Caribou distribution during calving in the northeast National Petroleum Reserve-Alaska , June 1998 to 2000

Barren ground caribou (Rangifer tarandus granti) of the Teshekpuk Caribou Herd (TCH) inhabit the western portion of Alaska's Arctic Coastal Plain within the National Petroleum Reserve—Alaska (NPR-A). Alaska's North Slope communities, management agencies, and private industry are interested in this herd because of its importance as a subsistence resource and location relative to potential petroleum development. From 1998 through 2000, we monitored caribou distribution during the calving period within the Northeast Planning Area of the NPR-A using systematic strip-transect aerial surveys, as well as VHF and satellite telemetry for cow caribou. Aerial survey and telemetry data indicated cows with calves were distributed around Teshekpuk Lake, with a concentration south of the lake in 1999 and 2000. Inconsistencies in weather conditions, survey timing (both strip-transect and VHF surveys), 100% coverage survey areas, and small sample sizes confound interpretations of our results. However, several patterns were apparent. Later transect survey timing (7—12 June versus 4—7 and 5—8 June) resulted in more cow/calf pairs recorded. Our 18% coverage area, originally based on VHF telemetry data for the extent of TCH calving, covered a consistently high proportion (95% to 100%) of the annual calving ranges (95% kernel utilization distributions), but accounted for only 24% to 46% of the adult cows in the TCH based on the current Alaska Department of Fish and Game population estimate (1999) and average 1998¬2000 herd composition. It appears that either our transect survey methodology significantly underestimated the true number of caribou cows in the study area, many cows calved outside the area or moved into the area and calved after our surveys, or we have over estimated the number of reproductive cows in the herd. Our 100% coverage transect areas covering oil and gas lease areas, contained 38% of the calving range with 23% of TCH cows in 1999; and 18% of the calv¬ing range with 8% of TCH cows in 2000. Based on 95% minimum convex polygon ranges, satellite collared cow/calf pairs were not stationary during either our survey period (14.7 ± 6.56 km2; mean ± standard error of the mean; 4—12 June) or during the calving period (86.9 ± 72.30 km2; 1—20 June) during 1998—2000. Site specific pre-development data on caribou distribution during calving in NPR-A will be useful for assessing the importance of specific areas to caribou during calving and for designing oilfields that minimize impacts should oil development occur.


Introduction
Barren ground caribou (Rangifer tarandus granti) of the Teshekpuk Caribou Herd (TCH) inhabit the western portion of Alaska's Arctic Coastal Plain and range primarily within the National Petroleum Reserve-Alaska (NPR-A; U. S. Department of the Interior [USDI], 1998).North Slope communities, management agencies, and private industry are par-Rangifer, Special Issue No. 14,2003 Rangifer,  ticularly interested in this herd because of its importance as a subsistence resource and location relative to potential oilfield development and existing oil and gas exploration.
The TCH was recognized as a separate herd from the Central Arctic (CAH) and Western Arctic (WAH) herds, which also range within NPR-A, in the mid 1970s (Davis & Valkenburg, 1978).Since Table 1.Caribou number, density (caribou/km 2 ), and mean group size (± standard error of the mean) recorded on 6 systematic aerial strip-transect surveys during calving in the National Petroleum Reserve-Alaska, 4-12 June 19984-12 June -2000.A sightability correction factor was applied to 18% and 100% coverage surveys during 1999 and 2000 to correct for poor sightability due to patchy snow cover (Lawhead et al., 1994).
Most TCH cows begin migrating to the Teshekpuk Lake area in May, and by early June most of the cows have moved into the calving area sur¬ rounding the lake (Carroll, 1999).Changes in the location of the calving area for the TCH have been documented since this herd was identified.During the mid 1970s, calving was reported on the west side of Teshekpuk Lake (Davis & Valkenberg, 1979;Silva, 1985).Since the mid 1990s, however, calving has been northeast, east, and southeast of Teshekpuk Lake (Carroll, 2001).
Although there has been oil and gas winter explo¬ ration within the Teshekpuk Lake area since the late 1940s (Silva, 1985), currently the TCH has no regular contact with industrial development (Murphy & Lawhead, 2000).Davis & Valkenburg (1978) recog¬ nized the importance of obtaining baseline data on the TCH prior to development of the NPR-A.Baseline data on herd size, distribution, and season¬ al habitat use will be useful for planning to mini¬ mize impacts of likely future petroleum exploration and development on the TCH.Current abundance and distribution data will allow resource managers to make informed decisions regarding management of the herd's habitat.
The objectives of this study were to: (1) determine the current (1998)(1999)(2000) number, sex/age composi¬ tion, and distribution of caribou during the calving period near Teshekpuk Lake and within areas of potential oil and gas development and, (2) to investi¬ gate the importance of these areas of potential oil and gas development to the calving range for the TCH as defined by VHF calving sites.VHF telemetry data describe the extent of calving for each year of our transect surveys.Satellite telemetry data describe the potential for movement of cows with and without calves during both our transect survey period (2-14 June) and the calving period (1-20 June).

