CHAPTER 1

Trends in Dall’s sheep populations in military overflight areas
mitigated and not mitigated for potential negative impacts

INTRODUCTION

Large remote landscapes are a scarce resource and are in demand from sources as diverse as wildlife species and the United States Air Force (USAF).  Interior Alaska is such a place.  In 1976, large portions of the airspace in eastern Interior Alaska were designated as “special use” for military operation areas (MOAs).  Within these areas, the USAF, as well as allies of the United States, practice air combat, and formation and aerobatic training (Final Alaska MOA EIS Record of Decision 1997).  Since 1976, expectations of the military and a variety of technological and other changes created a need to modify the existing airspace arrangement and an environmental impact statement (EIS) was written (Department of the Air Force 1995) followed by a Record of Decision (Final Alaska MOA EIS Record of Decision 1997).  The Record of Decision (1997) established a set of mitigation measures intended to minimize the impacts of military jet overflights on people and animal within the MOA structure.  The 11^th Air Force Noise/Flight Sensitive Areas List (2002) specifies how these mitigation measures will be implemented and is updated on a periodic basis.  Protection of Dall’s sheep in interior Alaska is the focus of 3 of mitigation measures.

Mitigation measures directed at Dall’s sheep focus on the lambing season and the rut.  In the Alaska Range, mitigation measures for lambing sheep are in place from 1 May - 30 June.  Military aircraft are to maintain a floor of 1,000 feet above ground level (AGL) or 5000 feet AGL depending on location (Fig. 1.1).  During the rut, military aircraft are restricted to flying above 5000 feet AGL in the mitigated area under the Fox MOA from 15 November - 15 December (Fig. 1.1).  In the Yukon-Tanana Uplands, military aircraft are to maintain a floor of  5,000 feet AGL from 10 May - 15 June over the Cirque Lake lambing area (FAA Aeronautical Study 95-AAL-042NR, Dept. of the Air Force 2002).  Because the floor of this MOA is 100 feet above ground level at other times of the year, as opposed to the 7,00 foot AGL ceiling in the Fox MOA (Dept. of the Air Force 1997), Dall’s sheep in this area may be exposed to a significant change in flying activity during the 1 month mitigation period.  Because Dall’s sheep mitigation measures in the Yukon-Tanana Uplands are some of the most restrictive, this area is the logical choice for initiating a study to examine the effectiveness of the mitigation measures.  An additional mitigation area, the Charley River, partially overlaps the Cirque Lake lambing area.  Although not specifically directed at Dall’s sheep, sheep in this area may still benefit.  The Charley River is mitigated 2 nautical miles either side of the riverbank from the surface to 2000 feet AGL from 15 April to 31 August (Dept. of the Air Force 2002).

Dall’s sheep populations are found in disjunct pockets throughout the Yukon-Tanana uplands.  Sheep occur in the limited patches of steep rocky alpine habitat interspersed throughout the dominant boreal forest (Durtsche et al. 1990, Boudreau 1996).  Interchange of individuals between habitat patches has been documented in a few instances (Durtshe et al. 1990, Burch and Lawler 2001).  History of disturbance from civilian aircraft, hunting pressure and recreation pressure is minimal (Boudreau 1996). 

Figure 1.1  Military Operations Areas (MOAs) in Interior Alaska.  Of the areas mitigated for protecting Dall’s sheep, the mitigation around Cirque Lake is the most restrictive with an exclusion area for military jets from the surface to 5,000 feet above ground level (AGL) from 10 May – 15 June.

Dall’s sheep in the Yukon-Tanana Uplands have been living with the MOAs since their creation in 1976.  The nature of the flying has changed however, as have the aircraft.   In the mid 1980s, F-15s and A-10 military aircraft replaced the F-4s that had been stationed in Alaska.  In 1991, F-16s were added to the air fleet.  In 1995, the current pattern of mitigation was put into place (Department of the Air Force 1995).  It is a goal of this document to evaluate the effectiveness of the Cirque Lake Dall’s sheep mitigation on sheep population levels and trends in comparison to areas within the Yukon-Tanana Uplands with no mitigation.      

