Sorex arcticus - Arctic Shrew

Physical Description:

Throughout most of its range, S. arcticus can be distinguished from other shrews by the presence of a well-developed postmandibular canal and lack of pigment on the ridges of their unicuspids.  The forth unicuspid is generally smaller than the third and the species lacks an accessory tine on the first upper incisor.  The tricolored pattern especially on adult pelage also distinguishes S. arcticus from other shrews in its range.  However, in the northern extent of the range of S. arcticus, S. cinereus (masked shrew) may also have lighter colored sides, but they are generally smaller than S. arcticus and lack a postmandibular canal (Kirkland and Schmidt 1996).  Also, individuals of S. fumeus (smoky shrew), often found with S. arcticus, may exhibit weakly pigmented unicuspids and a postmandibular foramen.  These two species can be distinguished from the lack of lighter coloration of the sides and a generally smaller skull size in S. fumeus, however general body size and weight does not differ significantly between the two species (Kirkland and Schmidt 1996).

S. arcticus exhibits two different color phases mostly correlated with age (Clough 1963).  Adults, most from the previous year, and some first litter young of that year are distinctly tri-colored.  The back is a rich dark brown, markedly distinct from the lighter brown of the side, which is also sharply contrasted from the paler brown color of the belly.  The pelage of the young is not as distinctly tricolored and may appear more bicolored.  The back is a medium brown that gradually fades to a lighter brown on the side, which then fades into a paler brown on the belly (Clough 1963).

Of shrews captured in Wisconsin the average total length for males were 113.6 mm (n=31) and 116.2 mm (n=22) for females.  The average tail length was 41.6 mm (n=31) for males and 41.9 mm (n=22) for females.  Length of the hind foot averaged 13.9 and 13.8 mm for males and females (n=30 and n=23) respectively.  The average weight for males (n=30) and females (n=24) was 8.0 and 8.3 g respectively (Clough 1963).

Distribution:

The genus Sorex has a holarctic distribution, and based on fossil evidence it is thought that ancestors of S. arcticus invaded the nearctic across the Bering Land Bridge possibly multiple times before the early Pleistocene (George 1988).  Most Sorex arcticus fossils have been found in Pleistocene deposits well south of the southern extent of the Wisconsin glaciation (central and southern Appalachian Mountains, and the Great Plains area) (Kirkland and Schmidt 1996).

Currently, S. arcticus is distributed throughout most of the boreal coniferous forest of North America.  The species occupies most of Canada from the Southern Yukon and MacKenzie valleys, east to Quebec, Nova Scotia and the Atlantic coast.  The range is bound by Hudson Bay to the North and extends south to the northern part of the Midwestern United States, (Eastern North and South Dakota to Wisconsin and the Upper Peninsula of Michigan) (Kirkland and Schmidt 1996).  S. arcticus has been found across most of the glaciated area of Wisconsin.  The southern extent of the species in Wisconsin is a population of S. arcticus found and studied just north of Madison in Dane Co. (Clough 1963).

Ontogeny and Reproduction:

For the most part breeding begins in May and lasts until August although some lactating females have been found into September and October and occasionally pregnant or lactating females were caught earlier in April (Clough 1963; Baird et al. 1983; Buckner 1966).  All males trapped from February to June had enlarged testes (Clough 1963).  Also, adult males stopped appearing in traps after July, whereas adult females could be found in traps through August (Clough 1963; Baird et al. 1983).

The average embryos or corpora lutea found in females was slightly greater than six; pregnant females trapped in Wisconsin averaged six embryos (Clough 1963; Baird et al. 1983; Buckner 1966).  Since corpora lutea are present in almost all pregnant females in early spring but are mostly absent from pregnant females in the fall, the presence of corpora lutea may indicate future reproductive periods.  Also, since corpora lutea have been found in lactating females but no embryos have been found in lactating embryos, the species may exhibit delayed fertilization or implantation until lactation has ceased (Baird et al. 1983)

Ecology and Behavior:

Most S. arcticus occur in many habitats but populations are primarily found in non-forested, hydric communities, primarily marshes or grassy clearings (Wrigley et al. 1979; Baird et al. 1983; Buckner 1966; Clough 1963; Kirkland and Schmidt 1996).  A population studied in Manitoba for many years seemed to vary directly with depth to ground water (Buckner 1966).

In one study it was reported that there was an inverse relationship between population size and density of S. arcticus and another shrew S. cinereus (Buckner 1966).  However, another study found no relationship between the abundance of both species (Clough 1963).

The diet of S. arcticus seems to consist entirely of insects especially the larch sawfly later into the season (Kirkland and Schmidt 1996).  However they have also been observed to eat meat when it was provided to them in captivity (Clough 1963).

Metabolic measurements for S. arcticus are mostly an intermediate between the lower metabolism of S. cinereus and the higher metabolism of Blarina brevicauda (northern short-tailed shrew).  When feed larch sawflies, it was found that the assimilation efficiency of S. arcticus was about 88% and stomach contents had a mean caloric value representing only 9.5% of estimated metabolic requirements of the species.  Thus an individual’s stomach must be filled about eleven times a day (Kirkland and Schmidt 1996).

