Identifying Kittlitz's Murrelet Nesting Habitat in North America at the Landscape Scale

The Kittlitz's murrelet Brachyramphus brevirostris, family Alcidae, is a small, diving seabird endemic to marine waters of Alaska and the Russian Far East where it spends its entire annual cycle (Day et al. 1999). During the summer breeding season, individuals occur in nearshore waters, where they feed primarily on small marine fish (Hatch 2011). During the past decade, the Kittlitz's murrelet has been the subject of conservation concern (U.S. Fish and Wildlife Service [USFWS] 2004, 2013; Butcher et al. 2007; BirdLife International 2014) as a result of evidence of population declines in some portions of its range (Kuletz et al. 2011a, 2011b; Piatt et al. 2011), as well as an apparent association with glacially influenced marine habitats that are changing rapidly (Kuletz et al. 2003; Arendt et al. 2009). Until recently, little information was available describing Kittlitz's murrelet natural history, particularly its breeding biology (Day et al. 1999).

Unlike most seabirds, Kittlitz's murrelets are not colonial breeders but instead nest solitarily, selecting sites that often are far inland from marine foraging areas (up to 74 km; Day et al. 1983; Gaston and Jones 1998). Nests typically are located on un- to sparsely vegetated talus slopes, barren ground, and cliffs and ledges in near-coastal uplands and mountains, where a single egg is laid in a nest scrape on open ground (Figure 1; Day et al. 1999). Both adults incubate at 24–48-h intervals for ∼30 d, followed by a 20–30-d chick-rearing period during which time the chick is left unattended while its parents forage between chick-meal deliveries. Owing to its secretive nesting habits and cryptic breeding plumage, presumably aimed at avoiding detection by predators (Figure 1), only 17 nests had been discovered in Alaska before 1999, and most were found accidentally (Day et al. 1999).

Since 2005, three key studies in Alaska have filled information gaps in our knowledge of the nesting biology of Kittlitz's murrelets. From 2006 to 2012, researchers located 95 nests by ground-searching foot-accessible areas on Agattu and Adak islands in the Aleutian Archipelago (Kaler et al. 2009; R.S.A. Kaler, USFWS, unpublished data). Using similar methods, researchers at Kodiak Island in south-central Alaska discovered 74 nests from 2008 to 2012 (Lawonn 2012; Corcoran et al. 2014). Finally, in Icy Bay, a heavily glaciated fjord system in southeastern Alaska, researchers found 35 nests from 2007 to 2012, primarily by using radiotelemetry to track murrelets captured at sea to inland nest sites (Kissling et al. 2015). Across these studies, 74% of monitored nests failed because of nest depredation, chick death (owing to starvation, exposure, or disease), or chick abandonment (USFWS 2013). Although the rates and causes of nest failure differed among study areas, overall reproductive output of Kittlitz's murrelets in these studies appeared alarmingly low, raising concern that this species may be experiencing reproductive problems, as proposed by Day and Nigro (2004). Although it is unknown if the ultimate cause of reproductive failure is associated with the marine or terrestrial habitats, or a combination of both, management actions that reduce or eliminate factors contributing to nest failure may benefit this species.

Although these recent studies (summarized in USFWS 2013) have shed new light on Kittlitz's murrelet nesting habitat in a few discrete areas, the location and extent of other possible nesting areas remains unclear. Our overall goal was to combine existing disparate data sets to identify potential Kittlitz's murrelet nesting habitat throughout its North American range to assist managers responsible for guiding conservation directives and to provide researchers with a meaningful starting point for discovering new nesting areas. Specifically, we 1) characterized Kittlitz's murrelet nesting habitat at the landscape scale; 2) defined criteria for delineating and identifying possible nesting habitat; and 3) mapped potentially suitable nesting habitat based on these criteria.

We delineated geographic boundaries to our study area based on the regular at-sea occurrence of Kittlitz's murrelets during the breeding season, which included nearly all of coastal Alaska (Figure 2). We did not consider the density of murrelets at sea as a factor in delineating boundaries, although we acknowledge that the greatest densities of this species occur in south-coastal Alaska (USFWS 2013). We excluded from consideration land east of 154°W (∼100 km east of Point Barrow) in northern Alaska because breeding Kittlitz's murrelets do not appear to use marine or terrestrial habitat along the Beaufort Sea coastline (Day et al. 2011). We did not include land in coastal British Columbia (south of 54.65°N and east of ∼130°W; Figure 2) because Kittlitz's murrelets are extremely rare south of Alaska (Carter et al. 2011). We considered all remaining land in Alaska and interior Canada within 100 km of the Alaska coastline in our analysis, following Day et al. (2011; Figure 2). We divided this area into four regions following Gallant et al. (1995): Northern Alaska (NOAK), Aleutian Islands (AI), Alaska Peninsula Mountains (APM, including Kodiak Island), and Pacific Coastal Mountains (PCM; Figure 2). These regions are ecologically distinct because of differing environmental factors such as topography, climate, and geology (Gallant et al. 1995).

