The Effects of Cheatgrass on the Relative Abundance of a Northern Utah Snake Community
by Lucas K. Hall                                            

 

Abstract


Cheatgrass (Bromus tectorum), an invasive exotic annual plant, is known to diminish many shrub-dependent organisms; however, relatively little research has been done on how it affects snake communities.  I conducted a study on Antelope Island State Park, Davis County, Utah, where cheatgrass is abundant and may be negatively impacting snake populations.  Four study sites with different cheatgrass cover (37.90 – 69.23%) were established to measure relative abundance of snakes. I trapped snakes using funnel traps attached to drift fence arrays (one array per site).  A total of 35 individuals were captured between June and September 2005 (221 trap-array days) representing two snake species: the western yellow bellied racer (Coluber constrictor mormon) and the Great Basin gopher snake (Pituophis catenifer deserticola).  Linear regression was used to determine if a relationship existed between the percent of cheatgrass cover and species relative abundances.  A negative relationship was found between the relative abundances of both snake species and increasing cheatgrass density.  These results suggest that cheatgrass may negatively affect the snake community in shrub-steppe habitat.

 

Introduction

Cheatgrass (Bromus tectorum), an invasive

annual plant introduced from Eurasia, was detected in the western U.S. as early as the 1890’s (Mack, 1981; Novak and Mack, 2001).  Since its introduction in the Intermountain West, cheatgrass has come to dominate at least 200,000 km² of the shrub-steppe landscape (Mack, 1989).  Characteristics of cheatgrass that allow it to out-compete native perennials include abundant seed production, rapid germination (Stewart and Hull, 1949) and its superior competitiveness (Holmgren, 1956; Harris, 1967; Melgoza et al., 1990).  Cheatgrass alters plant community structure (Hulbert, 1955; Brooks, 2000), soil nutrient cycling (Walker and Smith, 1997; Evans et al., 2001; Wolfe and Klironomos, 2005), microclimate (D’Antonio and Vitousek, 1992) and fire frequency (Stewart and Hull, 1949; Young and Evans, 1978; Walker and Smith, 1997). 

Cheatgrass has diminished many shrub-dependent organisms (Pimentel et al., 2000) such as shrubland birds (Wiens and Rottenberry, 1985; Knick and Rotenberry, 1995, 2000), small mammals (Yensen et al., 1992; Gitzen et al., 2001), and lizards (Newbold, 2005); however, little is known about how it affects snake communities.  Cheatgrass cover does reduce the detectability of snakes (Hirth et al., 1969; Mortensen, 2004), but there may be potentially negative implications to the management and preservation of snakes inhabiting cheatgrass areas that have yet to be studied (e.g., locomotive performance and prey carrying capacity may be limited in dense cheatgrass).  In northern Utah, cheatgrass is widespread in shrub areas with native grasses and may negatively impact snake communities.  My objective was to determine if cheatgrass adversely affects snake abundance.

 

Materials and Methods

Study Area. – This study was performed on Antelope Island State Park (UTM: Zone 12; 395,961 E; 4,545,924 N; elevation 1,310 m; 11, 311 ha) located in Davis County, Utah.  Antelope Island is the largest island in the Great Salt Lake (Figure 1) and is not a true island due to its causeway and a naturally occurring land-bridge during dry years.  Cheatgrass is one of the most abundant plants on the island; other common shrub vegetation includes grasses and plants such as bluegrass (Poa spp.), buckwheat (Eriogonum spp.), wheatgrass (Agropyron spp.), rabbitbrush (Chrysothamnus spp.), and sagebrush (Artemisia spp.) (Marshall, 1940). 


Field Procedures. –
From April to July 2005 four study sites (3 ha each) were selected.  Sites were located ≥ 1 km of one another; sites C and D were established in Bridger Bay and sites A and B in White Rock Bay (Figure 1).  These bays were chosen for their like habitat and variety of cheatgrass densities.  Cheatgrass is common throughout the island, and each site included some; however, sites differed in percent of cover.  Cheatgrass coverage was determined by walking 20 transects from the center of each plot, placing a 1-m² wooden frame on the ground every 10 m and estimating cheatgrass cover within the square to the nearest 5% (Daubenmire, 1959), for a total of 200 samples per site.  Samples were then calculated to find the mean cheatgrass percentage / m² for each site. 


