Testing the Effectiveness of Marine Reserves: A Multi-Species, Multi-Reserve Experiment

PRINCIPLE INVESTIGATOR
S. James Taggart, US Geological Survey, Alaska Science Center - Biological Science Office, Glacier Bay Field Station, P.O. Box 240009, Douglas, Alaska 99824. (907) 364-1577. Jim_Taggart@usgs.gov

CO-PRINCIPLE INVESTIGATOR
Philip N. Hooge, US Geological Survey, Alaska Science Center - Biological Science Office, Glacier Bay Field Station, P.O. Box 140, Gustavus, Alaska 99826. (907) 697-2637. Philip_Hooge@usgs.gov

INTRODUCTION

In 1999, after great controversy, the National Park Service (NPS) created America’s largest temperate marine reserve by closing commercial fishing in parts of Glacier Bay National Park, Alaska (Department of the Interior 1999) .  During the 1990's, collapsing fisheries around the world caused doubt about the long-term sustainability of certain fisheries (Ludwig et al. 1993, Rosenberg et al. 1993, National Research Council 1995, Agardy 1997, Hastings and Botsford 1999, Murray et al. 1999) . Alaskan crustacean fisheries are particularly prone to serial depletion and collapse (Orensanz et al. 1998) . An emerging theoretical and empirical body of information hypothesizes that “no-take marine reserves” may promote marine biodiversity and enhance the long-term sustainability of many fisheries (Plan Development Team 1990, Carr and Reed 1993, Rowley 1994, Agardy 1995, Ticco 1995, Bohnsack and Ault 1996, Allison et al. 1998, Coleman and Travis 1998, Lauck et al. 1998, Garcia-Charton and Perez-Ruzafa 1999, Murray et al. 1999, Ward et al. 1999, National Research Council In Press) . 

In order to be effective, a marine reserve must be big enough to retain a large proportion of the protected population long enough for positive effects such as increased size, density, or fecundity to be realized (Polacheck 1990) . In addition, an effective reserve must include relevant habitat for the protected species (Dugan and Davis 1993) . In practice, however, reserve areas are often chosen pragmatically, on very limited information, and many reserves are created with no monitoring and evaluation procedures (McNeill 1994, Hockey and Branch 1997) . Monitoring reserves to determine if they meet their objectives is essential (McNeill 1994, Allison et al. 1998) . 

Marine reserves are likely to be an effective conservation tool for organisms which have relatively sedentary adult life stages (compared to the size of the reserve) and highly mobile larval stages so that the reserve can seed surrounding areas (Nowlis and Roberts 1999, Chiappone and Sealey 2000, Martell et al. 2000, Murawski et al. 2000, Pitcher et al. 2000, Roberts 2000, Warner et al. 2000) .  Reserve size and configuration are also vital factors that influence whether a marine reserve will effectively protect adult breeding population (Polacheck 1990, Demartini 1993, Guenette and Pitcher 1999) .  A small boundary to reserve area ratio results in low movement across the reserve boundary and spawner stock biomass increases in the reserve, shifting the age structure of the population to older individuals.  In addition, in order for a reserve to supply larvae to a larger surrounding area, it needs to be located in a portion of the species’ range that acts as an ecological source, rather than a sink (Pulliam 1988, Pulliam and Danielson 1991, Carr and Perry 1997, Roberts 1998, Crowder et al. 2000) .  

Although theoretical concepts and simulation models are rapidly developing for marine reserves, their effectiveness at protecting breeding adults has been demonstrated primarily in tropical areas (Agardy 2000) . Beneficial effects on areas surrounding reserves have been demonstrated in only a few studies (Alcala 1988, Murawski et al. 2000) . Data on the effectiveness of marine reserves are especially limited from high latitude ecosystems (Rogers-Bennett et al. 1995, Murawski et al. 2000, Paddack and Estes 2000) . High latitude reserves may be less effective than tropical reserves because temperate fish have broader movement patterns than coral reef fish (Fogarty 1999) . Thus to be effective, temperate reserves may have to be much larger (Guenette et al. 2000, Martell et al. 2000, Murawski et al. 2000) . 

