The expansion of human settlements and land developments have been implicated in the decline of caribou Rangifer tarandus across their geographical range, but the extent to which specific activities affect their demography and movements is still disputed (Klein 1980, Bergerud et al. 1984, Miller & Gunn 1985, Cronin et al. 1998, Nellemann et al. 2003, Noel et al. 2004). Direct links between caribou declines and human activities or development are often difficult to isolate, but some intermediate and long-term impacts of disturbance have been reported.

Reduction in the abundance of caribou in the vicinity of disturbed areas has been documented repeatedly, often with an avoidance zone of 1-5 km (Cameron et al. 1992, Smith et al. 2000, Dyer et al. 2001, Nellemann et al. 2001, Vistnes & Nellemann 2001, Mahoney & Schaefer 2002, Cameron et al. 2005). Females with calves are less tolerant and more likely to avoid disturbance, than are other caribou (Roby 1978, Cameron et al. 1979, Smith & Cameron 1983, Dau & Cameron 1986, Chubbs et al. 1993, Nellemann & Cameron 1998, Nellemann et al. 2001), and caribou are most sensitive to disturbance during the calving season (Cameron et al. 1979, Cameron et al. 1992, Dyer et al. 2001, Nellemann et al. 2001). Group size also affects the response to disturbance. Large groups usually avoid disturbed areas more than smaller groups (Cameron et al. 1979, Curatolo & Murphy 1986, Murphy & Curatolo 1987, Nellemann et al. 2003). Furthermore, the timing and direction of migration may be altered by disturbance (Cameron & Whitten 1980, Curatolo & Murphy 1986, Mahoney & Schaefer 2002).

While isolated developments, affecting a small fraction of a population's range, seem to have few obvious effects on herd demographics, progressive industrial activity and habitat alteration can exert a serious cumulative effect, including reduction in available range and population size (Nellemann & Cameron 1998, Vistnes et al. 2001), decreased reproductive rates and population fragmentation (Nellemann et al. 2003).

Perhaps the greatest impact of development on caribou has come from indirect effects of disturbance. Greater access to caribou created by linear corridors, such as roads and seismic lines, has led to increased harvest by humans (Bergerud 1974, Edmonds 1991, Rettie & Messier 1998) and predation (James 1999, James & Stuart-Smith 2000, James et al. 2004) which are suspected of contributing to caribou population declines. Widespread logging in western Canada has increased the availability of early successional stages of vegetation, increasing range suitability for moose Alces alces, creating an alternate prey base for wolves Canis lupus, and possibly large reductions in caribou numbers (Edmonds 1991, Seip 1991, 1992, Wittmer et al. 2005).

Newfoundland has been the site of substantial habitat alterations, and the effects of logging, hydroelectric development and road construction on caribou have been addressed in several studies (Mercer et al. 1985, Chubbs et al. 1993, Mahoney & Schaefer 2002); these results broadly concur with patterns observed elsewhere. Although the impacts of such industrial activities are well studied, little is known about the effects of mining on caribou and other ungulates.

In this paper, we report on the effects of the Hope Brook gold mine on a portion of the La Poile woodland caribou herd in southwestern Newfoundland. This herd, one of the largest on the island, was the subject of a comprehensive study during 1985-1991, which examined caribou distribution in the vicinity of the mine site before disturbance, after the initiation of construction and throughout the operation of the mine during five seasons. We predicted that the disturbance created by the development and operation of the mine would result in a shift in the distribution of caribou away from the mine site, and we expected this response to be evident in all seasons.

Material and Methods

Study Area and Caribou Herd

The Hope Brook gold mine occupied an area of about 2 km², in the southwest corner of Newfoundland, within an important winter and calving/post-calving habitat for the La Poile caribou herd (Fig. 1). Our study focused on the impact of mine development and operation on a variable section of the La Poile caribou herd located within a 195 km² survey block centered on the mine site.

Figure 1

Study area and location of the Hope Brook gold mine within the La Poile caribou herd range boundary.

The total range of the La Poile caribou herd is approximately 7,000 km² (Mahoney 2000a) and is predominantly barren land interspersed with partially forested river valleys, rocky outcrops, bogs, fens and ponds. The coastline is generally rugged, and inland the region shows gentle relief with summits rising to 650 m above mean sea level and average elevations of 300 m a.s.l. In the barren regions, heath vegetation is characterized by alpine azalea Loiseleuria procumbens, diapensia Diapensia lapponica, pink crowberry Empetrum eamesii and sheep laurel Kalmia angustifolia. Forested areas are composed of black spruce Picea mariana, balsam fir Abies balsamea and occasionally yellow birch Betula lutea.

