WEEDY ANNUALS AND ESTABLISHMENT OF SEEDED SPECIES ON A CHAINED JUNIPER-PINYON WOODLAND IN CENTRAL UTAH James N. Davis Kimball T. Harper ABSTRACT United States is grazed. Heavy grazing in this type results in greater than average juniper-pinyon tree densities and invasion of this vegetational type onto a<ljacent plant communities (Aro 1971; Woodbury 1947). Loss of vegetation in the understory due to grazing appears also to have decreased the incidence of wildfires in these woodlands and allowed an increase in tree density (Arnold and Schroeder 1955; Johnsen 1962). An example of the problem that depletion of understory species can cause on many of Utah's juniper-pinyon winter ranges is evidenced by large proportions (up to 50 percent) of some deer herds that were lost during the severe winter of 1949-50. In contrast and during this· same time, juniper-pinyon winter ranges in good condition experienced deer losses that were only slightly higher than those expected in moderate winters (Plummer and others 1968). The severe winter of 1949-50 was not an isolated event; there have been many winters since (including 1964-65, 1972-73, 1978-79, 1979-80, 1983-84, and 1988-89) in which there have been heavy deer losses in many different geographic areas of the State (Jense 1989). In an effort to remedy damage on some pinyon-juniper critical winter ranges, cooperative work between two agencies, the Intermountain Forest and Range Experiment Station (now the Intermountain Research Station), Forest Service, U.S. Department of Agriculture, and the Utah State Department of Fish and Game (now the Utah State Division of Wildlife Resources, hereafter referred to as the Division) was initiated. The objective of the effort was to find plant materials and revegetation methods for artificially restoring forage production on depleted juniper-pinyon winter ranges. Such efforts were considered to have value for watershed protection and for improved livestock grazing and big-game habitat (Plummer and others 1968). Anchor chaining has proven to be the most effective and economical of several techniques tested for tree removal, seedbed preparation, and seeding success (Plummer and others 1964). The anchor chain is particularly effective on juniper-pinyon sites that are undulating, rocky, and sometimes steep. Such sites characteristically have shallow, poorly developed soils that make establishment of seeded species more difficult. Since the beginning of our rehabilitation efforts, the Division has treated more than 50 areas statewide, a total of over 24,000 ha (60,000 acres) on State-owned lands, and has cooperated with the Forest Service, Bureau of Land Management, U.S. Department of the Interior, and private landowners on chaining projects involving an additional Bur buttercup (Ranunculus testiculatus) and cheatgrass brome (Bromus tectorum) occurred in very large numbers on a juniper-pinyon (Juniperus spp.-Pinus spp.)-treated site. Of the herbaceous species planted (nine grasses and seven {orbs), only nine appeared in large enough numbers to occur in any of the 520 sample quadrats. Significant plot-to-plot variation in density of seven ofthe seeded species was explained by the initial densities of either bur buttercup or cheatgrass. Precipitation probably was not responsible for the poor establishment of seeded species, because it was 160 percent ofnormal in that year. Almost without exception, the native and seeded perennial grass species increased over the period of observation. The data suggest that, although annual weeds interfered with initial establishment of the seeded perennials, these species gradually became highly competitive with and strongly reduced the density of both bur buttercup and cheatgrass. INTRODUCTION Juniper-pinyon woodlands dominate almost 30 percent of Utah's land area (West and others 1975), and are estimated to cover from 17.5 to 32.5 million ha (43 to 80.2 million acres) of the western United States (Kuchler 1964; West and others 1975). In Utah, this type occurs primarily between 1,500 and 2,100 m (5,000 and 7,000 ft) elevation, but it is not uncommon for these limits to be transgressed (Woodbury 1947). For example, the woodlands occur as low as 980 m (3,200 ft) near St. George and as high as 2,560 m (8,400 ft) on south-facing slopes on the Book Cliffs in Carbon County. The current low carrying capacity (understory productivity) of this vegetative-type appears to be a consequence of many years of excessive grazing by domestic animals (Forest-Ra:nge Task Force 1972). Clary (1975) estimated that 80 percent of the juniper-pinyon type in the western Paper presented at the Symposium on CheatgTass Invasion, Shrub DieOff, and Other Aspects of Shrub Biology and Management, Las Vegas, NV, April 5-7, 1989. James N. Davis is research biologist for game range restoration studies with the Utah Division of Wildlife Resources stationed at the Shrub Sciences Laboratory, Intermountain Research Station, Forest Service, U.S. Department of Agriculture, Provo, UT 84606; Kimball T. Harper is professor of Botany and Range Sciences, Department of Botany and Range Science, Brigham Young University, Provo, UT 84602. 72 This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. The areas chained on this study site were not continuous, but were small areas of varied shapes ~d topographical relief along ridges, ravines, and in smallvalleys. Since traditional wildlife management has considered "edge" an important wildlife habitat variable, there was a conscious effort in the chaining operation to increase edge effect. The areas were chained and seeded in November of 1982. Elevation ranges from 1,740 m (5,700 ft) on the westerly edge of Black Hill to 2,070 m (6,800 ft) in the mountain brush zone at the head of Cane Valley. The area has an average slope of 13 percent (ranging from 7 to 26). Soils are shallow, 25 to 30 em (about 10 to 12 inches) deep over fractured parent material. Soils on the lower portions of the site are an Am toft flaggy loam series (loamy-skeletal, carbonatic, mesic, Lithic Xerollic Calciorthids); the upper sites are an Atepic-Badland association series (loamy, carbonatic, mesic, shallow, Xerollic Calciorthids) (SCS 1981). The shallow soils support scattered populations of black sagebrush (Artemisia nova), Wyoming big sagebrush (Artemisia tridentata ssp. wyomingensis), and mountain low rabbitbrush (Chrysothamnus viscidiflorus ssp. lanceolatus ). Sampling was done in midsummer along 10 permanently marked lines, each 270 m (about 900 ft) in length. The lines were located and read in 1982 before the area was chained. Lines were relocated after chaining and remarked for easier location and were read for the next 3 years. Each line consisted of five transects 30m (almost 100ft) long, except line 6, which had seven transects because of vegetational diversity along its length. Transects alternated with 30-m segments that were not inventoried. Quadrats (1.0 m 2 ) were placed at 3-m intervals along each transect beginning at the 0 point and alternating from the right to the left sides of the survey tape. This allowed placement of at least 50 quadrats along each of the 10 lines distributed throughout the chained area. Cover was determined within each 1.0-m2 quadrat for each species using a procedure slightly modified from that described by Daubenmire (1959). The modification consisted of adding one extra cover class with limits ofO to 1 percent. The modification provided a more accurate estimate of cover for small or subordinate species. Plant densities for grasses and forbs were determined by counting individuals rooted within the 1.0-m2 quadrats. Deer presence was determined before and after treatment with each of the l.O-m2 quadrats that contained recent pellet groups (Ferguson 1955; Neff 1968; Wallmo and others 1962). Shrub densities were estimated along a 0.005-ha strip plot centered on the 30-m survey tape (1.67 m wide). Frequencies for both forbs and grasses were based on species presence within any quadrat along each of the 52 transects. Frequencies for shrubs were based on occurrence within each of the 52 0.005-ha strip plots on each transect. Tree densities were determined with the quarter method (Cottam and Curtis 1956). Points for the quarter method were located at the beginning and end of each transect. This gave 40 point-to-tree distances per line. A soil penetrometer was used to estimate depth to obstructions in the soil at random points (Ostler and others 1982). Most obstructions were stones. Five depth measurements were taken at each end of each transect along each line. This gave 50 measurements per line. 84,000 ha (207,000 acres) of big-game winter range (Fairchild 1982). There are divergent views regarding whether such woodlands should be chained and seeded (Dalen and Snyder 1987; West 1984) or whether suitable sites should be treated at all (Gifford 1987; Lanner 1981), but these efforts have almost always enhanced soil stability, forage production, and habitat recovery. Typically, the success of juniper-pinyon treatments has been reported in terms of additional forage made available by tree removal. For example, Phillips (1977) considered 59 such projects initiated in the Forest Service's Intermountain Region between 1954 and 1975. He reported that on average, bare ground decreased by 11 percent and forage production (air-dry weight) increased from 46 to 320 kglha (100 to 710 lbs/acre). Clary and Jameson (1981) reported even greater increased production in Arizona. Typically, the woodlands in New Mexico and eastern Arizona receive more warm-season precipitation than areas in Nevada and northwestern Utah and have an inherently greater production potential. Other research has identified the species best adapted to specific treatment areas (Johnsen and Gomm 1981; Jordan 1981; Judd 1966; Monsen 1987; Plummer and others 1968; Renney 1972; Springfield 1965). Planting mixtures for seeding juniper-pinyon removal areas produce artificial plant associations in which species may not be fully compatible with each other or with resident native plants. Furthermore, it is difficult to produce and maintain a stand of specified composition, because each species responds differently to the natural and imposed environmental factors that affect seedling establishment and subsequent