STUDY AREA

The study area was eastern interior Alaska in the Yukon-Tanana Uplands and is centered in Yukon-Charley Rivers National Preserve. The most prominent feature of this area is the Yukon River which flows across the northern portion of this region (Fig. 1.2).  Elevations range from 305 m above sea level to 2000 m above sea level.  Climate is semi-arid continental.  Climate summaries are available from 2 weather stations in the vicinity of the study area (both are approximately 115 km from the center of the Cirque Lake survey unit). At Circle, Alaska, north-west of the study area, mean annual precipitation from 1957 – 1999 was 20.7 cm.  Mean daily maximum temperatures ranged from 22.8°C in July and –22.9°C in January (National Weather Service). At Eagle, Alaska, east of the study unit, mean annual precipitation from 1957 – 2002 was 30.4 cm.  Mean daily maximum temperatures ranged from 22.6°C in July and –19.7°C in January (National Weather Service).  Topography in the study area is rolling forested areas with rugged alpine tundra areas interspersed throughout.

The study area is within the subarctic boreal forest zone.  Dominant tree species include black spruce (Picea mariana) in low-lying areas, and white spruce (Picea glauca), aspen (Populus tremuloides) and paper birch (Betula papyrifera) in better drained locations.  Above 600 m, tundra vegetation dominates.  Plant species in these alpine areas are typical tundra plants and include dryas (Dryas spp.), dwarf willow (Salix spp.), and bearberry (Arctostaphylos spp.)  The Yukon-Tanana Uplands were largely unglaciated throughout the Quaternary and as a result, steppe vegetation persists on south facing bluffs along riversides.  In some instances, such as along the Charley River, these south facing bluffs are heavily utilized by Dall’s sheep.  More detailed descriptions of the physiography and vegetation of this area can be found in Roland (1996), Swanson (1999), and Ducks Unlimited (1998).

METHODS

Aerial Surveys and Study Area

Observers counted Dall’s sheep from a helicopter in 8 study units within the interior Alaska, Military Operations Areas in June or July during 1997 – 2002 (Fig. 1.2). Not all units were surveyed in 2000 as poor weather prevented the completion of the survey.  Some sheep counts were available prior to 1997 (Appendix A).  In most cases, this data were not used in analysis due to uncertainties regarding the area covered by the survey, the time spent surveying (sampling intensity), large gaps between years in data sets, and differences in timing of the survey.  From 1997 – 2002, the same 2 individuals acted as observers for the 8 units used for most analysis with 1 exception.  In 2000, a year when the survey was only partially completed, only 1of the 2 participated.  In 2 of the survey units (Cirque Lake and Mount Sorenson) data collected from 1993 through 1995   is consistent with the areas and the survey effort of the 1997-2002 data set but the observers collecting the data differed.  Although one observer during these 3 surveys was the same individual, the second observer was a different individual each year.  Different types of helicopter have been used during surveys.  Before 1998, surveys were conducted with a Bell 206 Jet Ranger, or a Hughes 500 and from 1998 – 2002, surveys were conducted with a Robinson R-44.

Figure 1.2.  Dall’s sheep survey units in the Yukon-Tanana Uplands, eastern interior Alaska.  Units were surveyed in mid-summer from 1997-2002 using a helicopter.

