsw112
ICES J Mar Sci, 73
AJ Hobday (2010)
10.1111/j.1365-2419-2010.00540.x
Fish Oceanogr, 19
LL Sadorus (2021)
10.1111/fog.12512
Fish Oceanogr, 30

Publisher: Springer Journals
Copyright: Copyright © The Author(s) 2022
eISSN: 2050-3385
DOI: 10.1186/s40317-022-00288-w
Publisher site: See Article on Publisher Site

Abstract

Background: Understanding connectivity is critical to the management of exploited fish stocks, but migratory dynamics of Pacific halibut (Hippoglossus stenolepis) in the Bering Sea and Aleutian Islands region are not well-under - stood. In the current study, 145 Pacific halibut ≥ 82 cm fork length were tagged with Pop-up Archival Transmitting (PAT ) tags to evaluate interannual dispersal, seasonal migration, and depth-specific habitat use. Results: Endpoint locations obtained after 1 year at liberty (n = 79), fishery recoveries after 2–3 years at liberty (n = 5), and at-liberty geopositions based on light data (n = 5313 estimates from 109 fish) indicated geographically distinct movement patterns: Pacific halibut tagged in the Western and Central Aleutian Islands remained within the island groups in which the fish had been tagged; fish in the eastern Bering Sea remained in that ocean basin, moving among International Pacific Halibut Commission (IPHC) regulatory areas and into Russian waters; those tagged south of Unimak Pass in IPHC Regulatory Area 4A displayed the greatest amount of emigration, dispersing eastward both seasonally and interannually to as far south as Washington State. Analysis of daily maximum depth and temperature data from 113 individuals demonstrated group-level variation in summer temperatures experienced by the fish and in the timing, duration, and synchrony of movement to deep-water wintering grounds. Conclusions: Depth-specific habitat use was suggestive of regionally explicit migratory contingents, while interan- nual dispersal patterns were consistent with the existence of multiple functional spawning units. The results may guide future research to examine cross-basin connectivity in the Northern Bering Sea and provide inputs for numeri- cal modelling of individual movements, larval advection, and recruitment analyses. Keywords: Hippoglossus stenolepis, Migration, Spawning, Satellite tagging, Bering Sea Okhotsk, and southward into the northern Sea of Japan. Background These fish are highly migratory, with complex and widely Pacific halibut (Hippoglossus stenolepis) is a wide-rang - dispersing life-history stages [1]. The species supports ing Pleuronectid flatfish species distributed from north - substantial subsistence, recreational, and commercial ern California, USA, northward throughout the Gulf of fisheries [2] and although managed as a single unit stock Alaska and Bering Sea, westward into the Russian Sea of throughout US and Canadian waters [3] research con- ducted in North American waters suggests internal pop- *Correspondence: tim.at.martingale@gmail.com ulation structure at roughly ocean-basin scales [4]. Data collected using Pop-up Archival Transmitting (PAT) tags Martingale Marine Ecological Research, 7019 14th Ave NE, Seattle, WA 98115, USA are consistent with the hypothesis that mature fish have © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Loher Animal Biotelemetry (2022) 10:18 Page 2 of 21 a higher probability of remaining within the basins of the known to conduct seasonal spawning migrations that are eastern Pacific Ocean (defined as the Gulf of Alaska, east - composed of both onshore–offshore movement [5, 12, ern Bering Sea, and the Aleutian Islands region) to spawn 13] and large-scale (> 1000 km) alongshore redistribution than dispersing among basins to do so [4, 5]. Consist- [5, 13–16]. Without constraining recoveries to a relatively ent with this, although population genetic analyses have short window representing the summer feeding period demonstrated little genetic differentiation throughout [12] or having the ability to evaluate each fish’s individ - North American waters [6, 7], best-available evidence ual’s location within its seasonal migration trajectory, suggests that the western Aleutian Islands may support seasonal movement may be confounded with interannual a population component that is significantly different dispersal, whether ontogenic [10] or due to adult straying than in either the eastern Pacific Ocean [6, 7] or in Asian [17]. Resolving among forms of dispersal can be impor- waters of the southern Sea of Okhotsk [7]. tant for understanding how fish movements interact with However, neither existing satellite-tag data nor genetic target [sensu 18, 19] and non-target [sensu 20, 21] fisher - analyses provide insight into connectivity at the annual ies, and for properly defining regional stock components to multi-annual scales over which Pacific halibut fisher - [sensu 3] intended to reflect effective spawning biomass ies are managed and prosecuted or the decadal-scale units [sensu 4]. processes comprising the ontogeny of a species in which In the current study, interannual connectivity in Pacific 8–16 year-olds are the most-represented demographic halibut was investigated using PAT tags. In brief, PAT in the directed fishery [3] and is known to live in excess tags are electronic tags that contain an automated release of 50 years [8]. Satellite tagging has primarily investi- mechanism, sensor package, and satellite-broadcast capa- gated seasonal dispersal (i.e., movements executed over bilities, allowing for environmental data to be collected the course of 6–7 months), while genetic studies speak while attached to the host fish. Recovery/broadcast dates to reproductive isolation over hundreds to thousands of (and, hence, period at liberty) may be pre-specified, and years. As such, there is a need to conduct adult connec- the tag’s final position is determined by the receiving sat - tivity studies ranging from interannual to generational- ellite [22]. This has the advantage of allowing for determi - scale to bridge this gap. nation of final location even if fish move to areas absent Perhaps the most substantial effort to better under - of fishery effort, such as where fisheries are excluded stand interannual dispersal in Pacific halibut was a large- (e.g., the IPHC Closed Area; Fig. 1), or where reporting scale Passive Integrated Transponder (PIT tag) study of physical recoveries is reduced (e.g., in Asian waters conducted along the North American coast between for fish tagged in North America). In addition, locations 2001 and 2009 [9]. Over the course of that study, 67,000 during time at liberty can be inferred from archived light fish were tagged and > 3000 were recovered via intensive data [sensu 23] and may allow for state-space modelling dockside sampling. Statistical modeling of the recovery of movement [24–26] that cannot be achieved using con- distributions, along with analyses derived from a long his- ventional mark-recapture data. PAT tags have been used tory of conventional tagging [10, 11], provided consider- to investigate seasonal movements in Pacific halibut; in able insight regarding migration among geographic areas, particular, identifying key spawning grounds in the BSAI including relationships between movement rates and fish region and their connectivity to components of the com- size. However, these results were subject to limitations mercial fishery [13]. However, due to their relatively high due to the nature of the tags and the recovery mecha- cost (~ $4000 US), the distribution of such releases has nism [1]. Conventional tag recoveries (including PIT been patchy and has lacked coverage in one region that tags) are dependent primarily upon commercial fisher - may be key to understanding basin-scale stock structure ies and, thus, relatively little information can be obtained in Pacific halibut: adjacent to Unimak Pass (Fig. 1), which in regions, where fishing effort is low or within which represents a primary connection between the Eastern the mixing of tagged fish into the fishable population is Bering Sea (EBS) and Gulf of Alaska (GOA), from both incomplete [9]. As a result, limited insight was obtained an oceanographic [27] and biological [28, 29] perspective. regarding dispersal within the Bering Sea and Aleutian Here, we use satellite-generated final positions, fish - Islands (BSAI) region or between the BSAI region and the ery-recovery locations, at-liberty longitudes estimated Gulf of Alaska (GOA). Also, the PIT tags were implanted from ambient light data, and depth trajectories during subcutaneously with no external markings placed on the time-at-liberty [sensu 12] to infer regional variation in fish; the tags could only be detected electronically. Thus, interannual dispersal, the timing and duration of sea- the location data were limited to the scale of entire fish - sonal migrations, and depth-specific regional habitat ing trips: i.e., it was known within which commercial use of Pacific halibut in the BSAI region and in the offload each fish had been recovered, but not the precise western GOA adjacent to Unimak Pass. recapture location or date. In addition, Pacific halibut are L oher Animal Biotelemetry (2022) 10:18 Page 3 of 21 Fig. 1 International Pacific Halibut Commission regulatory areas and geographic features referenced in this manuscript. Note that the western Aleutian Islands (i.e., the Near and Rat Islands) are located in the eastern hemisphere, with IPHC Regulatory Area 4B crossing the Antimeridian at approximately Amchitka Pass Mk10 PAT tags measured 170 mm in length and Methods 40 mm in maximum diameter, with a plastic-coated Tag deployments braided-cable antenna protruding from the distal end. A total of 145 Pacific halibut were tagged with either PSAT Flex tags measured 131 mm in length and 42 mm Wildlife Computers (Redmond, Washington, USA) in maximum diameter, also with a distal antenna. The Mk10 or with Lotek Wireless (St. Johns, Newfound- tags were programmed to record depth (at a resolu- land, Canada) PSAT Flex Pop-up Archival Transmitting tion of 4 m) every 30 s and ambient light levels every (PAT) tags during the boreal summers (June–August) minute and, upon surfacing, to transmitting their data of 2008 and 2009 (Mk10), and during June 2016 to the US National Oceanic and Atmospheric Admin- (PSAT Flex) (Table 1). All fish were captured during istration’s polar-orbiting satellites, administered by the the IPHC’s Fishery-Independent Setline Survey (FISS) Advanced Research and Global Observation System [30] using benthic longline gear rigged with 16/0 cir- (Argos). For tags that were not physically recovered cle hooks at 5.5 m spacing, baited with chum salmon prior to their programmed broadcast dates, archived (Oncorhynchus keta), and soaked for approximately 6 h environmental data were transmitted as aggregated prior to retrieval. Fish that were in excellent condition (“binned”) depth data, depth profiles, and light-based and of commercially legal size [i.e., ≥ 82 cm fork length twilight curves that could subsequently be used to (FL)] were tagged at pre-selected and regularly spaced define habitat use and estimate location during time- stations throughout the BSAI and far-western GOA at-liberty. Depth data were binned into consecutive (Fig. 2). Individual fish were randomly selected at each 8-h blocks that included the minimum and maximum station to achieve a tagged demographic that was rep- resentative of the surveyed population. Loher Animal Biotelemetry (2022) 10:18 Page 4 of 21 Table 1 Tag deployment data Tagging region Date range(s) Sample sizes FL range (mean ± SD) Total Displ. Light Depth Usable Western Aleutian 06 Jun–31 Aug, 2008 22 18 20 20 20 84–171 (106.9 ± 20.8) Central Aleutian 28 May–03 Jun, 2008 18 15 16 16 16 83–129 (95.2 ± 13.0) Transition Zone 18 Jun–13 Jul, 2008 24 17 23 23 23 82–122 (94.4 ± 10.87) EBS Shelf Edge 10 Jun–01 Aug, 2008; 49 19 32 32 34 82–135 (97.7 ± 17.7) 11–26 Jun, 2016 30 Jun–20 Jul, 2008; 32 9 18 22 23 82–139 (119.9 ± 28.6) EBS Shelf Islands 06 Jun–21 Aug, 2009 Dates over which fish were tagged, total number of tags deployed, number of tags that either reported remotely after 1 year at liberty or were recaptured multiple years after deployment (Displ.), produced light data suitable for estimation of at-liberty longitude or daily maximum depth data and, therefore, produced data used in analyses (Usable); and the range and mean forklengths of the fish included, for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the Aleutian Islands and Eastern Bering Sea (EBS), and a transition zone between those ecosystems and the western Gulf of Alaska depths experienced by the fish during each binning subunits at fishery-relevant scales. In particular, prior period. migratory [4] and genetic [7] research has suggested that During tagging, the fork length of each tagged fish was east–west movement of Pacific halibut may be limited recorded and tags were secured to the fish via 15–18 cm by deep-water passes through the Aleutian Island Chain leaders of 130-kg test nylon monofilament coated with (i.e., at Amchitka and Amukta Passes; Fig. 1). Tagging adhesive-lined polyolefin, anchored to the fish using studies have demonstrated different rates and distances surgical-grade titanium darts that were inserted through of dispersal for Pacific halibut found north versus south the pterygiophores on the eyed side of the fish, roughly of Unimak Pass [9]. Analyses also suggest that fisheries 2.5 cm medial to the dorsal fin. Tag assemblies weighed of shallow island ecosystems in the Eastern Bering Sea approximately 120 g in air, representing 0.15–1.5% of (EBS) may display different dynamics than farther off - the initial body weights of the tagged fish as estimated shore [32, 33]. Thus, for the subsequent analyses, data from their lengths [31]; however, the tags were slightly were grouped by tag-deployment location within geo- (i.e., < 5 g) buoyant in water. The tags were programmed graphically distinct regions (Fig. 2): (1) the Western Aleu- to detach and report during the non-spawning (feed- tian Islands, representing all waters south of 55.5° N lat. ing) period, 365 days after deployment, thereby produc- and west of a line extending from 51° N lat. by 180° W ing endpoint data representing interannual dispersal lon. (i.e., Amchitka Pass) to 54° N by 178° W; (2) the Cen- that was absent of seasonal migration to the greatest tral Aleutian Islands, representing waters south of 53° N degree possible. Mk10 PAT tags were programmed to and between 171.7° W (i.e., Amukta Pass) and a line report within 2 days of the occurrence of any premature extending from 51° N by 180° W to 54° N by 178° W; (3) release event (i.e., upon floating to the surface prior to the BSAI-GOA Transition Zone, representing southern the programmed date of detachment). PSAT Flex were IPHC Regulatory Area 4A and spanning 164–171.7° W so not equipped with surface-detection capabilities and, as to include all waters south, and within 40 km to the therefore, only initiated satellite broadcasts on the pro- north, of the Fox Islands (i.e., representing a region of grammed tag-detachment date. For both tag types, the mixing among the eastern Aleutian Islands, far-western dart-and-tether assemblies remained embedded in the GOA, and around Unimak Pass); (4) the Eastern Bering fish following tag detachment to serve as a conventional Sea Shelf Edge, representing waters of the continental tag that allowed for a third location to be obtained if the shelf edge from 135 to 500 m depth, north of 54.6° N, and fish were subsequently recaptured. Both the tag bodies spanning 165.4–178.5° W, and; (5) the Eastern Bering Sea and tethers were printed with tag numbers and contact Shelf Islands, representing shallow (< 105 m) water within information and unique identification numbers were also 40 km of St. Matthew Island and the Pribilof Islands. In engraved on the tagging darts. a Pacific halibut management context, the Western and Central Aleutian Regions comprise the western and east- Tagging regions ern halves, respectively, of IPHC Regulatory Area 4B; the The analyses contained herein are intended to describe Transition Zone comprises southern Regulatory Area dispersal patterns during the summer feeding period, 4A; the EBS Shelf Edge spans northern Area 4A and the within geographic regions that represent biological offshore extent of Regulatory Area 4D; and the EBS Shelf L oher Animal Biotelemetry (2022) 10:18 Page 5 of 21 Fig. 2 Deployment and endpoint locations of Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during the boreal summers of 2008, 2009, and 2016. Colors represent five geographic regions within which individuals were pooled for analysis. Only fish whose tags produced data are plotted. Circles indicate tagging locations, where closed symbols represent final locations that were amenable to analyses of interannual dispersal; open symbols are fish whose final locations were not included but which produced depth or light data. Open diamonds connected by solid lines indicate final positions after 1 year at liberty; dashed lines indicate fish that were physically recaptured after 2–3 years at liberty. The precise location of one fish that was recaptured in coastal Kamchatkan waters was unknown, and is denoted with a question mark. Note that Kamchatka and the Western Aleutians are located in the eastern hemisphere, with Antimeridian located at approximately Amchitka Pass in the south and the western Gulf of Anadyr in the north Islands are contained within Regulatory Areas 4C (Pri- successfully reported via satellite, positions were deter- bilof Islands) and 4D (St. Matthew Island) (Figs. 1 and 2). mined by the receiving satellite(s) from the Doppler shift of the transmitted radio frequency in successive uplinks Final locations, linear displacement, and interannual received during each satellite pass [22]. Each satellite- dispersal generated position estimate was assigned a “location Final fish locations were obtained for fish that were either class” by Service Argos that indicated the positional accu- recaptured in fisheries or whose tags reported through racy of the estimate; locations reported herein represent the Argos satellite system. Tag rewards were offered to the first broadcast for each tag for which positional accu - individuals who captured tagged fish, as well as for tags racy was reported to be < 1000 m. Mk10s that detached found adrift or awash following their detachment. Final earlier than their scheduled date were programmed to locations for the recaptured fish reported herein repre - broadcast upon surfacing if they remained at zero depth sent the coordinates that were provided to the IPHC by for a 2-day period; locations reported herein represent the individuals who returned those tags. For tags that the first position obtained thereafter. Tag drift that may Loher Animal Biotelemetry (2022) 10:18 Page 6 of 21 have occurred between tag surfacing and reporting was of individual twilight events can be biased relative to true not estimated. PSATs were not equipped with zero-depth twilight due to water clarity or cloud cover. In contrast, premature-release detection and were, therefore, pro- the mean of two similarly biased events during a single grammed to report only after 365 days at liberty. As such, day may still result in an accurate estimate of solar noon, reliable final locations for fish whose PSATs detached and so all days in which both sunrise and sunset could be prematurely could not be obtained and these locations characterized were included in the analysis. were omitted from analysis. Longitude was calculated from twilight data by: (1) For analyses of interannual dispersal, only fish whose computing local solar noon as the mean of the mid-sun- final positions were obtained after 1 full year at liberty rise and mid-sunset times; (2) adjusting local solar noon were included; herein defined as occurring within a win - to the nearest minute (i.e., at the same resolution as the dow of 360–370 days after tagging. Positions obtained light data) in accordance with the Equation of Time [34] after shorter or longer periods were omitted, because for the given date; (3) comparing the adjusted local noon they had the potential to be biased due to seasonal migra- to 1200 UTC (Coordinated Universal Time), wherein tion, which may be substantial in magnitude [4, 5]. Such each hour of offset between those two times indicated 15° endpoints may be more representative of spawning/win- of longitude relative to the Prime Meridian. tering locations than the locations to which those fish When longitude is known, latitude may be deter- would have migrated the following summer, whether mined based on day length on any given date. However, displaying homing behavior [14, 17], straying [17], or light-based latitudes were not estimated herein, because engaged in ontogenic redistribution [9]. The dispersal results for Pacific halibut have been found to be highly of individuals whose locations were obtained during the variable [35] due to reductions in estimated day length winter spawning period has been reported elsewhere that arise from cloud cover and water turbidity in the [4]. Herein, interannual displacement between mark and North Pacific Ocean. In addition, changes in the vertical reporting was calculated for each fish included in the distribution of the tagged fish from day-to-day and sea - analysis as the linear distance between its tagging loca- sonally can introduce intractable variance and bias into tion and either its physical recapture or first accepted twilight-derived estimates of local day length. satellite-derived reporting position, ignoring land masses Plots of light-based longitude estimates according to that might lie between those endpoints. date (i.e., demonstrating potential east–west disper- sal during time at liberty) were constructed for each At‑liberty position estimation fish and using regionally aggregated data. Inferring the Daily longitude estimates of tagged Pacific halibut dur - displacement of Pacific halibut from light-based longi - ing their times at liberty were estimated using archived tude is somewhat subjective, because factors that skew ambient light data. MK10 PAT tags were programmed light readings taken at depth can introduce variability to identify and broadcast “twilight events” in which light into those estimates, even for relatively stationary tags levels either increased or decreased at rates that were [35]. Isolated position estimates must be considered consistent with sunrise and sunset, respectively. Wildlife cautiously, whereas series of positions located consist- Computers’ proprietary software Global Position Esti- ently away from a point of reference are more likely to mator version 3 (GPE3) was used to extract the twilight indicate displacement of the fish. Individual outliers are data from each tag’s broadcast file, and each twilight often easy to identify within individual trajectories (e.g., event (i.e., a series of nine sequential light readings) was Fig. 3a); series of positions estimates were evaluated for visually inspected. Putative sunrise/sunset events that plausibility considering the magnitude of displacement did not exhibit smoothly sloping light levels or contained that they nominally suggested with respect to their dis- null values were rejected. In addition, twilight events tance from known release-recovery locations and other were rejected if the fish displayed a change in depth dur - light-based longitude estimates over the time elapsed ing the putative event such that the changing light levels between consecutive position estimates. Herein, series might have resulted from vertical movement as opposed of estimates that suggested smoothly trended east–west to true sunrise or sunset; e.g., light levels declining as a redistribution were considered plausible. Series of five fish descended. Curves