This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain.

Plant Compositional Change in a Colony of Black-Tailed Prairie Dogs in South Dakota 1

Richard P. Cincotta,2Daniel W. Uresk,3 and Richard M. Hansen4

Abstract.--Peak season multi-species cover of vegetation and burrow mound density were estimated for 3 years along a transect that ran from the geometric center of a black- tailed prairie dog (Cynomys ludovicianus) colony, to its edge. The forb dominated core of the colony expanded 25 m in radius during the study, while composition changed dramatically in a small (<0.3 ha) zone midway between the edge and the core. We determined that year to year func- tional increases in multi-species canopy cover, described by the natural growth function (~2=0.72, P<0.001), occurred only after shortgrasses were reduced below 75% cover. The density of burrow mounds was positfvely correlated to com- positional change (r=0.58; P<0.01). We observed that bur- row mounds provided-early sites for the eatablishment of forbs. However, after the canopy cover of shortgrasses receded below 75% (in this location, probably from 4 to 7 years after initial inhabitancy by prairie dogs), exten- sive compositional changes occurred between burrow mounds.

INTRODUCTION et al. 1983, Archer et al. 1987). The pattern of plant composition appears to recapitulate colony Largely because of the influence that the expansion, i.e. as prairie dog colonies expand black-tailed prairie dog (Cynomys ludovicianus Ord) outward, a forb-dominated community follows. exerts on trend in rangeland vegetation, the spe- The objective of our study was to deter■ ine the cies has been a target of extermination or inten- extent, rate, and pattern of prairie dog induced sive control since the early 1900's (Merriam 1902, changes in plant composition within a colony on Schenbeck 1982, Uresk 1987). Plant succession mixed-grass prairie habitat. within colonies has been described as consisting of the initial disappearance of perennial grass cover, followed by an increase in shortgrasses MATERIALSAND METHODS (Bonha■ and Lerwick 1976), and an eventual in- crease and dominance by annual forbs and dwarf Study Area shrubs (Koford 1958, Garrett et al. 1982, Coppock The study was conducted in Badlands National Park, in southwestern South Dakota, in a colony of black-tailed prairie dogs situated in ■ ixed-grass 1 Paper presented at the Ninth Great Plains prairie, north of the edge of the White River Bad- Wildlife Da■ age Control Workshop. (Fort Collins, lands. Vegetation of the area is wheatgrass-grama­ Colorado, April 17-20, 1989). •Richard P. Cincotta is a post-doctoral schol- buffalo grass type described by Kuchler (1975). Depart ■ent Occurrence of flora and fauna in the area is de- ar, of Anthropology, University of Cali- scribed by A~new et al. (1986). Descriptions of fornia, Los Angeles. climate 8 W. soils, topography and are available in Daniel Uresk is Project Leader, USDAFor- Cincotta et al. (1987), and Uresk and Schenbeck est Service, Rocky Mountain Forest and Range Exper- i ■ent (1987). The prairie dog colony selected was <10 Station, Rapid City, SD. years old at the beginning of the research. It was 'Richard M. Hansen is Professor Emeritus, Range located approximately 1 k■ north of the valley Science Department, Colordao State Univer- ridge, or "wall", that marks the northern extent of sity, Fort Collins.

