University of Nebraska Kearney

Springer, J. T., and A. W. Voigt.  1999.  Small mammal populations in a prairie/riparian forest ecotone.  Proceedings of the North American Prairie Conference 16:127-132.

SMALL MAMMAL POPULATIONS IN A PRAIRIE/RIPARIAN FOREST ECOTONE

Abstract:  The Platte River valley and the islands in it have become a riparian forest composed primarily of cottonwoods (Populus deltoides) and the eastern redcedar (Juniperus virginiana).  These lands are rarely cultivated or used to any extent for agriculture.  Therefore many nonegame wildlife species thrive in these areas without much disturbance by man.  Determining small mammal population sizes in this area was our main objective.  Live-traps were placed in 3 different habitats (open prairie, scattered redcedar and dense redcedar) on an island of the Platte River in south central Nebraska.  In the habitat we labeled open prairie, Kentucky bluegrass (Poa pratensis) dominated with big bluestem (Andropogon gerardii) in second place.  A total of 73 small mammals was captured in the open prairies:  primarily western harvest mice (Reithrodontomys megalotis) and deer mice (Peromyscus maniculatus).  The scattered redcedar usually surrounded the open prairie patches.  A total of 95 small mammals was captured in scattered redcedar:  primarily deer mice followed by white-footed mice (P. leucopus).  The dense redcedar habitat consisted of a Kentucky bluegrass understory with eastern redcedar leaves covering the floor.  A total of 91 small mammals was captured, consisting entirely of deer mice and white-footed mice.  The total of 259 captures represented 8 species of small mammals.  The removal of the eastern redcedar would reduce the population of deer mice and white-footed mice significantly.

PROCEEDINGS OF THE NORTH AMERICAN PRAIRIE CONFERENCE 16:127-132

Key words:  deer mouse, eastern redcedar, Juniperus virginiana, Peromyscus leucopus, Peromyscus maniculatus, Platte River, Reithrodontomys megalotis, Sherman live-traps, western harvest mouse, white-footed mouse.

Eastern redcedar (Juniperus virginiana) is not really a cedar but a member of the cypress family (Cupressaceae) and so named because of the distinctive cedar scent (Hugo 1990).  It is used to make “cedar” chests, pencils and other useful items in our everyday life.  Harvesting these trees can be an essential tool in managing their numbers.  It is very tolerant to drought and can grow in virtually any soil condition.  It is heavily resistant to herbicides and harmful insects (Buehring et al.  1971), and has been proven to be very tenacious. 

When these trees are destroyed for management purposes, what happens to the wildlife that use them for shelter and food? Holthiujzen and Sharik (1984) found that 71 species use the redcedar seeds as food, 63 of which were birds.  Other wildlife species use redcedars as shelter: for example deer (Odocoileus spp.), wild turkey (Meleagris gallopavo) and several species of small mammals.  Which species and in what numbers will often depend on the plants growing beneath the redcedars.  What they are will vary with soil condition, water and nutrient availability, and light intensity from the canopy cover of the trees. 

The objective of this study was to determine relative numbers of small mammal species present, plant densities of open prairie and understory plants of the eastern redcedar, and plant species diversity within 3 different habitats (dense redcedar, scattered redcedar and open prairie) near the Platte River. 

Although an ecosystem is always dynamic, removal of redcedars for habitat management will result in change in the small mammal community.  We suspect that the removal of the redcedars will result in a decline of some small mammal populations, particularly those that use the trees for shelter and food.  We further suspect that wholesale redcedar removal will affect small mammal populations that live on the forest edge and in the prairie.  

MATERIALS AND METHODS

This part of Nebraska has an annual precipitation of 63.35 cm, most of which occurs during April through October, and an average temperature of 9.7°C (Nagel et al. 1980).

The sites with open prairie had Kentucky bluegrass (Poa pratensis) and big bluestem (Andropogongerardii) predominating.  The 3 open prairie sites had not been grazed or cultivated for more than 15 years and the litter accumulation is up to 15 cm in some of the areas.