were not rejected if a fish’s depth or more “clumped” estimates were evaluated for plausi- change was inverse to the slope of the twilight event; bility on the basis of their distance from prior and sub- e.g., light levels increasing as a fish descended. The lat - sequent positions. Implied displacement was considered ter would be consistent with crepuscular vertical migra- implausible if it could only have been reached from the tions and provide additional evidence that the event was nearest series of prior plausible position estimates by the truly sunrise/sunset. Days on which only one twilight fish having swum continuously at ≥ 2 m per second for event was identifiable were omitted, because the timing consecutive 24-h periods without resting. L oher Animal Biotelemetry (2022) 10:18 Page 7 of 21 Fig. 3 Light-based longitude estimates (upper panels) and daily maximum depth profiles (lower panels) during periods at liberty (2008 and 2009) of: a one Pacific halibut tagged in the Western Aleutian Islands, from date of tagging (top of upper panels) to date of final tag reporting (bottom of upper panels), and; b, c two Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the far-western Gulf of Alaska. All data were obtained via satellite transmission (i.e., the tags were not recovered). The approximate longitudinal span of International Pacific Halibut Commission regulatory areas is indicated by shading; note that the Western Aleutians are located in the eastern hemisphere, with the Antimeridian bisecting IPHC Regulatory Area 4B at approximately Amchitka Pass. The fish depicted in a remained in shallow water throughout the year and the light data provide no evidence that it departed its tagging location. However, note the single outlier position in early June of 2009: this likely represents a biased daylight curve, as the position is farther eastward than a Pacific halibut could likely swim during the time elapsed relative to the remainder of the position estimates. The fish depicted in b and c emigrated from their tagging region and had final tag-reporting locations in the Gulf of Alaska and coastal Washington State, respectively; the timing of their migrations was coincident with their movement to deep water in autumn In addition to plotting estimates over time, group-level histograms. Those distributions were inspected to evalu - at-liberty redistribution was evaluated using longitude ate whether they were described by a single mode or estimates that were standardized relative to each indi- multiple modes. For each mode that was identified, vidual’s known initial location, as: (Lon − Lon ), where mean, median, and skewness were calculated. Median i t Lon = the longitude at the initial (tagging) location and and skewness values near 0 would indicate little net Lon = light-based longitude estimate at time . That is, directional shift in at-liberty fish distributions relative to t t each light-based longitude estimate was expressed as known tagging locations; negative values would indicate being X° either to the west (negative values) or east (posi- tive values) of the fish’s known initial location. The data were then pooled by region and plotted as frequency Loher Animal Biotelemetry (2022) 10:18 Page 8 of 21 net westward redistribution; and positive values eastward annual mean depth occupied; and the shallow-water redistribution. phase defined as the remainder of the year. Maximum depth data for each fish were then inspected to determine Seasonal depth distributions individual deep-water phases, similarly defined as the Depth data were analysed to characterize depth-spe- longest temporally consistent period (i.e., occurring dur- cific habitat use and the extent and timing of migra - ing boreal winter, as per the regional mean profile) dur - tion to deep-water spawning grounds. Herein, it was ing which that individual persistently occupied depths assumed that Pacific halibut spend considerable time at deeper than the pooled regional mean. These data were or near the seafloor each day, such that the maximum then used to identify the deep- and shallow-water phases depth recorded during each 8-h data-binning period was for each individual and compute mean depth within each likely to represent seafloor depth, and multiple analyses phase (i.e., deep-water and shallow-water) for those indi- were conducted using all tags within each geographic viduals. Individual mean depths were then averaged to region for which maximum-depth data were obtained derive regionally explicit mean depths occupied during (Table 1). A primary objective was to evaluate the degree each depth phase; this eliminated, to the greatest degree to which seasonal migration may vary by region. Such practicable, biasing the regional means in favor of fish movements are typically composed of offshore migra - whose tags transmitted more data than others. tion in the autumn with return to shallow waters in the To investigate whether water temperature might spring. Prior research [4] has compared geographic vari- influence the timing of the autumn offshore migra - ance in seasonal habitat use by defining a fixed spawning tion, conditions experienced by migratory individuals period among all regions of interest that was equivalent were compared among geographic regions. Migratory to the period over which active spawning (i.e., spawn- individuals were defined as those that moved to depths rise behavior) [36] has been reported from archival- below the regional mean annual depth in autumn and tagging data [12, 37]. However, the high-resolution data remained in deep water for at least 4 weeks thereafter. required to confidently conduct spawn-rise analyses [38] For each such individual, mean temperature experienced for Pacific halibut are primarily limited to the GOA and during the 6-week period spanning 28 July and 7 Sep- applying any single spawning period to all regions within tember was calculated; this period was chosen, because an analysis may fail to capture important geographic dif- it represented the period during which the tagged Pacific ferences in spawn timing; and, in the current context, off - halibut in all regions persistently occupied, on average, shore–onshore migrations that occur prior to, and after, their shallowest depths. For each migratory individual, the active spawning are of considerable interest. Herein, an autumn departure date was then identified that rep - the available depth data were used to define regionally resented the date on which the fish descended below the explicit summer shallow-water and winter deep-water regional average depth and persistently occupied deeper phases after standardizing data resolution among all water thereafter. Based on this date, the average tempera- tags analyzed by converting high-resolution (i.e., 30 s) ture experienced by the individual was then computed data from physically recovered tags to 8-h-binned mini- for the 1-week period (i.e., 168 h) immediately prior to its mum and maximum values. Each region’s pooled, aver- departure to deep water, as well as for each of the three age annual depth profile was divided into two periods 1-week periods prior to that (i.e., producing 4 weekly representing the occupation of habitat that was either values spanning the 4 weeks prior to autumn depar- shallower than, or deeper than, the region’s population- ture). Mean summer temperatures experienced among level annual mean. First, mean depth within each region fish within each region were compared to determine was computed among all individuals for all available bin- whether fish inhabiting different regions had experienced ning periods throughout deployment. Next, a region- significantly different pre-migration conditions. Within ally explicit annual mean depth occupied was calculated regions, the five sequential temperatures were compared using the longest duration for which data were contigu- to determine whether conditions changed significantly in ously available among regions: i.e., 21 June of the year of the weeks prior to offshore migration, and to potentially tagging through 3 June of the year of tag-reporting. Then, identify threshold temperature(s), or net change in tem- mean maximum daily depth (MMDD) profiles [sensu 12] perature, that might have initiated migration. were constructed for each region as 3-day moving aver- ages of the binned data (i.e., averaged across twelve con- Statistical analyses secutive 8-h bins), to smooth short-period fluctuations. For all fish for which a date of tag detachment was Each region’s deep-water (spawning) phase was then obtained, the relationship between mean fish length and defined as the longest period over which the smoothed tag-retention period (i.e., to test whether subsequent mean depths were continuously deeper than the region’s analyses might be size-biased) was examined via linear L oher Animal Biotelemetry (2022) 10:18 Page 9 of 21 regression. Regional differences in mean fish lengths, Table 2 Tukey HSD comparisons: fish lengths and tagging depths tagging depths, seasonal depth distributions (i.e., mean shallow- and deep-water phase depths), and linear dis- Comparison WA CA TZ SE SI placement were investigated via multiple analysis of vari- WA – 0.235 0.207 0.228 0.480 ance (MANOVA) in which region was specified as the CA 0.978 – 1.000 1.000 0.000 independent (categorical) variable and the other factors TZ 0.976 1.000 – 1.000 0.000 as dependent. Tukey HSD post hoc multiple comparisons SE 0.754 0.412 0.427 – 0.000 were conducted for factors that were determined to be SI 0.190 0.421 0.464 0.021 – significant. Regional variance in the proportion of satel - lite transmissions received per tag, mean temperatures Results of Tukey HSD pairwise post hoc comparisons of the mean length of fish contributing to analyses in the current study (above the diagonal) and the mean experienced by fish, and the mean number of light-based depths at which fish were captured and released (below the diagonal), for Pacific longitude estimates generated per tag were examine via halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Western Aleutian (WA) and Central Aleutian (CA) one-way ANOVA with Tukey HSD post hoc multiple Islands, the Southeast Bering Sea continental Shelf Edge (SE) and Shelf Islands comparisons conducted where significance was detected. (SI), and a Transition Zone (TZ) between those regions and the western Gulf of Statistical tests were conducted using TIBCO Statistica Alaska Bold values indicate comparisons that were significantly different at p ≤ 0.05 version 13.5.0.17 (Palo Alto, California, USA) and errors reported in the text of this manuscript will represent one standard deviation unless otherwise specified. had been tagged at St. Matthew Island. Tags from 119 fish generated endpoint locations via satellite transmission: Results 76 of these reported within the specified annual-displace - Tag deployments ment window of 360–370 days at liberty; 43 reported A total of 145 Pacific halibut were tagged: 115 from 28 after shorter periods, ranging from 14 to 343 days. Fork May to 31 August 2008; 17 from 6 June to 21 July 2009; length was not found to be a significant determinant of and 13 from 11 to 26 June 2016 (Table 1; Fig. 2). Deploy- tag retention periods; i.e., smaller fish did not shed their ments occurred throughout the study area in 2008; but, tags at a significantly higher rate than larger fish [linear only at the EBS Shelf Islands in 2009 and only on the regression: df (1, 112), F = 0.028, p = 0.870, R = 0.0002]. northern EBS Shelf Edge in 2016. Tags were deployed in The proportion of binned data that were received via each tagging region roughly in proportion to the abun- satellite broadcast from tags that transmitted data sum- dance of Pacific halibut ≥ 82 cm FL and ranged from a maries was highly variable, ranging from 7.0 to 96.7%. minimum of 18 individuals tagged in the Central Aleu- Mean proportion of data