short relict histories have been used to determine trends of vege- Vegetational Composition and Ground Cover of tational changes. In a vegetation Two Natural Relict Areas and Their Associated survey of the Fort Worth Prairie Dyksterhuis (1946) selected re- Grazed Areas in the Red Desert of Wyoming1 lict areas which had not been grazed for at least five years or

ROBERT MARQUISS AND ROBERT LANG longer, giving emphasis to one ( area which had not been grazed Department of Agronomy, College of Agriculture, Uni- by domestic livestock since versity of Wyoming, Laramie, Wyoming barbed wire became generally available. He did not consider any land that had been culti- The Red Desert of Wyoming field and Buetner (1938)) and vated or subjected to other types is a valuable resource because it others have used the method as a of severe disturbance which furnishes winter grazing for a key to both the past and future leave a relative permanent im- high percentage of the state’s in recording functions and press on soil or vegetation. sheep. Consequently, proper changes of plant communities. The earliest records of the grazing management of this and Clements (1934) described a vegetational composition of the other winter range areas is a relict in an ecological sense, as Red Desert area where this study problem continually confronting “a community or fragment of was conducted were made by the range manager. one that has survived some im- Aven Nelson in 1897 (Nelson, The early stages of range de- portant change, often to become 1898). The main objective of his terioration are difficult to detect in appearance an integral part survey was not merely to list without a guide to the vegeta- of the existing vegetation.” He plant species, but also to evalu- tional composition and produc- further describes a relict as pos- ate the forage value of the tion potential to use as a basis sessing evidence of “two indica- region. for comparison. Natural barriers tor values of the first impor- to grazing resulting in relict tance. These relate to the causes The findings of Nelson were areas are few. When they can concerned on the one hand and confirmed by Vass and Lang be found, they are invaluable for to the original vegetation and its (1938). In counts of living and the determination of composi- changes on the other. The dead plants and reproduction of tion and production potential of former may be climatic, edaphic, four shrubby species, they found similar range sites. or human, each exhibiting a that with few exceptions there Two natural relict areas occur- more or less distinctive impress.” was sufficient reproduction to ring in Wyoming’s Red Desert The relict method is a key to replace the plants which had were utilized in this study in an both the past and the future and died in 1936. A comparison of attempt to determine changes in places proper emphasis upon the the grazing capacity of the ma- vegetation which may be attrib- measurement and record of com- jor vegetation type between a uted to grazing on the shallow- munity functions and changes by controlled grazing area, as repre- soil range sites common to the means of quadrats and exclo- sented by the Rock Springs area. sures. Grazing Association lease, with In areas of high precipitation a non-controlled grazing area, as Previous Studies and mature soils, where vegeta- represented by the land sur- The range condition class tional composition changes are rounding the lease, showed no method of study is commonly relatively rapid, exclosures with significant difference. used in range analysis. It is based upon comparison of relict and grazed areas on the basis of vegetational composition, Beetle (1952)) Gardner (1950)) Whit- --BLACK ROCK BUTTE 14.07 Acres 17.98 Acres 1 Published with the approval of the Director, Wyoming Agricultural Experiment Station, as Journal Paper No. 108. Portions of qhe ma- terial in this paper were used by the senior author in partial fuZfiZZ- FIGURE 1. Scale diagram and area of the relatively flat summits of the two natural ment of the requirements for the relict areas studied in Sweetwater County, Wyoming. The two areas are separated by Master of Science degree. a physical distance of about 50 miles.

104 VEGETATION OF RELICT AND GRAZED AREAS 105

Table 1. Monthly and annual precipitation in inches and annual departure from average for fhe five years of study” Annual Departure Year Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Precipi- from 20- tation year mean 1937 -67 1.26 .72 1.01 .73 1.94 3.00 .62 .38 .71 .93 .70 12.67 +S.Ol 1938 .16 .38 1.24 1.68 1.64 .56 1.42 1.24 1.00 2.23 .54 .09 12.18 +5.52 1939 .55 .82 .17 .98 .65 .33 .45 .03 .84 .74 .oo .07 5.63 -1.03 1940 1.12 .65 .86 .64 .78 .08 .24 .18 1.57 .74 .90 .50 8.26 +1.60 1956 .38 .14 .05 .94 2.20 .Ol .27 .56 .35 1.18 .70 .36 7.14 + .48

20-year mean (1937-1956) 6.66 * Average precipitation data from the Farson and Rock Springs weather stations.

The Study Area desert range area were: mutton curs during the summer months, although both 1937 and 1939 Two relatively flat topped blue grass (Pou #endZeriam), were years with abundant sum- buttes2 located about 50 miles bottlebrush squirreltail @tan- mer moisture. (Table 1). apart in the northwestern quar- ion hystrix) Indian ricegrass ter of Sweetwater County, Wyo- (Oryzopsis hymenoides), blue- Meihods of Study bunch wheatgrass (Agropyron ming, provided the natural relict The vegetational composition spicutum), needle - and - thread, areas for this study. These and ground cover of the relicts and thickspike wheatgrass (Ag- buttes, locally known as Black and surrounding grazed land ropyron dasystachyum). Rock Butte and Pilot Butte, have were studied in June each year high perpendicular sides, and Although a great variety of by the point-observation plot their tops could not possibly forbs were listed from the study method as described by Stewart have been grazed within historic plots, those most frequently en- and Hutchings (1936). These time by any herbivorous animals countered were : Hoods phlox studies were conducted in 1937, other than rodents. The grazed (Phlox hoodii), cushion eriogo- 1938, 1939, 1940 and 1956. Ten areas studied in conjunction with num (Eriogonum ovulifolium), circular plots, each 100 square- the butte tops were located near Hooker sandwort (Arenaria feet in area, were studied on the level base of each butte, but hookeri), slenderbush eriogo- each location each year. On the still within the shallow-soil range num (Eriogonum microthecum), summits of each butte, plots site common to the relict areas. stemless goldenweed (Haplopup- were spaced at random but not The vegetation of both the pus acaulis), fringed sagebrush at the very edge of the butte. grazed and relict areas studied (Artemisia frigida), arrowleaf The randomly located plots on consisted mainly of desert shrubs balsamroot (Bulsumorrhiza sag- the grazed areas were near the as dominants and perennial ittata), stickseed (LappuZu spp.), level base of the butte but con- grasses as subdominants. Vari- bluebells (Mertensiu oblongifo- fined to the shallow soil range ous perennial and annual forbs Zia), and Fremont penstemon site similar to that of the summit. were evident from year to year (Penstemon fremontii). In addition to the listing of depending on climatic conditions. The elevation of Pilot Butte species and estimation of ground The dominant species on both is 7,923 feet, and it rises some cover, an actual count of the relict and grazed areas was big 450 feet above the surrounding number of live plants ,other than sagebrush (Artemisia triden- area. Black Rock Butte rises 658 seedlings), dead plants, and seed- tutu). Other common browse feet above the sourrounding area lings of four shrub species was species were: winterf at (Euroti to an elevation of 7,558 feet. The made on each study plot each Zunata), spiny hopsage (Grayia size and the outlines of the rela- year. The species thus counted spinosa), shadscale (A trip 1ex tively level tops of both buttes were big sagebrush, shadscale, confertifoliu), low rabbitbrush are shown in Figure 1. Gardner saltbush, and winterf at. (Chrysothamnus viscidiflorus A cool climate and low precipi- Samples of the top 6 inches of pumilus) bud sagebrush (Artem- tation are characteristic of the soil were taken from the sum- isia spinescens), and Gardner Red Desert region where this mits and from near the level saltbush (Atriplex gardneri). study was conducted. Average base of both buttes in the gen- Grasses most common to this annual precipitation is less than eral area where the vegetation 10 inches and much of the ef- study plots were located. These fective moisture comes during soil samples were analyzed to 2 Black Rock Butte is located in Sec. March, April, and May in the determine if differences in tex- 24, T. 22 N., R. 101 W., and Pilot Butte is located in Sec. 9, T. 19 N., form of snow and light showers. ture, pH, soluble salts, or air- R. 106 W. Generally little precipitation oc- water permeability existed be- ROBERT MARQUISS AND ROBERT LANG

brush, contributed 34 percent of the total vegetational composi- tion on the relict plots and 54 percent on the grazed plots. These figures would indicate that big sagebrush is an increaser on these shallow soil range sites. Greatest variation in ground cover of big sagebrush within the five-year period was from 3.87 percent in 1938 to 2.64 per- cent in 1940. When these data pertaining to ground cover of big sagebrush were analyzed sta- tistically, differences between buttes, locations, years and all of the interactions between these Relict Grazed Relict LlGrazed Pilot Butte Black Rock Butte factors were highly significant.

FIGURE 2. Average square-foot ground cover per plot for the Pilot Butte and Black The ground cover of winterfat Rock Butte relict and grazed areas. was found to be considerably lower on the grazed areas than tween the soil of the relict areas creased in total ground cover in on the relict areas. Black Rock and that of the associated grazed 1938 over 1937, decreased in the Butte had the greatest ground areas. following 2 years to a low in cover of this species, with an av- Results 1940, but again increased slightly erage of 0.60 square-foot per plot, by 1956. These total ground cover Total Ground Cover while the grazed area surround- fluctuations occurring on the ing the butte averaged only 0.04 Total ground cover was greater grazed areas were similar to square-foot per plot. Winterf at on each grazed area than on its those of the precipitation aver- contributed 12 percent of the to- associated relict. The grazed area ages of March through June of surrounding Pilot Butte had the the same 5 years. General views tal ground cover on the relict greatest total ground cover with of the vegetation on the two compared to 0.57 percent on the a five-year average of 10.54 per- relict areas in June 1956 are grazed land; under these condi- cent, compared with 6.02 percent shown in Figures 4 and 5. tions of shallow-soil range site, on the summit of the butte. The Separate analyses of the it would be considered a de- Black Rock Butte grazed area av- ground cover were made of the creaser. Pilot Butte had an av- eraged 6.97 percent total ground four most widely distributed erage ground cover of 0.18 cover compared with the top of species of major importance square-foot of winterfat per plot, the butte which averaged 4.84 within the grazed and relict while the grazed area surround- percent. The average total areas. The species with the ing the butte averaged 0.12 ground cover per plot on each lo- greatest ground cover, big sage- square-foot per plot. It contrib- cation is shown in Figure 2, and . the average percentage compo- F sition by vegetational groups for Shrubs and semi-shrubs these same areas is shown in El Figure 3. q Grass and grass-like The average total ground cover q Forb species per plot on the top of Pilot Butte decreased for each of the five years studied while on Black Rock Butte it decreased from 1937 to 1940 and increased again by 1956 to about the five-year u average. The two grazed areas 5 reacted somewhat differently z 0 each year from the two relict Relict Grazed Relict Grazed areas, but the reactions of the Pilot Butte Black Rock Butte grazed areas were similar to each FIGURE3. Average percentage total ground cover in three vegetational groups for the other. Each grazed area in- Pilot Butte and Black Rock relict and grazed areas. VEGETATION OF RELICT AND GRAZED AREAS 107 uted 3 percent of the total ground cover on Pilot Butte and only 1 percent on the grazed land. These differences in ground cover of winterfat between buttes, between locations, and among years were highly signifi- cant. The greatest variation in the ground cover of Indian ricegrass was between the Pilot Butte re- gion and the Black Rock Butte region. Little difference existed in ground cover of this species between each relict area and its associated grazed area, and vari- ations among years was small. Differences between the buttes were significant at the 0.01 level. FIGURE 4. Typical vegetation on the Pilot Butte relict area. Photo taken in June 1956. Differences between relict and grazed areas (locations), among Frequency of Specific Occurrence on their associated relicts. These years, and interactions of these There were only 9 species of species were generally of less factors were non-significant. the total of 50 which occurred on -than 10 percent frequency. Needle-and-thread was com- both locations of Pilot Butte and All species present on the two mon on the two grazed areas and Black Rock Butte. Big sagebrush buttes and the two grazed areas contributed about the same occurred with 60 percent fre- which had an average frequency ground cover as did Indian rice- quency on