The scattered redcedar sites had predominantly Kentucky bluegrass with occa-sional eastern redcedar.

The dense red-cedar sites consisted primarily of eastern redcedar with a few cottonwood trees (Populus deltoides) interspersed.  Kentucky bluegrass was virtually the only species making up the understory.

Sixty-six Sherman live-traps were placed in each area: 22 traps per habitat.  The total areas covered by the trap lines varied.

The traps were baited with peanut butter twisted in wax paper (each resembling a Hershey’s Kiss^®).  Cottonballs were placed in the traps as nesting material to prevent hypothermia.  There was a total of 792 trap-nights from 18 September through 31 October 1997.

Captured small mammals were identified, weighed, measured and marked with a permanent ink marker on the inside skin of both ears for recapture identification.
We used X² analysis to determine whether the population density of 1 habitat was significantly different from the other 2 habitats. Species distributions by habitat were also tested using the X² analysis. 

Plants were collected from 4 sites: 2 samples were taken from each 10-m² quadrat in 1-m² subplots. Four plant samples were taken from the open prairie, and 4 samples were taken from dense redcedar areas.  The plants were identified and density levels were established.  Light intensity was measured at each 10­m² quadrat.

RESULTS AND DISCUSSION

Table 1 shows how many of each species were captured within the 3 habitats.

Fig. 2 shows the average capture frequencies of all habitats. 

Plant Density

The dominant plant species was Kentucky bluegrass in the open prairie locations and the dense redcedar understory (Table 2 and Table 3).

Light intensity averaged 7625 ft-candles in the open prairie and 1363 in the dense redcedar understory. 

Litter accumulation of the understory was quite extensive consisting mostly of redcedar needles and a few cottonwood leaves.

Open Prairie

A total of 73 captures was obtained in the open prairie sites (Table 1).  This was very nearly the number (73.11) that would have been expected if all habitats had equal captures per total area of habitat. Compared to the scattered redcedar habitat, the total numbers were not significantly different (P < 0.5).  Also, compared to the dense redcedar habitat, the total numbers were not significantly different (P < 0.25).

The capture frequencies of the open prairie sites were greatest for the western harvest mouse (Reithrodontomys megalotis) at 20.4%, followed by the deer mouse (Peromyscus maniculatus) at 8.9%.

White-footed mice (Peromyscus leucopus) had a capture frequency of 3.3%.  This surprised us somewhat because the white-footed mouse is typically a woodland inhabitant (Kantak 1996). But trees were never very far away.

The other species caught included both the short-tailed shrew (Blarina brevicauda) at 0.8% and the masked shrew (Sorex cinereus) at 0.7%.

The analysis of species distribution showed that open prairie had significantly more short-tailed shrews than the other 2 habitats (P < 0.01), even though the number was small.  There were significantly more harvest mice (P < 0.001) as well.  There were statistically fewer numbers of white-footed mice (P < 0.001), and deer mice (P < 0.001) compared to the other habitats.

Scattered Redcedar

A total of 95 captures was obtained in the scattered redcedar sites (Table 1).  This was statistically far fewer captures (P < 0.025) than would have been expected if all habitats had equal captures per total area of habitat.  However, this total was not significantly different from the total for the open prairie sites.   Here the deer mouse had the highest capture frequency at 23.3%.  This was followed closely by the white-footed mouse with a capture frequency of 21.7%.  All other species of small mammals in the scattered redcedar habitat had capture frequencies below 2%: house mouse at 1.9%, western harvest mouse at 1.2%, and the least frequently captured species was the meadow vole (Microtus pennsylvanicus) at 0.4%.

The analysis of species distribution when compared to dense redcedar only showed that the scattered redcedar habitat had significantly fewer total captures (P < 0.01), as well as fewer white-footed mice (P < 0.025).

