Table Of Content

document archived Historic, Do not assume content reflects current scientific knowledge, policies, or practices. 4 7L iUn United States Comparing Revegetation Department of Agriculture Success on Monitored and Forest Service Intermountain Nonmonitored Gas Well Sites Research Station Research Note Southwestern Wyoming INT-417 in August 1993 Karri A. Smith CO Jeanne C. Chambers o 3: — Well sites and road and pipeline rights-of-way asso- Abstract Council on Environmental Quality regulations implementing the National Environmental PolicyAct au- ciated with oil and gas exploration and development thorize Federal land management agencies to monitor sur- cumulatively disturb large areas ofland. Council on face disturbance mitigation. But no standardized methods Environmental Quality (CEQ) regulations implement- exist for monitoring the reclamation ofsites disturbed by ingthe National Environmental Policy Act (NEPA) oil and gas exploration and development. This studyexam- (40 CFR 1505.2 and 1505.3, July 1989) authorize inedthe effects ofcompliance and revegetation monitoring Federal land management agencies to monitor sur- on reclamation success ofthe Riley Ridge Natural Gas face disturbance mitigation. But to date, no standard- Project in Wyoming. It also evaluated various vegetation ized reclamation monitoring procedures exist for oil monitoring methods, including those used by the Bureau and gas exploration and development such as those ofLand Management. Monitoring increased reclamation provided for under the Surface Mining Control and success as problems were identified and remedial meas- ReclamationAct (SMCRA) of1977. Because no speci- ures were promptly applied. Three to 5 years after revegetation, total ground cover on monitored well sites was less fications for monitoring revegetation success exist, (91 percent) than on adjacent reference areas (96 percent), current monitoring programs developed for the oil but higher than on nonmonitored well sites (44 percent) and gas industry are often vague and vary among (p < 0.05). Species richness and the percentage ofdesir- Federal land management agencies, regions, districts, able species were higher on monitored than on nonmoni- and projects. tored well sites. Revegetation seedmixtures contained few Lack ofenforced monitoring, noncompliance with reference area species, and species and life-form similarity permit stipulations, and incorrect implementation of (Spatz' index) between reference areas and monitored or prescribed reclamation often result in revegetation nonmonitored well sites were low. The single Bureau of failure and increased soil erosion (Ostler and others LandManagement transect for monitoringbasal cover 1985) In an internal report from the Bureau ofLand provided general trends, but multiple transects or a repli- . Management (BLM), U.S. Department ofthe Interior cated design using quadrats were necessary for variance estimates and statistical comparisons. (USDI), Bauer (undated) stated that the negative con- sequences ofnoncompliance generally increase with the environmental sensitivity ofthe project area. For Keywords: compliance, reclamation, vegetation surface mining activities, revegetation monitoring assessment, monitoring, oil data demonstrate compliance with performance criteria standards and are used to develop remedial measures necessary to ensure site stability (Edgar and others 1985; Grip and Bakken 1986; Office ofTechnology Assessment 1986). Changes in vegetation composi- A tion and structure monitored over time may also indi- Karri Smith was a graduate student, ForestResources De- cate whether reclamation was successful (Keamerrer partment, Utah State University, Logan, when the studywas conducted. Her current address is: KarriA. Smith Consulting, 1986). In an effort to establish a monitoring frame- P.O. Box 521107, Salt Lake City, UT. Jeanne C. Chambers is re- work for evaluating land reclamation success in the SMCRA search ecologist, Intermountain Research Station, ForestService, United Kingdom, Tomlinson (1984) viewed U.S. Department ofAgriculture, Reno, NV. as an important example oflegislated standards for Theuse oftrade orfirm names in this publication isforreader determining trends in soil stability and ecosystem information and does notimply endorsementbythe U.S. Depart- development. ment ofAgriculture ofany productor service. 