received varied significantly tian Islands region to a maximum of 49 individuals [ANOVA: df (4, 112), F = 3.893, p = 0.005] by region, tagged along the EBS Shelf Edge (Table 1). Fish ranged as follows: Western Aleutian Islands = 59.6 ± 13.1%; from 82 to 171 cm FL. For fish whose tags subsequently Central Aleutian Islands = 61.5 ± 13.1%; Transi- produced at least one category of useable data (i.e., final tion Zone = 57.8 ± 19.6%; Eastern Bering Sea Shelf location, light, or depth data; n = 116), significant differ - Edge = 73.2 ± 24.3%; Shelf Islands = 77.9 ± 24.3%. Tukey ences in mean fork length were detected among tagging HSD post hoc comparisons indicated that reception rates regions [MANOVA: df (4, 62), F = 7.052, p < 0.001]. Mean were similar among the regions located along the Aleu- fish lengths were not significantly different within the tian Ridge and significantly higher, and similar to each three regions occurring along the Aleutian Ridge or on other, for the Eastern Bering Sea Shelf Edge and Shelf the EBS Shelf Edge; fish tagged at the EBS Shelf Islands Islands. were on average larger than in all other regions (Tables 1 None of the tags that were deployed in 2016 (i.e., PSAT and 2). Flex) were recaptured within their first year at liberty, successfully communicated via satellite, or produced Tag and data recoveries: via fishery and satellite data downloads. However, three tags were recovered in Twenty-one tags were neither physically recaptured longline fisheries on dates that were later than the tags nor communicated via satellite and were, therefore, had been programmed to detach and report, their batter- lost entirely. An additional three tags produced satellite ies having failed prematurely. These tags were recovered uplinks that were of insufficient strength to determine after periods at liberty of 701, 777, and 1155 days at lib- their final positions or upload any data. Of the remain - erty. Although no environmental data were obtained for ing 121 fish, two that had been tagged with Mk10s were these fish, final coordinates were obtained for two that recaptured in commercial fisheries prior to scheduled tag were recaptured in US waters. The third was recaptured reporting, after periods at liberty of 11 and 343 days; both in Russian waters in the Western Bering Sea (i.e., coastal Loher Animal Biotelemetry (2022) 10:18 Page 10 of 21 Kamchatka) and a precise location was not obtained. and Rat Islands group, bounded between Near Strait to These fish were recovered in late spring and summer the west and Amchitka Pass to the east; for the Central (i.e., late May, July, and August, respectively) and so their Aleutians, interannual displacement was confined to recoveries will be plotted; however, their final locations the Andreanof Islands between Amchitka and Amukta will not be included in calculations of regional dispersal, Passes (Fig. 2). From a fishery-management perspec - having not satisfied the criterion of occurring within a tive, this represented retention within IPHC Regulatory multiple of 360–370-day post-release. In addition to tags Area 4B (Fig. 1). Pacific halibut that were tagged in the recaptured in fisheries, seven tags (all Mk10s) were found EBS were more dispersive than observed in the Aleu- awash after having broadcast their data and producing tian Island regions. Those that departed the EBS Shelf final positions. These tags were downloaded to recover Islands dispersed either eastward on the EBS shelf (n = 1) their full scheduled-broadcast records. or northwestward into Russia waters (n = 1). Those that departed the EBS Shelf Edge displayed similar behav- Displacement and interannual dispersal ior, moving into the shallow waters of Bristol Bay (n = 2) Endpoint-derived mean interannual linear displacements and into Russian waters (n = 2) from just westward of (Table 3) were greatest for the Transition Zone; some- the Russian EEZ boundary to the Gulf of Anadyr. The what lower for the EBS Shelf Edge and Shelf Islands; and fish that was physically recaptured in coastal Kamchatka lowest in the Western and Central Aleutian Islands. How- after roughly 3 years at liberty brought the total number ever, variance in displacements was high in all regions of Pacific halibut that had dispersed to Russian waters to and none of the observed differences were statistically four, resulting in ~ 14% interannual emigration to Russia significant [MANOVA: df (4, 62), F = 1.677, p = 0.167]. for fish tagged in the EBS. Interannual dispersal was characterized by the following Within the Transition Zone, Pacific halibut that were general patterns (Fig. 2): (1) for Pacific halibut tagged in tagged north of Unimak Pass produced final endpoints the Western and Central Aleutian Islands, all final tag- that were exclusively within the Bering Sea (n = 6), within reporting locations were within the region in which the IPHC Regulatory Area 4A. One of these fish was recap - fish had been tagged; (2) Pacific halibut tagged in the tured in August of 2010, after 2 years at liberty (771 days), EBS moved among tagging regions and among IPHC to provide a third known location: it was captured in regulatory areas with final locations that were exclusively northern Area 4A, approximately 2 km from its initial within the Bering Sea, ranging from the Gulf of Anadyr, release location. In contrast, fish that were tagged south Russia, to Bristol Bay, Alaska; (3) Pacific halibut tagged of Unimak Pass (i.e., in the far-western GOA; southern in the Transition Zone displayed the greatest amount of Regulatory Area 4A) emigrated at a relatively high rate emigration from their tagging region, with migrants dis- (~ 55%), exclusively eastward, dispersing among IPHC persing across the GOA as far south as coastal Washing- Regulatory Areas: 4A (n = 5), 3B (n = 4), 3A (n = 1), and ton State. 2A (n = 1). One fish that emigrated to Regulatory Area 3B For Pacific halibut tagged in the Aleutian Islands, after 1 year at liberty was subsequently recaptured in its interannual displacement was confined to the Island third summer at liberty (1117 days) and had moved far- group that defined each tagging region: i.e., for the West - ther eastward into Regulatory Area 3A. ern Aleutians, fish remained resident within the Near At‑liberty dispersal inferred from light data A total of 109 tags generated twilight data that allowed Table 3 Dispersal results for the estimation of daily local noon (Table 1), producing 5313 daily longitudes estimates, while those fish were at Tagging region Mean Light‑lon range Light‑lon N liberty (Table 3). The number of at-liberty position esti - displ. ± SD (mean ± SD) (km) mates per fish ranged from 1 to 152 and averaged from 41 to 66, depending upon region (Table 3). However, no Western Aleutian 79 ± 37 3–97 (41 ± 7) 819 significant differences were detected in the average num - Central Aleutian 39 ± 68 15–81 (46 ± 8) 742 ber of longitude estimates per fish among regions [one- Transition Zone 313 ± 70 1–113 (51 ± 7) 1175 factor ANOVA: df (4, 104), F = 1.541, p = 0.196]. EBS Shelf Edge 189 ± 19 1–152 (45–6) 1417 For the Western and Central Aleutian Islands, inspec- EBS Shelf Islands 100 ± 211 1–141 (66 ± 8) 1160 tion of individual light-based longitude trajectories (36 Mean interannual linear displacement, range in and mean number of light- of 40 tags; Table 1) did not yield sufficient evidence to based daily longitude estimates obtained for individual fish, and total number conclude that any individual fish had departed its tag - of longitude estimates produced for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the ging region during its time at liberty. That is, despite Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those those individuals having generated numerous light-based ecosystems and the western Gulf of Alaska L oher Animal Biotelemetry (2022) 10:18 Page 11 of 21 longitudes to both the east and west of their regions’ late September and early December, spent 4–5 months in boundaries, no individual generated five consecutive IPHC Regulatory Areas 3B and 3A, and returned to Reg- estimates that would have plausibly placed that individ- ulatory Area 4A in late April. Their longitudinal migra - ual outside of its tagging region. Similarly, out-of-region tions were coincident with movement to deep water in movement could not be identified for any fish tagged on autumn and return to the depths at which they had been the EBS Shelf Edge or at the Shelf Islands, including for tagged in spring (Fig. 4a, b, lower panels). Light-based individuals whose final positions unequivocally demon - longitudes also allowed for the timing of regional depar- strated that they had departed the region. In contrast, ture to be identified for two migratory individuals. The seasonal departure followed by return to the tagging first of these fish (Fig. 3b) departed the Transition Zone region was evident for two fish tagged in the Transition at the end of December. The second (Fig. 3c) produced Zone south of Unimak Pass (Fig. 4a, b, upper panels). considerably fewer position estimates, but the available These individuals departed the Transition Zone between data suggest November departure. As with the seasonally Fig. 4 Light-based latitude estimates (upper panels) and daily maximum depth profiles (lower panels) during 2008 and 2009 for: a, b two Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the far-western Gulf of Alaska, and; c one Pacific halibut tagged at the Pribilof Islands in the eastern Bering Sea, from date of tagging (top of upper panels) to date of final tag reporting (bottom of upper panels). All data were obtained via satellite transmission (i.e., the tags were not recovered). The approximate longitudinal span of International Pacific Halibut Commission regulatory areas is indicated by shading. All three fish moved to deep water during the winter and returned to shallow water in spring. Eastward seasonal emigration is clearly evident in a and b. The fish depicted in c is known to have emigrated from its tagging area on the basis of its mid-winter depths (i.e., no such depths occur at the Pribilof Islands), but neither the timing or magnitude of its emigration can be determined. Longitude estimates are absent during its deep-water phase; the remainder of the profile is characterized by apparent measurement error of 5–7° relative to the fish’s known endpoint positions Loher Animal Biotelemetry (2022) 10:18 Page 12 of 21 migratory individuals, the longitudinal migrations of eastward redistribution was apparent after October. As these fish were associated with movement to deep water with the individual trajectories, pooled unstandardized during winter. longitudinal data for Pacific halibut tagged in the EBS For Pacific halibut that were tagged in the EBS (on both regions were relatively uninformative with respect to sea- the Shelf Edge and at the Shelf Islands), light-based longi- sonal and interannual redistribution. tude estimates provided little information regarding their In all regions other than the Transition Zone, pooled movements. Many of their interannual displacements standardized at-liberty longitude estimates were were relatively short distance and, for fish whose migra - described by a single frequency mode (Fig. 6) that was tions were considerable, their movements were executed centered within one degree of longitude relative to the to a considerable degree along north–south vectors fishes’ tagging locations: displacement in the mean (Fig. 2). Such movements are not well-resolved by longi- and median was small (< 0.25°) in the Western Aleu- tudinal position estimates. tians and on the EBS Shelf Edge; and somewhat larger Pooled among all individuals tagged within each region, (0.75–0.91°) for the Shelf Islands and Central Aleutians unstandardized light-based at-liberty longitude estimates (Table 4). These frequency distributions displayed posi - for Pacific halibut tagged along the