Background
Fur-bearing mammals are considered to be particularly sensitive to habitat loss and fragmentation [
1‐
4]. Their habitats have been affected especially by major changes that are incurred through human activities, such as forest harvesting and agricultural development [
5,
6]. Climate change could further modify or reduce the quality of available habitats [
7,
8]. Yet climate change and anthropogenic disturbances such as forest harvesting may also increase access to new territories by modifying biotic and abiotic factors that otherwise would limit a species’ potential to utilize a territory [
9]. Given that climate change and anthropogenic disturbances occur over long periods of time, resulting in complex cumulative impacts, they are often difficult to understand and to document [
10,
11].
Populations of fur-bearing animals typically exhibit low densities. Consequently, monitoring these species to document their status is relatively rare [
12,
13]. Sales of trapped pelts have long been used by wildlife managers to track fluctuations in the abundance of certain wildlife populations [
14‐
16]. However, from one year to the next, fur sales can be influenced by animal population status and trapping efforts, which depend upon numerous social (e.g., employment-trapping conciliation, trappers’ health status), economic (e.g., variation in fur prices, available material resources), and environmental (e.g., weather, local habitat disturbance) factors [
17‐
19]. Beyond information that is provided by fur sales, the experience of trappers and the knowledge that they accumulate over many trapping seasons is an invaluable, frequently untapped source of information, which would enrich our understanding of species with relatively low densities, such as mustelids [
20‐
22].
The growing interest in local knowledge is due, among other things, to the potential saving on time and money needed to gather the scientific information required to meet environmental challenges [
23,
24]. In addition to the number of individuals that are caught, trappers’ experience allows us to learn more about the ecology of exploited species [
22,
25]. Trappers continuously monitor population dynamics as well as natural and anthropogenic forest disturbances. Trappers have developed a solid expertise regarding ecosystem responses, including wildlife responses to habitat changes, over broad spatiotemporal scales [
26,
27]. Knowledge on habitat requirements of those species which are most sensitive to forest practices, from both local and scientific sources, is essential for the conservation of key habitats [
28].
Both fisher (
Pekania pennanti) and American marten (
Martes americana) are mustelids trapped for their fur that play a key role in forest social-ecological systems in eastern North America. They are sensitive to forest management prescriptions and considered to be ecological indicators of forest health [
1,
29,
30]. American marten is one of the species most frequently sought by trappers because of its ease of capture and the high market value that is placed upon its fur [
20,
31,
32]. In addition, marten and fisher are important in some Indigenous cultures, notably those of the Anishnaabeg [
33] and the Cree [
34]. These species can be considered as cultural keystone species that are essential in maintaining the complexity of socio-ecological systems [
35].
During the 1970s, the fisher experienced periods of low abundance in North America [
36]. This was attributed to over-exploitation of its fur and habitat loss [
36]. Since the early 2000s, the populations appear to have recovered and the interest of trappers in the fisher has intensified, resulting in a substantial increase in fisher pelt sales, especially in Quebec [
15]. Sales of fisher pelts apparently indicate that their geographical distribution could be expanding towards the northeast. This change in fisher distribution would thus result in greater overlap with marten’s range in habitats that are undergoing natural [
37] or anthropogenic disturbances [
6,
38]. The objective of this study was to document the local knowledge of trappers in western Quebec (Canada), where distributions of fisher and marten are sympatric, to identify factors that could explain variation in populations of the two species and the interactions between them.
Discussion
Information that was derived from interviews with local trappers regarding the evolution of marten populations and pelt counts had exhibited the same trends as fur sales data that were compiled by the Government of Quebec (Ministry of Forests, Wildlife and Parks, MFFPQ) at the regional level (Fig.
2). Fisher abundance appears to have increased, especially since the 2000s at the northern limit of its range (Zones 2 and 4). Compared to Zone 3, Zone 2 is a region that contains more agricultural operations and more mixedwood stands than coniferous stands. However, trappers have noted a recent decline in fisher populations, which is particularly pronounced in Zone 3. The main cause of this decline is attributed by trappers to the decrease in large conifers that are used as resting and rearing sites. Marten populations are decreasing throughout the study area. However, trappers have acknowledged the discrepancy between sales of pelts that were recorded for each trapline and what was actually taken. One of the main reasons is the harvest on unstructured territories and exchanges between trappers to respect harvesting thresholds set by the government that would permit them to retain their rights to exploitation. Interviews were also used to identify changes in populations of other species for which the Quebec Ministry of Forests, Wildlife and Parks did not have data, such as unexploited (e.g., prey) species or those for which the fur sales data do not reflect these observed changes (lynx exploitation thresholds), or the particular nature of a species (e.g., wolves, which trappers are targeting for the challenge that its capture represents).