the summit of Black of 10 percent or more are listed grass. It was less common on the Rock Butte and 96 percent on the in Table 2. two associated relict areas. Black grazed land of this area. The re- Plant Counts of Four Species Rock Butte had an average of maining 8 species found on both 0.06 square-foot ground cover per locations of the two butte regions Counts of living plants (other plot, while its surrounding area consisted of one semi-shrub, five than seedlings), dead plants, and had 0.20 percent average ground grasses, and two forbs. seedlings were made of big sage- cover of this species. Pilot Butte Forbs varied the most between brush, winterfat, shadscale, and had only an average of 0.01 the two locations (relict and Gardner saltbush. Only two square-foot ground cover per grazed). Black Rock Butte area species, big sagebrush and win- plot and its associated grazed had 6 species of forbs occurring terfat were frequent on both area had 0.09 percent. Differ- on the relict plots and 14 were Black Rock Butte, Pilot Butte, ences among years, between lo- recorded on the grazed area and their adjacent grazed areas. cations, and between buttes were around this butte. Pilot Butte The average number of living highly significant. From these had 19 species of forbs and its plants of big sagebrush per plot data it appears that needle-and- surrounding grazed area had 31. on Black Rock Butte was 2.82. thread would be classified as an Wild onion (Allium textile) This compared with 6.80 living increaser under grazing. was present only on Pilot Butte, plants per plot on the surround- Three species, shadscale, Gard- and scarlet globemallow (Sphae- ing grazed land. Pilot Butte av- ner saltbush, and bud sagebrush, ralcea coccinea) was present eraged 12.92 plants per plot and were present on Black Rock only on Black Rock Butte. These its surrounding grazed area aver- Butte and the surrounding were the only two species found aged 17.72. The number of living grazed area, but were not pres- to occur on the relict and not on plants increased from an average ent in the Pilot Butte region. the grazed plots. of 5.40 per plot in 1937 to 15.70 Actually shadscale was the dom- One shrub, gray horsebrush per plot in 1956 on the two relict inant species on the top of Black (Tetrudymiu cunescens), three areas. The number of living Rock Butte and big sagebrush grasses, thickspike wheatgrass, plants on the grazed areas in- was the subdominant. Average slender wheatgrass (A. truchy- creased from an average of 11.75 ground cover of shadscale was cuulum), and subalpine needle- per plot in 1937 to 19.55 per plot 1.41 percent on the Black Rock grass (Stipu columbiunu), one in 1956. These differences were Butte relict and only 0.14 percent sedge (Curex eleochuris), and 13 highly significant. on the grazed area surrounding species of forbs occurred on the The number of dead plants of the butte. two grazed areas studied but not big sagebrush varied the most 108 ROBERT MARQUISS AND ROBERT LANG

Table 2. Percentage frequency of plant species occuring on fwo relict and plants averaged only 1.50 per two grazed areas in Sweefwafer County, Wyoming. (Species with average plot on the four locations. There frequency of 10 percent or more.) were fewer dead plants on the

Butte Avg. grazed than on the relict areas. Pilot Butte Species Black Rock %f All *ifi’ Relict Grazed Relict Grazed Relict Grazed Areas Variations among years, how- ever, were greater than among Artemisia tridentata ______.100 100 60 96 80 98 89.0 locations, with an average from Poa jendleriana ______.______._ 88 92 72 80 80 86 83.0 Sitanion hystrix ______..._ _.______. 72 80 78 76 75 78 76.5 0.32 dead plants per plot in 1939 Oryzopsis hymenoides ______..52 44 92 88 72 66 69.0 to no dead plants recorded in Phlox hoodii ______.____. __.______88 90 20 62 54 76 65.0 1956. Differences among years Agropyron spicatum ______96 86 42 6 69 46 57.5 were significant at the 0.01 level. Eurotia lanata .______._.______. 80 40 80 8 80 24 52.0 Other variations were not sig- Eriogonum ovalijolium ______80 90 ___. 26 40 58 49.0 nificantly different. Arenaria hookeri ______100 30 48 8 74 19 46.5 Seedling counts of winterfat Stipa comata ______.______8 48 34 48 21 48 34.5 showed no signif icant difference Grayia spinosa ______56 ______2 62 29 31 30.0 between buttes, between loca- Atriplex conjertijolia _.______._ 92 26 46 13 29.5 Chrysothamnus viscidijlorus tions, or among years. pumilus _.______6 74 ____ 30 3 52 27.5 The average number of living Artemisia spinescens.. ______86 14 43 7 25.0 plants of shadscale on Black Eriogonum microthecum ______2 88 ____ 2 1 45 23.0 Rock Butte was 7.18 per plot. Its Haplopappus acaulis ______12 74 2 ____ 7 37 22.0 associated grazed area had an av- Agropyron dasystachym ______.___ 42 ____ 42 ____ 42 21.0 erage of 0.76 living plants per Artemisia jrigida __.______66 16 ______33 8 20.5 plot. The greatest number of Balsamorrhiza sagittata -______2 70 ______1 35 18.0 seedlings on the relict areas was Lappula spp. ______.______10 38 ____ 24 5 31 18.0 recorded in 1938 with an average Mertensia obtongijolia .______2 42 ______1 21 11.0 of 1.50 per plot. Average number Atriplex gardneri ______.______32 8 16 4 10.0 of seedlings over the 5 years Penstemon jremontii ___.______2 34 ____ 4 1 19 10.0 studied was only 0.04 per plot. This species did not occur in the among years. The number of data were analyzed statistically Pilot Butte region. dead plants on the relict areas there was no significant differ- Soil Analyses averaged 0.15 per plot in 1940 ence between locations. How- and 2.0 per plot in 1956. Grazed ever, differences between buttes, Soil samples were taken from areas had a higher percentage of years, and the interaction be- grazed lands of shallow-soil dead plants than their associated tween buttes and locations were range sites adjoining the relict . relict areas. significantly different at the 0.01 areas for comparison with soil Variations in the number of level. samples removed from the re- big sagebrush seedlings were Counts of winterfat showed licts. These were analyzed for high among years, between the greatest variation between possible factors affecting differ- buttes, and between locations. locations. The number of living ences in vegetational composi- Pilot Butte had the greatest av- plants was less on each grazed tion and ground cover between erage number with 24.26 seed- area than on the non-grazed re- the two locations. lings per plot. Its associated licts. Differences in numbers of A mechanical analysis for tex- grazed area averaged 15.62 seed- plants varied considerably less ture showed that the soil from lings per plot. High variations, than the variations in ground the grazed areas near Pilot Butte but in the reverse order, oc- cover for this species on the dif- and Black Rock Butte contained curred in the Black Rock Butte ferent buttes. The Black Rock a higher percentage of sand and area, Seedling numbers aver- Butte relict area had the greatest a lower percentage of clay than aged only 1.18 per plot on the average of living winterfat their associated relict areas. Tex- relict and 12.92 per plot on the plants per plot (5.86) and the ture class changed from clay grazed area. The average num- Pilot Butte grazed area had the loam on Pilot Butte to loam on ber of seedlings for both areas lowest (1.98). Variations among the associated land, and from (including the relict and grazed years ranged from an average of clay loam on Black Rock Butte to lands) was greatest in 1938 (33.65 5.62 living plants per plot in 1939 sandy clay loam on the sur- per plot) and 1939 (20.92) and to 1.95 in 1940. Differences be- rounding grazed land. averaged 13.50 per 100 square tween locations were significant The pH of the soil was slightly feet for the 5 years of study. at the 0.05 level, and all other lower on all grazed locations, and There were no seedlings in any differences were non-significant. the amount of soluble salts, al- of the plots in 1956. When these The number of dead winterfat though low in the soils from the VEGETATION OF RELICT AND GRAZED AREAS 109

was greater on both grazed areas studied than on their associated relicts. Ground cover of shrubs and semi-shrubs on the grazed land was the main factor ac- counting for this greater total. Changes in ground cover were greater on the grazed areas than on the relicts for the years of extremes in precipitation. Shrubs, semi-shrubs, and forbs contributed a greater percent- age, and grass and grass-like spe- cies made up a smaller percent- age of the total ground cover on the grazed area than on the relict areas. FIGURE5. A view of the northwesternpart of the Black Rock Butte relict area showing One shrub, three grasses, one the tall shrubby vegetation common to this part of the butte. Photo taken in June 1956. sedge, and thirteen forb species occurred on plots of the grazed relict areas, was even lower in to the shrubby vegetation and areas, but were not found on the samples from the grazed. Air- primarily to big sagebrush. How- relict plots. Two forb species oc- water permeability tests indicate ever, ground cover of forbs also curred on the relicts but not on all soils sampled to be stable in varied considerably on some of the associated grazed land. structure. the grazed plots. Soil samples analyzed from Spring rainfall has consider- each of the relict areas and from Conclusions able effect upon establishment the grazed areas varied only Grazing has apparently caused of big sagebrush seedlings. This slightly between the two loca- a shift in vegetational composi- factor does not operate alone, as tions. no seedlings were evident in tion on the shallow soil sites of LITERATURE CITED 1956, which was a year with Sweetwater County, Wyoming, BEETLE,ALAN A. 1952. A relict area more spring precipitation than as evidenced by the greater num- on the Wyoming shortgrass plains. ber of forb species present on occurred in 1939 and 1940 when Jour. Range Mangt. 5: 141-143. grazed areas, greater percentage relatively large number of seed- lings were recorded. CLEMENTS,FREDERIC E. 1934. The re- of ground cover made up of lict method in dynamic ecology. shrubs, and smaller amounts of Under the conditions of this Ecology 22: 39-68. grass when compared with close- study it must be concluded that big sagebrush and needle-and- DYKSTERHUIS,E. J. 1946. The vege- ly associated relicts. tation of the Fort Worth prairie. thread are increasers, while The use of relict areas as a Ecol. Mono. 16: 12-15. basis for judging grazed range winterfat is a decreaser under grazing. The reaction of other GARDNER,J. L. 1950. Effects of thirty lands is a common practice. years of protection from grazing in species to grazing pressure was However, this study shows a desert grassland. Ecology 31: 44-50. not sufficiently definite for ac- wide variation in the vegetation- curate classification as increasers NELSON,AVEN. 1898. The Red Desert al composition and ground cover of Wyoming and its forage re- or decreasers. between relicts of similar shal- sources. U. S. Dept. of Agr. Bul. 13. low soil sites separated by ap- Agros. 33. 72 pp. Summary proximately 50 miles. This fact STEWART, GEORGE,AND S. S. HUTCH- emphasizes the importance of Black Rock Butte and Pilot INGS. 1936. The point observation using many relicts or protected Butte in Sweetwater County, plot (square foot density) method areas as a basis for judging the Wyoming, were studied as true of vegetation survey. Jour. Am. Sot. Agron. 28: 714-722. condition of grazed lands. natural relict areas to compare Ground-cover fluctuations on vegetational composition and VASS, A. F. AND ROBERT L. LANG. both relict and grazed areas were ground cover with that of grazed 1938. Vegetative composition, den- sity, grazing capacity, and grazing high among years. Fluctuations land of similar rangesite char- land values in the Red Desert area. in this factor were greater on the acteristics surrounding these Wyo. Agr. Exp. Sta. Bul. 229. 72 grazed areas than on the relicts buttes. This is an arid area used PP. for different years. These primarily for winter grazing by WHITFIELD, CHARLESJ. AND EDWARD ground-cover changes among sheep. L. BEUTNER. 1938. The need for years may be largely attributed Average total ground cover natural areas. Ecology 15: 328-329.