We mapped potential nesting habitat of Kittlitz's murrelets with a three-step procedure. First, we compiled available nest records and evaluated them for authenticity. Second, we used the available nest records to define thresholds for selected physical and biological parameters that were used to distinguish habitat-capable from nonhabitat-capable lands for each parameter. Third, we mapped potential nesting habitat where all parameters were habitat-capable.

We compiled a database of all reported Kittlitz's murrelet nests (n = 249; Table A1) between 1904 and 2012. We then determined the veracity and locational accuracy of these nests by consulting original information sources. We categorized a nest as 1) “Confirmed” if it was observed directly and species identification was unequivocal, 2) “Probable” if it was located using aerial telemetry tracking of radiomarked Kittlitz's murrelets with at least two nonconsecutive inland locations (suggesting incubation exchange between adults), or 3) “Possible” if it was either a terrestrial site where only one aerial telemetry location was obtained or a nest record with uncertain species identification (due to confusion with the marbled murrelet Brachyramphus marmoratus, a congeneric species that sometimes nests in similar habitat; Barbaree et al. 2014; Kissling et al. 2015). Additionally, we classified each nest into one of four locational accuracy categories: very low (≥1 km), low (500 m–1 km), medium (100–500 m), and high (≤100 m; Table 1). We used only confirmed or probable nests with medium or high locational accuracy (n = 215) for analysis.

We considered three physical parameters (distance to coastline, elevation, and slope) and one biological parameter (land cover) as potentially important characteristics of nesting habitat for this species (Day et al. 1999, 2011). We used the National Elevation Dataset digital elevation model (DEM) to represent elevation in Alaska (60 m × 60 m horizontal resolution; Gesch et al. 2002) and a DEM for Canada from Canadian Digital Elevation Data (25 m × 25 m horizontal resolution; Natural Resources Canada 1996). We filled small gaps in Alaska National Elevation Dataset DEM coverage with Advanced Spaceborne Thermal Emission and Reflection Radiometer DEM data (ASTER, 30 m × 30 m horizontal resolution; Land Processes Distributed Active Archive Center 2006). We derived slope from this composite DEM (maximum elevation change with any adjacent cell) and generated a raster representation of distance to coastline from a 1:63,000 coastline shapefile for Alaska (Alaska Department of Natural Resources 1998). For land-cover classification, we used the National Land Cover Dataset (NLCD; 30 m × 30 m horizontal resolution) for Alaska (Homer et al. 2007) and the Earth Observation for the Sustainable Development of Forests (25 m × 25 m horizontal resolution; Natural Resources Canada 2008) for Canada. We preprocessed data to a common projection and resampled to match the spatial resolution of the coarsest data set (National Elevation Dataset DEM, 60 m) using ArcGIS (v10.1; ESRI, Redlands, CA).

We obtained values for distance to coastline, elevation, slope, and land-cover class for nest sites based on their spatial intersection with habitat parameter layers. Although written descriptions and field measurements of nest habitat parameters were available for many nests, we did not use them because that information did not exist for all records and it often was measured or estimated at inconsistent scales or at finer resolutions than could be represented by the Geographic Information System (GIS) data sets. Instead, we used the values extracted from the GIS habitat parameters to ensure that we produced a nesting habitat map at a common, range-wide scale and with standardized habitat measurements and classifications.

Regionally, we determined minimum and maximum values of the three physical parameters among nests to establish thresholds for each parameter. We then expanded the range of these values by proportionally fixed amounts in each region under the assumptions that 1) nests have been found near, but not at the extreme limit, of each parameter; and 2) those extreme limits vary regionally based on how local climate, topography, and geology interact to create potential habitat that is accessible to Kittlitz's murrelets within their energetic limits (e.g., distance from coastline). These adjusted thresholds also absorbed the effects of uncertainty due to potential nest location or habitat parameter inaccuracies. We deemed land within these adjusted thresholds as habitat-capable for that particular parameter and land outside the threshold range we designated as nonhabitat-capable (following Raphael et al. 2006, 2011).