Dall’s sheep were counted in study units by flying contours around terrain features (Burch and Lawler, 2001).  Typically, 2 – 3 passes were made over a given feature.  Observed sheep were classified as rams, ewe-like, yearling, or lambs.  Rams included all animals with greater than 1/4 curl horns.  Rams < 1/4 curl and all adult ewes were categorized as “ewe like” because of the difficulty in distinguishing between these 2 classes.  Yearling ewes were categorized as such based on size differences relative to lambs and adult ewes as well as horn characteristics (Geist 1971).  Yearling rams were likely classified primarily as “ewe like” because of the similarity in appearance in yearling rams and adult ewes.  In some situations, however, yearling rams with small bodies and small horns may have been classified as “yearlings”.  Lambs of the year were conspicuous because of their small size and horns.  Locations of sheep were marked on 1:63,360 scale USGS maps and numbers and sex and age classifications were recorded.  This information was used to reduce the potential for “double counting” sheep during the survey.  Flight lines of areas surveyed were recorded as were survey times.

Survey units and areas surveyed for Dall’s sheep were areas known to consistently harbor sheep.  These areas were flown until observers and the pilot felt comfortable that all probable sheep habitat had been surveyed.  Partially due to different environmental factors, this comfort level varied from year to year. With experience, we were also able to eliminate improbable sheep habitat from the survey.  Therefore, the speed at which these units were surveyed varied as did the area.  We used flight time as a measure of survey effort as this measurement includes aspects of both flight speed and area.  To calculate sheep densities, we used the perimeter of our flight path during the 1999 survey as this area is a good approximation of known sheep habitat in the area.  All sheep observed during the surveys in all years fell within the perimeter of these boundaries with 2 exceptions.  Both exceptions occurred to the south of the 1999 Twin Mountain survey area.  In 1997, 2 ewes and 2 yearlings were observed south of the boundary and in 1998, 8 ewes, 2 yearlings and 4 lambs were observed south of the boundary.  

We evaluated sheep sightability and developed correction factors using radiocollared sheep and mark-resight methodology (Furlow et al. 1981, Leslie and Douglas 1979, 1986, Neal et al. 1993).  Two separate studies involving radiocollared sheep have occurred within the study units during years in which aerial surveys were conducted (Burch and Lawler 2001).  As a result, the number of radiocollars within our study area has varied (range = 8 - 24 animals), as has the distribution of collars.  Radiocollared animals in 1997 were evenly distributed throughout the units surveyed with the exception of West Point/Puzzle Gulch where there were no radiocollared sheep.  During subsequent years, radiocollared sheep have been scattered throughout the survey units but have been most concentrated along the Charley River and Cirque Lake units, and in the West Point/Puzzle Gulch area.  A fixed-wing aircraft (Piper Super Cub, PA-18) was used to locate all radiocollared sheep within 1 day of the helicopter sheep population survey and we assumed a closed population (no births, deaths, immigrations, emigrations during the 24 h period; Neal et al. 1993). Coordinates of radiocollared sheep were noted during the helicopter aerial survey and compared to the locations recorded by the fix-winged aircraft.  Only radiocollared sheep located within the perimeters of the study units were used in estimates of sheep populations in the survey units and only radiocollared sheep located within the perimeters of the study area were used in estimates of sheep populations in the study area.

Of the 8 study units surveyed within the MOA structure, 2 are mitigated for low-level military overflights (Fig. 1.2).  Beginning in 1995, military aircraft are to avoid the Cirque Lake Dall’s sheep lambing area from the surface to 5000 feet AGL (above ground level) from 10 May to 15 June.  The area to be avoided is a 7 nautical mile radius area (Fig. 1.1) centered at 64°48’00”N/143°45’00”W (11^th Air Force 1997).  Mitigation for the Charley River is an avoidance area 2 nautical miles either side of the riverbank from the surface to 2000 feet AGL from 15 April to 31 August.  The mitigated area extends from 64°41’00”N/143°38’00”W to 65°19’00”N/142°46’00”W (11^th Air Force 1997).