Understanding how marine species in Glacier Bay interact with the surrounding waters is not simply a fisheries enhancement issue; it is fundamental to determining effective approaches for solving most marine resource issues in the Park. For example, if adult halibut remain in Glacier Bay most of their adult life and it is demonstrated that sport fishing is depleting the population, the NPS could restore the population by restricting the sport harvest of halibut in the Park. Alternatively, if halibut frequently move between the Bay and the rest of northern southeastern Alaska, a unilateral decision by the NPS to limit sport fishing in the Park would be ineffective at protecting this population. In this second hypothetical case, reversing the population decline would require an interagency approach involving the NPS, Alaska Department of Fish and Game, stakeholders and the North Pacific Fisheries Management Council. The regional nature of many marine processes will require developing management and research collaborations across jurisdictional boundaries if marine resource issues are going to be solved (National Research Council 2000) .  Executive Order 13158 (May 26, 2000) directs the Department of Interior and the Department of Commerce to take a collaborative regional approach and develop a National Network of Marine Protected Areas. 

The fishing closures in Glacier Bay resulted in the creation of a network of five marine reserves that vary in shape and range in size from 40 to 280 km² (Figure 1).  Since so little is known about reserves in temperate waters and because the reserves created in Glacier Bay are potentially large enough to meet conservation objectives for many species, the opportunity in Glacier Bay to test the effectiveness of a marine reserve network as a marine conservation management tool is important on a local, regional, national and even global scale. 

The retention of breeding adults in marine reserves is quantified in simulation models as transfer rate;  these models demonstrate that transfer rate is central to reserve effectiveness (Polacheck 1990, Demartini 1993, Guenette and Pitcher 1999) . We propose attaching sonic tags to Pacific halibut (Hippoglossus stenolepis), Tanner crab (Chionoecetes bairdi), and red king crab (Paralithoides camtschaticus), and measuring the transfer rate of the newly created reserves in Glacier Bay by deploying an ultrasonic gate along the boundary of each reserve.  This study will allow us to quantify the effectiveness of the reserves at protecting the adult breeding portion of selected populations. The quantitative testing that we are proposing will allow managers, scientists and the public to evaluate the utility of reserves as a management tool in the conservation of local and regional marine resources in Alaska.  It will also set the stage for future studies, which will address the effect of the reserves on larval supply and the role of reserves as ecological sources vs. sinks.

BACKGROUND

Commercial fishing has occurred in Glacier Bay since at least the turn of the century (Taylor and Perry 1990) . Commercial fishing continued under federal regulation after the establishment of Glacier Bay National Monument in 1925 and its subsequent enlargement in 1939. Since 1966, however, federal regulation and legislation have prohibited commercial fishing in Glacier Bay. In addition, the Wilderness Act has prohibited commercial fishing within Glacier Bay’s wilderness waters since 1980 (USNPS 1998) . Despite these regulations, commercial fishing activities continued in Park waters. 

Starting in 1990, the Department of the Interior attempted to resolve the commercial fishing issue through litigation and administrative rulemaking. In October 1998, Congress passed legislation that immediately closed Wilderness Waters and certain other areas within Glacier Bay to commercial fishing. A year later, the Department of the Interior published the Final Rule implementing Glacier Bay National Park commercial fishing legislation (Department of the Interior 1999) . The legislation established special regulations for commercial fishing in the marine waters of Glacier Bay National Park and provides: “protection of Park values and purposes, prohibition of any new or expanded fisheries, and opportunity for the study of marine resources.” 

Specifically, particular areas of non-wilderness waters in Glacier Bay proper and all wilderness waters within Glacier Bay National Park were closed to commercial fishing (Figure 1). Commercial fishing in the central portion of Glacier Bay proper was limited to three specific commercial fisheries (Pacific halibut, Tanner crab and salmon) and there is a phase-out (or “grand-father”) process for these three fisheries. The phase-out-process allows qualifying fisherman in the three authorized commercial fisheries to continue fishing in specific areas of Glacier Bay proper with nontransferable, lifetime permits. At the end of the phase out period, all commercial fishing will be closed in Glacier Bay proper. Finally, the legislation clarifies that the marine waters of the Park outside of Glacier Bay proper will remain open to various existing commercial fisheries and will be cooperatively managed by the State of Alaska and the Department of Interior. 

