This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. Commercial Fuelwood Harvesting Affects on Small Habitats in Central Arizona

William H. Kruse

Abstract-In a central Arizona fuelwood harvest area, 75 percent by burning affects the newly formed slash habitat for small of the overstory was cut in a commercial harvest, resulting in large and plants and also reorganizes the nutrient base quantities of residual logging debris that altered habitat for many stored in the slash (Harrington 1989, Covington and DeBano wildlife species. Small mammals have intricate roles in ecosystem 1988). Natural decomposition of residual slash provides a function, and current fuelwood management practices have para­ slower and more complete return of nutrients to the system, doxical affects on small mammal habitats. In a small mammal while providing the protective effects of slash. J uni per slash, study, no differences in total capt,ured were detected among unlike pine slash, decomposes at a slower rate. treatment plots. Immediately following overs tory reduction or re­ Small mammal populations are impacted by overstory moval in 1992 and 1993, differences among species ca ptures, s pecifi­ disturbances (Turkowski and Reynolds 1970) while on-the­ cally deer mouse, increased significantly. The increased capture ground slash causes an increase in abundance of some rate remained significant throughout the remainder of the study. species regardless of overstory condition (Severson Pinyon mouse captures declined significantly immediately follow­ 1986). Kruse and others (1979) found that when the over­ ing treatments, but were not detectablydifferent from pretreatment story was removed or reduced, that preferred the levels a year later. woodland condition were fewer in number than those on the treated areas. This small mammal research was part of an effort to study the effect of nutrient cycling, other wildlife, and wood prod­ Commercial fuelwood harvesting generates greater and uct management on soil, water, tree, and range resources in more concentrated slash and has more affect on microsite pinyon-juniper woodlands. This paper discusses the effects conditions than noncommercial fuelwood cutting. This has of commercial fuelwood harvesting in an old-growth or late prompted some central Arizona USDA Forest Service Ranger seral pinyon-juniper woodland on small mammal popula­ Districts to assess the effects of commercial fuelwood re­ tions in central Arizona. moval and slash disposal, particularly by burning. Specifi­ cally, removal of slash habitat through burning is a concern. Fortunately, commercial harvest permits, however, can pro­ Study Area ______vide detailed slash management directions to meet specific management objectives. Burning is usually not performed The Heber Ranger District, Apache-Sitgreaves National by the fuelwood permit holder but instead is included in Forest in central Arizona was the study-site area. Average Forest Service management plans. tree basal area (diameter measured at root crown) was Fuelwood removal and slash management also affects 23.2 ± 5.4 m2/ha, which produced 35.3 ± 12.7 m3/ha of fuel­ microsite nutrient cycling, miderstory production (specifi­ wood (Kruse and Perry 1995). One-seed juniper (Juniperus cally protecting forages from large ungulate grazing), and monosperma) was the dominant species (54 percent). The regeneration of overstory species. Small mammal popula­ second most dominant tree (25 percent) was Colorado pin­ tions are also affected by removal of overstory, understory yon (Pinus edulis), followed by alligator j uni per (J. deppeana) composition and structure change, and slash accumulation (13 percent). Ponderosa pine (Pinus ponderosa) occasionally and subsequent manipulation. Basic ecological information occurred on moist sites (8 percent). Mean pretreatment is needed to support current harvesting plans (Gottfried canopy cover was approximately 40 percent, while the mean 1987). The least understood management option has been annual herbaceous and woody plant potential productivity slash disposition (Severson 1986, Baker and Frischknecht was approximately 562 kg/ha. 1973). The study area is relatively flat, dissected by several small Retention or removal of slash provides or eliminates ephemeral drainages. Elevations are between 2,000 and specific habitat characteristics for certain small mammals. 2,060 m. The primary soil subgroups, derived from lime­ In addition, retention or removal of slash affects the poten­ stone, are Lithic Ustochrepts, Udic Haplustalfs, and Typic tial protective cover for emerging new plants. Slash removal Eutroboralfs. The mean annual precipitation is between 34 and 46 cm.