The climate is strongly influenced by the ocean, particularly in the south (0-20 km inland; Maritime Barrens) where elevations are generally below 300 m a.s.l. Average annual temperature is about 7.5°C with mean minimum and maximum temperatures of around 2.0°C in February and 18.1°C in August. The mean duration of the frost-free season is 157 days near the coast and 101 days inland. Total annual snowfall is 266 cm near the coast and 339 cm inland (Mahoney 2000b).

Based upon 16 stratified random block, transect and mark-recapture surveys, the La Poile caribou herd increased from 500 animals in 1960 to a high of 11,210 in 1988, falling to 4,300 in 2004 (Mahoney 2000a, S.P. Mahoney, unpubl. data). Moose, the only other ungulate in the study area, are widespread in forested areas. Wolves Canis lupus have been absent from Newfoundland since about 1922, but lynx Lynx canadensis are present in low abundance and black bears Ursus americanus are common. Both species prey on caribou calves (Mahoney et al. 1990). Coyotes Canis latrans reached the island of Newfoundland in 1985 but were not sighted in the area, and no coyote kills were reported during the course of our study. More detailed descriptions of the climate, soils, flora and fauna of southwestern Newfoundland can be found in South (1983).

Mine Development

Construction of the Hope Brook gold mine began in July 1986, and the mine reached peak daily operating capacity in April 1991. In May 1991, the mine was closed due to economic and environmental problems. Royal Oak Mines Incorporated bought the property in April 1992, and commercial operations were resumed in July 1992 and continued until 1997. Our study focused on the period of the mine's history during 1985-1991 (Table 1).

Table 1

Time line of the development and operation of Hope Brook gold mine during 1985-1991. Mine activities were categorized into three phases of mine activity. Numbers and dates of flight surveys included in each season are presented for each phase of mine activity.

Results

Caribou distribution, in the pre-disturbance phase, was not related to distance from the mine centre in either season (Fig. 2). However, the number of caribou increased with distance during construction in all seasons, and this trend continued during the operation in all seasons except late winter (see Fig. 2). Between 51 and 92% of the variation in caribou distribution was accounted for by distance from the mine centre. There was an overall reduction in the number of caribou, within the 100-km² study area, as the mine activity progressed. During late winter, pre-calving and calving seasons, the number of caribou within 6 km of the mine centre declined progressively from pre-disturbance to operation, and only 17-27% of the number of caribou seen during pre-disturbance remained in the study area during operation in these seasons (see Fig. 2)

Figure 2

Relationship between caribou abundance and distance from the mine centre for each of the five phases of mine activity. Regression lines are shown for significant (α = 0.05) relationships only. The numbers of caribou (N) observed within the study area (100 km²) are averaged for each phase of mine activity within each season.

Group size increased with distance from the mine site during construction in the autumn (R² = 0.54, P = 0.015), pre-calving (R² = 0.48, P = 0.032) and calving seasons (R² = 0.75, P = 0.001), but not in the winter or late winter seasons (R² ≤ 0.02, P ≥ 0.7). There was no relationship between distance and group size in either season during operation R² ≤ 0.21, P ≥ 0.18), but the average group size, within the 100-km² study area, decreased from pre-disturbance to operation during the late winter, pre-calving and calving seasons (Fig. 3).

Figure 3

Changes in caribou group size (mean ± S.E.) as a function of the five phases of mine activity for each season. The letters above the bars represent significant differences between phases using Tukey's post-hoc tests.

In comparison with the pre-disturbance phase, most caribou avoided the area within 4 km of the mine centre during mine construction and operation (Table 2). This behavioural response was most evident during the winter, late winter and pre-calving seasons. During the autumn and calving seasons, fewer caribou occupied the area within 4 km during construction than during operation, but there was no difference in the distribution of caribou across the two distance zones between pre-disturbance and operation (see Table 2).

Table 2

Percentage of caribou distributed between the two distance intervals 0-4 km and 4-6 km from the mine centre. χ² was performed comparing caribou distribution during the pre-disturbance (PD) phase to their distribution during the construction (C) and operation (O) phases.

Discussion

Opportunities to quantitatively assess the impact of mining activities on ungulates have been limited (Oehler et al. 2005). To our knowledge, this is the first study to investigate the effect of mining on the distribution of woodland caribou, over multiple seasons, using pre-development control data.

