Flight Activity (FA)

A bird risks collision with OWEI while in flight and the amount of time a bird spends in flight can vary greatly among species. We quantified the overall percentage of time flying (Flight Activity, FA) for the 89 species in the POCS. We used information from the Birds of the World accounts (Billerman and others, 2022; https://birdsoftheworld.org/bow/home), Vulnerable Index Version 1 of our marine bird Vulnerability Index, at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017; Adams and others, 2020), and tracking data found during literature searches. In literature searches, we searched by species group (for example, loons), in association with each of the following phrases as standardized search terms to find all relevant sources:

We used boat-based and aerial seabird observations within the POCS (Adams and others, 2016; Takekawa and others, 2017) and the North Pacific more broadly (Drew and others, 2005) to quantify DFA (because surveys were performed during the day). We combined the two regional POCS aerial survey datasets to spatially represent most of the POCS: (1) off southern California (Takekawa and others, 2017) and (2) off northern California, Oregon, and Washington (Adams and others, 2016). In all at-sea datasets, birds were recorded as either flying or on the water. For each species (or species group) with greater than 100 sightings in the combined databases, we calculated the ratio of flying individuals to all individuals recorded to estimate the DFA. We did not use these data for species known to spend a considerable percentage of their time on land (for example, cormorants and terns) because the ratio of birds that were flying versus on the water is biased by the number of uncounted birds that were on land at the time of the surveys.

In Version 1, we quantified NFA and DFA values separately (eqs. 1, 2). In Version 2, we retained records of NFA and DFA values when available, and then averaged them for a final value, creating the metric FA_N&D (Nocturnal and Diurnal Flight Activity; eq. 3). However, we also included studies that quantified FA but did not differentiate between nocturnal and diurnal periods; in these cases, we recorded the median of all published 24-hour FA (FA24) values from a literature source (eq. 3).

Our method for determining the final FA value used in equation 3 considering FA_N&D and FA24 values necessitated an approach that varied among species and was dependent on which data inputs were available for a given species. One FA variable is the proportion of daylight, and thus the proportion of time spent flying during day and night, which can vary with season and latitude for a given species. We prioritized FA estimates that were most representative of the latitudes of the POCS and the seasons that a given species spends within the POCS. In some cases, when information on NFA and DFA was limited, we assumed equal periods (12 hours) of the day (light) and the night (dark) for a 24-hour day when determining NFA and DFA values. As an example of how FA was generated, consider FA for the storm-petrel species in the POCS below. Species-specific FA values were available for two POCS storm-petrels:

For all other storm-petrel species in the POCS, we used the above values for Leach’s and Townsend’s Storm-Petrels, in addition to published values from other similar species:

We averaged FA24 and FA_N&D values, when available, to generate FA for all other Storm-petrel species—FA=mean(FA_N&D, FA24)=mean(0.64, 0.61)=0.62.

Time Spent in Rotor Swept Zone

Birds are vulnerable to collision with OWEI turbine blades when flying within the rotor sweep zone. We collated estimates of RSZt for all species, or from similar species where species-specific estimates were lacking, and reported the median RSZt for each species. In Version 1, we defined the rotor swept zone at 10–200 meters (m). We chose the lower rotor swept zone limit of 10 m to be consistent with Ainley and others (2015). For our Version 2 update, we revisited the latest rotor swept zone height estimates for turbines proposed for the POCS. Due to the bathymetry of the POCS, all OWEI in this region will be built on floating, anchored platforms. One of the few existing floating offshore wind turbines as of 2023, Hywind Tampen, has a hub height of 105 m and a rotor diameter of 167 m, resulting in a 21–189-m rotor swept zone (ABSG Consulting, Inc., 2021). This rotor swept zone height is consistent with most marine bird RSZt studies, which estimate the amount of time birds of a given species spend between 20- and 200-m rotor swept zone height. Therefore, in Version 2, we defined the rotor swept zone as 20–200 m above the ocean to accommodate the most accurate information on RSZt.

Consistent with the other Vulnerability Index metrics, we used ordinal RSZt values in Version 1. Furthermore, due to the limited amount of flight height data available, we binned RSZt estimates into three large ranges in Version 1: (1) 5=greater than 20 percent, (2) 3=5–20 percent, and (3) 1=less than 5 percent. Since publishing Version 1, the number of studies quantifying marine bird flight heights has greatly increased, especially in the context of rotor swept zone height. Through a combination of Global Positioning System (GPS) tracking, range finders, and coupling radar and camera data, flight height estimations are becoming more accurate for a greater number of species. Therefore, in Version 2, RSZt values represent the estimated percentage of time spent flying within the 20–200-m rotor swept zone height for a given species (as a value between 0 and 1, or for a similar species when data for a given species were not available).