1 Due to the magnitude ofthe proposed development ranges from 90 to 100 days. The forested north-south m in the Riley Ridge Natural Gas project area in south- trendingridges at 2,833 to 3,000 receive 51 to 64 cm western Wyoming, the sensitive nature ofthe area, ofprecipitation, have a growing season of60 to 75 days, and limited agency personnel to do the monitoring, and a mean annual air temperature of to 2°C. Soil ajoint agency/operator monitoring program was de- textures include sandy loams, loams, and clay loams signed for Exxon Corporation's LaBarge Fogarty Creek ranging from very gravelly or very cobbly loams to Unit. The program ensures operator compliance with heavy clay loams. permit stipulations and successful reclamation of The gas well study sites are generally within the surface disturbances. In developing the program, the mountain big sagebrush (Artemisia tridentata BLM and Forest Service were guided by CEQ direc- vaseyanaVbluebunch wheatgrass (Agropyron spica- tives implementing the NEPA, the Onshore Oil and tum), and Wyoming big sagebrush (A. tridentata Gas Order No. 1 (43 CFR 3164) of1983, and the Riley wyomingensis)/b\uebunch. wheatgrass vegetation Ridge Natural Gas Project environmental impact state- types. In addition, a widespread sagebrush complex ment (USDI BLM and USDA FS 1984). This program comprising mountain big sagebrush, alkali sage- provided an opportunity to study whether monitoring brush (A. longiloba), and black sagebrush (A nova) influenced the success ofreclamation efforts at gas exists throughout the project area with each type well sites. dominating portions ofa community. Douglas-fir The Fogarty Creek Unit monitoring program speci- {Pseudotsuga menziesii) and aspen (Populus tremu- fiedjoint inspection ofthe well field development by loides) communities occur primarily on cool, north- the operator and authorizing agency to monitor suc- facing aspects in association with the mixed sage- cess and maintenance oferosion control structures and brush types. revegetation for at least two growing seasons or until attainment ofperformance criteria standards. Atask METHODS force ofagency personnel and other resource special- ists ensured that monitoring commitments made by Monitored and nonmonitored gas well sites selected the operator were carried out. for study were in similar elevational ranges and were A quality assurance/control program comprising reclaimed during a similar period according to final or BLM agency and Exxon Corporation personnel monitored production phase reclamation specifications. The construction activities ofaccess road and pipeline identified eight nonmonitored well sites, reclaimed in rights-of-way and well sites. A reclamation specialist 1984, 1985, and 1986, that were within or near the for Exxon Corporation was assigned to oversee imple- vicinity ofthe Riley Ridge project area. All eight sites mentation ofreclamation procedures to ensure compli- were evaluated and consisted ofsingle gas exploration ance with lease stipulations and to determine the effec- wells drilled by various operators. No formal operator tiveness ofconstruction and reclamation measures. or agency monitoring was conducted during project Following reclamation, an independent reclamation construction or following reclamation ofnonproduc- consultant monitored revegetation success and soil tive well sites or unused portions ofproducing well erosion condition annually until performance criteria sites. Five monitored well sites were selected from standards were met. Areas requiring remediation were Exxon Corporation's LaBarge Fogarty Creek Unit of identified in agency task