摘要：Movement data represent important inputs to both numerical and conceptual models that contribute to the assessment and management of sablefish (Anoplopoma fimbria) and Pacific halibut (Hippoglossus stenolepis) in the eastern North Pacific. While conventional mark-recovery data have greatly increased our understanding of these populations, tagging methods that produce only endpoint data (i.e., locations at tagging and recapture) cannot address dynamics such as the timing, frequency, and recurrence of movements. As a preliminary step in evaluating the feasibility of using acoustic biotelemetry for addressing management-relevant questions for relatively deepwater (≤ 600 m depth) species, we conducted tests of the detection range between VEMCO V16-5H transmitters and VR2W receivers. Six range-test stations were established that varied from a relatively shallow (154 m), narrow (3.8 km) ocean pass; through the mouth of a wide (20 km), deep (600 + m) fjord; and over the continental shelf. Each station consisted of one receiver mooring and five transmitter moorings deployed at regular distances from 200 to 1200 m from the base of the receiver mooring. Bottom depth varied from 195 to 588 m. Transmitters were suspended 0.5 and 5 m off bottom, and receivers at 145–400 m below the sea surface depend on bottom depth. Receivers were deployed for 162–595 days; however, maximum transmission time was 365 days. The rate at which acoustic transmissions were detected by the receivers decreased with distance from the transmitters, mostly monotonically; at first gradually, from transmitters at distances of 400–800 m from the receivers, and then more steeply at distances beyond 800 m. Considerable variability in detection rates was observed, including low detection rates from individual locations and distances. Temporal changes in detection rates were observed that included convergence/divergence between detection of transmitters at a given distance from receivers but that were suspended at different elevations above bottom; stepwise seasonal shifts in mean detection; diurnal periodicity that was likely due to diurnal winds; and tidally mediated detection cycles. We determined that a linear “gate” of receivers spaced 1000 m apart would have a very high probability (≥ 0.99) under most conditions of detecting an acoustically tagged Pacific halibut transiting the gate, and that tagged-halibut identification rates using 2400-m receiver spacing could approach or exceed 90%. However, during episodic periods of relatively poor tag detection, these rates are likely to decline to ~ 60–80%; and to below 10% under extreme conditions. Acoustic telemetry is a feasible technique for monitoring the movement of deepwater epibenthic fishes. However, temporal and spatial variability in detection rates, especially in seasonally stratified waters, has a strong potential to impart biases on study results and generate spurious patterns unrelated to tagged-fish behavior. The nature of such biases should be carefully considered during array placement and data interpretation. Range testing should be conducted at all geographic locations at which arrays are to be established, rather than extrapolating from a few test locations. Arrays should always contain sentinel tags to monitor detection rates and account for transient features that affect those rates. We also note the necessity of placing at least two gates at any given location in order to substantiate movement of tagged animals past any desired point of reference.