Aleutian Ridge (Fig. 5) tive (eastward) skew, which was relatively minor in the displayed some clear seasonal trends. A decrease during EBS (< 0.6) and moderate in the Aleutian Island regions winter in the frequency with which estimates were gener- (~ 1.4). Few longitude estimates occurred beyond ~ 5° of ated was apparent in all regions, and most evident in the the median and those observations were symmetrically Western Aleutian Islands (Fig. 5a). This was coincident distributed, consistent with the occurrence and mag- with movement to deeper water (see next section), with nitude of the apparent random measurement error that fish reaching depths at which ambient light levels are was evident in many of the individual longitude trajecto- expected to be below the tags’ detection thresholds. For ries (e.g., Fig. 4c). In contrast, fish tagged in the Transi - fish tagged in the Transition Zone (Fig. 5c), progressive tion Zone produced a distinctly bimodal distribution of Fig. 5 Light-based longitude estimates during periods at liberty for all Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags in the a Western and b Central Aleutian Islands, and c Transition Zone between the eastern Aleutians, Southeastern Bering Sea, and western Gulf of Alaska, from dates of tagging (top of panels) through final tag reporting (bottom of panels). Dashed lines indicate ocean passes separating regions located along the Aleutian Ridge and the approximate extent of International Pacific Halibut Commission Regulatory areas in the Gulf of Alaska. Note that the Western Aleutians and the Komandorskiye Ostrova are located in the eastern hemisphere, with the Antimeridian bisecting IPHC Regulatory Area 4B at approximately Amchitka Pass L oher Animal Biotelemetry (2022) 10:18 Page 13 of 21 Fig. 6 Frequency distributions of light-based at-liberty location (longitude) estimates obtained for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those ecosystems and the western Gulf of Alaska. Values have been standardized to reflect degrees westward (negative values) or eastward (positive values) of each fish’s known initial tagging location. The vertical dashed line in each panel indicates a value of 0, such that distributions not centered at that location indicate substantial residency of the tagged halibut at some distance away from their initial longitudes Loher Animal Biotelemetry (2022) 10:18 Page 14 of 21 Table 4 Characteristics of frequency distributions of in some regions, the tagged population’s deep-water standardized geolocations phase extended beyond that date. Mean tagging (fish-release) depth varied significantly Tagging region Mean ± se Median ± se Skewness ± se among regions [MANOVA: df (4, 62), F = 2.64, p < 0.042]. Western Aleutian 0.14 ± 0.11 0.23 ± 0.11 1.37 ± 0.09 This was expected, given that the regions in the EBS Central Aleutian − 0.87 ± 0.14 − 0.91 ± 0.14 1.38 ± 0.09 were chosen to represent depth-specific habitat. As such, TZ Mode 1 − 1.22 ± 0.09 − 1.23 ± 0.09 − 0.38 ± 0.08 mean tagging depth (Table 5) was significantly (Tukey TZ Mode 2 13.0 ± 0.23 13.1 ± 0.24 0.08 ± 0.25 HSD, p = 0.021) greater for fish released along the EBS EBS Shelf Edge 0.17 ± 0.10 0.15 ± 0.10 0.06 ± 0.07 Shelf Edge than at the Shelf Islands; no other pairwise EBS Shelf Islands − 0.61 ± 0.10 − 0.75 ± 0.10 0.55 ± 0.07 comparisons were significant (Table 2). Characteristics of frequency distributions (see Fig. 6) of light-based at-liberty Mean depths during the tagged individuals’ summer- location (longitude) estimates, standardized relative to initial locations, of Pacific time shallow-water phases (Table 5; Fig. 7) varied sig- halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags nificantly among regions. Mean shallow-water phase during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone (TZ) between those ecosystems and the western Gulf of Alaska depths for fish tagged on the EBS Shelf Edge were signifi - cantly greater than for fish in all other regions (all com - parisons: Tukey HSD, p < 0.001) and shallow-water phase standardized longitude estimates composed of a primary depths for fish tagged in the Western Aleutian Islands mode positioned ~ 1.2° westward of the fishes’ tagging were greater than at the EBS Shelf Islands (Tukey HSD, locations and a smaller secondary mode centered ~ 13° p = 0.038). No other pairwise comparisons were signifi - eastward, in the Gulf of Alaska. The secondary mode was cant at p ≤ 0.05. For the most part, relative differences consistent with winter locations of fish that were deter - in depth distribution persisted throughout the fishes’ mined to be seasonally migratory (Fig. 4). movements to deeper water during the winter (Table 5; Fig. 7), with one exception: Pacific halibut tagged at the Seasonal depth distributions EBS Shelf Islands exhibited the deepest mean deep- The timing of tagging and, therefore, the precise span water phase (563 ± 156 m), while in no other region did over which depth data were available varied by tagging the mean deep-water phase exceed 500 m. Fish tagged in region. For the purposes of plotting and analysis, the the Transition Zone displayed the shallowest deep-water “year” over which depth data were included was defined phase, averaging 369 ± 69 m. Differences were signifi - as the 365-day period over which the greatest amount cant between the Shelf Islands and the Central Aleutian of data was available among all regions: from 19 June of Islands (Tukey HSD, p = 0.028), between the Shelf Islands the tagging year through 18 June of the subsequent year. and the Transition Zone (Tukey HSD, p = 0.002), and This provided contiguous data for all regions except the between the EBS Shelf Edge and the Transition Zone Western Aleutian Islands. Fish were tagged earlier in the (Tukey HSD, p = 0.034). Western Aleutians than in other regions, resulting in data In addition to the observed differences in phase- that terminated approximately 2 weeks earlier than else- specific mean depths (Table 5), the timing of offshore– where: i.e., on 3 June of the year after tagging. The analy - onshore redistribution, length of deep-water residency, ses presented subsequently demonstrate that terminating and general form of the mean maximum daily depth all regions’ depth profiles on 3 June to maintain a unified (MMDD) trajectories differed among tagging regions. period among all regions would have been inappropriate: Table 5 Depth results Tagging region Mean depths (meters) Deep‑ water phase Tagging (± SD) Shallow (± SD) Deep (± SD) Arrive Depart Days Western Aleutian 144 ± 90 158 ± 40 453 ± 65 09 Nov 07 May 180 Central Aleutian 133 ± 103 139 ± 29 411 ± 85 12 Nov 20 May 190 Transition Zone 123 ± 62 141 ± 49 369 ± 69 25 Nov 04 May 161 EBS Shelf Edge 187 ± 80 214 ± 72 468 ± 93 12 Dec 04 Jun 175 EBS Shelf Islands 74 ± 35 107 ± 19 534 ± 156 27 Dec 11 Jun 194 Mean depths at which fish were captured and released (tagging), during the fishes’ summer shallow-water and winter deep-water phases, the mean dates of arrival and departure from and resulting population-level average duration occupying deep-water habitat for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea (EBS), and a Transition Zone between those ecosystems and the western Gulf of Alaska L oher Animal Biotelemetry (2022) 10:18 Page 15 of 21 Fig. 7 Three-day moving averages of maximum daily depth for Pacific halibut (Hippoglossus stenolepis) tagged with pop-up archival transmitting tags during 2008 and 2009 in the Aleutian Islands, Eastern Bering Sea, and a Transition Zone between those ecosystems and the western Gulf of Alaska. Note that these profiles include all individuals for which depth data were generated, even if they did not conduct an offshore migration (e.g., see Fig. 3a, lower panel) and, therefore, tend to result in group-level mid-winter depths that are shallower than the group’s average deep-water phase depth (i.e., Table 5) Pacific halibut tagged in both Aleutian regions behaved form of the MMDD trajectories was quite similar for all similarly, arriving on average in relatively deep water in three regions (Fig. 7a). In contrast, fish tagged in the EBS mid-November and returning to shallower water in May. arrived at deep-water grounds approximately 1 month Deep-water phase duration and depth were shorter and later than along the Aleutian Ridge and did not return shallower, respectively, in the Transition Zone than far- to shallow water until June (Table 5; Fig. 7). Unlike in ther westward in the Aleutians (Table 5); however, the all other regions, Pacific halibut tagged at the EBS Shelf Loher Animal Biotelemetry (2022) 10:18 Page 16 of 21 Table 6 Mean temperatures (degrees Celsius) experienced by region, Pacific halibut tagged in the Central and Western Pacific halibut (Hippoglossus stenolepis) tagged with pop-up Aleutian Islands remained within their island groups (i.e., archival transmitting tags during 2008 and 2009 in the Aleutian within the Andreanof and Near-Rat Islands, respectively), Islands, Eastern Bering Sea (EBS), and a Transition Zone between while those tagged on and along the EBS continental those ecosystems and the western Gulf of Alaska, that undertook shelf moved among Bering Sea regulatory areas and from seasonal offshore migrations during the autumn of the year in Alaskan to Russian waters. These dispersal patterns are which they were tagged consistent with summer-to-winter PAT-tagging studies Tagging Midsummer D‑4 D‑3 D‑2 D‑1 that have indicated basin-scale reproductive segrega- region tion with considerable mixing within those basins [4]; research that has shown Samalga Pass to be an oceano- Western 4.8 ± 0.6 4.5 ± 0.5 4.5 ± 0.5 4.5 ± 0.5 4.5 ± 0.5 Aleutian graphic [39] and ecological [40–42] boundary within the Central Aleu- 5.3 ± 0.4 5.5 ± 0.8 5.2 ± 0.7 5.2 ± 0.7 5.0 ± 0.6 Aleutian Island ecosystem; and population-genetic analy- tian ses that suggest relative isolation of Pacific halibut in the Transition Zone 5.2 ± 0.6 5.4 ± 0.5 5.2 ± 0.7 5.1 ± 0.6 5.0 ± 0.6 Aleutian Islands westward of Amchitka Pass [7]. In addi- EBS Shelf Edge 2.9 ± 0.4 3.0 ± 0.6 3.0 ± 0.6 3.0 ± 0.6 3.0 ± 0.6 tion, light-based geolocation demonstrates that connec- EBS Shelf 3.3 ± 0.9 2.6 ± 1.3 2.6 ± 1.3 2.7 ± 1.2 2.8 ± 1.1 tivity among regulatory areas may be seasonally cyclic, Islands with fish emigrating from their tagging region for the “Midsummer” was defined as 28 July through 7 September, when fish were winter and returning the following spring. This was evi - resident on shallow grounds. “D-1” represents mean temperature during the week prior to the fishes’ departure to deep water; D-2 is the week prior to D-1, dent at both the individual and population level for fish and so forth. The superscripts after the midsummer means indicate two groups tagged south of Unimak Pass. determined to be significantly different (p < 0.05) from one another. In no region In contrast to the large degree of out-of-area movement were mean temperatures experienced during the 4 weeks prior to departure found to be significantly different from the region’s midsummer temperature demonstrated by Pacific halibut tagged south of Unimak Pass, there was insufficient evidence to suggest move - ment of tagged fish out of either the Central or Western Islands displayed a distinctly bi-phasic offshore migra - Aleutian Islands, at either interannual or seasonal scales. tion, initially moving from December until late February Although numerous tags produced light-based longitude to depths of approximately 300 m, then descending to estimates that might have placed them outside of these > 400 m during March and April (Fig. 7b). regions, inspection of the individual trajectories failed to Mean temperatures experienced during midsummer reveal series of consecutive estimates indicative of east– by seasonally migratory Pacific halibut (Table 6) varied west redistribution beyond the regions’ boundaries, and significantly [ANOVA: df (4, 51), F = 