Rapid turnover rates occur in S. arcticus populations between May and October.  The S. arcticus populations tend to start small in spring then increase and peak during the summer and decline in fall, and individuals breeding in the spring were probably subadults the previous fall.  Other yearly trends have not yet been noticed or do not occur (Clough 1963; Baird et al. 1983; Buckner 1966).  Through mark and recapture sampling it was found that breeding adults from the beginning of the breeding season in the spring were often gone by the end of June, and the young from early spring litters were reproductively mature adults by August, yet most of these individuals did not survive through the final trapping periods (Buckner 1966).  This information suggests that S. arcticus has a generation turnover rate of about two generations every year (Buckner 1966).  This disagrees with the other opinion that only adults from the previous year breed (Baird et al. 1983; Clough 1963).

S. arcticus populations are marked by a high mortality rate.  Fifty percent of a litter dies between birth and leaving the nest.  Another fourteen percent of the original litter dies within the first month out of the nest.  Only twenty percent reach sexual maturity (80 percent mortality), and only ten percent reach ten years of age.  Most if not all die before reaching fifteen months of age.  However, one individual was determined to be eighteen months old from marked recapture data. (Buckner 1966).

The home range of S. arcticus is estimated at about 1.46 ± 0.14 acres (n=78).  S. arcticus exhibits a slight intolerance of home range overlap especially at high population densities (Buckner 1966).

S. arcticus activity is highest during the night and lower than expected in the morning (between six and ten am) (Clough 1963; Buckner 1966). When active, S. arcticus does everything on the run.  However, they will spend frequent periods of inactivity laying on their sides or stomachs (Clough 1963).

A pair of S. arcticus has been observed capturing grasshoppers early in the day.  The two shrews would climb a stem to about 31 cm and jump on grasshoppers about 25 cm away.  Thirty-seven such attacks were observed in a fifteen-minute period.  Thirty-three of these attacks were successful (89%), although as the air temperature increased during the day the grass hoppers were able to quickly escape and successful attacks diminished.  This observation indicates that S. arcticus may use vision to find food during daytime foraging (Kirkland and Schmidt 1996)

It has been found that S. arcticus is relatively unaffected by the occurrence of wildfires.  This is probably due to the preference of shrews for hydric habitats.  Fire can pass through these habitats but cannot combust all plant litter and ground cover due to high moisture levels.  After one burn, the shrew population increased, but after a different burn did not.  This increase may just be a result of immigration (Gruebele and Steuter 1988).

Remarks:

As with other members of the Sorex araneus-arcticus group, S. arcticus possesses trivalent sex chromosomes.  Males have the three sex chromosomes X, Y1, and Y2, but females only have two sex chromosomes X and X.  The diploid karyotype is thus 29 for males and 28 for females and the fundamental number is 38.  The diploid karyotype in S. a. maritimensis is the same (males 29 and females 28) but the fundamental number is 34, not 38.

There exists three subspecies of S. arcticus:  S. a. arcticus in the northern portion of the species range (most of Canada), S. a. laricorum in the southern extent of the species range (Northern Midwest), and S. a. maritimensis in the eastern extent of the species range (Nova Scotia and land east of the St. Lawrence Seaway).  The range of S. a. maritimensis does not overlap with the ranges of the other two subspecies.

Due to differences found in craniometric measurements, genetic sequencing, karyotypic analysis, and appearance between S. a. maritimensis and the other subspecies it has been argued that S. a. maritimensis should not remain as an S. arcticus subspecies (Kirkland and Schmidt 1996; Stewart et al. 2002; Volobouev and van Zyll de Jong 1988).  These differences, especially genetic differences in mitochondrial cytochrome b sequence, suggest that S. a. maritimensis has probably been separate from the other two subspecies since the early Pleistocene.  This separation may be a result of allopatric speciation in eastern North America due to the Appalachian Mountains and/or large continental glacial lakes or seas (Stewart et al. 2002).  Current range distribution is probably a result of more recent climatic events (Wisconsin glaciation) that occurred after the time of probable divergence 2 million years ago (Stewart et al. 2002)

Literature Cited:

Baird, D. D., R. M. Timm, and G. E. Nordquist.  1983.  “Reproduction in the arctic shrew, Sorex arcticus.”  Journal of Mammalogy.  64:298-301.

Buckner, C. H.  1966.  “Populations and ecological relationships of shrews in tamarack bogs of Southeastern Manitoba.”  Journal of Mammalogy.  47:181-194.

Clough G. C.  1963.  “Biology of the arctic shrew, Sorex arcticus.”  American Midland Naturalist.  69:69-81.

George, S. B.  1988.  “Systematics, historical biogeography, and evolution of the genus Sorex.”  Journal of Mammalogy.  69:443-461.

Gruebela, M. J. and A. A. Steuter.  1988.  “South Dakota records of pygmy and arctic shrews: response to fire.”  Prairie Naturalist.  20:95-98.

Kirkland, Gordon L. Jr., and David F. Schmidt.  1996.  Sorex arcticus.  Mammalian Species No. 524.  American Society of Mammalogists.

Stewart, D. T., N. D. Perry, and L. Fumagalli.  2002.  “The maritime shrew, Sorex maritimensis (Insectivora: Soricidae): a newly recognized Canadian endemic.”  Canadian Journal of Zoology.  80:94-99.

Wrigley, R. E., J. E. Dubois, and H. W. R. Copland.  1979.  “Habitat, abundance, and distribution of six species of shrews in Manitoba.”  Journal of Mammalogy.  60:505-520.

Volobouev, V. T. and C. G. van Zyll de Jong.  1988.  “The karyotype of Sorex arcticus maritimensis (Insectivora, Soricidae) and its systematic implications.”  Canadian Journal of Zoology.  66:1968-1972.

 

Reference written by Justin Fleischman, Biology 378 student.  Edited by Christopher Yahnke.
Page last updated 4-29-04.