understory consisted predomin- Effects on Associated Species of Burning, antly of bluebunch wheatgrass (Agrcpyron spicutum) , thick- Rotobeating, Spraying, and Railing Sagebrush spike wheatgrass (A. dusystuch- yum), fescue (Festucu ida- hoensis), plains reedgrass WALTER F. MUEGGLER AND JAMES P. BLAISDELL (Culumugrostis montunensis), subalpine needlegrass (Stipu Range Conservationists, Intermountain Forest and Range columbiuna), blue grasses (Poa Experiment Station, Forest Service, U. S. Department of spp.) , Agriculture, Ogden, . threadleaf sedge (Curex filifoliu), fleabane ( corymbosus) , lupines (Lupinus Control of big sagebrush effect of various selective spp.), and minor amounts of (Artemisia tridentata) on heav- methods of sagebrush control. other grasses and forbs. Besides ily infested areas has long been Blaisdell (1953) reported in de- the dominant big sagebrush, the recognized as an effective range tail the effect of sagebrush burn- shrubby vegetation included improvement practice. By kill- ing on residual vegetation on spmeless gray horsebrush (Te- ing this one undesirable species the Upper Snake River Plains of trad ymiu cunescens) , down y the grazing capacity of range- Idaho. Bohmont (1954), Hyder rabbitbrush (Chrysothumnus lands can often be increased and Sneva (1956)) and Blaisdell puberulus) , and antelope bitter- manyf old. and Mueggler (1956) are among brush (Purshiu tridentutu) . A During the past 20 years sev- those who have studied the more detailed listing of species eral methods have been devel- effect of the sagebrush-killing and relative amounts can be oped to reduce sagebrush num- 2,4-D sprays on associated vege- found in Table 2. bers and bring about such range tation. Burning, spraying, roto- Methods improvement (Pechanec et al., beating, and railing as well as 1954). Generally these methods other methods were described by A 54-acre area was divided can be divided into two main Pechanec et al. (1954)) but only roughly into quarters, and a dif- categories: those that destroy all general information was pre- ferent treatment was imposed existing vegetation, and selective sented on the relative effects of upon each. An untreated check methods that destroy sagebrush these methods on associated veg- was reserved adjacent to the without complete destruction of etation. Evidently little infor- treated areas. Field-scale treat- the herbaceous understory. mation is available on direct ments were used and cost data When there are few of comparisions of various meth- were kept. desirable species, consideration ods of sagebrush removal under One-quarter was sprayed in in the selection of a control carefully controlled conditions. early June 1952 using a ground method generally need” be given A comparison of four methods spray unit with a 16-foot boom only to effectiveness of sage- of sagebrush control that main- mounted on a truck. The 2,4-D brush removal, relative cost, and tained the native herbaceous ethyl ester (Weedone 48) was development of a satisfactory understory-burning, rotob,eat- applied in a water carrier at the seedbed. Establishment of for- ing, railing, and spraying with rate of 2 pounds of acid and ap- age species will be dependent 2, 4-D-was started in 1952 on proximately 20 gallons of water upon artificial seeding. On the the Upper Snake River Plains per acre. Unfortunately it was other hand, if a good stand of in southeastern Idaho. The study necessary to spray the area when desirable species is present, it is area is in a fairly homogeneous, winds were fairly high. Spray upon these that increased pro- dense stand of big sagebrush be- drift, combined with unexpected duction should depend; selection tween 2 and 3 feet in height, on rockiness of the area, resulted in of the eradication method should the spring-fall range of the U. S. poor spray coverage with the not only consider degree of sage- Sheep Experiment Station near ground vehicle. For this reason brush kill and cost but also the Dubois, Idaho. The topography the sagebrush kill on the experi- effect of treatment upon these has only slight relief with little mental area was much lower associated plants. It is with this surface drainage; the sandy- than would ordinarily be ex- question of the effect of treat- loam soil is underlain by basal- pected. Similar treatments on ment on associiated fbrage tic lava. Precipiation averages nearby areas have caused almost species that this paper is con- about 13 inches annually. complete eradication of big sage- cerned. A good understory of native brush. Considerable information has perennial species was present One of the areas was roto- already been obtained on the throughout. This herbaceous beaten in early August 1952. A