171 the White River Badlands, about 16 ka. south fro■ bracteata Lag. & Rodr.). All species with a cover Wall, South Dakota. When prairie dogs first be- value greater than 2.0% were retained, including caae established, the location was a privately nearly all local ■ajor forage species for livestock owned horse pasture. Livestock Bison (Bison.!!!.!!!!!!) (Uresk 1986) and prairie dogs (Uresk 1984, grazed on the colony, during the study, however Pagerstone et al. 1981). The absence of blue livestock did not utilize the area. graaa (Bouteloua gracilis (H.B.K.) Griffiths) fro■ this list is due both to its paucity on this site Changes in Plant Coaposition and our inability to distinguish it fro■ buffalo grass when the grasses were in a clipped condition. The colony was saapled once annually, fro■ 1981 to 1983, during the peak period for bioaass of ungrazed vegetation (the last week in July and The difference in cover betweens and S', was first week in August). In order to intersect the calculated as the euclidean distance ■easure be- full range of vegetation, the canopy cover of in- tween ■ulti-species ■eans: dividual plant species was evaluated along three parallel center-to-edge transects, which we have 20 collectively teraed a profile. Five aeters apart, 0.~(I (C£ ■ -C£a•)•)o.a the three replicates traced straight lines fro■ 1-1 the origin of the colony (the oldest burrows, pin- pointed by the foraer property owner) near the where C£ is the cover of a species. geoaetric center of the colony, to points (525 a away) on the western edge of the colony. To test the null hypothesis, H0 :D.=0 (n.=3, n.,•9), we used a per■utation technique, MRPP Along the profile, we chose an interval of (Mielke 1986, Mielke et al. 1981, Mielke et al. 25 • to separate saapling sites, S (Ss0,1,2, ... , 1976). MRPPutilizes the su■ of the euclidean 21), inside the colony (on-colony). At each s, distances (weighted for group size; Berry et al. we established three sampling transects, one on 1983), d, between all possible within-group pairs each replicate of the profile. To quantify local to express concentration within a particular ■utually exclusive, exhaustive grouping with group plant coaposition without prairie dogs, we ex- sizes g1,g., ... ,g. tended the profile 50, 75, and 100 a beyond the (Berry et al. 1983; for detailed exa■ple, Ziueraan colony, thus establishing a site, S', of nine see et al. 1985). The null saapling transects outside the colony (off-colo- hypothesis actually proposed with MRPP, that all per■utations of g1,g., ... ,g. are equally likely, ny). Bach transect was 29 • ••>long, along which is tested by: (1) ordering the coaputed values. of thirty (50 ca by 20 ca; 0.1 quadrats were d, (2) locating the relative position of the sta- placed, 1 • apart. Canopy cover per species was tistic on the list and (3) determining the P-value estiaated by recording the appropriate class, fro■ six possible cover classes (Daubenaire 1957), for as the proportion of all values of d less than or each plant species present in the quadrat. equal to the observed value of the a priori group- ing. We assu■ed that the group aean at each s was fro■ ■ore extreae .dis­ We defined plant coaposition as the canopy different S' when the P of cover of plant species encountered, i.e. aulti- tances was S,0,05. The full set of species encount- ered in sa■pling was used in this co■putation. apecies cover. To coapare aulti-di ■ensional data (on-colony vs off-colony) with a co■puter algo- Plant co■position was further described by rith■ for the ■ulti-reaponse per■utation procedure (MRPP;Berry and Mielke 1983), we reduced the data calculating the ratio of forb to grass cover at saapling sites. Forbs included all broad-leaved to 20 plant species (fro■ 75 encountered) by se- . Grasses included species fro■ the taxo- lecting the twenty species with the highest ■axi- no■ic faailiea of and Cyperaceae (sedges). ■u■ cover aaong pooled sites. These species (co■- All species encountered during sa■pling were in- ■on and scientific na■es according to Van Bruggen cluded in this computation. 1985) were: western wheatgrass (Agropyron s ■ ithii Rydb.), red threeawn (Aristida purpurea Steud.), dwarf sagebrush (Arteaisia cana Pursh) Aster fal- Co■positional ~ Lindl., Japanese chess (Broaus J~poiiTcus-- Rate of Changes tectoru■ Thunb.), cheat grass(.!!.:,_ L.), buffalo ti ■e grass (Buchloe dactyloides (Nutt.) Engel ■.), Carex Since we did not know the exact of eleocharis Bailey, Chenopodiu■ strictu■ Roth,--- prairie dog eatablishaent on all points along the horseweed (Conyza canadensis (L.) Cronq.), six­ profile, we calculated a growth rate relative to su■■ er the previous year's state (Green 1979). Thus, we weeks fescue (Pestuca octoflora Walt.), assu■ ed cypress (Kochia ·scoparia (L.) Schrad.), poverty that D[t+ t] was a function of D[t), where weed (Monolepis nuttalliana (Schultes) Greene), t equaled 1 year, and twas peak season during Russian each of the first two study years. Since D has a buckhorn (Plantago patagonica Jacq.), li ■ it ■axi ■u■ thistle (Salsola iberica Sennen & Pau), tu■ ble­ finite upper (the possible difference grass between co■pared cover saaples), we fit a si ■ple (Schedonnardus panniculatus (Nutt.) Trel.), aay■ptotic (Solanu■ trifloru■ growth curve, the natural growth func- cut leaf nightshade Nutt.), tion [f(x)•a(l-e-b•)J, to this data. Paraaeters scarlet ■allow (Sphaeralcea coccinea (Pursh) (a,b) were esti ■ated by least squares esti ■ate Rydb.;. sand dropseed ( cryptandrus using a non-linear regression algorith■ (Marquardt (Torr.) A.(jray), and prostrate vervain (Verbena 1963).