Macro-Avoidance (MA)

A bird’s likelihood of avoiding OWEI influences its potential CV (higher likelihood of avoidance leads to lower CV) and DV (higher likelihood of avoidance leads to higher DV). We recognize three broad types of avoidance behavior: (1) Macro-Avoidance (MA; change in flight course to avoid entering a wind farm area), (2) meso-avoidance (change in flight direction within a wind farm area to avoid wind turbines), and (3) micro-avoidance (last-minute flight movements to avoid a specific turbine; Cook and others, 2014). In our Vulnerability Index, we focus on MA, which affects CV and DV directly (albeit with the opposite effect). Cook and others (2014) highlight that gulls, in particular, display avoidance behavior but at a micro-avoidance level (for example, a Herring Gull is estimated to have an MA rate of 0 but a micro-avoidance rate of 0.99; Cook and others, 2014). Accordingly, we did not include values from studies that reported overall OWEI avoidance rates because these studies included micro-avoidance rates.

We define MA as the proportion of birds displaced from OWEI areas that would have been expected to be there otherwise (for example, MA=0.10 represents a decrease of 10 percent in species abundance compared to baseline abundance numbers; Cook and others, 2014). We retained existing MA data used in Version 1 of the Vulnerability Index and data from MA studies published since Version 1. Macro-Avoidance has been a major focus of post-construction studies at existing OWEI sites, and there were considerable new data available to incorporate. However, data are still deficient for many POCS species (for example, albatrosses and pelicans). When data were not available for POCS species, we estimated MA based on MA measured at OWEI for similar species.

In Version 1, MA was given inverse ordinal values for CV versus DV because MA has an inverse effect when considering the two types of Vulnerability (Adams and others, 2017; Kelsey and others, 2018). Herein, with revised, continuous MA values (0–1) derived from the literature, we quantify MA the same way for CV and DV (low value is low MA, and high value is high MA). For CV, the focus is the birds that are not macro-avoiding OWEI, and thus, we used the inverse (1–MA) in our revised CV equation (eq. 3).

Habitat Specificity (HS)

Habitat Specificity is a function of how diverse a species’ needs are (Foraging Specificity [FS]) and the likelihood that, if an individual of that species were displaced by OWEI, they would be able to find an unaffected available habitat elsewhere (Spatial Specificity [SS]). In Version 1, we called this metric HF; however, we have since realized that we were quantifying inflexibility (in other words, specificity), not flexibility per se. Furthermore, Version 1 used subjective and qualitative ordinal rankings, as shown below:

Thus, to update this metric to be more transparent and quantitative, we revised our methods and renamed the metric HS. We considered two components of the ordinal HF definitions of species’ habitat for quantification: (1) a spatial component (in other words, ranging from foraging habitats over a large area to more spatially constrained, habitat-specific requirements) and (2) a foraging component (in other words, ranging from a generalist/opportunistic forager to a species with narrow, prey-specific requirements). We determined that the spatial and foraging components could be quantified separately as SS and FS and then recombined to calculate HS.

We quantified SS using species distribution models developed by Leirness and others (2021) for the POCS. Specifically, for a given species, we quantified the skewness of the distribution of predicted density values across the POCS as a measure of how broad and dispersed (low skew of distribution of density values) or isolated and aggregated (high skew of distribution of density values) a species’ estimated distribution was. Figure 2 provides an example of skewness for two species in the POCS in the summer season. The estimated summer distribution of Pink-footed Shearwater is dispersed throughout the POCS and thus has a low skewness value. In contrast, Double-crested Cormorant (Nannopterum auritum) has a more aggregated distribution, only found in a few coastal locations along the POCS, resulting in a high skewness value (fig. 2). In this Vulnerability Index, 56 of the 89 species (63 percent) had enough spatial data for Leirness and others (2021) to generate seasonal species distribution models like those shown in figure 2. We calculated skewness for each of the 56 species in Leirness and others (2021), which represented the species’ SS, by taking the weighted average of all seasonal layers. We used the sum of predicted density values for each season and gave more weight to seasons when more birds were predicted/present. We quantified the skewness of the combined, seasonally-weighted layer for each species. Skewness values were then natural-log transformed and rescaled (0–1) by dividing by the maximum species-specific SS value. For the remaining 33 species without skewness values, we generated the best estimate of SS based on skewness values for similar species modeled by Leirness and others (2021), expert opinion, and by referencing the species’ distributions mapped in eBird data (eBird, 2021).

