|Product dimensions:||6.00(w) x 9.00(h) x 1.00(d)|
About the Author
David S. Maehr is assistant professor of conservation biology in the Department of Forestry at the University of Kentucky and author of The Florida Panther (Island Press, 1997).
Reed F. Noss is a consultant in conservation biology, past editor of the journal Conservation Biology, and president of the Society for Conservation Biology (1999 - 2001). He is the author of The Redwood Forest (Island Press, 2000), The Science of Conservation Planning (Island Press, 1997) and Saving Nature's Legacy (Island Press, 1994). He also wrote the foreword for Restoring Diversity (Island Press, 1996).
Jeffery L. Larkin is a post-doctoral scholar at the University of Kentucky.
Read an Excerpt
Large Mammal Restoration
Ecological and Sociological Challenges in the 21st Century
By David S. Maehr, Reed F. Noss, Jeffery L. Larkin
ISLAND PRESSCopyright © 2001 Island Press
All rights reserved.
Is the Return of the Wolf, Wolverine, and Grizzly Bear to Oregon and California Biologically Feasible?
CARLOS CARROLL, REED F. NOSS, NATHAN H. SCHUMAKER, AND PAUL C. PAQUET
Carnivores are indicators of ecosystem function and can serve as keystones in the top-down regulation of ecosystems (Terborgh et al. 1999). Although the strength of top-down processes varies widely among species and ecosystems (Noss et al. 1996), it is probably more prevalent than many ecologists have assumed (Terborgh et al. 1999; Crooks and Soulé 1999). Wide-ranging carnivores may serve as "bioassays" of emergent landscape characteristics such as connectivity and give us information on the optimal size and arrangement of reserves. Viability analysis of carnivore species may highlight potential reserve areas that are not targeted in other biodiversity assessments such as gap analysis (Scott et al. 1993).
The restoration of mammalian carnivore species to portions of their former range, either by restoration of habitat or through active reintroduction, presents new challenges. Besides the inevitable sociopolitical difficulties, large and medium-sized carnivores may be particularly sensitive to landscape configuration because of their low population densities and large area requirements. In these species, population processes operate on a regional scale. Thus, regional-scale habitat models can be useful management tools for prioritizing restoration efforts. Multispecies conservation strategies have many advantages over single-species strategies (Noss et al. 1997). The extent to which restoration efforts for a particular species will enhance viability of the broader carnivore guild can be assessed by considering the major factors—such as topography, forest structure, and risk of human-induced mortality—that limit their distribution and abundance.
We used predictive habitat models to develop a carnivore restoration strategy for Oregon and northern California for three species: the gray wolf (Canis lupus), grizzly bear (Ursus arctos), and wolverine (Gulo gulo)—species currently extirpated from most or all of the region. Natural range expansion from adjacent states or existing refugia may be possible for at least the wolf and wolverine, and reintroduction programs have been proposed for all three species. Knowledge of the current amount and configuration of habitat can help us to identify core habitat areas and dispersal routes for these species and predict the viability of restored populations.
The three carnivore species considered here differ in the degree to which they tolerate human-associated landscape change and direct persecution. Contrasts in behavior, demographic characteristics, and population or metapopulation structure result in differing levels of ecological resilience (Weaver et al. 1996). All three species avoid humans. The wolf is highly resilient demographically, but its social structure increases the area requirements for viable populations. In mountainous portions of the western United States, the wolf may be especially vulnerable because its avoidance of rugged terrain brings it into greater proximity to human settlements (Paquet et al. 1996). The wolverine has the lowest reproductive output despite its relatively small size. The grizzly bear's limited dispersal abilities make it the most vulnerable species at the metapopulation level (Weaver et al. 1996).
The three species differ as well in the ecological roles they have played in Pacific coastal ecosystems and the potential impact of their restoration on current ecosystem dynamics. Bears may be an important link between riparian and upland systems—especially in regions, such as the Pacific coast, with anadromous fish populations. Nutrient input from salmon may be a key factor in productivity of coastal forest ecosystems (Bilby et al. 1996). The grizzly bear in south-central Alaska redistributed 40 percent of the salmon entering a coastal stream and was the conduit for 20 percent of the nitrogen uptake in adjacent forests (Hilderbrand et al. 1999). Storer and Tevis (1955:17) comment: "Its numbers multiplied by its average daily metabolic requirement must have made the grizzly an outstanding factor in the total food consumption by mammals ... [and] a dominant element in the original native biota of California." The historic ecological role of the wolf in the Pacific Northwest is unknown. Wolf reintroduction appears to have strongly affected vertebrate communities in the Greater Yellowstone Ecosystem (Smith et al. 1999), and we might expect similar effects in portions of the Pacific states. The ecological influence of wolverine populations in the lower 48 states is almost entirely unknown.
