Modeling the Ecological Factors Affecting Northern Caribou in the Omineca Region, British Columbia

Study area location within British Columbia, Canada. INTRODUCTION This study was initiated to examine the ecological factors affecting northern caribou in central British Columbia. A large proportion of the Mackenzie Timber Supply Area (TSA) is comprised of high value caribou habitat, and contributes significantly to the provincial supply of timber. These potentially conflicting resource uses are of particular concern because caribou numbers demonstrated a decline following the range expansion by moose in the early 1990's into central B.C. (Heard and Vagt 1998). Increases in moose numbers are most pronounced where logging has created early seral habitats, providing higher supplies of vegetative browse species. This increase in primary prey density can lead to a larger wolf population, which does not exhibit a negative feedback to declining numbers of caribou (Bergerud et al. 1984, Seip 1992). This relationship can be expressed as:

In attempts to alleviate the interaction with moose and wolves, the Caribou Management Strategy (CMS) for the Mackenzie TSA recommends large patches of cutblocks, ranging from <40 to 5000 hectares, as well as large patches of unharvested areas, where caribou may be less susceptible to predation (Table 1). Relative to the current Forest Practices Code (FPC) (i.e. smaller patches <40-250 ha. - and closely juxtaposed), this strategy is thought to better mimic the natural disturbance patterns found on this landscape and to minimize fragmentation. The dominant natural disturbances in this area are characterized as wildfires of moderate size (20-1000 ha.) with mean return intervals of 200 years.

Terrestrial lichens such as Cladina rangiferina are the primary forage supply for Northern Caribou.

We developed a monitoring program to document the habitat use patterns and population parameters of caribou, moose and wolves as implementation of the CMS proceeds. This data will be incorporated into a CARIBOU MANAGEMENT MODEL, allowing us to test the above mentioned relationship between these 3 species. For example, if caribou populations decline, we expect to be able to attribute this to increased predation risk. We will also explore alternative hypotheses for population declines, such as habitat loss.

We will develop a model to forecast the distribution of potential caribou habitat across a landscape unit. Our model will be based on mapped ecological attributes and the five data sources we are currently collecting:

1. TELEMETRY:

Locations from 217 GPS and VHF radio-collared caribou, moose and wolves will provide information about spatial overlap between these species in relation to landscape patterns.

2. MORTALITY:

Investigations of mortality causes and their spatial distribution will help us assess the relationship between caribou, moose, and wolves.

3. HABITAT:

Seasonal assessments of habitat used by caribou will assist us in identifying habitat attributes important for caribou throughout the year.

4. MAP DATABASES:

Click here to view larger image (157K jpeg).

Attributes from map databases can be queried to determine the habitat use patterns of caribou, moose, and wolves.

5. CENSUS:

Aerial census of moose and caribou will provide information about population trends and calf:cow and bull:cow ratios.

We are now entering year 2 of data collection, therefore the development of a full scale model would be premature. We developed a simplified version to demonstrate the potential of future models. Based on current knowledge, we assumed that suitable early winter caribou habitat is characterized by stands of low site index (i.e dry), dominated by lodgepole pine, 60 - 140 years in age, which supports an abundance of terrestrial lichens (the primary forage for caribou). We selected the Klawli landscape unit within the Mackenzie TSA for it's relatively unharvested condition and because it had an approved forest development plan available, which has yet to be implemented (Fig. 1). The original development plan was modified to incorporate the recommendations of the CMS (Fig. 1, Inset A). We modeled the projected distribution of seral stages and suitable caribou habitat before and after implementation of the harvest plan. We then compared the distribution of early seral stands and patch sizes before and after harvest, relative to the BC Forest Practice Code and the CMS recommendations (Table 1). For areas where substantial amounts of suitable caribou habitat were removed, we examined the age structure breakdown of dry pine stands to compare areas of young stands that will become suitable habitat, replacing older stands that will become unsuitable.

As expected, the distribution of early seral stands changed substantially after a total harvest of 5000-7500ha (12,000-18,500 acres), however, caribou habitat was less influenced (Fig. 1). Note the relatively unfragmented condition of the landscape under the CMS compared to the original development plan (Inset A).