Methods In: Monsen, Stephen B.; Stevens, Richard, comps. 1999. Proceedings: ecology and management of pinyon-juniper communities within the Interior West; 1997 September 15-18; Provo, UT. Proc. RMRS-P-9. Ogden, UT: U.S. Field Methods Department of Agriculture, Forest Service, Rocky Mountain Research Station. The study area consisted of 33 units, 4 ha in size. Three William H. Kruse is Range Scientist (retired) USDA Forest Service, Rocky Mountain Research Station. The Southwest Forest Sciences Complex, units were treated with silvicultural prescriptions. Thirty 2500 S. Pine Knoll Dr. Flagstaff, AZ 86001. 4 ha study units were grouped into five blocks representing

USDA Forest Service Proceedings RMRS-P-9. 1999 215 Figure 1-Typical overstory, preharvest conditions of Figure 3-Type conversion. Slash accumulation but late seral old-growth pinyon- juniper woodland at Heber/ not burned following harvest of commercial fuelwood. Mud Tank small mammal study area. Noncommercial stems cut and also remain.

six overstory/slash treatments. Sixteen of the 30 were ran­ included were the silvicultural treatments and the commer­ domly selected for the small mammal study. Four of the cial harvest where the noncommercial stems were left uncut 6 overstory/slash treatments were replicated in the 16 units as advanced regeneration. Small mammal trapping was (Kruse 1995). conducted during July and August from 1990 through Trapping occurred on 4 overstory treatments: (1) controls, 1996. Before harvest, downed woody fuel was estimated at where the units were untreated (fig. 1); (2) burned, to 3.15 mtJha (Kruse and Perry 1995). Post harvest slash simulate a forest fire (fig. 2); (3) type conversion, where accumulation was estimated at 55.71 mtlha (fig. 3). fuelwood was harvested, the non-commercial residual trees Treatments were assigned randomly and were not neces­ cut, but slash was not burned (fig. 3); (4) type conversion, sarily contiguous; roads or drainage channels could separate where fuelwood was harvested, residual trees cut, and slash units within a given block. Harvesting began during fall! burned (fig. 4). Type conversion is clearcutting to convert a winter of 1991 and continued for 24 months. Burning com­ woodland to grassland. The two overstory treatments not menced when the slash was at least 2 years old. Treatment schedules are in Kruse (1995). A 100 m 2 trapping grid was located in the center of each unit. 8 x 10 x 25 cm Sherman live trap was placed at each grid point, 10 m x 10 m apart. At alternate points, a 10 x 12 x 40 cm Sherman live trap was placed near the smaller one. The bait was a mixture of chicken scratch and rolled oats. Each unit was sampled yearly with 150 traps for 3 nights and 2 days. Physical measurements were taken and recorded for each , then they were toe clipped

Figure 2-Burned standing green woodland to simulated wildfire for type conversion (clearcutting to Figure 4-Type conversion. All overstory harvested or convert a woodland to grassland). cut, and slash burned.

