Observations of cattle use of prairie dog towns

Transcription

1 J. Rnge Mnge. 56: Septemer 23 Oservtions of cttle use of pririe dog towns DEBRA A. GUENTHER AND JAMES K. DETLING Authors re Reserch Associte, Nturl Resource Ecology Lortory (NREL) nd Professor, Deprtment of Biology nd senior Reserch Scientist, NREL, Colordo Stte University, Fort Collins, Colo At the time of the reserch, the senior uthor ws Reserch Assistnt, Deprtment of Biology, Colordo Stte University, Fort Collins, Colo. Astrct We investigted the use of pririe dog towns y cttle (Bos turus) on the shortgrss steppe of northestern Colordo y conducting surveys of cttle nd vegettion from June to August Cttle presence nd ehvior were recorded 3 times week during driving surveys of 15 lck-tiled pririe dog (Cynomys ludovicinus) towns. A suset of 3 pstures with pririe dog towns ws intensively surveyed twice weekly wherein the hitt nd ctivity of rndomly chosen focl niml ws recorded every 6 minutes for 3.5 hours. Bite nd step counts of other individuls were recorded for 5-minute intervls. Vegettion height nd cover dt were collected monthly on ech of 6 hitts. Results from driving surveys nd intensively surveyed pstures were similr; cttle neither significntly preferred nor voided pririe dog towns. Bre ground cover on pririe dog towns did not significntly differ from most other hitts, ut vegettion on pririe dog towns ws significntly shorter on (men = 6.7 cm) thn tht off (men = 11.9 cm) pririe dog towns. Nevertheless, forging oservtions indicted tht there ws no significnt difference etween cttle forging rtes on swles (7.9 ites/min) nd pririe dog towns (69.5 ites/min). Thus, cttle on the shortgrss steppe pper to use pririe dog towns in proportion to their vilility nd, while there, they grze s intensively s they do on hitts not inhited y pririe dogs. Key Words: cttle, pririe dogs, Cynomys ludovicinus, grzing, shortgrss steppe Blck-tiled pririe dogs (Cynomys ludovicinus) gretly lter vegettion (Coppock et l. 1983, Fhnestock nd Detling 22) nd shre mny preferred forge species with cttle (Bos turus) (Hnsen nd Gold 1977), leding to specultion out competition etween these species for forge. On the shortgrss steppe, pririe dog ctivities increse the numer of oth for nd grss species nd, through selective grzing, exert selective pressure ginst lue grm (Boutelou grcilis [H.B.K.] Lg. ex Griffiths) nd in fvor of ufflogrss (Buchloe dctyloides [Nutt.] Engelm.) (Bonhm nd Lerwick 1976). In mixed-grss pririe, cnopy height decreses s pririe dogs clip vegettion Funding for this study ws provided y the Ntionl Science Foundtion Long Term Ecologicl Reserch (LTER) progrm (NSF Grnt Numer DEB ). We re grteful to Dr. Lrry Rittenhouse nd Dr. Dvid Steingreer for their ssistnce in developing this study. Dr. Tom Hos contriuted his time nd nlyticl dvice. The Shortgrss Steppe LTER nd USDA-ARS Centrl Plins Experimentl Rnge stff provided invlule knowledge nd logisticl support. We lso thnk Bo Peterson nd Mrk Bll of the Pwnee Ntionl Grsslnds for ccess to informtion. Mnuscript ccepted 24 Oct. 2. Resumen Investigmos el use de ls colonis de los perritos de l prder por el gndo (Bos turus) en l esteps de zctes cortos del nordeste de Colordo, medinte l conduccion de muestreos del gndo y l vegetcion de Junio Agosto de L presenci de gndo y su comportmiento se registro 3 veces por semn durnte el periodo de muestreo en el recorrido en vehiculo de 15 colonis de perritos de l prder de col negr (Cynomys ludovicinus). Un sugrupo de 3 potreros con colonis de perritos de l prder se muestreo intensivmente, 2 veces por semn, y cd 6 minutos durnte 3.5 hors se registro el hitt y ls ctividdes de nimles elegidos letorimente. Los conteos de mordids y los psos de otros nimles se registrron en intervlos de 5 minutos. Mensulmente se colectron dtos de coertur y ltur de l vegetcion en cd uno de los 6 hitts. Los resultdos de los muestreos en vehiculo y los de los potreros intensivmente muestredos fueron similres, el gndo ni prefirio ni evito significtivmente ls colonis de los perritos de l prder. El porcentje de suelo desnudo en ls colonis de los perritos de l prder no difirio significtivmente en l myori de los otros hitts, pero l vegetcion en ls colonis de los perritos de l prder fue significtivmente ms cort dentro (medi = 6.7 cm) de l coloni que fuer de ell (medi 11.9 cm). Sin emrgo, ls oservciones de forrjeo indicron que no huo diferenci significtiv entre ls tss de forrjeo del gndo fuer (7.9 mordids/minuto) y dentro de ls colonis de los perritos de l prder (69.5 mordids/min). Asi, el gndo