The historical factors leading to range contraction and extirpation differ among the three species, but all three were affected by predator control programs during the late 1800s and early 1900s (Schullery and Whittlesey 1999). To the extent that today's land management agencies are more tolerant of predators, the absence of these species from much of their former range is not an inevitable consequence of current human population density and land use. In eastern Europe, Italy, and China, populations of large carnivores coexist with much higher levels of human population density (Mattson 1990). In this chapter we present an approach for evaluating the biological feasibility of restoring populations of wolf, grizzly bear, and wolverine to areas within their former range in Oregon and California. We adapt habitat suitability models developed for these species in the Rocky Mountains, show potential core areas in a combined study region, and estimate potential population size.
The historical distribution and abundance of the wolf in the Pacific coastal states is uncertain. In California, wolves were probably most abundant on the northern coast, where elk (Cervus elaphus) were abundant, and in the northeastern corner of the state where they were found until 1922 (Grinnell et al. 1937; Schmidt 1991). Early extirpation from northern coastal California may have been due to human settlement patterns, including the gold rush of the 1850s, that briefly made the area one of the most densely populated in the western United States. Wolves were historically common in western Oregon (Bailey 1936), as well as east of the Cascades Range (Young and Goldman 1964). Most museum specimens were collected from the western foothills of the Cascades; the last wolf bounty in Oregon was awarded there in 1946 (Verts and Carraway 1998). Wolves reportedly persisted in the Oregon Cascades even after they were extirpated from the Rocky Mountain region (Young and Goldman 1964). While only scattered wolf reports exist from the latter half of the twentieth century, wolves have recently been documented dispersing into Oregon from the rapidly growing Idaho population (see chapter 6 in this volume).
Generally wolves locate their home ranges in areas where adequate prey are available and human interference is low (Mladenoff et al. 1995). The primary limiting factor for wolves has not been habitat degradation or prey depletion but direct persecution through hunting, trapping, and predator control programs. As human tolerance of large predators increases, however, wolves are well equipped to recolonize remaining areas of their former range. Because wolves reach sexual maturity at an early age and have large litters, the wolf has a high level of ecological resilience compared with other large carnivores (Weaver et al. 1996). The species' flexible social structure allows pack size, fecundity, and dispersal to respond to shifts in population density and prey abundance (Fuller 1989; Boyd et al. 1995; Weaver et al. 1996). Nonetheless, wolves were eliminated in areas of the western United States where grizzly bears persisted, suggesting that these compensatory mechanisms have limits. Population densities of wolves are usually far lower than population densities of sympatric grizzly bears. And as social animals, wolves are more susceptible to predator control than solitary animals.
Human activities affect wolf distribution and wolf survival (Thiel 1985; Fuller et al. 1992; Mladenoff et al. 1995; Paquet 1993; Paquet et al. 1996). In Wisconsin (Mladenoff et al. 1995) and Minnesota (Fuller et al. 1992), wolves selected areas with low human population density. The absence of wolves in human-dominated areas may reflect high levels of human-caused mortality, displacement resulting from behavioral avoidance, or some combination of both (Fuller et al. 1992; Mech and Goyal 1993). Roads, by increasing human access, negatively affect wolf populations at local, landscape, and regional scales (Fuller 1989; Thurber et al. 1994; Mladenoff et al. 1995). Wolves may avoid densely roaded areas because of traffic volume (Thurber et al. 1994), or their absence may be a direct result of mortality associated with roads (Van Ballenberghe et al. 1975). Even in areas where wolf harvest is prohibited, 80 to 95 percent of mortality is often anthropogenic (Fuller 1989; Mech 1989; Paquet 1993; Pletscher et al. 1997). Wolves in mountainous regions often concentrate their activities in forest valleys where snow conditions and prey availability are optimal (Paquet 1993; Paquet et al. 1996; Singleton 1995). Topography has not been incorporated in previous models (Mladenoff et al. 1995) developed in the north-central United States, due to the flatter terrain typical of that region.
