Moose ( Alces alces ) population size and density in the Inuvik Region of the Northwest Territories , Canada

Responding to community concerns, the Gwich’in Renewable Resources Board (GRRB) and the Government of the Northwest Territories Department of Environment and Natural Resources (ENR) conducted an aerial moose (Alces alces) survey in the Inuvik region of the Northwest Territories, Canada to estimate moose density and distribution. !e survey was "own in March 2011 and a random strati$ed sample design was used. Local knowledge was incorporated in to the strati$cation of survey cells. Moose density in survey blocks ranged from 9.66 moose/100 km in the Ikhil Pipeline block to 0 in the Peel River block with a coarse overall moose density 2.24 moose/100 km. Densities found were low but within expected range for the species in this region of North America based on past surveys.


Introduction
In the Gwich'in Settlement Area (GSA) and the adjacent Inuvialuit Settlement Region (ISR) of the Northwest Territories (NWT), Canada (Fig. 1), management of moose (Alces alces) populations is primarily the responsibility of co-management boards and of the territorial government.e Gwich'in Renewable Resources Board (GRRB) is the co-management board for wildlife in the GSA while the Wildlife Management Advisory Council (Northwest Territories) [WMAC (NWT)] is the co-management board for wildlife in the ISR.
A 2006 survey in GSA reported low and declining (from past surveys between 1980 and 2000) moose densities ranging from 0 to 3.78 moose/100 km 2 (Lambert, 2006).Moose in the ISR have not been surveyed since the mid-1980's (Jingfors & Kutny, 1989).Current local knowledge suggests moose numbers have increased in the Mackenzie Delta (Fig. 1).Local barren-ground caribou (Rangifer tarandus groenlandicus) herd population numbers have been low, in particular the Cape Bathurst herd, resulting in a harvest closure of that herd in 2007.So, despite a perception of current healthy moose population the declining caribou led to community concerns about impacts of possible predators and harvesters switching to moose.In response to these concerns and in order to inform possible management decisions by GRRB and WMAC (NWT), we conducted

Methods
Population estimates followed the strati ed random sampling methods of Gasaway et al., (1986) and was analyzed using the GeoSpatial Population Estimator Software (Delong, 2006).Density estimates (moose/100 km 2 ) were calculated for each of the eight survey blocks based on total number of moose sighted in selected cells (# moose/area surveyed * 100).We held workshops with local Renewable Resources Councils (RRCs) and Hunter and Trapper Committees (HTCs) to de ne the survey region and map areas of expected high and low moose density in the survey period.Local experts were used as a cost-e ective way to stratify the survey area while ensuring the involvement of local indigenous people.
Aerial survey methods generally followed those described by Kellie & Delong (2006).e survey region was divided into 2' latitude by 5' longitude (~ 4 km by 4 km) cells using ArcGIS 9.2 (ESRI, 2006).e cells were then strati ed as high or low moose density using workshop classi cations and habitat data.If cells were not classi ed as high or low moose density during the workshops vegetation cover maps were used to classify the cells.Area classi ed as open deciduous, closed deciduous, shrubs, wet herbaceous, emergent vegetation were considered areas were high density of moose would be expected.Areas with closed needle leaf, open needle leaf, non-vegetated soil, sparse vegetation or rock/gravel were considered low moose density classes.
Eight areas of interest were identi ed based on past surveys and input from HTCs, RRCs, GRRB, and WMAC (NWT) (Fig. 1).Survey blocks in the GSA were similar to the 2006 survey with slight modi cations to the Peel River and Arctic Red River survey blocks based on input from the RRCs. is includes adjusting the Arctic Red River survey block (Fig. 1) into a discontinuous block with a portion near the community of Tsiigehtchic and a portion up river.New survey blocks were created in the ISR.Cells were randomly selected for surveying, with 2% of cell selections made manually to ensure good coverage, such as the inclusion of both high and low survey blocks.Surveyed cells represented 4,368 km 2 and 16.1 % of all survey blocks (Table 1).
We ew the survey in March 2011 using a Cessna 206 and Cessna 185 xed-wing aircraft.Surveyed cells were to be covered in their entirely with the intent to detect all moose in the cell.Search intensity varied by block based on block vegetation cover; heavily treed areas were covered more intensely than open/tundra areas.Snow tracks were circled to determine if the moose was still located in the block.A pilot, navigator, and two observers spotted and classi ed moose inside each selected cell and noted any moose observed outside selected cells.Locations were recorded using GPS.Wolves and other wildlife observations inside or outside selected cells were also recorded.