competitiveness. The species that are best adapted to initial conditions favorable for germination tend to dominate rehabilitated plant communities; seeds of species that are less adapted fail to germinate or germinate late, and the seedlings are suppressed. Nevertheless, mixtures of species are almost universally seeded on juniper-pinyon chained areas, because environmental conditions in both space and time cannot be accurately predicted. With several species in a mixture there is less chance of failure. In this study, we evaluate initial establishment of seeded species on a chained juniper-pinyon site in central Utah (Sanpete County) with high densities of two weedy annuals: cheatgrass brome (Bromus tectorum) and bur buttercup (Ranunculus testiculatus). LOCATION AND METHODS The treated area 4s about 300 ha (750 acres) in size, and is located about 4 km (2.5 mi) northeast of Ephraim, UT. Most af the lower study area was privately owned before purchase by the Division; the upper reaches of the treatment area included lands managed by the Forest Service. Such lands are rehabilitated primarily for wildlife by the Division, because they lie within geographic areas considered to be critical winter range or in areas experiencing wildlife depredation problems. Many, if not most, of the lands purchased for rehabilitation have had a long history of overgrazing and therefore support communities with dense populations of weedy species. The weeds are often alien to the region and can interfere greatly with establishment of seeded species. 73 Statistics Analysis System (SAS 1985) was used for data analysis. Multiple regression was used to determine how seeding rate and density of naturally occurring plant species affected establishment of seeded species. Initially, the power of various subsets of the independent variables for predicting species response was evaluated by examination of coefficients of multiple determination (R2) and magnitude ofC(p) values (measure of total squared error). Intimately, forward selection of independent variables for multipleregression models was used and vif (variance inflation factors) and eigenvalues were employed to evaluate the effectiveness of the predictive models. Plant nomenclature follows Plummer and others (1977), except for Lewis flax (Linum perenne ssp. lewisii), which follows Welsh and others (1987). RESULTS AND DISCUSSION Chaining reduced Utah juniper (Juniperus osteosperma) tree density on the site from 2,230 to 186 treeslha (902 to 75 trees/acre) 3 years after treatment (92 percent reduction). Pinyon (Pinus edulis) tree density declined from 627 to 62 per ha (254 to 25 per acre) for a decrease of about 90 percent. Average tree kill with cabling or chaining normally ranges from 40 to 80 percent (Arnold and others 1964; Aro 1971, 1975). Thus, the anchor chain was unusually effective in removing the trees from the ground at the beginning of this study because moisture conditions favored uprooting of the trees and because trees were old and even aged. Precipitation near the study area during the first year after seeding was unusually high. In the treatment and establishment water year (October 1, 1982, to September 30, 1983), there was 437 mm (about 17.5 inches) of precipitation. Normal precipitation for the area is 272 mm (10.9 inches) a year. This above-normal precipitation pattern continued through 1986 (1983-84, 454 mm or 17.9 inches; 1984-85, 372 mm or 14.6 inches) (NOAA 1983-86). In the months of initial establishment (March-May 1983), temperatures were above normal for March, then abnormally cool for April and May. Total shrub density, irrespective of species, increased steadily throughout the period 1983-85 rising from 9,570 to 11,880 plants per ha (table 1). Shrub decadence (plants with ~25 percent of the crown dead) declined dramatically from over 20 percent in the pretreatment period of 1982 to less than 2 percent in the summer of 1985. Shrub seedlings (surviving established plants ignored) were 77 percent more numerous after treatment than they had been in the pretreatment sample. Grass and forb cover increased 26 and 40 percent respectively relative to pretreatment conditions by summer 1985. Bare ground, which averaged 4 7 percent before treatment, decreased to 14 percent by the summer of 1985. Percent rock cover showed essentially no change (8 to 9 percent) over the course of study. Litter was estimated at 17 percent before treatment, increased to 26 percent for 2 years following treatment, and then declined to 15 percent in 1985. Twenty-six species were seeded onto the area. Species of grasses (9), forbs (7), and shrubs (10) were included. Of the species planted, only nine appeared as established seedlings in any of the 520 1.0-m2 sample quadrats. This was not expected since precipitation was well above normal in the winter of 1982-83 and during the growing season of 1983 following the late fall seeding of the site. Of the nine perennial grasses sown, only four established in sufficient numbers to occur in the sampling quadrats (fairway wheatgrass [Agropyron cristatum], intermediate wheatgrass [Agropyron intermedium], bluebunch wheatgrass [Agropyron spicatum], and orchardgrass [Dactylis glomerata]) (tables 2 and 3). These four species combined were represented in 1983 (the year of establishment) with an average of about 7.6 seedlings/m2 even though approximately 359 perennial grass seeds were sown per square meter. The seed to seedling establishment rate in the first year was 2.1 percent. Considering only the four species that did establish seedlings, the establishment rate was 0.4, 12.4, 0.3, and 0.03 percent for fairway, intermediate, and bluebunch wheatgrasses and orchardgrass, respectively. Table 1-Average shrub density (plantslha) across the entire treated area for the 3 years following seeding. Densities are based on actual counts within strips 1.67 m wide by 30 m long centered on the 30-m transect (0.005 ha) Year Plant species Artemisia novel A. tridentata ssp. wyomingensis2 A triplex canescens2 Ceratoides lanatcl Chrysothamnus nauseosus2 ssp. albicaulis C. viscidiflorus ssp. lanceo/atus Gutierrezia sarothrae Leptodactylon pungens Opuntia species 19821 1983 1984 1985 Black sagebrush Wyoming big sagebrush Fourwing saltbush Common winterfat White rubber rabbitbrush 320 220 50 660 60 230 130 100 870 70 240 140 100 810 100 470 130 90 860 210 Mountain low rabbitbrush 2,910 2,830 2,500 2,280 Broom snakeweed Prickly phlox Prickly pear 1,940 2,760 1,200 2,800 1,480 1,060 5,520 1,360 1,890 4,480 1,120 2,260 Common name 1Counts 2 done before treatment. Seeded species. 74 Table 2-Average grass and forb plant density (plants/ha) across the entire treated area for the 3 years following seeding in November 1982. Densities are based on actual counts in 520 permanently marked 1.0-m 2 quadrats. The quadrats were read in mid-summer (about July) Year Species 1983 Common name 1984 1985 Grasses Agropyron cristatum 1 Agropyron intermedium1 Agropyron spicatum1 Bromus tectorum Dactylis glomerata1 Oryzopsis hymenoides Poa secunda Sitanion hystrix Stipa comata Fairway wheatgrass Intermediate wheatgrass Bearded bluebunch wheatgrass Cheatgrass brome Orchardg rass Indian ricegrass Sandberg bluegrass Bottlebrush squirreltail Needle-and-thread grass 780 74,570 160 1'130,420 0 9,080 1,090 12,420 10,450 690 110,090 90 229,560 650 11,830 1,310 7,320 10,410 1,660 71,420 660 140,720 300 17,740 5,260 17,700 14,070 Kings sandwort Musk bristle thistle Douglas chaenactis Lambsquarters goosefoot Cryptantha Prickly lettuce Lewis flax Narrowleaf gromwell Yellow sweetclover Ladak alfalfa Common sainfoin Hoods phlox Bur buttercup Small burnet Tumble mustard Common dandelion Yellow salsify 3,380 190 690 440 1,190 3,630 1,600 440 80 3,840 2,750 6,200 1,876,690 14,460 420 1,080 580 3,900 710 250 500 940 116,560 750 310 1,330 1,480 2,250 6,050 253,900 13,000 1,710 80 1,020 3,570 6,680 120 0 850 82,770 30 4,780 190 1,250 700 7,220 285,320 14,070 2,050 130 4,210 Forbs Arenaria kingii Carduus nutans Chaenactis douglasii Chenopodium album Cryptantha humilis Lactuca serriola Unum perenne ssp. lewisii 1 Lithospermum incisum Melilotus officinalis1 Medicago sativa1 Onobrychis viciifolia1 Phlox hoodii Ranunculus testiculatus Sanguisorba minor1 Sisymbrium altissimum Taraxacum officinale Tragopogon dubius 1Seeded species. Five of the six seeded perennial forbs were included in the sample of seedlings established in 1983 (tables 2 and 3). Cicer milkvetch (Astragalus cicer) failed to establish any seedlings despite an average seeding rate of about 386 seeds/m2• It was only seeded on the four upper lines (or sites) because it usually needs more moisture for establishment and the lower sites were thought to be marginal for its establishment and growth. Establishment rates (percent of seeds producing seedlings) for the other five forbs were 0.02, 0.13, 0.06, 0.55, and 0.39 percent for Lewis flax (Linum perenne ssp. lewisii), Ladak alfalfa (Medicago sativa), yellow sweetclover (Melilotus officinalis), common sainfoin (Onobrychis viciifolia), and small burnet (Sanguisorba minor), respectively. Grass and forb species densities and frequencies are summarized in tables 2 and 3 respectively for the 3 years following seeding. Cheatgrass brome, the weedy species generally considered to be the major competitor with seeded species for space in see dings in the juniper-pinyon zone (Krebs 1972), was represented by over 1.1 million plantslha in 1983. Cheatgrass density had declined 88 percent by the summer of the third year. The density of seeded species increased over the same period. This suggests, as did Stewart and Hull's (1949) work in southern Idaho, that seeded perennial species were offering considerable competition to cheatgrass by the third growing season and had forced a large reduction in its density, but cheatgrass still persisted in low numbers on most sites. All other grasses (both seeded and native perennials) except intermediate wheatgrass increased in density during the 1983-85 period. Despite the fact that intermediate wheatgrass density declined over the period of record, it is still present in greater numbers than all other perennial grasses combined (71,420 versus 57,400 plantslha). Three perennial grasses, bluebunch wheatgrass, orchardgrass, and Sandberg bluegrass (Poa secunda), had more than tripled their 1983 density by the 1985 growing season (table 2). Fairway wheatgrass and Indian ricegrass (Oryzopsis hymenoides) densities had increased by 112 and 95 percent respectively by the third growing season. Needle-and-thread grass (Stipa comata) and bottlebrush squirreltail (Sitanion hystrix) densities increased 35 and 43 percent respectively in that interval. By 1985, density of well-established perennial grass individuals averaged 131m2 across the entire treatment area. Seeded species accounted for only about 57 percent of the established individuals in the 1985 growing season. N onseeded perennial grasses increased their density by 75 Table 3-Grass and forb frequency across the entire treated area for the 3 years following seeding. Frequencies are based on actual presence within quadrats along each of the 52 transects Species 1983 Year 1984 1985 Fairway wheatgrass Intermediate wheatgrass Bearded bluebunch wheatgrass Cheatgrass brome Orchardgrass Indian ricegrass Sandberg bluegrass Bottlebrush squirreltail Needle-and-thread grass 16 65 3 87 0 73 33 86 35 31 73 9 85 20 75 39 73 25 47 83 22 96 20 81 46 81 31 Kings sandwort Musk bristle thistle Douglas chaenactis Lambsquarters goosefoot Cryptantha Prickly lettuce Lewis flax Narrowleaf gromwell Yellow sweetclover Ladak alfalfa Common sainfoin Hoods phlox Bur buttercup Small burnet Tumble mustard Common dandelion Yellow salsify 40 10 25 21 17 73 50 13 4 39 60 31 98 71 15 29 37 33 17 10 19 15 87 22 12 35 27 30 23 87 63 25 8 42 25 23 8 0 12 87 3 23 4 19 25 37 92 62 33 10 50 Common name Grasses Agropyron cristatum1 Agropyron intermedium1 Agropyron spicatum1 Bromus tectorum Dactylis glomerata1 Oryzopsis hymenoides Poa secunda Sitanion hystrix Stipa comata Forbs Arenaria kingii Carduus nutans Chaenactis douglasii Chenopodium album Cryptantha humilis Lactuca serriola Linum perenne ssp. lewisii 1 Lithospermum incisum Melilotus officinalis1 Medicago sativa1 Onobrychis viciifolia1 Phlox hoodii Ranuncu/us testiculatus Sanguisorba minor1 Sisymbrium a/tissimum Taraxacum officinale Tragopogon dubius 1Seeded species. a total of 66 percent (3.3 to 5.5 plants/m2 ) during the observation period. Native grasses and other understory species are usually assumed to decline in density after a chaining treatment followed by seeding (Tausch and Tueller 1977), but in this study the native grasses increased strongly. The native grasses have become an important part of the plant community now existing on the treatment area. Frequency data show that two of the native, nonseeded perennial grass species (squirreltail and needle-and-thread) have not spread to new areas on the treated site, but two other nonseeded perennial grasses (Indian ricegrass and Sandberg bluegrass) appear to have colonized new sites (they had higher frequency values in 1985 than in 1983). By 1985, all of the seeded perennial grass species appeared on sites not occupied in the 1983 growing season (table 3). That apparent range expansion could represent dispersal of seed into previously unoccupied quadrats, merely delayed germination, or slow development of seedlings that were too small to be detected in 1983. Forb densities and frequencies are also summarized in tables 2 and 3. Almost all perennial native forbs increased in density between 1983 and 1985. Two perennial native forbs, cryptantha (Cryptantha humilis) and dandelion (Taraxacum officinale) declined during this same interval. Musk thistle (Carduus nutans), prickly lettuce (Lactuca serriola), tumble mustard (Sisymbrium altissimum), and yellow salsify (Tragopogon dubius ), all introduced annuals or biennials, increased dramatically in density and frequency during the 1983-85 interval. Lambsquarters (Chenopodium album), and bur buttercup (Ranunculus testiculatus) both decreased in density and frequency. Indications are that these species are not able to compete with the species they are now forced to associate with. It is often assumed that perennial species, once established, will crowd out weedy annuals through time and in the absence of heavy grazing (Stewart and Hull 1949). Our data suggest that this is only partially true, since some introduced annuals and biennials steadily increased in the seeded area. Prickly lettuce, whose life cycle varies from annual or biennial to a short-lived perennial in the study area, experienced a large increase in density between 1983 and 1985. Density increased from 3,630 to 82,770 plantslha, while its average frequency increased from 73 to 87. Prickly lettuce was heavily grazed throughout its period of active growth. The species appeared to function as a short-lived perennial during the study period, a time of heavy animal use and above normal precipitation. Harper (1977) noted that high densities and severe competition cause Lactuca to function in this way. 76 Table 4-The effect of several independent variables on establishment of seedlings of selected species. The analysis is based on seedlings observed in the first sampling (July 1983} after treatment. The density of competing species was taken concurrently with dependent variable species in the same period