This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. Characteristics of Managed Forest Habitat Selected for Nesting by Merriam's Turkeys Brian F. Wakeling 1 and Harley G. Shaw2 I Abstract - Nest site selection by Merriam's turkey (Meleagris gallop avo memam/) was studied on the Mogollon Rim, Arizona, from 1987 through 1991. Compared with measured habitat availability, nests had higher shrub and deciduous tree seedling densities (e < 0.001). Nests also had more cover comprised of shrubs, deciduous trees, and rock (E < 0.001, E. < 0.001, and E 0.036, respectively). This cover averaged a greater height than at random sit~s (e 0.032). Green foliage volumes at nests werfJ greater (E< 0.05) and horizontal visibilities were lower (E < 0.001). Nest sites were selected in stands that had clumped understory (E 0.032) and overstory 0.003). Basal distributions (E < 0.001), and patchy forest canopies (e areas were greater on nest sites (E < 0.001). Steep slopes were selected (E. < 0.001) for nesting purposes and canyons were the selected landform (E. < 0.001). Although turkeys nested in logging slash, slash piles were avoided. = = = = To date, little timber harvest has occurred on slopes >40%, inadvertently protecting much nesting habitat. Timber treatments that promote small scale patchiness, such as uneven-aged management or group selection harvests, can emulate vegetational characteristics of nesting habitat. Leaving loosely scattered slash, especially near the base of trees, can provide suitable nest sites in stands with ~ 50% canopy coverage. Not all timber treatments, however, negatively influence turkey populations. In some cases timber harvest improved the suitability of tuIkey brood range (Mollohan and Patton 1991). Additionally, turkeys have used second growth timber extensively for feeding in some locations (Rumble 1990). Limited infonnation was available on nest site selection in southwestern Merriam's turkey range. Because timber halvest did not affect habitat selection by turkeys in a consistent manner (Scott and Boeker 1977, Mollohan and Patton 1991), loss of nesting habitat was considered to be a potential cause of reduced turkey numbers. Nesting habitat may be especially critical in affecting southwestern tUIkey populations. Generally, only adult (2 year old) hens nest in the southwest (Goemdt 1983, Crites 1988, Wakeling 1991, Stone 1993). In addition, relatively few hens live beyond the age of 3 (Wakeling 1991). We studied nesting habitat selection by Merriam's tUIkey to detennine how management activities affected habitat use. Funding for this project was provided through Federal Aid in Wildlife Restoration Act Project W-78-R. R. Day, lR. Wegge, M. Senn, D. Skinner, K. Berger, C.M. Mollohan, E. Ozog, L.M. INTRODUCTION Merriam's tUIkey populations in the southwest may have declined from historic levels (Shaw 1986, Green 1990). Alterations to forest habitat by land management practices probably contributed to the decline (Shaw 1986). Tutkeys select habitat characteristics based upon specific behavioral activity such as feeding or loafmg (Rumble 1990, Mollohan and Patton 1991). Manipulations to forest structure can reduce habitat suitability (Scott and Boeker 1977). 1 Brian F. Wakeling is a Research Biologst with the Arizona Game and Fish Department, 2221 West Greenway Road, Phoenix, AZ 85023. 2 Harley G. Shaw is a Wildlife Biologist with General Wildlife Services, P.O. Box 370, Chino Valley, AZ 86323. 359 Schiavo, C.A. Staab, C. Hart, B.S. Holt, W.A. Rosenberg, C.H. Lewis, and V. Fitzpatrick provided field assistance. We are grateful to R.A. Ockenfels for critical review of an earlier draft of this manuscript. CD. Chevalier made helpful suggestions that improved the poster presentation STUDY AREA 2 The 335 mi Chevelon Study Area (CSA) was located approximately 40 mi south of Wmslow, Arizona, along the Mogollon Rim. Elevations ranged from 5500 ft in the northern portion to 7900 ft in the southern portion Annual precipitation averaged 18.6 in, with 2 conc~ntrations, the first occurring during winter stonns in Janwuy through March, and the second during summer storms in July through early September (Natl. Oceanic and Atmos. Admin 1991). Five habitat associations were identified on the CSA based upon terrestrial ecosystem swveys (Laing et al. 1989). These associations were mixed conifer, ponderosa pine (Pinus ponderosa)-Gambel oak (Quercus gambelii), pinyon (P. edulis)-juniper (Juniperus spp.), aspen (Populus tremuloides), and forest meadow associations (fig. 1). Mixed conifer associations were dominant above 7600 ft, and extended along east facing slopes and drainages. This habitat included Douglas-fir (Pseudotsuga menziesii), white fir (Abies conc%r), limber pine (Pinus jlexilis), and Rocky Mountain maple (Acer glabrum). Ponderosa pine dominated west facing slopes below 7600 ft and above 6000 ft. Below 6000 ft, pinyon-juniper was dominant with ponderosa pine stringers in drainages. Gambel oak occurred in all associations, in pockets in the mixed conifer and pinyon-juniper associations, and as a widespread conspecific with ponderosa pine. At elevations below 7000 ft, pinyon and alligator juniper (Juniperus deppeana) became increasingly abundant. Logging and grazing have been and remain major commercial land uses on the CSA. Logging began in the late 1930's and most ponderosa pine stands on level terrain have been logged at least once. Little logging has occurred on steeper slopes of major canyons. Legend § Mixed Conifer fSZl Ponderosa Pinel Gambel Oak • • Ponderosa Pinel Pinyon-juniper ~ Pinyan/ Juniper R D Forest Meadows Aspen Water Figure 1. - Vegetation associations present on Chevelon Study Area, Arizona (based on Laing et at 1984). were not approached for 1-2 weeks to prevent premature nest abandonment. Nests were then located and monitored daily to detennine dates that hens left nests. Habitat Mensuration METHODS Habitat characteristics of each nest site were measured after the hen and brood abandoned the nest area following predation or hatching. Percent slope was measured with a clinometer. Landfonn was classified as minor canyon «200 ft wide), major canyon (~200 ft wide), or ridgetop-flat. Canopy structure was classified as single storied, multiple storied-uniform, or multiple storied-clumped (i.e. gave an uneven-aged appearance). U nderstOlY and overstory were classified as clumped or evenly distributed. Stem densities of shrubs and deciduous seedling «1 in diameter breast height [DBHD, sapling «1 to 5 in DBR), and mature (>5 in DBH) trees were determined on a 0.01 ac circular plot centered on the nest. Stem density of conifers was Merriam's tuIkey hens were captured during winters of 1987 through 1990, using box traps, drop nets, and rocket nets (described by Wakeling 1991). Hens were equipped with motion-sensing backpack radio telemetty units (felonics, Mesa, AZ and AVM Electronics, San Francisco, CA) and released at the capture site. Nest sites were located by monitoring radio instrumented hens ~2X weekly. Hens were suspected of nesting when >2 consecutive locations were within 114 mile of each other and motion-sensing transmitters indicated inactivity. Inactive hens 360 .u measured on a 0.1 ac circular plot, also centered on the nest. Basal area was detennined from DBH measurements taken on all conifers encountered on this plot. Four 25-ft line intercept transects were established, each radiating from site center at right angles to one another. The first transect was randomly oriented. These transects were used to determine canopy cover, within 18 in of ground, from rock, down wood, grass, foms, shrubs, deciduous trees, and coniferous trees. The distance from which a tuIkey silhouette, placed at site center, could no longer be seen was detennined in 4 directions, parallel to line intercept transects. Average height of cover was ocularly estimated. Green foliage 'volume was detennined at each site according to MacArthur and MacArthur (1%1). Identical measurements were -taken at random plots to represent habitat availability (Marcum and Loftsgaarden 1980). We generated random Universal Transverse Mercator coordinates by computer and plotted them on 7.5' USGS topographic maps. Once a point w~ located on the ground, we stepped off a random distance in a random direction This endpoint was considered random plot center. Table 1. - Mann-Whitney values, probabilities, and mean values for habitat parameters at nest and random sites on the Chevelon Stud~ Area. Mean Mean Nest Random P Habitat Parametefi 11 <0.001 5960 1017 485.5 Shrub Density Deciduous <0.001 2228 169 524 Seedling Density Deciduous 0.006 300 10 726 Sapling Density Mature Deciduous Tree 0.006 120 7 728 Density Conifer Tree 0.193 50 60 808.5 Density 10.8 503 0.010 6.0 Grass Cover 7.3 545.5 0.031 4.4 Forb Cover 0.036 14.4 8.1 Rock Cover 552.5 Down Wood 755 0.896 12.7 12.4 Cover Deciduous Tree <0.001 0.1 408 1.7 Cover Conifer Tree 718.5 0.621 2.0 2.0 Cover <0.001 3.0 0.7 394.5 Shrub Cover <0.001 49 143 111 Silhouette Visibility 187 <0.001 54 18 Slope Data Analysis The Mann-Whitney 11 test was used to test differences in non-oonnal continuous data (Zar 1984:138). We used Student t-tests to test for differences between nonnal continuous data (Zar 1984:126). Chi-square contingency table analysis was used to test categorical data (Zar 1984:62). Multiple categories were evaluated using Bonferroni simultaneous confidence intelValS (Neu