Bird scaring lines (BSLs) protect longline fishing gear from seabird attacks, save bait, reduce incidental seabird mortality and are the most commonly prescribed seabird bycatch mitigation measure worldwide. We collaborated with fishermen to assess the efficacy of applying BSL regulations from the demersal longline sablefish fishery in Alaska to a similar fishery along the U.S West Coast. In contrast to Alaska, some U.S. West Coast vessels use floats along the line to keep hooks off the seafloor, where scavengers degrade the bait and the target catch. Our results confirmed that BSL regulations from Alaska were sufficient to protect baits from bird attacks on longlines without floats, but not baits on longlines with floats. Longlines with floats sank below the reach of albatrosses (2 m depth) at a distance astern (157.7 m +/- 44.8 95% CI) that was 2.3 times farther than longlines without floats (68.8 m +/- 37.8 95% CI). The floated longline distance was well beyond the protection afforded by BSLs, which is approximately 40 m of aerial extent. Black-footed albatross attacked floated longlines at rates ten times more (2.7 attacks/1000 hooks, 0.48-4.45 95%CI) than longlines without floats (0.20 attacks/1000 hooks, 0.01-0.36 95% CI). Retrospective analysis of NOAA Fisheries Groundfish Observer Program data suggested that seabird bycatch occurs in a few sablefish longline fishing sectors and a minority of vessels, but is not confined to larger vessels. Analysis also confirmed fishermen testimonials that night setting reduced albatross bycatch by an order of magnitude compared to daytime setting, without reducing target catch. Night setting could be an effective albatross bycatch prevention practice if applied to the U.S. West Coast sablefish longline fishery and provide a practical alternative for vessels that elect to use floated longlines. These results highlight the importance of understanding region-specific longline gear modifications to identify effective bycatch reduction tools and the value of working collaboratively with fishermen to craft solutions.

We comprehensively tested combinations of three primary mitigation measures in a pelagic longline fishery with one of the highest rates of interaction with what may be the world's most challenging seabird assemblage (dominated by Procellaria genus petrels), aboard fishing vessels typical of the Asian distant water fleet. Multiple measures were used to compare the performance of weighted vs. unweighted branch lines set with two bird-scaring lines - hybrid lines with long and short streamers - during daytime and nighttime. The weights used were a novel double-weight configuration. Secondary attacks on baits brought to the surface by white-chinned petrels drove albatross mortality. Regardless of time of day, weighted branch lines with two bird-scaring lines, deployed and maintained with an aerial extent of 100m, reduced bird attacks by a factor of four, and secondary attacks and seabird mortality by a factor of seven, compared to unweighted branch lines, with little effect on fish catch rates and with no injuries to crew. This combination yielded zero bird mortalities when gear was set at night. We conclude that the simultaneous use of two bird-scaring lines, weighted branch lines and night setting meet our criteria for best-practice seabird bycatch mitigation for the joint-venture fleet targeting tuna and related species in the South African EEZ, To be successful, the aerial extent of bird-scaring lines should be aligned with the distance astern that baited hooks sink beyond the foraging depth of the dominant seabird - in this case white-chinned petrels to a depth near 5 m. Given that these measures were successful in one of the most challenging pelagic longline fisheries, they are likely to be widely applicable to pelagic longline fisheries using similar gear. (C) 2013 Elsevier B.V. All rights reserved.