Ungulates such as elk, deer (Odocoileus virginianus and O. hemionus), moose (Alces alces), and bighorn sheep (Ovis canadensis) make up the bulk of the wolf diet (Mech 1970; Fuller 1989), although they may take smaller prey such as snowshoe hare (Lepus americanus) and beaver (Castor canadensis). Ungulate biomass (Keith 1983; Fuller 1989), density, and species diversity (Corsi et al. 1999) are important habitat factors. In a review of studies from several regions, for example, prey density explained 72 percent of the variation in wolf density (Fuller 1989). A smaller core area can support a viable wolf population if prey biomass per unit area is high (Fritts and Carbyn 1995; Wydeven et al. 1995).
The current distribution of the grizzly bear in the Pacific coastal states is confined to a small remnant population in the North Cascades of Washington, where 17 records have been confirmed in recent decades (Almack et al. 1993). Distribution was widespread in Oregon except in the arid east (Bailey 1936). The last grizzly bear in the state was killed in northeastern Oregon in 1931 (Verts and Carraway 1998). As many as 10,000 grizzly bears occurred throughout California except in the northeast and the Mojave Desert (Storer and Tevis 1955). Although the California grizzly may have been most common in chaparral rather than dense forest, it was also abundant in oak woodlands, river valleys, and mixed-hardwood forests (Storer and Tevis 1955). In the 1850s, a settler counted 40 bears feeding along the Mattole River in coastal northwestern California (Grinnell et al. 1937). Despite its initial abundance, the species was extirpated from northern California by 1902. The last known grizzly in the state was killed in the southern Sierra Nevada in 1924 (Storer and Tevis 1955).
The grizzly bear has a combination of life history traits that contribute to its low resilience in the face of human encroachment (Bunnell and Tait 1981). Its low lifetime reproductive potential (as few as three female young per adult female) makes population viability sensitive to small declines in adult survivorship (Weaver et al. 1996). Although subadult males often disperse two home-range diameters (about 70 km), successful long-distance dispersal between subpopulations has not been recorded in the western United States.
The range of the grizzly has become increasingly fragmented (Craighead and Vyse 1996), exacerbating the demographic and genetic risks associated with small, isolated populations. Craighead and Vyse (1996) have compared the viability of bear populations on islands of varying size and conclude that isolated populations require at least 1000 bears for long-term persistence. Mattson and Reid (1991) found a similar size threshold for European brown bear populations. Roads and their traffic cause direct mortality, disrupt bear behavior, create barriers to movement (Archibald et al. 1987; McLellan and Shackleton 1988), and increase poaching and removal of habituated bears (Mattson et al. 1987; Weaver et al. 1996).
Although the grizzly is an omnivore, its resilience is limited by its seasonally high caloric needs (Weaver et al. 1996). The species was widespread in western ecosystems, and its diet reflects this distribution. Key foods range from soft mast (drupes, berries) and hard mast (acorns, whitebark pine nuts) to fish, vertebrate carrion, and insects (Storer and Tevis 1955; Mattson and Reid 1991).
Because wolverines exist at low densities and inhabit remote areas, it is difficult to judge whether the Pacific Northwest supports reproducing populations or just dispersing individuals. In California, the last confirmed specimens were collected from the Sierra Nevada in the early 1900s (Grinnell et al. 1937). More recent unconfirmed reports originate from the southern Sierra Nevada (Barrett et al. 1994). In Oregon, specimens were confirmed in the Blue Mountains of eastern Oregon in 1986 and 1992 and from Steens Mountain in southeastern Oregon in 1973 ( Verts and Carraway 1998). Specimens were collected from the central Oregon Cascades in 1965, 1969, and 1973 ( Verts and Carraway 1998). The 1969 specimen was a female, suggesting the possibility of reproduction. It seems clear, however, based on the sparse evidence, that wolverines are scarce in Oregon and California relative to the Rocky Mountains. Although little is known about the wolverine's habitat needs and distribution in the Pacific Northwest, fragmentation of the landscape by roads and human development may hinder natural recovery there as in other regions (Carroll et al. 2001).