Methods
Aerial Surveys Between 4-12 June 1998-2000, we con¬ ducted systematic, strip-transect aerial sur¬ veys (Caughley, 1977b) during the calving period from fixed-wing aircraft (Cessna 206).Transect centerlines were spaced at 9.6-km and 1.6-km intervals, using a 1.6¬ km transect width, resulting in 18% and 100% coverage of the survey areas, respec¬ tively.All transects were oriented north-south and centered on township and section lines mapped on 1:63 360-scale U.S. Geological Survey (USGS) topographic maps.Systematic surveys were flown 90 m above ground level at 185-200 km/h airspeed.
During surveys, two observers searched an 800-m wide area on each side of the transect centerline.A third observer entered data into a computer and assisted with spotting caribou during 1999 and 2000 surveys.Aircraft wing struts were marked with tape to enable visual control of transect strip-width (Pennycuick & Western, 1972) and estimation of group distances perpendicular from the transect cen¬ terline.Transect widths were also checked periodi¬ cally with inclinometers and maps to assist in train¬ ing observers.
As with previous aerial surveys in the Prudhoe Bay oilfield (Pollard et al., 1992;1996), global position¬ ing system receivers (GPS) were used to navigate the aircraft and provide position records during system¬ atic surveys.Locations of animals were estimated by using the GPS in combination with visual estimates of their perpendicular distance to the transect cen¬ terline.At the time of sighting, all data were entered directly into a notebook computer that was linked to the GPS receiver (Geolink version 6.1, Michael Baker Jr., Inc., Jackson, MS).The computer software associates a real-time GPS position (latitude and longitude of the survey aircraft) for each animal sighting entered with related attributes (e.g. , num¬ ber of individuals in the group and sex/age classifi¬ cation).When possible, behavior and habitat types, along with the group attributes and time of sight¬ ing, were recorded into the computer or an audio recorder.
We counted and classified caribou as bulls, cows, calves, or unclassified, based on body size, antler development, pelage, and calf presence.Unclassified caribou were adults or yearlings that couldn't be classified with confidence.Caribou near the outer margin of transect strips were most difficult to clas¬ sify.During patchy snow cover sightability of cari¬ bou was diminished and a sightability correction fac¬ tor (SCF=1.88)developed for 20-70% snow cover on Alaska's North Slope (Lawhead et al., 1994) was applied to survey results.The total number of cows within the 18% coverage survey area was extrapolat¬ ed and the variance was calculated based on the sam¬ ple of 11 transects within the 6509 km 2 area (Caughley, 1977a).June weather data was obtained from National Oceanic and Atmospheric Administration weather stations at Deadhorse, Alaska for 1998 and Nuiqsut, Alaska for 1999 and 2000.Data were unavailable for the Nuiqsut station for June 1998.

Satellite and VHF Telemetry
Through a cooperative project involving ADFG, NSB, and BLM, caribou were captured using a skidmounted net gun from a Hughes 500 helicopter (Carroll, 2001).Platform Transmitter Terminal (satellite radio collar transmitters or PTTs) or VHF (Very High Frequency) collars were attached to cap¬ tured caribou; and satellite collars were programmed to transmit on a 6-h per 48-h cycle (Carroll, 2001).Data were received from 6 collared animals by down¬ loading satellite location data (monthly summaries) from the Argos Data Collection and Location System (ARGOS) in Landover, Maryland.VHF radiotracking flights collected information on 27-36 animals year -1 between 2-16 June to determine caribou movements, distribution, and reproductive status (Carrol, 2001).
VHF calving locations defined as the first sighting of an individual cow with a calf (Carroll, 2001) were used to calculate "calving ranges" as the fixed kernel utilization distribution (Worton, 1989) with least squares cross validation (Silverman, 1986) using the Animal Movements extension (Hooge and Eichenlaub, 1997) for ArcView® Geographic Infor¬ mation System.Calving ranges for cow caribou based on VHF telemetry from 2-16 June were calculated separately for cows with (n = 16-23 cows year -1 ) and cows without (n = 4-14 cows year -1 ) calves.
All mean values are reported with the standard error (SE) of the mean.Spatial comparison of calving ranges and study area coverages was completed using MapInfo Professional ® .Mean range values were com¬ pared using 2-sample t-tests without the assumption of equal variance (Snedecor & Cochran, 1980).1).However, more cow/calf pairs were recorded in 1999 (Table 1).In general, cow/calf pairs were closer to Teshekpuk Lake in 1998 and 2000 than in 1999 when survey timing was later (Fig. 2).In 1999, more cow/ calf pairs were west of Teshekpuk Lake and at the southern border of the study area than in the eastern study area (Fig. 2).In the 18% coverage area, cows without calves generally occurred over a larger area than cow/calf pairs during 1998 and 2000, with more cows with¬ out calves occurring west and southwest of Teshekpuk Lake (Fig. 2).
Based on the 11 transects sampled consistently during 1998-2000 in the 18% coverage area, the estimated num¬ ber of cows declined from the number observed in 1998 by 47% in 1999, and 36% in 2000 (Table 2).Our estimated number of cows within this area ranged from 24% to 46% of the estimated total of 10 077 cows in the TCH, based on the 1999 ADFG population estimate (Table 2).