Freshwater discharge from large rivers into the coastal ocean creates tidally-driven frontal systems known to enhance mixing, primary production, and secondary production. Many authors suggest that tidal plume fronts increase energy flow to fish-eating predators by attracting planktivorous fishes to feed on plankton aggregated by the fronts. However, few studies of plume fronts directly examine piscivorous predator response to plume fronts. Our work examined densities of piscivorous seabirds relative to the plume region and plume fronts of the Columbia River, USA. Common murres (Uria aalge) and sooty shearwaters (Puffinus griseus) composed 83% of all birds detected on mesoscale surveys of the Washington and Oregon coasts (June 2003–2006), and 91.3% of all birds detected on fine scale surveys of the plume region less than 40km from the river mouth (May 2003 and 2006). Mesoscale comparisons showed consistently more predators in the central plume area compared to the surrounding marine area (murres: 10.1–21.5 vs. 3.4–8.2birdskm−2; shearwaters: 24.2–75.1 vs. 11.8–25.9birdskm−2). Fine scale comparisons showed that murre density in 2003 and shearwater density in both 2003 and 2006 were significantly elevated in the tidal plume region composed of the most recently discharged river water. Murres tended to be more abundant on the north face of the plume. In May 2003, more murres and shearwaters were found within 3km of the front on any given transect, although maximum bird density was not necessarily found in the same location as the front itself. Predator density on a given transect was not correlated with frontal strength in either year. The high bird densities we observed associated with the tidal plume demonstrate that the turbid Columbia River plume does not necessarily provide fish with refuge from visual predators. Bird predation in the plume region may therefore impact early marine survival of Pacific salmon (Oncorhynchus spp.), which must migrate through the tidal plume and plume front to enter the ocean. Because murres and shearwaters eat primarily planktivorous fish such as the northern anchovy (Engraulis mordax), aggregation of these birds in the plume supports the hypothesis that it is the plume region as a whole, and not just the plume fronts, which enhances trophic transfer to piscivorous predators via planktivorous fishes.

We used a combination of seabird data (both fishery-dependent and fishery-independent) and fishing-effort data to evaluate the relative fisheries risk of five west coast groundfish fisheries and one shrimp fishery to black-footed (Phoebastria nigripes), short-tailed (P. albatrus) and Laysan albatrosses (P. immutabilis). To assess risk, an overlap index was derived as the product of total fishing effort and at-sea survey density of black-footed albatross. This index was used as the primary tool to estimate overlap with the endangered, relatively rare short-tailed albatross, which show similar habitat utilization from satellite telemetry tracks. Telemetry data indicate Laysan albatross primarily occur offshore beyond observed fishing effort. Black-footed and short-tailed albatross-fishery overlap was highest at the shelf-break (201-1000 m) north of 36 degrees N. Overlap and reported albatross mortality indicate that the sablefish (Anoplopoma fimbria) longline and Pacific hake (Merluccius productus) catcher-processor fisheries pose the greatest risk to these species; the near-shore rockfish (Seabastes spp.) longline, pink shrimp (Pandalus jordani) trawl, California halibut (Paralichthys californicus) trawl, and non-hake groundfish trawl fisheries pose relatively little risk. Implementing proven seabird bycatch-reduction measures will likely minimize albatross mortality in the highest-risk fishery, sablefish longline. (C) 2013 Elsevier B.V. All rights reserved.

We compared the performance of two bird-scaring line designs (light lines with short streamers vs. hybrid lines with a mix of long and short streamers) deployed in pairs with unweighted branch lines on two joint venture tuna vessels typical of distant-water tuna fisheries in the South Africa Exclusive Economic Zone (EEZ). We also added weight to a subset of branch lines, and compared the effects of line weighting and night vs. day setting in combination with bird-scaring lines on bird and fish catch rates. White-chinned petrels (Procellaria aequinoctialis), a diving seabird, dominated the seabird assemblage; they were the most numerous seabird and they attacked baited hooks and were killed at the highest rates. Secondary attacks - surface foraging albatrosses stealing baits from white-chinned petrels - drove albatross mortality. With hybrid scaring lines deployed, both diving and surface foraging seabirds made fewer attacks (1.5 and 2 times, respectively) within the lines' 100 m aerial extent, where hooks are closest to the surface and birds are most vulnerable to hooking, than with light lines. However, all metrics of comparison between hybrid and light lines were not statistically conclusive, primarily because birds could access baited hooks in areas beyond the protection afforded by bird-scaring lines (aft and to port of their aerial extent). Seabird bycatch rates were 4.6 times higher during daylight hours (2.00 birds/1000 hooks; 52 birds) than at night (0.439 birds/1000 hooks; 28 birds) and night catch rates near the full moon doubled. Bird catch rates were 18 times higher on unweighted branch lines (1.07/1000 hooks; 79 birds) than on weighted branch lines (0.06/1000 hooks; 1 bird) with no detectable effect on fish catch. With respect to streamer lines, our results suggest that in Procellaria petrel dominated systems the aerial extent of bird-scaring lines (of any design) should span the distance that baited hooks are within 10 m of the surface to effectively prevent bird attacks on baits. Overall, results suggest that night setting, adequate branch line weighting, and proper deployment of two bird-scaring lines have the potential to reduce seabird bycatch in pelagic longline fisheries without reducing the catch rates of target fishes. Further development of safe, fast-sinking branch line weighting configurations and bird-scaring lines less prone to tangling on fishing gear is critical to this effort. (C) 2013 Elsevier B.V. All rights reserved.

approved: Robert E. Gresswell Michael A. Banks Because human land use activities often result in increased fragmentation of headwater stream habitats, a better understanding of the effects of fragmentation on the genetic heterogeneity of stream salmonids is useful for effective management. We used eight microsatellites to examine the genetic structure of potamodromous coastal cutthroat trout (Oncorhynchus clarki clarki) in Camp Creek, an isolated headwater stream in western Oregon. Our objectives were to determine if coastal cutthroat trout were genetically structured at fine spatial scales and to assess the effects of natural and anthropogenic barriers on coastal cutthroat trout genetic variation. Fish sampling occurred at 10 locations, and allele frequencies differed significantly among all sampling sections. Dispersal barriers strongly influenced coastal cutthroat trout genetic structure and were associated with reduced genetic L Redacted for privacy diversity and increased genetic differentiation. Results indicate that Camp Creek coastal cutthroat trout exist as many small, partially independent populations connected by low to moderate levels of gene flow. In headwater streams, increased habitat fragmentation can result in genetic and demographic isolation leading to reduced coastal cutthroat trout genetic diversity and compromising long-term population persistence. ©Copyright by John E.B. Wofford December 17, 2003

Relationships among landscape structure, stochastic disturbance, and genetic diversity were assessed by examining interactions between watershed-scale environmental factors and genetic diversity of coastal cutthroat trout ( Oncorhynchus clarkii clarkii ) in 27 barrier-isolated watersheds from western Oregon, USA. Headwater populations of coastal cutthroat trout were genetically differentiated (mean FST= 0.33) using data from seven microsatellite loci (2232 individuals), but intrapopulation microsatellite genetic diversity (mean number of alleles per locus = 5, mean He= 0.60) was only moderate. Genetic diversity of coastal cutthroat trout was greater (P = 0.02) in the Coast Range ecoregion (mean alleles = 47) than in the Cascades ecoregion (mean alleles = 30), and differences coincided with indices of regional within-watershed complexity and connectivity. Furthermore, regional patterns of diversity evident from isolation-by-distance plots suggested that retention of within-population genetic diversity in the Coast Range ecoregion is higher than that in the Cascades, where genetic drift is the dominant factor influencing genetic patterns. Thus, it appears that physical landscape features have influenced genetic patterns in these populations isolated from short-term immigration.