et al. 1974). All tests were considered significant at ~ ~ 0.05 with the exception of Bonferroni confidence intelValS. Because these confidence intervals take into account simultaneous tests which affect individual alpha levels, the . overall alpha level was set at ~ ~ 0.1 (Byers et al. 1984). aDensities are presented per acre, cover as percent canopy cover between 0-18 in of ground, visibility In ft, and slope as percent Feet Above Nest Bowl 100 90 80 70 60 50 40 30 20 RESULTS 10 o Habitat parameters were measured at 67 nest sites and 29 random plots. Shrub and deciduous seedling, sapling, and mature tree densities were higher at nest sites than at random plots (Table 1). Nest sites also had more 0-18 in cover comprised of shrubs, deciduous trees, and rock (Table 1). This cover averaged a greater height at nest sites (mean = 10.6 in) than at random plots (mean = 8.2 in) a-test, ~ = 0.032). At nest sites, green foliage volumes below 15 ft were greater (Fig. 2) and horizontal visibility distances shorter (Table 1) than at random plots. Nest sites were selected in stands that had clumped understory and overstory distributions (Table 2). Multiple storied-clumped forest canopies were favored (Table 2). Basal areas were greater in stands at nest sites (mean = %.1 triac) than at random sites (mean = 57.0 ft2/ac) a-test, ~ < 0.001). Steep slopes were 0.05 0.1 0.15 0.2 0.25 0.3 0.35 Foliage Volume (sq ft/cu ttl - - Nest Sites -t- Random Plots Figure 2. - Foliage volume (MacArthur and MacArthur 1981) on nest and random sites on the Chevelon Study Area, Arizona. abcdPoints With the same letter are significantly different from each other (I-test, f. 0.05). selected (Table 1) for nesting pwposes and canyons were the selected landfonn (Table 2). Although tuIkeys frequently nested in scattered logging slash, slash piles were only used twice. Less grass and fom cover was found at nest sites than at random plots (Table 1). No differences could be detected between nest sites and random plots in conifer tree densities or canopy cover from conifer trees or down wood. 361 Table 2. - Nest and random site proportions, Bonferroni confidence intervals around nest site proportions, Chi-square values, and probabilities for habitat parameters on the Chevelon Study Area. Habitat Parameter Observed Available CI X2 f. Landform Minor Canyon Major Canyon Ridgetop-Flat n Understory Distribution Even Clumped 0.473 0.400 0.127 55 0.069 0.138 0.793 29 0.330-0.616 0.259-0.541 0.031-0.223 0.222 0.778 0.448 0.552 28 0.111-0.333 0.667-0.889 0.127 -0.355 0.645-0.873 54 0.448 0.552 29 0.000 0.264 0.736 53 0.071 0.572 0.357 27 0.000-0.000 0.135-0.393 0.607 -0.865 n 54 Overstory Distribution Even Clumped 0.241 0.759 n Canopy Structure Single Story Multi-Uniform Stories Multi-Clumped Stories n 37.07 <0.001 4.58 0.032 6.72 <0.001 12.02 0.003 DISCUSSION MANAGEMENT IMPLICATIONS In our study, steep slopes were selected for nesting. Merriam's turkey appear consistent in selection for steep slopes throughout their range (Goemdt 1983, Mackey 1984, Crites 1988, Hengel 1990). This selection may assist nesting turkeys in eluding detection by ground predators. Additionally, hens may take advantage of the incline to gain flight if disturbed. Other characteristics selected at nest sites in our study seem to provide hens with hiding cover. Greater amounts of shrub and deciduous vegetation, rock, and green foliage volumes function to obscure nesting hens. Obscurity and camouflage would help hens avoid predator detection while nesting. Merriam's turkey consistently select nest sites that provide greater cover (Goemdt 1983, Hengel 1990, Rumble 1990). Although turkeys selected nest sites in stands that had clumped multiple canopies, it is difficult to determine whether this was a true selection for canopy characteristics or selection for steep slopes, where these characteristics predominated. The importance of clumped overstories is not readily apparent. Clumped overstories may simply be indicative of suitable understories. Nest sites in our study not located in canyons were frequently in stands that had been treated with a group selection halvest, yielding a clumped overstory. While steep slopes were selected for, clumped overstories may also be important, although to a lesser degree. Given adequate availability of other important features, such as water and suitable brood habitat, we believe Merriam's turkey select nesting habitat based upon 1) slope steepness and 2) suitable hiding cover. 