We defined FS as a function of a species’ Foraging Mode (FM; for example, deep diving, surface dipping, and plunge diving) and Prey Size/Trophic Level (ST; table 5). For a given species, we assigned one point for each unique FM used and ST in their diet (table 5); these final FM and ST scores were rescaled so that values were equally weighted across FM and ST, respectively. We averaged FM and ST and subtracted the average from one to get the final FS value for each species. We subtracted from one so that a high value would reflect high specificity, and a low value would reflect low specificity. For each species, final SS and FS values were averaged to get a final HF value between 0 and 1. Below, we provide an example using the same two species used to describe FS.

Pink-footed Shearwaters forage via shallow diving and surface seizing/aerial pursuit, gaining them two FM points. Double-crested Cormorants, which forage via deep diving and shallow diving, also got two FM points (table 5). Across all species, the maximum number of FM points was 4. Thus, the rescaled FM score for Pink-footed Shearwaters and Double-crested Cormorants was 2 out of 4, or 0.50. For ST, Pink-footed Shearwaters feed on small forage fishes and squids representing two prey sources, giving them an ST score of 2. In comparison, Double-crested Cormorants forage on small- and medium-sized forage fishes and invertebrates, gaining them an ST score of 3. Across all species, the maximum number of ST points was 4. Thus, Pink-footed Shearwaters got a rescaled score of 2 out of 4, or 0.50, and Double-crested Cormorants got a score of 3 out of 4, or 0.75. The final FS score was calculated by taking 1 minus the mean FM and ST scores and rescaling all scores from 0 to 1, resulting in a score of 0.67 for Pink-footed Shearwater and 0.5 for Double-crested Cormorant.

Values from the Actual Versus Similar Species

We assumed there was greater uncertainty associated with species for which no metric values were available in the literature, and thus, values were derived from similar species. To quantify the uncertainty associated with the difference between data from actual or similar species being used, we assigned uncertainty values between 0 and 1. Given that we used values from the most closely associated species when species-specific values weren’t available, the uncertainty values were quantified as follows:

Number of Literature Sources

We assumed that there was greater uncertainty associated with metric values derived from only one literature source, as opposed to metric values for which data from multiple literature sources were available. For each species, we quantified the number of sources (from the species in question or from similar species when data were lacking) used to determine the metric value. We tried to avoid referencing review papers; however, review papers were included in this count when the original source was unavailable. We found that most metric values were derived from five or fewer sources. Thus, we determined any metric value with six or more sources to be a high number of sources. When quantifying the number of sources used, we capped the number of sources at six (for example, any metric value with more than six sources used was given a source value of six). We then rescaled this value from 0 to 1 (with low numbers representing low uncertainty) by dividing the source’s value by 6 and subtracting that value from 1.

Range of Published Values

When multiple metric values were reported for a given species, we calculated the median of those values as our final metric value. We assumed that there was greater uncertainty associated with values derived from a wider range of published values, as opposed to metrics where all published values were close to one another; to quantify this uncertainty, we took the difference in the range of values for each species’ metric value (the difference between the maximum and minimum values reported). Multiple value ranges were used in the metric FA (value ranges for NFA, DFA, and FA24). In this case, we used the difference in values from the largest value range in our uncertainty quantification. Because HS values were generated from discrete components, there was no variance around the HS value, and thus, we were not able to quantify uncertainty for that metric.

Flight Activity

Flight Activity calculations varied slightly among species groups, based on data availability and species-specific behavioral differences. For many species in the POCS, we used DFA values generated from the ratio of birds seen flying and resting on the water during at-sea surveys. These surveys do not account for birds on land, which is especially relevant for breeding birds attending nest sites on land. We acknowledge this limitation when using at-sea survey data. However, the DFA values generated from at-sea survey data were the best available information for many species in the POCS. Some species also are known to roost on land during the night (for example, pelicans, cormorants, gulls, and terns). When no species-specific data were available for a night-roosting species, we used an NFA of 0.001. Below, we present the variations in FA calculations and the resulting FA values among species groups.

