Recent records of the cotton mouse (Peromyscus gossypinus) in Illinois.
The cotton mouse (Peromyscus gossypinus) is fairly common throughout most of its range from Virginia to Florida, W to eastern Texas and N through Tennessee to western Kentucky. The species reaches the northwestern periphery of its range in the Midwest in extreme southern Illinois. Historically, limited numbers of P gossypinus were reported from five counties in Illinois: Alexander, Johnson, Pope, Pulaski and Union (Hoffmeister, 1989). However, none have been documented in Illinois since 1909. As noted by Hoffmeister (1989:215), "What has happened to P. gossypinus in southern Illinois remains a mystery. Ample search within the last 30 years has been made specifically for these mice. Trapping has been done in habitat that should be suitable for the species but no specimens of P gossypinus have been found."
Cotton mice occur in a variety of habitats, but in general are associated with bottomland forests and swampy areas (Wolfe and Linzey, 1977). Identification of cotton mice can be problematic because of their morphological similarity to other species of Peromyscus, including the white-footed mouse (P. leucopus) (Linzey et al., 1976; Engstrom et al., 1982; McDaniel et al., 1983). Here we report captures of cotton mice in Illinois for the first time since 1909.
Small mammals were livetrapped in July, August and September, 1996, as part of a larger investigation on 27 sites throughout the Horseshoe Lake Conservation Area (HLCA), Alexander County, Illinois. Located in the Mississippi River floodplain, HLCA encompasses 3702 ha. Horseshoe Lake is a shallow, eutrophic, oxbow lake surrounded by emergent wetlands, and swamps dominated by bald cypress (Taxodium distichum) and tupelo (Nyssa aquatica). Mesic and bottomland forest areas of HLCA include a variety of overstory trees, including pin oak (Quercus palustris), red maple (Acer rubrum), sweet gum (Liquidambar styraciflua), hackberry (Celtis occidentalis), cottonwood (Populus deltoides), box elder (A. negundo) and willow (Salix nigra). Understory on most sites is very sparse and often includes poison ivy (Toxicodendron radicans), greenbriar (Smilax sp.) and grape (Vitis sp.). Several trapping sites had large numbers of flood-killed snags and fallen logs.
Live traps were set in 90-m long transects with 10 stations each 10 m apart. Three traps were set at each station: one large Sherman (8 x 9 x 23.5 cm), one small Sherman (5.4 x 6.3 X 16.5 cm) and one specially designed shrew trap with a main compartment (20.3 x 4.5 x 4.5 cm) and a removable rear portion (5.0 x 4.7 x 4.7 cm). Sherman traps were baited with cracked corn and sunflower seeds, shrew traps with canned cat food mixed with cooking oil. All traps were operated for 3 consecutive days. Variable numbers of large Sherman live traps also were operated at six additional sites throughout HLCA during the same period.
A second study area was located in a woodlot in Carbondale, Jackson County, for a long-term study of southern short-tailed shrews (Blarina carolinensis). Live traps were set in a 13 x 12 grid with 10 m between stations. One small Sherman live trap or one shrew trap was set at alternate stations. When open, all traps were baited with the cat food/oil mix and checked at 3-h intervals for 24 h. The overstory habitat included pole-sized shingle oak (Quercus imbricaria), black cherry (Prunus serotina), osage-orange (Maclura pomifera) and persimmon (Diospyros virginiana). Understory vegetation included honeysuckle (Lonicera japonica) and common pawpaw (Asimina triloba).
All captured animals were toe-clipped to identify recaptures. Sex, reproductive condition and body mass were recorded. For all Peromyscus, length of the hind foot was measured. Animals were released at the point of capture, except for those that died in traps, or were removed either for voucher specimens or as presumed P. gossypinus. Specimens were prepared as standard skins and skulls and deposited in the SIUC Mammal Museum.
We based identification of Peromyscus gossypinus on the mensural characteristics noted by Hoffmeister (1977). These included hind foot length as well as three skull measurements: condylobasal length, crown length of the maxillary toothrow, and length of the nasals. We also used two discriminant function equations of Laerm and Boone (1994), one based on hind foot and skull length, the other on condylobasal length and maxillary toothrow length, to distinguish P gossypinus from P. leucopus. Finally, we examined the first and second lower molars of each specimen under a dissecting microscope for an entolophulid or a mesostylid. Hoffmeister (1977) found this characteristic in most of the cotton mice he examined, whereas it was absent in most white-footed mice. We identified an individual as a cotton mouse only when there was agreement using the criteria of both Hoffmeister (1977) and Laerm and Boone (1994).