172 Effects of Burrowing ... 10 ►- 11181 During the saae period, the nu■ber of burrow - 11182 0 ----a-••·-- ■ounds were counted within a 25 • x 25 • square -- 11183 =&IS (0.0625 ha) of which the saapling site was the a: deter■ ined center. We the relationship between IIO .1 the density (ha-~) of burrow ■ounds (independent variable, rando■ effect) and D (dependent vari- c:,e able) by regression. Por calculations of r, .01 significance (P~0.05) was evaluated using the ,D P-statistic (Cacoullos 1965). ..0 .001 II. 0 100 200 300 400 500 CR IICZ RESULTS

Changes in Plant Co■position 1.0 1981 The off-colony site was do■ inated almost co■- 0.8 -- 1982 ----- !'-,. pletely by buffalo grass, western wheatgrass, ----a-··· 11183 .. Japanese bro■e . . (Table 1), while sand dropseed and 0.8 ■ P<0.05 ... . six-week minor . . fescue were co-unity constituents 0 . . (<5% . . cover). Other species present (<2.0% cover), 0.4 ~ though not a■ong the 20 species use to calculate ■ulti-species cover, were green needlegrass (Stipa viridula Trin.) and needle-and-thread (S. co■ ata 0.2 Trin. & Rupr.). On-colony sites near the edge had low forb:grass ratios (Pig. la) though ■oat peren- 0.0 nial aid-grasses, e.g. sand dropseed, green need- 0 100 200 300 400 500 legrass and needle-and-thread were virtually absent Distance from Colony Center (m) (<0.5% cover). Porb: grass ratios were highest (1.73:1 in 1981, 1.62:l in 1982, and 8.59:1 in Pigur~ 1.--Three year comparisons uf: (a) the 1983) in the core of the colony (fro■ Oto 50 • ratios of forb to grass cover along the fro■ the center). A taxono■ ically diverse array colony profile: (b) the differences in mul­ of annual forbs were do■ inant at the core, includ- ti-species (20 species) cover. D, between ing prostrate vervain, buckhorn, cut leaf night- on-colony and off-colony sites along the shade, rough pigweed (Allaranthus retroflexus L.), profile. Locations of zones of extensive tu■ble- tu■bleweed (A. albus L.), poverty weed compositional change, the core (CR) and and Russian thistle. Tumblegrass, a perennial mid-colony zone (MCZ). are indicated. gra■ inoid that frequents disturbed soils, was also an i ■portant constituent of this co■aunity. Un- expectedly, forb: grass ratio increased fro■ 0.08:1 During the study, D increased mainly in the (1981) to 1.21:1, two years later, in a small (<0.3 core and surrounding MCZ (Pig. lb). Also, Din- ha) ■ id-colony zone (MCZ) about 350 • fro■ the creased on sites adjacent to the original core, colony center. This zone was co■pletely surrounded thus enlarging its radius by about 25 • in two grass do■ inated co■■unities. by The dominant con- years. Although there was a positive correlation stituents of MCZwere prostrate vervain, horseweed, between forb:grass ratio and D (r=0.72, P=6.14, 64 and red threeawn (a perennial grass). df, P<0.001), forb:grass ratio was not always a true indicator of change; some compositional chan- ges involved the replacement of perennial grass species by other grasses (e.g., buffalo grass was soaeti ■es replaced by red threeawn or tumblegrass). TABLE 1. Cmopy cover of plan! species (>2.0'1, cover) in lbe off-colony lite. Rate of Compositional Change Canopy cover (± SD) Upon plotting D[t+l] as a function of D[t], off-<:Olooy we noted a difference between points representing "highly disturbed" zones of the colony (the core Species 1981 1912 1913 and MCZ), and those from the remaining locations n-9 n-9 n=9 along the profile. Highly disturbed zones showed a high positive correlation between yearly states ------'I, Buffalo gru1 10 (6) 79 (12) 16 (I) (r=0.84, P=ll.5, 12 df, P<0.001). Application of the the natural growth function to these data (Pig. Western wheatgra11 S (4) 6 (2) S (3) 2) yielded a good fit (R8 =0.72, P<0.001). The Sand dropseed 2 (2) 2(<1) 2(<1) re■aining points, where grasses dominated, were Six-week fescue 3 (l) 4 (1) 2 (2) clumped near the origin. On these sites, yearly states were 11egat1v,dy cur-i-1'1;:ited tr·• 0.53, F=3.32, Japanese brome S(I) 4 (3) 9 (S) 28 df, P

173 Effect of Burrowing 1.0 The density of burrow ■ ounds and D (Fig. 3) • 1981 ■ ■ 1982 were positively correlated (r=0.58; F=3.76, 64 df, 0.8 • ■ P<0.01). However, linear, exponential, and poly- ■ 1983 ■ nomial regressions of these variables yielded poor fits (Ra<0.38). Burrow mounds, and disturbed soil 0.6 • directly adjacent to ■ounds were observed to be Q Iii sites for initial establish■ ent of annual forbs. 0.4 • In highly disturbed zones, annual forbs, dwarf Iii • • ■ shrubs (e.g., pasture sagebrush [Arte■ isia frigida • • Willd.]) some ■ and "pioneer" perennial grasses 0.2 ■ Ill ; I ' Iii El I ■ ■ (e.g., red threeawn and tumblegrass) occupied the ■ II • • I I ■ ground surface between aounds. Iii 0.0 • I 0 100 200 300 Burrow Mounds (per ha) Discussion

The differential extinction, replacement and Figure 3.--The relationshp between the difference resilience of plant species during prairie dog in multi-species cover between on-colony and inhabitancy create the observed pattern of commu­ off-colony sites. D, and the density of nity change. Knowing the long-term response of prairie dog burrow mounds. siailarly behaving plant species (Harper 1977; rather than taxono ■ ic affinities) may help range and wildlife managers understand prairie dog in- duced succession. We considered species on the prairie dog colony to fall roughly into four cate- gories: (1) perennials that quickly disappeared after initial prairie dog inhabitancy; primarily aid-grasses, e.g. sand dropseed, green needlegrass, and needle-and-thread; (2) shortgrasses that were initially resistant to the impacts of prairie dog grazing, e.g. buffalo grass (Fig. 4): (3) annuals that became established on recently disturbed soil associated with burrow mounds, e.g. buckhorn, pros- trate vervain, and scarlet ■allow; (4) annuals and perennials that only became established on the most disturbed sites. e.g. oovertv weed. Although the

o y - 1.108 (1.0- exp{-2.4xD R•2. 0.72 1.0 .. 0 .... .•· •· 0.8 ..• ... 0 ...... •"' ,- 0.6 0 ...... "" .. -+ ...... 0 ..... - 0.4 ... Q .------, - • •• .-·· + >75% shortgrass canopycover 0.2 + w--·••• o c75% shortgrass canopy cover . --~♦ 1:1

0.0 4-:::·'°=-----.---~===;:===;==~ DISTANCE FROM COLONY CENTER (M) 0.0 0.2 0.4 0.6 0.8 1.0 D [t]

Figure 4. Mean canopy cover for six plant species Figure 2.--Year to year changes in the difference along the colony profile for three consecutive years. The species are western wheatgrass between multi-species cover, D, between on- (Agsm), colony and off-colony sites. The natural buffalo grass (Buda), poverty weed growth function was fit to points (n=14) that (Monu), Patagonia Indianwheat (Plpa), scarlet went below 75% shortgrass canopy cover during mallow (Spco), and prostrate verfain (Vebr). the study. A year to year increase in D was The Y-axis is transformed using Log10 (Y+l). Distance from the colony center to an edge not observed for points (n=30) above this m m threshold. 525 distant was measured in 25 intervals.

174 cover of western wheatgrass initially dec.reased was 75% cover: sites that went below tills level with prairie dog inhabitancy, the species was not experienced abrupt changes in coaposition during displaced entirely as were all other aid-grasses. the following year. Thus, sites below threshold The ability of this species to aaintain a presence slipped froa teaporary stability into the asyapto- under intense prairie dog grazing aay be due to the tic increase in D that we have described. survival of decuabent "grazing" ecotypes that are soae known to occur in prairie dog colonies Although coapositional changes in colonies are (Detling and Painter 1983). likely to be aost evident in an expanding core, ir- regular patches closer to the periphery aay undergo At the colony edge. plant coaposition was change, as well. Archer et al. (1987) concluded not radically affected by the loss of aid-grasses. froa that the foraation of forb doainated co-unities in Reduction of aid-grass cover, resulting prairie dog colonies could be attributed mostly to prairie dogs clipping tall plants for predator the length of time of sustained prairie dog activi- avoidance rather than for forage (King 1955), aay aaount may aore aesic ty. The initial of shortgrass cover have a greater consequence on grasslands also affect the rate of prairie dog induced succes- (Archer et al. 1987). The initial 2 yr period of fro■ sion. Since prairie dog colonies are aggregates of soil disturbance, resulting building burrow highly territorial family groups (coteries) rather mounds and allowing the introduction of soae annual ■ !nor coaaunity. than a cooperating colonial entity (King 1955, forbs, had a iapact on the plant Hoogland 1981), population and activity of prairie However, long-tera inhabitancy of prairie dogs on dogs is non-uniforaly distributed across the colo- aixed-grass prairie vegetation in our location (we fro■ ny. Whereas territories at the core are contigu- estiaate 11 to 13 yrs) can cause the coaplete ous, those at the edge of expanding colonies are disappearance of perennial shortgrasses between often spread spaciously aaongst relatively undis- mounds. turbed vegetation. Thus, compositional changes coa­ outside the core are likely to be patchy. In our The nature of soil disturbance and plant study site, coapositional changes outside the core aunity structure varied ■arkedly between the two area (in MCZ) appeared to result fro■ sustained characteristic types of burrow aounds (King 1955, inhabitancy of a single coterie on a site with Sheets et al. 1971), (1) doae aounds and (2) crater below average shortgrass cover. mounds. Whereas these structures may reach 1 • in height and 2.5 • in diameter in an old colony (King In aany colonies in the Badlands area, the 1955), all were less than half these diaensions in iaparts Doae presence of a shortgrass understory resil- our colony. aounds were coaposed of loosely iency to the plant community, delaying the eventual packed subterranean soil spread widely over the shift to annual forbs that is usually incoapatible ground surrounding the entrance. These mounds with range livestock management objectives. It becaae sites for the establishaent of a variety of should be noted, however, that prairie dog colonies forbs that assuaed a prostrate habit, either as a rumi­ fora provide valu~ble forage resources for native consequence of their natural growth (Warwick nants; pronghorns (Antilocapra aaericana) prefer and Briggs 1980) or because they were heavily clip- the high quality herbage in colony cores (Krueger ped by prairie dogs. Prostrate vervain, tuabHng 1986), while bison are attracted to the highly ■ustard, buffalo bur (Solan1111rostratua Dunal), and doae nitrogenous, low fiber regrowth of grasses at the cut leaf nightshade were frequent occupants of edges of colonies (Coppock et al. 1983). mound sites in our study colony. Extrapolation of results fro• this study to other prairie dog colonies should be done cautiously. In "Crater aounds" were narrow, roughly cone- fact, where plant composition and herbivore use is shaped structures that prairie dogs constructed auch different froa the single colony studied. fro• uprooted vegetation, displaced litter, huaus, extreae care should be used in extrapolating these and aineral soil which they scraped froa a patch results to other areas. adjacent to the burrow entrance. After a rain, prairie dogs packed the aaterial tightly with their noses. Thus, surfaces of the crater aounds aade ACKNOWLEDGEMENTS poor sites for seedling establishaent. However, adjacent "quarried" patches were invaded by annual This study was sponsored by a research grant forbs, aost frequently buckhorn, during the follow- froa the National Park Service (Contract No. CX12- ing spring. 00-1-B035) and logistic support froa the US Forest Service Forest and Range Experiaent Station in On our colony, the buffalo grass doainated Rapid City, South Dakota. We thank Superintendent co-unity experienced a resilient period (probably Gil Blinn, Chief Ranger Lloyd Kortje, District froa 4 to 7 yrs) during which little change occurr- Ranger Mike Glass and the staff of Badlands Nation- ed. This aay be a period during which root carbo- al Park, as well as the US Forest Service Experi- hydrate reserves were being depleted (Santos and aent Station technical staff, for their assistance. Trlica 1978), both to supply above ground regrowth Invaluable assistance was rendered by colleagues in and froa increased aicrobial grazers below ground the field, including Willia■ Agnew, Debra Paul, (Inghaa and Detling 1984). A decisive shift in Kelly Brown, Wangoi Migongo, and Bw. Mkuu S. Baun!. coaposition was experienced when shortgrass was Monte Garrett and Layne Coppock at Wind Cave N.P. replaced by a aixture of araed and/or sprawling provided helpful i;uggestions based upon ··their own· grasses and forbs, aroaatic dicots, and bare field observati0ns. Willia■ Agnew coapiled plant ground. The shortgrass "threshold" at our location speciaens invaluable to the research.