By closing commercial fishing in parts of the Park, the National Park Service has effectively created a network of five marine reserves that vary in size and shape.  Testing the effectiveness of a marine reserve depends largely on knowledge of movement patterns of the key species (Polacheck 1990, Carr and Reed 1993, Demartini 1993, Rowley 1994) . Little is known about the movements of species that are commercially harvested in Glacier Bay except halibut. A telemetry study of halibut in Glacier Bay suggests that Pacific halibut exhibit a developmental shift in home range patterns. Juvenile fish in Glacier Bay move widely but often remain within Park waters. Sexually mature fish occupy home ranges that are often less than 0.5 km² in size. These larger fish exhibit site fidelity and home ranges appear to be maintained within and between years (Hooge and Taggart 1998) . 

Information on Tanner crab movements is limited and there are no data from Glacier Bay. Donaldson (1983) tagged male Tanner crab tagged near Kodiak, Alaska and found that animals tagged in bays tended to move offshore while those tagged offshore remained in that general area. The mean net movement of male Tanner crabs was 27.9 km during the 4 year study. 

Movements of red king crab have not been studied in Glacier Bay. However, in nearby waters (Auke Bay, Alaska), female red king crab were tagged with ultrasonic tags were located at weekly intervals for over a year (Stone et al. 1992) . They gradually moved to deeper water in spring following mating and egg extrusion and remained at these depths through early November. Female crabs abruptly moved into shallow water during November where they resided until late February or March. Females molt and mate during March through May. The home range of primiparous female red king crabs (11.9 km²) exceeded that of multiparous crabs (avg. 3.6 km²) for a one-year period (Stone et al. 1992) . 

EXPERIMENTAL DESIGN

Marine reserve models demonstrate that transfer rate is central to reserve effectiveness (Polacheck 1990, Demartini 1993, Guenette and Pitcher 1999) . With low transfer rates, spawner stock biomass increases in the reserve, shifting the age structure of the population to older individuals. We propose to measure the transfer rate by attaching sonic tags to Pacific halibut, Tanner crabs and red king crabs and recording the departure of the tagged individuals by deploying (and maintaining) an ultrasonic gate along the boundary of each reserve for two years. The transfer rate between the entire Bay and Icy Strait will be measured with an ultrasonic gate deployed near the mouth of Glacier Bay. The ultrasonic gates will be constructed by anchoring data loggers along the boundaries at regular intervals so that 100% of the boundary is monitored (Figures 2-7). By tagging a random sample of individuals in each reserve, the proportion of the tagged animals that leave the reserve is an estimate of the transfer rate. After the ultrasonic gate and the tags are deployed, we will collect movement data with gates and boat-based band-transect sampling for two years. 

Halibut and Tanner crabs

Halibut and Tanner crab commercial fishing was closed in five reserves. We will measure the transfer rate from each reserve to the central portion of the bay by attaching sonic tags to a random sample of individuals and monitoring their departure with an ultrasonic gate. In the future, the entire Bay will be closed to commercial halibut and Tanner crab fishing as the “grand-fathered” fishermen leave the fisheries. Thus evaluating how the entire bay functions as a reserve for halibut or Tanner crabs is of interest for future management decisions as well as for the current management of fishing during the phase-out period. To measure the transfer rate of the entire Bay we will attach sonic tags to a random sample of animals throughout the Bay, including the central portion. We recognize that the transfer rate for the entire Bay may change when it is closed to all commercial fishing. If an increase in biomass occurs after the full closure, we predict that the transfer rate will increase as individuals move from higher abundance (Glacier Bay) to areas of lower abundance (Icy Strait). Consequently, the initial bay-wide transfer rates are likely to be biased low compared to transfer rates after the Bay closes to halibut and Tanner crab fishing. The animals tagged in the central part of the Bay will also allow us to estimate the transfer rate from the fished area into the reserve areas. 

Red King Crab

The entire bay was closed to red king crab commercial fishing in 1999. We plan, however, to attach the sonic tags to king crabs in each of the five reserves as well as the central part of the Bay. Since the area adjacent to the reserves (central part of the Bay) is closed to king crab commercial fishing we predict that the transfer rates from the “reserves” to the central part of the bay for king crabs is likely to be biased low. That is, if the central portion of the Bay was commercially fished and the biomass was higher in the reserves, we would expect more movement from areas of high to lower abundance (from the reserve to the fished area). We believe that the opportunity to measure king crab transfer rates is unique, so tags should be deployed among the five reserves as well as in the entire bay. 