216 USDA Forest Service Proceedings RMRS-P-9. 1999 Table 1-Species captured and percent of composition. 20

Species Common name Composition 16 Percent truei pinyon mouse 42 Peromyscus maniculatus common deer mouse 37 12 Eutamias dorsalis cliff chipmunk 9 Neotoma albigula white-throated wood rat 7 8 Neotoma mexicana Mexican wood rat 2 Peromyscus boylei brush mouse 2 Sylvi/agus auduboni desert cottontail 4 Spermophilus variegatus rock squirrel <1 Neotoma stephensi Stephens wood rat <1 0 Dipodomys ordi Ord kangaroo rat <1 90 91 93 94 95 96 Microtus mexicanus Mexican vole <1 90 vs 91 P = 1.0 mean 90,91 vs.93 P = < 0.001 mean 90,91 vs.94 P = < 0.001 mean 90,91 vs.95 P < 0.001 and released. Recaptures were noted. Relative abundance = and species composition of small mammals live trapped and mean 90,91 vs.96 P = 0.004 released on the study area are in table 1. Figure 6-The deer mouse captures. Captures were significantly increased immediately following treatments Analysis Methods in 1992-93 (p < .001). and remained higher through the remainder of the study (p < .004) while the slash habitat Replicated study units among 4 blocks were selected was on the ground aging. Some of the units were burned randomly as the experimental design layout (Ludwig and before the 95 and 96 trapping periods but it is unclear Reynolds 1988). Blocks were based on similarity of pretreat­ whether the burning impacted the deer mouse captures. ment overs tory conditions and characteristics. The treat­ ment units included combinations of no burning or cutting, burning standing green, cutting and no cutting (fig. 1-4). among treatments were tested by analysis of variance with The small mammal study replicated these 4 treatments in years used as a repeated measure. Significance oftreatment each of 4 blocks. effects was assessed based on temporal interaction (pre vs. The null hypotheses was that small mammal capture post treatment) in treatment responses. Comparisons among rates did not differ among treatments. Capture differences treatments, for individual post treatment years were ad­ justed by Bonferoni correction to maintain Type I error at 0.05. Tests were performed on total captures, deer mouse 80r------~ captures, and pinyon mouse captures only. The 1992 data _ Untreated ...... '1iiiiii"'T~~~t~d" ...... '" was also omitted from the analyses due to conflicting treat­ 60 ment effects and application. Results ______40 Before 1992 harvest treatments, there were no differences among the unharvested and the to-be-harvested units for 20 total captures (fig. 5). Although year-to-year differences occurred, these data demonstrated the homogeneous na­ o ture (p = 1.0) of all units in 1990 and 1991. Analysis of post 90 91 93 94 95 96 treatment responses, however, indicated significant treat­ ment effect (F= 7.17, p < 0.001). Following treatment, a 90 vs 91 P = 1.0 significant difference among treatments in 1993 was demon­ mean 90,91 vs. 93 P = < 0.001 strated (F = 5.79, SE 5.47, p < 0.001). Total captures treat­ mean 90,91 vs. 94 P = 0.537 ment effects were not significant for 1994 (p = 0.536), 1995 mean 90,91 vs.95 P = 1.0 (p = 1.0), and 1996 (p = 1.0). Deer mouse captures on the harvested and the unhar­ mean 90,91 vs. 96 P = 1.0 vested units (fig. 6) also showed the similarity among all units before treatment (p = 1.0), and the significant differ­ Figure 5-There were no differences in captures be­ ence between treatments in 1993 (p < 0.001). Unlike total cap­ tween unharvested and to-be- harvested units during the tures, differences for deer mouse captures were significant pre-treatment period (p > .5). Total captures of all spe­ for 1994 (p < 0.001), 1995 (p < 0.001), and 1996 (p = 0.004). cies significantly increased immediately following treat­ Half of the harvested units were burned in 1995 and 1996. ment in 1992-1993 (p < .001). but were not detected Pretreatment tests on pinyon mouse captures among all again (p > .5). study units were again similar (p = 1.0). Converse to the deer