In the center of each site, a hardware cloth drift fence trap array (0.635 cm² mesh, 40 cm high × 30 m long) was established in an “X” pattern.  To eliminate gaps between the fence and soil irregularities, fences were buried at least 5 cm.  Attached to each fence were 10 double-ended, hardware cloth funnel traps (0.635 cm² mesh, 38 cm high × 1 m long)(Cavitt, 2000) placed 5 m apart.  Traps were covered with white corrugated plastic to reduce heat stress of captured animals.  Funnel traps were used as the primary method for data collection (Fitch, 1987).  Snakes captured by hand within sites were also recorded.      


Captured snakes were identified (Collins and Taggart, 2002) and sexed by hemipenal probing (Schaefer, 1934) with the exception of some neonates, in which case hemipenal eversion was performed (Rosen, 1991).  Snout-vent lengths (SVL) were measured by contouring a metric vinyl tape along side the individual’s snout to its vent.  Tail lengths (TL) were also recorded by measuring from the vent to the tail tip.  Snakes were weighed to the nearest ±1 g using a Pesola® spring scale.  Captured snakes were individually marked by clipping unique combinations of ventral scales (Spellerberg, 1977).  Once marked, snakes were released at the point of capture. 


As a result of an abnormally wet and cold spring, sites C and D were first opened in early-June.  Sites A and B however, were relocated to White Rock Bay to avoid bison (Bos bison) grazing areas and were reopened in mid-June and early-July, respectively.  Traps were closed from mid-July to early-August to prevent heat stress and trap mortality due to air temperatures exceeding 38° C.  Traps were closed in late September after an early frost.

           
Data Analyses. –
Site-specific relative abundances (number of snakes trapped / 10 trap-array days) were calculated for each species.  Linear regression was used to determine if a relationship existed between species relative abundance and the percent cover of cheatgrass.  SVL measurements were calculated to find the species mean according to site.  Linear regression was performed for both species to determine if dense cheatgrass cover affected snake SVL.  Chi-square analyses were used to verify if sex ratios (males:females) differed significantly from parity.  Statistical tests for regression and nonparametric analyses were conducted using Statistical Package for the Social Sciences (SPSS) software, version 13.0.  Significance level was set at α = 0.05 for all statistical tests. 

 

Results

The cheatgrass cover percentage analysis determined individual percentages for each site.  All four study sites were distinct in cheatgrass cover percentage (Table 1).  Twenty-eight western yellow-bellied racers (Coluber constrictor mormon) and seven Great Basin gopher snakes (Pituophis catenifer deserticola) were the only species of snakes observed and trapped during 221 trap-array days.  Racers were captured in all four sites, whereas gopher snakes were found in only three sites (A, B and C; Figure 2).  The desert-striped whipsnake (Masticophis taeniatus taeniatus) was not observed during the course of this study despite its recorded presence on the island (Mortensen, 2004).
 

Relative Abundance. – Only one marked snake (C. constrictor; site C) was recaptured.  Consequently, no estimates of population size could be derived.  Relative abundance of racers was negatively associated with percentage of cheatgrass cover (F = 22.065; df model, error = 1, 2; P = 0.042; R² = 0.917; Figure 2).  Likewise, gopher snake relative abundance showed a significant negative relationship to increased cheatgrass cover (F = 18.582; df model, error = 1, 2; P = 0.049; R² = 0.903; Figure 2). 

SVL Analyses.
– We found no significant relationship between racer and gopher snake SVL with the percent of cheatgrass (F = 0; df model, error = 1, 2; P = 0.994; R² = 0; F = 4.895, df model, error = 1, 2; P = 0.270; R² = 0.830).

Sex Ratios.
– The sex ratio (males:females) for racers did not significantly differ from parity (11:16; χ2 = 0.46, df = 1, P > 0.05).  There were too few data to statistically compare the sex ratio for gopher snakes (3:4).


Discussion


These findings show that snake abundance is comparatively lower in cheatgrass on Antelope Island, suggesting that increasing cheatgrass density negatively impacts snake communities.  However, during a replicate study (one performed 20 y after the original to monitor changes in reptile status in Idaho) Cossel (2003) found that snake


abundance was not cheatgrass.  Yet, the difference between our results may lie in the method through which we ranked vegetation abundance.  Cossel used transect walk-through and point / line intercept surveys to classify all the major cover types of vegetation on a four-point scale.  Subsequently, he could not detect a percentage difference among his highest (four, on his scale) cheatgrass rankings.  Furthermore, since Cossel’s study was a replicate, he was required to utilize previous study sites which exemplified an array of habitat variables; whereas I was able to limit my sites to a similar habitat pattern with the only differences being cheatgrass densities.  Thus, he was unable to directly measure cheatgrass effects on snake abundance using similar habitat with the sole variable being cheatgrass percentages.

Gopher snakes exhibited lower relative abundances than racers in all sites, particularly in site D, where gopher snakes were absent.  Their absence could be attributed to lower overall vertebrate prey abundance upon which they are dependent (Fitch, 1949; Rodríguez-Robles, 1998).  Racers were perhaps more abundant because they can seemingly maneuver more easily in cheatgrass (Hirth et al., 1969) than the slower, more robust gopher snakes (Mosauer, 1935).  However, studies measuring garter snake (Thamnophis elegans) locomotion in differing push-point densities (representing vegetation stalk thickness by inserting nails into a board), speed was reduced in both experimental populations of garter snakes as push-(cont'd from the print edition) point densities surpassed intermediate levels (Jayne, 1986; Kelley et al., 1997).  Coupling this could be the impediment that dense vegetation would have on burst speed, a trait that is a selected for among garter snakes (Jayne and Bennett, 1990) and a primary method by which racers obtain prey (Fitch, 1963).  Field observations of racers and gopher snakes at sites with greater cheatgrass density suggest that speed is hindered in both snake species (personal observation), which may affect the ability to forage and escape predators.   

Alternatively, cheatgrass may negatively affect prey abundance, as seen in some lizard species.  Though lizards do not represent the majority of gopher snake and racer diets (Klimstra, 1959; Rodriguez-Robles, 1998), their negative association with cheatgrass is noteworthy.  Working in northern Utah, Newbold (2005) showed that cheatgrass decreased desert horned lizard (Phrynosoma platyrhinos) locomotive performance and abundance.  Mortensen (2004) found lizard abundance and diversity to be considerably lower in areas dominated by cheatgrass on Antelope Island.  This may tentatively explain the absence of the lizard-consuming whipsnake (Parker and Brown, 1980; Camper and Dixon, 2000) during my survey and provide some explanation for the reduced abundance of gopher snakes and racers in high cheatgrass sites.

Birds and their eggs make up a small percentage of
gopher snake and racer diets (Fitch, 1963; Parker and Brown, 1980; Rodríguez-Robles, 1998).  Decreasing populations have been observed among Sage Sparrows (Amphispiza belli), Brewer’s Sparrows (Spizella breweri), and Sage Thrashers (Oreoscoptes montanus) because of habitat loss due to wildfire and consequently, exotic grass invasions (Wiens and Rottenberry, 1985; Knick and Rottenberry, 1995, 2000).  This, in turn, would reduce nesting sites of shrubland birds in areas dominated by cheatgrass, discouraging snake foraging in them.

Small mammals are important in gopher snake and racer diets (Fitch, 1949, 1963; Klimstra, 1959; Rodríguez-Robles, 1998; Shewchuk and Austin, 2001).  In a small mammal survey performed in the shrub-steppe of Washington, Gitzen et al. (2001) found that small mammal capture rates were reduced in cheatgrass sites.  Furthermore, in southern Idaho, Townsend’s ground squirrels (Spermophilus townsendii) and their burrows were also less abundant in areas with cheatgrass (Yensen et al., 1992; Van Horne et al., 1997).  During spring and summer censuses on Antelope Island in 2006, I found deer mouse (Peromyscus maniculatus) relative abundances steadily decreasing as cheatgrass density increased (L. Hall, unpublished data).  Reduced small mammal abundance does not only limit prey abundance for snakes, but also diminishes burrow system availability.  Both snake species use rodent burrows for oviposition (Fitch, 1949; Parker and Brown, 1972) and thermoregulation (Brown, 1973; Huey et al., 1989). 

Insects, principally orthopterans, constitute the bulk of racer diet (Klimstra, 1959, Fitch, 1963; Brown, 1973), whereas gopher snakes rely much less on insects (Parker and Brown, 1980; Rodríguez-Robles, 1998).  Among orthopterans, acridid grasshoppers are favored by racers (Brown, 1973; Shewchuk and Austin, 2001.  In Colorado, Craig et al. (1999) discovered acridids in a variety of habitats, some of which included cheatgrass.  In a related study from south-central Idaho, Fielding and Brusven (1993) found relatively high density, but reduced diversity of acridids in exotic annual sites.  Unfortunately, partially digested grasshoppers recovered from racer stomachs are only identified to family level, probably because of the difficulty of recognizing any distinguishing species features beyond that level.  Therefore, it is problematic when concluding if racers are selecting the few acridid species abundant in high cheatgrass areas; nevertheless, it could be one reason for racer occurrence (although lowered) in sites C and D. 

Invasive plant species are important contributors to biodiversity declines (Parker et al., 1999; Davis, 2003).  Racers and gopher snakes are essential predators and prey to a variety of organisms in shrub-steppe food webs (Fitch, 1949, 1963; Parker and Brown, 1980).  Consequently, negatively impacted population dynamics of these snakes could lead to biological disruptions in shrub ecosystems.  The results from this study support my initial hypothesis that cheatgrass cover negatively affects snake abundance. It may do so by encumbering locomotion and by not supporting an adequate density / diversity of potential prey species.

 

Literature Cited

Brooks, M. L.  2000.  Competition between

     alien annual grasses and native annual 

     plants in the Mojave desert.  American

     Midland Naturalist 144:92-108.

 

Brown, W. S.  1973.  Ecology of the racer,  

     Coluber constrictor mormon (Serpentes,

     Colubridae), in a cold temperate desert

     in northern Utah.  Unpublished Ph.D.

     Thesis.  University of Utah, Salt Lake

     City.

 

Camper, J. D., and J. R. Dixon.  2000.  Food

     habits of three species of striped 

     whipsnakes, Masitcophis (Serpentes:

     Colubridae).  Texas Journal of Science

     52(2):83-92.

 

Cavitt, J. F.  2000.  Fire and a tallgrass

     prarie reptile community: effects on

     relative abundance and seasonal  

     activity.  Journal of Herpetology. 34:12-

     20.

 

Collins, J. T., and T. W. Taggart. 2002,

     Standard Common and Current

     Scientific Names for North American

     Amphibians, Turtles, Reptiles and

     Crocodilians (fifth edition): Publication

     of the Center for North American

     Herpetology, Lawrence, Kansas, pp. 1-

     44.

 

Cossel, J. O.  2003.  Changes in reptile

     populations in the Snake River birds of

     prey area, Idaho between 1978-1979

     and 1997-1998: The effects of weather,

     habitat and wildfire.  Unpublished Ph.d

     dissertation, Idaho State University,

     Pocatello.

 

Craig, D. P., C. E. Bock, B. C. Bennet, and J.

     H. Bock.  1999.  Habitat relationship

     amond grasshoppers (Orthoptera:

     Acrididae) at the western limit of the

     Great Plains in Colorado.  American

     Midland Naturalist 142(2):314-327.

 

D’Antonio, C. M., and P. M. Vituosek. 

     1992.  Biological invasions by exotic 

     grasses, the grass / fire cycle, and

     global change.  Annual Review of

     Ecology and Systematics 23:63-87.

 

Daubenmire, R.  1959.  A canopy cover

     method of vegetational analysis. 

     Northwest Science 33:43-64.

 

Davis, M. A.  2003.  Biotic globalization:

     does competition from introduced

     species threaten biodiversity? 

     Bioscience 53(5):481-489.

 

Evans, R. D., R. Rimer, L. Sperry, and J.

     Belnap.  2001.  Exotic plant invasion

     alters nitrogen dynamics in an arid

     grassland.  Ecological Applications 11(5)

     :1301-1310.

 

Fitch, H. S.  1987.  Collecting and life-

     history techniques.  In R. A. Seigel, J. T.

     Collins, and S. S. Novak (eds.), Snakes:

     Ecology and Evolutionary Biology,

     Macmillan, New York, pp. 143-164.  

 

 


------.  1963.  Natural history of the racer,

     Coluber constrictor.  Publications of the

     Museum of Natural History of the

     University of Kansas 15:351-468.

 

------.  1949.  Study of snake populations 

     in central California.  American Midland

     Naturalist 41(3):513-579.

 

Gitzen, R. A., S. D. West, and B. E. Trim. 

     2001.  Additional information on the

     distributions of small mammals at the

     Hanford site, Washington.  Northwest

     Science 75(4):350-362.

 

Hall, L. K.  2006.  The Effects of Cheatgrass

     (Bromus tectorum) on Deer Mouse

     (Peromyscus maniculatus) Abundance on

     Antelope Island.  Unpublished

     undergraduate data, Weber State

     University, Ogden.

 

Harris, G. A.  1967.  Some competitive

     relationships between Agropyron

    spicatum and Bromus tectorum

     Ecological Monographs 37(2):89-111.

 

Hirth, H. F., R. C. Pendleton, A. C. King, and

     T. R. Downard.  1969.  Dispersal of

     snakes from a hibernaculum in

     northwestern Utah.  Ecology 50(2):332-

     339.

 

Holmgren, R. C.  1956.  Competition

     between annuals and young bitterbrush

     (Purshia triedentata) in Idaho.  Ecology 37

     (2):370-377.

 

Huey, R. B., C. R. Peterson, S. J. Arnold, and

     W. P. Porter.  1989.  Hot rocks and not-

     so-hot rocks: retreat-site selection by

     garter snakes and its thermal

     consequences.  Ecology 70(4):931-944.

 

Hulbert, L. C.  1955.  Ecological studies of

    Bromus tectorum and other annual

     bromegrasses.  Ecological Monographs

     25(2):181-213.

 

Jayne, B. C.  1986.  Kinematics of terrestrial

     snake locomotion.  Copeia 1986(4):915-

     927.

 

-----, and A. F. Bennett.  1990.  Selection on

     locomotor performance capacity in a

     natural population of garter snakes. 

     Evolution 44(5):1204-1229.

 

Kelley, K. C., S. J. Arnold, and J. Gladstone. 

     1997.  The effects of substrate and

     vertebral number on locomotion in the

     garter snake Thamnophis elegans

     Functional Ecology 11(2):189-198.

 

Klimstra, W. D.  1959.  Foods of the racer,

    Coluber constrictor, in southern Illinois. 

     Copeia 1959(3):210-214.

 

Knick, S. T., and J. T. Rotenberry.  2000. 

     Ghosts of habitats past: contribution of

     landscape change to current habitats

     used by shrubland birds.  Ecology 81(1):

     220-227.

 

-----.  1995.  Landcape characteristics of

     fragmented shrubbsteppe habitats and

     breeding passerine birds.  Conservation

     Biology 9(5):1059-1071.

 

Mack, R. N.  1981.  Invasion of Bromus

    tectorum L. into western North America:

     an ecological chronicle.  Agro-

     Ecosystems 7(1981):145-165.

 

-------.  1989.  Temperate grasslands

     vulnerable to plant invasions:

     characteristics and consequences.  In

     Drake, J. A. (ed).  Biological Invasions: A

     Global Perspective.  London: John Wiley,

     pp. 155-179.

 

Marshall, W. H.  1940.  A survey of the

     mammals of the islands in Great Salt

     Lake, Utah.  Journal of Mammalogy 21(2)

     :144-159.

 

Melgoza, G., R. S. Nowak, and R. J. Tausch. 

     1990.  Soil water exploitation after fire:

     competition between Bromus tectorum

     (cheatgrass) and two native species. 

     Oecologia (1990)83:7-13.

 

Mortensen, G.  2004.  Antelope Island

     herptile study.  Report submitted to the

     wildlife management department of

     Antelope Island, pp. 1-24.

 

Mosauer, W.  1935.  How fast can snakes

     travel?  Copeia 1935(1):6-9.

 

Newbold, S.  2005.  Desert horned lizard

     (Phrynosoma platyrhinos) locomotor

     performance: the influence of cheatgrass

     (Bromus tectorum).  Southwestern

     Naturalist 50:17-23.

 

Novak, S. J., and R. N. Mack.  2001.  Tracing

     plant introduction and spread: genetic

     evidence from Bromus tectorum

     (cheatgrass).  Bioscience 51(2):114-122.

 

Parker, I. M., D. Simberloff, W. M. Lonsdale,

     K. Goodell, M. Wonham, P. M. Kareiva, M.

     H. Williamson, B. Von Holle, P. B. Moyle, J.

     E. Byers, and L. Goldwasser.  1999. 

     Impact: toward a framework for

     understanding the ecological effects of

     invaders.  Biological Invasions 1:3-19.

 

Parker, W. S., and W. S. Brown.  1980. 

     Comparative ecology of two colubrid

     snakes, Masticophis t. taeniatus and

     Pituophis melanoleucus deserticola, in

     northern Utah.  Milwaukee Public

     Museum Publications in Biology and

     Geology 7:1-104.

 

------.  1972.  Telemetric study of

     movements and oviposition of two female

    Masticophis t. taeniatus.  Copeia 1972(4)

     :892-895.

 

Pimentel, D., L. Lach, R. Zuniga, and D.

     Morrison.  2000.  Environmental and

     Economic costs of nonindigenous species

     in the United States.  Bioscience 50:53-62.

 

Rodríguez-Robles, J. A.  1998.  Alternative

     perspectives on the diet of gopher snakes

     (Pituophis catenifer, Colubridae): literature

     records versus stomach contents of wild

     and museum specimens.  Copeia 1998(2)

     :463-466.

 

Rosen, P. C.  1991.  Comparative ecology and

     life history of the racer (Coluber

    constrictor) in Michigan.  Copeia 1991(4)

     :897-909.

 

Schaefer, W. H.  1934.  Diagnosis of sex in 

     snakes.  Copeia 1934:181.
 

Shewchuk, C. H., and J. D. Austin.  2001.  

     Food habits of the racer (Coluber

    constrictor mormon) in the northern part

     of its range.  Herpetological Journal

     11:151-155.

 
Spellerberg, I. F.  1977.  Marking live snakes

     for identification of individuals in  

     population studies.  Journal of Applied

     Ecology 14:137-138.

 

Stewart, G., and A. C. Hull.  1949.  Cheatgrass

     (Bromus tectorum L.)—an ecological

     intruder in southern Idaho.  Ecology 

     30:58-74.

 

Van Horne, B., G. S. Olson, R. L. Schooley, J. G.

     Corn, and K. P. Burnham.  1997. 
     Effects of drought and prolonged winter
     on Townsend’s ground squirrel
     demography in shrubsteppe habitats. 
     Ecological Monographs 67(3):295-315.

 

Walker, L. R., and S. D. Smith. 1997.  Impacts
     of invasive plants on community and
     ecosystem properties.  In J. O. Luken and J.
     W. Thieret (eds.), Assessment and
     management of plant invasions, Springer-
     Verlag, New York, NY, pp. 69-86.

 

Wiens, J. A., and J. T. Rottenberry.  1985.  
     Response of breeding passerine birds to
     rangeland alteration in a North American
     shrub-steppe locality.  Journal of Applied  
     Ecology 22(3):655-668.

 

Wolfe, B. E., and J. N. Klironomos.  2005. 

     Breaking new ground: soil communities
    and exotic plant invasion.  Bioscience 55(6)
    :477-487.

 

Yensen, E., D. L. Quinncy, K. Jonson, K.
     Timmerman, and K. Steenhof.  1992.  Fire,
     vegetation changes, and population
     fluctuations of Townsend’s ground
     squirrels.  American Midland Naturalist
     128(2):299-312.

 

Young, J. A., and R. A. Evans.  1978. 
     Population dynamics after wildfires is
     sagebrush grasslands.  Journal of Range
     Management 31:283-289.

 




                     

                   
Verbal Equinox Home

Click here to view Figure 1, Figure 2, and Table 1 that accompany this article.






Verbal Equinox Home