Triage for conserving populations of threatened species: The case of woodland caribou in Alberta

Prioritization of conservation efforts for threatened and endangered species has tended to focus on factors measuring the risk of extirpation rather than the probability of success and cost. Approaches such as triage are advisable when three main conditions are present: insufficient capacity exists to adequately treat all patients, patients are in a critical state and cannot wait until additional capacity becomes available, and patients differ in their likely outcome and/or the amount of treatment they require. The objective of our study was to document the status of woodland caribou (Rangifer tarandus) herds in Alberta, Canada, with respect to these three conditions and to determine whether a triage approach might be warranted. To do this we modeled three types of recovery effort - protection, habitat restoration, and wolf control - and estimated the opportunity cost of recovery for each herd. We also assessed herds with respect to a suite of factors linked to long-term viability. We found that all but three herds will decline to critical levels (<10 animals) within approximately 30 years if current population trends continue. The opportunity cost of protecting all ranges by excluding new development, in terms of the net present value of petroleum and forestry resources, was estimated to be in excess of 100 billion dollars (assuming no substitution of activity outside of the ranges). A habitat restoration program applied to all ranges would cost several hundred million dollars, and a provincial-scale wolf control program would cost tens of millions of dollars. Recovery costs among herds varied by an order of magnitude. Herds also varied substantially in terms of their potential viability. These findings suggest that woodland caribou in Alberta meet the conditions whereby triage should be considered as an appropriate conservation strategy.     

Range fidelity: The missing link between caribou decline and habitat alteration?

Conservation of forest-dwelling caribou (Rangifer tarandus caribou) is of great concern across most of its range. Anthropogenic disturbances, primarily logging activities, have been identified as the most important cause of caribou decline, although the mechanisms underlying this decline are not fully understood. Caribou commonly display fidelity to calving sites or seasonal ranges, but the potential role of this life-history trait has been largely overlooked in research and conservation planning. This is surprising because sites and ranges with high inter-annual use should have high conservation value. We investigated the relationship between habitat disturbances and home-range fidelity of forest-dwelling caribou across three study sites in Québec, Canada, using a broad range of natural and anthropogenic disturbances. Between 2004 and 2007, we tracked 47 adult female caribou using GPS collars. Home-range fidelity varied between seasons, being higher during calving and summer, and lower during winter. Caribou reduced fidelity following natural and anthropogenic disturbances, the latter having a stronger negative influence. Anthropogenic disturbances had a strong negative impact on home-range fidelity during annual, summer and winter periods, whereas natural disturbance was the dominant factor during calving. Despite this negative influence on fidelity, caribou tended to demonstrate range fidelity even in study sites most impacted by human activities. Habitat disturbances could produce two possible outcomes for caribou conservation: (1) a trend for females to reduce home-range fidelity which could translate into lower calf and female caribou survival through reduced familiarity with food distribution, escape cover and predation risk and (2) a global tendency to maintain range fidelity even in a drastically modified landscape which could turn into an ecological trap, particularly for calves when predation risk increases due to increased black bear density in early successional forests. Taking range fidelity behavior into consideration during forest management planning could direct conservation efforts toward the best available sites and therefore facilitate caribou persistence in managed landscapes.     

Identifying indirect habitat loss and avoidance of human infrastructure by northern mountain woodland caribou

Understanding the impact of indirect habitat loss resulting from avoidance of human infrastructure is an important conservation priority. We evaluated resource selection for 10 global positioning system collared northern mountain woodland caribou (Rangifer tarandus caribou) in British Columbia, Canada, with seasonal resource selection functions (RSF) developed at the second-order (landscape) and third-order (within home range) scales. To estimate how much habitat was lost due to avoidance, we estimated the zone of influence (ZOI) around multiple developments and modeled realized and potential habitat. Potential habitat was approximated by removing the ZOI from RSF models. By calculating the spatial difference between potential and realized habitat we estimated the amount of indirect habitat loss. Caribou displayed hierarchical avoidance of development, with the greatest avoidance occurring at the second-order. During both seasons caribou avoided high-use roads by 2 km and low-use roads by 1 km. In winter, caribou avoided town by 9 km compared to 3 km in summer. However, in summer caribou avoided mines by 2 km and cabins and camps by 1.5 km, while in winter when human activity was low, avoidance of these features was minor. As a result of avoidance of the cumulative ZOI, approximately 8% and 2% of high quality habitat was lost in the study area in winter and summer, respectively. Our study provides an approach to identify the extent and quality of habitat influenced by indirect avoidance. Conservation efforts should prioritize protecting areas of high quality habitat degraded by avoidance in the vicinity of human development.     