61 62 ’ WALTER F. MUEGGLER AND JAMES P. BLAISDELL standard Case rotobeater with windward. The fire was set si- was of necessity made prior to chain flails was run over the multaneously along both wind- the early June spraying and con- area only once to shred the ward edges and a hot, clean burn sequently before maturity. brush. The machine was ad- was obtained. At this time it was difficult to justed so that the flails just The entire study area was de- segregate individual grass spe- cleared the ground surface. ferred from grazing the year of cies rapidly; therefore all grasses Ample power was supplied by a treatment and the year follow- were handled as one unit in the wheeled tractor. ing. Moderate grazing by sheep 1952 inventory. The 1953 and Railing was also done in early was permitted thereafter. 1955 inventories were made later August. The rail used was con- Each area, including the un- in the growing season, and structed of three 11-foot sections treated check, was sampled by grasses were segregated by indi- of heavy railroad rail loosely 20 permanent, 48-square-foot vidual species with the exception bolted together at the ends to circular plots spaced in a regular of two pairs, Agropyron dasysta- make a semiflexible 33-foot pattern. The weight estimate chyum-Calamagrostis montanen- length of rail. The center section method (Pechanec and Pickford, sis and Carex filifolia - Festuca was of double weight. Drags 1937) was used to obtain herbage idahoensis. Because of similarity were attached to the rear to keep production by species prior to of vegetative appearances, the rails upright. A deisel motor treatment (1952)) 1 year after grouping in these two pairs was patrol pulled this rail over the treatment (1953)) and 3 years necessary for rapid field identi- area twice, once each way. after treatment (1955). Esti- fication. The fourth area was burned in mates were also made of the per- Results And Discussion late August. Prior to burning, a centage of herbage available to standard fire line was con- sheep, that is, the percentage Eff ecfs of Treatments structed, consisting of a double that was not obstructed by All four treatments caused grader line and a lOO-foot back- brush. sizable reductions in sagebrush fired strip on the leeward sides In order to obtain pretreat- and increases in herbaceous spe- and a single grader line on the ment data, the 1952 inventory cies (Fig. 1). Burning and roto-

cl GRASS .I’ . .,.:j::. F 0 R B S 900 q others SHRUBS sagebrush

‘52 ‘53 /55 ‘52 ‘53 ‘55 ‘52 ‘53 ‘55 ‘52 ‘53 ‘55 ‘52 ‘53 ‘55 UNTREATED BURNED ROTOBEATEN SPRAYED RAILED FIGURE1. Grass, forb, and shrub herbage production prior to treatment (1952) , and 1 year (1953) and 3 years afterward (1955) , as compared to production on untreated range. EFFECTS OF SAGEBRUSH TREATMENTS ON ASSOCIATED SPECIES 63 beating reduced sagebrush con- caused better than a one-third damage from 2,4-D is common siderably more than spraying increase, and railing resulted in (Bohmont, 1954; Hurd, 1955; and railing. Burning brought approximately a one-fifth in- Blaisdell and Mueggler, 1956). about a much greater increase in crease of grass. Apparently Very pronounced reductions forbs than any of the other treat- burning was the only treatment of shrubs occurred with all ments, and rotobeating and that actually injured the grasses. treatments. This reduction re- spraying a greater increase in Although other treatments flects the effect of treatment grasses. Railing caused inter- caused no injury, only on the upon sagebrush which composed mediate increases in both grasses sprayed area were grasses able almost 95 percent of before- and forbs. Despite general in- to take immediate advantage of treatment shrub production. creases in grasses, forbs, and the release from sagebrush com- Greatest reductions in sagebrush shrubs other than sagebrush on petition. During the first grow- were apparent the first year all treated areas, total vegetal ing season after treatment after treatment; thereafter pro- production 3 years after treat- grasses were noticeably more duction increased slightly. After ment was still considerably less vigorous on the sprayed range 3 years sagebrush production on than on the untreated area. The than on adjacent areas, and herb- the burned area was still only 11 inability of a predominantly her- age production was greatly in- percent and on the rotobeaten baceous cover to produce as creased. Substantial gains in area only 14 percent of what it much foliage dry matter as a grass production the first year would have been with no treat- shrub-herb complex has been after spraying with 2,4-D is ap- ment; production on the sprayed previously observed by Blaisdell parently the rule (Bohmont, and railed areas was 50 and 43 (1953). 1954; Hurd, 1955; Hyder and percent, respectively. Produc- It should be noted that there Sneva, 1956). However, the rea- tion of the shrub group exclud- were differences in herbage pro- son for such increase on the ing sagebrush was increased by duction on the various areas sprayed area and not on the roto- all treatments with the excep- prior to treatment as well as beaten or railed area is not clear. tion of an initial reduction on natural year-to-year fluctuations This may have resulted from re- the burn. By the third year pro- on the untreated area (Fig. 1). duced evaporation and a greater duction of associated shrubs, as Therefore, to determine accur- accumulation of snow among a group, had doubled and tripled ately the effect of treatment up- sagebrush skeletons. A compar- on treated areas. Such increases on vegetation it is necessary to ison of March 1 snow accumula- are attributed to the sprouting compare relative rather than tion showed 0.9 inch more mois- habit of most of these species. absolute vegetal changes over ture in the form of snow on the Additional information on ef- the S-year period. sprayed than on the rotobeaten fectiveness of the various treat- As compared with production area, about 7 percent of the aver- ments for sagebrush control is on the untreated area, grasses age annual precipitation. It is supplied by plant counts (Table decreased slightly on the burn also possible that the vigorous 1). Before treatment the num- the first year, remained about grass production may have re- ber of sagebrush plants per 48- the same on the rotobeaten and sulted from an actual stimulus square-foot plot ranged from 9.8 railed areas, and increased mark- by the 2,4-D. on the area to be burned to 11.6 edly on the sprayed area. Three Production of forbs increased on the untreated area, or an av- years after treatment the delet- on the burned, rotobeaten, and erage of approximately 11 plants erious effect of burning on total railed areas during the first sea- per plot. The greatest reduction grass production had been over- son, but decreased on the in numbers was obtained by come, but there still had been no sprayed area in relation to the burning, which killed all the old increase attributable to burning. untreated area. By the third plants. The first year after Blaisdell (1953) found similar year relative forb production on treatment the burned area had injury to grasses the first year all treated areas was slightly only 1.5 plants per plot (all seed- after burning, but reported that less than the first year after lings) and by the third year rhizomatous species generally treatment. However, burning after treatment only 2.8 plants recover fully within 3 years after was still considered responsible per plot, or an over-all reduc- burning, and that losses incurred for a 61-percent increase in total tion in numbers of 72 percent. by most other grasses are re- forb production, rotobeating a Railing was the least effective covered in 12 to 15 years. Roto- 50-percent increase, and railing method for reducing sagebrush beating resulted in a 50-percent a 20-percent increase, and railing numbers, as there were 3.8 gain in grass production over it would have been with no plants per plot the first year what it would have been by the treatment. During this same after treatment and 6.2 by the end of the third year had there period spraying had caused a 39- third year, or a reduction of only been no treatment, spraying percent loss in forbs. Such forb 42 percent. Despite the rather 64 WALTER F. MUEGGLER AND JAMES P. BLAISDELL

Table 1. Average numbers of sagebrush plants per 48-square-foot sample crease. As pointed out previ- plot on five different areas before, 1 year after, and 3 years after using ously, there was a lag in re- different methods fo kill sagebrush. sponse of grasses to release by 1952 1953 1955 rotobeating the first year as prior to First year Third year compared to the forbs. This is treatment after treatment after treatment illustrated by relative produc- Untreated 11.6 10.4 11.9 tion figures between the two Burned 9.8 1.5” 2.8 groups over the entire period Rotobeaten 10.7 2.0 3.6 (Fig. 1). Before treatment pro- Sprayed 11.1 3.5 4.3 duction of grasses was almost Railed 10.7 3.8 6.2 -_--. ___- .-_ six times that of the forbs; 1 year after it was less than twice *Seedlings as great; and in 3 years, by which time grasses had taken low kills of sagebrush from the species. Species other than sage- advantage of reduced brush brush that were harmed by 2,4-D treatment in this test, it competition, it was over three burning were should be recognized that spray- Agropyron spica- times as great as the forbs. ing is an effective method for turn, Antennaria microphylla, The most pronounced effects sagebrush eradication; kills of Penstemon radicosus, and espe- of spraying were reductions in more than 90 percent are not un- cially Carex-Festuca and Pur- Lupinus and Erigeron, besides common (Blaisdell and Mueg- shia. Although Stipa corn&a sagebrush, and increases in most gler, 1956). may have suffered a loss, it was grasses, especially Stipa colum- probably not as severe as the biana. The indicated increase of Species Herbage Yields figures indicate; frequency data Stipa corn&a is of doubtful va- Herbage yields of individual suggest that there was less of lidity because of the erratic oc- species prior to treatment and 3 this species on the burned area currence of this species on the years afterward are shown in originally than on the untreated. sample plots. Both and Table 2. The data are grouped Burning appeared to benefit Purshia Tetradymia benefited by spray- into arbitrary desirability most Agrop yron-Calamogrostis, ing. Production of other species classes for sheep in order to ob- St ip a columbiana, Astragalus was not appreciably changed. tain a better understanding of convaZZarius, Lupinus, Erigeron, changes in amount and type of Chrysothamnus, and Tetradymia. With greater sagebrush kill forage. The classif ication is These observations agree essen- from spraying, which might or- based upon sheep preferences in tially with long-term observa- dinarily be expected, grasses this particular spring-fall range tions by Blaisdell (1953)) who that increased under this limited area; the same species may of found, however, that initial de- reduction in competition would course be rated differently else- creases in many of these species, probably show greater gains, and forbs would probably be where. Although pretreatment with the exception of Festuca damaged more severely. Blais- estimates were not made for in- and Purshia, were only tempo- dell and Mueggler (1956) found dividual grasses, the effect of rary and that they eventually that such species as each treatment is indicated by regained or exceeded their orig- Penstemon, Arnica fulgens, Comandra um- comparison of subsequent yields inal production. bell&a, Eriogonum heracleoides, with those on the other areas. None of the species on the and Chrysothamnus, not obvi- Caution must be exercised in de- rotobeaten area, other than sage- ously affected in this study, were ducing treatment effects by di- brush, was seriously injured by lightly damaged by spraying rect comparison with the un- this treatment; all appear to with 2,4-D. treated area. This is especially have responded favorably. The true in the case of the grasses Changes induced by railing apparent decrease in Agropyron where no before-treatment fig- were not as pronounced as with spicatum is deceptive, for fre- other treatments; ures are available for individual no species quency data indicate a mark- were greatly favored and none species. An attempt has been edly greater occurrence of this aside from sagebrush was notice- made to avoid erroneous inter- species in 1953 on the untreated ably damaged. The effects of pretations by utilizing supple- area than on any of the treated railing were similar to those of mentary information to indicate plots, suggesting an inequality rotobeating, but to a lesser de- inequalities in original produc- of original production. What gree because of lower sagebrush tion. seems a slight decrease on the kill. Species that appear to have Of all the treatments, burning rotobeaten area, and on the benefited most are Carex-Fes- caused some of the most pro- sprayed and railed areas as well, tuca, Poa, Lupinus, Chryso- nounced changes in individual may actually be a slight in- thamnus, and Tetradymia. The EFFECTS OF SAGEBRUSH TREATMENTS ON ASSOCIATED SPECIES 65

Table 2. Herbage production before treatment (1952) and 3 years after (1955) on comparable areas subjected to different meihods of sagebrush removal.

Untreated Burned Rotobeaten Sprayed Railed - 1952 1955 (Air-dry pounds per acre) Total Grass 207 268 229 295 250 490 245 430 200 310 Total Forbs 49 114 82 307 43 150 65 92 80 223 Total Shrubs 367 481 361 137 368 102 281 219 267 184 Desirable Species Agropyron spicatum 41 18 33 29 30 Agropyron dasystachyum- Calamagrostis montanensis 31 68 68 46 30 Carex filifolia- Festuca idahoensis * 106 * 36 * 204 * 146 * 143 Koeleria cristata 3 8 7 5 9 Poa spp. 18 35 66 25 34 Stipa comata 13 4 20 40 1 Stipa columbiana 51 123 91 138 63 Astragalus miser 1 7 - 3 1 8 Trace 4 1 6 Astragalus convallarius - 1 - 10 - 5 - 2 - 4 Lupinus spp. - 1 41 148 3 26 15 4 6 51 Penstemon radicosus 3 6 5 3 2 4 5 7 5 9 Purshia tridentata 8 8 14 Trace - - 5 13 - - Others 1 8 6 26 - Trace 2 2 5 8 Total Desirable 220 294 295 482 256 532 272 461 217 388 Moderately Desirable Achilles lanulosa 1 2 3 9 1 1 - Trace - 1 Antennaria microphylla 4 18 4 4 1 17 5 18 7 39 Arnica fulgens 8 9 1 2 4 7 5 6 23 20 Comandra umbellata Trace 1 3 2 1 3 5 3 1 1 Erigeron corymbosus 26 40 9 38 21 49 14 13 21 44 Eriogonum heracleoides 3 7 8 13 4 14 8 17 2 10 Chrysothamnus puberulus 4 11 5 68 4 16 3 12 3 20 Others 6 3 2 3 4 4 4 3 5 4 Total Mod. Desirable 52 91 35 139 40 111 44 72 62 139 Undesirable Species Artemisia tridentata 346 452 337 47 360 67 269 178 257 144 Tetradymia canescens’ 5 8 5 21 4 20 4 17 6 20 Others Trace 18 - 50 1 12 2 13 5 26 Total Undesirable 351 478 ,342 118 365 99 275 208 268 190 Total Production 623 863 672 739 661 742 591 741 547 717 *1952 production for total grasses given above. remaining species showed little pounds per acre on the untreated areas, respectively. It is appar- change that can be directly at- range. Production of the desir- ent, then, that all treatments tributed to railing. The apparent able class alone was 482, 532, 461, produced sizable increases in reduction in Stipa comata ’ is and 388 pounds per acre for the available forage and that in- probably not real, since frequen- respective treatments, while that creases were greatest on burned cy data suggest very erratic dis- for the untreated range was only and rotobeaten range. It can be tribution of this species. 294 pounds. Such changes were expected that availability of for- Changes in amount and type not immediate but occurred age on the sprayed area will in- of forage are indicated by class gradually over the 3-year period. crease as the sagebrush skeletons totals of desirable and moder- The rotobeaten area especially decay. ately desirable species in Table 2. was slow to respond with an in- Three years after treatment total crease in forage production. Eradication Methods air-dry production of desirable Only 75 percent of the herbage Cost of the various eradication and moderately desirable species on the untreated range was con- methods may vary considerably, was 621, 643, 533, and 527 pounds sidered available to livestock, depending upon size of area to be per acre on burned, rotobeaten, whereas 98,98,88, and 90 percent treated, equipment used, etc. The sprayed, and railed areas, respec- was available on the burned, following figures may be consid- tively, as compared to 385 rotobeaten, sprayed, and railed ered indicative of comparative 66 WALTER F. MUEGGLER AND JAMES P. BLAISDELL

costs. The cost of burning, which range at the U. S. Sheep Experi- Chrysothamnus puberuZus and includes construction of ade- ment Station near Dubois, Idaho. Tetrad ymnia canescens, which quate fire lines, actual burning, Burning, rotobeating, spraying sprout readily, increased under and the loss of grazing because with 2,4-D, and railing were com- all treatments. of l-year mandatory deferment, pared on adjacent areas to obtain All treatments produced siz- is about $0.70 per acre. Roto- a better understanding of rela- able increases in available for- beating costs vary from $3.00 to tive effects upon associated na- age. Three years after treatment $6.00, spraying with 2 pounds tive species and forage produc- total production of desirable and acid equivalent 2,4-D, $2.50 to tion. moderately desirable species was $3.00, and railing twice over, Burning was the only treat- approximately two-thirds great- $2.00 to $2.50 per acre. It is ment that injured any grasses. er on the burned and rotobeaten readily apparent that despite the The Carex filifolia-Festuca ida- areas, and better than one-third l-year grazing deferment, burn- hoensis group was most severely greater on the sprayed and railed ing is by far the least expensive reduced by burning. Other areas, than on the untreated method of sagebrush eradication. grasses, though set back tempo- range. Only 75 percent of the More specific information on rarily, soon recovered. Grasses grass and forbs was available to burning procedure and costs is as a group were greatly favored livestock on the untreated range, given by Pechanec, et al. (1954a). by the other treatments. There but 98,98,88, and 90 percent was When selecting a method best was a lag in increased grass pro- available on the burned, roto- suited for a particular area, it is duction the first year after roto- beaten, sprayed, and railed areas, necessary to consider not only beating and railing, but grass respectively. effect upon existing vegetation production increased immediate- and cost but also adaptability of ly with spraying. LITERATURE CITED the method to the site and effect Although burning brought BLAISDELL,JAMES P. 1953. Ecological upon erosion hazard. Rotobeat- about the greatest increase in effects of planned burning of sage- ing and railing leave a good litter total forbs, Antennaria micro- brush-grass range on the Upper cover and do not increase the phylla and Penstemon radicosus Snake River Plains. U. S. Dept. erosion hazard, but both methods were injured; Astragalus conval- Agr. Tech. Bull. 1075, 39 pp. are restricted to fairly rock-free Zarius, Erigeron corymbosus, and AND W. F. MUEGGLER.1956. Effect of 2,4-D on forbs and areas. Rotobeating should not be Lupinus spp. were most bene- shrubs associated with big sage- attempted where rocks protrude fited. Rotobeating and railing brush. Jour. Range Mangt. 9: 38- more than 3 inches above the soil tended to favor all forbs, perhaps 40. surf ace or excessive equipment Lupinus spp. more than the BOHMONT, DALE W. 1954. Chemical breakage may result. Spraying others. Spraying caused a pro- sagebrush control-good and bad. does not increase the erosion nounced reduction in forbs, with Agr. Exp. Sta. Cir. 54, hazard appreciably and is suit- Lupinus spp. and Erigeron co- 7 PP. HURD, RICHARD M. 1955. Effect of able for any terrain where air- being most severely rymbosus 2,4-D on some herbaceous range planes can be used, but may be affected. plants. Jour. Range Mangt. 8: 126- highly undesirable on sheep Sagebrush was greatly re- 128. ranges which have a high per- duced by all treatments. By the HYDER,D. N. AND F. A. SNEVA. 1956. centage of susceptible forbs. third year after treatment sage- Herbage response to sagebrush Burning is suited to any terrain brush production on the burned spraying. Jour. Range Mangt. 9: where the fire can be kept under area was only 11 percent of what 34-38. control; however this method it would have been with no PECHANEC,J. F. AND G. D. PICKFORD. 1937. A weight estimate method should not be used on readily treatment, 14 percent on the for the determination of range or erodible soils or on slopes steep- rotobeaten area, 50 percent on pasture production. Jour. Amer. er than about 30 percent, for the sprayed area, and 43 percent Sot. Agron. 29: 894-904. burning destroys all litter and on the railed area. Sagebrush --, et al. 1954. Control- exposes the soil to erosion. plants were reduced to 28, 33, 39, ling sagebrush on range lands. and 58 percent of their former U. S. Dept. Agr. Farmers’ Bull. Summary 2072, 36 pp. numbers under the respective ,et al. 1954a. Sage- Four different methods of se- treatments. Purshia trident&a brush burning -good and bad. lective sagebrush control were was severely injured by burning; U. S. Dept. Agr. Farmers’ Bull. studied on sagebrush-bunchgrass it was favored by spraying. 1948, 34 pp.