In the northeastern Pacific Ocean, the North American population of southern resident Killer Whales (Orcinus orca, hereafter ‘‘SRKW’’) was listed as ‘‘Endangered’’ by the United States government effective 16 February 2006 (NMFS 2005a). There is significant scientific interest in filling data gaps regarding SRKW habitat use and ecology from November to April because few data are available about the winter distribution and feeding behavior of SRKW (Ford and Ellis 2006; Krahn and others 2004). From May to October, these whales occur primarily in US and Canadian waters of Juan de Fuca Strait, the Canadian Gulf Islands, the US San Juan Islands, and Georgia Strait (Fig. 1), where their diet appears to be dominated by adult Chinook Salmon (Oncorhynchus tshawytscha) (Ford and Ellis 2006; Ford and others 1998). After October, the southern residents expand their range to include Puget Sound as well as the outer coast of Vancouver Island, Washington, Oregon, and California (Krahn and others 2004). Between 1975 and 2004, there have been only 17 confirmed sightings of SRKW between November to April (Krahn and others 2004), and 12 of those winter sightings were from the outer coast. To our knowledge, this is the first published report that describes SRKW behavior at the mouth of the Columbia River and also positively identifies individual SRKW associated with that behavior (Krahn and others 2004; NMFS 2005b). We report confirmed sightings of at least 13 photo-identified SRKW individuals from L-pod, the largest of the 3 pods in the southern resident population (NMFS 2005b). The Columbia River mouth is approximately 245 km south of the entrance to the Strait of Juan de Fuca. The river forms the border between southern Washington and northern Oregon, USA, and it supports spring, summer, and fall runs of Chinook Salmon, with the fall run being the largest (Healey 1991). Beginning in August 2004, 2 of the authors (JEZ and TJG) conducted year-round, weekly or bi-weekly surveys of marine birds and mammals from a land-based observation site 6 km north of the Columbia River mouth. From dawn to dusk we counted birds and mammals within a specific 1.8 km2 area every half-hour during all daylight hours using a 20 spotting scope from an overlook at the North Head Lighthouse in Washington State (NAD 1983 UTM zone 10: Easting 417191.55, Northing 5127731.94; elevation 59 m above sea level). Additionally, we used 8 binoculars and Fujinon 25 ‘‘Big Eyes’’1 to aid species identification and to observe behavior between the half-hourly counts. These surveys documented variation in marine bird and mammal abundance relative to changes in tidal phase, time of day, and the strength and position of a visible boundary between river water and oceanic water. This boundary between newly discharged river water and oceanic water, hereafter called the ‘‘plume front’’, often manifests as a continuous, turbulent white foam line extending from river mouth offshore for up to 46 km (Morgan and others 2005). On 22 March 2005, the survey began at 06:45 local time. At 06:51, the observers (JEZ

Small stream systems are complex networks that form a physicochemical template governing the persistence of aquatic species such as coastal cutthroat trout Oncorhynchus clarkii clarkii. To gain new insight into these interactions, we initiated an integrated program of land- scape-scale sampling that is focused on fine- and broad-scale relationships among upslope landscape characteristics, physical stream habitat, and the spatial patterns of cutthroat trout abundance. Our sample of 40 catchments (500-1,000 ha) represented approximately 15% of the 269 barrier-isolated catchments in western Oregon that support populations of cutthroat trout. Because data were collected in a spatially contiguous manner throughout each catch- ment, it was possible to collect biological and geographic information necessary to assess the spatial structure of cutthroat trout abundance. Results underscore the influence of the physi- cal habitat template at a variety of spatial scales. For example, cutthroat trout move through- out the accessible portions of small streams. Some cutthroat trout congregate in areas of suitable habitat and form local populations that may exhibit unique genetic attributes. At times, some cutthroat trout move into larger downstream portions of the network where they may con- tribute to the genetic character of anadromous or local potamodromous assemblages. Results underscore the advantages of viewing habitats that are critical to the fitness and persistence of cutthroat trout populations as matrices of physical sites that are linked by movement. It is apparent that human activities that impede movement among suitable habitat patches can have unanticipated consequences for metapopulations of cutthroat trout and may ultimately affect their persistence.

Current land-management decisions that affect the persistence of native salmonids are often influenced by studies of individual sites that are selected based on judgment and convenience. Although this approach is useful for some purposes, extrapolating results to areas that were not sampled is statistically inappropriate because the sampling design is usually biased. Therefore, in recent investigations of coastal cutthroat trout (Oncorhynchus clarki clarki) located above natural barriers to anadromous salmonids, we used a methodology for extending the statistical scope of inference. The purpose of this paper is to apply geospatial tools to identify a population of watersheds and develop a probability-based sampling design for coastal cutthroat trout in western Oregon, USA. The population of mid-size watersheds (500-5800 ha) west