DATA ANALYSIS

Sampling effort (survey time and area per unit) varied from year to year.  In each year, however, effort was spent in each unit to complete what the observers felt was a thorough survey.  We tested the assumption that the number of sheep counted during a survey was not affected by the sampling effort during surveys using a general linear model (GLM: SPSS 9.0 1999).  Because different sexes may vary in sightability (McCorquodale 2001), we separated total number of sheep into “ewe groups” and rams.  Yearlings and lambs were included with ewe group totals because of autocorrelation considerations (it is unusual to observe lambs or yearlings without ewes).  Survey units varied considerably in complexity and character.  Some had substantial tree cover whereas others were devoid of trees.  Differences in terrain complexity may result in differences in the amount of survey time expended in an area.  For this reason, we analyzed units separately.

Trends in Dall’s sheep population numbers following the initiation of mitigation measures within the MOA structure were examined using repeated-measures general linear model.  For analysis, we lumped ewe, lamb, yearling and ram numbers into “total sheep” to reduce the number of dependent variables within our model.  Seven survey units were classified into 1 of 3 categories: 1) no mitigation (n = 5); 2) Cirque Lake mitigation (n = 1); 3) Charley River mitigation (n = 1).  Sheep surveys from 1997 – 2002 were used for analysis with the exception of 2000 (n = 5).  The 2000 survey was excluded due to missing data.  An 8^th survey unit (West Point/Puzzle Gulch) was also excluded from this analysis because of missing data.  Data used in repeated-measures analysis were examined for sphericity with the Machly sphericity test.  Results of all statistical tests were considered significant when P < 0.05. 

We tested for differences in number of rams and total number of sheep in ewe groups before and after the initiation of mitigation measures at a mitigated area (Cirque Lake) and non-mitigated area (Mount Sorenson) with a multivariate analysis of variance (MANOVA).  These units were chosen for analysis because they are the only 2 units for which we have more than 2 surveys prior to 1995, the year mitigation measures were started (Appendix A).  When MANOVA detected differences in number of rams and number of sheep in the ewe group category, we used analysis of variance (ANOVA) to identify variables that differed.  

The disjunct nature of the sheep habitat in the Yukon-Tanana Uplands suggests numerous sheep populations isolated by large expanses of unsuitable habitat.  Evidence from the aerial sheep surveys, however, show great variability in numbers and composition of sheep populations that are inconsistent with reasonable mortality and production rates in closed populations suggesting interchange between these populations.  Evidence from radiocollared sheep also suggested interchange between these populations (Burch and Lawler 2001).  For this reason, following the examination of individual units, we looked at trends in sheep population in the Yukon-Tanana Uplands using a general linear model by combining sheep counts from multiple units into a yearly sheep population index.  Five years of data were available for analysis by combining data from the Charley River, Cirque Lake, Mount 5580, Twin Mountain, Mount Sorenson, Copper Mountain, and Diamond Fork survey units.  Known sheep areas within the Yukon-Tanana Uplands excluded from this analysis because of insufficient data, non-compatible or non-existent data were West Point, Puzzle Gulch, Glacier Mountain and Mount Harper (Fig 1.1).

The precision of sheep surveys since 1997 was evaluated using mark-resight methods (White and Garrott 1990, Neal et al. 1993).  We used the Lincoln-Peterson estimate to evaluate sheep numbers within units surveyed (Chapman 1951; Seber 1982, White and Garrott 1990) using the number of radiocollared sheep and total number of sheep observed during the helicopter survey.  In addition, we estimated population variance and constructed 95% confidence intervals for yearly population estimates (Seber 1982, White and Garrott 1990) in 7 survey units.  Sheep counts and radiocollared sheep at West Point/Puzzle Gulch were not included in mark-resight population estimates because data for this unit was only available for 3 years.

RESULTS

Survey Effort

Considerable variation existed in survey effort, both within survey units (Table 1.1) and between survey units.  Within survey units, there was a tendency for smaller units and units in which less time was spent surveying to have greater variation in survey times (Table 1.1).  The exception to this trend was Mount 5580 which had both a low coefficient of variation (9%) as well as a small mean survey time (0.5 h).  Of all units surveyed, the most effort per unit area was directed toward Copper Mountain and the least was directed at the Charlie River (Table 1.1).  Survey effort was over 80% greater per unit area covered at Copper Mountain in comparison to the Charley River. 