Brant (Branta bernicla) migrates into the POCS and overwinters in bays and estuaries along the western coast of the United States. We used at-sea survey observations of Brant (Drew and others, 2005) to generate an estimated FA of 0.69 for the POCS. Although at-sea surveys take place during the day (thus providing a DFA value), Brant are thought to fly throughout the day and night while migrating; thus, we assumed NFA equaled the DFA value calculated from at-sea survey data (Lewis and others, 2020). Although data exist for Brant FA at wintering areas, we did not incorporate those data into FA because Brant are largely resident during winter and confined within the bays and coastal areas; therefore, overwintering values were not related to FA within the POCS.

Sea ducks found in the POCS (Common Merganser [Mergus merganser]; Red-breasted Merganser [Mergus serrator]; Harlequin Duck [Histrionicus histrionicus]; and Long-tailed Duck [Clangula hyemalis]) all have an NFA of 0.001 based on species accounts that reported they do not fly at night. We derived DFA for these species from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). The resulting FA values were relatively low and ranged from 0.08 for Long-tailed Duck to 0.11 for Red-breasted Merganser.

Based on species accounts, scoters are thought to fly the same amount during the night as during the day while migrating through the POCS. Thus, we used values generated from daytime at-sea survey data for scoter DFA and NFA (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). The resulting FA values ranged from 0.07 for Surf Scoter (Melanitta perspicillata) to 0.10 for White-winged Scoter (Melanitta deglandi).

Grebes were seen in the POCS during the spring and fall migration periods, with large numbers wintering in nearshore waters (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017; D. Pereksta, written commun., Bureau of Ocean Energy Management, 2025). Grebes fly mostly at night, with an estimated NFA of 0.15 (J. Felis, oral commun., U.S. Geological Survey, 2023). Grebe DFA was generated from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017) and combined with the estimated NFA for the resulting FA value ranging from 0.08 (Western Grebe [Aechmophorus occidentalis] and Clark’s Grebe [Aechmophorus clarkii]) to 0.09 (Horned Grebe [Podiceps auritus], Red-necked Grebe [Podiceps grisegena], and Eared Grebe [Podiceps nigricollis]).

We estimated that phalaropes fly half as much during the night as they do during the day (R. Carle, Oikonos Ecosystem Knowledge, written commun., 2024). We derived DFA values for Red-necked Phalarope (Phalaropus lobatus) and Red Phalarope (Phalaropus fulicarius) from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017) and halved that value to generate their respective NFA values. The FA24 was estimated for Red-necked Phalaropes by van Bemmelen and others (2023). Considering FA_N&D and FA24, FA values for Red-necked and Red Phalaropes were 0.50 and 0.52, respectively.

Jaegers migrate through the POCS from breeding areas in the boreal and arctic north, with the majority of individuals overwintering areas south of the POCS (D. Pereksta, written commun., Bureau of Ocean Energy Management, 2025). South Polar Skuas (Stercorarius maccormicki) breed in the southern hemisphere and reach the POCS in relatively small numbers during their overwintering period (Furness and others, 2020). We generated DFA values for these birds from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). We estimated an NFA value of 0.05. There are accounts of jaegers and skuas observed flying infrequently at night. We used the median of spring and fall migration FA values for South Polar Skuas and Long-tailed Jaegers (Stercorarius longicaudus; Bonnet-Lebrun and others, 2021) to generate FA24. Considering FA_N&D and FA24, FA values of jaegers and skuas ranged from 0.35 (South Polar Skua) to 0.43 (Long-tailed Jaeger).

We derived DFA for alcids from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017), and we derived NFA values from tracking studies of similar species. Most historical at-sea surveys reported Xantus’s Murrelet (Synthliborampus hypoleucus), which is now recognized as two separate species: Scripps’s Murrelet (Synthliboramphus scrippsi) and Guadalupe Murrelet (Birt and others, 2012; Chesser and others, 2012). Both species are very similar in appearance to Craveri’s Murrelet (Snythliboramphus craveri) and difficult to identify during surveys; therefore, we use DFA values from survey records for Xantus’s Murrelet, Scripps’s, Guadalupe, and Craveri’s Murrelet combined. The resulting FA values for alcids ranged from 0.04 (Common Murre [Uria aalge] and Rhinoceros Auklet [Cerorhinca monocerata]) to 0.17 (Ancient Murrelet [Synthliboramphus antiquus]).

For gulls, we found published FA24 values from tracking studies for Black-legged Kittiwakes (Rissa tridactyla), Western Gull (Larus occidentalis), and Herring Gull (Larus argentatus) and from these, we derived FA values of 0.31, 0.11, and 0.15, respectively. For all other gull species in the POCS, we used the median of values from the published FA24 values of similar gull species (Black-legged Kittiwakes, Western Gull, European Herring Gull, and two other species not found in the POCS) for a final FA value of 0.27.

For tern species, available FA values were obtained from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017) and a tracking study of Caspian Terns (Hydroprogne caspia; Rueda-Uribe and others, 2022). The survey-derived DFA values are biased upwards because the ratio of terns resting on the water to terns flying does not take into account birds roosting on land, which is common behavior throughout the year for all tern species in the POCS. We did not find any relevant NFA estimates for tern species; therefore, we used the FA24 value of 0.38 published for Caspian Terns in Rueda-Uribe and others (2022) as the final FA value for all tern species.

For Red-billed Tropicbird, we used at-sea survey observations and tracking studies from a similar species (Red-tailed Tropicbird [P. rubricauda]; Spear and Ainley, 2005; Adams and others, 2020; Bonnet-Lebrun and others, 2021) to generate an FA of 0.32. For loons, we assumed an NFA of 0.001 based on species accounts that reported they do not fly at night. We did not find any data on NFA during migration periods, however, so this estimate of 0.001 may be low. The DFA for loons was derived from at-sea survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). The resulting FA ranged from 0.06 (Common Loon [Gavia immer]) to 0.20 (Pacific Loon [Gavia pacifica]).

For the three albatross species in the POCS, we used tracking studies and at-sea survey data to determine FA. Conners and others (2015) provided tracking-based NFA and DFA values for Laysan Albatross (Phoebastria immutabilis) and Black-footed Albatross (Phoebastria nigripes), and Adams and others (2020) provided tracking-based values for Laysan Albatross. At-sea survey data also provided DFA values for both species (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). The data from Laysan and Black-footed Albatross were used for Short-tailed Albatross (Phoebastria albatrus) resulting in an FA value of 0.67 for all three species.

We provided the literature values used to determine FA for storm-petrels as an example of how we incorporated data from a wide range of data sources (see the “Methods” section). The resulting FA values ranged from 0.52 for Townsend’s Storm-Petrel (Hydrobates socorroensis) to 0.62 for Wilson’s Storm-Petrel (Oceanites oceanicus), Fork-tailed Storm-Petrel (Hydrobates furcatus), and Leach’s Storm-Petrel. Pacific Northern Fulmar (Fulmarus glacialis rodgersii) DFA data were available from at-sea surveys (Drew and others, 2005); however, we did not find NFA data for this species. To determine NFA, we used the ratio of NFA to DFA for Pink-Footed Shearwater (Carle and others, 2019), which has similar FA to the Pacific Northern Fulmar; from this ratio, we estimated an FA value of 0.41.

Most of the gadfly petrel species (Pterodroma spp.) in the Vulnerability Index are present in the POCS during their nonbreeding seasons. For these species, nonbreeding season FA values from Bonnet-Lebrun and others (2021) were used to determine FA values of 0.36 for Mottled Petrel (Pterodroma inexpectata) and Cook’s Petrel (Pterodroma cookii), and an FA value of 0.40 for Murphy’s Petrel (Pterodroma ultima). Conversely, Hawaiian Petrel (Pterodroma sandwichensis) disperses to the western equatorial Pacific after breeding, and individuals are primarily observed in the POCS during breeding season (approximately April–October) foraging trips. Therefore, we used breeding season FA studies of Murphy’s Petrel (Clay and others, 2017) and Bermuda Petrel (Pterodroma cahow; Raine and others, 2021) to estimate Hawaiian Petrel FA in the POCS. These studies provided the percentage of time flying during breeding season foraging trips (incubation period [Clay and others, 2017] or chick rearing [Raine and others, 2021]). We used our knowledge of Hawaiian Petrel nesting behavior to estimate the complementary portion of time a bird was at its nest (as opposed to on foraging trips) to generate NFA and DFA value estimates. These values were used to determine an FA value of 0.61 for Hawaiian Petrel.