100% survey area
In 1998, cow/calf pairs occurred throughout the 100% coverage area cen¬ tered on the area northeast and southeast of Aerial Surveys Teshekpuk Lake (Fig. 3).In 1999 and 2000, cow/calf We completed 3 aerial surveys at 18% coverage and pairs were concentrated in the northwestern half of 3 aerial surveys at 100% coverage in the Teshekpuk the survey area, southeast of Teshekpuk Lake and Lake area during the caribou calving period between south of the Kogru River (Fig. 3).In 2000, we observed lower caribou densities, fewer total caribou, and fewer cow/calf pairs within the 100% coverage survey area than in the same area in 1999 (Table 1).Survey timing in 1999 was later (9-12 June) than in 2000 (5-8 June).Mean group size declined steadily during the 1998-2000 survey period (Table 1).
The eastern extent of calving in the 100% coverage area appears to be Atigaru Point; only 1 cow/calf pair was sighted east of Atigaru Point (Fig. 3).The 100% coverage area during 1998 included 19% of TCH cows.Survey coverage, focused on areas of potential oil and gas development, included 23% of TCH cows in 1999 and 8% in 2000 (Table 2).As in the 18% coverage area, a higher proportion of cow/calf pairs was recorded during 1999 than in either 1998 or 2000.
Based on the limited data available for this period, satellite-collared cows with calves (n=4 caribou) during 1-20 June 1998-2000 were generally south and southeast of Teshekpuk Lake and cows without calves (n=2 caribou) were  1998-2000(Carroll, 2001) )  area after our survey on 5-8 June, either before or after parturition.
More cow/calf pairs were observed on our transect surveys in 1999 than in either 1998 or 2000 appar¬ ently because survey dates in 1998 and 2000 were before most cows had calved.Patchy snow conditions and cooler temperature from 1-15 June were likely responsible for the decreased numbers of caribou recorded within our 18% coverage survey area in 1999 and 2000 (Table 2).Calving ranges were larg¬ er in 1999 and 2000 than in 1998, indicating that cows/calf pairs were spread across a larger area in these years.VHF determined calving ranges during 1998-2000 were potentially influenced by (1) sur¬ vey timing, (2) small sample sizes, and (3) weather conditions.
We recommend that calving period transect sur¬ veys be flown after 15 June, when most calves have been born, and the ADFG VFH calving surveys have been completed.This later timing may be less dis¬ ruptive to pre-parturient and parturient cows and should allow documentation of a higher proportion of cow/calf pairs.Also, because VHF determined calving ranges will already be established, this may allow a more accurate comparison between the pro¬ portion of TCH cows and the proportion of the TCH calving range represented by the area of interest for potential development.
Annual variation in calving distributions is evi¬ dent from the few years of survey data presented here, and is potentially influenced by snow cover, weather conditions, survey timing, and possibly sample size.Comparing the proportion of the calv¬ ing range within a particular transect survey area should indicate how important any particular area is for the calving period during a given year.Our com¬ parison of calving ranges based on VHF surveys to estimates for the number of TCH cows within the 18% coverage transect survey area suggests that the VHF determined calving ranges may underestimate the calving range of the TCH, for these years.Satellite telemetry data for 1998-2000 indicate cows were not static either during our survey period or during the calving period.

Fig. 4 .
Fig. 4. Caribou calving locations determined by VHF radio telemetry from 4-15 June 1998-2000 (Carroll, 2001) and fixed kernel utilization distributions for calving locations and for cows with¬ out calves in the National Petroleum Reserve-Alaska.