Factors affecting Eurasian lynx kill rates on semi-domestic reindeer in northern Scandinavia: Can ecological research contribute to the development of a fair compensation system?

Semi-domestic reindeer (Rangifer tarandus) are the main prey for lynx (Lynx lynx) in northern Scandinavia. This causes large, but poorly documented, losses of reindeer. Although the compensation schemes differ between Norway and Sweden, there is a legal requirement in both Scandinavian countries that losses of semi-domestic reindeer to large carnivores should be fully compensated. The current level of compensation payment is based on limited data on lynx kill rates on reindeer. The main goal of this study was to quantify lynx kill rates on reindeer across a range of ecological conditions in northern Scandinavia. A total of 35 lynx were monitored during 3667 days. All lynx individuals killed reindeer. However, we observed a high variation in individual kill rates. Kill rates were strongly affected by sex and social status (i.e. females with and without kittens), as well as season and reindeer availability. The highest kill rates were observed in summer for male lynx, and the lowest were observed during winter for solitary lynx at low reindeer density. In summer, several female lynx switched completely to small prey species even at high densities of reindeer. The present political goals for lynx conservation in Scandinavia require that lynx occur in the reindeer husbandry area where reindeer are the only suitable ungulate prey available. Accordingly, it is therefore impossible to have the total prevention of depredation as a management goal. Rather depredation levels must be limited such that they are considered acceptable, and fair compensation payments should be paid. Our study is the first empirical quantification of the magnitude of lynx depredation on semi-domestic reindeer. This is an important step necessary to enable the setting of fair compensation levels and will hopefully lead to a reduction in conflict level concerning the potential impacts that lynx have on reindeer.     

Site fidelity of female caribou at multiple spatial scales

Studies of site fidelity have been hampered by arbitrary designations of spatial scale and the lack of null models for comparison. We generated null expectations of fidelity at different scales from the distribution of radio-tracked animals in a population. We applied the models to space use of satellite-tracked caribou (Rangifer tarandus caribou), the most vagile nonvolant terrestrial animal, from populations representing sedentary and migratory ecotypes. We compared distances between consecutive-year locations of adult females to expectations based on the total range and seasonal range of each population. At the scale of the total range, sedentary and migratory caribou displayed remarkably similar philopatry, despite a 30-fold difference in size of their population ranges, from time of calving (late May) to breeding (late October). The most intense fidelity occurred during post-calving when, on average, sedentary and migratory females returned to as near as 6.7 km and 123 km, respectively, of locations occupied the previous year. At the scale of the seasonal range, the ecotypes differed. Sedentary caribou still displayed fidelity from calving to breeding; migratory caribou exhibited fidelity only during late autumn. For migratory, but not sedentary caribou, inter-year distances during winter were negatively correlated with age, implying that older females were more philopatric. We conclude that reproductive activities delimit the season of fidelity of female caribou of both ecotypes, and that scale-dependent ecotypic differences in fidelity may reflect different factors of population limitation. A spatially-explicit approach to site fidelity is essential for synthesizing patterns across studies.     

浅谈中国驯鹿种群

中国驯鹿有较高的经济价值和开发前景,是当地鄂温克民族赖以生存的物质基础。为了有效保护和利用这一珍贵资源,针对饲养技术落后、管理粗放而导致中国驯鹿种群品质下降、数量减少等方面的问题,采取相应对策。     

Genetic diversity and relatedness of boreal caribou populations in western Canada

We studied genetic diversity within and gene flow among six ‘threatened’ populations of boreal caribou (Rangifer tarandus caribou) inhabiting Alberta and British Columbia, Canada. Mean expected heterozygosity (H_E) across 11 loci spanned a narrow range between 0.74 and 0.79. Estimates of H_E were in the mid to high range of those typically observed in natural populations of large mammals, including caribou, and were not suggestive of any immediate threat to survival. We concluded that recent anthropogenic fragmentation of caribou range in western Canada has yet to affect genetic diversity of populations. Analysis of population structure identified a region of relatively low gene flow corresponding with the valley of the Peace River. The highest value of F_ST observed between populations on the same side of the river was 0.025, whereas the lowest value that spanned the river was 0.044. Confirming this result, an assignment test demonstrated that 96.5% of animals could be assigned to the correct side of the Peace River, though only 66.0% of animals could, on average, be assigned to populations of actual origin. Taken as a whole, our results support the existence of two discrete metapopulations bisected by the Peace River, within each of which there exist multiple populations, or at least multiple regions, which experience considerably higher levels of interchange. For caribou inhabiting the boreal plains, large rivers such as the Peace and Mackenzie may serve as biologically meaningful boundaries for managing metapopulations.     

