USE OF FOREST OPENINGS BY ELK IN NORTHERN MECHIGAN Thesis {‘or ”'00 Degree of M. S. MKCHIOAN STATE UNIVERSE“ James E. Knight 3975 «THESIS q; '1 ‘7 ABSTRACT USE OF FOREST OPENINGS BY ELK IN NORTHERN MICHIGAN BY James E. Knight ‘ iFactors Influencing the use of forest openings by elk (Cervus canadensis) were investigated on the property of a private-recreation club in Montmorency County, Michigan. Data gathered from track plots indicated several variables that influenced the degree of elk use an Opening would receive. Elk use of openings increased in fall, decreased during winter when snow accumulation reached l2 inches (30 cm) and increased again following snow breakup in early May. Excessive summer temperatures corresponded with a decrease in elk activity within openings. Large Openings were used most in fall and springzg small openings were used most in winter and all size Openings were used during summer. High percentages Of Open land in the 6ND acres (259.2 ha) surrounding a study opening were associated with high track counts. Openings having a slope less than l5 degrees received the highest elk use. Although different communities were utilized, openings favored by elk were also favored by deer. There was a definite relationship between difficult human access to a given Opening and high elk activity. Openings used most had dense vegetation somewhere on the edge, and at least some browsable species available. USE OF FOREST OPENINGS BY ELK q) IN NORTHERN MICHIGAN Q) I By James E. Knight A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife I975 ACKNOWLEDGMENTS I would like to thank Mr. James Porritt and the members of Canada Creek Ranch for their cooperation and hospitality. I would like to thank my graduate committee: Dr. Leslie Gysel, Dr. Rollin H. Baker and Dr. Jonathon w. Wright for their help and guidance. I am thankful to District 5 Department of Natural Resources' wildlife biologists: Mr. Bob Strong, Mr. Jeff Greene and Mr. Tom Carlson for their help and suggestions. A special thanks to my wife, Denise, for her help collecting data, for her patience with me and for her moral support throughout the study. TABLE OF CONTENTS. ACKNOWLEDGMENTS LIST OF TABLES. LIST OF FIGURES . INTRODUCTION. Study Area METHODS . ‘ Determination of Elk Use . Weather Data . Determination of Opening Characteristics Statistical Analysis RESULTS . DISCUSSION. . Temporal Variation . Weather. . Temperature . Snow. . Opening Characteristics. Size. . Open Land . Slope . Deer . . Human AccesS . Vegetation . Density . . . Edge Composition Shrubs, Ferns, Herbs, Soil . . . . LITERATURE CITED. APPENDIX. and Lichen. Page ii (Du-PF? Table LIST OF TABLES Variables considered in the study Per cent of each tree species in the edge vegetation of each opening for all size classes Percentage of grasses, shrubs, ferns and lichen within each opening. Density of trees and shrubs in opening edge communities., All size classes included Physical characteristics of the ten openings containing sample plots. Soil pH and chemical composition in sample openings . Woody plants Identified in elk rumens, December 1964-1965 (Buss I967) Results of multiple regression analysis with number of elk tracks as the dependent variable and characteristics of the openings as the independent variables Page l2 l3 l4 22 36 37 Figure LIST OF FIGURES Portion of Canada Creek Ranch used in the study. Elk tracks by month for each opening . Elk tracks related to snow depth in Openings l and 2 . .. Seasonal observations of elk tracks in openings, l97h- I975 . Track observations and snow accumulation . Elk tracks in lO openings and maximum daily summer temperatures Relative trends of elk and deer tracks Results of multiple correlation.. Correlation coefficients for elk tracks and variables measured daily. Page ‘5 I9 2A 26 27 3O 38 INTRODUCTION - Moran (I973) first reported the importance of forest openings ‘for elk (Cervus canadensis) in Michigan. This importance has also been recognized in other states. 'In Montana the percentage of elk observed was greatest in forb or grassland communities during the entire year (Knight I970). Opening characteristics important to maximum elk-use have been investigated in other western states (Craighead _E _1. I973), but many of these characteristics do not apply in Michigan, where tOpography, sOiI, vegetation, weather and elk behavior patterns are different. Forest Opening characteriStics which habitat managers can control (i.e., size, shape, soil inadequacies and vegetational composition) are important. In addition, uncontrollable variables (i.e., temperature, precipitation, presence of competing wildlife, or moon phase) must be considered (Harper l97l). Knowledge concerning these