For distance from coastline, we excluded land <200 m from shore in all regions because Brachyramphus murrelets probably avoid nesting immediately along the shoreline where predators are abundant (Albert and Schoen 2007). To set the maximum distance from coastline threshold in each region, we identified the distance of the nest farthest from shore and added 30%. We chose an adjustment value of 30% in all regions to allow for consistency with Day et al. (2011) in NOAK. We did not set a maximum distance to coastline threshold in AI because the maximum possible value was too small (<20 km). For elevation and slope thresholds, we determined the minimum nest elevation and slope of nests in our database in each region and reduced each parameter by 15%; this value was arbitrary but seemed reasonable given that many nests in our data set were subject to sampling bias. For similar reasons, we did not set upper limits to elevation or slope in any region. Kittlitz's murrelets will nest on cliff ledges surrounded by terrain that is inaccessible by foot and at elevations up to 2,555 m (Kissling et al. 2015). We assumed that high elevations in steep alpine areas have not been traversed as much as low elevations in flatter terrain, so it was less likely for nests to have been found opportunistically in the former areas. Although Kittlitz's murrelets undoubtedly encounter physiological limitations that prevent them from nesting above a certain altitude, those limits are unknown. Further, the amount of snow- and ice-free land at the highest elevations was minimal (<0.1% above 2,555 m), so our liberal approach to setting thresholds for elevation and slope probably had little influence on our results.

We used land-cover class as a biological parameter to designate habitat-capable land for Kittlitz's murrelets in all regions. We identified the NLCD land-cover class value at each nest site and then compared them to nest site descriptions to evaluate whether they were biologically valid or could be erroneous on account of potential nest-site locational inaccuracy or NLCD error. We excluded potentially erroneous land-cover classes (e.g., ice–snow).

In each region, we identified potential Kittlitz's murrelet nesting habitat as lands within the habitat-capable bounds of all four parameters combined. To determine regional variation, we calculated the total land area (km²) of potential habitat in each region and the proportion of that land by land-cover class considered habitat-capable. We then joined regional maps to produce a final map of potential nesting habitat for Kittlitz's murrelets in North America.

Across all regions, the maximum distance between a Kittlitz's murrelet nest and the coastline was 73.5 km, the minimum nest elevation was 128 m, and the minimum slope was 0° (n = 215 nests; Table 2). After adjusting thresholds to account for uncertainty, the maximum distance from coastline for habitat-capable land ranged from 36.9 km in PCM to 95.5 km in NOAK (Table 2). The minimum habitat-capable elevation was lowest in NOAK (109 m) and highest in AI (156 m); the minimum habitat-capable slope ranged from 0° in PCM to 5.6° in NOAK (Table 2).

Almost all nests (90%) were found on three land-cover classes: barren (45% of nests), dwarf-shrub (33%), and grassland–herbaceous (12%). The remaining 10% of nests were located on perennial snow–ice (n = 19) or shrub–scrub (n = 2). All nests within the perennial snow–ice land-cover class were located using aerial telemetry in the PCM region, were located adjacent to steep, barren habitat dominated by cliffs, and fell within the medium-level accuracy category. Hence, we assumed that these nests likely were located on barren habitat and did not consider perennial snow–ice to be habitat-capable. It also is possible that some nests in this land-cover class were discovered in years of less snow cover than the remote sensing data used to create the NLCD land-cover product, suggesting that these nest sites may only be available in some years. Similarly, we assumed that shrub–scrub was not habitat-capable because the written habitat descriptions for both nests located in this land cover were not consistent with GIS-derived data. Additionally, both nests were adjacent to one of the predominant three land-cover classes, indicating small-scale inaccuracies in the NLCD land-cover or nest locations. Therefore, we ultimately defined only barren, dwarf-shrub, and grassland–herbaceous land-cover classes as habitat-capable.