Diurnal Flight Activity values for all shearwaters were derived from survey data (Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017). Migratory and wintering FA24 values exist for many shearwater species in Bonnet-Lebrun and others (2021); therefore, we used the wintering values from this study as FA24 for each species studied (and similar species where relevant). The NFA and DFA values from tracking data were available for Pink-footed Shearwaters, Manx Shearwater (Puffinus puffinus) and Wedge-tailed Shearwater (A. pacifica). Wedge-tailed Shearwaters are not found in the POCS but have similar flight styles to the POCS shearwater species, except for Sooty Shearwater (A. grisea) and Short-tailed Shearwater (Puffinus tenuirostris), which spend less time flying. We estimated NFA for Sooty Shearwater and Short-tailed Shearwater by calculating the ratio of NFA to DFA for the other POCS shearwater species (NFA was 50 percent of DFA) and the DFA derived from at-sea survey data for these two species. We incorporated these published and calculated FA_N&D and FA24 values for each shearwater species for final FA values that ranged from 0.20 for Sooty Shearwater to 0.47 for Pink-footed Shearwater.

We used tracking studies of Red-footed Booby, Masked Booby, and Brown Booby during their breeding season and at-sea survey data of Cocos Booby (Sula brewsteri) to generate NFA and DFA for all five booby species (Drew and others, 2005; Weimerskirch and others, 2005; Adams and others, 2016, 2020). The resulting FA values ranged from 0.23 for Brown Booby to 0.42 for Red-footed Booby.

For cormorants, we assumed NFA was 0.001 based on species accounts and personal observation that they spend time roosting rather than flying at night. We did not use DFA values from at-sea survey data because a substantial proportion of cormorants are roosting on land during the day, resulting in the proportion of flying to resting birds observed at sea to be skewed high. From the DFA and FA24 values we found for Double-crested Cormorant and three cormorant species not found in the POCS, we estimated FA values of 0.03 for Double-crested Cormorant and 0.02 for Brandt’s Cormorant (Phalacrocorax penicillatus) and Pelagic Cormorant (Phalacrocorax pelagicus).

For pelicans, we also assumed NFA is 0.001 based on species accounts. Their DFA was derived from at-sea survey data and one telemetry study of Brown Pelicans (Pelicanus occidentalis; Croll and others, 1986; Drew and others, 2005; Adams and others, 2016; Takekawa and others, 2017), resulting in an FA of 0.27 for Brown Pelicans and American White Pelicans (Pelecanus erythrorhynchos).

Time Spent in Rotor Swept Zone (RSZt)

To generate estimations of RSZt for marine birds in the POCS, we used flight height studies that quantified the amount of time species spent between 20 and 200 m above the ocean surface (rotor swept zone height). Some studies set the lower limit of the rotor swept zone at 30 m instead of 20 m. We included these studies; however, they likely underestimated RSZt per our definition. For some species with limited flight height estimation, RSZt relied on broad estimations. We used these values because they are the best available information, and we acknowledge flight height data could be increased with future research. Most studies of RSZt were conducted in the North Sea and western Atlantic Oceans. Adams and others (2020) provided data for several POCS species from studies of breeding birds in Hawai‘i. In this section, we present the variations in RSZt data inputs and the resulting RSZt values among species groups.

We could not find any flight height data for Brant. Robinson Willmott and others (2013) estimated that Brant spent 18–42 percent of their flight time in the rotor swept zone; from this percentage, we estimated an RSZt value of 0.31 (median value of range). Similarly, we could not find any flight height data for any of the sea duck and merganser species in the POCS. Based on data from Common Eider (Somateria mollissima) flight heights around OWEI in the North Sea, we estimated 0.27 RSZt for the duck and merganser species in the POCS. For scoters, we used data from RSZt studies of Common Scoter (Melanitta nigra) in the North Sea for a RSZt estimate of 0.03 for the three scoter species present in the POCS.

We used flight height studies on grebe species near OWEI in the North Sea and western Atlantic Ocean to generate an RSZt estimate of 0.03 for the five grebe species present in the POCS. We could not find any published RSZt values for phalaropes; however, Robinson Willmott and others (2013) estimated 9–75-percent-RSZt. From this percentage range, we estimated both phalarope species in the POCS to have an estimated RSZt of 0.42 (median value of range). We used South Polar Skua and Great Skua (S. skua) flight heights near OWEI in the North Sea and western Atlantic Ocean to generate an RSZt estimation of 0.04 for the skua and jaeger species in the POCS.

For all alcid species except Common Murre, for which RSZt estimates were available, we used studies on alcid RSZt near OWEI in the North Sea and western Atlantic Ocean and applied those estimates to similar alcid species in the POCS. The resulting RSZt values ranged from 0.001 for Common Murre, Pigeon Guillemot (Cepphus columba), and the POCS puffin species, to 0.02 for the murrelets and small alcids.