35.264, p < 0.001] the distribution of pooled, standardized values in both according to tagging region. Tukey HSD post hoc com- Aleutian regions showed little skew and no secondary parisons indicated that the temperatures experienced peaks consistent with cross-pass movement. Rather, the were similar among regions located along the Aleutian nature of the longitude estimates to the east and west of Ridge and significantly lower for fish tagged on the East - the passes that define the Aleutian tagging regions was ern Bering Sea Shelf Edge and at the Shelf Islands. In no most consistent with estimation error induced by local region was a significant change in the experienced tem - environmental conditions biasing the perception of local peratures detected between midsummer and any of the noon relative to its true value [35]. four 1-week periods prior to the initiation of departure to Although light-based geolocation failed to resolve sea- deeper water in autumn. sonal movement among areas in the EBS—perhaps due to the ability of individuals to migrate from summer feeding sites to deeper-water spawning grounds without Discussion moving considerably along east–west axes—the end- The current study builds upon prior analyses of relative point data clearly demonstrated interannual migration spawning segregation [4] and genetic population struc- from the USA into Russian waters (n = 4). Ultimately, ture [6, 7] to enhance our understanding of connectiv- movement between North American and Asian waters ity within the Pacific halibut stock of the eastern Bering on the northern Bering Sea continental shelf is not sur- Sea and Aleutian Islands. The Pacific halibut that were prising: there are no known oceanographic or geologi- tagged herein exhibited basin-specific dispersal in which cal features that might impede such connectivity. Rather, fish tagged in the Bering Sea had summer distributions larval transport modelling predicts that some proportion that were within the Bering Sea, while fish tagged south of larvae spawned in the EBS are likely to be delivered of Unimak Pass were more dispersive and either occu- to Russian coastal habitat [43], and our understanding pied or transited all IPHC regulatory areas in the Gulf of ontogenic [9, 10] and seasonal [4, 5, 44] cross-basin of Alaska and US Pacific Northwest. Within the BSAI L oher Animal Biotelemetry (2022) 10:18 Page 17 of 21 movements in the GOA should lead to an expectation of With respect to seasonal migration and depth-specific considerable cross-basin mixing in the Bering Sea, par- habitat usage, regional differences were apparent. In the ticularly at young ages. Both systems are characterized EBS, Shelf Island fish moved to deeper winter spawn - by spawning that is concentrated in submarine canyons ing habitat than fish tagged along the Shelf Edge, despite along their eastern and central margins, and westward- having summered in relatively shallower water that was flowing shelf-edge currents [45, 46]. Basin-scale con- considerably farther from their shelf-edge winter destina- nectivity over the course of Pacific halibut life history is tions. In addition, the annual depth trajectories of Shelf likely to display approximately the same spatial structure Island fish were on average considerably more biphasic in both systems. Within the Bering Sea, individuals that (i.e., dual-stage) than observed on the Shelf Edge. This are derived from EBS spawning and which settle in nurs- dual-stage movement pattern is common in individu- eries along the Asian coast should be expected to return als and, in cases in which the data allow for a detailed to the North American spawning stock if the population evaluation of active spawning (i.e., putative egg release) is to maintain long-term stationarity. It is perhaps only [12, 36, 38], appears to represent an initial period of pre- surprising that the current study detected as much inter- spawn staging that is followed by active spawning at the annual migration to Russian waters as it did, given that deeper stratum. The depths associated with each stratum those movements are counter to the expected direction vary according to individual and examples of dual-stage of mean population-level dispersal of benthic-stage indi- movement occurred in all tagging regions in the current viduals: i.e., from west to east. This would suggest that study. Ultimately, the dual-stage nature of the group-level overall migration rates across the Russia–USA maritime MMDD profile for Shelf Island fish derives from these border are non-trivial, which is an important observation fish having exhibited more-synchronous movements from the perspective that stock assessment modelling, than were observed elsewhere. In other regions, indi- policy analyses, and management decisions have tradi- vidual dual-stage migrations were obscured by averag- tionally assumed that Pacific halibut in waters of Canada ing among individuals that did so with variable timing. In and the USA exist in a closed system in which there is addition, differences in sex ratio among tagged individu - no exchange with population(s) in the western half of als within each region could obscure this pattern, if one the species’ geographic range. The omission of this con - sex more consistently undertakes dual-stage migration nectivity when characterizing the function of EBS Pacific than the other. We are unable to address that hypothesis halibut stocks to-date [e.g., 3, 4, 9] is likely due to an here, because reliable techniques for evaluating sex with- absence of data regarding its magnitude and dynamics. out sacrificing the individuals had not yet been devel - Future work should seek to generate migration-rate data oped and sex was, therefore, unknown. However, taken along this axis. together, the displacement, at-liberty longitude estimates, The observation that out-of-area dispersal was highest and depth profiles suggest that the Pacific halibut tagged for Pacific halibut that were tagged in the far-Western along the Shelf Edge and at the Shelf Islands likely occu- GOA is consistent with prior PIT-tagging research [9] in pied the same regional slope spawning grounds (i.e., which approximately 90% of the fish that emigrated east - in Pribilof, Pervenets, and Zhemchug Canyons; Fig. 1) ward from IPHC Regulatory Area 4A into GOA regula- [12], but with somewhat different timing and depth tory areas were individuals that had been tagged south of preferences. the Aleutian Ridge. The spatial discontinuity in connec - Pacific halibut recruitment is believed to be environ - tivity within this region (i.e., very different mixing pat - mentally driven, via favorable plankton productivity terns on opposite sides of the Aleutian Ridge) highlights or larval transport [47] and the maintenance of broad the challenges associated with defining management spawning and migratory periods in marine fish popula - subregions that are intended to represent stock structure tions likely represents ecological bet-hedging [48] that and function [3] when constrained to the boundaries of ensures long-term recruitment success and stock pro- historical management units that may not have been ide- ductivity. Evidence of migratory contingency has been ally crafted to do so. For assessment and policy purposes, observed in Pacific halibut in the Gulf of Alaska [14] data obtained from Regulatory Area 4A are assigned to and US Pacific Northwest [16], and analyses suggest that Bioregion 4 [3], which largely describes stock status and Pacific halibut in the EBS have exhibited shifts in their dynamics of the eastern Bering Sea. However, the current distribution and habitat use over the last three decades results demonstrate that dispersal from IPHC Regulatory in response to changing ocean temperatures [49, 50]. Area 4A is more complex: 4A North represents the EBS, Maintaining a diversity of life-history strategies may be while 4A South represents the western GOA and the critical to the successful management of exploited stocks. individuals tagged within it displayed considerable move- Parameterizing the timing and duration of larval release, ment across the GOA and into the US Pacific Northwest. and the depths (i.e., current regimes) into which larvae Loher Animal Biotelemetry (2022) 10:18 Page 18 of 21 are released, is required for the construction of larval of onshore–offshore migration. For example, an early transport models and for evaluating relative recruitment trawl survey might sample inshore waters prior to the potential among population components. arrival of seasonal migrants and the setline survey sub- Pacific halibut tagged along the Aleutian Ridge, in both sequently occupy the slope after fish have departed, such the Aleutian Islands and the far-western GOA, were that neither survey fully indexes the population. A bet- observed to move to deep water and return to shallow ter understanding of the timing of these movement and habitat roughly 1 month earlier than fish tagged in the their interannual variability would be required to under- EBS and there was evidence of a gradual advancement stand to what extent such mismatch is likely in any given in the timing of deep-water occupancy moving anticy- year, but the ultimate result may be time-varying selec- clonically from the Western Aleutians to the EBS Shelf. tivity within each survey that should be accounted for in Although the available data did not provide any evidence the assessment models. The IPHC stock assessment does of a threshold temperature, or change in local tempera- have a structure that can account for time-varying selec- ture, that might initiate the autumn offshore migra - tivities [53] and the effects of seasonal migration on fish - tion, the observation that fish residing in cooler waters ery selectivity [54] and the incorporation of time-varying departed later in the year is consistent with findings from functions in assessment models [e.g., 55, 56] has received research conducted both to the south and north of the attention in the literature. current study. Pacific halibut tagged in the eastern Gulf of Inspection of Pacific halibut depth profiles also high - Alaska have been shown to initiate offshore migrations as lights the difficulty of evaluating at-liberty movements early as September and largely arrive at their wintertime for any given individual in the absence of highly resolved depths by the end of October [12]. Emerging data sug- location data. For example, fish tagged along the EBS gest that Pacific halibut that summer in Norton Sound Shelf Edge were found at shallower average depths in July (Fig. 2), in the northeastern Bering Sea, may move to of 2009 than they had occupied when they were tagged. deep water as late as March and spawn in April and May This suggests that their preferred summer habitat was (A Flanigan, University of Alaska Fairbanks, USA, per- likely shallower in 2009. However, from the available data sonal communication). Similarly, latitudinal gradients in is it difficult to identify the precise nature of the habitat spawn timing have been reported for North Pacific starry shift nor identify its driver(s). In particular, light-based flounder (Platichthys stellatus) and rex sole (Glyptoceph - longitudinal estimates provided little information regard- alus zachirus). Spawning in these species occurs approxi- ing their movements (e.g., relative to fish located along mately 4 months earlier off the coast of California than in the Aleutian Ridge), because large-scale redistribution the southeast Bering Sea [51]. may occur on the EBS continental shelf along north– In addition to its ramifications on interannual recruit - south axes. The associated changes in stock distribution ment potential and vulnerability to seasonal fisheries may be quantifiable through refined methodology, such [12], regional variance in seasonal migration timing can as advanced mobile acoustics [57, 58], ongoing develop- affect our