en ls esteps de zctes cortos prece usr ls colonis de los perritos de l prder en proporcion su disponiilidd y ellos pcientn tn intensivmente en ls colonis de perritos de l prder como en los hitts no colonizdos por est especie. (Archer et l. 1987, Dy nd Detling 1994); s tller grss species decline shorter species ecome reltively more undnt (Pinter et l. 1993). As the colony ges, older res ecome dominted y fors nd dwrf shrus, while more recently occupied res remin grss-dominted (Coppock et l. 1983, Whicker nd Detling 1988, Fhnestock nd Detling 22). On oth the shortgrss steppe nd mixed-grss pririe there is diet overlp etween cttle nd pririe dogs. On the Colordo shortgrss steppe, Hnsen nd Gold (1977) reported 64% nnul similrity in diet. In western Oklhom, O'Meili et l. (1982) oserved lower iomss of grss species pltle to cttle on pririe dog towns thn on control pstures, nd speculted tht there ws high degree of competition etween cttle nd pririe dogs. Despite this specultion, some studies hve shown tht pririe dog grzing my ctully positively ffect lrge herivores y 41 JOURNAL OF RANGE MANAGEMENT 56(5) Septemer 23

2 influencing forge qulity. On the mixedgrss pririe of South Dkot, Coppock et l. (1983) found greter crude protein concentrtions, in vitro digestiilities, nd live:ded rtios in plnts from pririe dog towns thn in plnts of the sme species from uncolonized res. Even though pririe dog towns hd 4% less forge stnding crop thn surrounding uncolonized res, ison (Bison ison) strongly selected pririe dog colonies to grze on whenever they were in the vicinity of one (Coppock et l., 1983). In nother study, Krueger (1986) concluded tht rther thn competitive reltionship, ison nd pririe dogs hve mutully eneficil reltionship. Vnderhye (1985) quntified diet qulity differences on nd off pririe dog towns nd used Swift's (1983) ruminnt nutrition model to simulte effects on ison from feeding on pririe dog towns. Her results indicted tht ison tht selectively feed on pririe dog towns in the summer would significntly increse ody weight reltive to those feeding exclusively off pririe dog towns. Cttle hve replced ison s the dominnt herivore on most of the Gret Plins. While some studies hve exmined the effects of pririe dog grzing on forge preferred y cttle or diet overlp etween the 2 herivores (Hnsen nd Gold 1977, O'Meili et l. 1982, Uresk 1983, 1984, 1985), little informtion is ville regrding whether cttle ctully preferentilly grze on pririe dog towns. The ojectives of this study were to: () otin estimtes of cttle use on vs. off of pririe dog towns cross the Shortgrss Steppe Long Term Ecologicl Reserch Site (SGS-LTER) through driving surveys; () to intensively study suset of the pstures from the driving surveys to quntify the mount of time cttle spend on pririe dog towns compred to ech of severl other hitts; (c) compre forging rtes of cttle on pririe dog towns with those on other hitts in the intensive survey pstures; nd (d) mesure vegettion cover y species nd height on nd off pririe dog towns in the intensive survey pstures to determine whether vegettion differences correlted with cttle grzing ctivities. Methods Study Are This reserch ws conducted t the SGS-LTER in northestern Colordo (Lt. ' ' 4 49 N, Long W) nd includes the Pwnee Ntionl Grsslnd (PNG) nd the USDA-ARS Centrl Plins Experimentl Rnge (CPER). The dominnt plnt species re lue grm nd prickly per cctus (Opunti polycnth Hw.), with lue grm mking up 9% of the sl plnt cover. Annul precipittion rnges from 3-55 mm, nd 7% occurs etween My nd Septemer s loclized thunderstorms (Luenroth nd Milchuns, 1991). Most soils re well drined loms or sndy loms, nd the topogrphy consists of rolling hills t n elevtion of 1,31m to 1,935m (Hzlett 1998). Cttle re present on the pstures used for this study from mid-my until mid- Octoer. Psture sizes for this study rnge from 13 to 2,378 h, nd the lck-tiled pririe dog towns rnge in size from L7 to 81.6 h nd occupy.3 to 1.7% of the those pstures (Tle 1). Most pstures contin stock tnks to supply cttle with wter. For this study, we chose pririe dog towns tht were not djcent to or surrounding wter tnks. Pstures re grzed t "moderte" stocking rtes (1.74 h/cow/month, Bo Peterson, pers. comm.). Most of these pririe dog towns re less thn 1 yers old ecuse of plgue-induced die outs nd susequent pririe dog recoloniztions (Mrk Bll, pers. comm.). Driving Surveys We conducted driving surveys to otin SGS-LTER-wide estimtes of cttle use of pririe dog towns pproximtely 3 times per week from 28 My to 15 Aug. 1999, for totl of 31 surveys of ech of 15 pririe dog towns in 12 pstures (Tle 1). Becuse cttle re most ctive the hlf hour efore sunrise until pproximtely 3 hours fter sunrise, nd from 3 hours efore sunset to hlf hour fter sunset (Arnold nd Dudzinski 1978), we restricted our surveys to those times. Time (morning vs. evening), strting loction, Tle 1. Chrcteristics of pstures nd pririe dog towns surveyed during summer 1999, on the Shortgrss Steppe LTER. All llotments on the Pwnee Ntionl Grsslnds hd cow/clf herds. On the Centrl Plins Experimentl Rnge pstures 5W, 22W, nd 28N hd steer herds, psture 29-3 hd yerling reeders, nd psture hd stock heifers. Age of towns were determined from U.S. Forest Service records nd indicte the numer of consecutive yers towns hve een ctive up until the time of this study. psture U.S. Forest re of size Service of town llotment town # y town Pwnee Ntionl Grsslnds (h) Roe Col Keot Simmons Fiscus Box Stonehm Centrl Plins Experimentl Rnge (h) W1 13 5W W NE N 13 28N 3 1Pstures lso used for intensive surveys. JOURNAL OF RANGE MANAGEMENT 56(5) Septemer

3 nd order of visittion of the towns were chnged with ech survey. Pririe dog towns were either entirely visile from the rod with 7 x 35 inoculrs or could e seen y wlking to nery vntge point. Pstures were only chosen for the survey if ll of the pririe dog towns on the psture could e viewed t the time of the survey. These pririe dog towns encompssed vriety of soil types nd uplnd nd lowlnd sites. The numer of cttle on ech pririe dog town nd their ehviors (grzing, resting, trveling, or stnding still) were recorded. Cttle were considered to e grzing if feeding while moving or stnding still. Resting ws defined s cttle lying down, nd trveling ws movement with the hed up. For this generl overview, we were only concerned with whether the cttle were on or off the towns. Pstures were often very lrge, nd not ll of the cttle could e seen t the time of the survey if they were off of the pririe dog town. Consequently, ehvior of cttle off pririe dog towns ws not recorded for the driving surveys. Driving survey dt were nlyzed using resource selection functions (Mnly et l. 1993). The rtio of the proportion of the cttle popultion occupying hitt (or) to the proportion of psture re the hitt occupies (iti) gives selection rtio (w1), where i = hitt (in this cse, on or off pririe dog towns): wi = of/mi (1) These selection rtios re stndrdized (Bi) to vlues etween nd 1 wherein: (2) Stndrdized selection rtios cn e interpreted s proilities, such s the proility of selecting hitt i next if ll the hitts were eqully ville. When only 2 hitts re considered (e.g. on nd off towns), vlues ove.5 indicte preference for hitt, vlues of.5 indicte rndom use, nd vlues elow.5 indicte voidnce of hitt. Chi-squre vlues were clculted for the selection rtios to determine significnt differences from rndom use (.5). Intensive Surveys To otin more detiled ssessment of where cttle were spending their time, suset of 3 smller pstures from the driving surveys, locted on the CPER, were chosen for more intensive surveys conducted from 4 June to 2 Aug (Tles 1 nd 2). In ech of these intensive study res, the entire psture could e oserved from severl vntge points so the entire psture ws considered the home rnge of the cttle nd ll hitt types could e oserved. Plnt communities on these pstures were divided into 6 hitts: pririe dog towns, swles, rocky ridgetops, uplnds, plnted crested whetgrss (Agropyron cristtum [L.] Gertn.) strips, nd Atriplex cnescens (Pursh) Nutt. dominted terrces (Tle 2). Tle 2. Hitts of the 3 intensive survey pstures on the Centrl Plins Experimentl Rnge nd reltive composition of the hitts within ech psture. Psture Hitt 22W 5W (%) uplnd Atriplex terrce pririe dog town swle rocky ridgetop crested whetgrss During 3.5 hour feeding out (morning or evening), rndomly chosen focl niml (Altmnn 1974) ws oserved every 6 minutes. To minimize disturnce, cttle were oserved from vehicle t distnce of t lest 1 km, using 7x35 inoculrs or 15 to 6 x zoom spotting scope. Hitt, instntneous ehvior (grzing, resting, trveling, or other), nd numer of other cttle within 5 m rdius were recorded. Frequency of ehviors ws nlyzed using tests of 2 percentges (Lehner 1996). In etween the focl niml oservtions, we conducted forging surveys to exmine forging efficiency. The numer of ites nd steps in 5-minute period were recorded for rndomly chosen individuls. Ech jw movement when the hed ws touching vegettion constituted ite nd ech movement of front leg ws considered step. Comprison of forging dt etween hitts ws conducted using n ANOVA nd Tukey's HSD procedure. The 3 pstures were mpped in Septemer with hnd-held Glol Positioning System (GPS) unit (Trimle Nvigtion Ltd., Sunnyvle, Clif. 9486) nd res of ech hitt were determined using PC ARC/INFO nd ArcView GIS 3.2 (Environmentl Systems Reserch Institute, Redlnds, Clif ). The hitt use dt for these 3 pstures were divided into 2 periods (Period I = 28 My to 9 July, Period II = 12 July to 15 August) nd lso comined to detect chnges in hitt use over the seson using log-rtio compositionl nlysis (Aeischer et l. 1993). This method exmines the proportionl hitt use y nimls in comprison to hitt vilility. Insted of the numer of loctions, it uses the numer of nimls s the smple size; therefore, it does not require independence of sequentilly collected loctions. This method lso ccounts for the unitsum constrint, common prolem when using compositionl dt, wherein hitt proportions re non-independent nd must sum to 1. For exmple, if n niml voids one hitt, it must spend time in nother, leding to n pprent preference for tht hitt type. Finlly, compositionl nlysis ccounts for the ritrry definition of hitt preference y llowing hitts to e rnked ccording to reltive use. This technique tests for rndom use on ll hitts. When there is devince from rndom use, the ville hitts cn e rnked from most preferred to lest preferred. The dt from this study were divided into 2 periods in n ttempt to detect ny sesonl difference while still hving lrge enough smple size for meningful sttisticl nlysis. Vegettion dt were collected in June, July, nd August on the 3 intensive survey pstures. Twenty,.1 m2 Duenmire frmes (Duenmire 1959) were plced rndomly in ech hitt on ech psture to determine plnt species composition nd cnopy cover. Averge vegettion height ws mesured y setting light crdord piece over the Duenmire frme re nd mesuring the height from the ground to the center of the crdord. Dt from ll 3 pstures were comined for vegettion nlysis due to the close proximity of the pstures (within 8 km) nd the similrity in soils nd vegettion types. Vegettion heights nd cnopy cover were nlyzed using n ANOVA nd Tukey's HSD procedure. Results Driving Surveys Stndrdized resource selection rtios for ech psture cross the seson for ll surveys displyed no cler trend in the use of pririe dog towns y cttle (Fig. 1). A Chi-squre test indicted slight preference (Bi =.527, p =.3) for the use of pririe dog towns y cttle for ll pstures over the entire smpling period. Of 789 cttle oserved on pririe dog towns over the seson, 91.4% were grzing, 5.7% were resting, 2.9% were stnding still, nd none were trveling. 412 JOURNAL OF RANGE MANAGEMENT 56(5) Septemer 23

4 o Psture Fig. 1. Stndrdized resource selection rtios (B;) nd stndrd error rs for cttle occurring on pririe dog towns from the driving surveys for ech psture nd ll pstures comined from 28 My to 15 Aug Vlues ove.5 represent selection for pririe dog towns, vlues elow.5 represent voidnce of pririe dog towns, nd.5 represents rndom use of pririe dog towns. Intensive Surveys towns. The town ws rnked lst in Period Ech of the 3 pstures ws surveyed 16 I (erly summer) nd first in Period II (lte times for totl of 25 surveys for Period I summer), wheres the swle switched nd 23 surveys for Period II. Focl ni- from the first rnk in Period I to second to mls were within 5 m rdius of hlf or lst in Period II. Comined dt over the more of the herd 54% of the time. Cttle seson for psture 29-3 rnked the town did not use hitts rndomly on ny of s second to lst. the 3 intensively surveyed pstures sed Behvior oservtions (n =1,72) of the on log-rtio compositionl nlysis (Tle focl nimls (Tle 4) indicted tht ct- 3). The 3 pstures used for the intensive tie spent the mjority of their time grzing surveys were nlyzed seprtely for hi- during the 3.5 hour oservtion periods. tt use due to the differing proportions of While on pririe dog towns, cttle spent ech hitt type on the pstures (Tle 6% of their time grzing, 13% resting, 2). The pririe dog town ws consistently 15% trveling, nd 12% doing other ctivrnked lst (lest preferred hitt) on ps- ities such s intercting with other cttle or ture 22W, ut ws not used significntly stnding still. Focl nimls for the logdifferently from the rocky ridgetops. On rtio compositionl nlysis spent more psture 5W, cttle used hitts lmost in overll time in the swles (Tle 3). the direct order of vilility (Tle 2), However, those sme nimls spent with the exception of the swle eing higher proportion of their time grzing rnked higher thn the pririe dog town. while on pririe dog towns, uplnds, Agin, there were no significnt sesonl Atriplex terrces, nd rocky ridgetops thn differences, with the town consistently in the swles (Tle 4). rnked in the middle. Psture 29-3 Forging oservtions of numers of reveled the only significnt sesonl dif- ites per step (Fig. 2) reveled few differference regrding cttle use of pririe dog ences etween cttle forging on pririe dog towns (men = 6.3 ites/step) nd most other hitts. Cttle did hve significntly fewer ites per step on the rocky ridgetops nd uplnd hitts (4.7, nd 4.8 ites/step, respectively) nd consequently moved t fster rte cross those hitts. Forging oservtions of the numer of ites per minute indicted tht there ws no significnt difference etween cttle forging on swles (7.9 iteslmin) nd pririe dog towns (69.5 iteslmin). The cttle on these 2 hitts hd significntly higher forging rtes thn on ll other hitts. Vegettion Chrcteriztion We compred cover for 5 plnt species pltle to cttle nd 2 non-pltle species, s well s cover of re ground nd litter cross ll 3 intensively surveyed pstures (Fig. 3). Of the pltle plnts, swles hd the highest cover of western whetgrss (Pscopyrum smithii [Ryderg] Love). Pririe dog towns nd the Atriplex hitt hd the next highest cover of western whetgrss nd they were not significntly different from ech other. Pririe dog towns contined significntly less lue grm cover thn the Atriplex terrces nd uplnd hitts, ut more thn swles nd crested whetgrss hitts. The pririe dog towns nd swles hd significntly more cover of ufflogrss thn the other hitts. Cover of needlelef sedge (Crex eleochris Biley) nd scrlet gloemllow (Spherlce coccine [Pursh] Ryderg vr. coccine) ws low (<5%) cross ll hitts. Of the non-pltle plnts (Fig. 3), red three-wn (Aristid purpure Nuttll vr. longiset [Steud.] Vsey) hd low cover (1.2%) on the pririe dog towns, nd this ws lower thn on either the rocky ridgetops or uplnd hitts. Sixweeks fescue (Vulpi octoflor [Wlt.] Ryd.) hd low (-2%) cover cross hitts nd few significnt differences etween hitts. Bre ground cover on pririe dog towns ws not significntly different from Tle 3. Results of the log-rtio compositionl nlysis for the 3 intensive survey pstures including hitt rnkings ccording to use. The designtion "cw" refers to the crested whetgrss hitt, "rocky" refers to rocky ridgetops, nd "town" refers to pririe dog town. Period I refers to surveys conducted from 3 June to 8 July 1999, nd Period II refers to surveys conducted from 13 July to 21 Aug Within row, nd within psture, vlues with the sme letter were not significntly different t P <.5. psture 22W psture 5W psture 29-3 Period I swle > uplnd > rocky > town" swle > Atriplex" > townc > cwc > rocky uplnd > Atriplex > rocky'" > town Period II swle > uplnd > rocky > town" Atriplex' > swle" > town > cw'> rocky Atriplex > uplnd > swle > rocky Comined swle > uplnd > rocky > townc Atriplex > swle > town > cw > rocky uplnd > swle" > town > rocky JOURNAL OF RANGE MANAGEMENT 56(5) Septemer

5 Tle 4. The proportion of ehviors of focl nimls verged over ll 3 intensive survey pstures on the CPER, summer, Within column, vlues with the sme letter were not significntly different t P <.5. The designtion "other" refers to ehvior such s stnding still, drinking wter, or intercting with other cttle The designtion "cw" refers to the crested whetgrss hitt, "rocky" refers to rocky ridgetops, nd "town" refers to pririe dog town. Hitt grzing resting n rocky.82 O.OO Atriplex.75.4' uplnd l town.6.13 cw.56c.3 swle.42.2 tht in other hitts with the exception of the crested whetgrss hitt nd rocky ridgetops which hd higher percentges of re ground (Fig. 3). Averged over ll 3 pstures nd over July nd August (Fig. 4), vegettion on pririe dog towns (men = 6.4 cm) ws significntly shorter thn tht in ll other hitts (men =11.9 cm). Discussion Although Chi-squre nlysis of the 1999 driving survey dt indicted tht cttle significntly selected for pririe dog towns on this Colordo shortgrss steppe site, the mgnitude of the effect ws quite smll (Bi =.527) nd differed little from rndom use (Bi =.5). Clerly, cttle did not exhiit strong preference for pririe dog towns s did ison on the mixed-grss pririe (Coppock et l. 1983, Krueger 1986). However, 91% of the cttle oserved on pririe dog towns during driving surveys were grzing. This my seem intuitive since we surveyed during pek grzing times; however this shows tht cttle did not spend their time simply trveling cross pririe dog towns to get to more preferred grzing res. Insted, they ctully spent significnt mount of time on the pririe dog towns