We will conduct field investigations around St. Paul Island, where a rich historical dataset is available to inform habitat and juvenile abundance from the 1980s (Armstrong et al. 1987). These data are now available in GIS and will be used to inform site selection. We will view habitat with drop cameras at 16 sites randomly selected from historical survey locations in previously identified shell hash habitat. These historical studies were conducted at a variety of depths, primarily between 40-60 m and did not explore shallow nursery habitat (0-20 m depth), where blue king crab juveniles have anecdotally been reported to occur, and therefore we will expand the scope of the historical studies to include 16 shallow sites (0-20 m depth), which are accessible for SCUBA diving. These shallow sites will be selected based on site accessibility and drop camera identification of shell hash or rock/gravel habitat. A total of 32 sites will be sampled, with 16 “historical” sites (40-60 m) and 16 “shallow” sites (0-20 m). Habitat and larval supply will be sampled at the historical sites and habitat, larval supply, dive surveys, and predation will be sampled at the shallow sites, as detailed further below.
Drop cameras will be used to re-verify select Armstrong et al. (1987) habitat distributions at historical sites and for preliminary site selection for dive surveys at shallow sites. Before any data collection with drop cameras, initial deployments will occur in tandem with dive surveys to confirm image recordings with diver observations. At each site, a camera will be deployed from a boat and lowered to the bottom with a weighted line, so that the camera will rest above the bottom to gather still images. The camera will be lowered ten times along a 1 km path to characterize the site. A tethered GoPro camera in a housing will be used as a low cost option to visualize the bottom habitat, as this new technology achieves good results in identifyingmicrohabitat (Schmidt and Rzhanov 2012, Struthers et al. 2015). This inexpensive method is used in lieu of benthic tows with bottom sleds (Rooper 2008, Williams et al. 2010) because of the availability of rich historical habitat data. Substrate will be classified following Armstrong et al. (1987) as rock, sand, mud, gravel, cobble with presence of shell hash noted. In the event that habitat re-verification does not reveal similar habitat types, then future studies could use more sophisticated and more expensive tools, such as sleds or sonar. Image processing will follow Schmidt and Rzhanov (2012) and will include prior calibration of camera(s) for depth of vision (single and/or stereo cameras), storage of images or recordings, image quality processing, and integration of computer based time-stamp and GPS location. Detailed habitat information will be collected by divers in shallow sites (<20 m) and methods are described further below.
Larval collectorsLarval collectors will be used to quantify larval supply and settlement around St. Paul Island at historical and shallow sites. Where shell hash habitat is confirmed at historical and shallow sites, larval collectors will be placed at 8 historical (40-60 m) and 8 shallow (0-20 m) sites, for a total of 16 sites. Collectors will be set at locations around St. Paul in April and collected in August. Collectors will be constructed of strings of 3 sacs of polyethylene netting stuffed with gillnet to form a sausage shape and attached to a groundline anchored on each end (Blau and Byersdorfer 1994, Pirtle & Stoner 2010). Hobo temperature loggers (Onset Computer Co.) will be affixed to the anchor at four sites, two historical and two shallow, to record temperature. Collectors will be retrieved and rinsed of contents with seawater. Sampled crabs will be counted, measured, and retained alive for predation studies (below). These individuals will be maintained alive in St. Paul at Trident Seafoods using their continuous flow seawater system.
Dive surveys will be conducted at 16 shallow sites (0-20 m) including the 8 shallow sites where the larval collectors were deployed. At each site, two 35 m survey transects will be surveyed, with one transect at each of 2 depths (~7 and ~14 m) running parallel to shore and/or the depth contour.During the survey, one diver will count pelagic fish along the 35 m swath within a 3 m cylindrical diameter visual field while the other diver will count demersal fish and large mobile macroinvertebrates along a 35 x 1 m swath. Then in 6 evenly spaced 0.5 x 0.5 m quadrats divers will quantify percent cover of substrate (rock, sand, mud, gravel, cobble and presence/absence of shell hash) and macroalgae and number of macroinvertebrates, including juvenile red and blue king crab, noting microhabitat type for each.
Dive surveys and experiments
Based on dive surveys from Year 1, two shallow sites will be selected for further experimentation. Transect and quadrat surveys will be repeated at each site at the beginning of Year 2 and permanent markers will be established at each transect.
Larval collectors – Year 2
Larval collectors will be deployed at the same sites as in Year 1 to investigate interannual differences in larval supply. In addition, we will increase the number of collectors at sites where larval supply was highest in year 1 in order to collect more blue king crab juveniles for tethering experiments. If blue king crab larvae and juveniles are not collected, then hatchery reared blue or red king crab will be used in tethering experiments, with modifications as needed by the state permitting agency.