Overall, mining affected the distribution of caribou. Prior to initiation of the Hope Brook gold mine development, caribou were dispersed throughout the study area, but during construction and mine operation, the number of caribou increased linearly with distance from the mine in all five seasons. Furthermore, there was avoidance or diminished use of the area within 4 km of the mine centre, which confirms the findings of numerous studies, which have reported lower abundance of animals near human disturbance, often with a diminished range use or avoidance within 4 km of the disturbed area (Nellemann & Cameron 1998, Vistnes & Nellemann 2001, Mahoney & Schaefer 2002, Nellemann et al. 2003, Cameron et al. 2005).

We provided evidence for the negative impact of mine disturbance on caribou distribution; however, evidence for demographic responses to mine activity was limited. While there was a decline in the number of caribou within the 100-km² study area as mine activity progressed during the late winter, pre-calving and calving seasons, this local decline was confounded by a wider population-level decline which was likely due to negative density-dependent effects (Mahoney 2000a, S.P. Mahoney & J.N. Weir, unpubl. data). The degree, if any, to which the Hope Brook gold mine may have precipitated these changes in the La Poile herd was unknown, but any loss of available habitat to a population experiencing food shortages is a cause for concern.

In addition to distributional changes, we detected changes in the structure of caribou groups with mine activity. Average group size, as far as up to 6 km from the mine centre, declined as mine activity progressed and increased with distance from the mine centre during the autumn, pre-calving and calving seasons. It is difficult to know whether the observed decline in group size, as the mine activity progressed, was an artifact of a reduction in the number of caribou in the study area, or whether large and small groups differentially avoided the mine area. Other researchers have reported that larger caribou groups avoided disturbed areas more than small groups (Cameron et al. 1979, Curatolo & Murphy 1986, Murphy & Curatolo 1987, Nellemann et al. 2003), and that the small groups, which occupy areas in the vicinity of the disturbance, are usually composed of bulls and yearlings who are more tolerant of disturbance (Cameron et al. 1979, Smith & Cameron 1993, Chubbs et al. 1983, Nellemann & Cameron 1998, Nellemann et al. 2003). Although we did not classify group composition during survey flights, researchers conducting the reconnaissance flights and mine workers generally reported an absence of females with calves near the mine site.

Mahoney & Schaefer (2002) called for land and wildlife managers and policy developers to consider a likely habitat loss of 1-4 km following industrial developments. The Hope Brook gold mine resulted in a direct habitat loss of 1.78 km² due to actual mine infrastructure and gravel overlay of vegetation during mine construction. However, by assuming a circular avoidance radius of 4 km from the mine centre, the construction and operation of the mine resulted in direct and indirect habitat loss of 50 km².

Although the loss of 50 km² from an annual home range of 7,000 km² may seem insignificant, it may have important ramifications depending on the location of the loss (i.e. if it is along a migration route, the impact may be substantial) and can contribute to cumulative effects of development. While significant localized impacts of hydroelectric development, forestry and road construction on caribou abundance, distribution and migration patterns have been reported for Newfoundland caribou herds (Mercer et al. 1985, Chubbs et al. 1993, Mahoney & Schaefer 2002), no study has addressed the cumulative effects of these industrial activities. However, cumulative and progressive development has led to significant habitat loss and fragmentation, resulting in reduced reproductive rates and increased predation risk for woodland caribou in Alaska and Alberta (Nellemann & Cameron 1998, James 1999, James & Stuart-Smith 2000, James et al. 2004) and wild reindeer in Norway (Vistnes et al. 2001, Nellemann et al. 2003).We argue that future environmental impact assessments of proposed industrial development must focus on the cumulative effects of the proposed activity and historical habitat alterations. This is especially relevant for woodland caribou given that most herds are experiencing population declines (COSEWIC 2002, Mahoney 2000a).

Although we focused on the effect of mine disturbance, Newfoundland may provide a unique environment for future caribou disturbance studies. A major argument explaining caribou avoidance of linear industrial features, such as roads and seismic lines, is that linear disturbances may enhance the ability of wolves to prey on caribou (James 1999, James & Stuart-Smith 2000). Because wolves were extirpated, Newfoundland is an ideal place to test predictions of habitat avoidance in response to linear disturbance in the absence of wolves. We would expect results to differ greatly from areas where caribou are limited by wolf predation, if caribou truly avoid these features because of increased predation risk.