Pacific water shrew (Sorex bendirii) COSEWIC assessment and status report: chapter 5

Habitat requirements

In Oregon and Washington, the Pacific water shrew is associated with skunk cabbage (Lysichiton americanum) marshes, red alder (Alnus rubra) riparian and stream habitats, and dense wet forests of western redcedar (Thuja plicata) (Maser et al. 1984; Johnson and Cassidy 1997; Verts and Carraway 1998). Both McComb et al. (1993) and Gomez and Anthony (1998) demonstrated significantly higher capture rates in riparian than upland habitats, and Gomez and Anthony (1998) concluded that this shrew was “an obligate species of riparian habitat”. Anthony et al. (1987), who compared small mammal captures in stream edge (~ 1 m from stream edge) and riparian fringe (15-25 m from stream), captured Pacific water shrews only in their stream edge transects. However, other researchers have captured it 25-350 m away from streams in riparian and mature upland forest (McComb et al. 1993; Gomez and Anthony 1998). West (1991) reported captures in dry Douglas-fir forests (Pseudotsuga menziesii) at considerable distances [no actual distance given] from water. It is unknown if these captures distant from water represent resident or transient animals. Because of the few captures for the Pacific water shrew, small sample sizes in most habitat studies limit statistical analyses. But the capture data strongly suggest that the Pacific water shrew is more abundant in older seral stages of forest than young forests (Aubry et al. 1991; McComb et al. 1993; Gomez and Anthony 1998). In the Olympic National Forest of Washington, Lomolino and Perault (2001) never detected this species in clear-cut habitats.

In Canada, habitat use is presumably similar to the Oregon and Washington populations, although habitat data are limited to incidental captures or observations. Brooks (1902) reported that it used thick woods and swamps in Chilliwack Valley. Habitat for the three captures taken by Zuleta and Galindo-Leal (1994) were creeks with conifer forest (western redcedar, western hemlock, Tsuga heterophylla) or mixed forests (bigleaf maple, Acer macrophyllum, Douglas-fir) with canopy covers exceeding 50%. Based on existing capture records for British Columbia with habitat data, Craig (2003) developed a habitat model for a four-class rating scheme. The model assumes that riparian habitats in mature forest (structural stages 4-7) of western redcedar or western hemlock, and mature (structural stages 4-7) deciduous or coniferous forests associated with marshes or wetlands are the ideal habitat for this species. The model also assumes that the species is usually found within 60 m of streams or wetlands. However, as noted by Craig (2003) more data on habitat are needed to refine the model including the coverage of wetlands and anthropogenic habitats. The model also needs to be tested and verified with actual field surveys.  Although forested habitats are important, this shrew has been taken in non-forested grassy habitats bordering ditches and sloughs in British Columbia (Craig and Vennesland 2004a). The most unusual habitat was the capture by Jackson (1951) of nine Pacific water shrews at Point Grey, Vancouver in ocean beach debris near small pools fed by a freshwater spring.

Habitat trends

There are no specific data on habitat trends for the Pacific water shrew and no quantitative data exist for the rate of habitat loss in the past 10 years. However, general trends can be inferred from studies of changes in land types (Moore 1990; Boyle et al. 1997) and loss of wetlands and streams (Fisheries and Oceans Canada 1998; Moore et al. 2003). A limitation of these studies is that they generally apply to only a portion of the total habitats occupied by the Pacific water shrew and they do not address the scale of habitat loss in the past decade. Nevertheless, they demonstrate that historically and in the past decade forest and riparian habitats have disappeared from this species’ range.

According to Boyle et al. (1997) much of the lower Fraser River basin before 1820 was covered in dense mature coniferous forest. The area also supported significant wetlands such as fens, marshes, bogs, and swamp forest. By the 1930s large areas had been logged and cleared for agriculture (Table 1). As a result of drainage and the construction of dykes, freshwater wetlands in this region declined about 87% from the period of first European contact in the 1820s to 1990. For example, Sumas Lake and its associated wetlands (an area with records of the Pacific water shrew in the late 1880s) was drained and dyked by 1924 with the loss of about 8,000 ha of marshland and slough (Moore 1990). Moore et al. (2003) conducted a detailed study quantifying the recent loss of wetlands in the region. Comparing 1989 inventory data from Ward et al. (1992) and 1999 orthophotos, they assessed changes in 320 freshwater wetlands. About 22% of the wetlands demonstrated some encroachment from urban or agricultural development with the loss of 965 ha (Table 2). Most of the loss resulted from agricultural activities such as cranberry production. Fisheries and Oceans Canada (1998) assessed the impact of habitat changes and water quality on 662 fish-bearing streams in the lower Fraser River valley (Table 3). They found that only 14% remained in their wild state; 15% of the streams were lost.

Table 1. Land cover changes in the LowerFraser BasinEcosystem from pre-1820 to 1990. Modified from Boyle et al. (1997).  Only land cover types used by the Pacific water shrew (Sorex bendirii) are listed. Values in parentheses represent the percent of total of land cover including categories not used by the shrew.
Land Cover Pre-1820 1930 1990
Coniferous
590,800 (71)
412,000 (50)
445,800 (54)
Deciduous/mixed
8,200 (1)
71,800 (8)
4,000 (0)
Fen
56,000 (7)
5,500 (1)
2,400 (0)
Swamp/bog/marsh
27,100 (3)
10,800 (1)
9,700 (1)
Agriculture
0
81,000 (9)
132,100 (16)
Urban
0
25,000 (3)
86,300 (10)
Cleared
0
79,200 (10)
8,600 (1)

 

Table 2. Changes from 1989 to 1999 in 320 wetlands of the Fraser River lowland region (Taken from Moore et al. 2003)
Encroachment Type Area Lost (ha) % of Total Area Lost Number of Wetlands Affected
Agriculture
404
42
26
Golf Course
244
25
4
Landfill
150
16
1
Industrial
64
7
19
In Transition
49
5
12
Residential
38
4
11
Commercial
9
1
6
Transportation
6
1
14
All
965
101
 

 

Table 3. Status of 662 fish-bearing streams in the lower Fraser River valley (Taken from Fisheries and Oceans Canada 1998)
Area Still Wild Endangered Threatened LostTable notea
Steveston-Langley
0
95
8
45
Abbotsford to Hope
27
142
58
6
Stave River to Hope
22
36
72
6
West Vancouver to Stave River
57
102
43
60
Total
106
375
181
117
Percent of total
14
48
23
15

Habitat protection/ownership

The precise amount of habitat legally protected is unknown. Protected areas are limited to 26 provincial and regional parks (see Existing Protection or Other Status Designations section). Federal lands within the known range include 4 Department of National Defence Properties (1,437 ha) and about 62 Indian Reserve lands (8,533 ha). These various federal lands are small and fragmented. The amount of habitat on provincial Crown land is small, probably less than 20%. Most Pacific water shrew habitat in Canada is on private land. The Best Management Practices Guidelines for Pacific water shrew in Urban and Rural Areas (Craig and Vennesland 2004b) developed by the province provide detailed guidelines for municipalities on protecting Pacific water shrew habitat,including a recommendation of 100-m buffers around streams or wetlands.

Page details

Date modified: