https://cpw.cvlcollections.org/files/original/f6a12defe1a57dbcf270436a59600ab6.pdf db99cc693ed6a3d346d3c5eab9b6cfb8 PDF Text Text | Revised: 9 November 2022 | Accepted: 11 November 2022 RESEARCH ARTICLE Plant and mule deer responses to pinyon‐juniper removal by three mechanical methods Danielle Bilyeu Johnston1 | Charles R. Anderson Jr.2 1 Colorado Parks and Wildlife, 711 Independent Avenue, Grand Junction, CO 81505, USA Abstract 2 Colorado Parks and Wildlife, 317 Prospect Avenue, Fort Collins, CO 80526, USA succession by removing pinyon (Pinus spp.) and juniper (Juniperus Correspondence livestock. Because prescribed fire carries inherent risks, mechan- Danielle Bilyeu Johnston, Colorado Parks and Wildlife, 711 Independent Avenue, Grand Junction, CO 81505, USA. Email: Danielle.bilyeu@state.co.us ical methods such as chaining, roller‐chopping, and mastication Funding information Colorado State Severance Tax; White River Field Office of Bureau of Land Management; XTO Energy; Federal Aid in Wildlife Restoration, Grant/Award Number: W‐185‐R; Shell Exploration and Production Company; Encana; Colorado Mule Deer Association; Colorado Mule Deer Foundation; Marathon; ExxonMobil Foundation; Williams Production; Colorado Oil and Gas Conservation Commission Land managers in western North America often reverse spp.) trees to reduce fire risk and increase forage for wildlife and are often used. Mechanical methods differ in cost and the size of woody debris produced, and may differentially impact plant and animal responses. We implemented a randomized, complete‐ block, split‐plot experiment in December 2011 in the Piceance Basin, northwestern Colorado, USA, to compare mechanical methods and to explore seeding (subplot) interactions. We assessed vegetation 1‐, 2‐, 5‐, and 6‐years post‐treatment, and mule deer (Odocoileus hemionus) response via GPS locations 3–8 years post‐treatment. By 2016, treated plots had 3–5 times higher perennial grass cover and ~10 times higher cheatgrass (Bromus tectorum) cover than untreated control plots. Roller‐ chopped plots had both the highest non‐native annual forb cover, and when seeded, the highest density of bitterbrush (Purshia tridentata), a nutritious shrub used by mule deer. Masticated plots had higher bitterbrush use during summer and fall, leaving less forage available for winter. Days of winter mule deer use from GPS point locations in chained and roller‐ chopped plots was ~70% higher than in control plots, while winter use in masticated plots was similar to control plots. Mule deer use appears related to a combination of hiding cover, resulting from residual woody debris, and winter forage This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2022 The Authors. Wildlife Society Bulletin published by Wiley Periodicals LLC on behalf of The Wildlife Society. Wildlife Society Bulletin 2023;e1421. https://doi.org/10.1002/wsb.1421 wileyonlinelibrary.com/journal/wsb | 1 of 21 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Received: 14 October 2021 DOI: 10.1002/wsb.1421 �| JOHNSTON and ANDERSON availability. Roller‐chopped plots provide the best combination of hiding cover and winter forage, but mastication or chaining, applied leaving dispersed security cover, may be better options at large scales or when invasive species concerns exist. KEYWORDS bitterbrush, chaining, mastication, mule deer, Odocoileus hemionus, pinyon‐juniper habitat, Purshia tridentata, roller‐chopping, security cover, utilization Over the past 150 years, woody plants have encroached on grasslands in ecosystems around the world, altering the services grasslands provide (Romme et al. 2009, Veldman et al. 2015, Ledesma et al. 2018). High shrub or tree cover can reduce the productivity of understory species, decreasing available forage for livestock and wildlife (Miller et al. 2000). In western North America, the pinyon (Pinus spp.) and juniper (Juniperus spp.) habitat type (hereafter pinyon‐ juniper) has expanded over the past 150 years (Soule and Knapp 1999, Miller et al. 2005). Loss of palatable understory forage has prompted land managers to remove trees to benefit large herbivores such as cattle (Bos taurus), mule deer (Odocoileus hemionus), and elk (Cervus canadensis; Howard et al. 1987, Skousen et al. 1989, Monaco and Gunnell 2020). Treatments have generally been successful in removing woody vegetation in the short term (Skousen et al. 1989, Ernst‐Brock et al. 2019), and increasing understory vegetation (Skousen et al. 1989, Owen et al. 2009, Coop et al. 2017, Rubin and Roybal 2018, Chambers et al. 2021). However, achieving the desired plant community and ensuring benefit to target wildlife species is more challenging. Following woody plant removal, sites can be invaded by exotic, undesirable vegetation (Bates et al. 2005, Coop et al. 2017, Havrilla et al. 2017, Monaco and Gunnell 2020). Furthermore, target wildlife species may not use sites for reasons unrelated to forage quality, such as predation risk (Bender et al. 2007), human disturbance (Dwinnell et al. 2019), proximity to other vegetation types (Watkins et al. 2007), or competition from non‐target species (Coe et al. 2004). Tree removal efforts for mule deer must account for the timing of mule deer nutritional requirements. Mule deer preferentially utilize a combination of grasses, forbs and shrubs throughout the year (Carpenter et al. 1979, Watkins et al. 2007). As snow deepens in winter, however, mule deer often rely on shrubs as herbaceous plants become less accessible (Kufeld 1973, Carpenter et al. 1979). In winter, mule deer often face a negative energy balance, although higher quality winter forage can reduce rates of winter weight loss (Watkins et al. 2007). Quality winter forage can also increase fawn survival the following year (Bishop et al. 2009, Bergman et al. 2014). We note that herbivores tend to browse the most palatable plants and plant parts first (Armstrong and Macdonald 1992), so herbivory during the growing season can lessen habitat quality the following winter (Jensen et al. 1972, Austin 2000). Thus, the quality and quantity of forage on winter ranges, especially palatable shrubs, can be a limiting factor for mule deer (Wallmo et al. 1977). High tree cover can crowd out understory grasses, forbs, and shrubs, depriving mule deer of essential resources (Howard et al. 1987, Bishop et al. 2009). However, an absence of trees can also be detrimental because mule deer require security cover (Watkins et al. 2007). The importance of security cover is so great that mule deer habitat selection has been shown to monotonically decline with the distance to the edge of tree cover (Northrup et al. 2021). Mule deer will select for pinyon‐juniper, even if having a high degree of it in their home range results in poor body condition (Bender et al. 2007). Therefore, to benefit mule deer, habitat treatments must provide additional forage within close proximity to security cover (Wallmo et al. 1977, Watkins et al. 2007). Options for reducing pinyon‐juniper cover include prescribed fire as well as a suite of mechanical methods including hand thinning, chaining, mastication, and roller‐chopping (Figure 1). Because prescribed fire can be difficult to implement and control, managers often choose mechanical methods (Tausch et al. 2009). Chaining involves pulling a ship anchor chain between 2 bulldozers, which uproots entire trees and leaves their skeletons 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2 of 21 �| 3 of 21 F I G U R E 1 Equipment, residual structure, and vegetation response 9 years post‐treatment for A) chaining, B) roller‐chopping, and C) mastication. intact, providing structures which may act as security cover (Howard et al. 1987). The soil is disturbed in localized fashion where uprooting occurs, and the tracks of the bulldozers create additional ground disturbance. Roller‐ chopping involves pushing trees over with a single bulldozer, which pulls a heavy rotating drum with protruding plates. The plates cut tree debris into large pieces, and the drum as well as the bulldozer tracks create ground disturbance. Mastication involves mulching entire trees to the ground level, providing fine debris which can help retain moisture (Owen et al. 2009, Young et al. 2013), but leaves behind no large woody debris structure. Mastication implements generally have rubber tires, which minimize ground disturbance, and seeding is accomplished by broadcasting seed prior to treatment. Chaining and roller‐chopping allow a seeding method not available with mastication; seeds that require deeper soil penetration can be dribbled on to the tracks of the bulldozer. This seeding method can be important because deep planting may benefit shrubs such as bitterbrush (Purshia tridentata; Paschke et al. 2003), a shrub highly utilized by mule deer in winter (Bartmann 1983). The differences between mechanical methods can influence vegetation and animal response (Short et al. 1977, Chambers et al. 2021). A growing body of literature has examined understory vegetation response of pinyon‐juniper to fire and mechanical treatments (Skousen et al. 1989, Coop et al. 2017, Rubin and Roybal 2018, Ernst‐Brock et al. 2019, Monaco and Gunnell 2020). Additionally, there have been studies that examined large ungulate response to one or more mechanical treatment types (Howard et al. 1987, Bowyer et al. 2001, Bergman et al. 2014, Ranglack and du Toit 2015, Smythe et al. 2015). However, comparisons of multiple mechanical treatment types with both vegetation and ungulate response data are