ctively using them s forging resource. On the more intensively surveyed pstures, focl nimls pprently represented herd movements well s more thn hlf the herd ws within 5m of them. Preference rnkings derived from log-rtio compositionl nlysis indicted tht over the seson, swles nd Atriplex terrce hitts were usully most preferred nd rocky ridgetops were often lest preferred. In generl, pririe dog towns were rnked ner the middle or elow, ut significnce tests did not indicte differences etween the pririe dog towns nd nywhere from 1 to 4 other hitts. Swles were highly rnked in ll 3 pstures indicting tht ct- tle were strongly selecting for 1 hitt low in proportionl re (Tle 2). This is consistent with Senft's (1983) findings tht swles were preferred during the growing seson. While cttle do not pper to strongly prefer the pririe dog towns, they lso do not void them. The compositionl nlysis of the intensive survey dt ws used ecuse of our ility to oserve nd smple the niml's entire home rnge, our ility to record n equl numer of oservtions for ech niml, nd ecuse ll hitts were used in ech psture. Aeischer et l. (1993) sug cw swle Atriplex gest tht cution is required for gregrious nimls. For this study, the focl nimls pper to represent the movements of the herd nd therefore we cutiously extrpolte our findings to the rest of the herd. Bison on mixed-grss pririe were found to hve significntly higher ite/step rtios on pririe dog towns compred to uncolonized res in 5 of 6 months of the My to Novemer study, excluding Novemer (Krueger 1986). This implied greter forge qulity on those pririe dog towns compred to the uncolonized re. In our study, the numer of ites/step on pririe dog towns ws not significntly different thn productive hitts such s swles nd Atriplex terrces. The forging rte (ites/minute) for cttle on pririe dog towns ws lso not significntly different from swles, preferred hitt. While ite size ws not mesured (nd could hve een smller on pririe dog towns), this nevertheless implies tht the plnts on pririe dog towns hve forge vlue. A study of cttle on the shortgrss steppe found tht forging velocities (rtes of wlking in steps/minute) only chnged if the differences in forge quntity were town uplnd rocky swle town Atriplex uplnd cw rocky Fig. 2. Lest squres mens nd stndrd errors of numer of ites per step nd ites per minute of forging cttle y hitt ("cw" refers to the crested whetgrss hitt, "rocky" refers to rocky ridgetops, nd "town" refers to pririe dog towns). Smple sizes: swle n = 118, pririe dog town n = 91, Atriplex terrce n = 211, cw n = 2, rocky ridgetop n = 24, uplnd n = JOURNAL OF RANGE MANAGEMENT 56(5) Septemer 23

6 1-8- L >6 v \4 2 4 L3 2 X L4 >3- "2 1 Pscopyrum smithii L. C swi twn Atr upi rky cw Boutelou grcilis Atr upi twn rky swi cw Buchloe dctyloides C swi twn rky upi Atr cw c C C d Sphrelce coccine e cw Atr twn upi rky swi _h Crex eleochris hi e upi swi Atr twn rky cw c e d d 8 6 ) `1- U 8 6 L X1 Aristid longiset li upi rky twn swi Atr cw Vulpi octoflor upi twn swi Atr rky cw C c c c cw rky twn swi upi Atr Litter BT e Atr upi twn swi rky cw Fig. 3. Averge cover (%) for 5 plnt species pltle to cttle (left column), 2 non-pltle species (Aristid longiset nd Vulpi octoflor), nd for re ground nd litter on ech hitt type. Cover is verged over June, July, nd August of 1999, nd cross the 3 intensively surveyed pstures on the CPER ("cw" refers to the crested whetgrss hitt, "rky" refers to rocky ridgetops, "Atr" refers to Atriplex terrces, "upl" refers to uplnd hitt, "swl" refers to swles, nd "twn" refers to pririe dog towns). Columns with the sme letters were not significntly different t P <.5. gret (Biley 1988, Biley et l. 1996). It my e tht the differences in forge quntity or qulity etween pririe dog towns nd swles on the shortgrss steppe were not sufficiently lrge for cttle to chnge their forging rtes. Our vegettion chrcteriztion indictes tht plnt species pltle to cttle Bre ground such s western whetgrss, lue grm, nd scrlet gloemllow (Vvr et l. 1977), re quite common on these shortgrss steppe pririe dog towns (Fig. 3). Non-pltle plnts often hd similr or higher cover off of pririe dog towns. The estimte for re ground cover on pririe dog towns (35.3%) in this study is essen- tilly identicl to wht Archer et. l (1987) found on the mixed grss pririe of South Dkot (35%). However, Archer et. l (1987) found more thn 3 times s much re ground on pririe dog towns s off, wheres on the shortgrss steppe, we found there to e s much or more re ground on the other hitts s on pririe dog towns. One importnt