Specific Hypotheses

Marine reserve models predict increases in spawner biomass as transfer rate decreases with the largest responses in spawner biomass occurring at rates less than approximately 25% (Polacheck 1990, Demartini 1993, Guenette and Pitcher 1999) . We will test the effectiveness of each reserve by measuring the transfer rate from the reserves to the surrounding water. Because models with specific response curves for crabs and halibut have not been developed, we propose to use a transfer rate of 25% as our null hypothesis. 

To refine our estimate of critical transfer rates we have initiated a collaboration with Dr. Carl Walters and Dr. Sylive Guenette (University of British Columbia, Canada) to develop a spatially and age-specific population model for Glacier Bay halibut, Tanner crabs and king crab. These models will be developed from a generic model (FISHMOD) developed by Walters. FISHMOD has recently been used to model the response of lingcod (Ophiodon elongates) (Martell et al. 2000) and Atlantic cod (Gadus morhua) (Guenette et al. 2000) populations to various management options including marine reserves. Cod have much higher transfer rates than lingcod and consequently much larger reserves are required for effective conservation. At the end of this study, we will update the models with the empirical transfer rates and use FISHMOD to predict changes in the population structure within the reserves and in Glacier Bay. These predictions will be used for developing hypotheses for future studies. 

Our first hypotheses will test the effectiveness of each reserve with respect to critical transfer rates: 

H_o: The transfer rate for female halibut (king crab, Tanner crab) from Geikie Inlet (Scidmore Bay-Charpentier Inlet, West Arm, East Arm, Beardslee Islands) to the central (open to fishing) portion of Glacier Bay is > 25%. 

H_a: The transfer rate for female halibut (king crab, Tanner crab) from Geikie Inlet (Scidmore Bay-Charpentier Inlet, West Arm, East Arm, Beardslee Islands) to the central (open to fishing) portion of Glacier Bay is <= 25%. 

Our second hypothesis will test the effectiveness of the entire Bay as a marine reserve with respect to critical transfer rates: 

H_o: The transfer rate for female halibut (king crab, Tanner crab) from the entire Bay to Icy Strait is > 25%. 

H_a: The transfer rate for female halibut (king crab, Tanner crab) from the entire Bay to Icy Strait is <= 25%. 

Our third hypothesis will test if the transfer rate out of the reserves is the same as the transfer rate into the reserves: 

H_o: The transfer rate for female halibut (king crab, Tanner crab) from the reserves (Geikie Inlet, Scidmore Bay-Charpentier Inlet, West Arm, East Arm, Beardslee Islands) to the central (open to fishing) portion of Glacier Bay is equal to the transfer rate from the central portion of Glacier Bay to the reserves. 

H_a: The transfer rate for female halibut (king crab, Tanner crab) from the reserves (Geikie Inlet, Scidmore Bay-Charpentier Inlet, West Arm, East Arm, Beardslee Islands) to the central (open to fishing) portion of Glacier Bay is greater than the transfer rate from the central portion of Glacier Bay to the reserves 

A fourth hypothesis will test the differences between the reserves based on boundary to area ratio:

H_o: There is no difference in transfer rate between reserves with high and low boundary: area ratio. 

H_a: The transfer rate will be higher in reserves with a high boundary: area ratio. 

Power Analysis

Since our goal is to simultaneously estimate the transfer rate of multiple species in multiple reserves, it is important to keep the tagged population sizes small without compromising power. The proportion of tagged animals that leave the reserve is an estimate of the transfer rate. We calculated power of a predicted proportion compared to a population proportion using a binomial distribution (one tailed test) with a program developed by Simple Interactive Statistical Analysis (Figure 8). We ran two analyses by varying the hypothesized effective transfer rate between 15% and 25%. For each analysis we contrasted population transfer rates of 0.01%, 0.05%, and 0.10%. From these analyses we selected a hypothesized transfer rate of 25% and a sample size of ten for species we expect to have very low transfer rates while a sample of 20 was selected for the rest of the species (Table 1). For the two crab species, the movement patterns of males and females are likely to be very different, so we treated males and females as separate samples. We anticipate female king crabs and female Tanner crabs to have very low transfer rates (0.01), so we selected a sample size of 10. We expect the male king and Tanner crabs to have higher transfer rates (0.05), so we selected a sample size of 20. We estimate the power for these samples is greater than 80%. 

