The Scientific Basis for Conserving Forest Carnivores: American Marten, Fisher, Lynx, and Wolverine in the Western United States

1. What they did

In 1993 the United States Department of Agriculture (USDA) Forest Service set out to assess what is known about several groups of animals that may be sensitive to forest management, including the harvest of late-successional (mature and old) forests. This report covers the four “forest carnivores”: American marten, fisher, lynx and wolverine, in the western United States outside Alaska. These species need large areas, are linked to older forests and had little information available for conservation planning.

Recognised authorities wrote a literature review for each species (Chapters 2 to 5). Each covers distribution and taxonomy, population ecology, reproduction, food and prey, habitat, home range, movements, interactions with other species and conservation status, and ends each topic with management considerations and research needs. A table in each chapter lists the existing studies, with their location, length, method and sample size, so readers can judge how solid the knowledge is.

The editors then drew the species reviews together into management considerations (Chapter 6) and a research strategy (Chapter 7). Appendices describe the ecoprovinces of the region, map what is known of the distribution of fisher, lynx and wolverine, and give National Forest status information. The editors’ brief did not include making specific management recommendations.

2. What they found

How reliable is the knowledge?

The editors list six ways research can mislead a conservation assessment: results from only one or a few places; many findings drawn from only a few studies; small samples or highly variable results; ambiguous measures and problems of scale (for example, an animal simply being in a habitat says little about what it needs); loose use of terms such as “old growth”, “dispersal” and “den”; and the wrong methods (for example, judging population structure from trapped carcasses). They point out that knowledge of lynx denning habitat rests on very few dens.

Where the knowledge comes from

Most of what was known came from Canada or Alaska (wolverine and lynx) or the eastern United States (fisher). Marten were the best studied in the western United States. Only two intensive radio-telemetry habitat studies of fishers had been published in the West, and only five lynx studies had ever been done in the western mountains.

Habitat

• Marten, fisher and lynx prefer forest with overhead cover and complex structure near the ground. Physical structure seems more important than tree species.
• Wolverines seem less tied to canopy cover, are often found in alpine and subalpine areas, and may above all need isolation from people.
• Only two fisher natal dens and four lynx natal dens had been described in the western mountains. Dens of marten, fisher and lynx are in stands with downed logs, snags, large and hollow trees and stumps. Wolverine natal dens have been found in snow tunnels, hollow trees and caves.
• Marten rest in winter almost entirely under the snow, getting there through logs that stick out of the snow. Fishers and wolverines dig snow tunnels.
• All four tend to avoid foraging in the open. Fishers and lynx usually cross openings quickly unless prey is plentiful there.

Food and hunting

• Marten mostly eat ground-dwelling voles. Lynx depend on snowshoe hares over much of their range. Fishers specialise on porcupines but eat almost any small to medium-sized mammal or bird they can catch. Wolverines are mainly scavengers of large ungulate carrion.
• Snow-tracking studies of lynx, reviewed in Chapter 4, showed hares were most of the diet even when scarce. Lynx hunted by following hare runways, concentrating in small areas of hare activity or using “waiting beds” (depressions in the snow). One study credited waiting beds with 61% of hares killed. Snow-tracking lynx for 416 km in Alberta gave 0.42 kills per day.
• Along 123 km of fisher tracks in Michigan, there were 14 kills and scavenges in about 21 days of hunting, many of them of little food value.
• In the western mountains, snowshoe hare populations do not cycle as they do in the north, so northern lynx and hare patterns do not apply there.

Home range and movement

• Home ranges run from under 16 km² for marten to several hundred km² for wolverines.
• Wolverines may travel 30 to 40 km a day. Distances measured by snow-tracking are better than telemetry; in one Alaska study actual movements were 33% longer than the straight lines between telemetry locations.
• Snow-tracking generally gives smaller lynx home ranges than radio-telemetry, so home ranges from different studies need care when compared.