difference found in vegettion on nd off pririe dog towns ws tht the plnts growing on pririe dog towns were significntly shorter thn those on ll of the other hitts. While cttle did not pper to e selecting pririe dog towns to grze on, they lso did not significntly void them, even though the vegettion ws shorter there. It is not cler why we did not see preference for pririe dog towns y cttle on the shortgrss steppe. Age of the pririe dog towns my e fctor. The towns used for this study were ll reltively young (men = 5 yers); however, res of pririe dog towns selected for grzing y ison in mixed-grss pririe were of this ge (Coppock et l. 1983). Slope nd the loction of wter tnks were not considered in the nlysis of this study; however, the pstures were reltively flt with rolling hills (-15% slope) nd cre ws tken to select pririe dog towns tht did not hve wter tnks on or djcent to them. Vriility cn lso occur in the ehvior of cows with clves versus steers. However, comintion of pstures with different cttle types ws used for this study to otin n overview of cttle ehvior in generl. Yer to yer wether differences might hve n effect on cttle selection of pririe dog towns on the shortgrss steppe. The summer of this study ws unusully wet. The totl precipittion for 1999 ws 55mm, t the upper end of the rnge of precipittion for the shortgrss steppe. Green (1998) reported tht during wet yer on the mixed-grss pririe, ison grzed on pririe dog towns preferentilly. He ttriuted this to the dditionl moisture cusing grsses on surrounding ungrzed hitt to grow tller nd seed out, wheres the grzed grsses on pririe dog towns were not seeding out nd hd high lef:stem rtios. During dry yer ison did not grze preferentilly on the towns. On the shortgrss steppe, Lerwick (1974) found tht pririe dogs nd cttle switched diets during drought yers with pririe dogs consuming more grss nd cttle consuming more fors thn in nondrought periods. Vrition in wether definitely hs n effect on plnt-herivore nd herivore-herivore reltionships, nd sur- JOURNAL OF RANGE MANAGEMENT 56(5) Septemer

7 cw Atriplex uplnd rocky town Fig. 4. Men verge height (cm) nd stndrd errors of vegettion for ll hitts verged over July nd August, 1999 ("cw" represents the crested whetgrss hitt, "rocky" refers to rocky ridgetops, nd "town" refers to pririe dog towns). veys conducted on the shortgrss steppe during drought yer might show very different results. However, since 1999 ws wet yer, which would e similr to norml wether yer on the mixed-grss pririe, we would expect to see cttle preference for pririe dog towns during the summer of our study if we were to see it t ll. Nitrogen content of plnts on nd off pririe dog towns ws not mesured for this study. However, if there ws indeed n increse in protein content in the plnts on pririe dog towns s shown in the mixed-grss pririe (Coppock et l. 1983), perhps this increse on the shortgrss steppe ( plce with less moisture) is not gret enough to cuse preference ut high enough to compenste for the loss in iomss. Therefore the hitt is used s expected nd not voided. Coppock et l. (1983) hypothesized tht response y lrge ungultes to pririe dog towns would likely only e seen in highly productive systems where the difference in hitt on nd off towns is gretest. The shortgrss steppe is very different system from the mixed grss pririe (Milchuns et l. 1998). Atsedu (1995) found differences in nitrogen content increses of grzed plnts on the shortgrss steppe to e dependent on grzing histories nd intensities. Also, cttle nd ison re similr in tht oth re generlist herivores, ut cttle re more selective feeders (Peden et l. 1974). The exmintion of cttle use of pririe dog towns on mixed-grss pririe nd of ison use of pririe dog towns on shortgrss pririe would provide insights s to whether it is cttle or the shortgrss system tht is driving the preferences for pririe dog towns. Furthermore, the issue of crrying cpcity is importnt (Hos nd Hnley 199). For exmple, how lrge or old does pririe dog town hve to e swle efore it egins to ffect the numer of cttle tht cn e sustined on psture in the shortgrss steppe? More informtion is lso needed on the effects of pririe dog grzing on the quntity nd qulity of forge for cttle on the shortgrss steppe. Literture Cited Aeischer, N. J., P. A. Roertson, nd R. E. Kenwrd Compositionl nlysis of hitt use from niml rdio-trcking dt. Ecol. 74: Altmnn, J Oservtionl study of ehvior smpling methods. Behv. 49: Archer, S., M. G. Grrett, nd J. K. Detling Rtes of vegettion chnge ssocited with pririe dog (Cynomys