The growth rate of male halibut is slower than females; male halibut rarely exceed 35 kg while females frequently exceed 100 kg reach (Schmitt and Skud 1978) . We plan to limit the transfer rate study of halibut to mature females for two reasons: 1) if transfer rates of females are determined to be low, we will design new studies which will focus on measuring changes in the size structure of the female portion of the population, and 2) it is difficult to differentiate mature males from immature females unless you examine the gonads which requires surgery or sacrifice of the animal. Based on the home range analysis of Pacific halibut (Hooge and Taggart 1998) , we predict that mature female halibut will have a low transfer (0.05) rate so we selected a sample size of 20 for this species. 

Table 1. Proposed number of sonic tags in the 5 reserves and open portion of Glacier Bay per species by gender. 

METHODS  

1. Study Area

The study area includes Glacier Bay and Icy Strait. Glacier Bay is an ecosystem characterized by change. The Little Ice Age glacial advance of about 80 km and its subsequent rapid retreat over the past 200 years are the largest glacial fluctuations of historic record. Both the physical and biological characteristics of Glacier Bay have shown sensitivity to climate change. The causes of glacial destabilization are not well documented, although most workers infer global warming as a cause. With glacial fluctuation, habitats and ecosystems can change dramatically. Advancing glaciers cover habitat and increase sediment and freshwater discharges. Retreating glaciers expose new areas, but these new habitats are composed of unstable substrates subject to rapid changes. The volume of freshwater entering the Glacier Bay changes markedly as glaciers advance and retreat, modifying salinity and water column structure. Increased freshwater discharge influences estuarine circulation processes, as well as nutrient influx. All of these factors result in a high variation in the location and amount of primary production, which in turn affects food webs in both the marine and terrestrial ecosystems. 

2. Marine Reserve Study Sites

Our study sites include the central part of Glacier Bay and five adjacent areas in the Park where commercial fishing was recently closed: Geikie Inlet, Scidmore Bay-Charpentier Inlet, West Arm, East Arm and Beardslee Islands. The central portion of the Bay remains open to commercial halibut and Tanner crab fishing. Thus, the five closed areas are reserves for Tanner crabs and halibut while the entire bay is a reserve for red king crab (and all other commercially fished species). 

3. Research Vessels

A 50’ USGS research vessel, R/V Alaskan Gyre, will be used to deploy and retrieve data loggers, catch and tag organisms, and conduct the band-transect searches. A 26’ USGS research vessel, R/V Eider, will also be used for retrieving data loggers and conducting band-transect searches. During band-transect searches, the location of the vessels will be continuously recorded by downloading Global Positioning System (GPS) fixes and times onto an on-board computer. 

4. Sonic Tags

Halibut and crabs will be tagged with sonic V16-5H RCODE sonic tags manufactured by VEMCO (Shad Bay, Nova Scotia, Canada). The cylindrical sonic tags are 92 mm long and 16mm in diameter (Figure 9). The tags will be programmed for a pulse interval of 2.5 minutes and, with lithium batteries, will have a life expectancy of 700 days. All tags will transmit at the same frequency (50 kHz). Unlike conventional sonic tags that transmit a single ping, RCODE coded tags transmit a burst of 6 pings followed by an off time interval. The burst of 6 pings encodes an identification number that can be decoded by a receiver and stored into memory. A total of 65,536 RCODE sonic tags can be programmed with unique codes. The duration of the off time interval is programmed to vary randomly about 10%, thus the pulses from multiple tags will only overlap briefly. A long pulse interval will conserve battery life as well as increase the number of tags that a single receiver can concurrently decode.  

For all species, tags will be attached to mature individuals. Since most female halibut are mature at 120-cm and few males reach 120-cm (Clark et al. 1999b) , we will select mostly mature females by simply selecting individuals greater than 120-cm. The sonic tags will be implanted into halibut using techniques previously developed at Glacier Bay (Hooge and Taggart 1998) .  