Figure 1.	Pre- and post-harvest distribution of early seral stands and caribou habitat in the Klawli Landscape Unit.  Click here to view larger map (172K jpeg)

In this landscape, there are few old growth stands, and the seral stage distributions differed substantially from the target distributions (Table 2).

Table 2. The seral stage distribution before and after harvest relative to the targets recommended by the B.C. Forest Practices Code (FPC).     % of Forested Area Within Landscape Seral Stage Age Range Pre-Harvest Post-Harvest FPC Targets Early <40 4.3 12.2 <27 Immature 40-120 44.7 41.3 N/A Mature 121-250 49 44.6 >29 Old >250 2.1 1.9 13

The majority of the land base (79%) falls within the largest patch size category (Table 3). This reflects the large natural disturbance patterns that have shaped the landscape. Both the existing and post-harvest patch size distributions differed significantly from the distribution recommended by the FPC (Table 1).

Table 3. Patch size distribution as percentages of total forested land before and after harvest. Patch Size Pre- Harvest Total (ha) Early Immature Mature Old Pre-Harvest Post-Harvest <40 0.3 1.8 1.1 0.6 3.8 5.7 40-80 0.2 0.9 0.5 0.2 1.8 2.5 80-250 0.9 2.0 2.2 0.8 5.9 8.0 250-1000 0.4 2.8 5.3 0.3 8.8 15.2 1000-5000+ 2.8 38.7** 38** 0.0 79.5 68.2 **Each star represents 1 polygon >5000ha

While very little caribou habitat was affected by this harvest, there was 1 location where a large patch of caribou habitat was removed (Inset B). We determined that the majority of the area cut was between 81-100 years old (Table 4). In addition, there is far less young habitat to replace old caribou habitat.

Table 4. Distribution of the dry lodgepole pine ecotype (ha) across age classes.   Age Class (Years)   41-60 61-80 81-100 101-120 121-140 Pre-Harvest 187 730 13,123 4,253 920 Harvested 0 0 1,925 719 43 Post-Harvest 187 730 11,199 3,533 877

If the assumptions implied by definitions of patches, seral stages, and caribou habitat could be accepted, management has provided us with some valuable insights already.

Considerable time and modification of the existing distributions of seral stages and patch sizes would be required to reach target distributions. This raises the question “are these targets appropriate, or should they reflect the pre-harvest distribution?”

Implementation of the CMS has led to several potential advantages:

• Fewer roads across the landscape
• Less fragmentation
• Larger unharvested patch sizes

Where caribou habitat was removed, the long-term impact remains equivocal. This habitat had the potential to remain caribou habitat for an additional 40-60 years, and perhaps the patch should have been located elsewhere to prevent fragmentation of the existing habitat. Alternatively, harvesting the area now will contribute to a more even supply of caribou habitat in the long-term. This reveals the importance of considering stand age and the potential for young stands to replace older stands as suitable caribou habitat.

There are many questions about management of caribou and caribou habitat that need to be addressed. We are taking an adaptive approach to answering such questions by applying multiple forest harvest tactics (different cutblock size ranges are currently proposed for 3 different herd areas), by developing descriptive management models, and by implementing monitoring programs which we will use to assess and refine the management model and the forest harvest tactics through time.

This Forest Renewal BC-funded project was implemented by Slocan Forest Products, Ltd., Donohue Forest Products, Inc., and Pacific Slope Consulting.

Bergerud, A.T., Buttler, H.E. and Miller D.R. 1984. Antipredator tactics of caribou: dispersion in mountains. Can. J. Zool. 62:1566-1575.

Heard, D.C. And K.L. Vagt. 1998. Caribou in British Columbia: A 1996 status report. Rangifer, Special Issue No. 10: 117-123.

Seip, D.R. 1992. Factors limiting woodland caribou populations and their interrelationships with wolves and moose in southeastern British Columbia. Can. J. Zool. 70: 1494-1503.