Effects of a power line on migration and range use of wild reindeer

Linear infrastructures such as roads, pipelines and power lines can hinder Rangifer migration and reduce the total amount of area available for foraging. We studied the barrier and aversion effect of a 66 kV power line transecting the range of wild reindeer in North Ottadalen, south central Norway using aerial surveys of reindeer distribution (direct measurement of reindeer use) and lichen measurements (indirect measurement of reindeer use) at varying distances from both sides of the power line. We present a clear definition of the expected barrier and aversion effects, and present specific hypotheses and predictions based on these two terms that are often used unclearly in the literature. The aerial surveys and ground observations showed that reindeer crossed underneath and grazed under and on both sides of the power line during 14 out of the 22 years (63.6%) surveys were conducted over the last 31 years. This behaviour was confirmed by the lichen measurements, indicating a higher use of lichen pastures along ridges close to and under the power line compared to those at increasing distances and up to 3 km from both sides of the power line. This winter grazing pattern probably reflects the topographical channeling of reindeer to the ridges along an 8 km wide topographical corridor encompassing the power line and cannot be attributed to the power line itself. We found that direct measurement data (visual observations such as ours or GPS/telemetry tracking) provide less ambiguous information for testing barrier and aversion effects than indirect measurement data (pasture/lichen measurements) that are likely influenced by numerous uncontrollable environmental variables independent of reindeer grazing. Our results contrast with recent studies indicating strong barrier and aversion effects of similar power lines for Rangifer migration and grazing behaviour in alpine terrain.     

Progressive impact of piecemeal infrastructure development on wild reindeer

We made monthly surveys of >2000 reindeer (Rangifer tarandus tarandus) from 1977 to 1987 before and after the construction of the Blue Lake hydroelectric reservoir in order to study the progressive impact of infrastructure development on wildlife. Following this development, reindeer densities within a 4-km radius declined gradually during winter to 8% of pre-development densities without significant changes in undeveloped control sites. During summer, reindeer gradually reduced use of areas within 4 km distance from roads and power lines to 36% of predevelopment density, with subsequent 217% increase in use of the few remaining sites located >4 km from infrastructure. Reindeer reproduction declined progressively as habitat was lost. Assessments of individual development projects seriously underestimate the long-term effects of the entire expanding infrastructure network. Piecemeal development has resulted in an estimated 70% loss of undisturbed reindeer habitat across the last century in Norway. During this time the reindeer population has become fragmented into 26 isolated sub-populations. Our results show that any further infrastructure development will put the remaining European population of wild mountain reindeer at great risk, as further habitat fragmentation will make the remaining undisturbed patches too small for holding viable populations. A general model of impacts of infrastructure on wildlife is presented. We discuss the importance of controlling piecemeal development in infrastructure for conservation of wildlife and argue that minimizing infrastructure development is likely one of the largest challenges in wildlife conservation ahead.     

Arboreal lichen in uncut and partially cut subalpine fir stands in woodland caribou habitat, northern Idaho and southeastern British Columbia

To better understand the effects of partial cutting on arboreal lichen production within woodland caribou (Rangifer tarandus caribou) habitat, lichen was hand picked from 1228 branches on 307 subalpine fir (Abies lasiocarpa) trees in Idaho (ID), and in British Columbia (BC). Lichen biomass from partially cut stands was compared with biomass on trees from adjacent uncut stands at each site. Arboreal lichen biomass did not differ significantly between uncut and partially cut stands. Total number of branches per tree did not differ significantly between uncut and partially cut stands. Live branches had more lichen than dead branches. Species composition of arboreal lichen changed in partially cut stands compared with adjacent uncut stands. The ratio of live to dead branches was substantially different within the BC partial cut.