Mature male Tanner crabs greater than 110-mm have a molt interval greater than two years (Paul and Paul 1995) and recently molted males can be identified by carapace condition (Jadamec et al. 1999) . We will select recently molted male crabs greater than 110-mm to minimized tag loss by molting during the two year study. Mature female Tanner crabs have a low probability of molting once they reach sexual maturity. Individual females that are sexually mature will be identified by the shape of the abdominal flap (Jadamec et al. 1999) .  

There is limited information on the molt interval of red king crab. Weber and Miyahara (1962) report that king crabs with carapace length of 126 mm, 142 mm, 158 mm, and 174 mm had molting probabilities of .87, .65, .37, and .03 respectively. To minimize tag loss from molting we will select recently molted king crabs and select crabs with a carapace length greater than 142 mm. To accomplish our goal of measuring the transfer rate of a sample of king crabs for two years, we will attach additional tags as tags are shed during molting. Tags will be glued to the carapace with epoxy (Stone et al. 1992) .  

Crabs will be collected with conical, top-loading 7.3 ft X 3 ft commercial Tanner crab pots which will have the same specifications as pots used by the Alaska Department of Fish and Game for king crab and Tanner crab surveys. We will use the same methods the ADF&G use on their surveys (soak time, bait type, bait quantity) (Clark et al. 1999a) , since standardized methods will facilitate interagency analyses of the pooled data. Halibut will be captured using snap-on longline fishing gear baited with herring. Longlines will be short (50 hooks spaced every 2 fathoms on 100 fathoms of ground line) so that we can set numerous lines and distribute the effort across the sampling grid.  

The effectiveness of a reserve depends on the reserve size and the home range of a species (Kramer and Chapman 1999, Walters 2000) . Even if a species has a small home range, individuals that are close to the reserve boundary will, through random movements, cross back and forth over the border. In order to avoid over-estimating the transfer rates from the individuals near the boundary, we propose setting an arbitrary 1-km “no-tag zone” on either side of the reserve boundaries.  

Sonic tags will be attached to a random sample of the mature portion of the population in each reserve and the in the entire Bay.  First, we will systematically set crab pots and halibut longlines at sampling stations on a 2-km grid for the large reserves, and a 0.5-km grid for the small reserves (Figure 10).  The number of sonic tags attached at each station will be proportional to catch per unit effort (CPUE) measured at each station compared to the rest of the stations.  After the number of sonic tags for each station has been determined by the CPUE, we will attach sonic tags to randomly selected crabs and halibut.  This procedure will distribute the tags randomly in each reserve. 

5. Data Loggers

The ultrasonic gates will be constructed by mooring VR2 Single Channel Monitors (Figure 9) along the boundary of each reserve and the mouth of Glacier Bay (Figures 2-7). The VR2 data loggers are dedicated remote monitors designed to detect RCODE sonic tags; the data loggers will record the sonic tags’ individual identification and the date and time when a tagged animal comes into range. The VR2 monitors have a battery life of 180 days and can store 300,000 sonic tag detections. They must be retrieved to be downloaded but can be redeployed at the same time. If we tagged 600 animals and those animals moved randomly in the Bay, each data logger on average would record 130,000 fixes. We will start with a frequent data logger retrieval schedule.  Based on the amount of memory filled during the first two months, we will develop an appropriate retrieval schedule for each gate to so that the VR2 memory will not be exceeded. 

Based on preliminary field tests in Glacier Bay (December 2000), the VR2 data loggers can detect and decode signals from any direction at 800 meters, so we estimate that the data loggers should be placed 1600 meters apart along the reserve boundary. There are three factors, however, that affect the reception radius of the data loggers: 1) distance the tagged animal is from the bottom, 2) swimming speed of the tagged organism, and 3) the pulse rate of the sonic tags. As all of these factors increase, the effective radius of the data logger reception decreases (Figure 11). If all of the organisms to be tagged were slow moving and stayed on the bottom, then we would place the data loggers 1600 meters apart. However, because we are also tagging halibut that can swim fast and may swim up into the water column, we will decrease the spacing of the data loggers to maintain a high reception probability. Exact spacing will be determined by Equation 1.

Where: ER= Effective Radius of the data logger; R= 800 m (the reception radius of the data loggers on the bottom); D= depth (meters); FS= fish speed (meter/min); and P= pulse rate interval of the sonic tags (min.). 

