REMUS was developed by Piotr Weclaw as a part of Master of Science research under the supervision of Dr Robert Hudson at the University of Alberta in the Department of Renewable Resources.
Woodland caribou populations in North America have declined in the last several decades (Bergerud, 1974; Bergerud, 1988; Edmonds, 1988; Dzus, 2001) and caribou range has contracted significantly (Edmonds, 1991). Since 1987, woodland caribou have been classified as threatened under the Alberta Wildlife Act. Current research suggests that industrial development (e.g. roads, oil wells, seismic lines) interacts with natural factors that may lead to population declines (James and Stuart-Smith, 2000; Dyer et al., 2001).
Effects of industry on caribou range act cumulatively with natural factors (mainly predation and forage availability) that shape caribou population dynamics. To understand and mitigate the effects of industry on caribou, cumulative effects assessment for the species is needed and is now mandated for major industrial projects. We developed a computer simulation model that might serve as a tool for such assessment. The aim of REMUS is to discuss simulation of woodland caribou conservation and management in northern Alberta, and to illustrate how computer simulation modeling can be used to assess cumulative effects influencing a threatened population.
The process of building the model and simulations completed in REMUS allowed us to suggest particular management strategies in caribou conservation and to make the following research recommendations:
1) Forage biomass available to caribou. Measurements of the total biomass of arboreal and terrestrial lichens, and vascular plants utilized by caribou are important, yet largely unavailable for the boreal plains of northern Alberta. One should also attempt to assess what percent of the total forage biomass is available to caribou, and finally, what percent of this available biomass can be potentially consumed by caribou (has nutritional value). To recommend mitigation in caribou habitat one needs to know forage carrying capacity of the habitat. Our simulations produced different thresholds for industrial development on caribou range for each level of lichen carrying capacity. These simulations also indicated that forage carrying capacity affects the ability of caribou to deal with predation. Revegetation and regeneration of seismic lines are other factors that would certainly affect forage carrying capacity. Consequently, this is also a key research need.
2) Functional loss of habitat. Simulations suggest that the most detrimental factor to caribou demography is the functional loss of habitat due to caribou avoidance of good quality (food) habitat in proximity of industrial infrastructures. We strongly recommend research on the mechanism governing avoidance by caribou.
3) Low-impact seismic lines. Future research on the effects of low-impact seismic lines on caribou and wolf behavior and ecology (e.g. predator-prey interactions, habitat use) would allow expanding the model by adjusting its structure to account for new industry practices.
4) Post-fire habitats. Examining the importance of fire and burned sites in caribou ecology would be an important contribution to model refinement and caribou conservation.
5) Bear predation. Potential effects of black bear predation on caribou need to be examined in field experiments. Although there is evidence that black bears prey on caribou (Ballard 1994), the information on black bear population dynamics in Alberta is scanty, and the effects of black bear predation on caribou populations are poorly understood.
6) Snow conditions may affect caribou mobility (Schneider at al., 2000) and this could increase caribou vulnerability to predation (Thomas, 1995). Some authors (Kelsall and Telfer, 1971; Adamczewski et al., 1988; Brown and Theberge, 1990; Thomas 1995) point out that snow conditions may play important role in caribou population dynamics in different locations. Yet, there is a lack of data on the effects of snow conditions limiting caribou access to forage in the boreal ecosystem. Research examining how snow cover influences caribou food habits and energetics would facilitate assessment of the impact of winter conditions on caribou demography.
7) Predator – prey interactions. Models attempting to describe predator – prey interactions are based on assumed numerical and functional responses of predators to prey availability. We think that the key elements in wolf population ecology are the functional response, social spacing behavior (territoriality and pack structure), reproductive potential, and prey switching. Although the literature on the above topics has grown in recent years (Messier, 1985; Messier and Crete, 1985; Ballard et al., 1987; Fuller, 1989; Gasaway et al., 1992; Messier, 1994; Eberhardt, 1998; Eberhardt and Peterson, 1999; Hayes and Harestad, 2000a; 2000b; Hayes et al., 2000; Messier and Joly, 2000), the understanding of the aforementioned factors and their interactions is relatively poor and evidence is sometimes inconsistent (Messier, 1994; Hayes and Harestad, 2000a). A better understanding of the processes that determine population dynamics of predators is important. This can be accomplished only by more research in this field. The most crucial questions seem to be: what factors determine the size of a wolf pack, and wolf territories, how prey availability affects wolf reproductive potential, and finally, what factors determine wolf kill rates?
The following management implications arise from our simulation study:
1) Thresholds for industrial development on some caribou ranges in northern Alberta may already be exceeded – model predicts that in areas with 1.22 km/ km2of linear corridors, caribou numbers would be reduced by 50% in 15 years, and the species would be extirpated in less than 40 years.
2) If there is no predation and no avoidance of linear corridors, caribou should be able to deal with very high impact of industry (the threshold is 40 km of linear corridors per 1 km2).
3) The threshold from (2) drops significantly if caribou show avoidance of linear corridors. All management practices that reduce caribou avoidance of industrial infrastructure are of highest importance in caribou conservation.
4) Significant reductions in moose may be more beneficial for caribou than direct wolf control. However, in order to recommend this management action, a spatial model describing distribution of wolves, moose and caribou and accounting for the Spatial Separation Hypothesis (James, 1999) needs to be developed to simulate the effects of relative distribution of the three species on wolf-caribou interaction.
5) Wolf control is not a viable solution in caribou conservation because:
If wolves are reduced to 0.004 /km2 (by more than 50% assuming the present density to be 0.0087 /km2 (Gunson, 1992)) and assuming that there is no human harvest of caribou and bear predation is insignificant, the threshold for linear corridors is 0.8 km /km2. Dyer (1999) reported the density of linear corridors in his study area to be 1.224 km/km2. Consequently, even reduction of wolves by more than 50% would not stop caribou decline on some ranges in northern Alberta;
If wolves are totally excluded from the ecosystem, and assuming that caribou avoid human-built structures as described by Dyer (1999) and Dyer et al. (2001), the maximum hypothetical threshold for linear corridors obtained in the simulations on caribou range is 1.65 km/km2 of seismic lines with no roads and less than that with roads. In some caribou ranges the density of linear corridors is already close to this threshold.
6) We believe it is very important to increase people’s awareness of caribou conservation. By educating and informing people about caribou status, we could reduce human-caused caribou mortality due to hunting/poaching. If people were not perceived as predators caribou would likely habituate to human presence. In such a case increased human activity on linear corridors might be beneficial to caribou by reducing natural predation. It is known that ungulates, if not threatened directly by people, can be attracted by human developments that provide refuge from predators (Banff-Bow Valley Study, 1996).
7) Immediate changes in land use practices and wildlife management actions need to be undertaken. The most beneficial to caribou seem to be:
significant reductions in moose population in areas where caribou populations are being managed;
industry practices that minimize caribou avoidance of human-made structures.
Many people contributed ideas and constructive criticism during the development and testing of this model. In particular, we thank Stan Boutin, Elston Dzus, Cormack Gates and Bob Wynes. We would like to thank Brad Stelfox, Forem Technologies, for helpful comments on the structure of the model and its relationship to ALCES, a widely adopted general landuse simulator. This project was funded by the Boreal Caribou Research Program through the Natural Sciences and Engineering Research Council Collaborative Research and Development Grant.
copyrightăPiotr Weclaw, 2003, 2004
Weclaw, 2003, 2004