lnde~endent varlable1 Seeded species 1 2 3 5 4 6 7 8 9 R2 Total Prob>f 0.86 .66 .87 .97 0.14 .13 .13 .04 - - - - - - - - - - - - - - - Contribution to R 2 value - - - - - - - - - - - - - Agropyron cristatum Agropyron intermedium Agropyron spicatum Dacty/is glomerata Linum perenne ssp. lewisii Melilotus officina/is Medicago sativa Onobrychis viciifolia Sanguisorba minor 2().38" 0.48" 0.19" 0.26" 0.23 0.24 0.61" .97" .91" .41" 0.09 .32" .25 .70 .04" .20" .19 .18 .54" 0.19 1.00 .77 .63 .89 .73 .002 .09 .15 .16 .14 1 1-Bur buttercup density, 2-Seed planted per square meter of the dependent variable species, 3-Total shrub density, 4-lntermediate wheatgrass density, 5-Mountain low rabbitbrush density, 6-Bottlebrush squirreltail density, 7-Cheatgrass brome density, 8-Total native grass density (includes: bottlebrush squirreltail, Indian ricegrass, and needle-and-thread grass), 9-lndian ricegrass density. 2 Superscript n shows the relationship is negative. The annual, bur buttercup, occurred in very large numbers in 1983 (1,876,690 plantslha or 190 plants/m2), but decreased dramatically over the 3 years of record to 283,320 plants/ha or 28 plants/m2, a decrease of 85 percent (table 2). The seeded forbs were all in decline in both density and frequency by the third growing season after seeding (table 2). Lewis flax decreased more in relative terms than any other seeded species during the first 3 years of study. Its density decreased by 98 percent (1,600 to 30 plants/ha), while its frequency decreased from 50 to 3 percent. Sainfoin also declined dramatically over the period 1983-85 (tables 2 and 3). Ladak alfalfa also declined, but its losses were smaller than those experienced by Lewis flax and sainfoin. Small burnet is the most persistent of the seeded forbs in this study by a large margin; it declined only 3 percent (tables 2 and 3). Performance of yellow sweetclover, a biennial, is more difficult to assess. Its numbers were low the first year after seeding (80 plants/ha), but that number increased to 1,330/ha in 1984 and then declined again to 190 plants/ha in 1985. Frequency data for yellow sweetclover reflected similar trends, from an initial value of 4 percent, to an intermediate value of 35 percent, and a 1985 value of 4 percent again. Precipitation is one of the most common factors blamed for establishment failures (Vallentine 1980), yet there was more than adequate precipitation during the establishment period. With almost 7 inches of above-normal precipitation in the year of establishment and above-normal precipitation for the next 2 years, lack of water should not have been a major limiting factor for establishment. Winter frost heaving also is not likely to have been an unusually serious problem for establishing seedlings at the site because snow cover was far above normal during that study period. Temperature conditions that may have destroyed new seedlings were probably not the warmer temperatures for March, which would have promoted germination, but the low temperatures in that same month. A minimum temperature 77 of -6.1 oc (21 °F) was recorded on March 20 and 23. That tern perature could have been lethal for some seedlings. The cold spell started in midmonth and continued with minimum temperatures (Fahrenheit) in the low- and midtwenties for 12 days (NOAA 1983). Therefore, the late March interval of severe cold could have had an important effect on the mortality of some species' seedlings. This would have in tum depended on how soon each species germinated and how developed they were at the time of the cold temperatures. It is possible also that the poor initial establishment observed for seeded species (some never became established) was related to interaction with the weed species (table 4). Buchanan and others (1978) reported that bur buttercup tissue exerted an allelopathic effect in in vitro experiments on germination and seedling development of many range grasses, such as tall wheatgrass (A elongatum), fairway wheatgrass, Russian wildrye (Elymusjunceus), and intermediate wheatgrass. Average bur buttercup density on the chained area was 190 plants/m 2 during the year of establishment. Such high densities of this weed appear to have adversely affected the establishment of five of the seeded species (table 4). This effect may well have been enhanced with the elevated levels of precipitation during the year of seeding. For example, since the allelopathic agent in bur buttercup is water soluble (Buchanan and others 1978), and it is known that the intensity and volume of rain influence the efficiency ofleaching, larger amounts of precipitation would have leached more allelopathic agents into the soil, where they may have exerted an adverse effect on germination and growth of seeded species (Tukey 1971). Normally, fairway wheatgrass is the dominant grass in seeding efforts in areas comparable to that studied (DePuit 1986; Rogier and Lorenz 1983). In this seeding, fairway wheatgrass was unimportant even though an average of over 42 seeds were sown per square meter. Conditions for establishment also appeared to be unfavorable for many other seeded species, since only nine of 16 grass and forb species seeded established in the quadrats inventoried. In only one case (common sainfoin) is there any strong indication that increased seeding rates would have improved establishment success (table 4). Reservation. Station paper No. 18. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station. 35 p. Aro, R. S. 1971. Evaluation of pinyon-juniper conversion to grassland. Journal of Range Management. 24: 188-197. Aro, R. S. 1975. Pinyon-juniper woodland manipulations with mechanical methods. In: Gifford, G. F.; Busby, F. E., eds. The pinyon-juniper ecosystem. Logan, UT: Utah State University: 67-76. Bradley, B. 1989. [Personal communication]. Data on file at: Utah Division of Wildlife Resources, Salt Lake City, UT. Buchanan, B. A.; Harper, K T.; Frischknecht, N. C. 1978. Allelopathic effects of bur buttercup tissue on germination and growth of various grasses and forbs in vitro and in soil. Great Basin Naturalist. 38: 90-96. Clary, W. P. 1975. Present and future multiple use demands on the pinyon-juniper type. In: Gifford, G. F.; Busby, F. E., eds. The pinyon-juniper ecosystem. Logan, UT: Utah state University: 19-24. Clary, W. P.; Jameson, D. A. 1981. Herbage production following tree and shrub removal in the pinyon-juniper type of Arizona. Journal of Range Management. 34: 109-113. Cottam, G.; Curtis, J. 1956. The use of distance measures in phytosociological sampling. Ecology. 37: 451-460. Dalen, R. S.; Snyder, W. R. 1987. Economic and social aspects of pinyon-juniper treatment-then and now. In: Everett, R. L., compiler. Proceedings, pinyon-juniper conference, 1986 January 13-16; Reno, NV. Gen. Tech. Rep. INT-215. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 343-350. Daubenmire, R. 1959. A canopy-coverage method of vegetational analysis. Northwest Science. 33: 43-66. DePuit, J. 1986. The role of crested wheatgrass in reclamation of drastically disturbed lands. In: Johnson, K. L., ed. Crested wheatgrass: its values, problems and myths; symposium proceedings. Logan, UT: Utah State University: 323-330. Fairchild, J. 1982. Wildlife habitat management on pinyonjuniper chainings. In: Johnson, K. L., ed. Proceedings of the second Utah shrub ecology workshop. Logan, UT: Utah State University, College of Natural Resources: 19-20. Ferguson, R. B. 1955. The weathering and persistency of pellet groups as it affects the pellet group count method of censusing mule deer. Utah Academy of Sciences, Arts and Letters Proceedings. 32: 59-64. Forest-Range Task Force. 1972. The nation's range resources, a forest-range environmental study. For. Resour. Rep. 19. Washington, DC: U.S. Department of Agriculture, Forest Service. 147 p. Gifford, G. F. 1987. Myths and fables and the pinyonjuniper-type. In: Everett, Richard L., compiler. Proceedings, pinyon-juniper conference; 1986 January 13-6; Reno, NV. Gen. Tech. Rep. INT-215. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 34-37. Harper, J. L. 1977. Population biology of plants. New York: Academic Press. 892 p. MANAGEMENT IMPLICATIONS AND CONCLUSIONS On this site, had not the seeding mixture contained intermediate wheatgrass and small burnet, this treatment may have been considered a failure. These two species contributed 94 percent of the total seeded population of grasses and forbs 3 years after treatment. Previous experience would not likely have enabled anyone to predict which species would become dominant and which would fail to establish significant numbers of seedlings even if the unusually favorable moisture conditions could have been predicted. This site and the resulting seeded population also prompt one to ponder the criteria one should use to rate the success of a revegetation project. Usually production of the seeded species is the only measure used. An alternative measure of success to be considered is increased onsite use by wildlife and an accompanying reduction in depredation of nearby agricultural lands, which in some areas is a major problem. Spring range ride transect counts in the vicinity of the study area and before the chaining usually yielded more than 100 dead deer for areas both to the north and south of the chaining. No dead deer have been observed on or in the immediate proximity of the chaining since treatment, evert though deer pellet group distribution patterns have ~ot changed significantly since treatment. Deer depredation calls once numbered more than 200 per winter-spring season; after treatment, they were reduced to two to five calls a year for areas near the chaining (Bradley 1989). In this respect, the chaining would have to be considered very successful, even though many revegetation efforts in similar ecological situations have resulted in far denser populations and greater production from seeded species (Aro 1971; Clary and Jameson 1981; Phillips 1977). ACKNOWLEDGMENTS Federal funds for wildlife restoration were provided through Pittman-Robertson Project W-82-R, studies 2, 4, and 6. Work was done in cooperation with the Utah State Division of Wildlife Resources and the Intermountain Research Station, Forest Service, U.S. Department of Agriculture. REFERENCES Arnold, J. F.; Jameson, D. A.; Reid, E. H. 1964. The pinyon-juniper type of Arizona: effects of grazing, fire and tree control. Prod. Res. Rep. Washington, DC: U.S. Department of Agriculture. 84 p. Arnold, J. F.; Schroeder, W. L. 1955. Juniper control increases forage production on Fort Apache Indian 78 '' .. ,·. Plummer, A P.; Monsen, S. B.; Stevens, R.1977. Intermountain range plant names and symbols. Gen. Tech. Rep. INT-38. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station. 82 p. Renney, C. W. 1972. Reseeding in the pinyon-juniper vegetation type of northern Arizona. Tech. Note Range 54. Phoenix, AZ: U.S. Department of Agriculture, Soil Conservation Service. 6 p. SAS. 1985. SAS user's guide: Statistics, version 5 ed. Cary, NC: SAS Institute Inc. 956 p. Springfield, H. W. 1965. Adaptability of forage species for pinyon-juniper sites in New Mexico. Res. Note RM-57. Fort Collins, CO: U.S. Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station. 4 p. Stewart, G.; Hull, A C. 1949. Cheatgrass (Bromus tectorum L.)-an ecologic intruder in southern Idaho. Ecology. 30: 58-74. Tausch, R. J.; Tueller, P. T. 1977. Plant succession following chaining of pinyon-juniper woodlands in eastern Nevada. Journal of Range Management. 30: 44-49. Tukey, H. B. 1971. Leaching of substances from plants. In: U.S. National Committee for IBP, eds. Biochemical interactions among plants. Washington, DC: National Academy of Science: 25-32. Soil Conservation Service. 1981. Soil survey of Sanpete Valley area, Utah, Parts of Utah and Sanpete counties. Washington, DC: U.S. Department of Agriculture, Soil Conservation Service; U.S. Department of the Interior, Bureau of Land Management, in cooperation with Utah State University, Agricultural Experiment Station and Utah State Department ofWildlife Resources. 179 p. Vallentine, J. F. 1980. Range development and improvements. Provo, UT: Brigham Young University Press. 545p. Wallmo, 0. C.; Jackson, A W.; Hailey, T. L.; Carlisle, R. L. 1962. Influence of rain on the count of deer pellet groups. Journal ofWildlife Management. 26: 50-55. Welsh, S. L.; Atwood, N.D.; Higgins, L. C.; Goodrich, S. 1987. A Utah flora. Great Basin Naturalist Memoir No.9. 894p. West, N. E. 1984. Factors affecting treatment success in the pinyon-juniper type. In: Johnson, K. L., ed. Proceedings, second Utah shrub ecology workshop. Logan, UT: Utah State University: 21-33. West, N. E.; Rea, K. H.; Tausch, R. J. 1975. Basic synecological relationships in juniper-pinyon woodlands. In: Gifford, G. F.; Busby, F. E., eds. The pinyon-juniper ecosystem. Logan, UT: Utah State University: 41-53. Woodbury, AM. 1947. Distribution of pigmy conifers in Utah and Northeastern Arizona. Ecology. 28: 113-126. Jense, Grant. 1989. [Personal communication]. Data on file at: Office of Assistant Chief of Game Management, Utah Division of Wildlife Resources, Salt Lake City, UT. Johnsen, T. N.; Gomm, F. B. 1981. Forage plantings on six Arizona pinyon-juniper subtypes. Journal of Range Management. 34: 131-136. Johnsen, T. N. 1962. One-seeded juniper invasion of northern Arizona grasslands. Ecological Monographs. 32: 187-207. Jordan, G. L. 1981. Range seeding and brush management on Arizona rangelands. Bull. T81121. Tucson, AZ: University of Arizona, College of Agriculture. 88 p. Judd, I. 1966. Range reseeding success on the Tonto National Forest, Arizona. Journal of Range Management. 19: 296-301. Krebs, C. J. 1972. Ecology-the experimental analysis of distribution and abundance. New York: Harper & Row. 694p. Kuchler, A. W. 1964. Potential vegetation in the conterminous United States. Spec. Publ. 36. New York: American Geographical Society. 111 p. Lanner, R. M. 1981. The pinon pine, a natural and cultural history. Reno, NV: University of Nevada Press. 208 p. Monsen, S. B. 1987. Shrub selections for pinyon-juniper plantings. In: Everett, R. L., compiler. Proceedingspinyon-juniper conference; 1986 January 13-16; Reno, NV. Gen. Tech. Rep. INT-215. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Research Station: 316-329. Neff, D. J. 1968. The pellet-group count technique for big game trend, census, and distribution: a review. Journal ofWildlife Management. 32: 597-614. National Oceanic and Atmospheric Administration. 1983-86. Utah climatological data annual summary (1983-1986). Asheville, NC: U.S. Department of Commerce, National Oceanic and Atmospheric Administration; National Climatic Data Center. Ostler, W. K.; Harper, K. T.; McKnight, K. B.; Anderson, D. C. 1982. The effects of increasing snow pack on a subalpine meadow in the Uinta Mountains, Utah, USA. Arctic and Alpine Research. 14: 203-214. Phillips, T. A. 1977. An analysis of some Forest Service pinyon-juniper chaining projects in Region 4, 1954-1975. Range Imp. Notes. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Region. Plummer, A. P.; Christensen, D. R.; Monsen, S. B. 1964. Job completion report for revegetation project W-82-R-9. Info. Bull. 64-14. Salt Lake City, UT: Utah State Department of Fish and Game. 60 p. Plummer, A. P.; Christensen, D. R.; Monsen, S. B. 1968. Restoring big game range in Utah. Publ. No. 68-3. Salt Lake City, UT: Utah Division ofFish and Game. 183 p. 79