force reports and monitored the Riley Ridge Natural Gas Project area. They had more intensively until problems were resolved. also been reclaimed during 1984, 1985, and 1986. This The objectives ofthis study were (1) to examine the unit contained the greatest number ofwell sites meet- effects ofcompliance, effectiveness, and revegetation ing the selection criteria. Compliance, effectiveness, monitoring on the success ofreclamation efforts and and revegetation success monitoring was conducted (2) to compare the BLM vegetation monitoring meth- at all monitored gas well sites from initiation ofcon- ods with a more intensive sampling regime. struction to attainment ofrevegetation and soil stability performance criteria standards. STUDY AREA Reclamation had been conducted from 3 to 5 years prior to this study and for monitored sites included The Riley Ridge Natural Gas Project is within the recontouring, seedbed preparation, seeding, fertiliza- Green River Basin ofthe Idaho-Wyoming Thrust Belt tion, mulching, and installation ofsurface runoffand and extends across Sublette, Lincoln, and Sweetwater erosion control structures. All nonmonitored well sites Counties in southwestern Wyoming. The gas well sites had been recontoured, but it was unknown whether evaluated were at elevations ranging from 2,381 to proper seedbed preparation, fertilization, or reseeding 2,962 m. Throughout most ofthe Green River Basin, had occurred. Seed mixtures for monitored well sites precipitation ranges from 17 to 23 cm per year, average contained six native and introduced grass species and annual air temperature is 6 °C, and the growing season 2 three to four forbs (table 1). Seed mixes for nonmoni- one reference area. Using a stratified random sam- tored well sites generally contained two to four native pling method (Chambers and Brown 1983; Knight 1978; or introduced grass species (table 2). Vogel 1987), six 33-m line transects spaced at 5-m in- tervals were established on each area. Five quadrats m Vegetation Monitoring (0.25 2 rectangular) were randomly placed along each transect (n = 30). Within each quadrat relative spe- During the 1989 growing season, we conducted veg- cies importance and frequency ofoccurrence were de- etation analysis ofmonitored and nonmonitored well termined by visually estimating the percentage aerial sites using two methods: (1) BLM line transect methods cover ofeach species. We used the following modified and (2) more intensive sampling methods (quadrats and Daubenmire cover categories: multiple transects) that would allow statistical compar- Category Percentag ison ofmonitored, nonmonitored, and reference areas. The BLM line transectmethods consisted ofrandomly 1 <1 1-5 placing one 33-m line transect in a revegetated portion 2 6-15 ofeach ofthe five monitored and eight nonmonitored 3 well sites and in an undisturbed adjacent reference 4 16-25 26-35 area. Evaluation points were located at 30.5-cm inter- 5 vals along each transect (n = 100). Basal ground cover 6 36-45 was recorded at each point as vegetation, litter, bare 7 46-55 ground, or rock/gravel. The BLM methods did not re- 8 56-65 quire species identification. For this study, vegetation 9 66-75 cover was identified by species to determine composi- 10 76-85 86-95 tion and desirable species present and to evaluate spe- 11 cies diversity and frequency. Desirable species were 12 96 - 100 defined as those prescribed in seed mixtures or as re- To determine aerial ground cover, a 35-mm photo- colonizing native perennials. graphic slide was taken vertically over the center of Intensive sampling methods were conducted on one each quadrat and analyzed according to methods de- monitored well site, one nonmonitored well site, and scribed in Chambers and Brown (1983). Slides were projected onto a 100-square grid, and relative per- centages ofvegetation, bare ground, litter, and rock Table 1—Seed mixtures specified for monitored and gravel were obtained by countingthe number of gas well sites squares covered by