unavailable. Furthermore, as short‐term responses to treatments often 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON differ from those observed later, studies with longer‐term responses (>5 years) are particularly valuable (Havrilla et al. 2017, Ernst‐Brock et al. 2019). The goal of our study was to compare both vegetation and mule deer responses between untreated (hereafter control), chained, roller‐chopped, and masticated plots, seeded and unseeded, in an area of moderate pinyon‐ juniper dominance that serves as winter mule deer habitat. An earlier study reported on vegetation responses 1–2 years post‐treatment (Stephens et al. 2016). Here, we report vegetation responses 6–7 years post treatment, use of palatable shrubs prior to winter, and days of mule deer winter use from point densities obtained from GPS collars deployed annually over a 5 year period. Our objectives were to determine: 1) which treatment type produced the best forb and shrub establishment from a concurrent seeding; 2) whether treatment types differ in cover of the undesirable annual invasive grass Bromus tectorum (hereafter cheatgrass) and less palatable native species; 3) the response of palatable forage shrubs to mechanical treatments and seeding; 4) which treatment type was most utilized during the growing season; and 5) differences among treatments in days of winter mule deer use. STUDY AREA The Piceance Creek Basin in Rio Blanco County, Colorado, USA, serves as winter range for one of North America's largest migratory mule deer populations (White and Lubow 2002). In recent decades, construction of well pads, roads and compressor stations for natural gas extraction has fragmented wildlife habitat and impacted mule deer habitat use (Anderson 2014, Northrup et al. 2015, Northrup et al. 2021). As mitigation for impacts on wildlife, extensive removal of pinyon‐juniper has occurred in efforts to increase forage quality and quantity. Although cattle, wild horses (Equus ferus), and elk were also present on the landscape, mule deer were the most numerous large herbivores during our study, even during summer, as some mule deer were non‐migratory and remained on winter range year round (Lendrum et al. 2014, Colorado Parks and Wildlife 2019). Within the Piceance Creek Basin, our study was conducted at 2 sites that were approximately 4.5 km apart within the area north of Piceance Creek colloquially known as Magnolia (north Magnolia site: UTM 738327 E, 4423141 N, 12S; south Magnolia site: UTM 733958 E, 4420956 N, 12S, Figure 2). The sites ranged from 2,000 to 2,100 m in elevation and from 43 to 45 cm in annual precipitation (PRISM Climate Group 2021). During the years when our mule deer data were collected (2014–2019), mean winter temperature was −1.0°C (PRISM Climate Group 2021). Soils at both sites were shallow and well‐drained, derived from sandstone and shale bedrock (Tiedeman 1978). Dominant understory vegetation included serviceberry (Amelanchier alnifolia), snowberry (Symphoricarpos rotundifolius), bitterbrush (Purshia tridentata), mountain big sagebrush (Artemisia tridentata), mountain mahogany (Cercocarpus montanus), plains pricklypear cactus (Opuntia polyacantha), phlox (Phlox spp.), hoary tansy aster (Machaeranthera canescens), Lewis flax (Linum lewisii), sedges (Carex spp.), wildrye (Elymus spp.), western wheatgrass (Pascopyrum smithii), bluegrass (Poa spp.), and Indian ricegrass (Achnatherum hymenoides). Natural predators of mule deer in this area included coyotes (Canis latrans), cougars (Puma concolor), bobcats (Lynx rufus), and black bears (Ursus americanus; Lendrum et al. 2014). The area was popular for hunting during the fall with an annual average of 511 mule deer harvested in the wildlife management unit (Game Management Unit 22) that encompassed both the north and south sites. METHO DS Experimental design and treatments Four blocks at the north site and 3 blocks at the south site were treated in October and November 2011 (Figure 2). Blocks consisted of a randomly assigned chained plot, roller‐chopped plot, masticated plot, and untreated control 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 4 of 21 �| 5 of 21 F I G U R E 2 Location of tree removal and control plots within north and south Magnolia deer capture areas (white outline) in the Piceance Basin, Rio Blanco County, Colorado, USA. Each block contained a control plot (pink), a chained plot (yellow), a masticated plot (teal), and a roller‐chopped plot (purple). Chained, masticated, and roller‐ chopped plots were divided into seeded (solid outline) and unseeded (dashed outline) subplots. Control plots were unseeded. 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON plot for a total of 28 plots across 7 blocks. Plots were approximately 137 m × 60 m (0.8 ha), but in some instances treated areas were slightly larger due to difficulty of operating heavy equipment within a precise area. Chained plots were treated with a shorter than typical 15 m smooth chain strung between 2 bulldozers, and the chain was dragged over the same area twice, with the second pass in the opposite direction of the first pass. Chaining generated woody debris and uprooted trees which were scattered and piled across the plot (Figure 1A). Roller‐ chopped plots were treated with a bulldozer towing a drum that was 3.6 m long with 25 cm blades and an operational weight of 9,100 kg. The length of woody material left by this treatment varied depending on diameter, with branches less than 10 cm in diameter chopped into 0.5 m sections or smaller, but larger boles left largely intact. Woody debris was scattered across the plot with some vertical structure, but with less than that of chaining (Figure 1B). Masticated plots were treated with a rubber‐tired tractor mounted with a Fecon Bull Hog mulcher (Fecon Inc., Lebanon, OH, USA). Masticated material was less than approximately 20 cm in length and provided no vertical structure; woody material scattered across the plot varied in depth between 0 and 25 cm (Figure 1C). All trees within the plot boundary experienced mechanical treatment. Plots were separated by strips of intact tree cover 20–60 m in width. No point within treated areas was more than 35 m from dense tree cover, well under the maximum of ~125 m recommended for mule deer habitat treatments by earlier studies (Short et al. 1977, Skousen et al. 1989). Mechanically treated plots were further divided into randomly assigned seeded and unseeded subplots. Control plots were not seeded because one of the main assumptions for the experiment was that the presence of pinyon‐ juniper overstory was contributing to reduced understory (Schott and Pieper 1985, Naillon et al. 1997) and therefore, adding seed to plots with intact overstory would not increase understory vegetation. Thus, each of the 7 blocks contained 4 unseeded subplots (control, masticated, chained, and roller‐chopped) and 3 seeded subplots (masticated, chained, and roller‐chopped), for a total of 49 subplots. The seeding rate was 600 pure live seeds per m2 and the mix included 10 shrub species, 14 forb species and 10 grass species. The seed mix contained a high proportion of native shrub seeds, particularly serviceberry, bitterbrush, and mountain mahogany, as these species comprise a large proportion of most mule deer winter diets in the Piceance Basin (Bartmann 1983). All seed in masticated plots, and most seed in chained and roller‐chopped plots, was hand‐broadcast just prior to treatment. In chained and roller‐chopped plots, several large‐seeded species that benefit from deep planting, including serviceberry, bitterbrush, mountain mahogany, skunkbush sumac (Rhus trilobata), Utah sweetvetch (Hedysarum boreale), and silvery lupine (Lupinus argenteus), were applied with a seed dribbler mounted on the tracks of the bulldozers (Plummer et al. 1968). For further details on the seed mix and seeding methods see Stephens et al. (2016). Vegetation Although plant cover has historically been quantified by ocular estimation (e.g. Daubenmire plots), later work showed that more precise and efficient plant cover estimates can be produced via the point intercept technique (Godinez‐Alvarez et al. 2009) as Daubenmire estimation commonly suffers from observer bias (Bergstedt et al. 2009, Godinez‐Alvarez et al. 2009). With point intercept, this problem is solved by constraining sampling to an unbiased, systematic selection of points in space. Plant cover is estimated by dividing the number of times a particular species (or species group) is encountered by the total number of points sampled. Over the past 15 years point intercept has become a standard technique used by land management agencies (Herrick et al. 2005, MacKinnon et al. 2011). In July 2016 and 2017, we measured plant cover in all 49 subplots via the line point intercept technique. We used 13, ~23 m transects per subplot, arrayed systematically perpendicular to the long axis of the subplot, ~8 meters apart. Our sampling points were spaced at 1 m intervals along each transect, resulting in 299 total points per subplot. A 5 m buffer from the plot edges was excluded from measurement. At each point, we aimed a laser sampling device (Synergy Resource Solutions, Belgrade MT, USA) at the ground and recorded all 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 6 of 21 �| 7 of 21 species of vegetation intersecting the path of the laser (Herrick et al. 2005). If necessary, upper canopy layers were brushed aside to reveal lower layers. Species were only recorded once per point, even if the laser hit the species more than once along its path. Cover data were collected similarly in 2012–2013 (Stephens et al. 2016), and we combined earlier data with these data for analysis. When possible, species were grouped into functional groups relevant to mule deer nutrition (Bartmann 1983): bitterbrush, serviceberry, snowberry, other shrubs, perennial grasses (all native), perennial forbs (all native with trace exotics), cheatgrass, and exotic annual forbs. With the exception of cheatgrass and exotic annual forbs, each of these groups is an important component of mule deer diet for at least some portion of the year (Bartmann 1983). If multiple species within a functional group were hit at a point (for instance, 2 kinds of perennial grass), this was counted as only one instance of the functional group. Percent cover of these groups was calculated at the subplot level. Although native annual forbs were sufficiently present to analyze in the prior study (Stephens et al. 2016), by 2016 and 2017 they were virtually absent, and were therefore not re‐analyzed for this study. We quantified density, leader length, and summer utilization of preferred forage shrubs using belt transects. We used 5 of the 13 transects used for plant cover, selected systematically so that data were collected evenly over the length of the subplot. In late September 2017, we counted and sampled all serviceberry and bitterbrush shrubs with at least 50% of their canopy within a 4 m wide belt centered on each of these transects. A total area of 240 m2 was sampled per subplot. Serviceberry and bitterbrush are both heavily preferred mule deer winter forage (Bartmann 1983) and sufficiently abundant for analysis. We measured the longest leader (current year shoot) on each plant from the bud scar to the apex. We recorded utilization of each shrub by benchmarked ocular estimation of the percentage of current year leaves and shoots that had been removed by growing season browsing. Growing season utilization was identifiable by the color of the wood at the bite mark (pinkish or tan rather than grey). Herbivores preferentially browse plant parts that are more accessible, palatable, and present larger bite sizes (Armstrong and Macdonald 1992, Danell et al. 1994). Therefore, the largest shoots at the apex of the plant are often browsed first, smaller lateral shoots are consumed next, and at utilization rates above 45%, entire apical shoots are typically removed, resulting in bite marks into less palatable older wood (Bilyeu et al. 2007). The old wood bites render most traditional browse utilization methods unreliable, but they are in themselves a useful benchmark for estimating utilization rates (Bilyeu et al. 2007). Old wood bites are distinguishable from current year shoot bites by bark color and branching pattern. We approximated the number of current year bites and old wood bites on each shrub, and used these benchmarks to ensure consistency in estimates of utilization among observers: only one to a few nibbles of current‐year shoot tips (5%); several nibbles of shoot tips (10%); many nibbles plus at least one old wood bite (30%); many nibbles of current year shoots, several old wood bites, most branches having some browsing evident (50%); many old wood bites, the only intact current‐year shoots occurring deep within the plant where they had protection from herbivory by woody branches above (70%). The efficiency of our approach was important because both shrub productivity and herbivore foraging behavior are highly spatially variable (Rutherford 1979, Senft et al. 1987), producing a high coefficient of variation in utilization rates, necessitating a large sample. Shrub count, mean length of longest leader, and average utilization were calculated at the subplot level. Because the design of the experiment was not fully factorial (there were no seeded control plots), 2 types of analyses were used to examine vegetation data: mechanical treatment effects analysis (MEA) which included mechanical treatment as a fixed effect, and seeding effect analysis (SEA), which included mechanical treatment, seeding treatment, and their interaction as fixed effects (Stephens et al. 2016). We used only unseeded subplots to examine effects of mechanical treatments relative to one another and to untreated control plots for MEA. We excluded control plots for SEA, which had no mechanical treatment, to allow analysis of the seeding treatment, and interactions involving the seeding treatment. We analyzed data as generalized linear mixed models using the glmmTMB package in R version 4.1.1 (Brooks et al. 2017, R Core Development Team 2020). In both the MEA and SEA, site was included as a fixed effect and block was included as a random intercept. In the SEA, plot was also included as a random intercept to account for the split‐plot design. Cover variables were analyzed as repeated 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON measures, with year, treatments, and year by treatment interactions as fixed effects. Year was analyzed as a factor because many of the responses, especially annuals, fluctuated widely from year to year. No cheatgrass cover was detected in 2012 so that year was dropped in the cheatgrass analysis. Accounting for correlations among repeated measures was accomplished with random intercept effects (Brooks et al. 2017) for subplot in the SEA and plot in the MEA. We analyzed cover and density as count data (count of point intercept hits or count of shrubs), utilizing a negative binomial model because data were overdispersed (Linden and Mantyniemi 2011), with offsets included to account for slight variations in sampling effort (log of total point intercept hits for cover data, and log of sampled area for density data). Utilization and leader length were normally distributed, and we therefore applied Gaussian models. Analysis of variance (ANOVA) computations and treatment comparisons were done using the emmeans package in R (Lenth 2020). We computed contrasts between mechanical treatments for the MEA (within years if repeated measures), and computed contrasts between seeded and unseeded subplots within levels of mechanical treatment for the SEA. Means are presented ±SE and we omit discussion of models in which no significant contrasts occurred (α > 0.05). Mule deer We opportunistically captured and recaptured adult female mule deer (>1 year old) twice each winter, respectively, from December 2014–March 2018 on the Magnolia winter range study area (Figure 2) using helicopter net gunning (Krausman et al. 1985, Webb et al. 2008, Northrup et al. 2014) to address additional research objectives (Peterson et al. 2018, Northrup et al. 2021). Upon capture, we administered 0.5 mg/kg of midazolam and 0.25 mg/kg of Azaperone (Wildlife Pharmaceuticals, Windsor, CO, USA) and transferred mule deer to a central processing site via helicopter (typically <4 km but occasionally up to 6.5 km). At the processing site, we took physiological data and fit each mule deer with a global positioning system (GPS) radio collar (G2110D Advanced Telemetry Systems, Isanti MN, USA) set to attempt a relocation once every 5 hours and equipped with a mechanism programmed to release during April of the year following deployment. Between 62 and 67 adult female mule deer were captured and collared each year, with approximately equal numbers caught at the north and south sites annually (Figure 2). For additional details on capture and handling methods, see Lendrum et al. (2014) and Northrup et al. (2021). We focused on winter because the study area serves primarily as winter range, with a much lower summer mule deer density (Lendrum et al. 2014, Northrup et al. 2015). Our study period was defined as occurring from October 15 through April 15 each year. Mule deer winter home ranges were calculated using minimum convex polygons. Although minimum convex polygons have been criticized as being too inclusive by potentially including unused habitats, inclusivity was appropriate in this case as we needed to determine the treatment plots to which each mule deer had been exposed, and GPS locations within the blocks confirmed exposure. At about 1 hectare, the scale of our treatment plots was small relative to mule deer home ranges. This scale precluded some analyses common in mule deer habitat use studies, such as resource selection functions. Nonetheless, when aggregated over 5 years, we had sufficient mule deer use days from location points within our study plots to perform a traditional ANOVA that gained strength from its replication and control plots. Each of the 7 blocks of 4 treatments were considered a possible trial for each mule deer, to be included only if the mule deer's home range completely encompassed the block. Our approach allowed us to distinguish between true zeros (the mule deer was exposed to all 4 treatments but did not use one or more of them) from missing data (for plots outside the home range). To minimize potential autocorrelation among sample points, we randomly selected one detection per mule deer per day (0000 to 2400 hours) that was at least 5 hours from the prior detection (i.e., mule deer use days). Next, we summed the number of use days for each mule deer in each plot, aggregating data from 2014–2019. We excluded blocks if the mule deer's home range did not completely encompass the block, if all values were zero for the block, or if GPS fix rates were <0.850. This process resulted in 26 mule deer in the analysis consisting of 13 using the 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 8 of 21 �| 9 of 21 southern plots and 13 using the northern plots (mean GPS fix success rate: 0.936 ± 0.005 SE), and a total of 208 individual × plot combinations. Because plot areas varied slightly, we calculated the number of mule deer use days per ha and analyzed this using a linear mixed model in package glmmTMB in R (Brooks et al. 2017). Block and individual ID were included as random intercept effects, while site and treatment type (chained, roller‐chopped, masticated, or control) were fixed effects. We tested for a site by treatment interaction term as a fixed effect but dropped it from the final model as it was non‐significant. We conducted ANOVA computations and treatment comparisons using the emmeans package in R (Lenth 2020). Our approach prevented mule deer with more detections from having undue influence on the results. While the range in the number of use days per individual was large (North Magnolia: 1 to 91 with a median of 17; South Magnolia: 1 to 85 with a median of 10), the range in the number of included blocks per individual was smaller (1 to 4). The inclusion of the individual ID random intercept accounted for variability due to individual in the total number of use days within a block, ensuring that on a block‐by‐block basis, the influence of each individual on treatment effects was the same (Bates 2005). An individual with 4 blocks would have more influence than an individual with one, but given the total number of block × individual combinations (52), no individual accounted for more than ~8% of modeled treatment effects. RESULTS Analysis details for the MEA are included in Supplementary Table 1 (ANOVA results), Supplementary Table 2 (contrasts between mechanical treatments for cover variables for each year) and Supplementary Table 3 (contrasts between mechanical treatments for density, utilization, and leader length variables). Statistical details for the SEA are included in Supplementary Table 4 (ANOVA results), Supplementary Table 5 (contrasts between seeded and unseeded for cover variables for each year) and Supplementary Table 6 (contrasts between seeded and unseeded for density, utilization, and leader length variables). We found no direct effect of mechanical treatment type on bitterbrush cover, but mechanical treatments did differ in how they interacted with seeding. Seeding increased bitterbrush cover 2 and 6 years post‐treatment within roller‐chopped plots and 6 years post‐treatment within masticated plots, but seeding never increased bitterbrush cover within chained plots (Figure 3, Supplementary Table 5). Seeding increased 2017 bitterbrush density only within roller‐chopped plots, where it brought about a 3‐fold increase (Figure 4, Supplementary Table 6). Growing season utilization of bitterbrush was about twice as high in unseeded masticated plots as in any other unseeded treatment, (Figure 5A, Supplementary Table 3). Bitterbrush leader length was ~35% lower in control plots than in unseeded mechanically treated plots (Figure 6A, Supplementary Table 3). All 3 mechanical treatments caused an initial reduction in serviceberry cover. One year after treatments were implemented, serviceberry cover was 7.1 ± 2.8% in control plots, but only 1.9 ± 0.6% in chained plots, 1.2 ± 0.6% in masticated plots, and 1.5 ± 0.6% in roller‐chopped plots. By year 2 only roller‐chopped plots were lower in serviceberry cover than the control plots. There was no significant effect in year 5, but in year 6 roller‐chopped plots were again lower than control plots (Supplementary Table 2). Roller‐chopped plots averaged 4.9 ± 2.3% serviceberry cover in year 6, while control plots averaged 10.0 ± 3.6%. Growing season utilization of serviceberry was lower in control plots than in unseeded mechanical treatments (Figure 5B, Supplementary Table 3), which were similar to one another. Within the masticated treatment only, serviceberry utilization was 32.2 ± 11.35% in unseeded subplots but only 18.3 ± 4.1% in seeded subplots (Supplementary Table 6). Serviceberry leader length was higher in unseeded roller‐chopped plots than in control or masticated plots (Figure 6B, Supplementary Table 3). Mechanical treatment impacted snowberry cover in unseeded plots, with the most apparent differences in year 6 (Figure 7A). By year 6, chained and masticated plots had higher snowberry cover than control plots. In addition, chained plots had higher snowberry cover than roller‐chopped plots (Supplementary Table 2). 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON F I G U R E 3 Percent cover of bitterbrush within seeded (solid lines) and unseeded (dashed lines) subplots 1–6 years after implementation of 3 pinyon and juniper removal methods. Stars indicate significant contrasts between seeded and unseeded subplots within years at α = 0.05. Error bars = SE. F I G U R E 4 2017 bitterbrush density within seeded (solid outline) and unseeded (dashed outline) subplots 6 years after implementation of 3 pinyon and juniper removal methods: CON (control), MAST (masticated), CHAIN (chained), and ROLLER (roller‐chopped). Star indicates a significant contrast between seeded and unseeded subplots at α = 0.05. Error bars = SE. The dominant species making up the other shrub category were species with mule deer forage value such as mountain mahogany, mountain big sagebrush, and green rabbitbrush (Chrysothamnus viscidiflorus). These shrubs were initially negatively impacted by all 3 mechanical treatments; one year after treatments were implemented, other shrub cover was 4.2 ± 1.4% in control plots, but only 1.2 ± 0.3% in chained plots, 1.5 ± 0.8% in masticated 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 10 of 21 �| 11 of 21 F I G U R E 5 Percent of current year growth removed by herbivory during the growing season for A) bitterbrush and B) serviceberry 6 years following implementation of 3 pinyon and juniper removal methods: CON (control), MAST (masticated), CHAIN (chained), and ROLLER (roller‐chopped), unseeded subplots only. Bars not sharing letters are significantly different at α = 0.05. Error bars = SE. plots, and 1.6 ± 0.6% in roller‐chopped plots (Supplementary Table 2). However, shrubs recovered quickly in treated plots and there were no significant differences between treated plots and controls in subsequent years. In year 6, other shrubs averaged 3.23 ± 0.6% cover across all plots. Perennial grasses were 98% native and dominated by the forage species bottlebrush squirreltail (Elymus elymoides), western wheatgrass, and Indian ricegrass. Perennial grasses responded to all 3 unseeded mechanical treatment types similarly; there was an initial negative impact of mechanical treatment, followed by a positive impact that increased in a monotonic fashion through year 6 (Figure 7B). In all 4 measurement years, perennial grass cover in unseeded mechanical treatments differed from control plots, but not from each other (Supplementary Table 2). Perennial forbs were 96% native with over 25 species and no single species exceeding 2% cover. The most dominant species had high forage value, and included white sagebrush (Artemisia ludoviciana), silvery lupine, hoary tansy aster, scarlet globemallow (Sphaeralcea coccinea), and bastard toadflax (Comandra umbellata). In year 1, unseeded chained and roller‐chopped plots had lower perennial forb cover than control plots. In addition, chained 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON F I G U R E 6 Average September length of the longest leader/plant for A) bitterbrush and B) serviceberry 6 years following implementation of 3 pinyon and juniper removal methods: CON (control), MAST (masticated), CHAIN (chained), and ROLLER (roller‐chopped), unseeded subplots only. Bars not sharing letters are significantly different at α = 0.05. Error bars = SE. and roller‐chopped plots had lower perennial forb cover than masticated plots. There were no further significant differences in yearly contrasts, but there were trends in years 2, 5, and 6 for higher perennial forb cover in masticated plots than in control plots (Supplementary Table 2). Seeding had an impact on perennial forb cover, with differences observed in year 5, when perennial forb cover in seeded subplots was higher than in unseeded subplots within all 3 mechanical treatment types (Figure 8, Supplementary Table 5). The difference can be attributed to the seeded species arrowleaf balsamroot (Balsamorhiza sagittata), Lewis flax, and Utah sweetvetch, all of which had 2–3 times higher cover in seeded subplots than in unseeded subplots. Exotic annual forbs were dominated by mullein (Verbascum thapsus), desert alyssum (Alyssum desertorum), tall tumble mustard (Sisymbrium altissimum), prickly lettuce (Lactuca serrulata), and flixweed (Descurainia sophia). Mechanical treatments, in the absence of seeding, impacted exotic annual forbs, although this was only evident in year 2. In year 2, exotic annual forb cover was higher in roller‐chopped plots than in chained or masticated plots, which were in turn higher than control plots (Figure 7C, Supplementary Table 2). Cheatgrass cover was higher in all 3 mechanical treatment types than control plots in years 2, 5, and 6 (Figure 7D, Supplementary Table 2). 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 12 of 21 �| 13 of 21 F I G U R E 7 Percent cover of A) snowberry, B) perennial grasses, C) exotic annual forbs, and D) cheatgrass 1–6 years following implementation of 3 pinyon and juniper removal methods, unseeded subplots only. Points not sharing letters are significantly different at α = 0.05 for within‐year contrasts between treatments. Error bars = SE. In addition, an effect of higher cheatgrass cover in roller‐chopped plots than in masticated plots in year 6 approached statistical significance (Figure 7D, Supplementary Table 2). Mule deer winter use days in roller‐chopped plots was 73% higher than in control plots (t204 = 2.593, P = 0.010, Figure 9). Winter use days in chained plots was 65% higher than in control plots (t204 = 2.868, P = 0.005, Figure 9). Winter use of masticated plots was statistically similar to control plots, and marginally lower (26%) than chained plots (t204 = 1.714, P = 0.088, Figure 9). Patterns in winter use days with respect to treatments were similar at North and South Magnolia (site × mechanical treatment interaction: t204 = 0.025, P = 0.904). DISCUSSION Consistent with prior studies (Skousen et al. 1989, Owen et al. 2009, Coop et al. 2017, Chambers et al. 2021) all 3 mechanical treatment types increased the herbaceous component, both of grasses, and, when combined with seeding, of forbs. The roller‐chopped treatment was the most successful at increasing the shrub forage resources 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON F I G U R E 8 Percent cover of perennial forbs within seeded (solid lines) and unseeded (dashed lines) subplots 1–6 years after implementation of 3 pinyon and juniper removal methods. Stars indicate significant contrasts between seeded and unseeded subplots within years at α = 0.05. Error bars = SE. F I G U R E 9 Winter mule deer use days from GPS locations (days/ha) over a 5‐year period in control plots and plots treated to remove pinyon and juniper trees by 3 different methods: CON (control), MAST (masticated), CHAIN (chained), and ROLLER (roller‐chopped). Bars not sharing letters are significantly different at α = 0.05. Error bars = SE. important for mule deer winter nutrition. Roller‐chopping facilitated a tripling in density of the desirable shrub bitterbrush, created the longest leader lengths in the desirable shrub serviceberry, and caused the least increase in the less‐desirable shrub snowberry. The vegetation responses of chaining and mastication were virtually indistinguishable from one another. Mule deer winter use days were as high in chained treatments as in 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14 of 21 �| 15 of 21 roller‐chopped treatments, while use days in masticated treatments were lower and not statistically different from control plots. This pattern was consistent between North and South Magnolia, even though an entirely different set of individuals inhabited the 2 study areas. Chained and roller‐chopped plots had large woody debris providing vertical structure, while the masticated plots had no vertical structure other than that of live shrubs. We observed that prior to winter, bitterbrush had been most heavily utilized in masticated plots. The primary large herbivore in the study area during summer and fall was mule deer (Colorado Parks and Wildlife 2019). Mule deer may have been attracted to masticated plots in summer because forbs responded well to mastication, and forbs are a particularly important part of mule deer summer diets (Short et al. 1977, Howard et al. 1987). However, there was no difference in forb cover between masticated plots and chained plots, so it was difficult to ascribe our results to forbs alone. Masticated plots are notably easier to walk through than either chained or roller‐chopped plots. Mule deer are sensitive to energetic costs of movements, and the importance of energetic costs can vary seasonally (Northrup et al. 2019). In summer, differences in energetic costs may have been a driving factor, as differences in security cover were minimal because of serviceberry which was present in all the plots. Serviceberry shrubs in our plots were sufficiently tall (~2 m) to provide security cover prior to leaf drop in late September. In winter, the absence of leafy shrubs may have made masticated plots less appealing, as they also lacked woody debris for security cover. In winter, security cover is especially vital, as both mule deer and predators are more concentrated in space (Watkins et al. 2007). In addition, the greater use of masticated plots in summer resulted in less high‐quality forage left over for winter use, as herbivores browse the most nutritious plants and plant parts first (Armstrong and Macdonald 1992). In the case of bitterbrush, utilization can lead to only larger diameter wood being available, which has lower nutritional quality (Bishop et al. 2001). These seasonal changes could make masticated plots less attractive to mule deer in winter. Because of the presence of large woody debris, these dynamics may have been reversed in roller‐chopped and chained plots. All herbivores utilized them only lightly during the growing season, possibly because they are more difficult to navigate. In winter however, the presence of snow makes energetic costs higher overall (Northrup et al. 2019), which could equalize differences in navigation difficulty among mechanical treatments. The greater nutritional resources and better security cover in chained or roller‐chopped plots may have made them more attractive winter foraging areas. Large woody debris may also have directly increased available forage in winter. Gleason et al. (2019) suggested that dark woody debris remaining after wildfire hastens snowmelt. Although our plots were not burned, both pinyon and juniper trees have dark bark and it is reasonable to conclude that snowmelt in the vicinity of large woody debris is also hastened, which would allow mule deer to access understory forage more easily. Mule deer will utilize a combination of grasses, forbs, and shrubs for as long as they are able to during winter (Carpenter et al. 1979); woody debris may extend the duration mule deer are able to maintain a varied diet. Thus large woody debris may act in several ways to increase winter mule deer benefit from tree removal treatments. Vegetation manipulations often entail tradeoffs, and our study was no exception. All 3 treatments caused an increase in undesirable cheatgrass, as has been shown elsewhere (Bates et al. 2005, Coop et al. 2017, Havrilla et al. 2017, Monaco and Gunnell 2020). However, the effect was reduced due to establishment of native species prior to the cheatgrass invasion, and dissipated by 2021. While roller‐chopping was best at facilitating establishment and productivity of key forage shrubs, it also produced the largest flush of non‐native annual forbs and resulted in a marginally higher response from cheatgrass. Both the ease of shrub establishment and the greater response of annuals are likely related to roller‐chopping creating the most extensive bare ground in the early years of the experiment (Stephens et al. 2016). The heavy rotating drum and protruding plates caused soil disturbance and a reduction in competition that allowed both desirable and undesirable plants to establish from seed. In the case of this study, the annual response was slight enough to be acceptable, but this may be a larger concern in drier environments (Chambers et al. 2021). Roller‐chopping impacted snowberry and serviceberry in ways that were favorable for mule deer. Snowberry was a less preferred forage species in our area that tended to increase with mechanical treatment. However, the 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON increase was more moderate with roller‐chopping than with chaining. Snowberry is a low‐growing, flexible shrub, and the chain simply passed over it, leaving snowberry free to benefit from a reduction in competitors. In contrast, the action of the heavy rotating drum in roller‐chopped plots set snowberry back. Roller‐chopping also set back serviceberry, a preferred forage species. By 6 years post‐treatment, serviceberry cover was still lower in roller‐ chopped plots than control plots, but leader lengths were longer, indicating that serviceberry plants were still recovering and remaining plants were more productive. Longer leader lengths are beneficial for large herbivores because they allow greater foraging efficiency (Shipley et al. 1999), thus the lower cover of serviceberry in roller‐ chopped plots was not necessarily a detriment for mule deer. The action of roller‐chopping on both snowberry and serviceberry may have been a factor in the success of bitterbrush seeding in roller‐chopped plots because of reduced competition. Managers should be aware of the pre‐existing shrubs prior to considering how vegetation may respond to alternative mechanical treatments. In our area, roller‐chopping provided a beneficial combination of vegetation reduction, plant establishment opportunity, and residual vertical structure. Perennial grasses increased steadily in response to tree removal, regardless of removal method or seeding treatment. Indeed, an increase in grasses is the most reliable result that can be expected from tree removal treatments (Jones 2019). Perennial forbs, on the other hand, increase less reliably (Jones 2019) and may need to be seeded (Havrilla et al. 2017, Ernst‐Brock et al. 2019). In our study, forbs increased only in seeded subplots, but increased similarly for all 3 removal methods. Large‐seeded forbs such as Utah sweetvetch (as well as large‐seeded shrubs such as bitterbrush) were seeded with a dribbler mounted to the bulldozers of the chained and roller‐ chopped treatments. The dribbler dropped seed onto the track of the bulldozer which then pressed the seed deeply and firmly into the