Numerous studies have reported Black-legged Kittiwake flight heights near OWEI in the North Sea and western Atlantic Ocean. We compiled the values from these studies, resulting in an estimated RSZt of 0.16 for Black-legged Kittiwake. For other gull species, we grouped RSZt estimations from the literature by species size into small gulls, medium gulls, and large gulls. We applied these groupings, and the resulting RSZt estimates, to similar species in the POCS: 0.07 for small gulls (Sabine’s Gull [Xema sabini] and Bonaparte’s Gull [Chroicocephalus philadelphia]), 0.27 for medium gulls (Heermannn’s Gull [L. heermanni], Short-billed Gull [Larus brachyrhynchus], Ring-billed Gull [Larus delawarensis], California Gull [Larus californicus], and Iceland Gull [Larus glaucoides), and 0.35 for large gulls (Western Gull and Glaucous-winged Gull [Larus glaucescens]). Herring Gull flight heights were well studied, and we used this gull as a similar species for the large gulls in the POCS; their estimated RSZt was 0.35.

Studies of RSZt exist for Common Tern (Sterna hirundo) and Arctic Tern (Sterna paradisaea). For all other tern species in the POCS, we applied these groupings and the resulting RSZt estimates: large terns, crested terns, and terns. The final category of terns included all tern species that did not fall within either of the other more specific categories. The resulting RSZt values were 0.10 for Common Tern, 0.01 for Arctic Tern, 0.10 for crested terns (Royal Tern [Thalasseus maximus] and Elegant Tern [Thalasseus elegans]), 0.08 for large terns (Caspian Tern), and 0.06 for terns (Least Tern [Sternula antillarum], Gull-billed Tern [Gelochelidon nilotica], Black Tern [Chlidonias niger], and Forster’s Tern [Sterna forsteri]). We could not find any data on Black Skimmer (Rynchops niger) flight heights for an RSZt value; therefore, we selected an RSZt value 0.08 ascribed to large tern.

We did not find any flight height information for Red-billed Tropicbirds. We used the Red-tailed Tropicbird RSZt value of 0.71 from Adams and others’ (2020) study of breeding birds in Hawai‘i for Red-billed Tropicbird within the POCS.

Flight heights of Red-throated Loon (Gavia stellata) near OWEI have been well studied in the North Sea and western Atlantic Ocean. We used the median of the values from these studies (0.06) as the RSZt value for all four loon species in the POCS. Adams and others (2020) estimated the RSZt of Laysan Albatross to be 0.03, given the similar size and flight styles, so we used this value for Black-footed and Short-tailed Albatrosses in the POCS.

Bradbury and others (2014) estimated RSZt values of 0.00 and 0.02 for Wilson’s and Leach’s Storm-Petrel, respectively. Deakin and others (2022) estimated an RSZt value range of 0.000–0.025 for European Storm-Petrel (Hydrobates pelagicus). We used the 0.02 value estimated for Leach’s Storm-Petrel as the RSZt for that species. For all other POCS storm-petrel species, we used the median of all values for an estimated RSZt of 0.016; the use of this calculation included Wilson’s Storm-Petrel because, although Bradbury and others (2014) generated an estimate for that species, it was an RSZt of 0.00 and we determined not to include zeros in our database, given the level of uncertainty in the estimations.

Flight heights of Northern Fulmars near OWEI have been studied in the North Sea and western Atlantic Ocean. We used the median value of 0.002 from these studies as the RSZt value for Pacific Northern Fulmar in the POCS. No RSZt data exist for gadfly petrel species; therefore, we applied the RSZt of 0.002 for Pacific Northern Fulmar to the gadfly petrels present in the POCS. Flight height data from Manx Shearwaters and other shearwaters outside of the POCS were used to generate the RSZt value of 0.0002 for shearwater species present in the POCS.

Adams and others (2020) estimated RSZt values of 0.034 and 0.027 for Red-Footed and Coco’s Boobies off Hawai‘i, respectively. However, these data are from chick-rearing birds and may not be representative of the flight heights of boobies throughout the year within the POCS. Therefore, we also applied data from Northern Gannet (Morus bassanus) RSZt studies in the North Sea and the western Atlantic Ocean to RSZt estimates for the POCS booby species. The RSZt values were 0.20 for Red-footed and Cocos Boobies and 0.13 for the other three booby species in the POCS.

We used studies on cormorant and shag species flight heights near OWEI in the North Sea and western Atlantic Ocean to generate RSZt estimations of 0.08 for the three cormorant species in the POCS. We could not find any flight height or RSZt estimations for either of the two pelican species in the POCS, so we used an RSZt value of 0.31 (median of the range of estimates) for Brown Pelican reported in Robinson Willmott and others (2013) and used in Adams and others (2017).