Status

• All four species have lost range in the western mountains. Some marten and fisher populations (for example, the two fisher populations in California) may be isolated. Lynx are gone from Oregon. Wolverines declined sharply over the past century but appeared to be recovering in some areas.
• Fragmentation of forest, mainly by roads and logging, is repeatedly named as the most important isolating force; only the wolverine seems less affected.
• Trapping can greatly affect these populations, including through incidental capture of fishers set for other furbearers.

3. What they concluded

• Huge knowledge gaps make it hard to judge the conservation status of these species. Even so, nothing in the assessment suggested that existing designations of them as species of concern were wrong, and conservation strategies are needed to ensure they persist in the western mountains.
• Complex, large physical structures typical of mesic late-successional forest will be important. There is little to suggest these structures can meet the animals’ needs outside their natural setting.
• Further loss or fragmentation of late-successional forest, especially clearcutting of continuous forest, may be harmful, most of all for marten and fisher, with effects depending on local context.
• Habitat changes that favour generalist predators may harm forest carnivores.
• Conservation will need ecosystem management at the landscape scale, cooperation across regions, agencies and the Canada and United States border, and protection of ecotypic variation, including in populations at the edges of the range.
• Isolated populations bring special challenges. Reintroduction makes sense only once the causes of loss have been fixed and suitable habitat restored, using animals from similar populations and forests.
• Monitoring after management actions cannot replace research. The editors recommend four types of study, replicated in chosen ecoprovinces: intensive radio-telemetry studies, intensive demographic studies, extensive studies of where the species occur, and ecosystem studies of prey, vegetation and interactions among carnivores. They suggest a Western Forest Carnivore Research Center in western Montana.
• The lack of information should not stop managers from developing conservative interim guidelines that keep future options open.

4. How tracking was or was not used

The report did not collect new field data. Tracking appears throughout the reviewed studies:

• Snow-tracking was used to study lynx hunting success, hunting methods, kill rates, diet, density and home ranges; fisher hunting success; wolverine daily travel distances; and marten habitat use (for example, 10 of 11 marten tracks in one Washington study were in older spruce, fir and lodgepole pine stands). One marten study in the species table used 35 km of tracks over two winters.
• Scats were the main source of diet information for all four species, often alongside stomach or gut contents. The authors note that remains in scats cannot always tell predation from scavenging (for example, deer hair in lynx scats, marten remains in fisher scats).
• Detection: snow-tracking, smoked track plates, sooted track boxes and baited cameras are named as ways to find out whether these species are present, but protocols for using them consistently were lacking. For wolverines, the review suggests confirming presence with remote cameras or confirmations of tracks, as sightings can mislead.
• The species review notes that marten tracks in snow “are distinctive to experienced observers”.
• Chapter 7 notes that radio-collaring lets researchers assign behaviour seen while snow-tracking to a sex and age class.

5. Tracker or observer certification

Not reported. As a review, the report does not describe who did the snow-tracking or scat collection in the studies it summarises.

6. Other technologies, skills or innovations

• Radio-telemetry, including aerial telemetry flights, for home range, habitat use, movement and dispersal.
• Baited camera stations, smoked track plates and sooted track boxes for detection.
• Geographic Information Systems (GIS) for landscape-scale studies.
• Genetic (DNA) analyses, mentioned as a way to study relatedness and gene flow.
• An ecological stratification of the region into ecoprovinces (Appendix A) used to plan where studies should be replicated.

7. Implications for conservation (from the paper)

• Forest carnivores are largely wilderness species, intolerant of human activity, with low reproductive rates and large area needs, so they are especially sensitive to forest management at the southern edges of their range.
• Because they are rare, mostly nocturnal, leave little sign and avoid people, they can easily go undetected. Failure to detect them does not prove absence, and standard detection methods were needed.
• Where trapping occurs, populations need close monitoring. Banci (Chapter 5) suggests that jurisdictions without the means to monitor populations closely, or without large refugia, cannot justify a harvest.
• Riparian corridors and links between populations may be important, as may large protected areas connected by travel corridors for wolverines.
• Fur harvest records, carefully documented, can give useful information on distribution and possibly abundance.