characteristics becomes more essential as the need to create and maintain openings be- comes an important management tool. This paper presents observations and measurements on the utiliza- tion of forest Openings by elk and the factors which may affect this U58. Study Area The study area is a portion of Canada Creek Ranch, a recreation and hunting club of l3,000 acres (5265.0 ha). The ranch is located l3 miles (20.9 km) south of Onaway (Montmorency County) in the northeast portion of Michigan's Lower Peninsula. -2- The history of land use on the area in which Canada Creek is located is similar to that of other large areas in northern Michigan: logging, burning, clearing, farming, land speculation and consolidation have occurred in sequence. Around I920 several individuals acquired the land for cattle raising (Crossmanl970). Title to the land passed to a series of owners until Canada Creek Ranch Association took over in ' 1940. "Approximately llOO acres (4h5.5 ha) of the property is subdivided into lots. The remaining ll,900 acres (A8l9.5 ha) is uninhabited. According to a survey by the Michigan Department of Natural Resources in I975, this area provides habitat for 36 per cent of Michigan's entire elk head (letter dated ll March I975 from Thomas CarlSon, Michigan Depart- ment of Natural Resources, Atlanta, Michigan). IApproximately 2600 acres (526.5 ha) are in Open areas. Most of these openings resulted from a fire in I939 (Crossman I970), but some were created at log landing and yarding points. A few of the smaller Openings resulted where cold air drainages caused frostpockets. About 9300 acres (3766.5 ha) have a cover of second-growth hardwoods and conifers. Approximately 3000 acres (I2l5.0 ha) have been cut over in the past l5 years. The Stands range in age from 0 to A0 years in the aspen type and from I5 to 70 years in the conifer and hardwood types. The very dense and shrubby jack pine (Pinus banksiana) stands found extensively in the northern part of the ranch are atypical of the stands found elsewhere on the ranch. Most of the stands in the central and southern portions are of moderate denSity and of high growth rate. Soils in the area are primarily Roselawn sand and sandy loam and Grayling sand (Veatch I930). -3- The hydrology of the study area relates to Canada Creek, which runs north through the property and into the Black River. Almost all swamps and feeder streams on the ranch empty into the Canada. Major woody species found throughout the study area include: trembling aspen (Pppulus tremuloides), bigtooth aspen (E. grandidentata), red oak (Quercus rubra), white oak (g, alba), sugar maple (Acer saccharum), red maple (A, rubrum), beech (Fagus grandifolia), black cherry (Prunus serdtina), basswood (Tilia americana), red pine (Pinus'resinosa), white pine (P, 3123), jack pine (P, banksiana), white birch (Betula papyrifera), elm (01mg; spp.), juneberry (Amalanchier sp.), balsam fir (Abies balsamea), white cedar (Thuia occidentalis), black spruce (Picea mariana), raspberry (figbg§_sp.) and blueberry (Vaccinium spp.). Principal herbaceous ground cover includes several grasses (Egg spp.), hawkweed (Hieracium sp.), strawberry (Frageria sp.), bracken fern (Pteridium aguilinum), and sweet- fern (Myrica asplenifolis). METHODS Ten openings were selected to provide varying geographical characteristics (Table l) in the 7000 acre (2835 ha) study area. The western quarter and central area of the ranch (Figure I) provided the most constant elk numbers based on the author's observations from prior years. For this reason the sample openings were located in this area. Determination of Elk Use Elk activity was determined by observations in 96 track plots. Track plots were 2.25 square foot (.209 m2) areas where the sod was re- moved to expose soil. These plots were located in ICU foot (30.h6 m) grids throughout each opening, varying from A to 36 plots in the ID openings. Size of the plot was checked periodically and any ingrowing vegetation was removed. A stake was placed 6 feet (I.8 m) north of each plot to serve as a reference point during snow coverings. Track plots were checked on 299 days between I September I974 and 3l August I975. Elk and deer (Odocoileus virginianus) tracks were re- corded on tally sheets and the tracks were then erased. Direction of the tracks was also recorded to ensure that the plot had not been placed on a major elk trail. Care was taken to approach track plots from different directions to avoid creating a path the animals might tend to follow. .Weather Data Weather data were collected daily from two different stations. Maximum and minimum temperature, precipitation, and snow accumulation were recorded at Pigeon River State Forest Headquarters, located 7.5 miles -A- -5- Table I. Variables considered in the study. Recorded Daily Recorded One Time Only Number of elk tracks Size of opening Number of deer tracks Boundary to opening ratio Maximum temperature Amount of open land in surrounding area Minfmum temperature Average Slope of opening Amount of rain' Difficulty of human access to opening Amount of snow Density of edge vegetation Amount of accumulated snow Number of edge communities Wind speed Per cent edible edge vegetation Wind direction Edge species in each Opening Moon phase . Density of opening vegetation Opening ground species Soil pH and nutrients Soil type in opening 2 N 0 1: z > . o ‘ A C5 5 ‘3 4‘4— 7 - -Trails l-Opening Number One Mile , - 1 Figure l. Portion of Canada Creek Ranch used in the study -7- (l2.07 km) west of the ranch. Anemometrical data were collected at Phelps-Collins Air Base 28 miles (AA.9 km) east of the ranch. Moon phase data were also recorded on a daily basis. Determination of Opening Characteristics Size of each opening and the percentage of Open land in the 640 acres (259.2 ha) surrounding it was determined from aerial photos. A numerical value was given for each Opening to indicate the ratio of boundary to total area of the opening. This value was obtained by: “=25: . where Et is the boundary to Opening ratio, E is the entire boundary in feet and A is the area of the openings in square feet. Edge communities were subjectively classified according to succes- sional stage and Species composition. .Ten yards (9.I m) outward from ' the opening boundary was that part of the edge communities considered in this study. Beyond this distance light penetration is limited and effect of the opening on vegetation is greatly reduced. Density and composition of woody vegetation in each community was determined by measurements of each tree within l0 yards (9.I m) of the opening boundary. Boundary in this study is that point at which the first woody stem is encountered going outward from the opening. Stems per acre were Used to indicatethe density of each community. Percentage of edible species along the edge of each opening was recorded. A list (Table A-l) established by Buss (I967) was used to decide which species were utilized by elk. Only vegetation lower than I0 feet (3.0km) was conSidered browsable. Herbaceous vegetation of openings was sampled, using the point- centered quarter method as described by Dix (l96l). In applying this -8- method, a single pin was placed vertically into the soil, guided by a (small notch cut into the toe of the sampler's boot. The area around the point was then divided into quarters by drawing two imaginary lines through the pin; one line was drawn parallel with the line of traverse while the second was drawn at right angles to the first. Species name and distance from the point to the closest living shoot was recorded. Thus, each sampling unit consisted of four shoots and four measured distances. From these data the relative densities of each species was determined for each opening. All lines of traverse were parallel and ID feet (3.02 m) apart starting IO yards (9.I m) southeast of the extreme. southeast point of the opening and heading north. There was one sampling point every l0 feet along the line of traverse. All traverse lines ex- tended into the edge l0 yards beyond the opening boundary. Number of sampling points for each opening ranged from 485 to 5A25. Soil samples were taken (3 per acre) in each Of the ten openings and analyzed for pH, phosphorus, potassium, calcium and magnesium. Testing was done by Michigan State University Soil Testing Laboratory. Human accessibility was categorized by the condition and proximity of the closest trail. Openings having trails whichprovided access for any type automobile made up one category. Openings which could be reached Only by trails suitable for off-road vehicles comprised the . second category. The most difficult access, openings which had no trails ‘connected to them, made up the third category. Statistical Analysis All data cOIlected were analyzed at the Michigan State University Computer Center, using the SPSS integrated system of computer programs -9... (Nie _£_§Js I975). The Pearson correlation coefficient was used to measure the association between two variables. A multiple regression analysis was used to measure the relative importance of the independent variables and the approximate amount of variance explained by the inde- pendent variables. RESULTS The results of the multiple correlation analysis (Figure A-l). shows the association between variables measured daily and numbers of elk tracks observed in each opening. The results of the multiple regression analysis (Table A-2) show the