Trappers who were interviewed in this study were concerned about marten populations, a species that is particularly vulnerable to habitat alterations caused by logging operations. Concerns were expressed about the lack of protective cover, which has been incurred through the increase in hardwood species [
64] and reductions in coarse woody debris (CWD) on the ground surface [
65,
66]. The marten has long been perceived as being dependent upon mature and aging conifer stands [
3,
67]. These stands generally provide greater quantities of CWD than do younger stands, forest plantations and intensively managed stands [
68]. CWD provides subnival spaces that can be used by marten [
69]. Subnival spaces are particularly important for prey accessibility and provide protection against predation, together with resting sites for thermoregulation [
70‐
73]. Consequently, habitat use by marten is mainly dependent upon the internal structure of the forest [
74,
75]. Fishers, in contrast, appear to be able to exploit a greater diversity of habitats, including more open forest environments, the edges of agricultural environments, and even urban areas [
22,
76]. Fishers are not necessarily old-growth forest specialists, and are known to utilize young forests, mixedwood stands and ecotones [
22]. A study that was conducted in 2016 in New York State [
77], however, showed that occupation of the territory by the fisher was positively influenced by proportions of coniferous and mixedwood forests, but negatively affected by the density and proportion of agricultural environments. Like marten, the fisher likely would be more dependent upon the structure of the forest than on the age or type of settlement [
22,
73‐
75]. Participants also pointed out that mustelid habitats were especially consistent with those of their main prey: snowshoe hares, squirrels and other small mammals. These prey species and, therefore, fisher and marten, require coarse woody debris, vegetation cover, and structurally complex forests [
72,
78]. This complex structure is encountered mainly in old-growth forests, but may also be present in younger stands that have suffered windfalls or insect outbreaks [
79,
80].
Participants stressed the importance of the availability of resting and rearing sites, which they did not consider to be sufficiently protected by forest management. Fishers depend upon the availability of large moribund trees that often have rotting heartwood, such as white pine, white cedar and yellow birch, in which they can make their dens and raise their young [
81,
82]. As noted by Bridger et al. in 2016 [
22], managers need to consider the importance of den availability for martens and fishers. The scarcity of dead wood is a further ecological challenge to forest management [
80,
83‐
85]. Protection measures for dead wood were previously mainly guided by FSC (Forest Stewardship Council) forest certification [
86]. With the introduction of ecosystem-based forest management, silvicultural practices have tended to create forest landscapes that contain all of the diversity of natural forests, including the composition and shape of stands and the presence of trees of different sizes, snags or woody debris [
87‐
90]. Participants also encourage forest development that creates small openings, similar to gaps that naturally drive the development of sugar maple-yellow birch stands [
43].
More generally, participants reported that the return of martens after logging was more rapid in mixedwood than in softwood stands. Most mixedwood harvest cuts are partial cuts, while softwood stands are cut more intensively. In 2005, Potvin et al. [
91] determined that cut mixedwood stands had higher lateral cover and higher regeneration compared to cut conifer stands. Forest management influences populations of competitors and even predators of marten and fisher. Lynx, wolf and coyote probably benefited indirectly from forest management, which favoured their main prey, including snowshoe hare and moose [
91‐
93]. Marten would thus be the victim of apparent competition [
94]. Whether through habitat alterations or its effects on prey or predator species, forest management appears to be detrimental to marten, while favouring the fisher.
The fisher is limited in its movements over snowpacks that do not fully support its weight and is dependent upon the coniferous cover that intercepts the snow [
95]. However, lower snow accumulation and ice crusting over a longer period of time result in greater load-bearing capacity of the snowpack, which seems to favour the fisher, thereby providing it with access to new habitats [
8,
96]. The experience of trappers also has made it possible to document this snow load-bearing capacity, a characteristic for which no scientific follow-ups have been performed over long temporal and spatial scales.
Climate change and human activity have led to changes in abundances of other species, which in turn may influence the population dynamics of fisher and marten. Raccoon and coyote populations have increased in the northern part of their ranges [
97,
98]. Populations of bald eagles and other raptor species have appeared to be recovering naturally throughout North America, since the use of DDT was discontinued [
99,
100]. All of these changes appear to be detrimental to marten, which would increase the abundance of predators and competitors while leading to marten habitat degradation at the same time. The fisher would be less affected, having fewer predators than marten.
The fisher may exert competitive dominance, thereby excluding marten from areas that the latter occupies [
101]. Indeed, competition between the two species has been reported by the participants. Yet interviews with trappers indicated that the principal relationship between fishers and martens was predation by the former on the latter species. To our knowledge, no study has yet been able to evaluate effectively the extent of predation of fisher on marten. In 2010, McCann et al. [
102] documented predation by fishers on martens in northern Wisconsin, which would occur only during the winter. Both fisher and marten eat squirrels, other small mammals and snowshoe hares, but in different proportions. They explained that their overlapping diets may lead to greater interactions in the winter when hunting by both predators decreases prey populations [
103,
104]. Additional studies would be required to identify the importance of marten in the diet of fisher in areas where the species are sympatric. Assessment of predation by fisher on marten would be all the more relevant as the abundance of fisher increases north of its current range, resulting in greater interactions with marten.
In this study, it was difficult to distinguish between effects of natural and anthropogenic disturbances on the population dynamics of marten and fisher. Whether it is forest rejuvenation, increasing the proportion of deciduous species, intensifying agricultural activities, climate change, or the consequences of their effects on other animal species, all of these factors are cumulative and, according to the experiences of the trappers who were interviewed, appear to favour the fisher to the detriment of the marten.