understanding of the distribution and demo - ments in geomagnetic-sensing electronic tags [59, 60], graphic structure of populations via their interactions and the adaptation of statistical models for tracking ben- with survey design. For example, the depth trajectories in thic and epibenthic marine species [24–26]. In addition, the EBS remained upward-sloping throughout June and the use of mark-report PAT tags, which generate pop-up into July of 2009 (i.e., the year after tagging), indicating locations in the absence of archived environmental data that these fish were still in the process of returning to and, therefore, cost considerably less than standard PAT their summer habitat when the assessment surveys com- tags, could be considered. Large-bodied species such as menced. In the EBS, two platforms are used to index the Pacific halibut can be tagged with multiple such tags that abundance of Pacific halibut: the IPHC FISS that inten - are programmed to release at any desired interval, pro- sively surveys the continental shelf edge and US National viding more-detailed migration histories than can any Marine Fisheries Service trawl surveys [52] within shal- single pop-up tag [sensu 61] as well as providing vali- lower continental shelf habitat. Because catchability var- dated locations to increase the accuracy of tracking mod- ies between the two survey techniques and they can be els [62]. conducted with slightly different timing in any given year, Ultimately, to fully understand population ecology in our perception of fine-scale distribution and habitat use Pacific halibut we will need to fill the scaling-gaps that exist may be affected. Differences in relative catchability can be within the existing data: in particular, a lack of informa- interpreted as differences in underlying abundance even tion describing generational-scale processes and informa- in a homogeneous population. Relative survey timing can tion on sex-specific behavior. With respect to the latter, either oversample or undersample the population if the ultrasonic [63] and genetic techniques [64] have now been timing of the surveys does not correspond with timing developed that allow for determination of the sex of Pacific L oher Animal Biotelemetry (2022) 10:18 Page 19 of 21 Consent for publication halibut during tagging and can be employed in future work. The author has given his consent to the manuscript being published. With respect to the former, the interannual time scales addressed herein are still far from generational in this Competing interests The author declares that he has no competing interests. species. Although it is rare for satellite-tagging studies to exceed 1 year due to hydraulic drag effects [65], tag shed - Received: 21 February 2022 Accepted: 20 April 2022 ding [66], and biofouling [67], the general lack of fouling and wear on PAT tags that have been physically recovered on Pacific halibut, after 2–3 years at liberty in the current study, suggests that PAT-tagging can likely be conducted References 1. Carpi P, Loher T, Sadorus LL, Forsberg JE, Webster RA, Planas J, Jasonowicz at multi-year time scales for this, and similar, species. In A, Stewart IJ, Hicks AC. Ontogenetic and spawning migration of Pacific addition, the tags used in the current study (i.e., Mark 10 halibut: a review. Rev Fish Biol Fish. 2021;31:879–908. https:// doi. org/ 10. PAT tags) were considerably larger than current-generation 1007/ s11160- 021- 09672-w. 2. Kong T, Tran H, Prem C. Fisheries data overview (2021). Meeting Docu- satellite tags and tended to suffer higher failure rates than ment IPHC-2022-AM098–06 Rev_1. Seattle: International Pacific Halibut more-recent studies on Pacific [e.g., 68] and Atlantic [38, Commission. 2022. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- 69] halibut. Furthermore, the IPHC has developed both 2022- am098- 06. pdf. Accessed 20 Feb 2022. 3. Stewart I, Hicks A, Webster R, Wilson D. Summary of the data, stock surgical [70] and external [71] tagging techniques resulting assessment, and harvest decision table for Pacific halibut (Hippoglossus in in situ retention periods in excess of 6 years and elec- stenolepis) at the end of 2021. Meeting document IPHC-2022-AM098-10. tronic archival tag manufacturers currently provide multi- Seattle: International Pacific Halibut Commission. 2021. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- 2022- am098- 10. pdf. Accessed 20 Feb sensor tags with operational life and logging capacities on the order of 7–9 years [e.g., Star Oddi (Reykjavik, Iceland) 4. Seitz AC, Loher T, Farrugia TJ, Norcross BL, Nielsen JL. Basin-scale DST-centi (www. star- oddi. com/ produ cts/ data- logge rs/ reproductive segregation of Pacific halibut (Hippoglossus stenolepis). Fish Manag Ecol. 2017;24:339–46. https:// doi. org/ 10. 1111/ fme. 12233. minia ture- depth- logger); Lotek Wireless (St. Johns, Can- 5. Loher T, Seitz AC. Seasonal migration and environmental conditions ada) LAT-2000 Series (https:// www. lotek. com/ produ cts/ experienced by Pacific halibut (Hippoglossus stenolepis), elucidated from lat20 00- series/) archival tags]. Ultimately, there is consider- pop-up archival transmitting tags. Mar Ecol Prog Ser. 2006;317:259–71. https:// doi. org/ 10. 3354/ meps3 17259. able potential to continue expanding the scope and scale of 6. Nielsen JL, Graziano SL, Seitz AC. Fine-scale population genetic structure research on individual behavior and population-level con- in Alaskan Pacific halibut (Hippoglossus stenolepis). Conserv Genet. nectivity using advanced electronic-tagging technology 2010;11(3):999–1012. https:// doi. org/ 10. 1007/ s10592- 009- 9943-8. 7. Drinan DP, Galindo HM, Loher T, Hauser L. Subtle genetic population and a variety of multidisciplinary approaches. structure in Pacific halibut. J Fish Biol. 2016;89:2571–94. https:// doi. org/ 10. 1111/ jfb. 13148. Acknowledgements 8. IPHC (International Pacific Halibut Commission). The Pacific halibut: biol- Kathy Bareza, Dan Rafla, Tucker Soltau Andy Vatter and Tom Wilson conducted ogy, fishery, management. Technical report no. 59. Seattle: International the tagging described in this study, aboard the fishing vessels Free to Wander, Pacific Halibut Commission. https:// www. iphc. int/ uploa ds/ pdf/ tr/ IPHC- Kema Sue, Pacific Sun, and Saint Peter. Erica Anderson-Chau, Claude Dykstra, 2014- TR059. pdf. Accessed 20 Feb 2022. Tracee Geernaert, Ed Henry, Eric Soderlund, and Evangeline White provided 9. Webster RA, Clark WG, Leaman BM, Forsberg JE. Pacific halibut on the invaluable logistical assistance through the IPHC Fishery-Independent Setline move: a renewed understanding of adult migration from a coastwide Survey. Ian Stewart, David Wilson, Josep Planas, and two anonymous reviewers tagging study. Can J Fish Aquat Sci. 2013;70:642–53. https:// doi. org/ 10. provided feedback and suggestions that improved its quality. 1139/ cjfas- 2012- 1371. 10. Hilborn R, Skalski J, Anganuzzi A, Hoffman, A. Movements of juvenile Author contributions halibut in IPHC regulatory areas 2 and 3. Technical report no. 31. Seattle: This study was conceived and planned by TL, who supervised all tag deploy- International Pacific Halibut Commission. 1995. https:// www. iphc. int/ ments, conducted all analyses, and wrote this manuscript. The author read uploa ds/ pdf/ tr/ IPHC- 1995- TR031. pdf. Accessed 20 Feb 2022. and approved the final manuscript. 11. Kaimmer SM. Pacific halibut tag release programs and tag release and recovery data, 1925 through 1998. IPHC technical report no. 41. Seattle: Funding International Pacific Halibut Commission. 2000. https:// www. iphc. int/ The sole Funding body for this work was the International Pacific Halibut uploa ds/ pdf/ tr/ IPHC- 2000- TR041. pdf. Accessed 20 Feb 2022. Commission (IPHC), and all tags were deployed from IPHC-chartered stock 12. Loher T. Analysis of match-mismatch between commercial fishing assessment survey vessels. periods and spawning ecology of Pacific halibut (Hippoglossus stenolepis), based on winter surveys and behavioural data from electronic tags. ICES J Availability of data and materials Mar Sci. 2011;68:2240–51. https:// doi. org/ 10. 1093/ icesj ms/ fsr152. The data sets generated and/or analyzed during this study are Public Domain, 13. Seitz AC, Loher T, Norcross BL, Nielsen JL. Dispersal and behavior of Pacific as per the International Pacific Halibut Commission’s (IPHC) current Data Use halibut Hippoglossus stenolepis in the Bering Sea and Aleutian Islands and Confidentially Policy (dated 2 August 2019), and may be obtained from region. Aquat Biol. 2011;12:225–39. https:// doi. org/ 10. 3354/ ab003 33. the IPHC upon request. 14. Nielsen JK, Seitz AC. Interannual site fidelity of Pacific halibut: potential utility of protected areas for management of a migratory demersal fish. Declarations ICES J Mar Sci. 2017;74(8):2120–34. https:// doi. org/ 10. 1093/ icesj ms/ fsx040. Ethics approval and consent to participate 15. Loher T, Blood CA. Seasonal dispersion of Pacific halibut (Hippoglos - All tagging was conducted under the auspices of the International Pacific sus stenolepis) summering off British Columbia and the US Pacific Halibut Commission. Northwest evaluated via satellite archival tagging. Can J Fish Aquat Sci. 2009;66:1409–22. https:// doi. org/ 10. 1139/ F09- 093. Loher Animal Biotelemetry (2022) 10:18 Page 20 of 21 16. Loher T, Soderlund E. Connectivity between Pacific halibut Hippoglossus 36. Seitz AC, Norcross BL, Wilson D, Nielsen JL. Identifying spawning behavior stenolepis residing in the Salish Sea and the offshore population, demon- in Pacific halibut, Hippoglossus stenolepis, using electronic tags. Environ strated by pop-up archival tagging. J Sea Res. 2018;142:13–124. https:// Biol Fish. 2005;73:445–51. https:// doi. org/ 10. 1007/ s10641- 005- 3216-2. doi. org/ 10. 1016/ jsear es. 2018. 09. 007. 37. Loher T, Seitz AC. Characterization of active spawning season and depth 17. Loher T. Homing and summer feeding site fidelity of Pacific halibut for eastern Pacific halibut (Hippoglossus stenolepis), and evidence of prob - (Hippoglossus stenolepis) in the Gulf of Alaska, established using satellite- able skipped spawning. J Northwest Atl Fish Sci. 2008;41:23–36. https:// transmitting archival tags. Fish Res. 2008;92:63–9. https:// doi. org/ 10. doi. org/ 10. 2960/j. v41. m617. 1016/j. fishr es. 2007. 12. 013. 38. Fisher JAD, Robert D, Le Bris A, Loher T. Pop-up satellite archival tag (PSAT ) 18. Rose GA, Leggett WC. Eec ff ts of biomass-range interactions on the temporal data resolution affects interpretations of spawning behavior catchability of migratory demersal fish by mobile fisheries: an example of and vertical habitat use. Anim Biotelem. 2017;2017(5):27. https:// doi. org/ Atlantic cod (Gadus morhua). Can J Fish Aquat Sci. 1991;48:843–8.10. 1186/ s40317- 017- 0137-8. 19. Rose GA. Cod spawning on a migration highway in the north-west Atlan- 39. Ladd C, Hunt GL Jr, Mordy CW, Salo SA, Stabeno PJ. Marine environ- tic. Nature. 1993;366:458–61. https:// doi. org/ 10. 1038/ 36645 8a0. ment of the eastern and central Aleutian Islands. Fish Oceanogr. 20. Gordoa A, Lesch H, Rodergas S. Bycatch: complementary information for 2005;14(s1):22–38. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00373.x. understanding fish behaviour. Namibian Cape hake (M. capensis and M. 40. Coyle KO. Zooplankton distribution, abundance and biomass relative to paradoxus) as a case study. ICES J Mar Sci. 2006;63:1513–9. https:// doi. water masses in eastern and central Aleutian Island passes. Fish Ocean- org/ 10. 1016/j. icesj ms. 2006. 05. 007. ogr. 2005;14:77–92. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00367.x. 21. Hobday AJ, Hartog JR, Timmiss T, Fielding J. Dynamic spatial zoning to 41. Jahncke J, Coyle KO, Hunt GJ. Seabird distribution, abundance and diets manage southern bluefin tuna (Thunnus maccoyii) capture in a multi- in the central and eastern Aleutian Islands. Fish Oceanogr. 