We classified 12% (70,411 km²) of the land within our study area as potential nesting habitat (Table 3), with dense, but discrete, patches located in NOAK and a more uninterrupted, narrow band extending across the coastal mountainous areas in AI, APM, and PCM (Figure 3; Data A1). Regionally, the proportion of potential habitat was greatest in AI (39%) and least in NOAK (9%; Table 3). Potential nesting habitat was present along much of the Pacific coast of Alaska (PCM, APM, and AI regions), although notable gaps existed in extreme southeastern Alaska (primarily on islands of the Alexander Archipelago) and in lowlands around upper Cook Inlet (Figure 3). Larger, higher elevation islands in AI tended to have more potential habitat (e.g., Adak Island) than smaller and lower islands. Along the Bering and Chukchi Sea coasts (NOAK), potential nesting habitat was present in the western Brooks Range, on the Seward Peninsula, east of Norton Sound in the Nulato Hills, and, to a lesser extent, in the Ahklun–Kilbuck Mountains between the Yukon–Kuskokwim Delta and Bristol Bay (Figure 3). Large habitat gaps occurred in the extensive lowlands near Bristol Bay, the Yukon–Kuskokwim Delta, Kotzebue Sound, and north of the Brooks Range on the Arctic Coastal Plain (Figure 3).

We found considerable regional variation in the proportional extent of habitat-capable land-cover classes (Figure 4). The “barren” land class was most widespread in PCM (80%), decreased westward through APM (43%) and AI (31%), and was least common in NOAK (14%). This pattern was opposite for the “dwarf-shrub” land class, which was most extensive in NOAK (86%) and least common in PCM (19%). “Grassland–herbaceous” habitat was only common in APM (9%) and AI (27%).

We provide the first comprehensive assessment of potential nesting habitat of the Kittlitz's murrelet for North America (Figure 4). Our criteria identified 12% (70,411 km²) of all land within regionally defined distance to coastline thresholds as potentially suitable for nesting, with most located in NOAK and PCM (72%; Table 3). Our classification scheme provides a tool to help resource managers and policy makers in North America make informed decisions for the management of potentially important Kittlitz's murrelet nesting habitat. In addition, our classification of nesting habitat provides a baseline to evaluate future shifts in the breeding range of this species (e.g., Raphael et al. 2011).

Our results indicate that the Kittlitz's murrelet uses a variety of habitats depending on availability. For example, Kittlitz's murrelets may nest farther inland in NOAK (≤96 km; Table 2) because the nearshore topography is more subdued and maximum elevations (≤1,450 m) are lower and farther from shore than in PCM (≤37 km; Table 2), where the Chugach–St. Elias Mountains (maximum elevation 5,489 m) present a significant nearshore topographic barrier limiting inland travel of murrelets to nest sites that are comparatively closer to the coast. Indeed, nesting Kittlitz's murrelets presumably would preferentially nest closer to the coastline to reduce flight energetic costs, especially during chick-rearing (Hatch 2011). The maximum distance flown inland by Kittlitz's murrelets, therefore, may be negatively correlated with the amount of suitable habitat located near the coast. Additionally, constant physical disturbance by extensive, active glaciation in PCM provides substantial barren and sparsely vegetated habitat at very low slopes (0° threshold), whereas low-slope areas in NOAK (5.6° threshold; Table 3) are more likely to be covered by wet, vegetated tundra. These terrestrial factors may explain in part why Kittlitz's murrelets reach their greatest densities at sea in PCM compared with other regions (Day et al. 2011; Kissling et al. 2011; Madison et al. 2011).

Kittlitz's murrelet nesting habitat generally follows the near-coastal (<100 km inland) distribution of dry and alpine “upland tundra” vegetation zones in Alaska (Viereck et al. 1992). These zones manifest as Dryas spp. dwarf-shrub tundra on exposed ridges and rocky sites in northern and western Alaska (NOAK); Dryas and ericaceous dwarf-shrub tundra above treeline in mountainous regions of south-central Alaska (eastern APM and western PCM); and Empetrum spp. heath, ericaceous dwarf-shrub, and mesic forb herbaceous vegetation in AI and, to a lesser extent, APM (Viereck et al. 1992). Barren habitat consisting of little to no vegetation is associated with these upland tundra zones across their range; however, it is most widespread in PCM where higher elevations, steep topography, extensive glaciation, and concomitant erosional processes prevent significant soil development (Viereck et al. 1992; Gallant et al. 1995). The regional proportion of potential habitat in each habitat-capable land-cover class varied similarly (Figure 4), as did field measurements of vegetation type and cover at extant nest sites (Kaler et al. 2009; Day et al. 2011; Lawonn 2012; M.L. Kissling, USFWS, unpublished data). The regional variability in how habitat parameters manifest as suitable habitat, as well as the species' apparent high level of behavioral plasticity in regard to where breeding birds will go to locate nesting habitat, likely contributes to the broad, but irregular, distribution of Kittlitz's murrelets at sea during the breeding season (USFWS 2013).