each category. To determine whether basal cover estimates varied Seed mix Drill rate PLS1 with multiple transects and to permit statistical analy- kg/ha sis ofthe BLM data, three ofthe line transects estab- #1 lished during intensive samplingwere randomly chosen *Bromus marginatus 1.4 within each monitored, nonmonitored, and reference BLM *Agropyron trachycaulum 1.4 area and evaluated according to the methods. *Deschampsia cespitosa .2 Astragalus cicer .7 Data Analysis Alopecurispratensis .5 *Linum lewisii .5 One-way analysis ofvariance (ANOVA) was used Festuca longifolia .2 to compare differences among the five monitored and Poa compressa .2 eight nonmonitored reclaimed well sites and 13 refer- Trifolium hybridum .2 ence areas for which cover data were collected using #2 the BLM line transect method. Differences among *Agropyron trachycaulum 1.4 the intensively sampled monitored and nonmonitored *Agropyron smithii 1.4 well sites and reference area, and between the quad- BLM *Stipa viridula .9 rat and line transect methods, were evaluated Sanguisorba minor .9 with two-way ANOVA. Arcsine transformations were Onobrychis viciaefolia .9 performed on percentage data. Mean comparisons *Agropyron spicatum .7 were performed with Fisher's Protected LSD's. Simi- Astragalus cicer .5 larity between reclaimed well sites and adjacent ref- *Festuca ovina .5 erence areas in species number and composition was *Poa ampla .2 calculated using Spatz' similarity index (Chambers Kochia prostrata .2 and Brown 1983). We compared species lists to spe- 'PLS = pure live seed. cies identified in prescribed seed mixtures for both 'Denotes native status. 3 ) )) ) — Table 2 Dominant species present in 1989 on nonmonitored well sites and species specified for revegetation in seed mixtures. Data are from BLM line transect methods with one transect per site. Dates specify year gas well was drilled Gas well Drill rate PLS1 (kg/ha) Dominant species (percent cover) Shar Alan #1 *Bromus marginatus (2.7) Polygonum spp. (16) (ABD2 1986) *Festuca ovina (2.7) *Sitanion hystrix(2) Astragalus cicer(0.9) 36X-12G *Agropyron dasystachyum (2.3) A. cristatum (17) (PGW3 1985) *Stipa viridula (2.3) Elymusspp. (12) *Elymus cinereus (0.9) A. intermedium (1) *Poa secunda (0.9) *A dasystachyum (P)4 21-28 *A. smithii(2.3) S. inermus (8) (PGW 1981) *A. dasystachyum (2.3) A intermedium (7) A * dasystachyum (5) A tricophorum (1 Polygonum spp. (1 22-28 *A. spicatum (1.8) A. cristatum (7) (ABD 1985) *A. dasystachyum (1.8) *A dasystachyum (7) *Oryzopsis hymenoides (1.8) Sa/so/a /ca//'(3) *S. comata (0.9) *0. hymenoides (P) *S. comara (P) 22-21 M. smithii(2.7) S. inermus (16) (ABD 1984) *A. dasystachyum (2.7) A. tricophorum (8) •A smithii(1) *Artemisia tridentata (1 20-24 A. cristatum (2.7) Polygonum spp. (12) (ABD 1986) "A. dasystachyum (2.7) S. /ca//(P) 22-32 B. inermus (2.3) fi. inermus (15) (ABD 1985) A. intermedium (2.3) Phleumpratense (1 Alopecurispratensis (2.3) *A dasystachyum (1) P. pratense (2.3) "Achillea millefolium (1) A. tricophorum (P) Sa/so/a /ca//' (P) 43-20 B. inermus (2.3) 8. inermus (19) (PGW 1984) *A. dasystachyum (2.3) *A. dasystachyum (5) A. tricophorum (0.9) 'PLS = pure live seed. 2ABD = abandoned. 3PGW = producing gas well. 4P = presentbut <1 percent. "Denotes native status. monitored and nonmonitored well sites. We com- and nonmonitored well site averaged 3.2 and 1.3, re- puted the number and percentage ofspecies pre- spectively. In most cases, dominant grasses on both scribed in seed mixes and total number ofdesirable monitored and nonmonitored well sites were not those species present on each well site. recommended at the highest seeding rates (pure live seed kg/ha). Two nonmonitored well sites (Shar Alan RESULTS OF BLM LINE TRANSECT #1 and 20-24), both abandoned in 1986, were either METHODS not artificially seeded prior to abandonment or reseeding efforts failed. Vegetation ground cover on these To evaluate whether prescribed seed mixtures two sites mainly comprised weedy species with mini- were applied, the dominant species on monitored mal recolonization ofnative species (table 2). and nonmonitored well sites were compared to those The mean percentage desirable species present, as species identified for seeding (tables 1, 2, and 3). The a proportion ofthe total number ofspecies, was 92.4 number ofseeded species present on each monitored percent on monitored well sites and 55.9 percent on 4 )) ) — Table 3 Dominant species present in 1989 on monitored Spatz' similarity indices comparing both monitored gas well sites from transect methods with one tran- and nonmonitored well sites with their adjacent refer- sect per site ence areas were low when computed on a species basis (table 4). Values ranged from 0.00 to 0.72 percent for Gas well Seed mix Dominant species (percent) monitored well sites and 0.00 to 1.97 percent for nonmonitored well sites. Spatz' similarity indices com- 14-05 #1,2 Poa compressa (7) puted by life-form (grasses, forbs, and shrubs) were *Agropyron trachycaulum (4) Phleum pratense higher for monitored well sites (ranging from 3.9 to (3) Agropyron intermedium 23.3 percent) and nonmonitored well sites (ranging (3) Astragalus cicer(2) from 0.0 to 6.2 percent with one site having a similar- *Agropyron smithii(2) ity index of21.0 percent) when compared with adja- Bromus inermus (2) cent reference areas. unknown annuals (1) Cover comparisons showed that basal ground cover *Linum lewisii(~\) was higher on monitored well sites than nonmonitored */A^nyrinunfjxy/irwnini Hi/aaGoVjrGotiadrwfhvlyiuiim11 t\iPVj well sites for both vegetation and litter (fig. 1). Vegeta- Alopecurispratensis (P) tion cover on both monitored and nonmonitored well 21-03 #1 *Agropyron trachycaulum (8) sites was lower than vegetation cover on adjacent ref- *Agropyron smithii(4) erence areas (p < 0.05). Litter cover on the monitored Astragalus cicer(1 well sites was higherthan that occurring on the refer- *Linum lewisii^) ence areas, while litter cover on nonmonitored well *Achillea millifolium (1) sites was lower than that occurring on reference areas unknown annuals (1) (p < 0.05). Rock/gravel coverwas similaron monitored, 15-28 #1,2 Bromus inermus (8) nonmonitored, and reference areas. Bare ground on Agropyron intermedium (7) monitored sites was similar to reference areas while *Agropyron smithii (2) nonmonitored sites had the greatest bare ground area unknown annuals (3) (p < 0.05). Total cover on monitored well sites and Agropyron cristatum (1 adjacent reference areas averaged nearly twice the Astragalus cicer(1) amount ofthat present on nonmonitored well sites Phleum pratense (P) (p<0.05). "Linum lewisii(P) *AaroDvron dasvstachvum (P) Alopecurispratensis (P) — 19-35 #1 *Linum lewisii(\5) Table 4 Spatz' similarity indices comparing monitored and *Agropyron smithii (7) nonmonitored gas well sites with their adjacent ref- Poa compressa erence areas. Values were computed by both spe- (2) unknown annuals (2) cies and life-form. Data were derived from BLM Astragalus cicer(1) line transect species lists and composition with one Phleum pratense (1 transect per site 17-16 #1 Agropyron intermedium (8) Gas well Species Life-form unknown annuals (6) site similarity similarity Poa compressa (5) Phleum pratense Percent (3) Astragalus cicer(2) Monitored *Achillea millefolium (P) Bromus inermus (P) 15-28 0.33 10.17 Elymus spp. (P) 17-16 .33 23.33 *Linum lewisii(P) 21-03 21.75 19-35 3.89 1P = present but <1 percent. 'Denotes native status. 14-05 .72 10.71 Nonmonitored 21-28 1.43 nonmonitored well sites. The mean number ofspecies 22-21 .14 3.65 present was 9.0 (n. = 5) on monitored well sites and 22-28 21.00 22-32 2.20 4.6 (n = 8) on nonmonitored well sites, while the average number ofspecies on adjacent reference areas was 36X-12G 3.56 14.0 (n = 13). Vegetation cover exceeding 5 percent Shar Alan #1 .85 weedy species occurred on one monitored well site and 20-24 four nonmonitored well sites. 