175 LITERATURECITED Ingham, R. E., and J. K. Detling. 1984. Plant- herbivore interactions in a North American Agnew w D W Uresk, and R. M. Hansen. 1986. aixed-grass prairie. III. Soil neaatode Flora and fauna associated with prairie dog populations and root biomass dynamics on colonies and adjacent ungrazed aixed-grass Cynomys prairie in western South Dakota. J. Range ludovicianus colonies and adjacent Manage. 39:135-139. uncolonized areas. Oecologia 73:307-313. s., M. King, J. A. 1955. Social behavior, social Archer, G. Garrett, and J. K. Detling. dynamics 1987. Rates of vegetation change associated organization, and population in a with prairie dog (Cynoays ludovicianus) black-tailed prairie dog town in the Black grazing in North American mixed-grass prairie. Hills of South Dakota. Contrib. Lab. Vert. Biol., No. 67, University of Michigan, Ann Vegetatio 72:159-166. Kvamme, Arbor. Berry, K. J., K. L. and P. W. Mielke. Koford, c. B. 1958. Prairie dogs, whitefaces, and 1983. Improveaents in the permutation test blue grama. Wild!. Monogr., No. 3. for the spatial analysis of the distribution Krueger, K. 1986-. Feeding relationships among of artifacts into classes. American Antiquity bison, pronghorn, and prairie dogs: an experi- 48:547-553. mental analysis. Ecology 67(3):760-770. Berry, K. J., and P. W. Mielke. 1983. Coaputation Kuchler, A. W. 1975. Potential natural vegetation of finite population parameters and approxi- mate of the conterminous United States. A■ . Geogr. probability values for multi-response Soc., Spec. Puhl. No. 36. perautation procedures (MRPP). Couunications Marquardt, D. W. 1963. An algorithm for least in Statistics-Simulation and Computation squares estimation of nonlinear parameters. B12:83-107. J. Soc. Industr. Appl. Math. 2:431-441. Bonham, c. D., and A. Lerwick. 1976. Vegetation Merriaa, H. 1902. The prairie dog of the Great changes induced by prairie dogs on shortgrass c. Plains. p. 257-270. In: Yearbook of the range. J. Range Manage. 29:221-225. between United States Department of Agriculture cacoullos, T. 1965. A relation the t and F (1901). US Gov. Print Off., Washington, DC. distributions. J. Amer. Statist. Assoc. Mielke, P. W. 1986. Non-metric statistical 60:528-531. W. M. analyses: some metric alternatives. J. Stat. Cincotta, R. P., D. Uresk, and R. Hansen. Plann. Inference 13:377- 387. 1987. Demography of prairie dog populations Mielke, P. W., K. J. Berry, and E. S. Johnson. reoccupying sites treated with rodenticide. 1976. Multi-response permutation procedures Great Basin Natur. 47:339-343. for a priori classifications. Commun. Coppock, D. L., J. K. Detling, J.E. Ellis, and M. Statist.-Theory Meth. A 5:1409-1424. 1. Dyer. 1983. Plant-herbivore interac- Mielke, P. W., K. J. Berry, P. J. Brockwell, and J. tions in a North American mixed-grass prairie, s. Williams. 