Because the wolverine's diet includes unpredictably distributed resources such as carrion, it has larger home-range requirements than equivalent-sized carnivores. Carrion use may link wolverines to other carnivores, such as wolves, that are much reduced in the western United States. Moreover, wolf poisoning campaigns have eradicated local wolverine populations as well in some regions (Banci 1994).
Female wolverines mature at three years of age and produce less than one kit per year until death at six to eight years (Copeland 1996). Large area requirements and low reproductive rates make the wolverine especially vulnerable to human-induced mortality and habitat alteration. Populations probably cannot sustain annual rates of human-induced mortality greater than 7 or 8 percent, a rate lower than that usually caused by trapping (Gardner 1985; Banci 1994; Weaver et al. 1996). Areas closed to trapping such as Yellowstone National Park and the Canadian mountain parks in Alberta and British Columbia appear to be wolverine refugia (Hatler 1989; Buskirk 1999).
Although the wolverine's long-range dispersal abilities (greater than 200 km in Idaho; Copeland 1996) may facilitate its persistence, females tend to settle closer to their place of birth (Banci 1994). The large home-range sizes of Idaho wolverine (a mean of 384 km2 in females) relative to those in Canada and Alaska suggest more limited food or denning resources (Copeland 1996). Female wolverines must leave their kits for lengthy foraging trips. In the lower 48 states, they often select natal dens in alpine areas where snow tunnels in talus can provide thermoregulatory benefits and safety from predators (Magoun and Copeland 1998). Natal dens in Alaska, by contrast, appear less limited by topography or human settlement (Magoun and Copeland 1998).
Using habitat quality to predict carnivore distribution is especially challenging because there is much to learn about the link between habitat and demography. Habitat models such as the habitat suitability index (HSI) system were developed primarily for site-level planning and may be poor templates for regional evaluations. Such conceptual models rely on qualitative relationships derived from expert opinion and published studies. Empirical models, in contrast, base predictions on statistical analyses of species occurrence data. Regional-scale empirical models have been used to predict range expansion of wolves in the north-central and northeastern United States (Mladenoff et al. 1995, 1999), as well as grizzly bear distribution in Idaho and Montana (Merrill et al. 1999; Mace et al. 1999; Boyce and MacDonald 1999).
Excerpted from Large Mammal Restoration by David S. Maehr, Reed F. Noss, Jeffery L. Larkin. Copyright © 2001 Island Press. Excerpted by permission of ISLAND PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
ContentsAbout Island Press,
Introduction: Why Restore Large Mammals?,
PART I - Feasibility,
Chapter 1 - Is the Return of the Wolf, Wolverine, and Grizzly Bear to Oregon and California Biologically Feasible?,
Chapter 2 - Feasibility of Timber Wolf Reintroduction in Adirondack Park,
Chapter 3 - Rewilding the Sky Islands Region of the Southwest,
Chapter 4 - Using Public Surveys and GIS to Determine the Feasibility of Restoring Elk to Virginia,
PART II - Practice,
Chapter 5 - Returning Elk to Appalachia: Foiling Murphy's Law,
Chapter 6 - Outcomes of Hard and Soft Releases of Reintroduced Wolves in Central Idaho and the Greater Yellowstone Area,
Chapter 7 - Health Aspects of Large Mammal Restoration,
Chapter 8 - Restoring the Mexican Gray Wolf to the Mountains of the Southwest,
PART III - The Human Link,
Chapter 9 - Translocation of Plains Bison to Wood Buffalo National Park: Economic and Conservation Implications,
Chapter 10 - Restoration of Grizzly Bears to the Bitterroot Wilderness: The EIS Approach,
Chapter 11 - Mountain Sheep Restoration Through Private/Public Partnership,
PART IV - Abetting Natural Colonization,
Chapter 12 - Black Bear at the Border: Natural Recolonization of the Trans-Pecos,
Chapter 13 - Restoring a Large-Carnivore Corridor in Banff National Park,
Chapter 14 - Tiger Restoration in Asia: Ecological Theory vs. Sociological Reality,
Chapter 15 - The Florida Panther: A Flagship for Regional Restoration,
Chapter 16 - The Biotic Province: Minimum Unit for Conserving Biodiversity,
Large Mammal Restoration: Too Real to Be Possible?,
Island Press Board of Directors,