USDA Forest Service Proceedings RMRS-P-9. 1999 217 mouse, pinyon mouse captures (fig. 7) significantly reflected 35 a negative effect following treatment in 1993 (p =0.052). This ...... IMJ ..l)n~~~?:t.~9; ...... negative effect for the pinyon mouse between the harvested 30 _ Treated and non-harvested units was not evident in 1994 (p = 0.322) 25 but was evident in 1995 (p =0.017) and 1996 (p = 0.040). The 1995 and 1996 harvest unit data included some units that 20 were burned in 1994 and 1996; 1996 was before sampling. 15 Discussion 10 ------5 This study was designed to evaluate fuel wood harvesting affects on small mammal capture rates following the harvest o 90 91 93 94 95 96 of old-growth or late-seral pinyon-juniper woodland over­ story. The treatments were 1) overstory removal; 2) creation 90 vs 91 P = 1.0 of large quantities of slash that was generally lopped and mean 90,91 vs.93 P = 0.052 scattered, although some piling was _done to allow cutting, mean 90,91 vs.94 P = 0.322 loading, and hauling, and 3) eventual slash burning. Data mean 90,91 vs.95 P = 0.017 were insufficient to determine which treatment activity had the greater affect on the small mammal capture rates. mean 90,91 vs.96 P = 0.040 Therefore, the "treatment" units included overstory removal, slash deposition, and some were burned. These "treatment" Figure 7-Pinyon mouse captures. Captures were sig­ units were tested against the uncut units. nificantly decreased immediately after treatments were Although not tested, it appeared that: 1) woodrat (Neotoma implemented in 1992-93 (p = .05), but not detectably lower a year later (p .32). Lower captures on the spp.) middens burned; 2) numbers of captures of all species = treated units were again apparent in 1995-96 (p < .04). may have decreased following burning; and 3) captures of brush mice (Peromyscus boylii) appeared to be greater on unburned slash units. Initially, both Peromyscus species responded positively or negatively to the harvest treatment and some differences continued for the following 3 years. References ______Half of the harvested units were burned in 1994 and 1995. Baker, Maurice F. and Neil C. Frischknecht. 1973. Small mammals Burning standing green, during a wildfire simulation, oc­ increase on recently cleared and seeded juniper rangeland. J. curred in 1994 and 1996. The study suggests that overstory Range Manage. 26: 101-103. removal and slash accumulation probably had more benefi­ Covington, W. Wallace and Leonard F. DeBano. 1988. Effects of fire on pinyon-juniper soils. pp 78-85. In Proceedings: Effects of Fire cial effects for the deer mouse, specifically, and to other species in Management of Southwestern Natural Resources. Tucson, as well. These 2 effects were detrimental to the pinyon Arizona. Nov. 14-17, 1988. mouse, however, particularly in the year following cutting. Gottfried, Gerald J. 1987. Regeneration of pinyon. In: Proceed­ This study corroborates earlier work showing that overstory ings-pinyon-juniper conference; 1986 January 13-16; Reno, NV. Ogden, UT: U.S. Department of Agriculture, Forest Service, is important to the pinyon mouse (Severson 1986). Burning, Intermountain Research Station: 249-254. which would not have correctea the lack of overstory condi­ Harrington, M.G. 1989. Soil nutrient and pinyon seedling response tion, was of further detriment to the pinyon mouse. to fire severity, p.143-147. In: Proceedings of the 10th conference on fire and forest meteorology, Maciver, D.C., ed. and others Fewer captures of all species were evident in the latter Chalk River, Ont.: Forestry, Canada, 1989,469 p. 3 years that included the burning portion of the study. Tests Kruse, W. H. 1995. Effects offuelwood harvesting on small mammal between treated and untreated units suggest little signifi­ populations in a pinyon-juniper woodland, p91. In; Douglas W. Shaw; Earl F. Aldon; Carol LoSapio, tech. Coords. Desired future cant difference, while treatment effects to the deer mouse still conditions for pinyon-juniper ecosystems; proceedings of the appear beneficial. Field observations suggest that unburned symposium; 1994 August 8-12; Flagstaff, Arizona. GTR-RM-258. units still contained viable woodrat middens and higher Fort Collins, CO. USDA-For. Serv., Rocky Mtn. For. & Range capture numbers for other species such as the cliff chipmunk Exp. Stn, 226 p. Kruse, W. H. and H. M. Perry, 1995. Ecosystem management and (Eutamias dorsalis). Baker and Frischknecht (1973) found fuelwood harvesting in an "old growth" pinyon-juniper woodland, no effect from slash on mice populations, except where it was p219.In; Douglas W. Shaw; Earl F. Aldon; Carol LoSapio, tech. windrowed, while Severson (1986) found that treatments Coords. Desired future conditions for pinyon-juniper ecosystems; proceedings of the symposium; 1994 August 8-12; Flagstaff, leaving slash benefited woodrats and brush mice following Arizona. GTR-RM-258. Fort Collins, CO. USDA-For. Serv., Rocky canopy removal. Therefore, the detrimental burning effects Mtn. For. & Range Exp. Stn, 226 p. in this study could have off-set the beneficial effects of the Kruse, W. H., R. P. Balda, M. J. Simono, A. M. Macrander, and C. D. slash. In addition, slash provided improved site protective Johnson. 1979. Community development in two adjacent pinyon­ juniper eradication areas twenty-five years after treatment. J. characteristics for plant regeneration and development. Environ. Manage. 8: 237-247. Population densities of small mammals are related to Ludwig, John A. and James F. Reynolds. 1988. Statistical ecology. overstory adjustments and/or slash composition in assess­ New York: John Wiley and Sons. 337 p. Severson, K. E. 1986. Small mammals in modified pinyon-juniper ing site productivity and quality. This study contributes to woodlands, New Mexico. Journal of Range Management. 39: 31-34. small mammal and basic ecological information to improve Turkowski, F. J. and H. G. Reynolds. 1970. Response of some rodent guidelines for harvesting fuel wood in pinyon-j uni per ecosys­ populations to pinyon-juniper reductions on the Kiabab Plateau, tems in the Southwestern United States. Arizona. Southwest. Natur. 15:23-27.

218 USDA Forest Service Proceedings RMRS-P-9. 1999