relative importance of the independent variables and indicate the approximate amount of variance exptained by the independent variables. The results of the vegetative analysis for the edge (Table 2) and for the ground cover (Table 3) are listed for each opening. Potential of the edge as escape cover as well as the relative amount of browse species in each community is indicated in Table A. Other physical characteristics for each opening are compiled in Table 5. Opening l is in the northern section of the study area. Soil here is Roselawn sand. The opening is surrounded by extensive scrub jack pine found throughout the area. The north-facing slope of this opening is 2l degrees. Elk tracks were not observed in Opening l until after the first snowfall in November (Figure 2). Occasional use continued until spring, but summer activity was recorded only on one occasion. Only a small portion of the 6A0 acres (259.2 ha) surrounding Opening l was grassland areas (Table 5)- Opening 2 is in an area dominated by scrub jack pine northward and oak ridges to the south. Slope of the study area is 20 degrees south- erly and most of the surrounding area is moraines interconnected with shallow valleys. Soil in this area is Roselawn sand. -10- Table 2. opening for all size classes. -I]- Per cent of each tree species in the edge vegetation of each Species Opening I 2 3 A 5 6 7 8 9 l0 Pinus banksiana 75 55 2 - 2 I l0 9 - - ,5. Fesinosa l - - - 28 86 - - - - .3. strobus - - - - - 3 , - l - - Populus tremuloides - l5 l5 5 - - 25 30 NO NO P. grandidentata - - 23 IO - - 25 25 59 50 Quercus rubra 2A 30 IO 30 35 IO 20 IO l I Q, glb§_ - - IS. 20 35 - l0 l0 - 5 Acer rubrum - - l5 l0 - - - l0 - l A, saccharum - - IS 20 - - - 5 - I Prunus serotina - - 5 5 - ~ l0 - ~ - -12- Table 3.. Percentage of grasses, shrubs, ferns and lichen within each Lichen IO opening. Species Opening 1 2 3 A 5 6 7 8 9 Io Pteridium aquilinum 75 30 35 70 85 - 32 8 l3 IO 30 Myrica asplenifolis IS 25 35 ll 2 A0 l0 7 3 l0 VacciOium spp. - 20 - 3 l0 1 2 - - - “Rubgg spp. - - lo I - - 8 l 5 2 Frageria sp. I 2 l 5 - - I 2 2 5 Hieracium sp. - - - l - l - 20 l - M25 SP. - 3 - 2 - - 3 - - - Gaultheria procumbens - - - - - I - - - _ Prunella vulgaris - - - - - - - 20 - - Asclepias sp. - - - - - - - l - l Chrysanthemum leucanthemum l - - 2 - e - 1 - ] Poaceae 7 20 I9 6 3 A7 12 3o 50 50 Trifolium sp. - -' - - - - - 5 30 - -13- Table A. Density of trees and shrubs in Opening edge communities. All size classes included. Opening Communitya Stems Per Acre I IC 1151 I N0 113 2 IC 1395 2 MA 101 2 N0 110 . - 3 SA . 9037 3 NA 96 3 No 1110 3 NH ‘ 183 1+ SA . > 7193 A N0 109 1+ NH 177 5 IC 1291 5 MC , 60 5 No 76 6 IC 1605 6 No ' , 107 7 SA 7131 7 N0 , ‘ , 130 _ 7 1c 1091 8 SA 11101 8 SH 8761 8 sc ' 11910 9 SA 12019 10 SA - 9011 10 NH 160 aCommunity designations: SA-sapling aspen; MA-mature aspen; SC-sapling conifer; IC-immature conifer; MC-mature conifer; SH-sapling upland hardwood; MH-mature upland hardwood;.MO-mature oak. -14- Table 5. Physical characteristics of the IQ Openings containing sample plots. Boundary Per Cent Degree Per Cent Opening Size to Opening of Section of Human b of Edge (ac.) Ratioa in Open Land Slope Accessibility Edible 1 1.12 .0231 8 21 1 20 2 1.29 I .0169 h 20 1 15 3 10.0A, .0101 A5 ' 15 3 A0 A ’ 1.38 .0158 27 7 2 35 5 .77 .0237 22 23 p i 2 2 6 I.86_ .Ol66 2H ll _ I l 7 2.12 .0157 25 9 3 60 8 .95 .019A 25 ' 10 2 l 99 9 1.32 .0426 26 ' 10 3 99 10 1.98 .0145 2A 12 1 95 E 8Determined by Et = Agwhere E is the ratio of boundary to total area of the Opening, E is the Opening boundary in feet, and A is the area in square feet. bNumerals denote increasingly difficult access. ..]5.. H e 10 Elk Tracks Elk Tracks .TTIIITHH -._‘._ _SONOIFMAMIIA SONOIFMAMIIA Month Month OPENING I OPENING? 10 10 1—7 1— e .: r5" . '1 .35 ., ET I 0 _ 3 _ SONOIEMAMIIA SONDIFMAMILA Month Month OPENING 3 OPENING 4 Figure 2. Elk tracks by month for each opening. -15- 0—. a . H11 _SON OF] EMA NII’I A Month OPENING 5 Elk Tracks .— O Elkvllacks l HINT-I STINIII'FNIANII IA Month OPENING 7 Figure 2 (continued). 10 1:5 F. SONOIFMAMIIII Month ' OPENING 6 10 P“ :s —-I w 0 Lil- 3 S OIIIII FINA M.|I,A Month OPENING a -17- IO 10 .2- '7 =3 3 ‘1‘ z _ 1 1:5 1 "5- _l a: , '5 a m ri— o H. __ o _n ,_ SONDIFMAMIJA SONDIFMAMIIA Month Month OPENING 9 ' OPENING to Figure 2 (continued). -l8- Elk track observations were similar to those in Opening I (Figure 2). Snow accumulation seemed to initiate movement to the cover provided by the extensive jack pine in this area (Figure 3). Opening 3 was the largest opening used in this study. Aspen re- production was encroaching the opening in some areas. These areas were interspersed with mature upland hardwood stands around the rest of the opening edge. The I5 degree west-facing sIOpe of Opening 3 was repre- sentative of the topography of the surrounding area. Forty-five per cent of the surrounding 6A0 acres (259.2 ha) was openings (Table 5).‘ The rest . of the section was mature upland hardwood and conifer stands. Soil type here was Roselawn sandy loam. Track observations in Opening 3 were constant for all seasons except winter (Figure 2). The decrease in tracks after permanent snow accumulation corresponded with an increase in use of some smaller openings and Openings with denser edge vegetation. Opening A was a clearing situated on a south-facing slope within a mature upland hardwood stand. A small area of sapling aspen was encroach- ing the south end of the opening. Soil was Roselawn sandy loam, which has a higher natural fertility and greater potential productivity than the Grayling and Roselawn sand found in some of the other study openings (Veatch I930). This could account for the high diversity of ground vegetation within the opening (Table 3). The low number of elk tracks observed in Opening A were relatively constant throughout the year (Figure 2). The small size and presence of winter cover south of the opening probably accounted for conStant use even after snow accumulated. -19- Y= I.I7jI- .0678X ”23.653 r=.808' ELK TRACKS ' N D 0 Opening I o Openingz 061016202630 SNOW DEPTH Figure 3. Elk tracks related to snow depth in Openings I and 2: -20- Opening 5 was located on an east-facing slope at the rim of a long valley. The low diversity of vegetation within and on the edge of the opening may be a reflection of the low content of moisture and the low fertility of the Grayling sand found here (Tables 2 and 3). _Woody vegetation found in this area is primarily mature hardwoods and conifers. Track counts in Opening 5 were lower than any other opening having greater than 20 per cent open land in the surrounding section. Use in thIS'opening was highest in the fall and spring and was lowest during the winter months (Figure 2). Opening 6 was almost entirely surrounded by an immature red pine plantation. The soil here, Roselawn sand, has been parched by repeated grass fires and nutrient content was very low. This I.86 acre (.75 ha) opening was situated on an ll degree south-facing slope. Elk tracks observed in Opening 6 were low during all seasons com- pared to other openings in the area (FigUre 2). The amount of open land in the surrOUnding sectiOn was average for the study openings in the central portion of the ranch (Table 5). Opening 7 was a secluded area surrounded primarily by seedling and sapling aspen and immature jack pine with mature oaks bordering the eastern side. The opening had an east-facing slope and was situated on Grayling sand. Opening 7 had high track counts for the year and was the only open- ing that had more use during the summer than during any other season (Figure 2). This could be due to the shade provided by the edge vegeta- tion, which would prevent dessication of the forage ground species. Opening 8 was an area of diverse vegetation, both within the Opening and on the edge (Tables 2 and 3). Roselawn sandy loam on an east-facing -2]- slope provided a substrate for the dense aspen and hardwood reproduction, which composed 90 per cent of the edge. The opening was situated in a section of 25 per cent open land. The high calcium in the soil here (Table 6) may contribute to the success of the diverse vegetation (Veatch I930). Elk track observations were relatively high in Opening 8 during all Seasons (Figure 2). Only Opening 9 had more tracks per unit area during the entire year (Figure 2). "Opening 9 was long and narrow, having a greater proportion of edge to total area than any of the other Openings (Table 5). The opening wound through a sapling aspen stand, created by a logging operation 7 years ago. The area surrounding this opening is a mixture of sapling aspen and mixed hardwoods. Clover (Trifolium spp.) was abundant in Opening 9 and high pH (Table 6) of the soil (ROselawn sandy loam) pro- vided an adequate substrate for this low acid tolerant species (Christenson and Doll I973). More elk tracks were observed per track plot in Opening 9 than in any other opening investigated. Use was high in all seasons, but dropped in winter (Figure 2). Opening lO was divided by a trail running north and south. Edge vegetation on the western side was primarily sapling aspen. The eastern edge was composed of sapling aspen and mature upland hardwoods. Soil was Roselawn sandy loam and had the highest pH and calcium content. The l2 degree sIOpe was south-facing. Total tracks for the year were high in Opening l0. However, winter and summer observations were low (Figure 2). -22.. Table 6. Soil pH and chemical composition in sample openings. 