1981. A class of nonpara- 1. Effects of black-tailed prairie dogs on metric tests based on multiresponse permuta- intraseasonal aboveground plant biomass and tion procedures. Biometrika 68:720- 724. nutrient dynamics and plant species diversity. Santos, G. L., and M. J. Trlica. 1978. Clipping Oecologia (Berl) 56:1-9. effects on production and carbohydrate Detling, J. K., and E. L. Painter. 1983. Defolia- reserves of blue grama and western wheatgrass. tion responses of western wheatgrass popula- p. 394-386. In: Proc. First Int. Rangeland tions with diverse histories of prairie dog Congr. (ed., D.N. Hyder). Soc. Range Manage., grazing. Oecologia 57:65-71. Denver, CO. Daubenmire, R. 1959. A canopy coverage method of Schenbeck, G. L. 1982. Management of black-tailed vegetational analysis. Northwest Science prairie dogs on the national grasslands. p. 35:43-64. 207-213. In: Proc. Fifth Great Plains Fagerstone, K. A., H. P. Tietjen, and O. Willius. Wildlife Damage Control Workshop (eds. R.M. 1981. Seasonal variation in the diet of Ti•• and R.J. Johnson). University of black-tailed prairie dogs. J. Mammal. Nebraska, Lincoln. 62:820-824. Sheets, R. G., R. L. Linder, and R. B. Dahlgren. Garrett, M. G., J. L. Hoogland, and W. L. Franklin. 1971. Burrow systems of prairie dogs in South 1982. Demographic differences between an old Dakota. J. Mammal. 52:451-453. and a new colony of black-tailed prairie dogs W. (Cynomys Uresk, D. 1984. Black-tailed prairie dog food ludovicianus). AIII. Midl. Natur. habits and forage relationships in western 108:51-59. South Dakota. J. Range Manage. 37:325-329. Green, R.H. 1979. Sampling design and statisti- Uresk, D. W. 1986. Food habits of cattle on cal methods for environmental biologists. aixed-grass prairie on the northern Great John Wiley & Sons, New York. Plains. Prairie Natur. 18:211-218. Harper, J. L. 1977. The population biology of Uresk, D. W. 1987. Relations of black-tailed plants. Academic Press, London. prairie dogs and control programs to vegeta- Hoogland, J. L. 1981. The evolution of colonial- tion, livestock and wildlife. p. 312-323. ity in white-tailed and black-tailed prairie In: J.L. Capinera (ed.), Integrated pest dogs (Sciuridae: Cynomys leucurus and C. manageaent on rangelands: A shortgrass prairie ludovicianus). Ecology 62:252-272. no~anoctive. Westview Press, Boulder, Colorauo.

176 Uresk, D. w.. and G. L. Schenbeck. 1987. Effect Warwick, S. I., and D. Briggs. 1980. The geneco- of zinc phosphide on prairie dog colony ex- logy of lawn weeds. V. The adaptive signifi- pansion as determined from aerial photography. cance of different growth habit in lawn and Prairie Natur. 19:57-61. roadside populations of Plantago ■ajor L. New Van Bruggen, T. 1985. The vascular plants of Physiologist 85:289-300. South Dakota. Iowa State University Press, Ziuer■an, G. M., H. Goetz, and P. W. Mielke. Ames. 1985. Use of an improved statistical ■ ethod for group co■parisons to study effects of prairie fire. Ecology 66:606-611.

177