When compared to open prairie only, there were significantly fewer short-tailed shrews (P < 0.025) and harvest mice (P < 0.001), and significantly more deer mice (P < 0.001) than would have been expected.  None of the other differences proved to be statisically significant.

Dense Redcedar

A total of 91 captures was obtained in the scattered redcedar sites (Table 1).  This was statistically far more captures (P < .025) than would have been expected if all habitats had equal captures per total area of habitat.  However, this total was not significantly different from the total for the open prairie sites in comparing just those 2 sites.

Only 2 species were captured within dense redcedar: deer mice and white-footed mice.  The average capture frequencies for deer mice was 25.9%.  White-footed mice had an average capture frequency of 18.3%. 

The analysis of species distribution when the dense redcedar habitat was compared to scattered redcedar habitat only showed that dense redcedar had significantly more total captures (P < 0.01), as well as more white-footed mice (P < 0.025).

When compared to open prairie only, there were significantly fewer short-tailed shrews (P < 0.05) and harvest mice (P < 0.001).  Also there were significantly more white-footed mice (P < 0.001) and deer mice (P < 0.001) than would have been expected.  None of the other differences proved to be statisically significant.

CONCLUSIONS

When woody plants invade prairies, it is inevitable that the tree/shrub canopy will result in environmental changes that affect the prairie flora and fauna (Gehring and Bragg 1992).  The same holds true for the reverse of this action.  When woody plants are removed from a habitat, environmental changes occur.  Most of this occurs naturally without human intervention.  The changing of flora means the loss of fauna due to death or emigration.  In our case, the removal of eastern redcedar would mean a significant loss or emigration of deer mice and white-footed mice.

Mitchell et al. (1995) found high-diversity low-density populations of small mammals within a pine community.  But when disturbed, the populations declined in numbers.  Their study suggested that in response to management practices of pine communities, the relative abundance of white-footed mice and other species would result in local disappearance. Those species would later recolonize if the habitat were left undisturbed.

Sekgororoane and Dilworth (1995) said that deer mice, short-tailed shrews, and masked shrews are a forest edge species.  They based this finding on the fact that these species showed no attraction to or avoidance of the forest edge.  Destruction of woodland habitat may result in substantial increase in resource competition among small mammals of the prairie and forest edge by small mammals from the woodland.  This could result in the loss of not just 1, but several species.

According to Kantak (1996), white-footed mice are microhabitat specialists in that they only inhabit areas that are more structurally diverse and that have a characteristic of a woodland.  Their success in grassy habitats is less than that of their relative, the deer mice, which is a microhabitat generalist.  Since most of our captures were of these 2 species (especially deer mice), the removal of the riparian forest could result in the decline of both species.

Since these 2 species serve as food to several other wildlife species (raptors, owls, coyotes, foxes, weasels, various snake species, etc), loosing significant numbers of them might not be desireable from a wildlife management standpoint.  Kantak (1996) suggested that such lose could be minimized by introducing some shrub species into the formerly forested area.

Prairie species captured during this sampling consisted of western harvest mice and short-tailed shrews.  Both prefer prairie with more than 2 years of litter built up (Springer and Schramm 1972).

Of the 2 species of voles typically found in mixed grass prairie, only the meadow vole was caught in fall 1997.  Prairie voles (Mictotus ochrogaster) have been caught on this property, just not during this study.  Both species prefer some litter buildup, but probably prefer a habitat with more typical prairie grass species and much less Kentucky bluegrass (Springer and Schramm 1972).  Removal of the eastern redcedar followed by the seeding of prairie grasses plus the use of occasional fire for management might increase numbers of western harvest mice (Springer 1986) and voles. This could offset the loss of white-footed mice and decline of deer mice.

LITERATURE CITED

Springer, J. T., and A. W. Voigt.  1999.  Small mammal populations in a prairie/riparian forest ecotone.  Proceedings of the North American Prairie Conference 16:127-132.