50” Pounds Per Acre Opening pH Phosphorus Potassium CalCium Magnesium I 5.3 38 I32 873 I88 2 A.8 57 2A 29], A2 3 5.6 A5 8A 582 73 “A S 5 38 36 . 727 82 5 5.0 20 A8 582 9A 6 5.5 22 A8 29I 52 7 S.l 3I A8 29I 63 8 5.6 I6 72 ll6A 83 9 5.9 62 .36 H36 52 l0 6.A 32 2A 2036 73 DISCUSSION Several of the variables investigated in this study seemed to have. no influence on the number of elk tracks observed. These variables included: wind, moon phase, boundary to opening ratio, rain and nutrient content of the soil. I The following is a discussion of those variables which did seem to have-some influence on the number of elk tracks observed. Temporal Variation Most changes in overall elk track counts were associated with changes in season (Figure A). Use of openings increased in the fall, correspond- ing with the peak of rutting activity in late October, decreased during winter, and increased.again following snow breakup in early May. Open- ings play an important nutritional role in spring when animals already stressed by the winter diet must quickly recover for the impending demands of coat change, calving, nursing, etc. (Moran I973). The indicated I increases in seasonal elk activity agree with data reported by Mackie (I970), COOp (I97I), Craighead _t_§l, (I973) and Moran (I973). Mid- summer use of openings in this study was higher than expected, but tapered off in late summer. Kirsh (I962) and Stevens (I966) also noted decreaSed use of exposed areas in late summer and attributed it in part to dessication of forage. Overall winter use of openings was inversely related to snow depth (Figure 5). Openings in dense coniferous communi- ties that received minimal use in other seasons showed a small increase in early winter (Figure 2). -23- 25 20 NUMBER OF ElII TRACKS IN 10 QPENINGS Figure A. St) 45 4C) .35 3C) 10 -2A- MON TN Seasonal observations of elk tracks in openings, l975-1975. -25- Weather Temperature A relationShip was indicated between elk tracks and maximum daily temperatures. When daily temperatures were above 800 F (260 C), elk use of openings dropped Sharply (Figure 6). When temperatures dropped, daily observations increased. Moran (I973) noted that on very warm days, the shady inside edge of the Openings were favored. This could account for the indicated reduced use of the opening proper. M Intensity of falling snow or total amount of snow from a single snow storm had no effect on elk observations. Elk pay little attention to snow in Michigan, and only extreme storm conditions drive them to temporary cover (Failing l95l). In this study the effect of individual storms was not noted; however, there was a sharp decline in elk tracks in openings after snow had accumulated to a depth of l2 inches (30.A cm) (Figure 5). (When the level of thawing snow again receded to this level, elk activity began increasing. Observations of elk in the openings at the time this threshold was reached indicate that elk utilize oniy sun- warmed areas, where snow was at the lowest levels. Some elk activity persisted in the openings throughout the wihter, regardless of snow accumulation. Most of this activity was movements associated with browsing. Utilization was noticed most in areas where drifting of snow had lowered levels so that elk could paw through what snow remained to feed on shrubs or grasses. -26- l00- — Accumulated Snow ---E|k Tracks (I C) l > o 1 1 1 1 n l l 1 1 I PERCENT OF THE MAXIMUM TRACKS AND SNOW Nov. Dec. Dec. Jan. Jan. Feb. Feb. Mar. Mar. Apr. Apr. 30 I5 30 I5 30 I5 30 I5 30 IO 30 DATE Figure 5. Track observations and snow accumulation. -27- IOO' MAVAWRVAVv/IVAVMW 80. E /‘.g I a ”\ ’ T‘s/s /’ \-a" O // \\\ ’l s . b—s‘"\v/ \’ \~, —— Maximum Daily Temperature --- Elk Tracks O I ' l l l l l l I I l L l 0715230 7l522A7I5238 DATE PERCENT OF THE MAXIMUM TRACKS AND TEMP Figure 6. Elk tracks in IQ openings and maximum daily summer temperatures. -28- Opening Characteristics Size The larger openings (Nos. 3, 6, 7 and IO) were used most during fall and in spring (Figure 2 and Table 5). Activity was highest after snow breakup, when exposed areas melted and grasses were made available. Moran (I973) found that in spring, elk made ”definite and predictable movements toward large openings, concentrating there throughout the green-up period.” In winter months, elk shifted to smaller coniferous openings (Nos. I, 2, 5 and 8), where some protection could be obtained from surrounding vegetation (Figure 2). During summer all sizes of openings were used (Figure 2). Edges were used almost exclusively around the larger openings and the vege- tative type of surrounding areas seemed to be more important than the size of the opening. One important factor provided by small openings during the summer is some protection of vegetation from dessication. Open Land The percentage of open land in the 6A0 acre (259.2 ha) section surrounding each opening seemed to be important. The seven openings that were in sections of greater than 22 per cent open land accounted for 9l per cent of the recorded elk tracks (Figure 2 and Table 5). Areas with less Open land seemed to be more important during the winter, when more cover was necessary. Slope 'Seven of the openings were part of a flat plain with an average slope of I0.5 degrees (Table 5). Three of the Openings had at least some slope -29- of 20 degrees or greater. Highest elk usage was on lepes of I5 degrees or less (Figure 2). In a study in Utah, Jeffery (I963) reported the highest elk usage on slopes of I7 degrees or less. Deer There were 8l per cent more deer tracks recorded than elk tracks and there was some correlation between the number of elk and deer tracks (Figure 7). Openings that were used extensively by deer were also used extensively by elk. Deer use was associated primarily with the edges, whereas in all but late summer, elk activity was associated with the mid-portions of the openings. Eighty-six per cent of the days on which elk tracks were recorded in any opening, deer tracks were also in the same opening. Lack of excessive broswing or grazing, and lack of observed aggressiveness or territoriality between the two species, indicates any existing competi- tion is probably minimal. Human Access There was a definite relationship between difficult human access and increased numbers of elk tracks. Openings 3, 7 and 9 had the highest elk tracks observed'and were also the most difficult to gain access to (Table 5). Results of this study follow those of Kowalsky (l96A), who observed a very low use of mountain meadows by elk which were not secluded from man. Coop (l97l) reported that AA per cent of the parks used by elk in Montana had no vehicle roads or trails. -30- IOOt II — Elk Tracks 1 l --- DeerTracks I (I) X 0 . < a: I I- I 2 I l D l l a ' I X l < I 2 1 1.1.1 ‘ I I .. I- I" ‘\ I, ll. I \‘ l‘ ‘l o ,/ \ x \ LU I I \\---- 3 I- Z LIJ 0 0: LL! 0. O 1 1 n 1 1 l l _L 1 1 S O N D J F M A M J MONTH Figure 7. Relative trends of elk and deer tracks. -3]- Vegetation Density Density of woody edge vegetation was an important characteristic. influencing elk use. Openings used most had at least some dense cover (>7,000 stems per acre) available on the edges (Figure 2 and Table A). Most summer activity of elk centered on openings having more open vege- tation, primarily the mature upland hardwood stands, on the edges. 0n extEeme high temperature days, elk were observed using south-facing slopes on the edges of these Openings, where breezes gave relief from insects. This follows elk behavior observed in Colorado (Boyd I970), Idaho (Balke t al. I965), Wyoming (Murie l96l) and Montana (Bruzda I953). .Edge Composition Seventy-five per cent of the elk tracks recorded were observed in openings having edges of early stage aspen and hardwoods (Figure 2). Clearings found in coniferous types had the least number of tracks ob- served. Openings with edges composed of more than 35 per cent edible vegetation (Table A-l) accounted for 77 per cent of the tracks observed (Figure 2 and Table 2). Shrubs, Ferns, Herbs, and Lichens The complex distribution of the ground cover (Table 3) made any association with elk tracks in the plots difficult to detect. Two open- ings, which had frequencies of 5 and 30 per cent clover, did show relatively ~ heavy elk use. Raspberry (Rubg§_spp.) was present only in the six open- ings having highest annual elk use (Figure 2). Any further indication of preference or use of ground vegetation was not determined in this study. -32- Soil In this study soil type and pH seemed to be more important than soil nutrient content when compared with numbers of elk tracks (Figure 2 and Table 6). The five openings having soils of Roselawn sandy loam accounted for 6A per cent of the total elk tracks recorded. Four open- ings having a soil pH greater than 5.5 had 56 per cent of the total tracks (Table 6). Both soil type and pH are interrelated and are prob- ably most important in determining the quality of vegetation, but they do seem to indirectly influence elk use of openings. LITERATURE CITED LITERATURE CITED Boyd, R. J. I970. Elk of the White River Plateau, Colorado. Colo. Div. of Game, Fish and Parks. Proj. No. W-38-R, Pp. l-l26. Brazda, A. R. I953. Elk migration patterns and some of the factors affecting movements In the Gallatin River drainage, Montana. J.Wildl. Mgmt. 17(1):9-23. ' Buss, M. I967. A study of food habits and habitat utilization by elk. M.S. thesis, University of Michigan. 