At the nest-site scale, field studies have shown that Kittlitz's murrelets tend to nest among the least-vegetated areas available locally (Kaler et al. 2009; M.L, Kissling, unpublished data) and sometimes avoid nesting near vegetated edges, perhaps to avoid predators (Lawonn 2012). The amount of vegetation cover measured within a 25-m radius of nest sites ranged from 0 to 75% (Kaler et al. 2009; Day et al. 2011; Lawonn 2012; R.S.A. Kaler, unpublished data; M.L. Kissling, unpublished data). At the landscape scale, grid cells in the NLCD product needed only to have >15% vegetation cover to be defined as a vegetated land-cover class and realistically also contained barren or sparsely vegetated patches that were lost at the coarser resolution of these data. Therefore, we likely overestimated potentially suitable habitat by including these vegetated land-cover classes. Our map product, however, provides a general representation of the spectrum of potential nesting habitat throughout the majority of this species' range.

Although we were not able to ground-truth the final product, we present three points that substantiate our regional habitat assessment. First, the known at-sea distribution of Kittlitz's murrelets during the breeding season generally mirrors the distribution of potential nesting habitat we identified (see USFWS 2013 for summary of at-sea distribution). Second, our results in northern Alaska are similar to those of Day et al. (2011), who mapped nesting habitat for this species by using only land-cover classes and elevation; the addition of slope (as recommended in that study) restricted the extent of potential habitat in our study, but the two efforts generally produced similar results. Third, the maximum annual nest density recorded in suitable, well-searched terrain is 0.118 nests/km² (Kodiak Island, APM; Lawonn 2012), which produced a breeding population estimate of ∼8,200 pairs (16,600 individuals) when applied uniformly to the areal extent of potential habitat we identified. Including the number of nonbreeders (potentially 80% of the population; Kissling et al. 2015) and a Russian population of ≥1,000 murrelets (Artukhin et al. 2011) could increase this number to be similar to current range-wide population estimates of ∼25,000 to ∼42,000 individuals (USFWS 2013), generally supporting our mapping criteria. Additionally, the calculation of true surface area (instead of planimetric area, as used here) would increase the total area of potential nesting habitat, most of which is sloping, resulting in a greater estimate of nesting murrelets. Of course, it is unlikely that Kittlitz's murrelets nest at similar densities across their entire range or use all habitat mapped in this study; the extent of suitable nesting habitat may drive regional distribution, whereas favorable at-sea conditions drive local abundance, or vice versa (Arimitsu et al. 2012; Raphael et al. 2014).

We did not calculate the probability associated with potential nesting habitat by quantitatively modeling the distribution of habitat parameters at extant nest sites because of spatial sampling bias (Raphael et al. 2011). Nests with reasonable location accuracy that were discovered accidently were few (n < 20), and, of the three focused nesting studies, those in AI (Kaler et al. 2009) and APM (Lawonn 2012) were biased toward restricted, foot-accessible areas that do not represent the full range of habitat parameters available in the regional landscape (Figure 3). In comparison, nest sites located using aerial telemetry around Icy Bay were sampled across the landscape without spatial biases. Those nests present the best opportunity for quantitative modeling of nest habitat suitability at the landscape scale, albeit only for PCM because of interregional variation in the availability of habitat parameters (e.g., land cover; Figure 4).

The framework and map presented herein provides the first management tool for identifying potential Kittlitz's murrelet nesting habitat across the species' North American range. We used an approach that isolated common regional and range-wide characteristics for four habitat parameters, with an emphasis on using extant nesting information within its limit of inference. Although we probably have overrepresented the true extent of nesting habitat with our criteria, the final map product can help focus future survey and research efforts to locate potentially important areas for conservation that can be refined further by ground-truthing. For example, the use of guano-facilitated vegetation growth at previously occupied nests has recently been used to locate active murrelet nests in the Aleutian Islands (Kenney and Kaler 2013); vegetation cues such as this could provide an important fine-scale tool in areas of potential nesting habitat identified herein. Our results suggest further analyses coupling at-sea distribution data with our nesting habitat map would assist investigations regarding relative importance of other factors—such as glaciation, habitat patchiness, or distance to vegetated edge—that may influence the distribution and abundance of Kittlitz's murrelet nesting habitat at the landscape scale (Kuletz et al. 2003; Lawonn 2012). Future nest surveys that employ nonbiased, spatially random, or stratified-random sampling designs (e.g., telemetry, point-, or line-transect designs) would allow for quantitative modeling of nesting habitat and better extrapolation to areas where nests have not been located. In particular, very few nests with reliable locational accuracy have been discovered in the largest region in this study (NOAK; n = 7). Focused research in NOAK (as well as AI and APM) that evaluates the relative importance of the barren and vegetated land-cover classes to nesting Kittlitz's murrelets could help to refine further the overall extent of potential habitat estimated in our study.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Reference S1. Corcoran RM, Mackey HL, Piatt JF, Pyle WH. 2014. Breeding ecology and behavior of Kittlitz's murrelet in Kodiak National Wildlife Refuge, Alaska: 2012 progress report. Kodiak, Alaska: Kodiak National Wildlife Refuge. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S1. Also available at: ehttp://www.fws.gov/uploadedFiles/Region_7/NWRS/Zone_2/Kodiak/PDF/Report2014.1_Kittlitz'sMurreletNestingEcology2012ProgressReport_KodiakNWR.pdf (2053 KB PDF).