Macro-Avoidance

Existing data on MA came from visual observations and radar studies at offshore wind developments in the North Sea. Therefore, for many species in the POCS, we used data from similar species to assign MA metric values. Below, we present the variations in MA data inputs and the resulting MA values among species groups. Some value ranges are large, highlighting the uncertainty present in MA estimations.

We used the one MA estimate for Brant (0.68) from Krijgsveld and others (2011). Using MA estimates for sea ducks from Cook and others (2012) and Krijgsveld and others (2011), we generated an MA of 0.73 for Red-breasted Merganser, Harlequin Duck, and Long-Tailed Duck in the POCS. Bradbury and others (2014) estimated an MA of 0.90 for Common Scoter, and we used this value for all three scoter species in the POCS.

Using MA estimates for grebes from Cook and others (2012) and Krijgsveld and others (2011), we generated an MA of 0.39 for all five grebe species in the POCS. Because we are not aware of any published data on phalarope MA specifically, we used the estimated MA of 0.27 for shorebirds from Krijgsveld and others (2011) for the two phalarope species in the POCS. Based on MA estimates for jaegers and Great and South Polar Skuas from Cook and others (2012) and Krijgsveld and others (2011), we generated an MA of 0.14 for the South Polar Skua and the three jaeger species in the POCS.

For alcids, several studies reported Common Murre MA and a few studies have reported, more generally, on alcid MA. From these studies, we estimated an MA value of 0.69 for Common Murre and 0.68 for all other alcid species in the POCS.

Macro-Avoidance has been well studied in Black-legged Kittiwakes, with a very wide range of estimated values. Vanermen and others (2017) estimated kittiwake MA to be 0.75, whereas Cook and others (2014) estimated an MA rate of 0 for kittiwakes. We took the median of all published estimates to assign an MA of 0.42 for Black-legged Kittiwakes. We used the same gull size groups described in the “Time Spent in Rotor Swept Zone” section to assign MA estimates among size groups in the POCS. From these findings, we estimated MA values of 0.28 for small gulls (Sabine’s and Bonaparte’s Gulls), 0.30 for medium to large gulls (Heermann’s, Short-billed, Ring-billed, California, and Iceland Gulls), and 0.30 for large gulls (Western, European Herring, and Glaucous-winged Gulls).

We found MA estimates for Sandwich Terns (Thalasseus sandvicensis) and for terns generally. We used the median of these reported values to generate an MA of 0.49 for all tern species in the POCS.

Robinson Willmott and others (2013) assigned Red-billed Tropicbirds an MA range of 18–29 percent. Using the midpoint of 18–29 percent for this estimation, we assigned an MA value of 0.24 for Red-billed Tropicbirds.

Macro-Avoidance in loons has been well studied in the North Sea. We took the median of all published MA estimations for loons to assign an MA value of 0.81 for the loon species present in the POCS.

We used the MA estimate of 0.28 from Krijgsveld and others (2011) for Pacific Northern Fulmar. Cook and others (2012) estimated the MA of tubenoses in general (including albatrosses, petrels, shearwaters, and storm-petrels) to be 0.50. Therefore, we estimated an MA value of 0.39 (the median of these two estimates) for all tubenoses in the POCS except Pacific Northern Fulmar.

We used Northern Gannet MA estimates for the five booby species in the POCS, resulting in an MA estimate of 0.87. Using MA estimates for cormorants and shags from Cook and others (2012) and Krijgsveld and others (2011), we generated an MA of 0.21 for the three cormorant species in the POCS.

No MA data were available for pelicans; therefore, we used MA estimates from Northern Gannet MA, resulting in an MA of 0.87. Northern Gannet has the most similar flight style and foraging patterns to pelicans of any species for which MA has been quantified.

Habitat Specificity

Habitat Specificity is the average of SS and FS. We calculated SS from the species distributions in Leirness and others (2021). Distribution models were not generated by Leirness and others (2021) due to insufficient data for 33 infrequently encountered species. For these species, we used distribution values for similar species and eBird data (eBird, 2021) to estimate SS. We generated FS values from the Birds of the World accounts for each species (Billerman and others, 2022). From the species accounts, we extracted all relevant information from the “Data and Foraging” sections in Birds of the World to enumerate FS, as outlined in table 5. Habitat Specificity for each species was the mean of their SS and FS values (table 5).