Data loggers will be suspended 5-10 meters off the bottom on short moorings with subsurface floatation (Figure 12). Subsurface floatation eliminates numerous problems associated with surface buoys (e.g. navigational hazard, fouling with kelp or logs, visual impact to visitors, freezing in ice during the winter). Disposable, degradable anchors will be used to secure the moorings to the bottom. The mooring configuration (i.e. anchor, hardware, floatation, line, etc.) will be modeled by Marinna Martini, an ocean engineer with the USGS Woods Hole Field Center to determine the necessary specifications required for the currents in Glacier Bay. Dr. Gary Bowen (University of Alaska Southeast, Juneau) will also be providing input on anchor design and construction. We will initially take a dual approach for deployment of these moorings. Acoustic release units (Figure 13) will allow individual moorings (Figure 12) with data loggers to be set and retrieved through the remote activation of the release mechanism. As a backup, we will attach subsurface moorings to a retrievable longline (Figure 14), which will allow us to retrieve the data loggers if an acoustic release fails. The inclusion of both the individual mooring method and the long-lining method of deployment allows for added security and flexibility in retrieval. 

6. Band-transect Searches

Band-transect searches will be performed inside and outside the reserves every 6 months to determine each tagged animal’s location. The band-transects will also be used to detect if each sonic tag is still attached to a live animal by determining if the animals have moved since the last search. We are also exploring with VEMCO the implementation of a mortality mode which would be triggered when an animal stops moving. The mortality tag would incorporate an accelerometer that would be queried by the microprocessor at a frequency and duration appropriate for each species.  

VEMCO manufactures a system for accurately monitoring the position of sonic tagged animals with a moored hydrophone array called Radio-Linked Acoustic Positioning (RAP). This system determines an animal’s location by measuring the difference in arrival time of the sonic signal to pairs of hydrophones. We are working with VEMCO to modify this system so that it will work with mobile hydrophones. We are also exploring the modification of a similar positioning system developed by Dr. Doug Wartzok (University of Missouri-St. Louis) and Dr. Brendan Kelly (University of Alaska Southeast, Juneau) to study the foraging ecology of ringed seals (Phoca hispida) (Kelly and Wartzok 1996) . 

Band-transect searches will be performed by towing a VH65 omni-directional hydrophone attached to a VR60 receiver. During preliminary testing in December 2000, we were able to tow a VH65 hydrophone 10 meters below the surface at 8km/hour by trailing the hydrophone behind a side-scan sonar housing. At this speed, the VR60 was able to decode 50 kHz tags up to 1050 meters away. However, the reception distance of the VR60 is influenced by water depth and boat towing speed, as shown in Equation 2, and band-transect width will be adjusted accordingly.

Where: ET= Effective Band-transect Width (m); R= 800 m (the reception radius of the data loggers on the bottom); D= depth (meters); BS= boat speed (meter/min); and P= pulse rate interval of the sonic tags (min.). 

Figure 15a shows a hypothetical band-transect search for the entire bay. There are some inlets in the Bay that are narrow enough to be searched in one strip and the speed of the boat would not be limited on the return. These “fast track” areas are shown in Figure 15b. We estimate that the time required to accomplish the band-transect searches will range from 104 to 157 hours depending upon the vessel(s) used and the towing speed (Table 2). 

7. Databasing and archival

Data from this project will be incorporated into the Glacier Bay Ecosystem Project data model. All attribute data will be placed in databases linked to the ARC/Info Geographic Information System. All purely spatial databases will be in ARC/Info format. Spatial data will also be archived as SDTS (Spatial Data Transfer Standard) format. All completed data sets will also be archived on CD-ROMS at the Glacier Bay Field Station. Project field notes will be written according to Glacier Bay Data Plan protocol standards and archived at the Glacier Bay Field Station. 

Table 2. Estimates of band-transect tracking time.

PRODUCTS

Scientific Journal Publication: Design and implementation of ultrasonic gates to measure the movement of sonic-tagged fish and crabs. 

Scientific Journal Publication. The distribution of Tanner and king crab, in a recently deglaciated fjord ecosystem. 

Scientific Journal Publication: Effectiveness of marine reserves: The results of the bay-wide multi-species movement experiment. 

Scientific Journal Publication: The distribution of Pacific halibut in a recently deglaciated fjord ecosystem. 

Scientific Journal Publication: Halibut reproduction: Timing and duration of spawning migration in a recently deglaciated f