At HLCA, a total of 7587 trap nights resulted in 618 individuals of nine different species captured 1498 times. White-footed mice comprised 81.1% of the individuals captured. Five male cotton mice (0.8% of all captures, 0.9% of all Peromyscus) were removed from three sites on HLCA (voucher specimens #3717 to #3721 SIUC Mammal Museum). The mean hind foot length of the five cotton mice (22.4 mm; SD = 0.89) was significantly greater than the hind foot length of 73 male P leucopus (mean = 20.6 mm; SD = 0.75) from the same three sites (t = 5.02, df = 76, P [less than] 0.0001). Likewise, mean body mass of the cotton mice (26.7 g; SD = 3.1) was greater than that of male P leucopus (mean = 21.0 g; SD = 2.5) from the same sites (t = 4.84, df = 76, P [less than] 0.0001). A mesostylid occurred in four of the cotton mice; the teeth of the fifth individual were too worn to make this determination.
Among 62 individual Peromyscus trapped on the Jackson County site, one confirmed cotton mouse was removed (SIUC #3722). Another individual we believe was a cotton mouse was marked and released. Hind foot length of both individuals (22 mm) and mean body mass (27.0 g) was significantly greater than the mean hind foot length (20.5 mm; SD = 0.56) and body mass (19.9 g; SD = 3.5) of 38 adult P. leucopus on the site (t = 3.74, df = 36, P [less than] 0.001 and t = 2.60, df = 8, P [less than] 0.032, respectively). The confirmed specimen also had a mesostylid. Although cotton mice have never been reported from Jackson County, the Carbondale site is only 29.0 km NE of Wolf Lake, Union County, a site where this species occurred in 1909 (Hoffmeister, 1989).
The five confirmed specimens of Peromyscus gossypinus from HLCA probably underrepresent its actual relative abundance at that site. Several other individuals we believe to be cotton mice were captured and released at HLCA. These include 13 individuals (12 males) with mean body mass of 27.4 g (range 26-29; Sd = 1.24) and mean hind foot length of 21.3 mm (range 19-23; SD = 0.95). They were marked and released on the same three trapping sites from which the known P gossypinus were removed.
Boone and Laerm (1997) reported a "complex pattern of clinal size variation" for Peromyscus gossypinus. Individuals at the periphery of the geographic range generally are smaller than those in the center of the range. Mean body and skull measurements of cotton mice from HLCA, although well within ranges of measurements reported by Hoffmeister (1977) and Laerm and Boone (1994), appeared to conform to this geographic pattern.
Despite extensive small mammal trapping in the state during the last 50 yr, cotton mice have not been reported in southern Illinois since 1909. It is possible they always have been in the region, but simply were overlooked or misidentified. To test for misidentified specimens, we examined all Peromyscus leucopus in the SIUC Mammal Museum with hind foot [greater than]21 mm or body mass [greater than]23 g (n = 8). Results were inconclusive. Hoffmeister's (1977) criteria are very conservative and referred all these specimens unquestionably to P leucopus. However, four specimens were marginally P gossypinus, based on at least one of the equations of Laerm and Boone (1994). None of these had a mesostylid, however. Conversely, cotton mice may have returned to the region relatively recently, perhaps in association with pronounced environmental changes such as the large-scale flooding in 1993 and 1994. Because the species is on the periphery of its range and commonly inhabits bottomland forest, cotton mouse populations in southern Illinois likely are ephemeral, and undergo periodic extirpation as a result of flooding or other extrinsic factors.
Acknowledgments. - This was part of a study funded by the Illinois Department of Natural Resources, Division of Natural Resource Review and Coordination. We thank C. Weickert for assistance with the field work. J. Whitaker and J. Laerm confirmed identification of cotton mice, and J. Laerm provided access to unpublished manuscript material.
BOONE, J. L. AND J. LAERM. 1997. Geographic and non-geographic morphological variation in the cotton mouse (Peromyscus gossypinus). J. Mammal., in press.
ENGSTROM, M. D., D. J. SCHMIDLY AND P. K. Fox. 1982. Nongeographic variation and discrimination of species within the Peromyscus leucopus species group (Mammalia: Cricetinae) in eastern Texas. Tex. J. Sci., 34:149-162.
HOFFMEISTER, D. F. 1977. Status of the cotton mouse, Peromyscus gossypinus, in southern Illinois. Am. Midl. Nat., 97:222-224.
-----. 1989. Mammals of Illinois. Univ. of Illinois Press, Urbana. 348 p.
LAERM, J. AND J. L. BOONE. 1994. Mensural discrimination of four species of Peromyscus (Rodentia: Muridae) in the southeastern United States. Brimleyana, 21:107-123.
LINZEY, A. V., D. W. LINZEY AND S. E. PERKINS, JR. 1976. The Peromyscus leucopus species group in Alabama. J. Alabama Acad. Sci., 47:109-113.
McDANIEL, V. R., R. TUMLISON AND P. MCLARTY. 1983. Mensural discrimination of the skulls of Arkansas Peromyscus. Proc. Arkansas Acad. Sci., 37:50-53.
WOLFE, J. L. AND A. V. LINZEY. 1977. Peromyscus gossypinus. Mamm. Species, 70:1-5.