Although variable amounts of effort were spent surveying individual units, we were not able to attribute variation in the number of sheep counted in ewe groups and ram groups within a unit to survey effort (Table 1.2).  Only at Mount 5580 did we approach a significant relationship (P = 0.078). Because none of the MANOVA results were significant, follow-up univariate tests for the effect of survey effort on number of sheep counted in ewe groups and the effect of survey effort on number of sheep counted in ram groups were not examined.

Trends in Dall’s sheep numbers in aerial survey units

Since 1997, both the total number of sheep counted as well as number counted in each sex and age class varied considerably within units and this was true of both the mitigated as well as the unmitigated areas. (Fig. 1.3; Appendix A).  This variation was considerable as illustrated by the number of sheep observed while surveying the Mount Sorenson area.  Observed sheep declined by 44% from the 1997 survey compared to the 1998 survey, and this was followed by a 180 % increase in observed sheep when comparing the 1998 survey to the 1999 survey (Fig 1.3).  Coefficients of variation for observed Dall’s sheep in individual units ranged between 55% and 20% for Diamond Fork and Mount 5580, respectively.  Densities of sheep also varied considerably between units with estimated sheep densities of up to 0.52 sheep per km² during 1999 and 2000 in the Twin Mountain unit and as few as 0.03 sheep per km² during 1997 in the Diamond Fork unit (Fig. 1.4).  Mean Dall’s sheep densities for all units over this time period was 0.18 [95% confidence level = 0.16, 0.19] sheep per km² of areas surveyed.  From 1997 through 2002, no differences in trends in Dall’s sheep numbers within the interior Alaska MOA were identified for those areas mitigated and not mitigated for low-level military overflights with a repeated-measures general linear model (Pillai’s trace: F = 1.079, d.f. = 4, 2, P = 0.533).

Trends in Dall’s sheep populations before and after mitigation

MANOVA results indicated a significant difference in the number of rams and number of sheep in the ewe group category in the years before and after the initiation of the mitigation measures at Cirque Lake and Mount Sorenson (Pillai’s trace; F = 5.04, d.f. = 2, 18, P = 0.029).  In addition, there was a significant interaction between location (Cirque Lake or Mount Sorenson) and mitigation (before or after)(Pillai’s trace F = 4.91, d.f. = 2, 18, P = 0.020).  Sheep numbers increased at Mount Sorenson following the initiation of mitigation and sheep numbers at Cirque Lake stayed relatively unchanged.  Examination of univariate results indicate that these differences were due to a significant difference in the number of sheep in the ewe groups (F = 10.63, d.f. = 1, 22, P = 0.004) but not in the ram groups (F = 0.28, d.f. = 1, 22, P = 0.605) before and after the initiation of mitigation measures.  The univariate results for the location * mitigation interaction also support a difference in the number of sheep in the ewe group category (F = 9.305, d.f. = 1, 22, P = 0.007) and not the number of rams (F = 1.982, d.f. = 1, 22, P = 0.175).  In general, following 1995 (the year mitigation measures were started), sheep numbers and sheep densities were higher at Mount Sorenson (unmitigated) in comparison to Cirque Lake (mitigated)(Fig. 1.3 and 1.4).

a) Non-mitigated

b) Mitigated

Figure 1.3.  Counts of Dall’s sheep in Interior Alaska Military Operations Areas.  Some survey units may be exposed to low-level military jet overflights throughout the year (a) and others have seasonal mitigations (b).  Military aircraft may fly as low as 100 feet above ground level but 10 May - 15 June, military jets must stay 5000 feet above ground level at the Cirque Lake area and 15 April - 31 August, military jets must stay 2000 feet above the Charley River area.

a) Non-mitigated

b) Mitigated

Figure 1.4.  Densities of Dall’s sheep in Interior Alaska Military Operations Areas.  Some survey units may be exposed to low-level military jet overflights throughout the year (a) and others have seasonal mitigations (b).

Trends in Dall’s sheep populations in the Yukon-Tanana Uplands

An examination of the sheep population in the Yukon-Tanana Uplands that included data from multiple survey units (see methods) did not indicate a significant trend in the sheep population from 1997 to 2002 (F = 0.416, d.f. = 1, 4, P = 0.565; Fig. 1.5).  Mean (+ SE) number of sheep observed during the 5 years of surveys was 308 (+ 9.8) animals and there was little variability in this count from year to year (coefficient of variation = 7%).

Figure 1.5.  Numbers of Dall’s sheep observed in 7 survey units within the Interior Alaska Military Operations Areas during summer aerial surveys, 1997-2002.  A survey was not completed in 2000 due to poor weather conditions.

Precision of Dall’s sheep surveys in the Yukon-Tanana Uplands

With the exception of 1998, sheep sightability was good (Table 1.3).  The number of sheep counted during that survey however, was similar to other years (Fig. 1.5). Using the Lincoln-Peterson estimate to evaluate sheep numbers within units surveyed, we estimated a mean (+SE) of 365 (+23.5) sheep within the 7 surveyed units within Yukon-Charley Rivers National Preserve (Table 1.3).  If 1998 is excluded from this analysis, the mean (+SE) estimate becomes 343 (+11.7).  In some instances, radiocollared sheep were outside of survey units during the survey, but were still within the boundaries of Yukon-Charley Rivers National Preserve.  Using these radiocollared animals allowed us to make an estimate of the sheep population within the preserve (Table 1.3).  The mean (+SE) estimate of sheep within Yukon-Charley Rivers National Preserve was 389 (+31.0).  Excluding 1998 from this data sets result in a Yukon-Charley Rivers National Preserve estimate of 360 (+11.0) sheep.   

DISCUSSION

Although considerable variability has existed in the amount of effort spent surveying for sheep in the Yukon-Tanana Uplands, this variability does not appear to have affected the sheep counted in individual units (Table 1.1).  This conclusion is supported by the lack of a relationship between survey effort and number of sheep counted (Table 1.2) as well as stability of the sheep count from 1997 – 2002 (Fig. 1.5).  This is encouraging as it appears that varying survey time based on pilot and observer comfort level allowed for consistent survey results.

No differences in trends in the number of observed sheep between areas mitigated and not mitigated for low-level military aircraft were identified from 1997 – 2002.  It is possible that trends in individual units were obscured by the amount of variation in counts from one year to the next.  Another interpretation of this result is that any changes in trends caused by mitigation measures were realized in the 1st year they were implemented (1995) and current trends were established at that time.  Alternatively, this result can be interpreted as no effect of mitigation measures on Dall’s sheep population trends in the Yukon-Tanana Uplands.  This latter explanation is supported by other evidence presented in this document such as population trends before and after the initiation of mitigation measures at Cirque Lake and Mount Sorenson.

The fluctuations in Dall’s sheep numbers within individual units from 1997 – 2002 was surprising.  Dall’s sheep are typically thought to be highly traditional animals that show great fidelity to specific ranges at specific times of the year (Geist 1971, Nichols and Bunnel 1999).  Results from the aerial surveys and radiotelemetry data (this study and  Burch and Lawler 2001) indicate this is not the case in the Yukon-Tanana Uplands in early July as it appears sheep are regularly moving from one suitable patch of habitat to another.  Therefore, long-term affects of mitigation measures at one site may be difficult to identify due to mixing of sheep from different survey areas.  Alternatively, the mitigation of a small area, such as Cirque Lake, may be inappropriate as sheep in the Yukon-Tanana Uplands appear to be using habitat at a larger scale than the small patches (survey unit) of suitable habitat scattered through out the Yukon-Tanana Uplands or defined by the 7 nautical mile radius mitigation area.  These movements do, however, emphasize the importance of travel routes between habitat patches.  Knowledge of these routes is thought to be traditional with younger animals learning the routes from following older animals (Nichols and Bunnel 1999). Therefore, disruption of these routes would hinder exchange between areas of suitable habitat.  Geist (1975) proposed this as an explanation for the lack of sheep in many apparently suitable areas. 

We did identify a significant long term change in the number of sheep when comparing a mitigated site (Cirque Lake) to a non-mitigated site (Mount Sorenson).  Specifically, the area that isn’t mitigated for low-level military aircraft experienced an increase in the numbers of ewes, lambs and yearlings since the initiation of mitigations measures.  This pattern is the reverse of what one would expect if population levels of Dall’s sheep are positively affected by the presence of the mitigation measures.  This supports the conclusions that there is no measurable effect of mitigation measures on population trends in the Yukon-Tanana Uplands.

The five year trend in Dall’s sheep numbers in the areas surveyed in the Yukon-Tanana Uplands indicates a stable population of sheep.  Due to differences in areas surveyed prior to 1997 it is difficult to know if the current population level is similar to population levels prior to mitigation measures 1995.  Based on values calculated from the Lincoln-Peterson estimate and the confidence intervals constructed around these values, we can have a high degree of confidence that the majority of sheep in the area of interest are occurring in the survey units and that we are seeing the majority of sheep.  Only during 1998 was there a problem with sheep sightability and this was likely due to problems with identifying marked animals (i.e., we missed seeing radiocollars on sheep) verses missing sheep in general because the total sheep count in that year was similar to other years. 

In conclusion, aerial sheep counts provided a relatively consistent measure of Dall’s sheep populations in the areas surveyed.  Although population counts varied considerably within individual units, sheep population levels in the entire study area varied little (Fig. 1.5).  There is no evidence that the current mitigation measures have altered Dall’s sheep population trends in areas mitigated for military overflight activity in comparison to areas not mitigated.  This conclusion is based on Dall’s sheep counts conducted before and after the implementation of mitigation measures as well as population trends following the initiation of the measures.  Movement of Dall’s sheep between units, however, complicate this conclusion as negative effects in one area may be off-set by movement of sheep from less impacted areas.  Finally, mitigation at one site (Cirque Lake) may be difficult to justify as this is not a closed population.  Evidence suggests a large amount of movement of Dall’s sheep between habitat pockets.  Future consideration of Dall’s sheep in this area should treat Dall’s sheep in this portion of the Yukon-Tanana Uplands as 1 population (Fig 1.1).

SUMMARY

We evaluated sheep population levels and trends in areas mitigated for low-level military overflights in comparison to areas with no mitigation.  The sheep population in the Yukon-Tanana Uplands was estimated using aerial surveys.  Areas surveyed (units) were separated by atypical sheep habitat.  From 1997 to 2002, the total number of sheep counted as well as number counted in each sex and age class varied considerably within units and this was true of both the mitigated as well as the unmitigated areas.  No differences in trends in Dall’s sheep numbers within units were identified for those areas mitigated and not mitigated for low-level military overflights.  For two units for which we have an estimate of Dall’s sheep numbers before and after the initiation of mitigation measure, sheep numbers increased at the non-mitigated site following the initiation of mitigation and sheep numbers at the mitigated site stayed relatively unchanged.  Counts of all units did not indicate a significant trend in the sheep population from 1997 to 2002 in theYukon-Tanana uplands and there was little variability in the total count from year to year (coefficient of variation = 7%).  There is no evidence that mitigation measures altered Dall’s sheep population trends in areas mitigated for military overflight activity in comparison to areas not mitigated.

Table of Contents | Chapter 2

2004 Sheep Report
http://www.nps.gov/yuch/Expanded/key_resources/sheep/sheep_2004.htm
Doug Beckstead
December 8, 2004