Tethering experiments – Year 2Tethering experiments will be conducted to identify predation potential for blue king crab juveniles and to identify blue king crab juvenile predators. Twenty-four hours prior to tethering experiments, individual crabs will be tethered with a 15 cm length of 2 lb monofilament fishing line (Pirtle et al. 2012); one end attached to the crab with cyanoacrylate glue and the other a small fishing swivel. Tethered crab will be isolated prior to field experiments in small 250 ml containers of seawater held in a larger tank to maintain temperature. In the field, all tethering experiments will take place nearby permanent transects and will last approximately 24 hours with 1 crab trial per plot. Initial setup will include placing a concrete block with an eyebolt drilled into the center buried below the natural substrate. At the beginning of the trial, the swivel attached to the end of the monofilament will be cable tied to the eyebolt and a temporary predator exclusion cage will be placed over the crab. Tethered crabs will remain under the exclusion cage for a minimum of 4 h to allow acclimation and prevent high initial predation (inferred artifact seen in red king crab experiments; Christopher Long, pers. obs.). After acclimation, at or near sundown, the cage will be removed and the crab will be exposed to predators. Divers will return at dawn the next day (after ~12 h of exposure) and again at sundown (after ~24 h total) and will record crab state (alive, molted, consumed by predator) along with the behavior and microhabitat of each live crab. To identify predators, a subset of crabs will be monitored with video cameras for 24 h. Cameras will be equipped with infrared LED lights for illumination that avoid artifacts of with predators being attracted to (or avoiding) visible light sources. Video will be recorded and later viewed in their entirety noting time and ID of any predators in sight, predator behavior, predator attacks, and crab consumed, crab habitat use, and crab behavior (as per Pirtle et al. 2012 and Daly et al. 2013). We aim to achieve 100 trials, with video observations for 40 trials; however, we recognize the difficulty of this goal and will be happy with smaller sample sizes, recognizing that any data we can achieve in this environment is very worthwhile.
Predation – fish stomach contents
Additional data on predation of juvenile blue king crab by fish predators will be attained through small scale, small individual targeted pot fishing. Under the assumption smaller, nearshore fishes are eating the smallest stage juvenile crab (Pirtle et al. 2012), we will attempt to secure permits for small mesh pots to capture age-0 to age-5 demersal Pacific cod (Gadus macrocephalus) and flatfishes yellowfin sole (Limanda aspera), rock sole (Lepidopsetta polyxystra), and flathead sole (Hippoglossoides elassodon), arrowtooth flounder (Atheresthes stomias), and Greenland turbot (Reinhardtius hippoglossoides). Fish will be captured with modified crab pots (smaller mesh and escapement mechanisms) using SCUBA after 24 hours pot soak. Fish will be corralled into dive bags and taken to surface and laboratory; while bycatch, including any larger juvenile blue king crab, will be released without handling through opened pot escapement mechanisms. Fish will be morphologically measured and stomachs removed. Prey within stomachs will be identified to lowest possible taxonomic level and weighed for proportional contribution to diet. This study on vertebrate fishes was approved by the UAF-IACUC committee, number 1035800.
All data collected will be organized in standard spreadsheet and txt format for database storage and maintenance. QA/QC will be done immediately upon data entry and periodically as required and backed up on multiple devices. Year 1 data, being almost entirely natural, non-experimental abundances or categorical variable designates, will be analyzed with basic statistics, ANCOVA, and Principle Component Analysis (PCA) to determine how habitat, crab, predator, and oceanographic measures are correlated. Biogeospatial maps will be created by habitat type and crab species abundances. An index of settlement probability will be calculated for each site by dividing the average number of blue king crabs per larval collector (a measure of larval supply) by the average density of crabs in the quadrats (a measure of settlement). This will be analyzed with the corresponding habitat data to identify EFH for settling glaucothoe. Micro-habitat preference and use by red and blue king crabs will be used to assess the potential degree of overlap between juveniles of the two species to allow estimation of the degree of competition and predation that may occur between the species (Long et al. 2015). Additional fractal dimension analyses will be performed on quadrat, transect, and intra-site transect levels to determine spatial patterns associated with habitat scaling and will be subject to PCA (Falconer 2003, Seuront 2010). Tethering data from the discrete diver checks will be fit to a base model assuming a constant rate of predation such that S = e-pt where S is the probability of survival, p is the predation rate and t is the time in hours (Long et al. 2012). Using maximum likelihood, a series of models will be fit to the data assuming a binomial distribution in which the predation rate will be allowed to vary with time-of-day (Day or Night), Habitat, or their interaction as well as Site. The Alkaike’s Information Criterion corrected for small sample size (AICc) will be calculated for each model and be best model selected (Burnham and Anderson 2002). From the video footage the identity of predators that attack and those that consume the tethered crab will be determined. The probability of a species being a predator will be normalized to the abundance observed in the transects and used as a rank of predator importance. This index could then be used to look at long-term trends in predator abundance. In addition, the behavior of the tethered crab will be categorized (moving, motionless, climbing, foraging; Pirtle et al. 2012) and analyzed with general linear models with time-of-day and habitat as factors.