ground. Although the spatial extent of seeding with this method is restricted, the success of forbs and bitterbrush within the chained and roller‐chopped treatments demonstrates its utility. In masticated plots, all seed was broadcast prior to mastication to ensure that seed would be in contact with soil. The success of forbs and bitterbrush within masticated plots demonstrates that this is also a useful method. However, broadcasting prior to treatment can be expensive, as it requires small‐scale machinery or hand‐held broadcasters capable of navigating between trees. Treatments utilizing bulldozers offer an alternative seeding option which may be more cost‐effective. The differences in winter mule deer use days observed in this study may be a conservative estimate of differences in foraging behaviors. We cannot reliably distinguish foraging behaviors from other activities in our GPS collar data. However, from a companion camera study conducted in summer in the same area (Colorado Parks and Wildlife 2019), we know that only about half of locations involve foraging behavior, and that differences in use between treated and control areas are much more apparent when only foraging behaviors are considered. Benefits to mule deer can and do occur in the absence of differences in changes in mule deer density (Bergman et al. 2014, Bergman et al. 2015). Therefore, while we did not detect a difference in mule deer winter use days between control and masticated plots, masticated plots may still provide meaningful benefits to mule deer winter range. The masticated treatments successfully increased grasses and forbs, which is important because mule deer prefer to use a combination of grasses, forbs and shrubs throughout the year (Carpenter et al. 1979, Watkins et al. 2007). Furthermore, the masticated treatments received a high degree of use prior to winter, which may have helped improve body condition for non‐migratory mule deer. Mastication can easily be adapted to leave trees within the boundary of treatment areas for security cover, which may improve winter use. Finally, while we did not detect a difference in non‐native species response between chaining and mastication at our mesic sites, it has been previously noted that mastication may result in less non‐native species cover than chaining in a drier ecosystem (Monaco and Gunnell 2020). Our study supports earlier work indicating that additional nutrition must be supplied within a context of security cover to best benefit mule deer (Watkins et al. 2007). We found that the roller‐chopping treatment, which facilitated establishment of a key winter forage shrub and provided at least some vertical structure, was highly utilized by mule deer in winter. The chained treatment, which provided some additional nutrition and the most structure, was also highly utilized. The mastication treatment, with some additional nutrition but no structure, was 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 16 of 21 �| 17 of 21 less well utilized. Our treatment areas were small and in most cases less than 35 m wide, well within distances to security cover previously recommended (Short et al. 1977, Skousen et al. 1989). However, there were no pinyon or juniper trees left within the boundaries of the treatment in our experimental design. Our results support the idea that some tree cover or vertical structure should be left within even small, linear treatments to provide additional security cover within habitat treatments intended to benefit mule deer. While managers occasionally burn piles of woody debris, we do not recommend this practice, as it eliminates beneficial structure (Howard et al. 1987) and can exacerbate outbreaks of cheatgrass (Havrilla et al. 2017). Furthermore, woody debris has additional benefits in ecosystems. Woody debris can reduce soil erosion on slopes (Noelle et al. 2017), and can improve microbial biomass and arbuscular mycorrhizae (Stoddard et al. 2008). Woody debris can also create microsites of lower soil temperatures during the growing season, facilitating plant establishment (Brown and Naeth 2014, Goldin and Hutchinson 2015). The role of large woody debris in facilitating earlier snowmelt and thereby facilitating winter foraging should be studied further. Although large woody debris may seem unsightly, we concur with earlier work indicating that treatments providing large woody debris should be reconsidered as a viable management option (Provencher and Thompson 2014). M A N A G E M E N T I M P L I C A TI O N S The ideal habitat treatment for mule deer would provide additional nutrition with as little loss of security cover as possible. Residual structure such as tree skeletons are beneficial in this regard, and may also increase the value of winter range treatments by discouraging use during the growing season, allowing more forage to be conserved for winter use. The relative forage value of dominant shrub species, and their probable response to different types of equipment, should be considered. Low growing, flexible shrubs can respond aggressively to chaining, and this may or may not be desirable. Equipment choice can also impact annual species response as well as the ease of establishing desirable species from seed. Tools that create more bare ground, such as roller‐chopping (and to a lesser extent chaining) impart a higher risk of annual weed invasion, but also provide greater opportunity for desired plant establishment. Cost per acre for mastication is about double that of roller‐chopping, which is in turn about double that of chaining. We note that mobilization cost of roller‐chopping is higher than that of chaining, which is in turn about double that of mastication. Roller‐chopping is well suited to a project in which a large acreage is intended to be treated in a mosaic style. Chaining is also well suited to treating large acreages, but a mosaic will be more difficult to produce because chaining requires 2 bulldozers to work in concert with each other. Mastication is well suited for smaller acreage projects where cost per acre is less of a concern than minimizing mobilization cost, but some trees and tall shrubs should be left for security cover within treatment areas. Our study was applied at relatively small scales with uniformly treated plots to focus primarily on vegetation responses, but larger scale treatments are necessary to enhance ungulate nutrition on winter range. While roller‐chopping produced the best combination of vegetation and mule deer responses in our study, both chaining and mastication may be appropriate in other contexts. Chaining is a cost‐effective tool that can result in a good response from mule deer in winter. Mastication was less favored by mule deer in winter, but produced good vegetation responses and may be a desirable option in drier environments with greater invasive species concerns. A C KN O W L E D G M E N T S We thank G. Stephens, M. Paschke, B. Wolk, and J. Jonas for assistance with the early years of the experiment, XTO Energy for initial funding, L. Belmonte and E. Hollowed of the Bureau of Land Management for access, T. Graham of Ranch Advisory Partners for help with site selection, J. Gammonley and S. Billings for administrative support, and M. Willie of T&M Contracting for treatment implementation. S. Van Nortwick, K. Kain, H. Trowbridge, and S. Thompsen collected vegetation data. C. Bishop helped initiate the experiment and provided advice on interpretation. C. Sundermann helped revise the paper. We thank L. Wolfe, D. Collins, M. Fisher, C. Bishop, 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON E. Bergman, D. Finley, D. Freddy, and numerous field technicians for mule deer capture coordination and field assistance. Fixed wing pilots L. Gepfert and L. Coulter aided with aerial telemetry flights and Quicksilver Air, Inc. assisted with mule deer captures. We thank A. Rodgers (Associate Editor), A. Knipps (Editorial Assistant), A. Tunstall (Copy Editor) and J. Levengood (Content Editor), and the anonymous reviewers for constructive comments that improved our manuscript. Funding and support for mule deer captures and monitoring were provided by Colorado Parks and Wildlife, White River Field Office of Bureau of Land Management, ExxonMobil Production/XTO Energy, Williams Production/WPX Energy, Shell Exploration and Production, EnCana Corp., Marathon Oil Corp., Federal Aid in Wildlife Restoration (W‐185‐R), the Colorado Mule Deer Foundation, the Colorado Mule Deer Association, Colorado Oil and Gas Conservation Commission, and the Colorado State Severance Tax. CO NFL I CTS OF I NTEREST The authors declare no conflicts of interest. ETHICS STATEME NT Mule deer capture and handling procedures were approved by the Colorado Parks and Wildlife Institutional Animal Care and Use Committee (protocol numbers 17‐2008 and 01‐2012) and followed the American Society of Mammalogists recommendations (Sikes et al. 2016). D A TA A V A I L A B I L I T Y S T A T E M E N T The data that support the findings of this study are available from the corresponding author upon reasonable request. ORCID Danielle Bilyeu Johnston Charles R. Anderson http://orcid.org/0000-0002-6885-5369 http://orcid.org/0000-0002-3063-1757 REFERENCES Anderson, C. R., Jr. 2014. Population performance of Piceance Basin mule deer in response to natural gas resource extraction and mitigation efforts to address human activity and habitat degradation. Colorado Parks and Wildlife. 26 p. https://cpw.catalog.aspencat.info/Files/412/ViewPDF Armstrong, H. M., and A. J. Macdonald. 1992. Tests of different methods for measuring and estimating utilization rate of heather (Calluna‐Vulgaris) by vertebrate herbivores. Journal of Applied Ecology 29:285–294. Austin, D. D. 2000. Managing livestock grazing for mule deer (Odocoileus hemionus) on winter range in the Great Basin. Western North American Naturalist 60:198–203. Bartmann, R. M. 1983. Composition and quality of mule deer diets on pinyon‐juniper winter range, Colorado. Journal of Range Management 36:534–541. Bates, D. 2005. Fitting linear mixed models in R. R news 5. Bates, J. D., R. F. Miller, and T. Svejcar. 2005. Long‐term successional trends following western juniper cutting. Rangeland Ecology & Management 58:533–541. Bender, L. C., L. A. Lomas, and T. Kamienski. 2007. Habitat effects on condition of doe mule deer in arid mixed wood land‐ grassland. Rangeland Ecology & Management 60:277–284. Bergman, E. J., C. J. Bishop, D. J. Freddy, G. C. White, and P. F. Doherty. 2014. Habitat management influences overwinter survival of mule deer fawns in Colorado. Journal of Wildlife Management 78:448–455. Bergman, E. J., P. F. Doherty, G. C. White, and D. J. Freddy. 2015. Habitat and herbivore density: Response of mule deer to habitat management. Journal of Wildlife Management 79:60–68. Bergstedt, J., L. Westerberg, and P. Milberg. 2009. In the eye of the beholder: bias and stochastic variation in cover estimates. Plant Ecology 204:271–283. Bilyeu, D. M., D. J. Cooper, and N. T. Hobbs. 2007. Assessing impacts of large herbivores on shrubs: tests of scaling factors for utilization rates from shoot‐level measurements. Journal of Applied Ecology 44:168–175. Bishop, C. J., E. O. Garton, and J. W. Unsworth. 2001. Bitterbrush and cheatgrass quality on 3 southwest Idaho winter ranges. Journal of Range Management 54:595–602. 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 18 of 21 �| 19 of 21 Bishop, C. J., G. C. White, D. J. Freddy, B. E. Watkins, and T. R. Stephenson. 2009. Effect of enhanced nutrition on mule deer population rate of change. Wildlife Monographs 172:1–28. Bowyer, R. T., B. M. Pierce, L. K. Duffy, and D. A. Haggstrom. 2001. Sexual segregation in moose: effects of habitat manipulation. Alces 37:109–122. Brooks, M. E., K. Kristensen, K. J. van Benthem, A. Magnusson, C. W. Berg, A. Neilsen, H. J. Skaug, M. Maechler, and B. M. Bolker. 2017. glmmTMB balances speed and flexibility among packages for zero‐inflated generalized linear mixed modeling. The R Journal 9:378–400. Brown, R. L., and M. A. Naeth. 2014. Woody debris amendment enhances reclamation after oil sands mining in Alberta, Canada. Restoration Ecology 22:40–48. Carpenter, L. H., O. C. Wallmo, and R. B. Gill. 1979. Forage diversity and dietary selection by wintering mule deer. Journal of Range Management 32:226–229. Chambers, J. C., A. K. Urza, D. I. Board, R. F. Miller, D. A. Pyke, B. A. Roundy, E. W. Schupp, and R. J. Tausch. 2021. Sagebrush recovery patterns after fuel treatments mediated by disturbance type and plant functional group interactions. Ecosphere 12:e03450. Coe, P. K., B. K. Johnson, K. M. Stewart, and J. G. Kie. 2004. Spatial and temporal interactions of elk, mule deer and cattle. Pages 656–669 in J. Rahm, editor. Transactions of the Sixty‐Ninth North American Wildlife and Natural Resources Conference. Colorado Parks and Wildlife. 2019. Wildlife research reports: Mammals, July 2019. 48 p. https://spl.cde.state.co.us/ artemis/nrserials/nr616internet/nr616201819internet.pdf Coop, J. D., T. A. Grant, P. A. Magee, and E. A. Moore. 2017. Mastication treatment effects on vegetation and fuels in pinon‐juniper woodlands of central Colorado, USA. Forest Ecology and Management 396:68–84. Danell, K., R. Bergstrom, and L. Iedenius. 1994. Effects of large mammaliam browers on architecture, biomass, and nutrients of woody‐plants. Journal of Mammalogy 75:833–844. Dwinnell, S. P. H., H. Sawyer, J. E. Randall, J. L. Beck, J. S. Forbey, G. L. Fralick, and K. L. Monteith. 2019. Where to forage when afraid: Does perceived risk impair use of the foodscape? Ecological Applications 29:e01972. Ernst‐Brock, C., L. Turner, R. J. Tausch, and E. A. Leger. 2019. Long‐term vegetation responses to pinyon‐juniper woodland reduction treatments in Nevada, USA. Journal of Environmental Management 242:315–326. Gleason, K. E., J. R. McConnell, M. M. Arienzo, N. Chellman, and W. M. Calvin. 2019. Four‐fold increase in solar forcing on snow in western US burned forests since 1999. Nature Communications 10, 2026 Godinez‐Alvarez, H., J. E. Herrick, M. Mattocks, D. Toledo, and J. Van Zee. 2009. Comparison of 3 vegetation monitoring methods: Their relative utility for ecological assessment and monitoring. Ecological Indicators 9:1001–1008. Goldin, S. R., and M. F. Hutchinson. 2015. Thermal refugia in cleared temperate Australian woodlands: coarse woody debris moderate extreme surface soil temperatures. Agricultural and Forest Meteorology 214:39–47. Havrilla, C. A., A. M. Faist, and N. N. Barger. 2017. Understory plant community responses to fuel‐reduction treatments and seeding in an upland pinon‐juniper woodland. Rangeland Ecology & Management 70:609–620. Herrick, J. E., J. W. Van Zee, K. M. Havstad, L. M. Burkett, and W. G. Whitford. 2005. Monitoring Manual for Grassland, Shrubland and Savanna Ecosystems Vol 1: Quick Start. Volume 1. USDA‐ARS Jornada Experimental Range, Las Cruces, New Mexico, USA. Howard, V. W., K. M. Cheap, R. H. Hier, T. G. Thompson, and J. A. Dimas. 1987. Effects of cabling on pinyon‐juniper on mule deer and lagomorph use. Wildlife Society Bulletin 15:242–247. Jensen, C. H., G. W. Scotter, and A. D. Smith. 1972. Guidelines for grazing sheep on rangelands used by big game in winter. Journal of Range Management 25:346–352. Jones, A. 2019. Do mechanical vegetation treatments of pinyon‐juniper and sagebrush communities work? A review of the literature. Utah Wild Project. https://doi.org/10.13140/RG.2.2.12538.13760 Krausman, P. R., J. J. Hervert, and L. L. Ordway. 1985. Capturing deer and mountain sheep with a net‐gun. Wildlife Society Bulletin 13:71–73. Kufeld, R. C. 1973. Foods eaten by Rocky Mountain elk. Journal of Range Management 26:106–113. Ledesma, R., C. Kunst, S. Bravo, M. Leiva, L. Lorea, J. Godoy, and V. Navarrete. 2018. Developing a prescription for brush control in the Chaco region, effects of combined treatments on the canopy of 3 native shrub species. Arid Land Research and Management 32:351–366. Lendrum, P. E., C. R. Anderson, Jr., K. L. Monteith, J. A. Jenks, and R. T. Bowyer. 2014. Relating the movement of a rapidly migrating ungulate to spatiotemporal patterns of forage quality. Mammalian Biology 79:369–375. Lenth, R. V. 2020. emmeans: Estimated marginal means, aka least‐squared means. R package version 1.5.3. https://CRAN. R-project.org/package=emmeans Linden, A., and S. Mantyniemi. 2011. Using the negative binomial distribution to model overdispersion in ecological count data. Ecology 92:1414–1421. 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL �| JOHNSTON and ANDERSON MacKinnon, W. C., J. W. Karl, J. J. Roevs, M. Taylor, C. S. Karl, C. S. Spurrier, and J. E. Herrick. 2011. BLM core terrestrial indicators and methods: Tech Note 440. US Dept. of the Interior, Bureau of Land Management. Miller, R. F., J. D. Bates, T. J. Svejcar, F. B. Pierson, and L. E. Eddleman. 2005. Biology, ecology, and management of western juniper (Juniperus occidentalis). Volume Technical Bulletin 152, Oregon State University Agricultural Experiment Station. Miller, R. F., T. J. Svejcar, and J. A. Rose. 2000. Impacts of western juniper on plant community composition and structure. Journal of Range Management 53:574–585. Monaco, T. A., and K. L. Gunnell. 2020. Understory vegetation change following woodland reduction varies by plant community type and seeding status: A region‐wide assessment of ecological benefits and risks. Plants 9, 1113. Naillon, D., K. L. Memmott, and S. B. Monsen. 1997. A comparison of understory species at 3 densities in a pinyon‐juniper woodland. in S. B. Monsen and S. Stevens, editors. Ecology and management of pinyon‐juniper communities within the Interior West. USDA Forest Service, Provo, Utah, USA. Noelle, S. M., C. A. Carr, T. K. Stringham, and M. A. Weltz. 2017. Slash application reduces soil erosion in steep‐sloped pinon‐juniper woodlands. Rangeland Ecology & Management 70:774–780. Northrup, J. M., C. R. Anderson, Jr., B. D. Gerber, and G. Wittemyer. 2021. Behavioral and demographic responses of mule deer to energy development on winter range. Wildlife Monographs 208:1–37. Northrup, J. M., C. R. Anderson, Jr., and G. Wittemyer. 2014. Effects of helicopter capture and handling on movement behavior of mule deer. Journal of Wildlife Management 78:731–738. Northrup, J. M., C. R. Anderson, Jr., and G. Wittemyer. 2015. Quantifying spatial habitat loss from hydrocarbon development through assessing habitat selection patterns of mule deer. Global Change Biology 21:3961–3970. Northrup, J. M., A. Avrin, C. R. Anderson, Jr., E. Brown, and G. Wittemyer. 2019. On‐animal acoustic monitoring provides insight to ungulate foraging behavior. Journal of Mammalogy 100:1479–1489. Owen, S. M., C. H. Sieg, C. A. Gehring, and M. A. Bowker. 2009. Above‐ and belowground responses to tree thinning depend on the treatment of tree debris. Forest Ecology and Management 259:71–80. Paschke, M. W., E. F. Redente, and S. L. Brown. 2003. Biology and establishment of mountain shrubs on mining disturbances in the Rocky Mountains, USA. Land Degradation & Development 14:459–480. Peterson, M. E., C. R. Anderson, Jr., J. M. Northrup, and P. F. Doherty, Jr. 2018. Mortality of mule deer fawns in a natural gas development area. Journal of Wildlife Management 82:1135–1148. Plummer, A. P., D. R. Christensen, and S. B. Monson. 1968. Restoring big‐game range in Utah. Utah Division of Fish and Game, Ephraim, UT, USA. PRISM. Climate Group 2021. Oregon State University. http://prism.oregonstate.edu Provencher, L., and J. Thompson. 2014. Vegetation responses to pinyon‐juniper treatments in eastern Nevada. Rangeland Ecology & Management 67:195–205. R Core Development Team. 2020. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/ Ranglack, D. H., and J. T. du Toit. 2015. Wild bison as ecological indicators of the effectiveness of management practices to increase forage quality on open rangeland. Ecological Indicators 56:145–151. Romme, W. H., C. D. Allen, J. D. Balley, W. L. Baker, B. T. Bestelmeyer, P. M. Brown, K. S. Eisenhart, M. L. Floyd, D. W. Huffman, B. F. Jacobs, et al. 2009. Historical and modern disturbance regimes, stand structures, and landscape dynamics in pinon‐juniper vegetation of the western United States. Rangeland Ecology & Management 62:203–222. Rubin, R. L., and C. M. Roybal. 2018. Plant community responses to mastication and mulching of one‐seed juniper (Juniperus monosperma). Rangeland Ecology & Management 71:753–756. Rutherford, M. C. 1979. Plant‐based techniques for determining available browse and browse utilization ‐ review. Botanical Review 45:203–228. Schott, M. R., and R. D. Pieper. 1985. Influence of canopy characteristics of one‐seed juniper on understory grasses. Journal of Range Management 38:328–331. Senft, R. L., M. B. Coughenour, D. W. Bailey, L. R. Rittenhouse, O. E. Sala, and D. M. Swift. 1987. Large herbivore foraging and cological hierarchies. Bioscience 37:789–799. Shipley, L. A., A. W. Illius, K. Danell, N. T. Hobbs, and D. E. Spalinger. 1999. Predicting bite size selection of mammalian herbivores: a test of a general model of diet optimization. Oikos 84:55–68. Short, H. L., W. Evans, and E. L. Boeker. 1977. Use of natural and modified pinyon pine juniper woodlands by deer and elk. Journal of Wildlife Management 41:543–559. Sikes, R. S., and the Animal Care and Use Committee of the American Society of Mammologists. 2016. Guidelines of the American Society of Mammalogists for the use of wild mammals in research and education. Journal of Mammalogy 97: 663–688. Skousen, J. G., J. N. Davis, and J. D. Brotherson. 1989. Pinyon‐juniper chaining and seeding for big game in central Utah. Journal of Range Management 42:98–104. 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 20 of 21 �| 21 of 21 Smythe, S., D. Sanchez, and R. Mata‐Gonzalez. 2015. Temporal effects of mechanical treatment on winter moose browse in south‐central Alaska. Alces 51:135–147. Soule, P. T., and P. A. Knapp. 1999. Western juniper expansion on adjacent disturbed and near‐relict sites. Journal of Range Management 52:525–533. Stephens, G. J., D. B. Johnston, J. L. Jonas, and M. W. Paschke. 2016. Understory responses to mechanical treatment of pinyon‐juniper in northwestern Colorado. Rangeland Ecology & Management 69:351–359. Stoddard, M. T., D. W. Huffman, T. M. Alcoze, and P. Z. Fule. 2008. Effects of slash on herbaceous communities in pinyon‐ juniper woodlands of northern Arizona. Rangeland Ecology & Management 61:485–495. Tausch, R. J., R. F. Miller, B. A. Roundy, and J. C. Chambers. 2009. Pinon and juniper field guide: asking the right questions to select appropriate management actions. U. S. Geological Survey circular 1335. U.S. Department of the Interior, U.S. Geological Survey, Rangeland Ecosystem Science Center. Tiedeman, J. A. 1978. Phyto‐edaphic classification of the Piceance Basin. Colorado State University, Fort Collins, CO. Veldman, J. W., G. E. Overbeck, D. Negreiros, G. Mahy, S. Le Stradic, G. W. Fernandes, G. Durigan, E. Buisson, F. E. Putz, and W. J. Bond. 2015. Tyranny of trees in grassy biomes. Science 347:484–485. Wallmo, O. C., L. H. Carpenter, W. L. Regelin, R. B. Gill, and D. L. Baker. 1977. Evaluation of deer habitat on a nutritional basis. Journal of Range Management 30:122–127. Watkins, B. E., C. J. Bishop, E. J. Bergman, A. Bronson, B. Hale, B. F. Wakeling, L. H. Carpenter, and D. W. Lutz. 2007. Habitat guidelines for mule deer: Colorado plateau shrubland and forest ecoregion. Mule Deer Working Group, Western Association of Fish and Wildlife Agencies. Webb, S. L., J. S. Lewis, D. G. Hewitt, M. W. Hellickson, and F. C. Bryant. 2008. Assessing the helicopter and net gun as a capture technique for white‐tailed deer. Journal of Wildlife Management 72:310–314. White, G. C., and B. C. Lubow. 2002. Fitting population models to multiple sources of observed data. Journal of Wildlife Management 66:300–309. Young, K. R., B. A. Roundy, and D. L. Eggett. 2013. Plant establishment in masticated Utah juniper woodlands. Rangeland Ecology & Management 66:597–607. Associate Editor: A. Rodgers. S UP P O R T I N G I N F O R M A T I O N Additional supporting information may be found in the online version of this article at the publisher’s website, including tables of ANOVA results, contrasts between different types of mechanical treatments, and contrasts between seeded and unseeded subplots within mechanical treatments. How to cite this article: Johnston, D. B., and C. R. Anderson, Jr. 2023. Plant and mule deer responses to pinyon‐juniper removal by three mechanical methods. Wildlife Society Bulletin e1421. https://doi.org/10.1002/wsb.1421 23285540, 0, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1421 by Colorado Parks & Wildlife, Wiley Online Library on [24/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License PLANT AND MULE DEER RESPONSE TO TREE REMOVAL � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Journal Articles Description An account of the resource CPW peer-reviewed journal publications Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Plant and mule deer responses to pinyon-juniper removal by three mechanical methods. Description An account of the resource Land managers in western North America often reverse succession by removing pinyon (Pinus spp.) and juniper (Juniperus spp.) trees to reduce fire risk and increase forage for wildlife and livestock. Because prescribed fire carries inherent risks, mechanical methods such as chaining, roller-chopping, and mastication are often used. Mechanical methods differ in cost and the size of woody debris produced, and may differentially impact plant and animal responses. We implemented a randomized, complete-block, split-plot experiment in December 2011 in the Piceance Basin, northwestern Colorado, USA, to compare mechanical methods and to explore seeding (subplot) interactions. We assessed vegetation 1-, 2-, 5-, and 6-years post-treatment, and mule deer (Odocoileus hemionus) response via GPS locations 3–8 years post-treatment. By 2016, treated plots had 3–5 times higher perennial grass cover and ~10 times higher cheatgrass (Bromus tectorum) cover than untreated control plots. Roller-chopped plots had both the highest non-native annual forb cover, and when seeded, the highest density of bitterbrush (Purshia tridentata), a nutritious shrub used by mule deer. Masticated plots had higher bitterbrush use during summer and fall, leaving less forage available for winter. Days of winter mule deer use from GPS point locations in chained and roller-chopped plots was ~70% higher than in control plots, while winter use in masticated plots was similar to control plots. Mule deer use appears related to a combination of hiding cover, resulting from residual woody debris, and winter forage availability. Roller-chopped plots provide the best combination of hiding cover and winter forage, but mastication or chaining, applied leaving dispersed security cover, may be better options at large scales or when invasive species concerns exist. Bibliographic Citation A bibliographic reference for the resource. Recommended practice is to include sufficient bibliographic detail to identify the resource as unambiguously as possible. Johnston, D. B. and C. R. Anderson Jr. 2023. Plant and mule deer responses to pinyon-juniper removal by three mechanical methods. Wildlife Society Bulletin e1421. https://doi.org/10.1002/wsb.1421 Creator An entity primarily responsible for making the resource Johnston, Danielle Bilyeu Anderson Jr, Charles R. Subject The topic of the resource Mule deer Pinyon pine Juniper Extent The size or duration of the resource. 21 pages Date Created Date of creation of the resource. 2023-02-25 Rights Information about rights held in and over the resource <a href="http://rightsstatements.org/vocab/InC-NC/1.0/">IN COPYRIGHT - NON-COMMERCIAL USE PERMITTED</a> Format The file format, physical medium, or dimensions of the resource application/pdf Language A language of the resource English Is Part Of A related resource in which the described resource is physically or logically included. Wildlife Society Bulletin Date Accepted Date of acceptance of the resource. Examples of resources to which a Date Accepted may be relevant are a thesis (accepted by a university department) or an article (accepted by a journal). 2022/11/11 Date Modified Date on which the resource was changed. 2022/09/11 Date Submitted Date of submission of the resource. Examples of resources to which a Date Submitted may be relevant are a thesis (submitted to a university department) or an article (submitted to a journal). 2021/10/14 Type The nature or genre of the resource Article