ludovicinus) grzing in North Americn mixed-grss pririe. Vegettio 72: Arnold, G. W. nd M. L. Dudzinski Ethology of free-rnging domestic nimls.elsevier Scientific Pulishing Co., Amsterdm, The Netherlnds. Atsedu, M Defolition responses of shortgrss steppe plnts in reltion to longnd short-term grzing history. Ph.D. Diss. Colordo Stte Univ., Fort Collins, Colo. Biley, D. W Chrcteristics of sptil memory nd forging ehvior in cttle. Ph.D. Diss., Colordo Stte Univ., Ft. Collins, Colo. Biley, D. W., J. E. Gross, E. A. Lc, L. R. Rittenhouse, M. B. Coughenour, D. M. Swift, nd P.L. Sims Mechnisms tht result in lrge herivore grzing distriution ptterns. J. Rnge Mnge. 49: Bonhm, C. D. nd A. Lerwick Vegettion chnges induced y pririe dogs on the shortgrss pririe. J. Rnge Mnge. 29: Coppock, D. L., J. K. Detling, J. E. Ellis, nd M. I. Dyer Plnt-herivore interctions in North Americn mixed-grss pririe I. Effects of lck-tiled pririe dogs on intrsesonl oveground plnt iomss nd nutrient dynmics nd plnt species diversity. Oecologi 56:1-9. Coppock, D. L., J. E. Ellis, J. K. Detling, nd M. I. Dyer Plnt-herivore interctions in North Americn mixed-grss pririe II. Responses of ison to modifiction of vegettion y pririe dogs. Oecologi 56:1-15. Duenmire, R A cnopy-coverge method of vegettionl nlysis. Northwest Sci. 33: Dy, T. A. nd J. K. Detling Wter reltions of Agropyron smithii nd Boutelou grcilis nd community evpotrnspirtion following long-term grzing y pririe dogs. Amer. Midl. Nt. 132: Fhnestock, J. T. nd J. K. Detling. 22. Bison-pririe dog-plnt interctions in North Americn mixed-grss pririe. Oecologi 132: Green, R. A Nitrogen distriution in perennil grsslnd: the role of Americn ison. Ph.D. Diss., Colordo Stte Univ., Ft. Collins, Colo. Hnsen, R. M. nd I. K. Gold Blcktil pririe dogs, desert cottontils, nd cttle trophic reltions on shortgrss rnge. J. Rnge Mnge. 3: Hzlett, D. L Vsculr plnt species of the Pwnee Ntionl Grsslnd. Generl Tech. Rep. RMRS-GTR-17. Fort Collins, Co: U.S. Dept. of Agr, Forest Serv., Rocky Mountin Res. St. 26 pp. Hos, N. T. nd T. A. Hnley.199. Hitt evlution: do use/vilility dt reflect crrying cpcity? J. Wildl. Mnge. 54: Krueger, K Feeding reltionships mong ison, pronghorn, nd pririe dogs: n experimentl nlysis. Ecol. 63: Luenroth, W. K. nd D. G. Milchuns Short-grss steppe, p In R. T. Couplnd (ed.) Ecosystems of the world. Elsevier, New York, N.Y. Lehner, P. N Hndook of ethologicl methods. Second edition. Cmridge Univ. Press. Cmridge, Gret Britin. Lerwick, A Effects of the lck-tiled pririe dog on vegettion composition nd their diet in reltion to cttle. M.S. Thesis, Colordo Stte Univ., Ft. Collins, Colo. Mnly, B., L. McDonld, nd D. Thoms Resource selection y nimls. Chpmn nd Hll. London, Gret Britin. Milchuns, D. G., W. K. Luenroth, nd I. C. Burke Livestock grzing: niml nd plnt iodiversity of shortgrss steppe nd the reltionship to ecosystem function. Oikos 83: O'Meili, M. E., F. L. Knopf, nd J. C. Lewis Some consequences of competition etween pririe dogs nd eef cttle. J. Rnge Mnge. 35: Pinter, E. L., J. K. Detling, nd D. A. Steingreer Plnt morphology nd grzing history: reltionships etween ntive grsses nd herivores. Vegettio 16: Peden, D. G., G. M. Vn Dyne, R. W. Rice, nd R. M. Hnsen The trophic ecology of Bison ison L. on shortgrss plins. J. Appl. Ecol. 11: JOURNAL OF RANGE MANAGEMENT 56(5) Septemer 23

8 Senft, R. L The redistriution of nitrogen y cttle. Ph.D. Diss., Colordo Stte Univ., Ft. Collins, Colo. Swift, D. M A simultion model of energy nd nitrogen lnce for free-rnging ruminnts. J. Wildl. Mnge. 47: Uresk, D. W Pririe dogs s ecosystem regultors on the northern high plins, p In: Seventh North Americn pririe conference, Proc. Aug 4-6, 198.Southwest Missouri Stte Univ., Springfield, Mo. Uresk, D. W Blck-tiled pririe dog food hits nd forge reltionships in western South Dkot. J. Rnge Mnge. 37: Uresk, D.W Effects of controlling lck-tiled pririe dogs on plnt production. J. Rnge Mnge. 38: Vnderhye, A. V. R Interspecific nutritionl fcilittion: Do ison enefit from feeding on pririe dog towns? M.S. Thesis, Colordo Stte Univ., Fort Collins, Colo. Vvr, M., R. W. Rice, R. M. Hnsen, nd P. L. Sims Food hits of cttle on shortgrss rnge in northestern Colordo. J. Rnge Mnge. 3: Whicker, A. D. nd J. K. Detling Modifiction of vegettion structure nd ecosystem processes y North Americn grsslnd mmmls, p In: Werger, M. J. A., P. J. M. vn der Art, H. J. During nd J. T. A. Verhoeven (eds.), Plnt form nd vegettion structure. SPB Acdemic Pulishing, The Hgue, The Netherlnds. JOURNAL OF RANGE MANAGEMENT 56(5) Septemer