Reference S2. Gallant AL, Binnian EF, Omernik JM, Shasby MB. 1995. Ecoregions of Alaska. Denver, Colorado: U.S. Geological Survey. Professional Paper 1567. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S2; also available at: https://pubs.er.usgs.gov/publication/pp1567 (16273 KB PDF).

Reference S3. Hatch NR. 2011. Foraging ecology and reproductive energetics of the Kittlitz's murrelet (Brachyramphus brevirostris) in Southeast Alaska. Master's thesis. Corvallis: Oregon State University. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S3; also available at: https://ir.library.oregonstate.edu/xmlui/handle/1957/27862 (1671 KB PDF).

Reference S4. Lawonn MJ. 2012. Breeding ecology and nest site selection of Kittlitz's murrelets on Kodiak Island, Alaska. Master's thesis. Corvallis: Oregon State University. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S4; also available at: https://ir.library.oregonstate.edu/xmlui/handle/1957/36245 (1974 KB PDF).

Reference S5. Raphael MG, Falxa GA, Dugger KM, Galleher BM, Lynch D, Miller SL, Nelson SK, Young RD. 2011. Northwest Forest Plan—the first 15 years (1994–2008): status and trend of nesting habitat for the marbled murrelet. Portland, Oregon: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-848. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S5; also available at: http://www.treesearch.fs.fed.us/pubs/38494 (2602 KB PDF).

Reference S6. Raphael MG, Galleher B, Huff MH, Miller SL, Nelson SK, Young RD. 2006. Spatially-explicit estimates of potential nesting habitat for the marbled murrelet. Pages 97–146 in Huff MH, Raphael MG, Miller SL, Nelson SK, Baldwin J, technical coordinators. Northwest Forest Plan—the first 10 years (1994–2003): status and trends of populations and nesting habitat for the marbled murrelet. Portland, Oregon: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-650. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S6; also available at: http://www.treesearch.fs.fed.us/pubs/23161 (3976 KB PDF).

Reference S7. Viereck LA, Dyrness CT, Batten AR, Wenzlick KJ. 1992. The Alaska vegetation classification. Portland, Oregon: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station. General Technical Report PNW-GTR-286. Found at DOI: http://dx.doi.org/10.3996/112015-JFWM-116.S7; also available at: http://www.treesearch.fs.fed.us/pubs/6941 (2602 KB PDF).

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any archived material. Queries should be directed to the corresponding author for the article.

To cite this archived material, please cite both the journal article (formatting found in the Abstract section of this article) and the following recommended format for the archived material.

Data A1. Geographic Information System (GIS) raster file of potential nesting habitat for the Kittlitz's murrelet Brachyramphus brevirostris in North America. Found at DOI: http://dx.doi.org/10.5066/F71V5C2W

Table A1. Locations, veracities, and locational accuracies of all Kittlitz's Murrelet Brachyramphus brevirostris nest sites reported through 2012. Found at DOI: http://dx.doi.org/10.5066/F71V5C2W

We are grateful to Robert H. Day, Robin Corcoran, John Piatt, and Tom Van Pelt for providing feedback, data, and assistance updating obscure nest information from long ago. Martin Raphael provided initial insight into how best to proceed with these data. We thank Josh Adams for mentoring support and Ryan Carle for review of an earlier version of the manuscript. This project received funding support from U.S. Fish and Wildlife Service and U.S. Geological Survey Western Ecological Research Center.

Any use of trade, product, website, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.