57 PP. ChriStenson, D. R. and E. C. Doll. I973. Mich. State Univ. Coop. Ext. Serv. Bul. No. A7I. 6 pp. Coop, K. J. l97l.. Habitat use, distribution, movement and associated behavior of elk, Little Belt Mountains. Mont. Fish and Game Dept. Final Report Proj. No. W-l20-R-l and R-2. 6l pp. Craighead, J. J., F. C. Craighead, R. L. Ruff, and B. W. O'Gara. I973. Home ranges and activity patterns of nonmigratory elk of the Madison drainage herd as determined by biotelemetry. Wildl. Monogr. 33: l-SO. , G. Atwell, and B. W. O'Gara. I972. Elk migrations in and near Yellowstone National Park. Wildl. Monogr. 29:l-A8. Crossman, C. 'I970. Canada Creek Ranch, it's history. Canada Creek Ranch, Atlanta, Mich. 62 pp. Dalke, P. 0., R. D. Beeman, F. C. Kindel, R. J. Robel and T. R. Williams. I965. Seasonal movements of elk in the Selway River drainage, Idaho. J. Wildl. Mgmt. 29(2):333-338. Dix, R. C. I96]. An application of the point-centered quarter method to sampling of grassland vegetation. J. of Rng. Mgmt. lA:63-69. Failing, O. l95l. Pigeon River Elk. Mich. Conservation 20(5):5-23. Harper, J. A. l97l.' Ecology of Roosevelt Elk. Oregon State Game Comm. AA pp. , J. H., W. W. Bentley, and C. F. Yocom. The status and ecology of the Roosevelt Elk in California. Wildl. Monogr. l6:I-A9. Jeffrey, D. E. .l963. Factors influencing elk and cattle distribution on Willow Creek summer range, Utah. M.S. thesis, Utah State Univer- sity. 52 pp. -33- -3D- Kirsch, J. B. I962. Range use, relationship to logging, and food habits of the elk in the Little Belt Mountains, Montana. Mont. Fish and Game Dept., Compl. Rept., Proj. W-98-R-2. AA pp} Knight, R. R. I970. The Sun River elk herd. Wildl. Monogr. 23zl-66. Kowalsky, S. I. I96A. Ecology of mountain meadows and use Of elk. M.S. thesis, University of Idaho, Moscow. ~ Mackie, R. J. I970. Range ecology and relations of mule deer, elk and cattle in the Missouri River Breaks, Montana. Wildl. Monogr. 20: l-79. McCaffery, K. R. and W. A. Creed. I969. Significance of forest Open- ‘ ings to deer in Northern Wisconsin. Wis. Dept. of Nat. Res. Tech. Bull. No. AA. IOA pp. Moran, R. J. 1973. The Rocky Mountain elk in Michigan. Mich. Dept. Nat. Res. and Dev. Rep. No. 267. 93 pp. Murie, O. J. l95l. The Elk of North America. The Stackpole Co., Harrisburg, Pa. 376 pp. . Nie, N. H., C. H. Hull, J. J. Jenkins, K. Steinbrenner and D. H. Bent. I975. Statistical package for the social sciences. 2nd Edit., McGraw-Hill Inc., New York. 675 pp. Stevens, D. R. I966. Range relationships of elk and livestock, Crow Creek drainage, Montana. J. Wildl. Mgmt. 30(2):3A9-363. Veatch, J. 0. I930. Soil survey of Montmorency County, Michigan. U.S. Dept. of Ag. Bur. of Soils and Cam. A3 pp. APPENDIX -36- Table A-l. Woody plants identified in elk rumens, December l96A-l965 (Buss I967). Species Percent Volume Gaultheria procumbens I3.2 Thuia occidentalis 9.8 Tsuga canadensis 6.l 7 .Rhusityphina A.3 Pinus banksiana 7.2 lflyrus malus 6.6 Populus tremuloides l.5 Amelanchier spp. V . trace Acer rubrgn ‘ l7.2 iii-1112.21.29. ' «5‘ Acer saccharum 2.0 Prunus serotina trace Tilia americana traCe .Fagus grandifolia trace -37- Table A-2. Results of multiple regression analysis with number of elk tracks as the dependent variable and characteristics of the openings as the independent variables. Independent Variables B8 F . R2 Changeb Human access . 2.0II .005 .3l Pereent of section open .527 .005 , .IO Soil pH ' - 3.7ll .OA2 .06 Deer tracks .A3l I .OOI . .OA Edible edge vegetation .3lO .Ol7 .03 Density of edge vegetation .I6O .033 .03 Amount of accumulated snow ' .539 .005 ..02 Slope J _.017 .050 .01 Boundary to opening ratio .270’ .003 .Ol Size of Opening' .009 ' NS ' .Ol Wind speed .000 NS .OI Temperature .003 NS .Ol Soil nutrients ' .OOO _ NS .Ol aB is a measure of the relative importance of the independent‘ variables. . sz change is the approximate amount of variance explained by the independent variables. -38- mo_nm_cm> vcm mxumcu x_o Lok mucu_u_mmoou co_um_oLLoQ .:o_um_oLLou O_a_u_:E mo mu_:mom .>__mb cuczmmoa .-< 0600_a. ,_o.v a mo.v a mxumch Como Omega coo: .c-0 0c-3 00600 0c_3 30cm .E:ou< 30cm c_mm .QEOF .c_z .QEOh .xmz eN_. «N-. am-. 60-. tea_. e-. PEA-. «m_. «em-. 0-. 60. 00. -0.- 00. -0. -0. :0. 60.- A0. m0. 00. 00. 60. 60. m0. m0.- m0.- -0.- 60. 60. -0. eekm. «ANN. N0. m0. 60.- 60.- -0. N0.- -0.- N0. .m0. 00. e-N. 60.- e-_.- em_.- 00.- 0-.- e-.- eea-.- 0-. e-. em-. eN-. 00. A0.- 60.- 0-.- 60.- N0.- -0.- 60. m0. Eamm. N0. N0. “0.- 00. 00. m0. _0.- 60. -0. e-. 00. N0.- 60. e__. m0. 60. A0. 00. A0. 00. 0-. em-. tea_. «-. eea-. em_. «60-. E~_. 60. 60. em-. Em-. a-. 0-. -0bo» 0- m m A 6 m a m N -. 0 z _ z m a 0 0-60_c0> "IIIIIIIIIII'IIIIIIIII