43-20 1 .97 6.28 5 RESULTS OF INTENSIVE SAMPLING ground on the monitored well site was less than that METHODS ofthe nonmonitored well site (p < 0.05) but similar to that ofthereference area. The quadrat samplingmeth- Relative species importance was evaluated on inten- ods resulted in higher bare ground on both monitored sively sampled monitored and nonmonitoredwell sites and nonmonitored well sites. Mean total cover (veg- and a representative reference area for both quadrat etation, litter, and rock/gravel) differed among sites sampling and multiple line transect methods (table 5). andmethods (both,p < 0.001). Total cover onthe mon- In general, dominant species on monitored, nonmoni- itored well site was higher than that on the nonmonitored, and reference areas were accurately identified tored well site but similarto that on the reference area by both aerial and basal cover methods. Mean cover (p < 0.05). On the nonmonitored well site, total cover percentage varied by species for each method. Fewer was lower than that on the reference area (p < 0.05). species were identified by basal cover evaluation (line Comparisons ofmeanbasal cover onthe intensively transects). Species frequency was higher for quadrats sampled monitored, nonmonitored, and reference areas because these values were determined from an average using one, two, and three line transects were compared m of30 plots per site (0.25 2), while the values obtained to the single BLM line transect data collected on each using multiple line transects were determined from site (fig. 3). Ground cover values varied for each cover the average ofonly 300 points (three line transects category among each transect and among samplers. with 100 evaluation points per transect) per site. Differences among samplers were greater than among Mean vegetation cover differed among monitored, transects forthe single sample. For the multiple tran- nonmonitored, and reference areas (p < 0.001) but not sects, the results were usually within ±1 standard between sampling methods (fig. 2). Vegetation cover error. on the monitored well site was higher than that on the nonmonitored well site (p < 0.05). On the reference DISCUSSION area, vegetation cover was higher than that on either BLM monitored and nonmonitored well sites (p < 0.05). Vegetation sampling data from both line tran- Mean litter cover differed amongboth sites and meth- sect methods and intensive sampling methods indi- ods (both,p < 0.001). Litter cover on the monitored cated that, 3 to 5 years after revegetation, ground cover well site was higher than that on the nonmonitored similarity between monitored well sites and adjacent well site and reference area (p < 0.05). Multiple line reference areas was higher than between nonmoni- transect evaluation resulted in higher litter cover than tored well sites and reference areas. Monitoring of the quadrat method on both the nonmonitored and compliance, effectiveness, and revegetation success reference areas (p < 0.05). Mean ground cover for appears to have contributed to the higher ground cover rock/gravel differed among sites (p < 0.05) but not values observed on monitored gas well sites because among methods. Mean bare ground cover differed problem areas were identified and prompt remedial among both sites and methods (both,p < 0.001). Bare measures were taken. Factors contributing to low vegetation cover on nonmonitored well sites may include improper site preparation (mixing oftopsoil with sub- soils, compaction, low fertility), improper seeding mix- 100-1 Monitored tures or methods (not planting at optimal seeding i iNonmonitored depths), low seeding rates, low seed viability, or low EB88888 Reference moisture availability during gennination. Remedial measures at monitored well sites, as documented in annual agency task force reports, included maintain- ing surface runoffand erosion control structures, reseeding areas with low perennial cover, and controlling noxious weed invasion, livestock, and wildlife grazing. Nonmonitored gas well sites where such measures were not taken appear more susceptible to revegetation failure and surface runoffand erosion control problems. Qualitative evaluation ofsoil erosion con- Vegetation Litter GRroacvkel/ GBraoruend dition at both monitored and nonmonitored well sites Cover Category revealed substantial rill and gully formation at a ma- — jority ofthe nonmonitored well sites, whereas moni- Figure 1 Comparison of mean percentage cover collected using BLM line transect tored well sites exhibited minimal rill formation and methods for monitored (n = 5), nonmonitored no gullying. (n = 8), and reference (n = 13) areas. Values Generally low similarity indices between monitored are means ±1 standard error. and nonmonitored well sites and adjacent reference 6 — Table 5 Mean aerial and basal cover and frequency of common species present on intensively sampled monitored and nonmonitored gas well sites and adjacent reference areas Quadrats1 Transects2 Aerial Basal cover Frequency cover Frequency - Percent - Monitored (14-05) Phleumpratense -I1AU.1IOQ bo.U 11 17 Oy.Au Agropyron dasystachyum* 7.84 32.0 5.17 4.0 Bromus inermus 5.72 24.0 2.50 3.0 Poa compressa 4.42 84.0 7.83 6.7 00 n Agropyron trachycaulum* 4.26 36.0 2.50 3.7 Agropyron smithii* 1l.Ci>O2 hIbc.aU .0000 -i -7 Alopecurispratensis* l.2b i12o.nU .1lcb .O Linum lewisii* ./4IQ0 11fbi.aU uescurania spp. .44 44.aU Asiragaius cicer 19 n 1.nunu 1.9O I I oWa'"i>/s/"»oi/a-""I KL'ainll .U4 9 n Achillea millefolium* .02 4.0 Nonmonitored (43-20) Bromus inermus 19.08 93.3 13.80 13.0 GO Agropyron intermedium 1H .82 30.0 Agropyron oasystacnyum 11./4I5E £0.0 11.QO9O n no Agropyron smitnii .y<£ £0.0 Agropyron trichophorum .-77-77 2O0A.A0 i1.AUAU i1 .Au Agropyron trachycaulum* .4o Cb."/7 Poa compressa .Jo 0.0 Mtwrnisia inusniaia f0t.7/ oiianiun nysirix .UO 0O..90 .1I7/ .0 Reference (43-20) r\fIGilllola U/Uc7llaLa °.fi 9.9. Qft 7 10 7 1 1miniis Qnn * 2 97 56 7 7/•I1f7 5 3 c_/lyuiiiuin opp. fifi 7 7/ .f0t0T 10 7 L.IIIUIII ICWloll 1i.9000 1109.90 oGiiitaoiniiinunn nh\y/scitrnixv* 9ODK.7/ 1i.P0.09 Phloxhoodii* .30 20.0 Astragalusspp.* .28 30.0 1.00 1.0 Castilleja spp.* .15 13.3 Geranium spp.* .13 26.7 .16 1.0 Sf/pa comata* .05 3.3 Poa secunda* .02 3.3 .83 1.0 Agropyron smithii* 2.50 3.0 Arenaria spp.* .83 1.3 Achillea millefolium* .16 .3 Amelanchieralnifolia* .16 .3 'Monitored (n= 25quadrats); nonmonitored (n= 30 quadrats); reference (n= 30 quadrats). HTiree transects (n= 300 points). 'Denotes native status. areas can be attributed to many factors including few In studies in the Bridger-Teton National Forest, reference area species in seed mixtures, substituting Smith and others (1988) also found that Spatz' simi- introduced species for natives prior to seeding, non- larity indices were low on seeded and nonseeded well establishment or nonpersistence ofseeded native spe- sites compared to adjacent reference areas after 3 to cies, competitive exclusion ofnative recolonizing spe- 33 years. Introduced species are typically seeded on cies by persistent introduced species or annual species, disturbed lands because they exhibit high seedling and changes in soil physical and chemical properties. establishment and rapid growth rates and provide 7 Vegetation Rock/Gravel 100- B B B C A 80- 60 40 C a a 20- c s a. M NM R R M NM R i Bare Ground 5 B C A C oO ioo- 80- 60- M 1 40- 20- 1 m. Basal H M N1M::: a R III Aerial — Figure 2 Comparisons of monitored (M), nonmonitored (NM), and reference (R) sites for the two intensive sampling methods. Values are means ±1 standard error. Uppercase letters indicate significant differences among sites (p < 0.05). Lowercase letters indicate significant differences between methods within sites (p < 0.05). Nonmonitored Nonmonitored Reference Vegetation Litter Rock/ Bare Total Vegetation Litter Rock/ Bare Total Gravel Ground Cover Gravel Ground Cover Cover Category — Figure 3 Comparisons of mean percentage cover of intensively sampled monitored, nonmonitored, reference areas using one, two, and three transects, and one BLM line transect. Values are means ±1 standard error. 8