Results & Discussion
Survey ights were conducted from March 16-24, 2011 with a total of 61.9 hours own.We observed a total of 168 moose: 79 within surveyed grid cells and 89 moose outside surveyed cells.We classi ed 63% of observed moose: 40 cows, 32 calves and 34 bulls, resulting in bull to cow and calf to cow ratios of 85:100 and 80:100, respectively.Composition estimates may be biased as the presence of calves aided in classi cation of cows, such that cows without calves may have been more often unclassi ed than by chance.Other wildlife observed included; moose, 33 sheep, 38 wolves, and ve caribou.
A total area of 3519 km 2 was surveyed making the coarse overall moose density 2.24 moose/100 km 2 .e highest density, 9.66 moose/100 km 2 , was found in the Ikhil Pipeline survey block (Table 1).
e Arctic Red River block was found to have very low densities with 0.53 moose/100 km 2 .No moose were observed in the Peel River block; however we do not believe there are no moose in the area, as there were tracks observed.e Peel River block was the smallest survey area and we believe that the sample size was too small to detect moose at the low densities they occur in this area.e densities found in the other blocks were: 4.49 moose/100 km 2 in Delta North, 1.08 moose/100 km 2 in Kugaluk-Miner Rivers, 1.94 moose/100 km 2 in Inuvik-Tsiigehtchic, 3.33 moose/100 km 2 in the Mackenzie Gas Pipeline Route, and 2.23 moose/100 km 2 in the Richardson Mountains (Table 1).
A sightability correction was not determined for our survey.Moose sightability varies by season, snow cover, habitat, and size of the survey unit (Gasaway et al., 1986).Habitat in the study region ranged from alpine/tundra to semi-open coniferous forest with sightability higher in more open habitats.It is possible to estimate a sightability correction factor using radio-collared moose (Gasaway et al., 1986).However, since there were no collared moose in our study area we could not obtain a correction factor.
e Delta North and Ikhil Pipeline sur-  Kutny, 1989) which is higher than the density found in this survey (1.08 moose/100 km 2 ).It is not known if this is a real trend because we only have two data points that were obtained using di erent survey methods.
e survey block area in 2011 was not as large as the 1988 survey.Compared to the 2006 survey in the GSA, the Richardson Mountains and Peel River blocks were found to have lower densities in 2011.e densities for the Richardson Mountain block were also lower than a 2000 helicopter survey of the Richardson Mountains that included the Yukon where a density of 4.8 moose/100 km 2 was found (Yukon Government, unpublished data).Methods for the 2000 survey were quite di erent, as optimal habitat was own instead of randomly sampling areas.
e Inuvik-Tsiigehtchic, Mackenzie Gas Pipeline route, and Arctic Red River survey blocks had higher densities in 2011 compared to the 2006 survey (Table 1).
Moose in the ISR and GSA are at the northern edge of their range.As such, environmental factors and range conditions may partially explain observed lower densities than in other portions of the species' range.Observed densities appear generally consistent with those reported for other subarctic regions.Franzmann & Schwartz (1998) summarized general densities (moose/100 km2) across the species range as < 12 in subarctic areas, 12-31 in better ranges, and 40-100 in excellent ranges.
Densities in the Inuvik Region appear lower than other areas of the NWT, except the North Slave region where density ranged from 2.0 to 3.5 moose/100 km2 in 2005 (Clu , 2005).
e highest densities recorded in the NWT have been 17 moose/100 km 2 around Fort Good Hope and Norman Wells (Maclean, 1994;Veitch et al., 1995).
Participation of knowledgeable community members and harvesters in the strati cation of survey areas was important to improve accuracy of population estimates.Natural low densities make it di cult to detect trends between surveys.Composition estimates could be improved if surveys were conducted before moose bulls shed their antlers.We advocate that information on moose habitat, recruitment, and mortality, as well as increased coverage of future surveys would help to increase precision and con dence of estimates and would help to explain changes in moose distribution, density and number.

Table 1 .
Number (#) of moose observations, moose densities, and population estimates by survey block.
1 Standard Error.2Couldnot compute estimate -insu cient samples in one stratum.