1\ukeys select nest sites in habitats that are steep (>30% slope) and have short (<70 ft using a turkey silhouette) horizontal visibility distances. Characteristics of habitat where nests are' generally found include modemte to high basal area of conifer trees and high densities of shrubs and deciduous trees, arranged in a small scale mosaic. For management purposes, horizontal visibility distances is simpler to evaluate than cover composition and quantity. The distance at which a turkey silhouette is obscured is an appropriate mpid measure of habitat suitability for nesting purposes. Mollohan and Patton (1991) found a correlation (r = 0.853) between horizontal visibility distance of a turkey (HVDt) and that of a person (HVDp). The relationship between the 2 measures was HVOt = 0.39 X HVDp. Thus, horizontal visibility distance at which a person is obscured can be used to evaluate habitat suitability for nesting. Obscuring cover may be comprised of herbaceous vegetation, shrubs, trees, steep slopes, or rocks. Whenever possible, timber treatments should be avoided on slopes>30%. Tunber treatments that leave modemte basal areas (70-90 ft?lac) in small (1-4 ac) patches favor retention of nest site characteristics. Scattering logging slash following harvest, mther than leaving slash piles, will result in habitat more suitable for nesting. Any treatment that increases horizontal visibility distance will reduce site suitability for nesting. Our results are consistent with nesting habitat recommendations suggested by Hoffman et al. (1993). Timber treatments that result in characteristics described for nest sites can be used to retain or create suitable nesting habitat. Treatments such as group selection harvests or uneven-aged management appear to facilitate this management goal. Even aged management approaches appear less suitable. 362 Mackey, D.L. 1984. Roosting habitat of Merriam's turkey in south-central Washington Journal of Wildlife Management. 48: 1377-1379. Marcum, C.L.; and Loftsgaarden, D.O. 1980. A nonmapping technique for studying habitat preferences. Journal of Wildlife Management. 44: 963-968. Mollohan, C.; and Patton, D.R. 1991. Development of a habitat suitability model for Meniam's turkey. Arizona Game and Fish Department; Technical Report Number 9. 217 pages. National Oceanic and Atmospheric Administration. 1991. Arizona climatological data. Volume 94. Neu, C.W.; Byers, G.R.; and Peek, J.M. 1974. A technique for analysis of utilization-availability data. Journal of Wildlife Management. 38: 541-545. Rumble, M.A. 1990. Ecology of Merriam's turkeys (Meleagris gallopavo merriam i) in the Black Hills, South Dakota. PhD. Dissertation; University of Wyoming; Laramie. 169 pages. Scott, V.E.; and Boeker, E.L. 1977. Responses of Merriam's turkey to pinyon-juniper -control. Journal of Range Management. 30: 220-223. Shaw, H.G. 1986. Impact of timber harvest on Meniam's turkey populations. A problem analysis report. Arizona Game and Fish Department; Federal Aid in Wildlife Restoration Report. 18 pages. Stone, S.B. 1993. Habitat selection and use by Meniam's turkey in the Prescott National Forest, Arizona. M.S. Thesis; University of Arizona; Tucson 102 pages. Wakeling, B.F. 1991. Population and nesting characteristics of Meniam's turkeys along the Mogollon Rim, Arizona. Arizona Game and Fish Department; Technical Report Number 7. 48 pages. Zar, J.H. 1984. Biostatistical analysis. Prentice hall; Engelwood Cliffs; New Jersey. 718 pages. LrTERATURE CITED Byers, C.R.; Steinhorst, R.K.; and Krausman, P.R. 1984. Crlarification of a technique for analysis of utilization-availability data. Journal of Wildlife Management. 48: 1050-1053. Crites, M.l 1988. Ecology of the Merriam's turkey in north-«ntra1 Arizona. M.S. Thesis; University of Arizona; Tucson 59 pages. Goerndt, D.L. 1983. Meniam's tuIkey habitat in relation to grazing and timber harvest in southcentral New Mexico. M.S. Thesis; New Mexico State University; Las Cruces. 96 pages. Green, H.P. 1990. Long tenn population trends and habitat use by Merriam's turkey on summer range in the White Mountains, Arizona. M.S. Thesis; Northern Arizona University; Flagstaff. 108 page~ Hengel, D.A. 1990. Habitat use, diet, and reproduction of Merriam's turkeys near Larnmie Peak, Wyoming. M.S. Thesis; University of Wyoming; Laramie. 220 pages. Hoffman, R.W.; Shaw, H.G.; Rumble, MA.; Wakeling, B.F.; Mollohan, C.M.; Schemnitz, S.D.; Engel-Wilson, R; Hengel, D.A. 1993. Management guidelines for Meniam's wild turkeys. Division Report Number 18; Colorado Division of Wildlife; F9rt Collins. 24 pages. Laing, L.; Ambos, N.; Subirge, T.; McDonald, C.; Nelson, C.; and Robbie, W. 1989. Terrestrial ecosystem survey of the Apache-Sitgreaves National Forests. USDA Forest SelVice; Southwestern Region 453 pages. MacArthur, R.H.; and MacArthur, J.W. 1961. On bird species diversity. Ecology. 42: 594-598. 363 -(> u.s. OOVERNMENI' PR.INl']]'.K; OFFICE: 1994-575-629/05084