2005;14:160– species longline fishery. Fish Oceanogr. 2010;19:243–53. https:// doi. org/ 77. https:// doi. org/ 10. 1111/j. 1365- 2419. 2005. 00372.x. 10. 1111/j. 1365- 2419- 2010. 00540.x. 42. Konar BH, Edwards MS, Bland A, Metzger J, Ravelo A, Traiger S, Weitzman 22. Keating KA. Mitigating elevation-induced errors in satellite telemetry BP. A swath across the great divide: kelp forests across the Samalga Pass locations. J Wildl Manag. 1995;59:801–8. https:// doi. org/ 10. 2307/ 38019 60. biogeographic break. Cont Shelf Res. 2017;143:78–88. https:// doi. org/ 10. 23. Block AB, Dewar H, Williams T, Prince ED, Farwell C, Fudge D. Archival tag-1016/j. csr. 2017. 06. 007. ging of Atlantic Bluefin tuna (Thunnus thynnus thynnus). Mar Technol Soc 43. Sadorus LL, Goldstein ED, Webster RA, Stockhausen WT, Planas JV, Duffy- J. 1998;32:37–46. https:// doi. org/ 10. 1007/ 978- 94- 017- 1402-0_3. Anderson JT. Multiple life stage connectivity of Pacific halibut (Hippoglos - 24. Pederson MW, Righton D, Thygesen UH, Andersen KH, Madsen H. Geolo- sus stenolepis) across the Bering Sea and Gulf of Alaska. Fish Oceanogr. cation of North Sea cod (Gadus morhua) using hidden Markov models 2021;30(2):174–93. https:// doi. org/ 10. 1111/ fog. 12512. and behavioural switching. Can J Fish Aquat Sci. 2018;65(11):2367–77. 44. Leaman BM, Geerneart TO, Loher T, Clark WG. Further examination of https:// doi. org/ 10. 1139/ F08- 144. biological issues concerning an extended commercial fishing season. In: 25. Le Bris A, Fisher JAD, Murphy HM, Galbraith PS, Castonguay M, Loher Sadorus L, editor. Report of assessment and research activities 2001. Seat- T, Robert D. Migration patterns and putative spawning habitats of tle: International Pacific Halibut Commission; 2002. p. 53–73. Atlantic halibut (Hippoglossus hippoglossus) in the Gulf of St. Lawrence 45. Stabeno P, Schumacher JD, Ohtani K. The physical oceanography of the revealed by geolocation of pop-up satellite archival tags. ICES J Mar Sci. Bering Sea. In: Loughlin TR, Ohtani K, editors. Dynamics of the Bering Sea. 2018;75:135–47. https:// doi. org/ 10. 1093/ icesj ms/ fsx098. Fairbanks: Alaska Sea Grant; 1999. p. 1–28. https:// doi. org/ 10. 4027/ dbs. 26. Nielsen JK, Mueter FJ, Adkinson MD, Loher T, McDermott SF, Seitz AC. 1999. Eec ff t of study area bathymetric heterogeneity on parameterization and 46. Stabeno PJ, Bond NA, Hermann AJ, Kachel NB, Mordy CW, Overland JE. performance of a depth-based geolocation model for demersal fishes. Meteorology and oceanography of the Northern Gulf of Alaska. Fish Ecol Model. 2019;24:18–34. https:// doi. org/ 10. 1093/ icesj ms/ fsx040. Oceanogr. 2004;10:81–98. https:// doi. org/ 10. 1016/J. CSR. 2004. 02. 007. 27. Stabeno P, Ladd C, Napp JM. Transport through Unimak Pass, Alaska. 47. Clark WG, Hare SR. Eec ff ts of climate and stock size on recruitment and Deep-Sea Res. 2002;II(49):5919–30. https:// doi. org/ 10. 1016/ S0967- growth of Pacific halibut. N Am J Fish Manag. 2002;22:852–62. 0645(02) 00326-0. 48. Tamario C, Sunde J, Petersson E, Tibblin P, Forsman A. Ecological and 28. Lanksbury JA, Duffy-Anderson JT, Mier KL, Bisby MS, Stabeno PJ. Distribu- evolutionary consequences of environmental change and management tion and transport patterns of northern rock sole, Lepidopsetta polyxystra, actions for migrating fish. Front Ecol Evol. 2019;7:271. https:// doi. org/ 10. larvae in the southeastern Bering Sea. Prog Oceanogr. 2007;72(1):39–62. 3389/ fevo. 2019. 00271. https:// doi. org/ 10. 1016/j. pocean. 2006. 09. 001. 49. Vestfals C, Ciannelli L, Hoff GR. Changes in habitat utilization of 29. Wright DL, Castellote M, Berchok CL, Ponirakis D, Crance JL, Clapham PJ. slope-spawning flatfish across a bathymetric gradient. ICES J Mar Sci. Acoustic detection of North Pacific right whales in a high-traffic Aleutian 2016;73(7):1875–89. https:// doi. org/ 10. 1093/ icesj ms/ fsw112. Pass, 2009–2015. Endanger Species Res. 2018;37:77–90. https:// doi. org/ 50. Webster RA, Soderlund E, Dykstra CL, Stewart IJ. Monitoring change in 10. 3354/ esr00 915. a dynamic environment: spatio-temporal modelling of calibrated data 30. Ualesi K, Wilson D, Jones C, Rillera R, Jack T. IPHC fishery-independent from different types of fisheries surveys of Pacific halibut. Can J Fish setline survey (FISS) design and implementation in 2020. Meeting Aquat Sci. 2020;77:1421–32. https:// doi. org/ 10. 1139/ cjfas- 2019- 0240. document IPHC-2022-AM098-07. Seattle: International Pacific Halibut 51. Castillo GC. Latitudinal patterns in reproductive life history traits of North- Commission. 2021. https:// www. iphc. int/ uploa ds/ pdf/ am/ am098/ iphc- east Pacific flatfish. In: Baxter B, editor. Proceedings of the international 2022- am098- 07. pdf. Accessed 20 Feb 2022. symposium on North Pacific Flatfish. Fairbanks: Alaska Sea Grant; 1994. p. 31. Courcelles D. Re-evaluation of the length-weight relationship of Pacific 51–72. https:// repos itory. libra ry. noaa. gov/ view/ noaa/ 38469. Accessed 7 halibut (Hippoglossus stenolepis). In: Sadorus L, editor. Report of assess- Apr 2022. ment and research activities 2011. Seattle: International Pacific Halibut 52. Lauth RR, Acuna A. Results of the 2006 eastern Bering Sea continental Commission; 2012. p. 459–70. shelf bottom trawl survey of groundfish and invertebrate resources. 32. Hare SR. Investigation of the role of fishing in the Area 4C CPUE decline. NOAA technical memorandum NMFS-AFSC-176. Seattle: US National In: Sadorus L, editor. Report of assessment and research activities 2004. Marine Fisheries Service. 2007. https:// apps- afsc. fishe ries. noaa. gov/ Publi Seattle: International Pacific Halibut Commission; 2005. p. 185–97.catio ns/ AFSC- TM/ NOAA- TM- AFSC- 176. pdf. Accessed 20 Feb 2022. 33. Loher T. Investigating variability in catch rates of halibut (Hippoglossus 53. Stewart I, Martell S. Development of the 2015 stock assessment. Meeting stenolepis) in the Pribilof Islands: is temperature important? Deep-Sea Res. document IPHC-2015-srb06-02. Seattle: International Pacific Halibut 2008;II(55):1801–8. https:// doi. org/ 10. 1016/j. dsr2. 2008. 04. 002. Commission. https:// www. iphc. int/ uploa ds/ pdf/ srb/ iphc- 2015- srb06- 02. 34. Lynch P. The equation of time and the analemma. Irish Math Soc Bull. pdf. Accessed 7 Apr 2022. 2012;69:47–56. 54. O’Boyle R, Dean M, Legault CM. The influence of seasonal migration on 35. Seitz AC, Norcross BL, Wilson D, Nielsen JL. Evaluating light-based geolo- fishery selectivity. ICES J Mar Sci. 2016;73(7):1774–87. https:// doi. org/ 10. cation for estimating demersal fish movements in high latitudes. Fish Bull. 1093/ icesj ms/ fsw048. 2006;104:571–8. L oher Animal Biotelemetry (2022) 10:18 Page 21 of 21 55. Wilberg MJ, Thorson JT, Linton BC, Berkson J. Incorporating time-varying catchability into population dynamic stock assessment models. Rev Fish Sci. 2010;18(1):7–24. https:// doi. org/ 10. 1080/ 10641 26090 32946 47. 56. Nielsen A, Berg CW. Estimation of time-varying selectivity in stock assess- ments using state-space models. Fish Res. 2014;158:96–101. https:// doi. org/ 10. 1016/j. fishr es. 2014. 01. 014. 57. Cimino M, Cassen M, Merrifield S, Terrill E. Detection efficiency of acoustic biotelemetry sensors on Wave Gliders. Anim Biotelem. 2018;6:16. https:// doi. org/ 10. 1186/ s40317- 018- 0169-4. 58. Lembke C, Lowerre-Barbieri S, Mann DA, Taylow JC. Using three acoustic technologies on underwater gliders to survey fish. Mar Technol Soc J. 2019;2(6):39–52. https:// doi. org/ 10. 4031/ MTSJ. 52.6.1. 59. Stockhausen H, Guðbjörnsson S. The earth’s geomagnetic field and geolocation of fish: first results of a new approach. ICES CM. 2009/B:19. 60. Nielsen JK, Mueter FJ, Adkinson MD, Loher T, McDermott SF, Seitz AC. Potential utility of geomagnetic data for geolocation of demersal fish in the North Pacific Ocean. Anim Biotelem. 2020;8:17. https:// doi. org/ 10. 1186/ s40317- 020- 00204-0. 61. Hussey NE, Orr J, Fisk AT, Hedges KJ, Ferguson SH, Barkley AN. Mark report satellite tags (mrPATs) to detail large-scale horizontal movements of deep water species: first results for Greenland shark (Somniosus microcephalus). Deep Sea Res. 2018;I(134):32–40. https:// doi. org/ 10. 1016/j. dsr. 2018. 03. 62. Gatti P, Fisher JAD, Cyr F, Galbraith PS, Robert D, Le Bris A. A review and tests of validation and sensitivity of geolocation models for marine fish tracking. Fish Fish. 2021;22:1041–66. https:// doi. org/ 10. 1111/ faf. 12568. 63. Loher T, Stephens SM. Use of veterinary ultrasound to identify sex and assess female maturity of Pacific halibut in nonspawning condition. N Am J Fish Manag. 2011;31:1034–42. https:// doi. org/ 10. 1080/ 02755 947. 2011. 64. Drinan DP, Loher T, Hauser L. Identification of genomic regions associated with sex in Pacific halibut. J Heredity. 2018;109(3):326–32. https:// doi. org/ 10. 1093/ jhered/ esx102. 65. Grusha DS, Patterson MR. Quantification of drag and lift imposed by pop-up satellite archival tags and estimation of the metabolic cost to cownose rays (Rhimnoptera bonasus). Fish Bull. 2005;103(1):63–70. 66. Jepsen N, Thorstad EB, Havn T, Lucas MC. The use of external electronic tags on fish: an evaluation of tag retention and tagging effects. Anim Biotelem. 2015;3:49. https:// doi. org/ 10. 1186/ s40317- 015- 0086-z. 67. Hammerschlag N, Cooke SJ, Gallagher AJ, Godley BJ. Considering the fate of electronic tags: interactions with stakeholders and user responsibil- ity when encountering tagged aquatic animals. Methods Ecol Evol. 2014;5:1147–53. https:// doi. org/ 10. 1111/ 2041- 210X. 12248. 68. Loher T, Dykstra CL, Simeon A, Hicks AC, Stewart IJ, Wolf N, Harris BP, Planas JV. Estimation of post-release mortality using acceleration-logging archival pop-up tags for Pacific halibut (Hippoglossus stenolepis) discarded from longlines. N Am J Fish Manag. 2021;42(1):37–49. https:// doi. org/ 10. 1002/ najfm. 10711. 69. Le Bris A, Fisher JAD, Murphy HM, Galbraith PS, Castonguay M, Loher T, Robert D. Migration patterns and putative spawning habitats of Atlantic halibut (Hippoglossus hippoglossus) in the Gulf of St. Lawrence revealed by geolocation of pop-up satellite archival tags. ICES J Mar Sci. 2018;1(1):135–47. https:// doi. org/ 10. 1093/ icesj ms/ fsx098. 70. Loher T, Rensmeyer R. Physiological responses of Pacific halibut, Hippoglossus stenolepis, to intracoelomic implantation of electronic archival tags, with a review of tag implantation techniques employed in flatfishes. Rev Fish Biol Fish. 2011;21(1):1027–37. https:// doi. org/ 10. 1007/ Re Read ady y to to submit y submit your our re researc search h ? Choose BMC and benefit fr ? Choose BMC and benefit from om: : s11160- 010- 9192-4. 71. Loher T, Geernaert TO. Captive holding to develop long-term archival fast, convenient online submission tagging protocols in Pacific halibut. In: Sadorus L, editor. Report of assess- thorough peer review by experienced researchers in your ﬁeld ment and research activities 2015. Seattle: International Pacific Halibut Commission; 2016. p. 445–63. rapid publication on acceptance support for research data, including large and complex data types Publisher’s Note • gold Open Access which fosters wider collaboration and increased citations Springer Nature remains neutral with regard to jurisdictional claims in pub- maximum visibility for your research: over 100M website views per year lished maps and institutional affiliations. At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

Animal Biotelemetry – Springer Journals

Published: Jun 2, 2022

Keywords: Hippoglossus stenolepis; Migration; Spawning; Satellite tagging; Bering Sea

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

Dispersal and seasonal movements of Pacific halibut (Hippoglossus stenolepis) in the eastern Bering Sea and Aleutian Islands, as inferred from satellite-transmitting archival tags

References (59)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies