Loading...
HomeMy WebLinkAboutAPA4128 This report, Susitna Hydroelectric Project. Final Report, Big Game Studies. Vol. II, Moose-Upstream, had been distributed erroneously with series statement on label on cover: Harza-Ebasco Susitna Joint Venture document ; no. 3494. In addition, the microfiche version was erroneously printed with "Date: 870500" and with the number 3494 in the header. The number 3494 is really assigned to a report with a similar title listed as Big Game Studies, Final report. Volume 2, Moose, Downstream published in 1987 (cited as 0 pages in length in the Susitna Hydroelectric Project document index). That report is referred to as a progress report in the preface of this report. Alaska Library Resources and Information Services has assigned APA no. 4128 to this document for the Susitna Hydroelectric Project collection. DMIIU•IJIMCO .__ - !k11itn1 Joint Venture ~ ........ SUSITNA HYDROELECTRIC PROJECT FINAL REPORT BIG GAME STUDIES VOL. II HOOSE-UPSTREAM WARREN B. BALLARD and JACKSON S. WHITMAN ALASKA DEPARTMENT OF FISH AND GAME 333 RASPBERRY ROAD ANCHORAGE, ALASKA 99518 APRIL 1988 PREFACE Between January 1980 and June 1986, the Alaska Power Authority (APA) contract'-d with the Game Division of the Alaska Department of Fish and G ... (ADF&G) to provide field data and recommendations to be used for assusina potential i111pacts and developing options for •itigating t.pacts of the proposed Susitna Hydroelectric Project on .case, caribou, brown bear, black bear, Dall sheep, wolf, wolverine, and belukha whales. ADF&G was only one of many participants in t his progra.. Information on birds, small mammals, furbearers, and vegetation was collected by the University of Alaska and private consulting firms. Fon~ally, ADF&G's role was to collect data which could be used to describe the baseline, pre-project condHions. This information was suppl .. ented with data from other ADF&G studies. Baseline conditions were defined to include processes which might be sufficiently sensi- tive to either direct or indirect project induced impacts to alter the dynamics of the wildlife populations. The responsibility of iapact assessment and mitigation planning was assigned by APA to several private consulting firms. ADF&G staff worked closely with these firms, but only in an advisory capacity. The project was cancelled before the i mpact assessment and mitigation planning processes were complete. In an effort to preserve the judgements and ideas of the authors at the termination of the project, the scope of this report bas been expanded to include material relating to impact assessment and mitigation planning. Statements do not necessarily represent the views of the APA or its contractors. Conjectural statements sometimes are included in the hope that they may serve as hypotheses to guide future work, should the project be reactivated . The following list of reports completely cover all of the Game Division's contributions to the project . It should not be necessary for the reader to consult the many progress reports. Hoose Modafferi, R. D. 1987. Susitna Hydroelectric Project, Big Game Studies, Final Report Vol. I -Hoose -Downstream. Alaska Dept. of Fish and Game. Ballard, W. B. and J. S. Whitman. 1987. Susitna Hydroelectric Project, Big Game Studies, Final Report, Vol . II -Hoose - Upstream. Alaska Dept. of Fish and Game. Becker, E . F. and W. D. Steigers. 1987 . Susitna Hydroelectric Project, Big Game Studies. Final Report, Vol. III -Hoose forage biomass in the middle Susitna River basin, Alaska. Alaska Dept. of Fish and Game. Becker, E. F. 1987. Susitna Hydroelectric Project. Big Game Studies. Final Report. Vol. v: -Moose Carrying Capacity Estimate. Alaska Dept. of Fish and Game. CaQbou Pitcher, K. W. 1 ~ 7 . Susitna Hydroelectric Project, Big Game Studies. Final Report. Vol. IV -Caribou. Alaska Dept. of Fish and G8118. 59pp. Black Bear and Brown Bear Hiller, S. D. 1987. Susitna Hydroelectric Project, Big Game Studies, final Report. Vol. VI-Black Bear and Brown Bear. Alaska Dept. of Fish and Game. Ballard, W. B., J. S. Whit•an, L. D. Aumiller, and P. Hessing. 1984. Susitna Hydroelectric Project, Big Game Studies. 1983 Annual Report. Vol. V -Wolf. Alaska Dept. of Fish and Game. 44pp. Ballard, W. B., J. S. Whitman, and C. L. Gardner. 1987. Ecology of an exploited wolf population in southcentral Alaska. Wildlife Monographs No. __ (In press). Wolverine Whitman, J. S. and W. B. Ballard. 1984. Susitna Hydroelectric Project, Big Game Studies. 1983 Annual Report. Vol. VII - Wolverine. Alaska Dept. of Fish and Game. 25pp. Dall Sheep Tankersley, N. G. 1984. Susitna Hydroelectric Project, Big Game Studies. Final Report. Vol. VIII -Dall Sheep. Alaska Dept. of Fish and Game. 91pp. Belukha Whale Calkins, D. 1984. Final Report. Game. 16pp. Susitna Hydroelectric Project, Big Game Studies. Vol. IX -Belukha Whale. Alaska Dept. of Fish and SUSITNA HYDROELECTRIC PROJECT BIG GAME STUDIES VOL. II -HOOSE-UPSTREAM DYNAMICS OF HOOSE POPULATIONS ALONG TilE MIDDLE SUSITNA RIVER IN RELATION TO PROPOSED HYDROELECTRIC DEVELOPMENT IN SOUTHCENTRAL ALASKA BY WARREN B. BALLARD, Alaska Department of Fish and Game, P. 0. Box 1148, Nome, AK 99762 JACKSON S. WHITMAN, Alaska Department of Fish and Game, 333 Raspberry Road, Anchorage, AK 99518 SUMMARY From 1976 through 1985, moose (Aices a/ces) demography, movements, and habitat use were studied in relation to a proposed hydroelectric development project along the middle Susitna River in southcentral Alaska. History of the moose population from the 1940s to initiation of these studies was reviewed. The moose population increased in the 1940s and 1950s due to mild winters, favorable range conditions, and low rates of mortality from hunting and predation. The population peaked in 1963 and began declining following a series of se11ere winters and high predation. Record low levels were reached by 1975. Between 1976 and 1985, 463 moose (61 5-to 10-month-old calves, 184 adults and 218) neonates were captured, processed, and equipped with either radio-collars or visual collars to aid in determining the causes of population decline and to assess potential impacts of hydroelectric development. Movements of radio-collared animals in relation to two proposed impoundments were used to delineate the boundaries of zones where moose would be impacted. The moose popu- lation within the zones was censused in 1980 and 1983, and data concerning sex-age composition were collected annually. Within a 6,522 km 2 area the moose population was estimated at 4,500 in 1980 (0.69 moose/km 2 ), whereas in 1983, the moose population was estimated at 4,573 within a 7,586 km 2 area (0.60 moose/km 2 ). Average age of adult cow moose was 7.7 years. Although average age of captured moose increased as the study progressed, differences were attributed to sampling biases associated with study of different subpopu 1 at ions. Pregnancy rates were initially high, averaging 81%, but apparently declined as the project progressed due to inaccurate diagnoses and study of the same individual moose which became older and less pro- ductive. Parturition occurred between 18 Hay and mid-June with 96% occurring between 24 May and 10 .Tune. Twinning rates averaged 38%. Overall, neonatal sex ratios were skewed in favor of males, but this difference was due to a large unexplained difference in 1977. TWo hundred and eighteen neonates were captured and radio-collared to determine causes of mortality within 4 areas during 1977, 1978, 1979,· and 1984. Predation accounted for 83~ of total mortality. Host mortality occurred during the first 6 weeks of life. Brown bears (Ursus arctos) were the greatest (73~) cause of mortality followed by •iscellaneous factors (12~). Rates of mortality between collared and uncollared calves were similar. Within the impoundment zones, black bears (Ursus amer/canus) were more dense than brown bears, but the latter were still the most important source of calf moose mortality. TWin calves had lower survival rates than single calves. Survival through 5 months of age averaged 26~. From 6-12 months of age during severe winters, males had lower survival rates than females. There were no differences in survival rates between sexes during mild winters. Annual calf survival rates avPraged 22 and 17~ for females and males, respectively. Yearling and adult female annual survival averaged 95~. Lowest annual survival (92~) occurred during a severe winter. Predation accounted for 8 of 11 mortalities when cause of death was known. Mortalities were equally divided between snow and snow-free periods. Adult bulls had lower survival rates than yearlL.~ bulls because the latter were protected from human harvest from 1980-86. Adult bulls (S2 yrs) had low rates of natural mortality (excluding hunting). Mean group size was greatest in October and lowest in August. Three major periods of moose movement were readily identifiable: autumn, spring, and during the rut. In late September and early October some moose made distinctive movements for breeding purposes. Dates of autumn migration to winter range were variable, but appar- ently coincided with first major snowfall. Spring migration was also variable and appeared related to snowmelt. Resident moose had over- lapping seasonal ranges, whereas migratory moose had nonoverlapping ranges separat~d by as much as 93 km. Home range use was traditional. Seasonal and total howe range sizes of resi ~nt moose were correlated with number of relocations and appeared a dequately defined when numbers of relocations , 13 and , 39, respectively. Migratory moose home range sizes were not positively correlated with numbers of relocations. Summer, autumn, and total home range sizes of migratory moose were larger than those of resident moose, but winter home range sizes were not different. Total home range size of migratory moose averaged 505 km 2 , whereas resident home ranges averaged 290 km 2 ; both were larger than those reported in the literature. A more repre- sentative method of estima~ing home range size was described and compared with the traditional method. Average age of separation of offspring from adults was 14 months. Following initial separation, thirty-three percent of offspring temporarily reassociated with parents from 1-6 occasions. Sixty percent of 15 offspring partially or fully dispersed from the parental home range. More males than females dispersed. Home range sizes of parents and offspring were correlated. Hales had larger home ranges than females. i Greatest seasonal chan~es in moose distribution and density within th~ proposed impoundment3 occurred in the Watan& and Jay Creek drainages. Numbers of moos~ within the Watana impoundment during winters of moderate severity ranged from 42·580 (0.2-2.3 moose/km1 ). In c~ar­ ison, numbers of moose within the Devil Canyon itapoundment were relatively low, ranging from 0.5-1.0 moose/mi 1 (0.2-0.4 moose/km 1 ). Both spruce (Picea spp.) and willow (Salix spp.) vegetation types were used disproportionately to their availability. Moose did not select habitats strictly on the basis of browse availability. During winter, areas with relatively low browse biomass were hecvily used by moose, apparently because the browse that was present was more available due to shallower snow depths. Moose occurred at lowest elevations during April and highest elevations during the rut. Elevations from 1,800- 3,000 ft (549-914 m) were used by moose disproportionate to avail- ability. Annually, north and south facing slopes w~re preferred. Relocations of radio-collared moose were heavily biased toward day- light observations during which time they were usually bedded. Highest frequency of feeding observations occurred during summer. An index for estimat ing winter severity early in the year and which also allowed comparisons of individ11al winters was developed and described. Use of different elevations by moose during winter was correlated with winter severity . Lower elevations were used as wi.nters became more severe in terms of total snow depth. It was predicted that during a severe winter 50~ of the radio-collared moose would occur within the areas to be inundated. Potential impacts to moose as a result of the proposed project were classified into 3 categories: important, potentially important, and unimpor•.ant. Thirteen important impacts to moose were identified and discussed. These included such impacts as permanent and temporary habitat loss, displacement and disruption of movements, increases in accidental and human-caused mortality, and increased mortality from predati.on. Of seven identified potentially important impacts, possible chang~s in climate within an unknown radius of the impound- ments could be the most important. Five unimportant impacts were i.dentified and discussed. Several approaches were used in an attempt to quantify the numbers '>f moose which potentially would b6 lost if t h~ h~drQel@ctric project were built. A subjective appraisal of the numbers of moose t<\ pe lost from 12 moose subpopulations indicated that about 1300 might be l~s~ as a result of the project. This latter estimate was similar to an est1~9t e (second appr64ch) of the habitat carrying capaci.ty within the impouftciJJ;<>&t!! during a severe w:i n t er. Population modeling (third approach) indi cated that minor changes ~n several key population parameters as a result of the project would be sufficie t to either cause or accentuate a population decline and perhaps help to maintain the population at lower levels. Actual losses to the moose population, however, can not be a c curately pre- dicted at this time. Importance of the impoundments to moose du~ing severe winters could be significantly different than. that observed during this study. A number of post impoundment studies that will be necessary to adequately quantify losses to the moose population as a result of the project are briefly summarized. ii TABLE OF CONTENTS INTRODUCTION Acknowledgements STUDY AREA HETIIODS Tagging and Relocating Hoose Population Trends and Density Survival and Mortality Rates Home Ranges, Distribution and Vegetation Use Statistical Tests DYNAMICS OF THE HOOSE POPULATION Population Trends Population Density Age Structure Productivity Survival and Mortality Calves, 1-5 months of age Calves, 6-12 months of age Annual calf survival rates Yearling and adult females Yearling and adult males MOVEMENTS, DISTRIBTUTION, AND HABITAT USE Group Size Movement Patterns Autumn Migration Spring Migration Seasonal And Total Home Range Sizes New Method of Home Range Calculation Dispersal And Home Range Formation Timing of Separation Types and Rates of Dispersal Home Range Formation and Size Adult Dispersal River Crossings Timing Location Winter se Of The Impact Zones Watana Impoundment Devil Canyon Impoundment iii .age TABLE OF CONTENTS Vegetation Use Outside of Impoundments Watana Impoundment Devil Canyon Impoundment All Areas Combined Elevational Use Slope Use Aspect Use Activity Patterns Daily Monthly Effects Of Snow On Hoose Distribution Winter Severity Index Elevational Use Versus Winter Severity IMPACT MECHANISMS AND PREDICTIONS OF IMPACTS DUE TO HYDROELECTRIC DEVELOPMENTS Impact Mechanisms Classification and Identification of Impacts Important Impacts Potentially Important Impacts Unimportant Impacts iv Page TABLE OF CONTENTS Prediction of Project Impacts on Moose Subpopulations I11portance of Impoundment Areas SWIIIary of Project Impacts Monitorin& Proarus Necessary for Refinement of Impact Assessment LITERAnJRE CITED v Paae INTRODUCTION Historically, Game Management Unit (GHU) 13 has been one of the most t.portant moose hunting and viewing areas in Alaska. Between 1963 and 1975 about 1n of the statewide harvest came from the area. The moose herd was thought to have increased during the 1940s and 1950s (Bishop and Rausch 1974). Estimates of sex-age composition were initiated in 1952, and annual surveys have been conducted in selected areas sinc e 1955. Hoose numbers were thought to have increased, apparently in response to favorable range conditions, low numbers of predators, &nd relatively low human harvests. Bishop and Rausch (1974) stated the concensus was human harvests only slightly affected sex and age ratios during that time period. The moose population apparently peaked in 1960 and then began declining (Bishop and Rausch 1974). There appeared to be an inverse relationship between numbers of wolves (Canis lupus) and moose. Wolf numbers were reduced to about 12 in the entire basin through predator control and aerial hunting activities (Rausch 1967, 1969.). Termin- ation of those activities resulted in a large wolf population increase (peaked 1965) and an apparent moose population decline (B i shop and Rausch 1974). Numbers of both brown and black bears were also thought to have been reduced during predator control activities, which may also have contributed to the moose population increase. Severe winters cr.ntri buted to the moose population decline (Bishop and Rausch 1974). With the exception of winter 1955-56, moose produc- tivity was thought t o be high and mortality low until winter 1961-62 when the population began declining. A severe winter also apparently occurred in 1965-66, but its effects were poorly understood (op. cit.). A severe winter with record snowfall occurred in 1971-72, and mortality was high; subsequent calf production and calf survival in 1972 were low. Between 1962 and 1974, hunters became more efficient at harvesting moose due to increased use of aircraft and all terrain vehicles. Thus, while the moose population declined, moose harvests remained "almost constant" (Bishop and Rausch 1974). They concluded that, after severe winters, the combined effects of mortality by humans and wolves had the capacity to preclude moose population growth and could have contributed to further moose population declines. A severe winter occurred during 1974-75, further reducing calf sur- vival, and the moose population appeared to continue its decline. Drastic reductions in human harvests appeared necessary for the moose population to recover. If predation was responsible for keeping the moose population at low levels, reductions in predator numbers would also result in a moose population increase. Predator-prey investi- gations were conducted from 1976-1985 and have been summarized by Ballard et al. (1981~.~. 1982a, and Ballard and Whitman (1987). While the GKU 13 moose populat on was undergoin& these f luctuations, studies were conducted concerning the feasibility of hydroelectric developllellt along the Susitna River. ln 1948 kaiser AhminUII Co. first exained the feasibility of hydroelectric development of the Susitna River. Since that time development proposals have ranged from a 2-12 dam system (Taylor tdld Ballard 1979). Host recentlt, the Devil Canyon-Watana Creek 2-dam s ystem was selected by the U. S. Army Corps of Engineers (Corps) as the most viable of several development alter- nat ives. Limited funds beca.e available for studies of moose distri- bution in 1975 in relation to the proposed impoundments (Mcilroy 197 5). The Corps increased the amount of fund ing in 1976, and results of these e fforts were present e d by Taylor and Ballard (1979) and Ballard and Taylor (1980). During the se ere winter of 1978-79, few 'nnds were available for studyin g moose; this became imr,>ort ant because e proposed impoundment areas were thought to be iJa1?ortant habitat during severe winters. During the late 1970s, the ~tat~ of Alaska took over res ponsibility from the Corps for power develoJ..ment along the Susitna Ri ver. The State, recognizing the importance nf wildlife resources in the area, initiated a series of studies in 19 ,0 . Detailed baseline information on moose numbers a 'd ecology was so ·tght to both adeq uat ely predict and .onitor the effects oe la ge-scale 1ydroelectric development on moose populations and to mitigate impacts . The pre.s ent study was conducted for two reasons: (1) to determine the causes of moose population decline in portions of GHU 13 sine 1960, and (2) to determine the potential impact of Susitna hydroelectric development (2-dam system, Watana, and Devil Canyon impoundments) on moose. This report summa.rizes the results of studies from October 1976 through January 1986, including data from other GMU 13 studies pertinent to e v aluating potent ial impacts of hydroelectric develop- ment. 2 ACD«M.EDGEMENTS We extend our thanks to a large number of individuals who participated in the design and ex,ecution of these studies. R. Rausch, J. Vania, R. Soeerville, IC. Schneider, and J. Faro were instruaental in getting the studies initiated. L. Aumiller, T. Balland, D. Cornelius, A. Cunning, J. Dau, S. Eide, C. Gardner, P. Hessing, J. Hughes, L. Hetz, T. Spraker, and J. Westlund all assisted with collection of data during the study. A. Franzmann, S. D. Hiller, S. H. Hiller, and ~ Becker provided valuable technical and statistical support through- out the study. E. Goodwin and C. Lucier provided laboratory support. B. Strauch assisted with preparation of figures. IC. Schneider pro- vided administrative support. A. and J. Lee, Lee's Air Taxi, contrib- uted greatly to the project not only by safely piloting tracking aircraft but also by donating time and expertise to insuring the success of the projects. V., C., and B. Lofstedt, Kenai Air Alaska, provided many safe hours of helicopter support. Both K. Bunch, Sportsmen_s flying Service, and H. McMahan, McMahan's Guide Service, also provided many hours of safe and efficient fixed-wing aircraft tracking and spotting. K. Adler provided bookkeeping and clerical support throughout many aspects of the study. S. Lawler typed the final version. S. Peterson, K. Schneider, B. Towns ~!ld, edited the report. G. Couey, Watana Camp Manager, provided helpful logistical support. I. Parkhurst assisted with final preparation of the manu- script. This manuscript is possible by contributions from the Alaska Department of Fish and Game (ADF&G), the Alaska Power Authority, and several Alaska Federal Aid in Fish and Wildlife Restoration Projects. 3 snmY AREA The ortainal study area included most tribut aries which drain into the Suaitna River upstrea of the mouth of Portage Creek (Fig. 1). The boundary generally followed the Denali Highway on the north; the Maclaren River and Tyone, Susitna, and Lou i se Lakes systems on the east; the Glenn Highway and Little Nelchina River on the south; and drainages upstreu of Portage Creek on the west (Ballard et al. 1982b). Reductions in the study area were made in 1983 (BallB.rd et al. 1983) when different zones of impact were identified . Data from radio-collared moose, which either seasonally or annually occupied areas to be direc tly altered by operation and maintenance of the Watana and Devil Canyon impoundments, were used to delineate an area where moose would be directly impacted. Home range polygons (Mohr 1947) were delineated for each moose which utilized either land to be inundated or lands which were to be altered by major facilities, encampments, or borrow pits. Outermost points of all these polygons were connected and used to delineate the border of a primary illpact zone (Fig. 1). In addition, a secondary impact zone was delineated on the assumption that moose displaced from the primary impact zone would caapete with moose occupying the secondary zone. Although moose in the secondary i mpact zone were not known to use areas directly impacted by the proposed project , their home range polygons overlapped home ranges of moose that used. the primary impact zone. Similarly, a tertiary impact zone was delineated where overlaps with the secondary zone occurred, assuming further competition from displaced moose (Fig. 1). Vegetation, topography, and general climate of the area were described by Skoog (1968), Bishop and Rausch (1974), Ballard and Taylor (1980), Ballard (198,2) and Ballard et al. (1987). Specific vegetatio descriptions of the impoundment areas were provided by Becker and ,Steigers (1987). 4 HE1110DS Taaaina and Relocating Hoose Hoose were darted froa a Bell 206-B (Jet Ranger) h •l licopter, except neonatea which were captured on foot (Ballard et al. 1979). Three coabinations of ~ were used to t..obilize adult and short yearlina 800se: (1) succinylcholine chloride with hyaluronidase (Wydase), (2) etorphine hydrochloride (M-99) with and without xylazine hydro- chloride (Roapun), and (3) carfentanil (Franzmann et al. 1984). Captured 800Se were marked with a radio-collar, a visual numbered canvas collar (Franzmann et al. 1974), or both. Sixty-one 5-10 month old calves, 115 adults, and 218 neonates were radio-collared while 69 adults were equipped with only canvas collars. All adults were aged by extracting a lower incisor tooth which was processed according to -tbods described by Sergeant and Pimlott (1959). Each moose waa ear-taased with numbered Monel metal tags. During spring, all female yearlina and adult moose were rectally palpated (Roberts 1971) to determine pregnancy status. Two types of receivers were used during the course of the study: (1) 4-band, 48-channel partable receiver manufactured by AVM Instru- ment Co. (Champaign, IL), and (2) portable progr81D1Dable 2,000-channel scanni113 receiver manufactured by Telonics (Mesa, AZ). Radio-collared ..aose were relocated from Bither a Piper PA-18 (Supercub) or STOL- equipped Cessna 180 or 185 fixed-wing aircraft. Each aircraft strut was equipped with a 3-element yagi antennae. A control box within the aircraft allowed monitoring o£ the strength of radio signals from both antennae or from either side of the aircraft. By switching fr011 antenna to antenna, the direction of strongest signal was determined and the aircraft piloted in that direction until the signal became stronger on the opposite antenna. This resulted in an initial series of broad slo•.., turns until the animal was close, at which time the search pattern developed into steep , sharp turns to visually observe the animal. Moose relocations were plotted on 1:63,360-scale USGS maps. Time, behav i or, numbers of associates (gr oup size determined for animals wit h j 3pproximately 1,300 ft (400 m) of instrumented individuals) by sex and age class, and vegetation type according to Vier'!ck and Dyrness (1980) were recorded on standardized forms. Activity patterns were divided into 4 categories: foraging, bedded, standing, and other. Sixty-five moose originally captured as 5 10-month-old calves and 115 adults were located on 5,421 occasions (! = 30 relocations per moose) from October 1976 through January 1986. Numbers of relocations per individual ranged from 2, for radio-collared short year lings which slipped collars or starved, to 104 for an adult female. Neonate calves were relocated and visually observed on hundreds of occasions and their signals monitored on thousands of occasions (Ballard et al. 1979). 5 Population Trends and Density Autu.n 1100se sex-age composition surveys conducted fr011 fixed-wins aircraft have been conducted annually in GMU 13 since 1955 in 16 different count arAas (Fig. 2). These low-intensity fliahts generally last •t minute (min) per mi 1 (0.4 min/km 1 ). Flight patterns consist of transects flown at 0.8-1.2 km widths between 300-500 ft (91-152 m) altitude on flat terrain or transects flown along contour intervals in billy and 110untainous regions . Such surveys are conducted after the first major autumn storm which provides complete snow cover usually durina late October tbrouah early December. Surveys are usually c011pleted before bulls shed their antlers. Moose are sexed and aged according to relative size, presence or absence and configuration of antlers, and vulva patch. Bulls with spiked, forked, or small pal- mated antlers less than 30 inches wide were assumed to be yearlings. Total moose observed per hour, bull: 100 cow ratios, calf: 100 cow ratios, and percent of herd c0111prised of yearling bulls are routinely used by managers as indicators of population trend. Such surveys are not used to estimate population size or density except when minimum est~ates are desired. Stratified random sampling (Gasaway et al. 1981) was used to estimate moose population size and density in autumn 1980 and 1983. Such counts were conducted in the same pattern as those described for sex-age surveys, but search intensity usually exceeded 4 min/mi 1 (1 .54 min/km 1 ). Total counts at search intensities 4 min/mi 1 (1.54 min/b1 ) were conducted in selected small areas, particularly the impoundment zones where documentation of winter moose densities in selected habitats was desirable. Survival and Mortality Rates Survival rates of radio-collared calf, yearling, and adult moose were calculated using methods described by Trent and Rongstad (1974). Neonates were monitored daily, allowing calculation of daily survival rates up to 1 November. All other rates were es t imated on a monthly basis. When dates of last observation and known death spanned several months, the median date was used. TWo survival rates were calculated for each age class and time period when appropri ~te: (1) only those animals whose fate was known, e.g., the animal was either dead or alive when last observed, and {2) the average of two dates--one calculated which assumed all missing animals were alive and another which assumed all missing animals we r e dead. Moose were excluded from survival calculations if it could not determined •.o~hether the radio- collar had fallen off o r the animal was dead. Causes of mortality were determined according to methods described by Ballard et al. (1979) and Stephenson and Johnson (1972, 197 3 ). Monitoring frequency was not sufficient t.J determine cause of death for most adult 1110rtalities, so cause of death was classified as unknown. When monitoring intensity toras frequent, such as once or twice per week, it was often possible tn classify c ause of death based on ground examination at the site or actual observation of a predator on the carcass. Causes of death were classified as unknown, brown or 6 black bear predation, wolf predation, bunting, miscellaneous (IDOr- talities such as being stepped on by their cow, pneU80nia, auto collisiona, etc.), or winter-kill. The latter cJ.assification included all mortality involving starvation or other winter-related conditions. Ha.e Ranges, Distributions and Vegetation Use Yearling and adult home range sizes were calculated using the minimum home range method (Mohr 1947). This aetbod may be adequate for estimating home range sizes of animals occupying flat terrain and ba.ogenoua habitat but may not be appropriate when large blocks of nonhabitat, e.g. mountains, areas ,4,000 ft (1,219 m) elevation or lakes are included within polygons. Consequently for s01ae analyses Mohr's (1947) method was modified as follows: 1. Seasonal, annual, and total home ranges were calculated. a. For home range calculations 3 seasons were recognized: Summer -Hay through August, Autumn -September through December, and Winter -January through April. b. Total and seasonal home range sizes were not calculated when numbers of relocations were S or <24, respectively. c. Selected relocations from different seasons ~ere included in another seasons home range calculations if there was a clear relationship with earlier or later points. 2. Linear lines connecting outermost relocations were used except in the following cases: a. When elevations ,3,600 ft (1,097 m) were involved the boundary followed the contour line. b. Slopes ,30 degrees were excluded. c. For outlying relocations, the polygon was drawn from the closest two perpendicular points to the outlier. d. When all relocations occurred on drainage, the boundary followed crossing it. one the side of drainage a major without Dates and timing of migrations and movements were determined by examining sequential observations of individual radio-collared moo~e. When s~quential moose relocations deviated from a cluster of points, migration or movement to another range was judged to have been initiated. Hoose were considered to have arrived at a seasonal range when a point fell within a home range cluster. 7 Seasonal and total home ranges between resident (overlapping seasonal ranges) and migratory (nonoverlapping) hOIH ranges were compared. Distances between winter and SWiller home ranges of aigratory 1100se were determined by measuring the closest points between seasonal home range polygons. Availability of overstory vegetation types as well as elevations, slopes, and aspects were assessed by measuring these variables at section corners of 1:63,360 scale topographic and vegetation maps. Use of these variables by moose was determined from radio relocations plotted on the maps. Elevations were determined by extrapolating between contour lines to the nearest 50 ft (15 m) interval. Slopes were classified into 3 categories: flat = Sl0° with contour line intervals , 0 .19 inch ( . 49 em), gentle = 11-30° with contour line intervals ranging from 0.03-0.19 inches (0.08 -0.49 em), and moderate = ,30° with contour line intervals <o.03 inches (0.08 em). Aspect was classified as flat or one of 8 compass directions from a line per- pendicular to the contour lines through the moose location point. Methods used to quantify moose browse and understory vegetation were described by Becker and Steigers (1987). Browse quantities were divided into seven categories, from high to zero, depending on browse quantity. Point locations of radio-collared moose (N = 2,930) were also divided into one of the corresponding browse categories by season of use. Selectivit y (preferred :>r avoided) of habitat types was determined by chi-square analyses ~imilar to Neu et al . (1974). Relative distribution of moose was determined in 1980 and 1985. Aerial distribution surveys differed from other types of counts and censuses in that less survey effort was expended per unit area, and no precise population estimates could be derived. Between 1-2 min/mi 2 (0.4 -0.8 min/km 2 ) was expended searching for moose. All moose observations were recorded on 1:63,360 scale USGS topographical maps. Similar to autumn censuses, winter distribution data were used to stratify areas into relative density strata, i.e. high, medium, low, and zero density. No attempt was made to estimate population size in the study area during late winter, because no reliable density esti- mates existed. Only the relative differences in density were cal- culated. In-depth total counts (no variance estimate) of the actual impoundment areas are provided in subsequent sections of this report. Statistical Tests Differences between means were compared by t-test and analysis of variance (Snedecor and Cochran 1973). Count data and proportion data were analyzed with Chi-square tests (op. cit.). Relationships between independent variables were examined by correlation analysis. Dif- ferences in mortality rates of neonate twins versus singles and between sexes were compared with a Logit model. Unless specifically stated, P 0.05 was required for statistical significance . 8 DYNAMICS OF THE MOOSE POPULATION Population Trends Trends in the moose population were assessed by examining GHU 13 moose sex-age composition count data collected from 1952 through 1984 and correlating survey year with moose/hour, bulls: 100 cows, calves: 100 cows, and percent yearling bulls in the herd. Prior to 1963, numbers of moose counted per survey hour were • -!able and not correlated with survey year. However, numbers of moose observed per hour of survey declined annually during the period 1963 through autumn 1975 (Fig. 3). Other moose population indic:' es •mch as bulls: 100 cows, calves: 100 cows, and percent yearling bulls in the herd, began exhibiting declines in the 1950s and declined through autumn 1975 (Figs. 4-6). In addition to the severe winters described by Bishop and Rausch (1974) severe winters occurred in 1974-75 and 1978-79 (see Winter Severity section). Apparently Bishop and Rausch 1 s (1974) assessment of the moose population peaking in 1960 was a subjective appraisal based on their experience in the area . Numbers of moose observed per hour surveyed suggest the population peaked in 1963 . Other population indicies suggest the population was already declining when composition counts were started in 1952. Moose counted per hour of survey, bulls:100 cows, calves:lOO cows , and percent yearling bulls in the herd all reached their lowest levels about 1975 (Figs. 4-6). After 1975, all population indicies suggested a moose population increase (p <o.Ol). Population modeling (see Ballard et al. 1986) suggested that reduced wolf and bear densities, mild winter conditions, and reduced bull harvests resulted in an annual moose population increase of about 3-5~. The moose population within the Susitna River Study Area exhibited virtually the same trend as the GHU 13 population (Figs. 7 through 10), except that productivity (as expressed by calf: 100 cow ratios) was quite vari able within the Susitna River Study Area prior to 1976. The moose population reached its lowest level in 1975. Thereafter, the moose population increased, although mortality (as reflected by calf: 100 cow ratios and percent yearling bulls) increased during 1 year following the severe winter of 1978-79. Proportionately more calves were produced from 1976-84 than from 1963-75 (p >o.OOS). Reduced wolf and brown bear densities, mild winter conditions, and reduced human harvests apparently contributed to a moose population increase. Population Density Two moose population censuses were conducted using Gasaway et al. 1 s (1981) survey methods. The first census was conducted in autumn 1980 before the final hydroelectric project study area had been delineated. Moose count areas 7 and 14 (Fig. 2), adjacent to the Susitna River east of Delusion and l~osina Creeks were censused from 5-8 November 1980. The remainder of the hydroelectric study area lying west of 9 Delusion and Kosina Creeks was not censused because of poor snow ' conditions but was stratified. A moose population estimate was derived by applying density estimates fr011 the census area to the stratified area. In addition, moose count area 3 was censused and count area 6 (Fig. 2) stratified so that a total moose population estimate could be derived for the area where long-term predator-prey studies were conducted (Ballard et al. 1986, 1987). The latter area (SRSA) was censused to partially validate a population model adapted to the hydroelectric study area. The estimated autumn 1980 moose population for count areas 7 and 14 was 1,986 (Table 1). A total of 743 moose were censused within 26 sa;p!e areas comprising 948 km 2 (39% of count areas 7 and 14). Of 945 mi 2 (2 ,448 km 2 ) within the count areas, 35% was classified as low moose density, 38% as medium moose density, and 27% as high moose density. Not all moose were observed during the census where survey intensity was 4.4 min/mi 2 (1.7 min/km2 ). Consequently, portions of 10 sample areas were randomly selected and resurveyed at 11.9 min/mi 2 (4.6 min/km2 ) to generate a sightability correction factor of 1.03 (Table 2). It was estimated that 98% of the moose were observed at the higher survey intensity. The corrected population estimate for count areas 7 and 14 was 2,046 moose, of which 22% were calves. Moose densities west of Kosina and Delusion Creeks were estimated following the regular c e nsus. One hundred seventy-nine moose were counted, which provided the basis for stratifying the remaining 830 mi 2 (2, 150 km 2 ); 562 mi 2 (1 ,456 km 2 ) were classified as low moose density, 256 mi 2 (663 km 2 ) as medium moose density, and only 12 mi 2 (31 km 2 ) as high moose density. The size of each stratum was then multiplied by the individual density stratum estimates (Table 1) to derive an approximate population estimate of 1,151 moose. Combining the latter estimate with that obtained for count areas 7 and 14 provided a total population estimate for the hydroelectric project area in autumn 1980 of approximately 3,197 moose. Stratification flights were also conducted in moose count area 6 (Fig. 2) on 9 Nov . 1980 with a Piper Supercub. This area was surveyed because it contained several subpopulations of migratory moose which occasionally utilized the impoundment zones even though the area was not within the boundaries of the hydroelectric project area. Of 470 mi 2 (1,217 km 2 ) stratified, 204 mi 2 (528 km 2 ) were classified as low moose density, 207 mi 2 (536 km 2 ) as medium moose density, and only 59 mi 2 (153 km 2 ) as high moose density. Extrapolating the average moose densities per stratum for count areas 7 and 14 (Table 1) to count area 6 provided an approximate population estimate of 830 moose. Moose count area 3 was also censused to help validate the moose population model. Four hun,ired seventy-three moose were estimated in the area. Combining moose ~opulation estimates for count areas 3, 6, and 7 yielded a moose population of 2, 772 during autumn 1980 within the 2,804 mi 2 (7,262 km 2 ) area. Within this area, 1,858 mi 2 (4,812 km 2 ) lies below 4,00 ft (1,219 m) elevation. Since moose rarely 10 utilize areas above that elevation, moose density on usable habitat in autu.n 1980 within SRSA was 1.49/mi1 (0.58 moose/km 1 ). Combining all areas which were either censused (count areas 3, 7, and 14) or stra- tified (area west of Kosina and Delusion Creeks and count area 6) in aut\lal 1980, a total of 4,500 moose were estimated within 2,518 mi 1 (6,522 km 1 ) of usable (1,219 m) habitat or 1.79 moose/mi 1 (0.69 moose/b1 ). During 1983 autumn moose population estimates for the hydroelectric primary impact zone, the SRSA, and several other count areas were made. The other areas were censused under other funding sour~es but are included here for comparative purposes. Distribution of moose in autumn 1983 was different from that in 1980 in that relatively fewer moose were present m open alpine areas. Consequently, moose were harder to observe as reflected by the sightability correction factor (1.19 in 1983 versus 1.03 in 1980). A total of 2,836 moose were estimated to occur within the hydro- electric primary impact zone during autumn 1983 (Table 3). Within the 1,156 mi 1 (2,994 km 2 ) of usable habitat within the primary impact zone, 16~ was classified as high moose density, 39~ as moderate moose density, and 45~ as low moose density. Overall, autumn density wJ.thin the impact zone in autumn 1983 was 1.82 moose/mi 2 (0.70 moose/km 1 ). A total of 2,795 moose were estimated within the SRSA (Table 4). The confidence interval about that estimate included the estimate gene- rated by population modeling (see Ballard et al. 1986) and validated the model for use under preproject conditions. Moose densities within this area were similar (1.94 moose/mi 1 or 0.75 moose{km:) to those in the hydroelectric primary impact zone, further strengthening its application for assessing population trends within and outside of the project area. The census estimate~, like the sex-age composition data and the population model, suggest the moose population had increased since 1975. During autumn 1983 a total of 2, 929 mi 2 (7, 586 km 2 ) of usable moose habitat within the SRSA, the primary impact zone, and one other count area was censused and 4,573 moose were ~stimated to occur (Table 5). Average moose density within this area was 1.55 moose/mi 2 (0.60 moose/km 2 ). Comparison of these average densities with those found within the hydroelectric project area (east of Tsusena Creek and Stephan Lake) suggests that the area to be impacted by the project contains relatively high densities of moose in relation to many other areas within GMU 13. Age Structure Average age of adult cow moose captured during 1976-1982 was 7.7 years (S.D. = 3.8 yr) (Fig. 11). Average ages among years were different (P > 0.05). Average ages of cow moose by year of capture were: 1976 = 7.5 years (SD = 3.4), 1977 = 7.0 years (SD = 3.8), 1980 = 9.4 years (SD = 3.8), and 1981 = 7.6 years (SD = 2.9). Cows captured in 1976, 1977, and 1981 were younger (p ~0. 05) than those captured in 1980. 11 Corrected for year of capture, cows ~10 years of age comprised 25~ of the sample in 1976 and 1977, whereas in 1980 they comprised 62~. This suggests that the age structure of the moose population had become composed of older individuals since 1976 and 1977. The exact opposite would have been expected based on autumn calf:cow ratios; 1976-1977 age structure was expected to be relatively old following several years of low recruitment, while a relatively young age structure was expected in 1980 following several years of improved recruitment due to predator reductions and mild winters . The former type of age structure was observed in the easter n portion of GMU 13 in 1975 where Van Ballenberghe (1978) reported 49% of tagged moose were ~ 10 years old. Although calves and yearlings were avoided during capture, no attempt was made to avoid other age classes and no biases would have been expected. These annual differences are attributed to differences in subpopulations and sampling variation. Average age of 3 captured adult (<1.5 yrs) bulls was 4.3 years (SD = 0.6 yr). Adu t bulls were avoided during capture for radio-collaring because of their relatively high mortality rates from hunting. Productivity Pregnancy rates among years were variable, bu-relatively big during the study; 88% in 1977 (N = 59), 73% in 1980 (N = 37), 79% in 1981 (N = 14), 82% in 1984 (N = 11), and 72% in 1985 (N = 19). Lower pregnancy rates after 19 77 were due to inaccurate diagnoses and lower productivity of older recaptured moose. For example, in 1980 four cows which had been diagnosed as not pregnant subsequently had calves. Of eight biologists participating in the tagging effort that year, only 2 were experienced and considered current (palpated 1 10 moose within previous 2 years) at assessing pregnancy rates. Also, many of the cows examined in latter years were recaptures from previous years. Since older moose are generally less productive than younger indi- viduals (Markgren 1969), the rates reported here should be considered minimal. Overall, pregnancy r a tes averaged 81%. GMU 13 pregnancy rates were similar to those reported elsewhere in Alaska and North America: 88% for eastern portion of GMU 13 (Van Ballenberghe 1978), 90% for GMU 9 on the Alaska Peninsula (Faro and Franzmann 1978), 90% in GMU 5 near Yakutat (Smith and Franzmann 1979), 88% in GMU 20 of interior Alaska (Gasaway et al. 1983), and 71-90% for other North American moose populations (Blood 1974). Yearling productivity was less than that of adul ts ; 2 (40%) of 5 yearlings physically examined produced calves. Earliest observations of moose parturition were 18 May in 1979 and 24 May in both 1977 and 19 78 for uncollared cows. During 1977, 1978, and 1980 timing of parturition and subsequent calf loss was determined by visually observing radio-collared cows and t h eir calves at 3 5-day intervals beginning on 24 May each year. No attempt was made to determine causes o f calf mortality for these animals. The earliest date a t which radio-collared cows were observed with calves was 25 May. Sixty percent of all calves were born between 29 May and 3 June of each year. Parturition wa s 96 % p e rce nt c omplete by 10 June 12 each year. In 1 case we observed a calf born in mid -August. The timing of parturition was similar to that reported in Alberta (Hauge and Keith 1981). Losses of radio-collared calves and calves of radio-collared cows in 1977, 1978, and 1980 were nearly identical (F'ig. 12), suggesting that the causes of mortality between the 2 groups were similar. Ninety- four percent of the natural mortality occurred before 19 July each year. After that date nearly all calves survived to at least 1 November each year. There.after, survival was dependent on wi nter severity and predation. Sex ratios and twinning rates at parturit i on were determined by examining neonates during calf mortality s tudies conducted in 1977, 1978, 1979, ad 1984 (see calf survival section). Observed twinning rates by year were as follows: 1977-19%, 1978-31%, 1979-52%, and 1984-63%. Overall, observed twinning rates averaged 38%. Pimlott (1959) reported that moose twinning rat es in North America ranged from 5-28%, while in Sweden twinning rates ranged from 17-65% (Markgren 1982). Franzmann and Schwartz (1985) s ggested that twinning rates reported in the literature had been collected by several different methods over several months and were not comparable. For example, Pi mlott' s (1959) rates were obtained in autumn after most neonate mortality had occurred (Ballard et al. 1981b, Franzmann et al. 1980). Markgren (1982) attributed differences in twinning rates in Sweden to climate and nutrition. There were no noticeable changes in habitat quality to account for the threefold differences in twinning rates among years for GMU 13 moose. Also, if winter severity prior to parturition had strongly influenced twinning rates, the 1979 (fol- lowing the severe winter of 1978-79) rate should have been low, while the 1977, 1978 and 1984 rates (following mild winters) should have been high. Only 1 year fit the expected pattern, and consequently the observed annual variations in observed twinning rates could not be explained. Overall, sex ratios of newborn calves (114 males to 91 females) were skewed in favor of males (X 2 =8.8, p=0 .07). This difference was due to the heavily skewed ratio wh i ch occurred during 1977; 35 of 50 calves (70%) were males (X 2 =8.0, p <o.OOS). Excluding 1977, sex ratios were not significantly different from 50:50 (79 males versus 76 females, X2 =0 .8, p=0.85). There were no differences in mortality rates among sexes (X 2 =17 .4, 14df, p=0.24). Verme and Ozoga (1981) determined that for white-tailed deer (Odocoileus virginianus) there was a relation- ship between interval following onset of es t rus and subsequent insemination to the sex ratio of fawns produce<!; does bred late in est rus produced higher proportions of male ca]·.•es. The implication was that in heavily hunted populat i ons where bull densities were greatly reduced, cows may have to wait to mate until they find s bull, resulting in higher male sex ratio at birth. Although speculative, there may have been a relationship between adult sex ratios and neonate sex ratios during this study. The lowest adult bull:lOO cow ratio in the calf mortal i ty study areas occurred in 1977 (11 males:100 females). Thereafter, bull:cow ratios increased from 17:100 in 1978 to 18:100 in 1979 and 24:100 in 1984. 13 Several investigators have expressed concern that low bull:cow ratios could influence conception rates and neonate sex ratios in ungulates (Mcilroy 1974, Bishop and Rausch 1974, Bailey et al. 1978, Verme and Ozoga 1981, A. Franzmann pers. comm., and many others). Differences in fetus size have been noted in several Alaskan moose populations where bu1l:cow ratios have been relatively low (Rausch 1967, J. Didrickson pers commun., V. Van Ballenberghe pers. commun., this study). Whether observations of small fetuses and skewed neonate sex ratios during some years were the result of relatively low bull: cow ratios was not known, but further investigation appears warranted. Survival and Mortality Calves 1-5 Months of Age. Causes of neonatal moose calf mortality were studied within 4 areas of GHU 13 during 1977-79 and 1984 (Fig. 13). Area 1 was studied during 1977-79, Area 2 during 1977-78, Area 3 during 1978, and Area 4 during 1984. A total of 218 moose calves were captured and radio-collared (Table 6). Twenty calves (9%) died of being abandoned or trampled by their cow due to capture act ivities. These calves were excluded from survival and mortality calculations. Predation by brown bears was the largest cause of calf moose mortality (Table 6) accounting for 73% of total mortality. The second largest cause of mortality was miscellaneous factors (12%) such as injury accidentally inflicted by the cow, drownings, and pneumonia. Wolf predation and unknown causes each accounted for 4% of the mortality. Predation from all causes accounted for 83% of total mortality during the first 5 months of life. Sixty-one percent of the calves died during the first 5 months of life. Ninety-six percent of the natural mortality occurred before 9 July of each year. Because the rates of calf loss between collared and uncollared calves of radio-collared cows were similar (Fig. 9), neither the collars nor the capture process predisposed the calves to death. There was considerable variation in survival rates among study areas (Table 6). Lowest survival rate occurred within the SRSA (Area 4) during 1984. That area was selected for study because it harbored dense populations of black bear (Miller 1984) which could have been an important source of calf mortality (Franzmann et al. 1980) not pre- viously documented in GHU 13. Because black bears would likely be eliminated as a result of hydroelectric development (Miller 1984), it was conce i vable that elimination of black bears could be beneficial to the moose population if they were a significant source of calf mor- tality. Black bears were found to be responsible for 11% of the total calf mortality in 1984 (Table 6). Similar to previous studies, predation by brown bears was the largest source of calf mortality (62%). There were differences in calf survival rates among areas and years (p<0.05). Mortality rates in all areas were greater for twins than single calves (p <o.Ol). Survival rates during the first 5 months of life varied from 3% in Area 4 in 1984 to 56% in Area 1 during 1977 (Table 6). Differences in survival rates among areas and years may 14 have been related to differences in densities of predator species, although not all observed d i fferences could be explained. For example, in 1977 wolf densities were greatly reduced in Area 1, but not in Area 2. Calf survival that year was greater in Area 1 than in Area 2. The same trend was not evident in 1978, but wolf populations in Area 2 had been greatly reduced and wolves were not abundant in Area 3 (Ballard et al. 1981a). In 1979 predation from brown bears was expected to have been greatly reduced in Area 1 since brown bear populations were temporarily reduced by about 60% (Miller and Ballard 1982). Although other data suggested that reductions in bear density greatly increased calf surviva l (Ballard and Miller 1987), radio- collared calf survival data suggested no improvement. This discre- pancy occurred largely because not all bears were removed from Area 1, and 2 bears which had not been removed killed at least 67% of the calves killed by bears. Overall, from 1977 through 1984 calf survival during the first 5 months of life averaged 26% (74% mortality). Calves 6-12 Months of Age. Starvation (or winter-kill) was the largest overall source of calf mortality from 1 November-May of each year, accounting for 79% (11 of 14) of the deaths. Nine of the winter kills occurred during the severe winter of 1978-79. Predation by brown bears was suspected in 2 cases while an unknown predator made 1 kill. From 1 November-May of each year, female calves (Table 7) had greater survival rates than male calves (Table 8). This was due lar gely to differences during the severe winter of 1978-79. During that year, male calf mortality was 72% ( 1. 00 minus survival rate) while known female calf mortality was only 6%; female calf mortality could have been as high as 30% (1 -0. 703) as suming half of the missing calves died. Regardless, during severe winters, male calves suffer higher (p < 0.05) rates of mortality than female calves. There were no differ- ences (p 0.05) between male and female calf mortality rates during years of moderate winter severity (5% for each sex). Annual Calf Survival Rates . Average annual calf survival rates for female and male calves were 22 and 17%, respectively (Table 9: deter- mined by multiplying rate from Table 6 times rates from either Table 7 or Table 8). Although male calf survival rates were lower than female rates from 1 No vember-May during severe winters, overall annual survival rates were not diffe rent (P >o.05). Combinations of different summer and winter calf survival rates were calculated to estimate ranges of annual survival rates which could occur among different areas and years in GMU 13 (Table 9). Highest calculated calf survival rates for male and female calves was 56% each, while the lowest rates were 2 and < 1% for females and males, respectively. The latter situation occurred during a year of high neonatal losses such as in Area 4 in 1984 and following a relatively severe winter such as in 1978-79 (Table 9). Higher survival rates occurred during years of low nee a te losses (such as in Area 1 in 1977) followed by low winter losses (such as in either 1979-80 or 1980-81). Summer and winter data were collected consecutively within 1978 and 1979, and the estimated survival rates for those years were 15 within the calculated extreme values (40 and 30% for female calves in 1977 and 1978, respectively, and 12 and 30% for male calves in 1977 and 1978, respectively). Yearling and Adult Females. Yearling and adult radio-collared cow moose survival rates were based on 43 and 532 moose years, respec- tively. Overall, yearling and adult cow survival each averaged 95% (Table 7). Lowest adult survival occurred in 1985-86 but that rate was only through Jan 1986 and probably not representative of the entire year. Adult female survival was also relatively low in 1978-79 (a relatively severe winter), 1979-80, and 1981-82, averaging about 92%. Adult survival might have been as low as 77% in 1978-79 if half of the missing animals (N=17) were assumed dead. Lowest yearling survival (75%) occurred in 1981-82. In general, radio-collared yearlings and adults were not monitored frequently enough to accurately detera.l ne causes of mortality. However, here were periods when monitoring intensity was sufficient to allow causes of mortality to be determined: during parturition in 1977 and 1978 when cows were monitored 3-5 times per week, and short yearling mortality studies in 1978-79 when cows were monitored once per week. From October 1976 through January 1986, twenty-one adult radio-collared females died. Of that total, 10 died from unknown causes. Predation accounted for 8 of 11 (73%) mortalities where cause of death was determined; brown bears killed 5, wolves killed 2, and unknown predators killed 1. Three adults starved. Cause of death for 2 yearlings was starvation and wolf predation. Contact with 37 adult radio-collared females was lost, so their fates were unknown. Dates of lost radio-contact were equally divided between snow-free (1 May-31 October) and snow-cover periods. Yearling and Adult Males. Yearling and adult radio-collared bull moose survival rates were based on 34 and 72 moose years, respectively (Table 8). Overall, adult bulls ha ' lower survival rates (65 to 74%) than yearling bulls (87 to 90%). Prior to 1980 any bull was legal for human harvest. Following that date only bulls with 3 brow tines on at least 1 antler or antler spreads 36 inches (91 em) were legal. Also, in 1984 only spiked or forked antlered males were legal in the SRSA, while after that year the regulation applied only to the area lying west of Lake Louise Road in Subunit GMU 13A. Yearling bulls had their lowest survival in 1979-80 when they were legal for human harvest (2 of 3 mortalities). Thereafter, yearling bull survival was relatively high ranging from 86-100%. Lowest adult bull survival rate occurred in 1985-86 (Table 8). How- ever, the rate applied only through January 1986 and may have been biased. The next lowest rate occurred in 1984-85. Adult bull survival declined as the study progresseu ·.:-.94, p >o.Ol), sug- g sting increased vulnerability with age when only bulls with antlers 36 inches (91 em) were legal. Radio-collared bulls had relatively low rates of natural mortality after they attained 2 years of age. Of 13 adult bull mortalities, 12 (92%) were due to human harvest and 1 (8%) was from unknown causes. 16 MOVEMENTS, DISTRIBUTION AND HABITAT USE Group Size Differet ces in average group size per month were determined for radio-collared adults from 1977-1985. There were no differences among years, so all years were pooled. During January through August, )301 of all observations of instrumented moose were of single animals (Fig. 14). In September that proportion began declining, and by October only 19% of the observations were of lone individuals, reflecting rutting concentrations. Proportions of one moose again increased in Nov ember and December. Average group size exhibited similar trends. Average group size was 2 moose from January through July (Fig. 15). Average group size increased to 3.0 in August, 4.9 in September, and 7.6 in October. After October group size decreased to 3.2. These results were similar to many other studies indicating that moose are not highly gregarious (except cows with calves) during much of the year. Largest (N = 52) group sizes occur during the rut and in post-rut aggregations. Generally, cows with calves do not associate with the large rutting groups. Movement Patterns Moose exhibited all of the movement patterns described by LeResche (1974) and many variations not described. Moose were classified into 2 basic categories based on overlap or nonoverlap of winter and summer home ranges: (1) residents--individuals with movements confined to relatively small areas and with portions of their winter and summer home ranges overlapping, and (2) migratory--individuals which moved over relatively large areas and whose winter and summer home ranges did not overlap. Three periods of significant movement were identifiable. These included autumn and spring migration and movements to rutting areas. Movements during the rut were most pronounced for resident moose. During late Septemb~~ and October, several moose made distinct move- ments to upland areas not used during other seasons of the year. These areas appeared to have greater numbers of large-antlered bulls than other areas, and consequently, bull density and behavior may have been an attractant. Both major identified rutting areas within the project area (Clark Creek and Tsisi Creek) had poor human access, and fewer bulls were killed there than in other areas. Migratory moose may also have moved to specif c r ut ting areas, but such areas were not easily identifiable because of the relatively large areas they occupied. Autumn Migration. Dates of autumn migrat i on were variabl a . LeResche (1974) and Van Ballenberghe (1978) both reported that weather, par- ticularly snowfall, was a mediating factor in moose migrations. Heavy snow accumulations ()1 ft or 0.3 m) stimulated autumn migration if it had not already been initiated. Response to lesser intensity storms or accumulations was not predictable. During years of low snowfall, 17 migratory and resident ~se did not move to lower elevational areas until early winter (January-February). Autuum migration occurred as early as October and as late as January. Host moose appeared to initiate autumn migration at about the same time; however, the speed at which they arrived on winter range was variable, ranging from a few days to several weeks and in some cases not at all. Rapid movement to winter range coincided with heavy initial snowfall, while the s 1.ower movements occurred when there was a gradual accumu ation of snow. LeResche (1974) suggested that winter snow depths, forage availability and quality, habitat suitability, and their various combinations dete~ined whether particular winter habitats were used. In years of moderate snowfall, forage and habitat were available at upland sites because snow depths were shallow. During these types of winters, moose did not arrive on winter range until February or March, if at all, and then they may have only remained on winter range for 2-4 weeks. During 1978-79, a relatively severe winter, several moose utilized winter areas they had not used during previous years. For example, from 1976 through 1978 an individual moose maintained a summer range near MaClaren River and a winter range along the Susitna River. Between 21 December 1978 and 14 April 1979, she was relocated 82 km to the south along Mendeltna Creek. In subsequent years (1980- 1984) she used her traditional winter and summer ranges and did not return to the winter 1979 location. Although this moose did not use the impoundment zone, it suggests that other moose might use the area during severe winter conditions. The importance of the impoundment zones to moose during a severe winter could be greatly different from that observed during this study when winters were relatively moderate. Spring Migration. Dates of spring migration were as variable as those observed during autuum months, ranging from March through mid-July. LeResche (1974) suggested that spring movements were in response to disappearance of snow and/or plant greenup. Spring movements during this study apparently were more related to disappearance of snow than to plant greenup. Rate of movement to summer range was also variable. Van Ballenberghe (1978) reported that in the eastern portion of G~W 13 moose departure to summer range occurred from mid-April through mid-June. During some years movements to winter range were completed in 1-2 weeks while in other years, 4-6 weeks were required. Most moose were on summer range by late April or early May where they calved. During some years moose remained on winter range for calving, with migration to summer range not occurring until mid summer; these movements may have been in relation to vegetation greenup. Seasonal and Total Home Range Sizes All moose exhibited seasonal movements within their total home ranges. Distances between winter and summer ranges of migratory moose ranged from 0.6-58 mi (1-93 km). The longer distances were associated with moose which summered in the upland areas of the Clearwater Mountains and wintered along the Susitna and MaClaren Rivers. 18 Use of seasonal home ranges by adult moose was traditional although at least 1 adult changed its home range permanently (see Adult Dispersal section). During severe winters moose may use areas which were not used during winters of moderate severity. LeResche (1974) suggested at traditional use of home ranges persisted over several gener- ations, but whether these conditions persist during severe winters is not known. Also, because yearling bulls disperse more often than females (Dispersal section), traditional usage of parental home ranges as suggested by LeResche (1974) is probably much lower for male than female moose. Seasonal and total home range sizes of resident adult -:ow moose increased (;> <o.OS) with numbers of relocations (Figs. 16, 17, 18, and 19). There was no (p >o.OS) relationship for migratory moose between seasonal home range sizes and numbers of relocations, but there was a (p >o.OS) negative correlation for total home range size (Fig. 20). Apparently, there were large areas between seasonal ranges not used by moose; additional relocations reduced th amount of unused area included in home rang~ calculations. Seasonal and total home ranges for resid~nt moose appeared adequately identified when •. umbers of relocations 13 and 39, respectively, (Table 10). Using these criteria, winter, summer, autumn and total home ranges for resident moose averaged 44, 40, 61, and 112 mi 2 (113, 103, 1S7, and 290 kms 2 ), respectively. Home range sizes were compared by ANOVA (Snedecor and Cochran 1973). Resident winter home ranges were not different in size (p <o.OS) from summer and autumn home ranges, but autumn home ranges were larger (p <o.iO) than summer home ranges. Winter, summer, and autumn home range sizes of migratory adult cow moose averaged 58 , 102, and 124 mi 2 (1S1, 263, and 322 km 2 ), respec- tively (Table 10). Total home range sizes did not appear to be adequately defined until numbers of relocations >40 (Fig. 20). Using those criteria total home ranges a veraged SOS km 2 (195 mi 2 ). There were no differences (p > 0. OS) between winter and summer ranges of m1gratory moose, but autumn ranges were larger than both winter (p < 0.05) and summer (p <0.10) ranges (Table 10). Migratory moose had larger (p < 0. OS) total home range sizes than resident moose (Table 10). They also had larger (p <o.OS) autumn and summer home ranges, but there was no difference (p > 0. OS) between sizes of winter ranges. The larger autumn home ranges of both groups reflected increased movements of moose during the rut (Houston 1968, Phillips et al. 1973, LeResche 1974, Hauge and Keith 1981, this study). LeResche (1974) reported that seasonal home ranges of moose were consistently small regardless of how far a moose moved between seasons. He reported that all studies consistently reported home x:anges that seldom exceeded 2-4 mi 2 (S-10 km 2 ). Home ranges for resident and migratory moose in this study were larger than those reported in the l i terature . 19 LeResche (1974) indicated that cows with calves had smaller home ranges than other moose. Ballard et al. (1980b) reported that home ranges of cow-calf pairs in late spring and early summer averaged 9.7 mi 1 (25 km 1 ). This average was larger than reported in the literature but smaller than those of other sex-age classes, providing additional verification that this group occupies sma l ler areas. New Method of Home Range Calculation. Many of the studies reported by LeResche (1974) concerning home range sizes occurred in areas where elevational relief was usually less than that found in GMU 13. Less than 1~ of 4, 700 relocations of radio-collared adults occ.urr'9d at elevations >4,000 ft (1,220 m) and only 3% occurred at elevations above 3,600 ft (1,097 m). Thirty-one percent (7 ,259 mi 2 or 18,800 km 1 ) of GMU 13 (23, 784 mi 2 or 61,600 km 2 ) is comprised of unusable habitat for moose (lakes, glaciers, or areas > 1, 220 m elevation). Large areas of nonhabitat are included in seasonal and total home range calculations using Mohr's (1947) method. To provide a refined estimate of actual home range size, Mohr's method was modified by basing calculations on actual habitat use according to methods described earlier . Home ranges for 13 adult cows (9 residents and 4 migrants) were calculated using the modified method. Estimates of seasonal and total home range sizes were smaller (Table 11) than those calculated using Mohr's method but still larger than those r e ported in the literature. Wi .nter and summer home ranges calculated by each method were similar (p >o.OS), but total home range sizes were not (p >o.OS). Winter and summer home ranges did not increase (p >o.OS) with numbers of reloca- tions as with Mohr's method of calcu lation, whereas autumn home ranges were negatively correlated (p <o .OS). Winter home range sizes were not different (p >o.OS) from summer ranges for resident and migratory moose using Mohr's me hod but with the modified method winter home ranges were larger (p > 0 .10) than summer home ranges for resident moose. Also, there was no difference (p >o.OS) between winter ranges of migratory versus resident moose. Possibly winter snow depths restrict movements of both types of moose. Both methods indicated that summer and total home ranges of migratory moose were larger (p < 0.10) than those of resident moose. Dispersa l and Home Range Fo r mation During March 1981 sixteen calves t 8 males and 8 females) and 1 yearling associated with radio-collared cows were captured and radio-collared in an attempt to investigate timing of parent-offspr i ng separation, rates of dispersal, and home range formation of subadults. Immediately follcrNing capture, radio contact with 2 calves was lost due to unknown causes. Timing of Separation. Average age of separation from parents was 14 months (SO= 4.2). Gasaway et al. (1985) reported that in interior Alaska, only 2 of 20 vearlings remained with the i r cows after 1 year of age. In this study, 81% of 16 yearlings remained with their cows > 1 year (Fig. 21). Thirt y-one percent of the separations occurred 20 during late June and July while 501 occurred dur g September and October. Separations at that time appeared induced by aggressive behavior of either cows or bulls during the rut. Gasaway et al. (1985) reported that once initial separation of parent-offspring occurred in interior Alaska it was permanent. In this study, 5 of 15 (331) yearlings and the 2-year-old were observed in temporary reassociation& with their cows from 1 -6 times (~ = 2, SD = 2). During parturition, adult cows that were still in associ- ation with the previous year 1 s calf exhibited varying degrees of aggressive behavior toward the yearli!l g . If the new calf survived, separation between cow and yearling was usually permanent. However, if the new calf died, there was a tendency for the yearling to remain with the cow at least through summer months. Types and Rates of Dispersal. Gasaway et al. (1985) reported that offspring selected home ranges that partially overlapped those of their parent; offsprings 1 home ranges overlapped at least half of parental home range. The maximum distance that offspring were observed from parental home ranges was 6.2 mi (10 km). Subadults in this study exhibited a different pattern. Dispersal was classified into 3 categories based on subsequent movements and home ranges of offspring in relation to those of the parent: (1) No Disperal -- Offspring mimicked movements of both summer snd winter home range of parent . Exploratory movements outside of traditional home ranges may occur during autumn of 1st and 2nd years following separation; (2) Partial Dispersal--Offspring share either winter or summer range of parent, but at least one of seasonal ranges is separate and dis- tinct from the parent. Offspring may ultimately mimic home range of adult but only after extensive movements outside of historical parental home range for at least 1 year; and (3) Full Dispersal-- Offspring established separate winter and summer home ranges which were not shared or, if shared, separated t~mporarily from that of the parent. Development of new home ranges may occur over several seasons. Nine of 15 (60%) offspring partially (N = 4) or fully (N = 5) dis- persed from the parental home range. More male than female (p <o.05) offspring dispersed. No male offspring remained fully within the home ranges of their dams. Females us u ally (75%) occupied the home ranges of their dams. Dispersal rates were comparable to those reported by Houston (1968) in Wyoming but higher than those reported for interior Alaska (Gasaway et al. 1985) and portions of Sweden (Cederlund in press). Several factors influence dispersal in moose populations (Houston 1968, Gasaway et al. 1985, Cederlund in press), but density may be par~icularly important. In interior Alaska where full dispersal ratas were'low, moose densities ranged from 0.2 moose/km 2 in 1975 to 0.3-0.6 moose/\\m 2 in 1978 and 1984, respectively (Gasaway et al. 1985). Moose densities during this study ranged from 0.6-0.8 moose/km 2 and were increasing. Moose densities were 3-4 times greater than those in interior Alaska which may partially account for the higher dispersal rates. 21 Dispersers appeared to move to areas of lower moose density. The receiving areas bad greater bunting pressure and lower bull densities than areas from which dispersal occurred. Host dispersers were bu L s which moved to either the Denali Highway or Lake Louise flats. If the proposed hydroelectric project results in lower moose densities and if there is a relationship between moose density and rates of yearling dispersal, then fewer moose will d i sperse. Therefore, not only will fewer moose be available for harvest in the project area but also in areas far removed from the project where heavy hunting pressure may have depleted a population. Home Range Formation and Size. Average home range sizes (modified method) of cows and their offspring were positively correlated (p <o .05) (Fig. 22). Offspring of cows with relatively large home ranges also had large home ranges. Hale offspring had larger (p < 0. 05) seasonal and total home r anges than females. Winter, summer and total home ranges for male offspring averaged 13, 10, and 34 mi 2 (34, 27, and 87 km 2 ), respectively, while females averaged 7, 6, and 29 mi 2 (19, 16 and 76 km 2 ), respectively. Changes in seasonal offspring home ranges were variable, and some changes did not occur until about 2.5 years following separation from the cow (Table 12). Adult Dispersal Use of seasonal home ranges by moose is traditional (LeResche 1974). During this study only 1 of 101 (1%) radio-collared adult females dispersed from their traditional home range. The single dispersing moose occupied a relatively small home range in the vicinity of the Susitna River from March 1977 through mid-August 1978. By 26 October 1978 she was relocated at the Dadina River, 110 mi (177 km) from her previous location. She maintained a resident home range in the Dadina area at least through 1981 when last re l ocated. Prior to this move- ment, the longest reported movement was 170 kms (106 miles) from the Northwest Territories (Barry 1961). River Crossings Timing. Fifty-nine of 113 (52%) radio-collared moose crossed the Susitna River in the vicinity of the impoundments on at least 170 occasions during 1976-1984 (Fig. 23). Thirty-five (59%) of 59 moose crossed the river at least once or twice. Greatest number of docu- mented crossings was 8 by 4 moose. Monitoring intensity was too low to detect all crossings, particularly when animals crossed over and back within a 10-14 day period. River crossings occurred during all months of the year, but most occurred during mid late winter (peak number in April) when moose were on winter range at lower elevations (Fig. 24). A second peak in crossings occurred during September and October, presumably because of increased movement during the rut. 22 Location. Crossin& locations in relation to the proposed impoundments were examined by plotting straight lines between consecutive moose locations which crossed the river. Lines bisecting the river were assumed reflective of crossing locations. Although that assumption was less accurate as time interval between relocations and the dis- tances between relocations increased, the analysis provides an indi- cation of areas where crossings were concentrated. Hoose crossed the Susitna River along the impoundment corridor from Devil Canyon damsite to the mouth of the Oshetna River. Crossings were concentrated (Fig. 25) in areas which bad characteristics con- ducive for easy movement. Several areas in the immediate vicinity of the impoundments were used extensively. These included: the mouth of Tsusena Creek just down- stream from the proposed Watana damsite, the area midway between Watana Creek and Jay Creek, and the areas adjacent to the mouths of Kosina and Jay Creeks. On the upper end of the Watana impoundment, crossings were also concentrated just downstream from the mouth of Goose Creek and immediately abcve the Oshetna River mouth. Areas where few or no river crossings occurred, such as in Devil Canyon and around the gauging station, were characterized by steep terrain which apparently restricted access. Because river flow characteristics were similar among areas crossed and not used by moose, actual fording areas may be influenced by surrounding terrain. Where adjacent terrain gradually slopes to the river and moose movements are not restricted by cliffs or steep embankments, more crossings were recorded. Winter Use of the Impact Zones Dur ing winters of moderate severity, radio-collared moose were seden- tary on winter range. Comparison of density stratification maps between autumn censuses (with population estimate) and winter distri- bution surveys (no population estimate) depicts seasonal use of habitats. Comparison of fall 1980 with winter 1981 distribution (Figs. 26 and 27, respectively) and fall 1983 with winter 1985 dis- tributions (Figs. 28 and 29, respectively) suggest that tb<> greatest change in seasonal distributions occurred in the Watana Creek-Fog Creek areas, the Watana Lake-Jay Creek areas, and the vicinity of the big bend of the Susitna River. The latter areas were characterized by low moose d e nsities in autumn, but large densities during winter. This was due to shifts from high elevations in autumn to lower ele- vations adjacent to the Susitna River during winter. Watana Impoundment. During winters 1981-1983 and 1985, total counts of moose were conducted within the Watana impoandment zone at &.n average survey intensity of 3.8 min/mi 2 (1.5 min/km 2 ). Comparison of annual counts suggests that late winter use of the Watana Impoundment during winters of moderate severity was highly variable, ranging from 42 moose in 1981 to 580 in 1983 (Table 13). Moose densities in the impoundment zone during these years ranged from 0. 4-6.0 moose/mi 2 (0.2-2.3 km 2 ). 23 Observability of moose in the Watana impoundment zone was low because of large topographical variation and dense overstory vegetation. Also, snow and lighting cor-ditions during the study were rarely optimal. Counts were conducted in spite of poor conditions because telemetry studies indicated that the largest numbers of moose occurred in the impoundments during those time periods. Calculated correction factors were often high because of low observability. Telemetry data s•pport the use of high sightability correction factors during these season . For example, only 2 of 7 and 2 of 8 radio-collared moose in 1983 and 1985, respectively, were observed during the counts. Devil Canyon Impoundment. The Devil Canyon impoundment zone was also counted in late winter but only in 3 years (Table 14). Count con- ditions were always poor, and moose observability was hampered by dense overstory vegetation. In 1983 and 1985 only 14 and 16 moose were observed, respectively. In comparison to the Watana impoundment zone, moose densities were low, ranging from 0.5-1.0 moose/mi 2 (0.2-0.4/km 2 ). Vegetation Use Preliminary analyses based on overstory vegetation indicated that spruce and willow vegetation types were selected out of proportion to their availability while tundra types were avoided (Ballard et al. 1985). The latter analyses did not indicate why a particular type was selected. If moose select habitats based primarily on the quantity of food, such analyses could provide misleading conclusions. Avail- ability and use of browse species, in addition to overstory vegetation analyses, were compared. The entire moose primary impact zone was divided into 3 subsegments based on proximity of the proposed impoundments. Because differing (p <o.05) quantities of browse occurred between the impoundments and outside them, it was not appro- priate to combine the areas for comparison of moose use versus avail- ability (Steigers and Becker 1987). Outside of Impoundments. Moose used browse vegetation types outside of the impoundments in proportion to their availability (Te.hle 15) except in the following cases : in winter (January-April) and summer (May-August) the medium shrub category was avoided (p <o.05). It was also avoided annually (X 2 = 28.9, P <o.005) while the very low strata was preferred (X 2 = 16.8, P <o.Ol). Watana Impoundment. Similar to area s outside the impoundments, there was no selection for any of the vegetation strata within the Watana Impoundment either by season or pooled (Table 16). Devil Canyon Impoundment. Because browse productivity was substan- tially lower within the Devil Canyon Impoundment than further up- stream, only 4 categories of browse strata were defined (Table 17). Low use of the area by moose was reflected by only having a total of 40 moose point relocations for utilization calculations. During all three seasons, there was no selectivity for browse by quantity strata (Table 17). 24 All Areas Combined. Based on the preceding analyses, moose did not appear to be selecting habitat on the basis of browse biomass. A · different interpr tation of seasonal habitat use was obtained wheL all 3 populations were pooled. During winter, moose exhibited a (p <o.OS) preference for areas with relatively little browse (Low, Very Low, ~~d Scarce browse biomass strata). Within the Watana Impoundment, all browse areas appeared important, although there was no statistic preference or avoidance for High, Medium, or Zero biomass strata. Outside t he impoundments during winter, there was an avo'dance of all strata except Medium Forest s trata, where there was an apparent but nonsignificant preference. During summer and autumn, the pooled data analyses were similar to those based on individual browse populations pres e nted earlier. Only Very Low and Scarce biomass strata were prefer red in summer and only in the Watana Irnroundment population. Winter was the only time peri od when moose appeared to be selecting particular habitat types based on browse biomass. They did not, however, indicate a preference for areas of high browse biomass (usually upland types), suggesting that other factors were important. During summer moose were widely distributed over the basin and did not avoid upland vegetation types. In autumn, food availability appar- ently does not limit moose distribution, so areas are apparently ~elected on the basis of factors other than food. Moose were not selecting areas based solely on quantities of bro~se. Other factors, such as thermal and escape cover, tradi tional use, snow depths, elevation, slope and aspect, and behavior, all affect where moose were located. The only area outside of the impoundments that a s not avoided in winter was the M·d'um-Forest strata while most of h e Watana Impoundment area was '0r.. " t ed by spruce. This strongly implies that the areas preferred by moose are dominated by spruce overstory. Earlier analyses based on overstory vegetation alone (Ballard et al. 1985) supper ' the hypothesis that spruce cover types are important habitats for wint e ring moose in southcem:ral Alaska. Nineteen percent of the basin is composed of spruce stands and 35% of the tctal moose obser ations ga ; erecl 'luring 1976 through 1981 were located in spruce overstory habitat s (Bdllard et al. 1982b). Elevational Use Diffarent elevations were used seasonally and annually by Susitna area moose. Use of lowest elevatioual strata occur.r d in April. As snow melted and retreated, moose moved to 1.Lgher elevations in May and June (Fi,'t. 30). After calving, they mov e1 to h i gher elevations in July, with downward movements during August and September. During the rut, higher elevations were again selected, reaching their highest level by October . In November, moose begar1 movements toward lower elevations which continued into March and April. The latter movements were apparently in response to deepenin& snows and/or lower browse avail- ability (Fig. 30). 25 Shifts in elevational use were also evident among seasons whe . percent frequency of occurre ce of relocations were compared with elevation (Fig. 31). Peak elevational use during winter occurred near 2,600 ft (792 m) elevation, while jn summer anrl autumn, the peaks shifted from 2,800 ft (853 m) to 3,000 ft (914 m) elevation, respectively. Elevational and vegetation use by Susitna moose in winter depends to a large extent on the severity of individual winters. As winter severity increases, the percent of moose utilizing lower elevations increases (see Effects of Snow-Elevational Use section). Elevations from 1,800-3,000 ft (549-914 m) were used disproportion- ately to their occurrence (Fig 32). Elevations over 3,000 ft (914 m) were used less, indicating an avoidance of the higher elevations where food and cover were less abundant. Only 16 of 2, 984 observations (0.5%) fro~ 1981-1984 were at e levations >3800 ft (1,158 m). Slope Use During winter, slopes were used by moose in proportion to their occurrence (X 2 = 0.01 to 0.10, P >o.05) (Fig. 33). During summer, flat areas were preferred (X 2 = 11.73, P = 0. 005) and gentle (X 2 = 5.17, P = 0.07) and moderate (X 2 = 6.20, P = 0.04) slopes avoided. During autumn, gentle (X 2 = 10.4, P = 0.01) and moderate (X 2 = 9.00, P = 0.02) slopes were preferred and flat areas avoided (X 2 = 21.41, P = 0.005). During autumn moose utilized higher elevations where terrain was more varied. Dur ing winter, snows apparently forced moose to use lower elevations and whatever slopes were available. Aspect Us e Annually moose preferred north-and south-facing slopes, whereas east, southwest, or west aspects were neither avoided nor preferred (Fig. 34). Other aspects (flat , northeast, southeast, and northwest) were avoided. No significant differences (p >o.05) in aspect use occurred among seasons (Table 18). All seasons combined, southwest-facing slopes were avoided (p <o .05) (Table 18). Activity Patterns Daily. Moose activity was recorded on 4,078 occasions during 1977-1985. Because all observations were from fixed-wing aircraft, they were b ias ed toward daylight hours between 0700 and 2400 hrs (Fig. 35), with the majority between 0800 and 1800 hrs. Moose were observed bedded on over half (52%) of the observations. Standing and foraging activities accounted for only 31 and 12%, respectively, of the activity categories. If moose had been monitored more often during nocturnal and crepuscular hours the percent of foraging obs ervations probably would have increased. There was a slight increase in the proportion of time moose spent bedded during the middle of the day (Fig. 36), with early morning observations more heavily weighted t oward other activities. 26 Monthly. Number of activity observations per month canged from 665 in March to 167 in January (Figs. 37 and 38). TbrJ lowest (X = 44%) occurrence of bedded observations occurred during summer, increasing in autumn (X= 55%) and winter (X= 58%). Conversely, the proportion of observations where moose were observed foraging was greater during summer (X= 17%) than at other times of the year (X= 7.5%) (Fig. 39). Physiologically, summer is the time of greatest energy intake for moose. Females with calves need high intake of food, both for milk production and for deposition of body fat reserves to sustain them through the winter. Males also take advantage of increased avail- ability of forage to depoait fat reserves for the rut and for over- wintering. During winter, moose are relatively sedentary, reflecting a negative energy balance which partially accounts for the higher frequency of bedded activity. Effects of Snow on Moose Distribution Assessment of winter severity is critical to understanding movements and population dynamics of moose. In the Susitna Basin, the winter of 1971-72 caused substantial mortality in the population, especially within calf and yearling cohorts. From 1977-1985 elevational use by moose was correlated (p <o.OS) with winter severity; during deep snow years moose used lower elevations. To fully assess impacts of the proposed project, moose movem~nts and habitat use during a relatively severe winter will have to be monitoreo. Because severe winters occur on an average of 3 out of 22 years and very severe winters only 1 of 22 years (see Ballard et al. 1986), it was necessary to develop the capability to predict winter severity by early February each year . The abi lity to predict winter severity early in the year would be beneficial because it could alert managers and researchers that potentially serious conditions existed. A method for quantitatively assessing winter. severity in relation to other winters was developed. The following subsections explain the relationships. Winter Severity Index. The winter severity index (WSI) for the middle Susitna River Basin was based on Soil Conservation Service (SCS) snow survey data collected from winter 1963-64 to 1985-86. Four SCS snow sites were used for the i ndex because of their proximity to the moose study area. They included: (1) Fog Lakes, (2) Square Lake (pri ~ to 1982 known as Oshetna Lake), (3) Monahan Flats, and (4) Lake Louis e. Tht'ee snow depth readings (January-March) from each of the 4 snow courses were summed and div i ded by the number of courses reporting. The WSI was compr ised of the average 3 month cumulative snow depths (Table 19). The index was based on the following assumptions: (1) the amount of snow cover during mid to late wint e r (January-April) was more impor- tant in terms of moose mortality than early winter snow depths; and (2) snow depths were t'te most important factor causing malnourishment in moose through 2 mechanisms--as depths increase, browse species are covered, necessitating cratering by the moose and more energy use per unit of food, and movements are restricted, requiring increased energy use to travel. 27 Three categories of winter were identifiable using the WSI; (1) severe winters when the WSI ~28.0, (2) mild winters when the WSI S18.0, and (3) moderate winters when WSI ranged from 18.1-27.9 (Fig. 40). Moose mortality data suggested that the 3 categories of winter severity were justified. Winters 1971-72 and 1978-79 were considered severe and resulted in substantial moose mortality (Stephenson and Johnson 1973, Ballard and Gardner 1980, Eide and Ballard 1982). Winter 1974-75 was also thought to have been rela- tively severe based on autumn moose calf survival (unpubl. data). All 3 severe years bad relatively high WSis (Table 19). Prediction of winter severity in the Susitna impoundment area during a current winter by early February was obtained by the following method: January snow depths from the 4 SCS snow surveys were averaged for each of 22 years. Annual WSis were then plotted against January snow depths f or the same 22 years. Correlation analysis was used to predict final winter severity (Fig. 41). Revised winter severity predictions could also be made following February snow course readings using t 1e same procedures (Fig. 42). Elevational Use versus Winter Severity. Monitoring intensity of radio-co l lar ed adult moose was increased during winters 1981-1984 to determine winter use of impoundment zones. There appeared to be a relationship betwe en elevat i onal use by moose and winter severity. Proportions of monthly relocations at elevations S2,200 ft (671 m) (h i gh pool level of Watana Impoundment) were compared with monthly winter severity indices (Fig . 43). The proportion of radio-collared moose at elevat i ons ~2,200 f t (671 m) was correlated (p <o.OS) with the WSI. The corre lation was used to predict percentages and numb e rs of moose wh i ch would potentially use th ~ area planned for inunda tion (high water level at 2,200 feet) during wi nters o f varying deg r ees of severity. Sixteen percent of rad i o-collared moose relocat i ons were at elevations S2 ,200 ft (671 m) dur i ng May through December . Thes e probably repre- sent year-round res i d e nt moose occurring along low e r elevations of the middle Sus i tna River Basin. As snow accumulates, moose which occur at higher elevations move downward and the proportion of the moose population utilizing the impoundment zone i ncreases. During moderate (average) winters, the p r oportion of radio-collared moose in t e impoundment zone increased to 17% in J anuary, 29% in February, 35% lD March, but then declined to 17% i n Ap r il when snows begin t o melt a d recede. If a severe wi nter sim il ar to 1971-72 were to occur, t he regression pred i cts that over 5 0% of t he middl e basin moose populat on would utilize the i mpoundment zones (Fig. 43). Assuming that the radio-collared moose were representative of the 2,400 estimated wi thin the middle Susitna Ba sin, the correlation would predict that during moderate winters an average of 590 (Jan ua r y = 17 %, February = 29%, March = 35 %, April = 17 %; Average = 25 %; 2 ,400 X 0.246 = 590) moose would use the i mpoundment zone. During severe wi te r s , t correlation p r edicts an average of 1,552 moose would us e the impoundments f::-om January t hrough April (January = 61%, February = 55%, March= 67 ~, April = 71 %; Average = 63%; 2,400 X .634 = 1 ,522 ). 28 No field data exist to support these estimates except 1 count of the Watana Impoundm e nt in winter 1983 when 580 moose were estimated during a winter of moderate severity (Table 13). Counts conducted during three other winters resulted in estimates of about 40 to 300 moose. To fully test these predictions, rad1o-collared moose should be monitored and a winter census conducted during a r~latively severe winter. 29 IMPACT MECHANISMS AND PREDICTION OF IMPACTS DUE TO HYDROELECTRIC DEVELOPMENT The project is expected to affect moose through a number of different mechanisms. These effects would vary greatly over time and space. It is particularly difficult to predict population changes where several mechanisms may have c umulative effects and the magnitude of these effects may vary depending ~n the si~e of the population or current environmental conditions. For example, a given set of impact mecha- nisms might cause a permanent reduction in moose densities in 1 drainage . yet densities in another drainage may decline only during severe winters. The net impact of those mechanisms on the entire populatio. would vary according to patterns of winter severity , loca- tion of t~e mechanisms, and movement patterns of moose. There is no perfect way of incorporating this variability into impact predictions. We selected 2 separate approaches. The first approach used descriptions of subpopulations to portray spatial vari.ability. The second involved estimating the number of moose which could be supported by the vegetation to be destroyed by the project. The third approach, not covered in this report was to develop a population model that could be used to portray temporal variability (see Ballard et al. 1986). Impact Mech~~isms Development of hydroelectric power on the Susitna River would impact moose populations both directly and indirectly through a number of different mechanisms. Impacts on moose can be classified into 3 broad categories: (1) habitat alteration, (2) impacts on population dyna- mics processes, and (3) socio-political-economic consequences. In this discussion we do not attempt to discuss socio-political-economic consequences except as related to reductions in moose hunting and viewing opportunities. Both beneficial and detrimental impacts on moose are likely to occur, but available literature is inadequate to guide assessment of impact magnitude due to many of the mechanisms. Consequently, until comparative pre-and post-i mpou ndment studies document the nature and exte t of impacts, prediction of i mpacts would remain speculative. We formulated hypotheses to aid in assessing how hydroelectric development mi ght impact moose populations. Hydro- electric development was divided into components of the biotic and abiotic environment which directly and indirectly influence factors regulating moose populat i on dynamics. The hypothesized processes are summarized in a matrix-type table (Table 20). Construction and operation aspects of the proposed project were categorized into 12 major project actions. Effects ~f these actions on the moose's environment were then categorized into major impact mechanisms with predicted negative and beneficial influences on moose population processes. How these impact mechanisms are likely to impact moose and ultimately manifest themselves in t:he moose population is detailed in subsequent sections of this repor t. 30 Most impacts are expected to occur during construction and the first 25 years of operation. However, several impacts would occur over the length of the project . Specific impact mechanisms may impact moose positively or negatively and may involve only certain ·;egments or subpopulations of moose. Changes in a moose herd due to hydroelectric development may be difficult to measure and may occur very subtlety over time. An impact which would have been unimportant under normal healthy preproject situations may become important, particularly if it occurs tdth other impacts. Classif i cation And Identificat i on Of Impacts For discussion purposes, the importance of various types of impacts in relation to th i:3 specific project were classified into 3 c ategories. These categories are based on the potential significance of an impact and on our current ability to detect significant changes in specific moose population parameters. The three categories are: Im portant Impacts: Important pru j ect-induced impac t s are those which available ev i dence indicates, individually or in summation, have a high probability of causing measurable change in reQose population size and/or productivity . Such change is often manifested through reductions in moose natal i ty or increases in moose mortality, or may ind i rectly alter a process which affects a key moose population parameter; e .g., alter ing predator/prey ratios may increase moose mortality . Such i mpac t s a r e usually significant and result in lower population size and/or c h a nges in distribution, ultimat ely reducing human consumptive and non c onsumptive uses. Potentially Important Impacts: These are project-induced impacts, which individually or in summ at ion, potentially could alter moose population s i ze or product i v i ty but for wh i ch insufficient evidence exists to confirm the i r s i gn i f i cance or potential to limit the popu- lation. Poten tially i mpo r tant i mpacts may be difficult to conf i rm and quantify because impact mechani s ms may mask their effects, or our abi l ity to detect changes ma y be i n adequate . Un i mportant Impacts. Un i mport an t impacts are those which data and logic indicate would have a low probability of altering moose popu- lation size and wh i ch would not const i tute a s i gnificant lim i ting factor. These impac t s may affect the s urvival or behavior of indi- v i dual animals. Based on baseline biological dat a pr esented i n previous ·ect ions and on general classifi cat i on of i mpacts descr ibed in the previous 2 sections, speci fi c i mp acts are described (not in order of antic i pated magnitude ): Important I mpacts (I .I .). 1. Permanent h a bit at loss due to i mpoundments and othe r permanent facilit i es would have an adv ers e permanent impact on ar ea moose populat i ons . 31 Rationale--Loss of ungulate habitat is not necessarily detrimental, e.g. habitat lacks components that contribute to potential ungulate carrying capacity. The proposed Susitna Hydroelectric Project impoundments and other facilities would eliminate habitat used by moose during winter and spring. This loss would significantly reduce carrying capacity because winter and early spring are critical periods for moose. Moose usage of wintering areas is highly traditional, so although ad ,quate habitat may be available in other areas, moose would still suffer high rates of mortality. In the long term moose numbers would be permanently lower because of this habitat loss. Timing of habitat usage is an important consideration when determining relative value of winter habitat. Although some moose subpopulations may utilize the same winter habitat annually, others may only use it during severe conditions. Intensive use during severe winters may vary from a few days to several weeks, during which time the long-term capacity of the habitat may be exceeded. Even if carrying capacity is exceeded, the overall mortality rate of the moose population may be less than if the habitat were not available. Slight reductions in mortality rates during severe winters can allow rapid recovery during subsequent years of mild winters. Moose Population Parameters to be Altered--If adjacent habitat is either at capacity or not available, e.g., deep snow, several popu- lation parameters could be altered. The magnitude of some impacts might be masked because they involve moose not directly affected by an impact mechanism. Both seasonal and year-round residents would be displaced fro~:~ the project area. Numbers of moose dying from star- vation (winter kill) are expected to be relatively high for several years. Winter-weakened moose would suffer higher rates of mortality from predation by wolves in winter and bears in spr ·ng. Calf moose mortality would be especially high, and annual recruitment may be less than mortality. Survival rates of displaced adults are expected to be relatively low. Surviving adults would be in poorer physical condition resulting in lower rates of calf production. Calves would be smaller and less viable, hence more vulnerable to predation and increased early spring and summer accident -caused mortality and other nonpre- dation losses. Because the current moose population is slowly increasing at a rate of 3-5% annually, increases in mo~tality would likely cause the population to decline. Reductions i n calf survival would preclude dispersal to other areas. The culmination of all events may result in extinction of several subpopulations of moose which reside in impact areas or depend on these areas for critical winter range. Adjacent subpopu- lations of moose would compete with cisplaced animals and would suffer increased winter kill and predation but at lesser rates than displaced moose. Impacts on Human Uses of Moose--A significant decrease in the numbers of moose available for subsistence and recreational harvest in the project vicinity is expected. Dispersals would be reduced, so numbers of moose available for harvest and viewing in surrounding areas would 32 also be reduced. This particularly applies to the Denali Highway bull moose population which is heavily hunted and is partially dependent on dispersals from the Susitna Pr ject area. I.I.-2. During and following reservoir-filling, displacement of moose and disruption of seasonal movement patterns would c r eate abnormal concentrations of moose adjacent to the impoundments. This displacement would attract and concentrate predators, resulting in higher predation r t es. Rationale--Predat i on by brown bears and wolves are currently the largest sources of mortality affecting dynamics of the Susitna area moose populat i on (Ballard et al. 1980, 1981, 1985). Typi c..ally, the sex and age of moose killed by predat ors is determin d by the vulner - ability of the prey. Usually predation focuses on th e y01 :ng and old of a population (Mech 1970). Exceptions to this rule commonly occur when deep snow results in animals bP.coming vulnerable to surplus killing (Eide and Ballard 1982) by impedence of movement, or espe- cially weakened by malnutrition or disease. Bears are typically facu l tative predators, whereas wolves are considered ob l igate pre- dators (Bal J.ard and Larsen 1986). Because moose and pr d aters would be concent !'ated at abnormal densities, both displaced and resident moose woulf be subjected to i ncreased levels of pre dation. Displaced moose wou l d be particularly vulnerab le because of stress, weakened condition , and lack o f familiarity with esca?e routes. Although displaced moose may d i e as a result of other impact mechanisms, moose of all ages are expected to suffer inc r eased mort:ality from pred .. c i on. Resident moose would be less vu lnerable than displaced moose but more vu l nerable than prior to the project due to increased competition for forage and living area and an increased number of predators. Resident calves wou l d be more vulnerable than adults becaus e this age class is usually subjected to higher !Dortality rat e s. In con j unction with other mortality f a ctors, incre ased preda t i on could significantly decrease the moose population and hold it at a lower densit y . Bee use there are no fast act i ng fe edback mechanisms between large ungulates and t heir prin cipal predators (wolves and bears), such population decl i nes and the r e sult i ng lower threshold l evels could span decades (Gasaway et al. 1983; Ballard and Larsen 1986 ). Although black bears c an also be significant predators of moose (Fra nzmann et al. 19 8 0), the y are currently significantly less impor- tant than brown bear and wo l ves in the Sus i tna Project area (Ballard et al. 1985). However, much of the current b l ack bear habitat would be eliminated by the project (M i ller 1984), potentially causing displaced black bears to be an add i tional source of predator mortality. Even though black bears would probably be eliminated from the area , the short-term add i t ive mor~ality to the moose population could accentuate a mo o se population dec line . Moose Population Parameters to be Altered-~ ortality from predation would i ncrease during a ll seasons, but particularly during late wi n ter and spring . 33 Impacts on Human Uses of Moose--Wolf and bear predation are generally considered to be additive ources of mortality, hence these predators compete directly with humans for the moose harvest. If predation contributes to a moose populat ion decline or maintains the population at low densities, human harves t s of moose would be greatly curtailed or eliminated unless a harvestable surplus is regained. I. I. -3. Open water below Watana Dam and downstream from the Devil Canyon Impoundment, in addition to ice shelving, may block access to traditional winter and calving areas. Rationale--Presence of open water during winter when ambient air temperatures are relatively low is expe cted to impede and possibly halt river crossings. Under pre-project conditions moose were found to cross the river during all seasons of the year (see River Crossings section). During periods when i c e is either forming or thawing, movements across the river are p r obably most hazardous . Moose may not cross major rivers when ice is of varying thicknesses, and thawing conditions occur. These types of hazardous conditions would exist in the vicin i ty of impoundments throughout autumn, winter, and spring. Opposing views exist as to t he potential importance of this impact factor. Bonar (1985) reported that moose crossed o pen water near Revels toke Dam at air tempera t ures of -20 degrees C. However, air temperatures in the Su sitna pro j ect area are quite ofte n lower than those found in southern Brit i sh Columbia. Harper (1985 at Fort St. John, Brit i sh Columbia (several hundred kms north of Revel s t oke), believed that open water downstream from t h e Bennett Dam was a major barrier to moose movements during win t er. He provided ob_ervations which suggested that moose were not would i g to cross open wa ter when air temperatures were about -30 to -4 0 degrees C. During winter 1979-80, moose refused to leave an i sland which was i nunda ed by 1 meter of s lus h ice caused by ice jams downstream and surrounde d by open water. The net result was tha t a t least 23 moose d i ed f rom exposure. Hi gh moose mortality in the vic i nity of reservoirs duri ng and after i ce format i on has been reported from the Sov i et Union (Danilov 1986). These lat ter observations s uggest t hat blockage o f movements may severely i mpact moose directly by mortality o r i ndir- ectly by preventing access to i mportant habitat. Use of seasonal habitats by moose is trad i t i onal (LeResche and Davis 1974, Van Ballenberghe 1978, Ballard and Taylor 1980, and Gasaway et al. 1983). This usage pattern suggests that individual moo se have developed successful strat egies for using seasonal environments. Although remaining hab i tat s u rround i ng the Sustina project may be c apable of supporting more moose, displaced moose with current survival strategies may not adapt qu i ckly enough to the l oss of hab i tat to avoid mortality. A similar scenario ex i sts for white- tai led deer populations which yard up during winter and may stay in an overb rowsed area and starve even though sui table hab i tat e x i sts in adjacent areas (Taylor 1965). 34 Hoose Population Parameters to be Altered--Mortality due to starvation would increase. Relatively large moose die-offs may occur during severe winter conditions because of blockage to winter range. Even- tually moose may adapt to th i s phenomenon, but populations may be held at low levels by artificially high densities of predators. Lower numbers of moose may become pregnant due to d i sruption of social behavior and poorer physical condition as a result of malnutrition . Calves are expected to experience greater rates of natural mortality due to accidents, pneumonia, and other nonpredator forms o f mortality. Rates of mortality from bear and wolf predation wo u ld also be higher due to weakened condi tion and crowding. Short-term mortality from hunting harvests may increase due to moose concentrating in relatively accessible areas. This latter impact assumes hunting regulations are not modified to reduce moose vulnerability . Impacts on Human Uses of Moose--Numbers of moose available for harvest lJO':mld decrease. In addition to reduced densities in the immediate p roject area, surrounding areas would exper i ence population reductions •,ecau •. e of the lack of emigration .rom the project area. I.I.-4. Ice shelving, open water, thin ice, and floating debris would ~ause direct mortality to moose attempting to cross impoundments. Rat i ~nale--Most moose populations experience direct morta l ity from nat .ral factors such as falling through thin ice or injuries resulting from slipping on ice (W. Ballard, A. franzmann, R. Hodaferri, and others, unpubl. data). Such accidents normally occur when moos A encounter bodies of water. This type of mor ality is usually ins ig- nificant in population dynamics and is considered density independent. These types of accidents would cont i nue to occur regardless of whether the project is built, but because more area would be covered by water and ice, conditions would be less stable. The Susitna Ri ver below the Watana and Devil Canyon dam sites is not expected to freeze as it h as under natur.<il conditions. If freezing does occur, thickness and stab ility of resulting i ce would be sub- stantially different than p resently occurs. Moose generally cross waterbodios dur i ng ice-free periods or when ice is sufficiently thick to support them (this study). Rega r dless of whether this behavior is learned or innate, moos t?: may not ad a p t to a bnormal thawing and icing conditions. For example, moose may a ttempt to cross ice-covered areas during time per i ods when such crossings were normally safe, thereby increasing mortality due to ice-related acc i dents aud drowning. I ce shelving along impoundment edges may pose hazards to moose attempting to c r os s o p en water or ice-covered areas. Depending on steepness and surface characteristics of ice shelves, moose may be unable to escape from open water. They may a l so be unable to escape if they fall through i ce wh i le crossing. Fatal injuries due to slips on ice shelves occur naturally, but frequency would incre ase as a result of the project. Float i ng debris may also increase moose mortality due to drown i ng. 35 Direct ungulate mor t ·lity attributable to thin ice, ice shelves, and floating debris is not well documented because no studies have been conducted to measure differences before and after creation of impound- ments in northerly latitudes. However, several references exist which document occurrences of such mortality. In Colorado, R. Lindsey (unpubl. data) documented that about 60 elk (Cervus elaphus) fell through ice while attempting to cross Blue Mesa Reservoir. Bonar (1985) indicates at least 10-20 moose fall through the ice each year at Revels toke Dam in southern British Columbia (BC), where temper- atures are considerably more moderate than those found in the Susitna area. In the latter case, Bonar (1985) had not analyzed any data, but he considered such losses insignificant to the populat ion. However, river crossings had been reduced to some unmeasured extent as a result of the hydroelectric project. In the Soviet Union, mortality caused by falling into impoundments during and after ice formation is vari- able by area and year but may reach 10-45% of the moose population (Danil~v 1986). F. Harper (pers. commun.) reports several instances of newborn moose becoming entangled in shoreline debris and being unable to escape from Williston Reservoir, BC. Under normal circumstances mortalities from these types of impacts are not be significant. However, because this is an additive source of mortality acting on an already stressed population, it should be viewed as a significant adverse impact . Population Parameters t be Impacted--Accidental mortality rates of adults and calves would be increased. Increased mortality rates within some subpopulations would be sufficient to cause population declines or reduce the rate of population growth. Impacts on Human Uses of Moose--This impact factor would be an additive source of mortality resulting in fewer moose for harvest or viewing. Fewer moose may also be available for dispersal into other areas which are partial y dependent upon Susitna moose populations to replace losses due to heavy hunting pressure. I .I .-5. Train and highway vehicle collisions due to new trans- portation access routes and traffic increases on existing routes would result in increased moose mortality. Rationale--Roads and rai l o ad corridors which are plowed free of snow during winter attract moose because travel is easier than in adjacent unplowed areas (Rausch 1959, Childs 1983). Plowing roads and rail lines results in steep banks and deep snow on either side of the tracks. MoosP. in snow-free areas seem reluctant to ent~r deep snow. Moose typically exhibit anti-predator behavior to oncc.ming trains: because they charge or hold their ground, they are killed y the train (Childs 1983). Access for the Susitna project would be achieved through a combination of railroad and road construction. A new road would be constructed from the Denal i Highway to the Watana construction camp. The existing Anchorage-Fair anks rail line would be connected by a spur line to the Devil Canyon Campsite. A road would also be constructed from the Devil 36 Canyon dam site to the Watana dam site. Because these new features would be built at elevations used by moose during winter, mortality from collisions may be relatively high during cor.struction and the initial years of operation. Moose typically migrate or move from high elevation areas in response to the first heavy snowfall each autumn. Depending on magnitud e and severity of the first storm, large numbers of moose could congregate on snow-free roads and rail lines. Mortality could be suffic i ently high to remove the annual surplus of mo0se and, in conjunction with other factors, could cause a population decline. Experience with railroad/moose collisions between Houston and Talkeetna support this scenario. During the severe winter of 1984-85, over 300 moose were killed (J . Didrikson, pers . commun.). Moose Population Par ameters to be Altered--Accid enta l mortality to calf and adult moose would increase. During most years the numbers of collision mortalities would be insignjficant. How c v t , ring years of deep snow, mortality could be s i gnificant. Large losses during relatively severe winters could alter the growth of the moose popu- lation in subsequent years. Once moose densities are lowered, other mortality factors s u ch as predation may prevent the popu l ation from increasing. Impacts on Human Uses of Moose--Because morta l ity from this i~pact is additive, its importance depends on i ts magnitude each year and on the population density of both predators and moose. Following severe winters with high losses, hunting harvests may be greatly reduced to allow the moose population to recover. Fewer moose may be available for viewing or disperal to other a r e as. I. I. -6. Snow drifts from impoundm e n t s and other major developments may impede moose movements and/or s u b j ect moose to higher risk of collision mo tality and may reduce the value of some areas as wi nter range . Rat i onale--Snow b l ow i ng off t h e i mp oundmen t s and other major facil- ities or developments is expect ed to create s u bstantial snow drifts, particularly along port i ons o f the shorel i n e . Areas which were prone to drifting prior to the project would li kely accumulate more snow with the project. Because moose avoid areas of deep snow , creation of new drifts would r e sult i n loss of hab i tat. If moose mov e ments are impeded and if moose avo id deep snow area s, some add i t i onal hab i tat may be unavailable. Snow drifts in th i s l atter case c ould also constitute a barrier to moose movements. Pred i ction of the exact loca tion and e xtent of snow drifting i s impossible because numerous factors influence its occu rrence. LGL (1985) predicts that i t would occur only in localized areas and par- t i cularly along the south and southwest a reas of the impoundments . In relation to the total project area, the term "localized" is appro- pri ate; however, these sma l l, local i zed impacts may become extremely i mportant to subpop lat i ons of moose if migration corridors are 37 blocked. Snow drifts may also occur along newly created transmission line corri dors, but prediction of the importance of this impact is even more difficult than predicting impacts of the impoundments. Areas covered by snow drifts retain snow longer than n ondr ift areas. Consequently, greenup of vegetation covered by drifts could be delayed in relation to other areas. Depending on the amount and type of habitat, loss o f early spring habitat could be important because moose are typically in relatively poor nutritional condit ion this time of year. Moose Population Parameters to be Altered--Mortali t y from starvation may increase due to disruption or impedance of movements and migra- tion, and to loss of habitat. Some moose may become more vulnerable to predation becaus e their escape may be delayed by snow drifts. Reproduction may be impacted because moose that do not die from star vation would be in poorer physical condition. Impacts on Human Uses of Moose--The total number of moose may be reduced. Although diff i cult to measure because the population could be stressed from a number of impacts, this particular impact is an additive source of winter mortality. As with other impacts, fewer moose would be available for human use and dispersal. I. I. -7. Drifted snow along railroad and road access corridors and roadway berms may impede movements of moose and/or subject them to higher risk of collision mortality. Rationale--In most respects this particular impact is s imil ar to and closely interwoven with I. I. -5 and 6 which have been discussed in preceding paragraphs. No further discussion is warranted. I.I.-8. Clearing of vegetation in the impoundment area would reduce carrying capacity prior to filling of the impoundment. Rat i onale--Clearing vegetation prior to fillin g the impoundment wou ld modify and destroy browse which traditionally has served as important moose winter range. Loss of winter range would occur as a result of reservoir filling; therefore, many impacts identified under I. I. -1 would occur here, with a few d ifferences in initial reaction . .Moose may continue to seek traditionally used habitat during winte r and spring. The area would be denuded of both escape and thermal cov~r, so moose may be more vulnerabl e to predation and exposure to severe weather . Social stress may occur because lack of spruce cover would allow moose a ~: relatively high densit ies to be in visual contact. Such contact can re~ml t in aggressive behavior among moose for avail- ab le forage (T. Sw eanor and F. Sandegren, unpubl. data). Population Parameters to be Impacted--Same as under I.I.-1. Impacts on Human Uses of Moose--Same as under I.I.-1. In addition, moose may initially be more vulnerable to hunting a nd poach ing. 38 I.I.-9. Increases in mortality of moose may occur due to increases in J.egal subsistence hunting and poaching. Rationale--Creation of impoundments and roads would create additional, easier access to the project area, so increases in hunting pressure may occur. Total harvests are expected to increase because moose would be more vulnerable due to stress and a combination of project impacts. Whether increased legal harvest is detrimental or even occurs depends on type of season and regulations in effect. For example, recent moose harvest regulations in the project area only allow harvest of moose with antler spreads of ~36 inches (91 em) or with 3 brow tines . The regulation provides protection to yearling and 2-year-old age classes while allowing unlimited hunter participation and assumes that not all large bulls would be harvested. If all larger bulls were harvested, most or all breeding would be done by young bulls, which could create social and possibly genetic pr blems. Under current hunting pressure not all large bulls aave been har- vested. Additiona! access could facilitate harvest of more older bulls whi c h would necessitate revised regulations to limit or redis- tribute harvest. Increased hunting pressure may increase crippling losses. Increased access would create a situation more conducive to illegal harvests. Whether increases in moose mortality due to poaching would be of sufficient magnitude to affect a moose population is not known. Because the moose population would be stressed from a number of other impacts, increases in hunting and poaching mortality wo uld be additive sources of mortality which cou t d contribute to a population decline. Moose Population Parameters to be Altered--Legal hunting mortality, crippling loss, and poaching may increase as a result of the project. Impacts on Human Uses of Moose--IP-it i ally, larger numbers of moose may be harvested in the project area. Unregulated access may create unpleasant hunting conditions because of hunter density. Ultimately, however, the number of moose available for harvest and other uses would decline. Hunter sur:cess wo-uld initially be high but would also decline. Poaching is Jikely to increase. I.I.-10. Both temporary and permanent loss of win t er h a bitat would occur as a result of borrow site development. Rationale--Creation and excavation of borrow pits would remove a l l vegetation and destroy summer and winter hab i tat. Access roads would create additional access for hunt i ng and poach ·ng. LGL (1985) pre- dicted that this loss C'E vegetat j on may only last from 2-20 years hecause all sites would be recovered with topsoil and shou l d become revegetated with useful moose forage species . Regardless, loss of these sites would contribute to a moose population decline through the same processes described under I.I.-1, with some differences . 39 Although ctual lost; of vegetation may be short term because of revege ta i on efforts (LGL 1985), there could be long-term impacts if the area s are revegetated by browse spec es l ess palatable to moose, o r the areas aro unavailable due to drifting snow. Also, once the moose population decHnes due to loss of hilb t tat and other factors described in the preceding and foll owing sect i ons, the moose popu- lat i on may then be limited by factors other than wint e r forage. Moose popu l a t ions are often r e gulated by factors other than forage (Gasaway et al. 198 3; Ballard and Larsen 1986). Numerous threshold levels exist which could keep the moose population below actual range ::arryi.ng capacity . Therefore, i.n a theoretical sense this i.mpact would b e short term, but i n reali.ty, once the population declines, it could be long term unless changes i n other f actors allow a populat i on increas e. Moose Population Parameters t o be Alt e red--S8llle parameters as those listed under 1 .1.-1, 2, 6, 7 , and 8 . Impacts on Human Uses of Moose--Initially moose would be more vulner- able to hunting and poaching due to im p roved access . Loss of habi~at with the resulting population decline would result in fewer moose available for hunting and viewing. The latter could be a short-term impact if the moose population is able to recover and take advantage of t he revegetated areas. In the latte c a se, improved access could a llow for i.ncreased har vests. I . I. -11. Permanent l oss and alteration of moose hab i tat would occu r as a result of acces s corridor construction, ma i ntenance, and use. Rationale--Construction, maintenance, and use of roads and r ail facilities would require add"tional grav el p i ts and berm construction beyond thos e needed for actual construction of the dams . Use of the areas, and maintenance , would create disturban ces that cause moose to avoid some areas. The problems encoun t e r ed with this impact are integral pa r ts of those discussed for other i mpacts. Moo s e Population Pa rame t ers to be Alter ed--S i milar to those describe d under 1 .1.-1 , 2, 5, 7, 8, 9, and 10. Impacts on Human Uses o f Moose-Ult i mately, the total numbers of moose available for human use would be reduced due to bot d irect and i ndirect l oss of habitat and increased mortality. Init i ally greater numbers of moose may be legally an d i llega lly harvested due to increased access. 1.1.-12. Due to improved access created by the project, the entire ~as in may be subject to increased commer c ial development which would result in loss of moose habitat and increases in moose mortality . Rationale--The project area lies within an area surrounded by the Parks, Glenn, Denali, and Richardson Highways. Because of remoteness, the area would probably not be commercially developed for decades. With the advent of the proposed project, Native corporations selected 40 land needed by the project and adjacent areas to take advltntage of new access routes. Creation of access and resu 1 ting secondary private developments are considered negative impacts on wildlife. In some cases secondary developments could have a greater impact on moose than the actua l project tself. Depending on the nature and location of developments (e.g., mining activities, lodge facilities), significant losses of habitat and increases in direct moose mortal ty due to auto collisions, poaching, and hunting could occur. Population Parameters to be Impacted--Because additional developments often result in direct loss of habitat and/or direct mortality, the effects on various moose population parameters would be identical to those described under many of the impacts previously described except that the degree of impact would vary. Impacts on Human Uses of Moose--Impacts would be similar to those described under previous impacts. Ultimately, fewer moose would occur. I.I.-13. Habitat quantity and quality for moose would improve along the tra.nsmission corridor because vegetation would be maii)tained in early successional stages. Rationale--Clearing transmission corridors and maintaining early successional states of spruce and mixed spruce-deciduous vegetation are expected to result in an improved browse biomass. This is expected to increase the carrying capacity for moose wintering along the transmL~sion corridor. Winter mortality may be reduced for some subpopulations and increases in productivity may occur. Human access into previously inaccessible areas would be greatly improved. Moose Population Parameters to be Altered--Due to improved nutrition, some increase in productivity might occur. Mortality due to winter starvation may be reduced. Mortality during severe winters would not be reduced because much of the improved habitat would be inaccessible during a severe winter. Impacts on Human Uses of Moose--Increased numbers of moose should be available for harvest and view~ng. Transmission lines would also provide additional access for all-terrain vehicles, facilitating both additional legal harvests and poaching. Potentially Important Impacts (P.!.). P.I.-1. Local climatic changes resulting from the impoundments would include increased summer rainfall, increased winds, cooler summer temperatures, increased early winter snowfall, hoarfrost depo- sition on vegetation in winter, delayed spring plant phenology, and changes in plant growth and species composition. These chenges would reduce habitat carrying capacity for moose and increase vulnerability to a number of forms of mortality. 41 Rationale--It is well documented that creation of large artificial bodies of water alters the climate o f the surrounding area. This "warm-bowl" and "cold-bowl" effect can significantly alter climate to such an extent that large differences in preci pitation and temperature can occur. LGL (1985), suggested the effects would be ''localized" and would not extend beyond 1-5 miles from the shoreline . If measurable changes in climate occur within this zone the impacts of the potent ial changes could be significant. LGL (1985) suggests that because the effects of climatic change would be "localized" the effects would not be measurable. In earlier studies for Rampart Dam and Reservoir, Henry (1965 ) modeled available climatic data and pre d i cted that a 10% change in precipitation would occur up to several hundred kilometers away from the impoundment. A number of other climatic changes were also predicted. Although the Susitna Project would be considerably smaller than the Rampart pro- posal, it appears reasonable to assume, bas Ad on studies such as Henry's and others (Taber and Raedeke 1976 -Ross Lake in Washington), that measurable changes in some climatic parameters would occur. To determine the magnitude of change, systematic p re-and post-impound- ment studies would be necessary to discount this potential impact. Climatic changes which could potentially be most important to moose include cooler summer temperatures, increased snowfall, increased hoarfrost deposition on vegetation, delayed spring melt, delayed spring plant phenology, and possible changes in plant growth and species composition. Detailed discussion of these potential effects follow: a. Cooler summer temperatures --This change could make conditions less favorable for s urvival of newbo r n moose calves due to exposure to cooler temperatures in conjunction with delayed snow melt and delayed plant phenology. b. Increased snowfall--Increases in snow depths adjacent to the impoundments due to i ncreased evaporation c t h :i make important wintering areas less desi r able as winter range. The area adjacent to the impoundments receives highe r. u s e than areas where browse may be more abundant but less available due to greater snow depths. Increasing snow depths within a 1-5 mile zone from the reservoir could significantly decrease the value of the remaining important winter range. For example, a 10% increase in snow depth over a 1-to 5-mile-wide zone in critical moose winter range could reduce the capacity of the area to support moose. c. Hoarfrost deposition on vegetation--Hoarfrost and rime ice naturally occur on vegetation along the Susitna River during some time periods. Where open water would occur year-round due to the impoundments (downstream of Watana and Devil Canyon dam sites), the frequen cy of frost and rime ice deposition on moose browse would increase. Although difficult to measure, the addition of substantial amounts of frost and rime ice on vegetation requires additional energy for moose to melt the ice. If frosting or 42 icina repeatedly occurs over the winter, this energy exrenditure could increase stre ss on the moos e population, given that their physiological condition is downwaxd e ven during moderate winters. In northern British Columbia, Harper (1 985) sugested that the occurrence of ice fog from the creation of the Bennett dam and reservoir on the Peace River may have been an additional factor causing reduced moose populations on the nort h side of the river. The Peace River Valley is now "foged·in" mos t of the winter due to warmer water coming from the dam, effective l y eliminating the insulation benefits of south-facing winter ranges (Op. cit.). d. Delayed spri ng melt--Cgoler temperatures in conjunction with increased snow depths could delay onset of spring thaw and increase length of time nee ss a ry for snow melt. This would also delay availability of some food plants. Moose would avoid areas which retain snow, resulting in a change in moose dist ribution and habitat selec tion and increasing pressure on adjacent habitats and populat i ons. e. Delayed spring plant phenology--Plant phenology is i nfluenced by a wide variety of factors (LGL 1985). With lowsr air temper- atures and increased snow depths • plant development would be slower than in areas with high t emperatures and less snow. Moose are usually in their poorest physiological state just befor e onset of greenup. Delay of greenup could significantly a ffect moose survival. LGL (1985) speculated that greenup would be delayed by a maximum of 3-5 days. The length of this time period would be dependent on the ac c umulation of snow and spring temperatures. f. Precipitation and temperature are among several facto.cs which influence composition, distr i bution, and growth of vegetation. Growth of existing vegetation may be altered due to cooler temperatures, increased snow depths, delayed spring melt, etc., all of which lead to a shorter growing season. This may alter the growth rates of wouldows and reduce the range carrying capacity. Changes in plant species composition would likely be very subtle and take several decades to be detected. Moose Population Parameters which could be Altered--Due to a loss of critical late-winter/early-spring habitat and delayed greenup of vegetation, survival of calves would be reduced. Poorer physiological condition of cows results in production of less viable calves . Increased mortality may result from exposure to a less suitable c limate. Moose may be more vulnerable to predation because of the poorer phys i cal condition and disp l acement from desirable hab i tat. Winter mortality from starvation may i ncrease due to loss of habitat and increases in energy expenditur es necessary for finding sufficient forage. Impacts on Human Uses of Moose--Becaus e this impact ultimately reduces habitat carrying capac i ty and increases mortality , fewer moose wo u ld be available for harvest, viewing, and dispersal. 43 P.I.-2 . War.er Wdter in downstream areas would result in open water and •ay alter plant phenoloay and affect sprina foraae and c pver for 1100se. Rationale--LGL (19.85) speculated tha t warm water condi ons would retard river ice development in late winter and melt ex i sting river ice faster. However, existing hydroelectric developm~nts provide scenarios for projecting iNpacts on mo >se. For example, on the Peace River below Bennett Dam in northern BC during 19 79-80, flow ice piled up in downstream areas, creating ice d am s. These dams then caus ed flooding and inundation of upstream riparian areas (Harper 1985). The inundated habitat was not suited for moose during the remainder of the winter. We suspect that these areas freeze and thaw more slowly, thus eliminating winter habitat ar.d retarding spring plant growth. Moose t hat become trapped on the inundated areas su f fer increased mortality d ue to exposure bec ause they do not move from the i s lands (Harper 1985). Moose Popu l ation Parameters Which Could Be Altered--Overall, carrying capacity for moose would be reduced and rates of mortality would increase (see di~cussion for I.I.-3 and 4). Impacts on Human Uses of Moose--Because the total number of moose would be reduced, fewer moose would be available for human use and dispersal. P .I.-3. Habitat quality may temporarily decrease near the reservoir as a result of locally high densities of moose dispersing from inundated areas. Rationa l e--Moose which become displaced d ue to inundation would con- centrate on adjacent habitat and utilize vegetation which currently supports other moose. The amount of forage present in and immediately adjacent to the impoundments is less than that outside the impound- ments. However, it receives much greater utilization (Becker and Steigers, unpubl. da t a), apparently because it i s more available due to shallow snow depths. Because this vegetation is heavily used, additional usage by displaced moose would probably exceed annual growth and r e duce carrying capacity. Moose Population Parameters which could be Altered--Starvation mor- tality would increase due to increased competition and reductions in carrying capacity. Remaining moose would experience decreased pro- ductivity along wi th increased mortality of calves. Impacts on Human Uses of Moose--Increases in natural mortality and declines in produ ction would result in fewer moose for human uses and dispersal. P . I. -4. Continued loss of moose habitat due to eros ion of impoundment shores . Rationale--Ero sion of shorelines would destroy an unknown quantity of moose habitat. Some areas may become revegetated with spec ies more useful as m<'ose forage. LGL (1985) considered this impact to be a 44 slight adverse iiBpact which could be offset by colonization of new veaetatio~, assumina the steepness of newly colonized areas would not preclude .aoae uae. This, with other impacts, is an additive impact which ~uld be relatively insignificant but, because it would occur in conjunction with other impacts, may result in additional loss of habitat and accidental deaths. Population parameters and human uses to be impacted are similar to those already discussed under P.I.-1, 2 and 3. P .I.-S. Drift i ng snow in the transmission line corridor may preclude use of winter browse. Rationale--Areas vegetated by short plant species appear more prone to snow drifting. This effect may negate some of the positive benefits derived from increases in browse production as a result of clearing corridors. New browse may be unavailable due to snow drifting. Hoose Population Parameters Which Could Be Altered--Increases in moos e productivity due to increased browse supplies described ~der I.I.-12 may not occur to the degree anticipated. Portions of the increased browse may not be available because of snow drifting. Consequently, starvation mortality during mild winters may not be reduced to the level anticipated under I.I.-12. Impacts on Human Uses of Moose--There may not be an increase in the numbers of moose available for harvest as a result of improvements in browse quantity predicted under I.I.-12. P.I.-6. Accidental fires resulting from human activities may rejuvenate decadent moose habitat. Rationale--Increases in human activities during construction and operation may result in accidental fires. Because many portions of GHU 13 have historically been subjected to wildfire, much of the moose habitat is fire-depeudent. If accidental fires occurred, moose habitat quality and quantity would improve resul t ing in increases in range carrying capacity. Whether the moose population could respond to tle improved habitat may dictate whether it becomes used. Improve- ments in habitat could be expected to last about 25 years before additional habitat improvement would be needed. Assuming vegetation and moose respcnd as they did to wildfires in Interior Alaska, no short-term detrimental impacts are anticipated (Gasaway and Dubois 1985). However, with increased private and commercial developments fire suppression programs usually intensify and the potential for habitat improvement from wildfire and controlled burning would probably never materialize. Hoose Population Parameters Which Could Be Altered--Depending on the size of the area imrolved, improvements in quality and quantity of forage could benefit moose. Cow moose could be in better physio- logical condition resulting in proro~ction of vigorous, healthy calves. Hoose of all age classes could be :in better physical condition and less prone to predation. Numbers of starvation mortalities could decline . 45 !~~pac t a on HWian Use s of Moose--If not l i mited by other factors, numbers of moose available for harvest and v i ewing could increase. If annual surpluses are not removed by hunting and predat i on, surplus animals may dis rse to less populated areas serv i ng to r es tock areas dep leted by hunting o r o.ther facto r s. P .I.-7. Increases in ground-based activity (road traffic, village activities, dam construction) may preclude use of some areas by moose, particularly sensitive areas such as c alving sites and winter habitat. Rationale --Increased human presence, particularly at v i llages and at areas where major habitat alterations are occurring, would result in disturbance to moose. Disturbance can manifest itse f in many forms; e.g., ungulate heart rates and other body functions increase when confronted with unnatural stimuli. Addit i onal stress does not neces- sarily result in an outward change in behavior o r in direct harm to t h e tmimal, but is an additive stress factor t o be considered in energy dynamics of moose . The most outward result of disturbance would be avoidance of areas where noise and visual stimuli cause harassment. Hoose are expected to avoid habitat areas near the damssites during active construc tion and other areas between dam sites, •rillages, and gravel borrow pits. Continued high-intensity use of villages, rail facil ities , airports , and dam sites may result in permanent avoidance. Hoo~e Population Para~eters Which Could Be Altered--Avoi dance of specific sites which h is torically served as wi nt e r habitat is equated with at least a temporary loss of habitat. This loss wo uld affect several moose population parameters, particularly those mentioned un der I. I. -1. Unimportant Impacts (U. I.). U. I. -1. Alteration of moose distribution may occur due to corridor traffic and disturbance . Rationale--Initially, activities associated with construct i on and operation of transportation corridors would cause moose to avoid these areas. This may result in short-term habitat loss if the avoidance occurs during winter. However, moose should become acclimatized to this disturbance, so no long-term impacts are anticipated. The greatest amount of disturbance may occur during hunting season through use of access corridors . Di sruption of movements in autumn could alter rutting behavior and force moose into less desirable areas. Potent i ally, this could affect reproduction and result in a short-term loss of productivity. In the short term, moose may suffer increased rates of starvation mortal i ty until they become accustomed to traffic and noise . Rutting behavior may be temporarily disturbed. U.I.-2 . Prior to filling, clearcut areas in the impoundment may inhibit movements due to slash piles and human disturbance . 46 Rationala--Bacauaa .aose may x 11ct negatively to creat i on of open arau without cover, temporar y retention of slash piles may 11it i gate part of the avoidance impact. Howe ver, continued human ~resence aay, in the short ter., cause temporary avoidance of the area until logging crews and oth er project personnel leave the area. Although not iaportant in itself, this impact is another additive source of nega- tive st~li for .aose. No long-term i mpacts on moose, or their uses, are anticipated from this particular impact . U.I .-3. Impeded drainage caused by road and railroad berms may alter moose habitat as a result of flooding of forest and shrub areas. Rationale--Water draina ge would be altered by construction of berms . In many cases this alteration would be minimized by proper installation of culverts and bridges. However, some alterations (such as temporary inundation of small, loca lized areas) which would kill vegetation, would occur. LGL (1985) ma intains that t h ere would be equal proba- bility of creating higher quality habitat as a result of berm con- struction. Although it is probably correct to assume plant species desired by moose would colonize the berm areas 1 this attractant would make moose more susceptibi • to death from vehicle coll i sions. Impacts on moose forage that are caused by from berm construction would be localized and would probab !y not result in measurable impact on the .aose population. Howev r, like many other impacts associated with this project, it may not be individually important but in s11111111ation with other impacts may be significant . u. I. -4 . Increase in aircraft overfli ~ t may stress animals or preclude use of some areas . Rationale--Experience with moose populations occurring in close proximity to airports suggests that this impact should not have permanent, long-t erm effects . However, there may be differences between air traffic at airports and that wh i ch might occur with the project . Although moose become accustomed to aircraft overflights at airports, these areas are usually fenced, so little additional human disturbance occurs. The proposed Watana airport would be adjacent to village sites, transportation corridors, gravel extraction, etc., possibly resulting in some avoidance due to other disturbtmces in add ition to a i rcraft. Predicti on of Project Impacts on Moose Subpopulations Based on studies of m v em ents of radio-col l ared moose from 1976 through 1986 (data presented earlier), a t least 12 subpopulation s of moose were identified which either utilize the proposed irn oundments or could be impacted by the project. For purposes 'ilf t s report a subpopulation is defined as a group of moose wh ich util ize similar winter and summer range and which mo v e to and from such areas in general synchrony. Generally, members of subpopulations breed nnd calve in the same area, but subpopulations are not discrete and many gradations exist. Certain subpopulations of moose would be impac-.ted 47 110re than others and discussion concernina specific subpopulations follows. For subpopulations with similar exposure to the project, di~cus s on of project impacts were pooled . Size of ~se subpopulations was determined by examinina locatio ns of radio-collared moose from each subpopulation during the 1983 census. The entire impact zone had been divided into discrete 12-20 mi 1 sample units. Each unit was stratified into one of 4 density classificatio.ns based upon sian and numbers of moose observed (Gasaway et al. 1981; Ballard et al. 1982, 1983; see Population Density section). Followina this process, randomly selected quadrats were intensively surveyed and the population densities of moose were estimated within each density classification. By adding the numbers of quadrats where radio- collared members of each subpopulation were located and then usina average density estimates we were able to estimate the relative size of each subpopulation based on autumn distributions. All estimates were corrected to exclude radio-collared moose which d i d not reside in the primary impact zone. Descriptions of characteristics, size, and predicted impacts of the proposed Susitna Hydroelectric project on 12 identifiable subpopu- lations of moose: 1. DEVIL CANYON TO FOG AND DEADMAN CREEKS MOOSE SUBPOPULATION Characteristics--This subpopulation is composed of resident individuals which generally have overlapping summer and winter range. Moose from this group move to a rutting area along Clark Creek each autumn. Elevational movements occur apparently in response to climatic factors, particularly snow depths. A sipificant relationship exists between winter severity and use of various elevations by moose, with lower elevations being utilized more frequently during years of deep snows. Moose utilization of the Devil Canyon impoundment is primarily restricted to the area east of Devil Creek. Several moose apparently calve in or immediately adjacent to the impoundment each year. Only a few moose use the luwer Devil Canyon impound- ment area, apparently due to the steepness of the canyon walls. Moose often cross the Susitna River during January through April to utilize south-facing slopes locat ed between Deadman Creek and opposite Stephan Lake. This subpopulation occurs wi thin the territories of at least 2 wolf p a cks (Portage Creek and Stephan Lake Packs) which prey heavily on moose (Ballard et al. 1982, 1983). Black bears are quite numerous in this area (Miller 1985) and consequently this particular subpopulation of moose probably receives the greatest amount of predation by black bears of any of those studied. However, brown bears are the most important predator. The area is lightly hunted by humans because of poor access. Conse- quently, the area has a relatively higher proportion of large- antlered bulls than many other moose subpopulations. Based upon censuses conducted in 1980 and 1983 and on interpretation of 48 radio-collared 1100se IIOV•41llt da a, this subpopulation is esti- mated to ca.prise 420 ind iv.d ala (181 of the primary imp~ct zone population). At least 701 of this subpopulation resides east of Devil Creek, with most occupying tho area between Deadman Creek and the area opposite Stephan Lake. Iapacts--Because a large number of developme.nts such as the Watana Dam, village facilities, railroad and access road cor- ridors, several borrow sites, etc., would occur within the range of this subpopulation, it would be one of tha most severely t.pacted. Loss of habitat would increase mortality due to winter-related starvation. The amount of habitat lost would be greateT than reported because moose would likely avoid additional areas due to disturbance, harassment, increases in snow depths brought about by changes in microclimats, drifting snow, etc. Also, year-round open water below the Watana Dam site would bisect the annual ran.ges of many individuals, making portions of the range unavailable in winter. Although moose are known to cross open water at air temperatures of about 0° F, they apparently have an aversion to crossing at colder temperatures. If open water during late autumn and winter results in increased snow depths within several hundred meters of the impoundment, additional habitat would be lost. During construction and early operation of the project, the physiological condition of wintering moose would decline, resulting in an increase in winter mortality. Moose that do not die from winter-related causes would be in poorer physical con- dition, resulting in production of fewer calves through reduc- tions in pregnancy and twinning rates. Calves would be less healthy and would suffer higher rat:es of natural mortality. Development of the Tsusena Creek borrow site, road development from t he Denali Highway and Devil Canyon, and establishment of camp facilities are likely to disrupt use of the Clark Creek rutting area. Increased access would result in increased poaching and hunting activity. As a result, the relatively high propor- tion of large-antlered bulls in this subpopulation would decline . Although black bear predation does not limit the moose popu- lation, inundation of black bear den sites and habitat would concentrate bears in the same hab i tats in which moose are forced to concentrate. Miller (1985) suggested that black bear popu- lations would eventually decline in the area. However, until those declines occur, predation by black bears would become a significant source of moose mortality. Brown bears and wolves would also take advantage of the increased prey concentrations. In the absence of increased predation, lowered productivity and increased mortality resulting from habitat loss and avo idance would cause the population to decline. 49 O.vel~nt of borrow s i tes would result in loss and avoidance of .ooae hab i tat. Althouah this habitat may eventually be replaced throuab natural recolonizat i on or reveaetation followina retire- •ent of the site, it is unlikely that the moose population would be able to respond to the increased Md ~roved forage . Once productivity declines and mortality increases, t his moose sub- population •ay never be able to increase because factors other than veaetat i on would prevent populat i ?n growth . Only throuah drastic changes in predator-prey rati s, changes in waterflow rea~es to allow freezing of open water below the dam sites, and large reductions in the levels of human disturbance can t his subpopulat i on be expected to recover. The area may serve as a "sink" by attracting moose from adjacent areas of hiah density, but these incomina moose would be subjected to the same factors that caused the original population decline. The subpopulation is expected to eventually stabilize at a very low level in comparison with pre-project conditions. We predict that this subpopulation of 420 moose would decline by at least two-thirds as a result of the project . 2. UPPER FOG AND TSISI CREEKS HOOSE SUBPOPULATION Characteristics--This subpopulation is composed primarily of a migratory group of individua l s which occupies Tsisi and upper Fog Creeks during late sUIIIDer and aut umn. Depending on t iming and extent of snowfall, these moo s e move to lower elev ations within or adjacent to the Wata n a i mpoundment zone where they may remain through all or pate of winter. In many cases they calve on wintering areas before r turning to sUIIIDer range. A segment of this subpopulation resides year•round in the Watana Lake-Kosina Cr eek area where they share winter range with a migratory segment. This subpopulation lies primarily within the range of the Watana wolf pack. Other wolf packs sometimes exist to the south of the Watana pack but are usually elim i nated by a i rcraft-assisted hunting . Althou$h black bears occur along the Sus i tna Ri ver, they are not cur'r,ently a significant source of moose mortality. Brown baars occur ~h roughout the area and are the most important mortality factor . Hunting pressure is generally light due to limited access; however, heavy hunting pressure sometimes occurs at Watana and Fog Lakes due to floatplane access. Based upon moose censuses and interpretation of radio-collared moose movements, this subpopul a tion was estimated at 350 indi- viduals. Host, if not all, of these moose winter in or adjac ent to the proposed impoundment. Impacts--Loss of winter habitat from direct inundat i on plus losses from drifting snow and climatic changes are likely to be the most important impacts initially affecting this subpopu- lation. As mentioned earlier, these impact mechanisms would 50 likely aanifest th-elves thro~gh increased winter-and early- sprin& .artslity and throuah decreased natality brouaht about by nutritional stress. The exact •aanitude would be dependent on the quantity of foraae lost through drifting snow and changes in microcliaate. Hoose displaced fr011 the impoundment and the snow-drift zone would be subjected to increased crowding c011petitio.n fr011 adjacent 1100se . They would also be ~ubject to increased levels of pre- dation from displaced predators. Both types of impacts and ot ~rs not specificly discussed here would be sufficient to cause the subpopulation to decline and to eventually stabilize at a lower level. Fluctuating water levels and resultant ice shelving may pose a problem for this particular subpopulation bec ,.use many membors cross the Susitna River where they share winter range with other subpopulat i ons. This impact mechanism would be an additional source of mortality to a group already suffering declines from other project-induced causes. Due to improved boat access from the im}'oundment and improved access created by road construction to the dam site, both legal and illegal harvest of moose would increase. In addition, private commercial developme t s are likely to occur with resulting impacts such as loss of habitat and disturbances. Based on ou= evaluation of impact mechanisms, this su b population of 350 individuals would decline by 50~. Short-term los ses may be even gre.ater dur ing severe winter condit "ons. Population response to a severe winter would be different from that prior to the project due to lower rates of reproduction and poorer overall health of the subpopulation. 3. KOSINA CREEK MOOSE SUBPOPULATION Characteristics--This subpopulation consists of nonmigratory moose which occupy the lower elevational drainages and mainstem of Kosina Creek. Hoose from this subpopulation demonstrate altitudinal movements similar to those of other subpopulations in the study area; high elevational areas are occupied dur i ng summer and autumn and low elevational areas are used during winter and early spring. Typically, most moose in this group move short distances up and down creek bottoms. The overall winter habitat carrying capacity of this area is relatively low ln relation to that of many other areas, due to heavy snow accumulations. This subpopulation has probably remained relatively stable over the past decade. Dispersal of 1 radio~collared yearling suggests that the population may con- tribute emigrants to other areas. Hunting pressure in the are~~o is light due to the relatively low moose population and poor access. We estimate this subpopulation at from 100-200 indi- viduals. 51 !Bpacts--No direct impacts as a result of the project are antici- pated. However, several indirect impacts could occur .. Hoose dispersing fr011 the area and attempting to cross the Susitna River mi ght su f fer h i gher rates of mortality by falling through thin ice or they could be blocked by ice shelvi ~g. If climatic changes were to occur at greater d i stances away from the impound- ment than are currently predicted, additional habitat may be unavai lable to moose. Perhaps the greatest threat to this subpopulation is increased commercial development, such as mining and lodge development, which could result from the boat and road access provided by the project. Loss of habitat, increased d.isturbance, increased poaching and hunting activity, etc., could be of sufficient magnitude to cause this marginal subpopulation to decl ine . Because all of the pos.sible impacts are speculative and beyond prediction, no attempt has been made to quantify them .. 4. WATANA CREEK -MONAHAN FLATS MOOSE SUB POPULATION Characteristics--This s ••bpopulation consists of a group of individuals which occasionally migrate to the impoundment zones during some winters. During years the impoundment zone ~s used, moose migrate to Monahan flats (60 km s to the north) i n late spring where they c alve and rem~in throu gh summer. B ~tween late summer and early spring these moose may migrate to ~he im pound - ment zone wh ere they overwinter. During other yee.rs they over- winter between Mon!lhan flats and the divi de betwe('-a Brushk.ane. and Deadman Creeks. Why they only periodi cally uti:ize the impound- ment zone is not k.n own . The total range occupied by this subpopulat ion falls within the range of 3 wolf packs (Watana , Jay, and Seattle Creek Packs). Brown bears occur throughout the area anc have been documented as the most important source of moose tr.ortality (Ballard et al. 1981). Black bears occur infrequent!. in the Monahan flats area but are relatively numerous along che mainstem of the Susitna River; they are often denning at the time that they could poten- tially come in contact with th .s moose subpopulat i on. Recre- ational and subsistence hunt i.Il& pressure is heavy along the Denali Highway . Impacts--Because moose from tt his subpopulation a~pear to utilize the impoundment zone as a wtntering area, loss of critical range would result in increased starvation of both calves and adults during winter. Reduced physical condition of surviving cows would result in reduced natality due to lower twinning and pregnancy rates. This particular subpopulat i on would also be directly impacted by the Denal i -to-Watana Camp road system through dir ect less ~f hab i tat and collision mortality, and indirectly through c hanging snow patte~ns brou ght about by drifting. Additional a ccess created by this system would subject this subpopulation to incre~sed ~evels of legal Bh~ illegal hunting harvest. 52 Becauae this subpopulation was not present durina autUIIO cen- suses, no count data exist for estimatin& relative size. Based on nuabers of animals associated with 1 radio-collared animal and on other miscellaneous observations, the aroup was estimated to contain no more than 50 individuals. Loss of winter ranae may result in an averaae reduction of 50\. 5. DEADkAN-WATANA CREEK HOOSE SUBPOPULATION Characteristics--This aroup of moose comprises migratory and nonmigratory individuals. The nonmigratory subpopulation is a continuation of the group at Deadman Creek which exists throuah- out the project area. The migratory group (which winters along Wataua Creek but miarates to Butte Creek during summer and autUIIO) was clumped with the nonm i gratory moose for discussion purposes. Hoose from this group utilize the impoundment zone adjacent to Watana Creek primarily during winter and in early spring for calvin&. Elevations above the proposed high pool level are used in late summer. Win ter range is shared with the migratory Wat!Uia-Coal Creek subp pullltion. Upper subalpine and tundra vegetation are used in aut~ during the rut. The group occurs within th.a territory of the Watana wolf pack which preys almost entirely on moose (Ballard et al. 1982, 1983). A second pack may occas i onally prey on these moose when in the Butte Creek area. Brown bears are the most important mortality factor, accounting for the deaths of about 46~ o f the moose calves produced (Ballard et al. 1985). Black bears also occur within the range of this subpopulation but only account for 8~ of the calf mortality . The area currently receives light hunting pressure because of poor ac~ess. Based on interpretations of radio-collared moose movements and aucumn censuses in 1980 and 1983, this group of 2 subpopulations was estimated a t 290 individuals with migratory moose numbering about 150. This subpopulation comprises 12~ of the moose popu- lation which comes in contact with the proposed i:mpoundments. Impacts--The greatest impact on this subpopula·tion is direct loss of winter habitat due to inundat i on and clima~ic chang~s resulting in deeper snow accumulations and drifting snow. Loss of winter habitat would result in relatively large increases in winter mortality. Hoose which do not die from starvation would be in poor physiological condition, resulting in lower calf pro- duction due to lower rates of pregnancy and twinning. Calves which are produced would be less likely to survive because of their lower physical conditicn. Hoose displaced from the impoundment would concentrate next to the impoundments . Predators, which would also be conc·entrated, would cause increased mortality and contribute to a population decline. 53 High densities of predators could maintain moose numbers at lower levels (Ballard and Larsen 1986) t han would have occurred other- wise. Hoose attempting to cross the Susitna River during winter and late spr::.ng may suffer higher rates of mortality. Blockage of movements because of fluctuating ice levels and accidental mortality associated with crossing this ice, would contribute to a subpopulation decline. The southwest shoreline and other areas are predicted to exhibit increased snow drifting (LGL 1985). Snow drifting could have the same effect as direct losses of habitat because the existing habitat would be less available. Because this moose subpopulation depends heavily on the riparian habitat of \iatana Creek, this moose subpopulation may decline an average 50-75%, from 290 to about 75 individuals. 6. WATANA-COAL CREEK HOOSE SUBPOPULATION Characteristics--This group of migratory moose uses the proposed \iatnna impoundment from middle \iatana Creek to the mouth of Jay Creek as winter range. This winter range is shared with at least 4 other subpopulations whi·ch include the migratory and nonmig- ratory \iatana Creek subpopulations, the upper Fog-Tsisi Creek moose, and nonmigratory moose from Jay-Kosina Creek east. This particular group probably utilizes the impoundment zone more than any other subpopulation studied except for nonmigratory moose from Kosina-Jay Creek east to Clearwater Creek. Use of the area appears governed by winter severity as reflected by snow depth. During years of below-average snow, these moose may not use the impoundment zone but confine their wintering activities to the knobs along the north side of the river immediately adjacent to the impoundments. During some years they may stay on summer range at and near Coal Creek .. During average snowfall years, they utilize the impoundments from 1-3 months depending on snow depths and other factors . Habitat use during a severe winter has u<lt been documented, but heavy usage of the impoundment zone is ?redicted. Hoose typically leave the impoundment zone in April or May. Movement to the calving grounds on Coal Creek usually occurs within 2 weeks . There are few calving concentration areas in GHU 13; however, Coal Creek is one of the most important. Cal-ling occurs in late May through early June. ~ese moose remain on the calving l:lr~a through summer. During autumn they occupy the upland knobs along Jay to \iatana Creek. Several dispersals by subadnlts have been documented for this subpopu- lation which may help restock area s heavily hunted or depleted by other factors . The subpopulation occurs within the range of at least 2 and sometimes 3 wol~ pack territories. These include tile \iatana Creek, Jay Creek, and B-S Lakes wolf packs (Ballard et al . 1982, 1983). Winter range of the moose subpopulatio n is predomi nantly 54 within the Watana pack territory whi l e summer range lies within the B-S Lake and Jay Creek packs. All of these packs depend primarily on 1100se, and wolf predation is the largest cause of winter mortality. Predation by brown bears on calves is high in this area, and roughly half the calves are killed by th .. annually (Ballard et al. 1981). No black bears have been observed on the calving grounds, although they occur in .timbered areas along Watana Creek. They are not a significant source of mortality at the present time. Unlike the Watana Creek sub- population, this subpopulation is heavily hunted. Access occurs from the Denali Highw~y from several all-terrain-vehicle trails. Based on autumn 1!183 census data, the area has the highest density of any of the subpopulations studied: 2 .7 moose per mi 1 (1.04/km 1 ) in autumn 1983. The subpopulation is estimated at 610 individuals or about 25'% of the moose occurring within the primary impact zone . Impacts--Tue greatest impact on this subpopulation would be loss of important winter habitat, hence large increases in winter- related mortality. Calving and twinning rates are expected to decline, and the total population would be reduced. Competition for forage, in addition to increases in predation and other direct mortality, may increase mortality rates until mortality exceeds natality. Many other impacts described for the resident and migratory Wa~ana subpopulation are expected to occur on this group as well. Because much of the winter habitat for this subpopulation is immediately adjacent to the irepoundments, changes in browse availability due to changes in climate, snow drifting, etc. could have serious effects on this moose sub- population. This subpopulation is e xpected to decline by an average of 50'% from 610 to about 300 moose. 7. JAY-KOSI NA CREEK TO CLEARWATER CREEK HOOSE SUBPOPULATION Character istics--This subpopulation consists primarily of non- migratory moose with relatively small home ranges. Considerable overlap in both seasonal and total home ranges among individuals occurs. Many of these moose have home ranges which are bisected by the Susitna River. Although nonmigratory, they move seasonally from higher elevations in autumn to low elevations during winter. A large number of moose remain on or close to winter range where they calve. Probably more individu.als f rom this subpopulation calve in the impoundment zone than any other groups studied. Approximately half of the locations within the impoundment zone for this group occur during Hay through August. Several subadults have dispersed from this area, thus it may also be important for recruitment to other areas . The area located between the Susitna River Gauging Station and the mouth of Clearwater Creek serves as a wintering area for several subpopulations of migratory moose. Resident moose from this subpopulation share winter range with moose from the upper Clearwater-Maclaren subpopulation, the Butte Creek-Susit.na 55 River subpopulation, and Moose fr011 the latter Susitna River above the do not winter in the on knobs immediately aubpopulation, the upper Oshetna-Black the Lake Louise-Susit.na subpopulation. subpopulation concentrate no;.:th of the big bend durina autumn. Moos~ which t.poundaent zone appear to overwint~r adjacent to the proposed impoundments. This subpopulation area may be occupied by up to 4 different wolf packs, all of which prey heavily on moose (Ballard et al. 1982, 1983). At least 3 packs have territorial boundaries which meet at the upper end of the impoundment where several moose sub- populations winter. Wolf predation is an important source of adult and calf mortality. However, predation by brown bears is the largest source of calf mortality (Ballard et al. 1981, 1985). Black bears are present in timbered areas along stream bottoms, but they are not numerous and do not constitute an important source of moose mortality. The area is heavily hunted, with access provided by numerous all-terrain-vehicle trails in addition to float plane access at several small lakes. Based on autumn census data, the subpopulation was estimated at 700 individuals: 1.9 moose per mile2 (0 .7/km2 ). Impacts--Loss of winter habitat and calving areas would be the most significant impact affecting this subpopulation. These and other impacts described earlier would affect this subpopulation but to a much lesser degree than other subrwpulations. The magnitude of the impacts would be less because the impoundmen·t becomes substantially smaller as it reaches the big bend where several subpopulations concentrate. However, the degree of climatic change and the amount of snow drift could be as important as inundation in the amount of habitat made unavail- able. This subpopulation may decline by an average of 2St (N=l75) as a result of the project. If the important winter habitat immediately adjacent to the impoundment is also impacted, the subpopulation could decline by more than SO%. 8. BLACK AND OSHETNA RIVER MOOSE SUB POPULATION Characteristics--This migratory group of moose was not studied as part of the Susitna Hydroelectric Project. However, it was studied earlier as part of a winter calf mortality study (Ballard et al. 1982) so limited movement information is available. Moose from this subpopulation share winter range with several others in and adjacent to the impoundment zone along the Susitna River from Goose Creek to the mouth of the Tyone River. Dep e nding on snow melt, these moose move to the upper portions of the Black and Oshetna Rivers where they calve and remain through summar and autumn. The subpopulation occurs mos t ly within the territories of 2 wolf packs, both of which prey heavi ly on moose (Ballard et al. 1982, 1983). Like other subpopu1atio~s in the study area, predation by brown bears accounts for most ~alf mortality (Ballard et al. 56 1981, 1985). The area is heavily bunted. Access is provided by n~roua all-terrain-vehicle trails, several airstrips, and by float plane. The subpopulation was censused in 1985 and has been surveyed annually to determine sex and age composition. The subpopulation was estimated at 400 . Impacts--The largest impact to this particular subpopulation could be crowding on winter range created by the presence of displaced moose from other subpopulations. This could result in high rates of mortality due to winter-related causes and pre- dation. Loss of habitat, incidental mortality, and increases in poaching and hunting activity could also contribute to declines in this subpopulation. This subpopulation may decline by an average of 101 as a result of the project. 9. CLEARWATER CREEK-MACLAREN RIVER HOOSE SUBPOPULATIONS Characteristics--This group of moose is composed of 2 separate subpopulations which breed in different drainages: Clearwater Creek and the upper Maclaren River. However, because both groups utilize the impoundment zone similarly, they are considered jointly. These 2 subpopulations of moose winter along lower Clearwater Creek and the lower Maclaren River to the big bend in the Susitna River. Both subpopulations are highly migratory. During some years moose calve on the wintering ar~a and then slowly move northward to &WEer range in the Clearwater Mountains. These moose remain in alpine areas through September and October. Heavy snow storms appear to stimulate migration and movement to winter range. Several subadults from the Watana-Coal, Fog-Tsisi, and Watana Creeks subpopulations have dispersed to this area. Because of the migratory nature of this subpopulation, it is exposed to predation by several wolf packs (Ballard et al. 1982, 1983). Predation by brown bears accounts for most summer calf mortality (Ballard et al. 1981) while wolves account for most winter mortality not attributable to starvation. Black bears are rare in the area. This subpopulation ls heavily hunted primarily because of its proximity to the Denali Highway. Based on autumn moose compos ition surveys and a census conducted in 1983, this subpopulation was estimated at 675. However, movements of radio-collared adults suggested that not all of the moose from this subpopulation winter near the impoundment zone. Two (151) of 13 radio-collated cows utilized the impo!mdment area. Based on this ratio it was estimated that 100 moose winter in or adjacent to the proposed impoundment. Impacts--This subpopulation of moose would be impacted similarly to tbe Black-Osbetna River subpopulation. Increased rates of mortality due to crowding, increased predation, and elimination 57 ' of dispersal into the area aay contribute to a decline. Nu8ber of 1100se which utilize the impouncment area •ay decline by an average of 50'1 (50 of 100 11100se). 10. BUI"l'E CREEI·SUSlTI~A RIVER HOOSE SUBPOPULATION Che.racteristics--Hovements of 2 radio-collared moose sugested that a relat i vely small subpopulation calved on Butte Creek wher~ it remains through sw.er. During J ate autumn or early winter the group migrates to winter range alona the big bend of the Susitna River. Winter ranae was shared with several other subpopulations. During some years, moose from this group sp3nd winter on SWIIIIIer range. The subpopulation occurs within the territories of 2 wolf packs (Ballard et al. 1982, 1983). However, brown bears were the most important cause of moose calf mortality. Black bears were rarely observed. Hunting pressure along the Denali Highway is heavy. Census data sugest this subpopulation numbers about 135. Impacts--This subpopulation would be impacted similarly to other migratory subpopulations wi ntering along the upper impoundment zone and may decline by about 10\ (N=14). 11. LAKE LOUISE FLATS-SUSITIJA RJVER HOOSE SUBPOPULATIONS Characteristics--During autumn, moose from his subpopulation move to areas along the big bend of the Susitna River where they remain through winter. During some years t h ey do not migrate. Parturition occurs on the Lake Louise flats, particularly along wateno.1ys. The subpopulation lies within the territori al boundaries of at least 4 wolf ks (Ballard et al. 1982, 1983). Brown bears annually kill about half of the calves. Bl ack bears were rare and not an important mortality factor. The area is heavily hunted with access provided principally by b oat, float plane, or road. The Lake Louise flats were partially censused in autumn 1983, and the moose population within a 632 mi 1 (244, km 1 ) arec. was esti- mated at 430. Not all of these moose are migratory and only about 50 utilize the impoundment zone. Impacts--This subpopulation would be impacted similarly to the other migratory subpopulations which utilize the upper Watana impoundment zone and may decline by about 10~. Predicted Impacts Based On Habitat Carrying Capacity Durin& 1981. and 1985, browse quantity and quality were determined within and outside the impoundment zones (Becker and Steigers 1987). 58 Browse production was areater outside the illpoundltent zones than within thu (Fia. 44). About half of the total browse production occurred between elevations of 2,450-2,970 ft (747-905 m) but areatest browse ~.:tilization by .aose occurred at lower elevations where less browse was producted (Fia. '•5). Utilization of browse within the illpounct.ents (2 . 200 ft) during 1985 (a winter of moderate severity) was about 7~. :owsing intensity was areater within both impoundment zones than outside (Fia. 46). The impoundment zones may even be more illportant to moose during severe or moderately severe winters. Unfortunately, severe winter conditions never occurred durin& years when radio-collared 1100se were adequately monitored. Durin& 1978-79 (a relatively severe winter) low-intensity monitorin& i n dicated that moose utilized different winterin& areas than those used during years of aid or moderate winters (Ballard and Taylor 1980). i s suaaests that .aose from other . ~ populations which normally do no use the impoundments may use tham during severe winters. Host (971) relocation of radio-collared moose occurred at elevations S3,400 ft (1,036 m) (Fig. 47). Hoose rarely utilize habitat elevations over 4,000 feet; such that did use occur took place durin& su.aer months. Hoose use was correlated (p c 0. 05) with browse production at different elevations, with proportionately more moose use than expected occurrin& at elevations S2,200 ft (671 m). Winter use of the impoundment zones appeared partially dependent on snow depth. Browse appeared less available at higher elevations d11ring years of moderate snowfall. When snow accumulations made browse unavailable at high elevations, moose moved into the impound- ment zones where browse was more available. As snow receded in spring, moose moved out of the impoundment zones. Annual use of the impoundment zones by moose was variable. Averaae elevation of 74 radio-collared adults was lowest during winter and spring and highest during autumn. Use of impoundment zones by indi- vidual ~oose was also variable, ranging from no use to 1-3 months use. Mild winter conditions were probably responsible for the large amount of annual variation in numbers of moose observed during winter cen- suses and distribution flights (approximately 40-600 moose estimated from censuses during 1 -2 day periods during March). During a severe winter the impoundments are expected to support larger numbers of moose. A winter moose distribution survey and snow measurements (Steigers et al. 1985) conducted duri.ng late winter 1985 supported the general concept that moose were avoiding areas of deep snow. Althou&h moose seek areas of hi&h browse production, availability during wi.nter may be the most important factor determining moose distribution. Esti.mates of the numbers of moose that could be sustained for a 90-day winter period within habitats that would be lost by construction of the project were variable, depending on assumptions concerning diet composition and degrees of browse utilization (Tables 21 and 22). The 50 and 601 browse utili.zation categories were inte nded to represent the lona-term carrying capacity. The 1001 utilization estimates were intended to represent the severe winter capacity. · 59 Althouah estiaates of the numbers of moose which could be sustained by the habitat under a given set of assumptions were useful for att.-ptina to interpret differences between population and habitat baaed data, potential for under-or over-estimating the importance of an area existed. Between 40 and 600 moose were estimated from counts within the impoundment zones during at least 1-to 2-day periods i n March during the study . Actual use of an area by moose, and their physical con- dition at the time can alter estimates of habitat carrying capacity. Also, high rates of winter mortality might be interpreted as indi- cating a partic ular habi tat was not important to the population. However, small differences in winter survival due to the presence or absence of key winter habitats can drastically alter the recovery of an ungulate population in future years . Historical counts of moose and tracks within the Susitna impoundments (Mcilroy 1975) suggest that the area is important as winter habitat durin& severe winters. The hypothesis has not been tested that the iapoundment zones provide key winter range that allows the moose population to recover more quickly from severe winters than if the habitat did not e x ist. Actual carrying capacity of the area could be several times larger than estimated if moose utilize the areas for either shorter periods or at different levels of physical condition. If browse resources become overutilized during severe winters and if those conditions only occur once or twice every 25 years, there may be no long-term harm to the plant community from overbrowsing. However, loss of critical habitat could be so important that the size and health of the moose population in future years could be substantially altered. Summary Of Proj ect Impacts Three d i fferent methods were used for predicting the impacts of the proposed project on moose. The first method est i mated specific pro ject actions on specific moose subpopulat i ons. This method predicted that a total of about 1,300 moose would be lost as a result of the proj ect . The latter figure included not only direct losses but looses attributable to secondary effects. These estimates were similar to the estimated numbers of moose which might be supported by habitat (the 2nd method) within the impoundments during severe winter conditions (assuming 100% use of annual browse and a digestibi lity factor of 1. 00, about 1, 182 moose could be supported for 90 days) . The 3rd method, population modeling, would have demonstrated that the decline would not be a static number. The population would continue to fluctuate but over a lower range of sizes . All of the methods suggest that losses to the moose population could be great. This finding is consistent with the hypotheses of biologists i n other areas of North America where riparian habitats important to moose have been inundated or altered (E. Warren, pers. comm .; K. Childs, pers. comm.; F. Harper, pers. comm .). Actual losses can not be predicted and would not be known until pre-and post-impoundment data can be compared . If built, this project offers the best opportunity for compar ing preproject populations wi th those occurring after the project becomes 60 ~rational. To docu.ent with accuracy and precision the expected illpacts on 1100se, several studies should be conducted duriJla con- struction or after the project is operational. Monitor ina Progra.s Necessary For Refinement Of Impact Assess..nt. The illpacts of hydroelectric development on wildlife, and particularly 1100se, have never been quantified because either post-impoWldment studies were not co.parable to data prior to inundation, oF because no pre-inundation studies were conducted. Consequently, estimates of losses have been speculative, as are the estimates presented in t h i 8 study. To properly assess actual losses, it would be necessary to conduct in-depth post-impoundment studies for comparison. A large nuaber of potential mechanisms of impact have been identified as a result of this study. Unfortunately, many of t he specifics would remain speculative, but the net results of several impacts should be measurable. For example, any effects on the moose population from drifting snow would be difficult to separate from other types of habitat loss or alteration. However, the cumulative effects of those illpacts could be quantified by comparing estimates of numbers of moose in the study area, before and after the project, with those in control populations. Therefore, for efficiency of study, several similar impact mechanisms should be grouped and evaluated by similar study methods. A ti8e table of when various impacts on moose might first be observ- able and when those impacts mi ght be most severe is summarized in Table 23. All estimates are speculative and serve as guides for initiating post-impoundment studies. To properly document the impacts from this project, post-impoundment studies should use similar methods, of an intensity equal to those used in pre-impoundment studies. Types of studies needed for proper documentation of impacts on moose are also presented. ~a most important components of post-impoundment studies consist of w~se censuses, maintenance of a pool of radio-collared calf and adult moose, and predation rates studies (Table 23). This level of study would allow documentation of tot al losses of moose and ident i fication of major impact mechanisms. Proper and adequate documentation of impacts due to this project could guide future assessments for other projects which should then require less exhaustive studies for adequate prediction of impacts. 61 -. LITERAnJRE CITED Bailey, T. N., A. W. Fran~ann, P . D. Arneson, and J. L. Davis. 1978. Kenai Peninsula moose population identity study. Alaska Dep . Fish and Game, Fed. Aid in Wildl. Restoration. Final Rep. Proj. W-17-3, W-17-5 , W-17-6 , W-17-7 , W-17-8, and W-17-9 . 84pp. Ballard, W., B. 1982. Gray wolf-brown bear relationships in the Nelchina Basin of southcentral Alaska. Pages 71-80 in F. H. Harrington and P. C. Paquet, eds. Wolves of the world. Noyes Publ., Park Ridge, N. J. ---=--=---' A. W. Franzmann, IC. P. Taylor, T . Spraker, C. C . Schwartz, and R. 0. Peterson. 1979. Comparison of techniques utilized to determine moose calf mortality in Alaska. Proc. North Am. Hoose Con£. Workshop 15 :363-387. -------,---::-:-' and C. L. Gardner . 1980. Nelchina yearling moose mortality study. Alaska Dep. Fish and Game, Fed. Aid in Wildl. Restoration Prog . Rep., Proj. W-17-11 and W-21-1. 22pp. , , and S . D. Hiller. 1980_!. Influence of ___ p_r_e_d:-a-:t-ors on sUIIIDer movements of moose in southcentral Alaska . Proc. North Am. Hoose Conf. Workshop 16 :339-359. -----:--• ------=-' and S. D. Hiller. 1982a. Nelchina yearling moose mortality study . Al aska Dep. F i sh and Game, Fed. Aid i n Wildl. Restor ation Final Rep ., Proj. W-21-1 and W-21-2 . 37pp . ----=--,...---' , J . H. Westlund, and J. R. Dau . 1982b . Susitna Hydroelectric Project. Phase I Final Report , Big Game Studies, Vol. V, Wo lf. Alaska Dep . Fish and Game . Anchorage. 220pp . ----:--:-::-' and D. Larsen. 1988. relationshi ps to moose manageme.nt. Implications of predator-prey Viltrevy :in press . --~-,.--• and S. D. Hiller. 1987 . Effects o f reducing brown bear d e nsity on moose calf survi val i n southcentral Alaska. Trans. Int. Congr . Game Biol. In press . -------::-:-' , and T. H. Spraker . 1980b. Hoose calf mortality study. Alaska Dep . Fish and Game. P.R. Proj. Final Rep., W-17-9, W-17-10, and W-21-1. 123pp. -------,..,.--• S. H. Hiller , and J. S. Whitman. 1986. Modeling a southc entral Alaskan moose population. Alces 22 :201-243 ---=---~· R. 0. Stephenson, and T. H. Sp r aker. 1981a. Nelchina Basin Wolf Studies . Alaska Dep. Fish and Game, Fed. Ai d in Wi ldl . Restoration Rep ., Proj. W-17-9 and W-17-10. 201pp . ------' T. H. Spraker, and K. P . Taylor . 1981~. Causes of neonatal moose calf mortality in southcentral Alaska. J . Wildl. Manage. 45:335-342. 62 , and 1. P. Taylor. 1980. Upper Susitna Valley moose ---pop--u-=-1-a-:-tion study. Alaska Dep. Fish and Gae, Fed. Aid in Vildl. Restoration Rep., Proj. V-17-9, V-17-10, and V-17-6. 102pp. --~~-=--~· and J. S. Vhit•an. 1987. Susitna Hydroelectric Project Final Report. Big Gae Studies. Vol. III. Moose-upstreu. Alaska Dep. Fish and Game. Anchorage. In press. --....,..-:---=-' , and C. L. Gardner. 1985. Susitna Hydroelectric Project. 1984 Annual Report. Big Game Studies. Moose Upstreu. Alaska Dep. Fish and Game. Anchorage. 48pp. ~-....,..---• , N. G. Tankersley Hessing. 1983. Susitna Hydroelectric Report. Big Game Studies. Vol. III. Dep. Fish and Game. Anchorage. 147pp. L. D. Aumiller, and P. Project. 1983 Annual Moose-Upstream. Alaska Barry, T. V. 1961. Some observations of moose Bathurst Peninsula, Northwest Territories. 75:164-165. at Vood Bay and Can. Field Nat. Becker, E., and V. Steiger. 1987. Browse report. Alaska Dep. Fish 1111d Game. Anchorage. In Press. Bishop, R. H., and R. A. Rausch. 1974. Moose population fluctuations in Alaska, 1950-1972. Naturaliste Can. 101:559-593. Blood, D. A. 1974. Variation in reproduction and productivity of an enclosed herd of moose (Alces alces) . Trans. Int. Congr. Game Biol. XI :59-66. Bonar, R. 1985. Interview with V. Steigers and R. Fairbanks, April 18, 1985. Jackson Hole, Wyoming. LGL, Anchorage. Cederlund, G., K. Larsson, and F. Sandegren. 1987. Summer movements and dispersal of female moose and their offspring in a migratory population in Sweden. J. Wildl. Manage. :in press . Child, K. N. 1983. Railways and moose in the central interior of British Columbia: a recurrent management problem. Alces 18: 118- 135. Danilov, P. I. 1988 . Population Dynamics of moose in the USSR. Viltrevy. In press . Eide, S., and W. B. Ballard. 1982. Apparent case of surplus killing of caribou by gray wol v es. Can. Field-Nat. 96 :87-88. Faro, J., and A. W. Franzmann. 1978. Productivity and physiology of moose on the Alaska Peninsula. Alaska Dep. Fish and Game, t 9d. Aid in Wildl. Restoration Rep., Proj. W-17-9 and W-17-10. 29pp. 63 Fran.-ann, 1974. IIeUt. Final 54pp. A. W., P. D. Arneson, R. E. LeResche and J. L. Davis. Developing and testing new techniques for 1100se •anage- Alaska Dep. Fish and G .. e, Fed. Aid in Wildl. Restoration Rep., Proj. W-17-2, W-17-3, W-17-4, W-17-5 and W-17-6. ------~· and C. C. Schwartz. possible population condition 49:394-396. 1985. Hoose twinning rates: a assessment. J. Wildl. Manage. _, and R. 0. Peterson. 1980. Causes of J . Wildl. SUDIII8r 1100se calf mortality on the Kenai Peninsula. Manage. 44:764-768. ---=--:-:--;' ~-=-=---=---;' D. C. Johnson, J. B. Faro, and W. B. Ballard. 1984. IIIIDobilization of moose with carfe.ntanil. Alces 20:259-281. Gasaway, W. C., and S. D. DuBois. 1985. Initial response of moose, Alces alces, to a wildfire in interior Alaska. Can . Field-Nat. 99:135-140. ----:-~~· , and K. L. Brink. 1980. Dispersal of subadult moose from a low density population in Interior Alaska. Proc. N. Amer. Moose Conf. and Work~hop 16:314-337. -----:-' , and S. J. Harbo. 1981. Moose survey procedures development. Alaska Dep. Fish and Game, Fed. Aid in Wildl. Restoration Final Rep., Proj. W-17-11, W-21-1, and W-21-2. 66pp. --~~~-· R. 0 . Stephenson, J. L. Davis, P. E. K. Shepherd, and 0. E. Burris. 1983. Interrelationships o f wolves, prey and man in Interior Alaska. \Hldl. Honogr. 84. 50pp. Harper, F. E. 1985. Personal correspondence toW. Ballard. 3pp. Hauge, T. H., and L. B. Keith. 1981 . Dynamics of moose populations in northeastern Alberta. J. Wildl. Manage. 45:573-597. Henry, D. H. 1965. Possible precipitation changes resulting from the proposed Rampart dam reservoir. Cold Regions Res. and Eng. Lah., U. S. Army Tech . Rep. 147. 17pp. Houston, D. B. 1968. The Shiras moose in Jackson Hole, Wyoming. Grand Teton Nat. Hist. Assoc. Tech . Bull . No. 1. llOpp. Hubbert, H. E. , and C. C. Schwartz. model to predict intake based Manage. In prep. Ruminant Simulation on body condition. J. Wildl. Kruuk, H. 1972. 166:233-244. Surplus killing by carnivores. J. Zoology. 64 LeRescbe, R. E. 1974. Moose 11igrations in North America. Naturalist• Can. 101:393-415. , and J. L. Davis. 1973. Importance ~f ---f7oods-.,...--to moose on the Kenai Peninsula, Alaska. Manage. 38:279-287. nonbrowse J . liildl. LGL Alaska Resource Assoc. Inc. 1985. Mitigation plan for wildlife and botanical resources, moose species account. Contract to Narza-Ebasco Susitna Joint Venture. Anchorage. 42pp. Harkgren, G. 1969 . 6(3):1-299. Reproduction of moose in Sweden. Viltrevy Hech, L. D. 1970. The Wolf: the ecology and behaviour of an endangered species. The Natural History Press, New York, N. Y. 384pp. Mcilroy, C. 1974. Hoose survey-inventory progress report-1972, Game Hanag-ent Unit 13. 66-74pp. In D. E. McKnight, (Ed.) 1974. Annual report of survey-inventory activities, Part II. Moose, caribou, marine mammals and goat. Alaska Dep . Fish and Game, Fed. Aid in Wild1. Restoration Prog. Rep., Proj. li-17-5. 269pp. 1975 . Preauthorization of the Susitna River Hydroelectric Projects: a limited wildlife study. Alaska Dep. Fish and Game. Anchorage . 17pp. Hiller, S. D. 1984. Big game studies. Vol. VI. Black Bear and Brown Bear. Susitna Hydroelectric Project, 1983 Annual Report . Alaska Dep. Fish and Game. Anchorage. 174pp. 1985. Susitna Hydroelectric Project. 1984 Annu&l report . Big game studies . Black bear and brown bear. Alaska Dept. Fish and Game. Anchorage. --"""":":---:---:' and 11 . B. Ballard . 1981a. Homing of transplanted Alaska brown bears. J . liildl. Manage . 46 :869-876. --~---' and 1982b . Density and biomass estimates for an interior Alaskan brown bear Ursus arctos population. Can. Field-Nat. 96 :448-454 . ----:---' E. F. Becker, and li . B. Ballard. 1986. Density estimates using modified capture-recapture techniques for black and brown bear populations in Alaska. Int. Conf. Bear Res . and Manage. In press . Mohr, C. 0. 1947. Table of equivalent populations of North Americ an small mammals . Am. Midl. Nat . 3 7: 223-249. Neu, C. li., C. R. Byers, and J. M. Peek. 1974. analysis of ut:ilization -availability data. 38:541-545. 65 A techn i que, for J. liildl. Manage. Phillips, R. L., W. E . Berg, and D. B. Sniff. 1973. l'lovament patterns and range use of moose in northwestern Minnesota. J. Wildl. Manage. 37:266-278. PU.lott, D. H. 1959. Reproduction and productivity of Newfoundland aoose. J. Wildl. Hanage . 23:381-401. Rausch, R. A. 1959. Some aspects of population dynamics of the railbelt moose populations, Alaska . H.S. Thesis. Univ. Alaska. Fairbanks. 8lpp. 1967. Some aspects of the population ecology of wolves, Alaska. Am. Zool. 7:253-265. 1969. A Alaska, 1957-1968. Con£., 34:117-131. summary of wolf studies in southcentral Trans. North Am. Wildl. and Nat. Resour. Roberts, S. J. 1971. Veterinary obstetrics and genital diseases. Edwards Brothers, Inc. , Ann Arbor, HI. 776pp. Sergeant, D. W., and D. H. Pimlott. 1959. Age determination in moose from sectioned incisor teeth. J. Wildl. Hanage. 23:315-321. Schlegel, H. 1976. Factors affecting calf elk survival in northcentral Idaho. A progress report. Proc. 56th Ann. Conf. W. Assoc. State Game and Fish Comm . pp342-355 . Skoog, R. 0. Alaska . 1968. Ecology of caribou (Rangifer tarandus granti) in Ph.D. Thesis. Univ. California, Berkeley. 699pp. Smith, C. A., and A. W. Franzmann. 1979. Productivity and physiology of Yakutat Forelands moose. Alaska Dep. Fish and Game, Fed . Aid in Wildl. Restoration Final Re p., Proj. W-17-10 and 11. 18pp. Snedecor, G. W., and W. G. Cochran. 1973. Statistical methods. Iowa Stat~ Univ. Press, Ames. 593pp . Steigers, W. P . , W. B. Ballard, and E. F. Becker. 1985. March 1985 Snow Depth Survey in the middle Susitna River Basin, Alaska. Harza-Ebasco Susitna Joint Venture. Draft Report. 33pp. Stephenson, R. 0., and L . J. Johnson. 1972. Wolf report. Alaska Dep. Fish and Game, Fed. Aid in Wildl. Restoration Prog. Rep., Proj. W-17-3. 51pp (~imeogr.). -----:--::--' and 19 73. and Game, Fed. Aid in WilJl. W-17-4. 52pp (Himeogr.). Wolf report. Alaska Dep. Fish Restoration Prog. Rep ., Proj. Taber, R. D., and K. Raedeke. 1976. Biotic survey of Ross Lake Basin. Seattle, Wash. 46pp. 66 Taylor, W. P. 1965. The deer of North A.erica. The S a ckpole Co., Washington, D. C. 668pp. Taylor, K. P., and W. B. Ballard. 1979. Hoose movements and habitat use alona the Susitna River near Devil Canyon, Alask~. Proc. N. Aa . Hoose Conf. Workshop, Kenai, Alaska. 169-186pp . Trent, T. T., and 0. J . Rongstad. 1974 . Home range and survival of cottontail rabbits in southwestern Wisconsin. J. Wildl. Manage. 38:459-472. Van Ballenberghe, V. 1978. Final report on the effects of the Trans-Alaskan pipeline on moose movements. Alaska Dep. Fish and Game. Special Rep. Anchorage. 44pp. Verme, L. J., and J. H. Ozoga. 1981. Sex ratio of white-teiled deer and the estrus cycle. J. Wildl. Manage. 45 :710-715. Viereck, L. A., and C. T. Dyrness. 1.980. A preliminary classifi- cation system for vegetation of Alaska. U.S.D.A. Forest Service. Pac. N . W. For~st end Range Expt. Sta. Report. PNW-106, 38pp. Wallmo, 0. C. (eds). 1981. Mule and Black-tailed Deer of North America. Univ. of Nebraska Press . Lincoln. 605pp. 67 Table 1. su ... ry of .oose census data and subsequent population estiaates for Count Areas 7 and 14 derived froa survey s conducted 5-8 November 1980 along the Susitna River in southcentral Alaska. Moose Density Stratum Lou No . sample areas censused 11 Total S.A. per stratum 26 Area of each stratum 333.8 Moose density per stratum 1.125 Pop. estimate per stratum 375 90% confidence interval • 1986 + 371 Sightability correction factor; 1.03 Corrected population estimate • 2046 ~ 382 Medium 9 27 355.3 1.847 656 High 6 18 256.1 3.726 954 . Table 2. s~ ... ry of sample areas reaurv,yed to deteradne aiahtability correction factor for the Susitna .oose census conducted 5-8 Noveaber 1980 in southcentral Alaska. Stratified Density L M H H L H M H L L TOTALS Sample Area ·· 21 49 15 34 9 16 71 64 47 23 10 Survey First Date Time Count 11/7/80 10 0 11/8/80 11 12 11/8/80 31 7 11/5/80 19 4 11/5/80 5 0 11/5/80 5 0 11/6/80 20 10 11/5/80 5 4 11/6/80 5 3 11/6/80 19 0 130 40 SIGBTABILITY CORRECTION FACTOR • 1.03 No. Moose Observed Intensive Count 0 13 7 4 0 0 10 4 ') J 0 41 Percent Observ. 100 92.3 100 100 100 100 100 10 0 100 100 98 Table 3. Moose census results froa 4-9 Noveaber 1983 and subsequent population estiaates for the priaary iapact zona. lli&h Mecliua Low Saap. No. Area Saap. No. Area Saap. No. Area Unit Moose (a12) Unit Moose (a12) Unit Moose (1112) 30 67 19.6 48 43 13.9 41 25 8.1 51 55 13.2 45 24 17.7 3 9 11.3 42 80 8.7 6 27 11.2 9 4 13.5 36 32 13.5 4 2 10.0 21 3 12.3 27 41 15.9 5 37 14.9 10 2 12.9 18 42 13.1 28 35 21.5 32 10 11.2 34 29 14.7 29 18 11.6 150 3 10.8 53 69 9.8 22 12 10.9 154 7 11.9 135 9 11.9 13 32 16.3 125 3 11 .8 139 30 12.5 11 12 12.5 133 7 11.0 168 72 ).3. 7 39 76 11.6 130 12 12.4 140 38 12.9 123 12 19.9 158 10 10.0 184 41 11.6 129 30 9.7 205 2 10.0 131 25 11.8 202 0 15.9 172 19 13.7 56 10 15.1 177 18 11.0 88 0 11 .8 204 8 15.5 60 18 13.1 170 18 14.1 203 5 11.3 58 33 24.0 187 12 13.8 153 29 13.3 190 14 11.4 TOTALS 13 605 171.1 2 1 524 296.5 19 142 228.2 'HIGH MEDIUM LOW TOTAL TOTAL SAMPLE UNITS 19 45 58 122 TOTAL AREA 248.9 602.3 704.8 1556 MOOSE DENSITY 3.536 1. 767 0.623 1. 53 MOOSE POPULATION EST I MATE 880 1064 439 2383 TOTAL MOOSE POPULATION ESTIMATE + 90% c. I. -2130-2636 SIGHTABILITY CORRECTION FACTOR --1.19 CORRECTED TOTAL MOOSE POPULATION ESTIMATE • 2836 + 301 (10.6%) • 2,535 TO 3,137 . Table 4. Hoose census data and subsequent population estiaate for the Susitna River Stc dy Area, Noveaber 1983. Hi&h Hediua Lov s-p . Ro. Area s-p. No. Area Saap. No. Unit Hoose (ai 2 ) Unit Hoose (a1 2 ) Unit Hoose 30 67 19.6 48 43 13.9 41 25 51 55 13.2 45 24 17.7 3 9 42 80 8.7 6 27 11.2 9 4 36 32 13.5 4 2 10.0 2! 3 27 41 15.9 5 37 14.9 10 2 18 42 13.1 28 3 21.5 32 10 34 29 14.7 29 1 11.6 14 0 4 48 17.8 22 12 10.9 18 1 1 49 19.0 13 32 16.3 19 3 9 64 22.2 11 12 12.5 16 0 12 57 19.3 39 76 11.6 10 0 17 J9 21.5 7 15 15.0 8 7 13 71 14.5 12 9 22.2 18 0 14 25 15.0 6 47 22.5 5 4 1 72 9.6 8 33 20.1 16 2 25 13 23.9 3 7 11 24 11.6 9 3 12.1 19 20 9.9 15 74 13.5 6 55 13.7 2 6 13.9 TOTALS 15 771 237.6 22 617 330.5 16 77 HIGH MEDIUM TOTAL NUMBER OF SAMPLE UNITS 20 43 AREA OF EACH STRATUM 320.5 606.5 HOOSE DENSITY PER STRATUM 3.245 1.867 MOOSE POPULATION ESTUIATE PER STRATUM 1040 1132 TOTAL MOOSE PO PULAT ION ESTIMATES+ 90% C.I. • 2349 + 256 SIGHTABILITY CORRECTION FACTOR • 1.19 CORRECTED TOTAL MOOSE POPULATION ESTIMATE • 2795 (2491 -3101) OR 2795 ± 306 (11 .0%) Area (ai 2 ) 8.1 11.3 13.5 12.3 12.9 11.2 20.6 17.2 19.1 10.8 8.3 17.4 7.1 14.7 19.7 20.0 224.2 LOW 36 515.2 0.343 177 ~able 5. su ... ry of .ooae cenaua data and population eatiaate for Coapoaition Count Areas 3, 6, 7, and 12 and the Pr iaary Mooae Iapact Zone within GHU 13 of aouthcentra l Alaska, Noveaber 1983. Saap. No. Area Saap. No. Area Saap. No. Aru Unit Hooae (a1 2 ) Unit Hooae (a1 2 ) Unit Hooae (a1 2 ) 30 67 19.6 48 43 13.9 4 1 25 8.1 5 1 55 13.2 45 24 17.7 3 9 11.3 42 80 8.7 6 27 11.2 9 4 13.5 36 32 13.5 4 2 10.0 21 3 12.3 27 41 15.9 5 37 14.9 10 2 12.9 18 42 13.1 28 35 21.5 32 10 11.2 34 29 14.7 29 18 11.6 150 3 10.8 53 69 9.8 22 12 10.9 154 7 11.9 135 9 11.9 13 32 16 .3 125 3 11.8 139 30 12.5 11 12 12.5 133 7 11.0 168 72 13.7 39 76 11.6 130 12 12.4 140 38 12.9 123 12 19.9 158 10 10.0 184 41 11.6 129 30 9.7 205 2 10.0 12 57 19.3 131 25 11.8 202 0 15.9 17 39 21.5 172 19 13.7 56 10 15.1 13 71 14.5 177 18 11.0 88 0 11.8 1 25 15.0 204 8 15.5 60 18 13.1 1 72 9.6 170 18 14.1 203 5 11.3 4 48 17.8 58 33 24.0 187 12 13.8 1 49 19.0 153 29 13.3 10 0 8.3 9 64 22.2 190 14 11.4 8 7 17.4 12 77 18.6 11 24 11.6 18 0 7.1 14 53 19.5 9 3 12.1 5 4 14.7 26 44 17.1 19 20 9.9 16 2 19.7 15 74 13.5 3 7 20.0 6 55 13.7 14 0 20.6 2 6 13.9 18 1 17.2 7 15 15.0 19 3 19.1 12 9 22.2 16 0 10.8 6 47 22 .5 2 17 19.3 8 33 20.1 4 12 20.4 25 13 23.9 6 8 21.2 8 24 18.7 11 0 19 .5 10 19 21.2 19 4 16.6 22 26 18.5 24 9 15.2 23 24 18.6 25 7 17.6 27 8 14.8 31 0 16.3 32 0 19.5 34 0 19.4 TO!ALS 24 1,204 365.2 36 916 551.9 40 231 582.9 Tabla i(coatiauad): TOTAL NUMBEI OF SAMPLE UNI TS AREA OF EACH STRATUM MOOSE DENSITY PEl. STRATUM POPULATION ESTIMATE/STRATUM HIGH 34 514.5 3.297 1696 MED LOW 66 102 941.7 1473.5 1.660 0.396 1563 584 TOTAL MOOSE POPULATION ESTIMATE -90% C.I.•3843 (3562-4124) SIGHTABILITY CORRECTION FACTOR • 1. 1 CORRECTED POPULATION ESTIMATE • 4573 (4239-4908) 4573 + 335 (7.3%) Table 6. tllllber• of mose calves collared and subsequent causes of .,rtallty and survival rates in G1IO .lr.3 of sout.bc»atral Al..U c!Urinq 1977-79 and 1984. AUA I AUA 2 XDX 3 XDX 4 Itt ARBNi tiiOQi) fOfXL Calves I!J" I!J'7R I!J7!J !t7!' I!J" I!J711 !t7!' 1m"" -ym-lg,, lg'111 R ~ Radio-collared 25 31 29 85 31 26 57 24 52 56 81 218 AbaodoDec! 2 4 1 7 4 2 6 1 6 6 7 20 Alllll1D1Dq 23 27 28 78 27 24 51 23 46 so 74 198 100.0 'Death froa: Brown bear predation 8 11 12 31 16 10 26 7 24 24 28 88 72.7 Nolf predation 0 0 0 0 1 0 1 1 3 1 l 5 4.1 tJDkDown predat ton 0 0 0 0 1 1 2 1 0 1 2 3 2.5 llt.cellaDeou.s 1 1 4 6 2 1 3 1 5 3 3 15 12.4 UnkDown 0 J. 0 J. 2 0 2 J. 1 2 2 5 4.1 Black bear predation 0 0 0 0 0 0 0 0 4 0 0 4 3.3 Coyote predat ton 0 0 0 0 0 0 0 0 1 0 0 1 0.8 All causes 9 13 16 38 22 12 34 11 38 31 36 121 61.1 SurThlDq to 1 Nov. 14 14 12 40 5 12 17 12 8 19 38 17 39.9 Calf dafS 2,384 2,259 2,174 6,817 1,186 2,175 3,361 2,033 1,612 3,570 6,467 13,823 Dc t lf survival rate 1 JuDe-31 October .996 .994 .993 .994 .982 .995 .990 .995 .976 .991 .994 .991 Sunhal rate 1 JuDe-31 October .561 .414 .323 .425 .057 .429 .211 .U6 .026 .263 .426 .260 Table 7. Survival rates of radio-collared cow moose in GMU 13 of southcentral Alaska during 1976-86. Adults e Yearlinss Calves a Year No. Method 1 Method 2 No. Method 1 Method 2 No . Method 1 Method 2 1976-77 39 1.000 1.000 2 1.000 1.000 1977-78 44 0.976 0.965 1 0.567 178-79 45 0.922 0.768 25 0.936 0.703 1979-80 53 0.924 0.890 18 1.000 0.9701 16 0.938 0.938 1980-81 77 C.966 0.9:11 15 1.000 0.966 9 1.000 0.884 1981-82 84 0.920 0.889 8 0.749 1982-83 81 0.968 0.939 1983-84 48 0.957 0.947 1984-85 39 0.944 0.931 1985-86 22 0.849 0.849 X 53 0.943 0.911 11 0.937 0.921 13 0.958d o. 773d FOOLED 122 0.948 0.907 43 0.949 0.925 51 0.871 0.883 a Seven month rate from 1 November through May. b Survival rate calculated only for those animals whose final fate is known. c Survival rate calculated for both those animals whose fate was known, and for those whose fate was unknown, the average of two dates were used: one calculated which assumed all missing d animals were dead and another which all assumed all were alive. Rates excluding severe winte~ of 1978-79 were 0.949 and 0.906, respectively. Table 8. Survival rates of radio-collared bull moose in CHU 13 of southcentral Alaska during 1978-86. Ad~lts Yearliys Calves a Year No. Method Methode No. Method 1 Method 2 No. Method 1 Method 2 1978-79 26 0.279 0.279 1979-80 3 1.000 1.000 11 0.643 0.611 18 0.942 0.918 1980-81 10 0.835 0.722 17 1.000 1.000 7 1.000 0.856 1981-82 22 0.803 0.668 6 1.000 1.000 1982-83 19 0.734 0.598 1983-84 8 0.738 0.738 1984-85 6 0.641 0.641 1985-86 4 0.397 0.397 X 17 0.735 0.681 11 0.881 0 .870 17 0.740d 0.684d PoOLED 32 0.735 0.649 34 0.900 0.874 ~1 0.684 0 .669 a Seven month rate from 1 November through May. b Survival rate calculated only for those animals whose the final fate is known. c Survival rate calculated for both those antmals whose fate was known and for those whose fate was unknown, the average of two dates were used: one calculated which assumed all missing d animals were dead and another which all assumed all were alive. Rates excluding severe winter 1978-79 were 0.949 and 0.907, respectively. Table 9, Calculated annual survival rates of radio-collared calf moose in GHU 13 of southcent ral Alaska, 1977-84 (from Tables 7 and 8). Survival Rates June throu&h October Nove mber throuah Hal 2b Annual Method 18 Method Method 1 Method 2 FEMALES Pooled (19 77-84) Pooled (1976-86) Pooled (1976-86) 0.260 X 0.871 0.833 0.226 0.217 Lowest (Area 4-1884) Lowest (1978-79) Lowest (1978-79) 0.026 X 0.9 36 o. 703 0.024 0.018 Highest (Area 1-1977) Highest (1980-81) Highest (1979-80) 0.561 X 1.000 0.938 0.561 0.526 Pooled (1977-84) Lowest (1978-79) Lowest (1978-79) 0.260 X 0.936 0.703 0.243 0.183 Lowest (Area 4-1984) Pooled (1976-86) Pooled (1976-86) 0.026 X 0.871 0.833 0.023 0.022 MALES Pooled (1977-84) Pooled (1978-81) Poo l ed (1978-81) 0.260 X 0.684 0.669 0.178 0.174 Lowest (Area 4-1984) Lowest (1978-79) Lowest (1978-79) 0.026 X 0.279 0.279 0.007 0.007 Highest (Area 1-1977) Highest (1980-81) Highest (1979-80) 0.561 X 1.000 0.918 0.561 0.515 Pooled (1977-84) Lowest (1978-79) Lowest (1978-79) 0.260 X 0 .279 0.279 0.073 0 .073 Lowe s t (Area 4-1984) Pooled (1978-81) Pooled (1978-81) 0.026 X 0.684 0.669 0.018 0.01 7 Seven month rate from 1 November through May. a Survival rate calculated only for those animals for which the final fate is known. b Survival rate calcula ~ed f or both those animals whom fate was known, and for those whose fate was unknown , the average of two dates were used: one calculated which assumed all missing animals were dead and another which al l assumed all we r e alive. l hble 10. •• -1 ed total 00.. range st.au for adult resident and •tCJZ"atory cow-studied during 1976-1984 tn GIIU 13 of sout.bc::ntral Alaaka. 0 a-Rcl9e Winter Hc.e Rani! ~r HOM Rani! Pall Ito. RanGe Total Mo-. RanQe No. NO. NO. NO. NO. IG. 1ocatlou 1100ae ! SD Range IIOOH ! so Range IIOOH ! SD Range Locations IIOOSe ! SD Range Ruident 4-13 35 67 57 4-217 11 46 35 19-121 33 105 141 6-720 25-39 25 209 1:23 63-545 14-23 6 93 53 lQ-146 34 87 77 23-456 20 144 103 43-462 4Q-54 14 261 149 113-568 24-33 4 106 77 34-209 7 168 88 29-262 l 440 00 SS-69 7 280 231 123-787 34-43 7 137 157 35-430 2 144 111 65-222 7Q-84 2 301 so 266-337 44-53 2 107 6 101-111 85-104 6 366 234 111-739 'l'otal or Ave8 19 113 101 lQ-430 43 103 84 23-456 21 157 115 43-462 39-104 29 290 182 lll-787 lll9t' atory 4-13 8 173 144 15-375 2 389 330 156-622 12 333 224 89-435 25-39 s 1061 510 454-1703 14-23 9 248 2U 73-605 3 280 128 133-371 4Q-54 3 603 171 411-740 24-33 3 193 155 38-347 4 234 210 6Q-266 55-69 4 497 170 263-667 34-43 4 75 24 48-106 7Q-84 1 398 'l'otal or ltre8 15 151 127 15-375 15 263 213 60-622 15 322 205 89 4Q-104 10 505 165 263 -740 a For .resident aooae only tncUvidu.als witb 14-53 tota.l relocations were used while for •~CJZ"atory .,ose al.l relocations were used. Table 11. Co~riaon of .. an aeaaonal and total home range sizes by .. thod of calculation for radio-collare d resident and migratory adult cow .oose studied in GKU 13 of soutbcentral Alaska durina 1978-1984 (s andard deviation in parenthesea). Residents ~N•92 --Migratory (N•4) Kohr 1 s New Mohr's New Season Method (ka2 )a Method (laa2 )b Method ( laa 2 ) Method (laa2 ) Winter 58.0 36.5 134.9 52.6 (29.8) (16.3) (144.8) (65.0) Su...er 55.9 21.0 152.6 43.8 (32.6) (15.3) (79.8) (7.8) Total 258.0 81.8 507.9 173.5 (204.6) (33. 7) (168. 4) (59.6) a Kiniaum home range or convex polgon method (see Mohr 1947). b Modified minimum home rnage method (see methods sections). Table 13. Comparison among years of moose counts conducted each March within the Watana Impoundment Zone, 1981-85. Survey Estimated time No. moose no. E timated Year (min.) observed S.C. F. a moose moose/m1 2 1981 3 74 42 l.OOb 42 0.4 1982 264 174 1.67 290 2.9 1983 396 161 3.600 580 5.0 1984 NO SURVEY 1985 436 173 1. 70 3 295 3.0 a Sightability correction factor. b Fewer moose were observed on recount. Table 14. Coapa r ison aaong years of moose counts conducted each March within the Devil Canyon Impoundment Zones from 1981-85. Survey Estimated t'l.me No. moose no. Estimated Year (min.) ob e erved s.c.F.a moose moose/mi 2 1981 190 28 1.06 30 1.0 1982 1983 123 14 1.0 34 .5 1984 NO SURVEY 198~ 166 16 1.40 22 • 7 a Sightability correction factor. \.' ble 5. Comparison of browse quantity with usage by radio-collared moose outaide f Watana and Devil Cany o n Impoundments in the Suaitna River Basin of southcentral Alaska, 19 76-1985. Strata WINTER High d-For Med-Shr Low Very Low Scarce Zero SUMMER High Med-Fot" Med-Shr Lo w Very Low Scarce Zero AUTUMN High Hed-For Hed-Shr Low Very Low Scarce Zero TOTAL Hig h Med-For Med-Sb= Low Very Low Scarce Zero Expect d nuaber of moose Area(ha) relocations 14,420 4,486 12,644 52,065 56,647 80,674 9,070 14,420 4,486 12,644 52,065 56,647 80,674 9,070 14,420 4,486 12,644 52,065 56,647 80,674 9,070 14 ,'20 4,4 86 12,644 52,065 56,647 80,674 9,070 57 28 31 286 307 337 26 60 26 25 225 310 348 45 59 23 37 211 214 237 38 176 77 93 722 831 922 109 F.x pect e d number of moose relocatio ns ~h i 2 Se lection 67.5 21.4 'i 9.0 24 2 .3 263.7 376.3 41.8 65.5 20.8 57.1 234.8 255.6 364.7 40.5 51.6 16.4 45.0 185.1 201.5 287.5 31.9 184.6 58.6 161.2 662.2 720.8 1028.4 114.3 1.6 2.0 13.3 7 .9 7 .1 4.1 6.0 0.5 1.3 18.0 0.4 11.6 0.8 0 .5 Not Significant . N o ~ Significant A\ IDED P-0.04 Not Significant Not Significant Not Significant Not Significant Not Significant Not Significant AVOIDED P-0.005 Not Significant Not Significant Not Significant Not Significant 1.1 Not Significant 2.7 Not Significant 1.4 Not Significant 3.6 Not Si gnificant 0.8 Not Significant 8.9 Not Significant 1.2 Not Significant 0.4 5.8 28.9 5.4 16.8 11.0 0.2 Not Significant Not Significant AVOIDED P•O. 005 Not Significant PREFERRED P•0 .05 Not Significant Not Significant Table 16. Comparison of browse quantity with usage by radio-collared moose Watana Impoundment area along the Susitna River of southcentral Alaska, 1976-1985. Expected Expected number of number of moose moose Strata Area(ha) relocations relocations Chi 2 Select1on WINTER High 1290 2 2.9 0.3 Not significant Med 819/ 10 17.5 3.2 Not significant Low 32858 61 69.6 1.1 Not significant Very Low 66795 142 141.4 o.o Not significant Scarce 81870 193 173.5 2.2 Not significant Zero 5978 9 12.5 1.0 Not significant SUMMER High 1290 3 1.5 1.5 Not significant Med 8197 9 8.8 o.o Not significant Low 32858 27 35 .1 1.9 Not significant Very Low 66795 79 71.2 0.9 Not significant Scarce 81870 90 87.4 0.1 Not significant Zero 5978 2 6.3 2.9 Not significant AUTUMN High 1290 0 0.3 0.3 Not significant Med 8197 1 1.9 0.4 Not significant Low 32858 10 7.5 0.8 Not significant Very Low 66795 15 15.3 0.0 Not significant Scarce 81870 19 18.7 o.o Not significant Zero 5978 0 1.4 1.4 Not significant TOTAL High 1290 5 4.7 0.0 Not significant Med 8197 20 28.2 2 .4 Not significant Low 32858 98 112.2 1.8 Not signific ant Very Low 66795 236 227 .8 0.3 Not significant Scarce 81870 302 279.6 1.8 Not significant Zero 5978 11 20.2 4.2 Not significant Table 17. Comparison of browse quantity with usage by radio-colla red moose in the Devil Canyon Impoundment area along the Susitna River of southcentral Alaska, 1976-1985. Expected Expected number of number of moose moose Strata Area(ha) relocations relocations Chi 2 Selection WINTER Low 2558 1 1.3 0.1 No selection Very Low 39508 30 19.6 5.3 No t~election Scarce 32923 6 16 .5 6.7 No selection Zero 4839 3 2 .4 0.2 No selection SliMMER Low 2558 3 1.1 3.3 No selection Very Low 39508 17 15.8 0.1 No selection Scarce 32923 9 13.2 1.3 Nn selection Zero 4839 3 1.9 0.6 No selection AUTUMN Low 2558 0 0.5 0.5 No selection Very Low 39508 10 6.9 1.4 No selection Scarce 32923 3 5.8 1.4 No selection Zero 4839 1 0.8 0.1 No selection TOTAL Low 2558 4 2.8 0.5 No selection Very Low 39508 57 42.6 4.9 No selection Scarce 32923 18 35.4 8.6 AVOIDED P•O.OS Zero 4839 7 5.2 0.6 No selection Table 18. Chi-aquare analyaia of aapect aelection durina 3 aeaaona in the priaary .aoae iapact zone of the Suaitna River &aain, aouthcentral Alaaka, 1977-1984. Seaaon WINTER SUMMER FALL TOTAL Chi 2 SEL* Chi 2 SEL* Chi 2 SEL* Chi 2 SEL* ASPECT FLAT 99.9 A 78.1 A 68.6 A 243.9 A N 290.0 p 330.0 p 234.5 p 858.1 p NE 17.0 A 44.0 A 19.4 A 78.0 A E 5.6 6.1 1.6 12.7 SE 57.6 A 75.0 A 29.7 A 159.4 A s 446.1 p 395.6 p 281.4 p 1118.0 p SW 11.6 9.0 14.5 33.6 A w 0.3 1.8 11.2 8.1 NW 58.0 A 25.1 A 21.1 A 98.9 A EL Selection is denoted by A for avoidance and P for pre f erence. All significance levels are at P <0.05 with 8 degrees of freedom. hble 19. loU Co...neUOD Stnta.--......., data frat .aow ~ ... lD tM el&u.. a,_, t:. !U"" laalJI, 1964-lMS. !'UII -tb II ,, " '' 61 '' 1ti n 12 1! ,. 1! 71 11 71 H .., II 12 IS R IS • LOCArUII ,..~ Joa 10 16 10 ll ]l u ll lO , 13 u 11 1t 31 17 H 39 u u 31 19 ., I) , ... 1t u u " 11 lt u ll 11 u :10 39 11 n 17 37 1l lt n ll 3 1l 14 ... u u 33 36 H 11 lt )7 ,. )4 H lO 33 l 2 ll H )3 lO lO 34 H )3 .llpr " u ,. 1.9 39 0 ' ll )7 37 u 11 lt 39 u l3 16 10 n 31 35 31 ._.. ..... .loa 14 u 17 17 14 u 12 ll :16 u u 19 u 17 16 19 16 16 31 n ll 16 19 -16 " 16 36 16 17 u 30 n ll 31 n ll 30 l7 39 19 11 lO 25 15 31 25 .... 20 :10 11 24 :10 l6 ll 19 21 ll 20 25 16 35 11 35 ll 22 l2 26 16 11 .llpr 20 u 16 1t u 10 6 u ll lt ll 2> 10 ll 16 14 16 17 ]l 17 u 17 --.loa H ll H 19 !l ll ll 10 l6 l> 11 )6 n lO 31 ll lO 2. 20 lO ll 34 ll ,... )0 :16 )3 35 , n 19 •o 1l l5 n l6 25 13 lO H 12 ll ., ll )6 H :16 .... n 35 37 :16 ll 30 :10 .. ll ,. n •o ll .l !l u lO H ll ll l5 ll .llpr 10 24 )0 ,. ll u ll u tl ll 19 u 24 ll )0 )9 H 36 ll lO l5 l6 ~.-.... Joa 16 14 ll II 19 16 ll 16 lO 34 u n u 11 30 30 23 11 16 17 17 16 18 -l1 u 30 34 23 16 u 16 lO 33 2> ,. 14 31 23 ll 36 u ., lO 31 23 19 .... II 16 .l lO 31 17 u :10 ll 23 2> l7 19 39 31 31 24 u 30 31 20 31 .llpr lt II I 33 :10 0 5 16 )0 10 u 31 10 :10 u u 17 1 11 lt u :10 !OrAL.,. 11 .9 11 .1 :10.6 U.l l7.3 ·'·' U.l l7.l ll.l 31.3 19.9 U .1 18.] :16.1 31.) 31.1 lS.I 19.0 33 .8 24.6 23 .3 l l.> •' lfteter ....,utr t.c1u PROJECT ACTIONS z.o . TABLE •• SUUMARY OF ACTIONS AND IUPACTS OF T HE EFFECTS OF HYDROELECTRIC DEVELOPUENTS ON MOOSE POPULATION DYNAUICS IN NORTH AUERICA. ENVIRONMENTAL IMPACTS RESULTS OF IMPACTS ~-+-+-+-+-+-+~X~~~_. .... ~~~+-~~~-+~~~y~~~-~=::::. .-loll i 0 , "II~ »< ,. m X ,. < lnoroaood oonoontrotlono ot -ooo lnoroaood oonoontratlono ol prodatoro EFFECTS ON MOOSE !.l POPULATION DYNAMICS c ~ 0 :0 , 0 ! • ! • 0 , .. c 0 , u 0 • :;; ~ c i c .! • l 0 A ~ • , • ~ .. c .! • .. • ~ • • • • • ~ i ~ .. ~ • ~ .. • .c ;o : • • zz -o 1------f--+--+---+---1 w-col-~~ ~~ tnz -4 ID ~ ,. -4 0 X ,. 1-----+-----+-----~--~~---.~5 lnoroaood quontltrlqulltr ol wogototlon + + + ~ ~ \ ...._.-+-+-+;:1 X;.t lf X I X I X I X X liliiTor HH-f-f~X'-f X X X X X X oprlng ~-O > ,.. ~--++~X X i X X X f~ -"" < o i X I X I X I X I X I X X. loll :u Q CoD X X X X f-!.!w;::lfl:...tw-+-~ ,. c c.o 1 z,. 0 X I X X X X oprlflo g Z » z Q m CoD lflorouod oonoontrotlono ol Moooo ln oroaood oonoontratlono ol prodotore doorooood QUonutr/quolltr ol vogototlon ~ ~ ~ ~ , ~ ::::'"' m ~ lnorouod 111oooo wulnoroblllty ~~~~~~~~At,~~~~~~--------~.-------------------~--_. __ _. __ ~--4-~z I-+-+-~++-+4~.!X~~X~-a•now doptll lncr ooooo o ~+-+-~++-+4~--11-J.:X::.+-1"-''"tll lftcrouoo 0 g •oro ooworo I longer luting wlntoro -- - -~ I-+-+-~++-+4~--II-J.:X::.+-1011•• watoror tllln leo ~ ,.. ~ 1---~----~---~---11---1 ~ X -;; ,.. doorouod quolltr ot vegetation ---HHI-t-+-~+~-+~~~f--lloo ellolwlng z ; 0 lt---+---+---~----1~--. ~ X vogototlon lolng or Q ~ Q dolorod oprlng groon-IIP -+ -D.. lou ol lneulotlon m 0 z 1-~~+-+--+--+-4-4~---11-'X-+_.'"crouod Mud lloto c.o 0 .. borrlore to •owoononto t-----+--_--t--+-_-+---1 ~ X lncrooeod oroolon 10 ~-r1rvt-~~-r~,-~~~~~~~-~~~~~~~~----------------------~----t---~--~---+--~~ f,=;~;~;~=;;:;;~;;=;~xi~x;:X~~~~=~ .. :~::,:~:=:~::::~::~::~~~:·~~~~~n~g~ .. •D•I:~E~O~~~~~A~UL~~~~~YS .. E ............................ -. ...... -.. .+ ..... ~ .... +-.. ~~---1* t:X+X':+X"'+X'-:t-11-'X+X"+~.:::X+-4-=X.:..f-X:.:..toll-rood volllclo c u .. t'X+~-+;;.::X~+X+X"+X~.:::X:..t-4-=X=+X=-IInorouod hunting X X X •:::."~::i~!:p",'1':,~ !ooo "'=x+++~-tx:-:+4':":x~x+++x-l•lrorott X X boot• INCREASED HUMAN USE OF AREA lnarooeod dlaturbenoo .!.I + Poeltlwo Clleneo -Hoeatlvo Cllon11o 51 ··~ .. I • . 111111 spy M'MT UMtbliJtY•MT 0 ®10 0 20 ------ 10 ' ....... Fig. 2. Boundaries of fall moose sex-age composition count areas within GMU 130 120 Y• 126.27X -0.43 c r• -0.98, p<0.01 110 ~00 I 10 c I eo I c 70 41.72 + 2.43X eo 0.86, c c 50 40 0 c 30 52 VE'AR Fig. 3. Moose observed per hour during sex-age composition surveys conducted within 15 count area3 in GMU 13 of southcentral Alaska, 1956-1984. Note: Fitted curves does not include data points from 1956 through 1962. 110 0 100 10 0 ao i 70 0 8 D 10 -~ 50 ~ ..0 Y• 97.98-3.62X Y• 14.33 + 1.20X i r• -0.92, p(0.01 r• 0.93, p~0.01 30 20 10 0 52 Fig. 4. Bulls per 100 cows counted during moose sex-age composition surveys within 15 count areas in GHU 13 of southcentral Alaska, 1952-1984 . D Y• 61.04 -1. 90X r• -0.73, pc.O.Ol I 0 8 -I Y• 21.86 + 0. 65X D ~ D r• 0.57, ~0.05 ·c D D 0 D 'I'E'AR Fig. 5. Calves per 100 cows counted during moose sex-age composition surveys within 15 count areas in GMU 13 of southcentral Alaska, 1952-1984. 12 11 10 • j 8 • Y•10.30-0.29X I a r•-0.79,·· P<O.Ol • I 7 • Y•3.00+0.60X r•0 .91, P<O.Ol e • • 5 • •• 4 • • 3 52 5. 72 7. eo Fig. 6. Percent small bulls within the moose population as determined from sex-age composition surveys c onducted in GMU 13 of southcentral Alaska, 19 52-1984 . 48 44 .. 2 40 38 fi 38 ~ 8 32 .. ~ t 28 ~ 28 2 .. i 22 20 18 18 1 .. 12 83 Y•49.8-2.77X r•0.95, P<O.Ol 87 88 • • • 71 73 75 YEAR Y•l0.5+1.42X r•0.87, P<O.Ol 77 78 81 Fig. 7. Bulls per 100 .cows observed during moose sex-ge composition surveys in the Susitna. River Study Area of GMU 13 of southcent ral Alaska, 1963-1984. 140 130 120 -0.47 et: Y•l25.8X ~ 110 ~ r•0.91, Jl(Q.Ol 100 I 80 • eo ~ 70 15 Y•42.9+1.18X ~ eo r•0.47, P<o.OS z • 50 40 • 30 • 83 8S 87 88 71 73 7S 77 78 81 8.3 YEAR Fig. 8. Moos e observed per hou.r during sex-age composition surveys in the Susitna River Study Area of GMU 13 of southcentral Alaska, 1963-1984. ~ 42 40 38 ~ 38 34 8 32 • 15 30 • Q. 28 • ~ Ys23.69+1.13X 28 r•0.82, P<O.Ol 24 • • 22 20 18 • • 18 83 8~ 87 88 71 73 7~ 77 78 81 83 YEAR Fig. 9. Calves per 100 cows within the Susitna River Study Area of GMu 13 of southcentral Alaska as determined during moose sex-age composition survP.ys, 1963-1984. 10----------------------------------------------• e • 7 e 5 4 3 2 Y•6 .86-0 .25 X r •-0 .67, P<O.O S • • 8 7 88 71 • 7 3 75 YEAR • • ~ / •2 .52+0 .74X • r•0.82, P<O.O • 77 78 81 Fig. 10. Percent of yearling bulls within the moose population in the Susitna River Study Area of GMU 13 of southcentral Alaska as determined from sex-age composition surveys, 1963-1984. L&.l ~ ~ f5 ID ::1 ::;) z 21 20 18 18 17 18 15 14 13 12 11 10 8 8 7 8 5 4 3 2 1 0 N = 15'f 2 3 4 5 8 7 8 8 10 11 12 13 14 15 18 17 AGE (YE'ARS) Fig. 11. Age structure of adult moose captured within GMU 13 , 1976-1984 . > !:: ...J c( ~ a: 0 ~ u.. 0 w <:) c( 1-z w 80 80 ~ 40 w Q. w > 1- c( ...J ::l ~ ::l 0 20 J U N E J U L Y • • calf of collared cow-1977 & 1978 X··· .. X calf of collared cow -1980 o ---o radio-collared calf-19 7 7 & 197 8 26-30 31 - 4 5-9 1CH4 15-19 20-24 26-29 30-4 5-9 1ir14 15-19 20-24 26-29 30-3 Aug Sept Oct Fig . 12. Da t es of mortalities of collared and uncollared mo ose calve s wi t hin GMU 13 , 19 77 -19 80 (mod ified from Ballard e t a l. 1981). ( ,\\11 E3 lSio s o 25 50km Fig. 13. Location of 4 study areas within GMU 13 where causes of moose calf mortality were determine d, 1977-1984. c 24 22 20 "' 0 18 z "' lr lr ;::) 16 0 0 14 0 "-12 0 t 10 z "' ;::) 8 0 "' lr "-6 ~ 4 2 0 2 5 6 MONTH 7 8 9 10 11 12 Fig . 14. Frequency of large (> 9) group sizes of radio-collared moose in GMU 13 of southcentral Alaska, 1977-1985. 7 6 3 2 2 3 5 6 MONTH 7 9 10 Fig. 15. Average group size of radio-collared moose by month of observation in GMU 13 of southcentral Alaska, 1977-1985. 1 1 12 450 0 400 ~ Ya39.351nX-16 .00 2 3SO r•0.35, P<O.OS ~ --"' 300 N iii 0 "' 2SO C) ~ 0 200 0 "' 2 0 0 0 0 :I: 1SO ~ 0 0 0 z 100 0 ! li 0 0 8 0 8 0 0 so I ooo* 0 0 0 0 0 0 0 00 oO 0 0 0 0 0 0 10 20 .30 40 NUMBER OF OBSERVATIONS Fig. 16. Relationship between numbers of relocations and size of winter home range s f or r esident adult f emale radio-collared moose in GMU 13, 1976-1985. ~00 4!50 if' 400 ::IE :ll:: .._, .... 3!50 N iii .... 300 Cl ~ 2!50 .... ::IE 200 0 :a: ffi 1~0 + ::IE ::IE + :;:) 100 en + !SO 0 8 10 14 + + + + + * + + + + + + + + + •• + **++ HI 22 Y•8 .16-+4 .32 r•0.33, P<O.OS + + + + + 28 30 NUMBER OF OBSERVATIONS + + 34 38 Fig. 17. Relationship between numbers of relocations and size of summer home ranges for resident adult femal e radio-collared moose in GMU 13, 1976-1985. 800 700 ~ 800 :IE :IC ....., ... N !500 iii ... Cl 400 ~ ... :IE 300 0 :J: ~ 200" 100 0 6 4 Y•35. 3+7 .16X r•O .32, P<O. 05 12 18 20 24 NUMBER OF OBSERVATIONS --- 28 Fig. 18. Relationship between numbers of relocations and size of fall home ranges of resident adult fema l e radio-collared moose in GMU 13, 1976-1985. 32 800 Y•lSO .09+2 .13X 700 r•0.28, P<O.OS r eoo 2 0 X ........ soo 0 w 0 N iii w 400 oo (!) ~ 0 0 w 300 0 2 0 ::1: 200 0 0 • 0 ' 0 ~ • 0 00 100 0 0 0 0 0 0 0 20 40 60 eo 100 NUMBER OF OBSERVATIONS Fig. 19. Relationship between numbers o f relocations and total home range sizes of resident arl ult female r adio-collared moose i GMU 13, 1976-1985. 1.8 1.7 a us 1.~ \II 1.4 ~ ~ 1.3 ~ 1.2 YJ '";;, !::!'ll 1 .1 me 0 1 ~o.~• (.):I j,l 0.9 "" 0.8 ~ 0.7 0 J: a 0.8 0.5 0.4 0 .3 0.2 20 a a a a -0 .88 Y•l8138 .SX r•0.70, f'<O.OS a D 40 60 NUMBER OF OBSERVATIONS 0 a a 80 Fig. 20. Relationship between numbers of relocations and total home range sizes of migratory adult female radio-collared moose in GMU 13, 1975-1 985. ~ ~ ~ ~ ~ Ill 3 7.: -~ 15( P< () r; •C (;. ui ~ ~ ~ ~ Ill m 2 0 -I> I&.. I> 0 ffi m· ::1 ~ z u 0 10 12 14 115 115 20 22 215 28 CALF ACE AT SEPARATION (MONTHS) Fig. 2 1 . Ages at which moose offspring separated from adults in GMU 13 of southcentral Alaska, 1981-1984. eo 55 so ~ ~ -45 ....... w N ii5 -40 w Cl ~ 35 w 30 2 0 :J: 5 25 20 + 15 + 10 s + + + + + + + Y•l0.01+0.90X r•0.87, P(Q,OS + 15 25 35 COW HOME RANOE SIZE (MI2.) + Fig. 22. Relationship between parents' and offsprings' total home range sizes following separation of offspring from adult moose in GMU 13, southcentral Alaska, 1977-1985. ec ~0 w 40 Ill 8 :I II.. 0 30 ffi ID ~ 20 :z: 10 0 0 NUMBER OF CROSSINGS Fig. 23. Numbers of occasions radio-collared moose crossed the Susitna River in the vicinity of the proposed impoundments in GMU 13, southcentral Alaska, 1976-1984 (percentages listed above bars). 13 12 t..l 1 1 (J ffi 10 It: It: :;:) e (J 8 e ~ t; 7 i5 s :;:) s s If 4 ~ (J 3 ffi 2 0.. MONTHS Fig. 24. Frequency of occurrence, by month, of river crossings by r~dio­ collared moose in the vicinity of the 2 proposed impoundments along the Susitna River in southcentral Alaska, 1976-1984. 21 20 19 18 17 115 II) 1S Cl 14 z u; 13 II) 0 12 0: 0 11 II.. 10 0 15 9 ID 8 :2 7 ::;) 6 z s 4 3 2 1 0 0 RIVER MILE ABOVE DEVIL CANYON DAMSITE Fig. 25. Number of river crossings at specific locations for radio-collared moose in the vicinity of the 2 proposed impoundmen t s along the Susitna River in southcentr al Al aska, 1976-1985. ~ over 4000 ft. l~~~ l ow density 1111 medium density Jill high density Fig. 26. Rel a tive densities of moose within the 9 rimary moose i mpact zone along the Susitna River of GMU 13, as determined from stratification and census flights, November 1980. ~ over 4000 ft. or not surveyed ~. low density 1111 medium density 1111 high density Fig. 27. Relative densities of moose within the primary moose impact zone along the Sus i tna River in GMU 13, as determined from stratification and census flights, November 1980. 0 over 4000 ft. ~~ low density II medium density ·11 high dens i ty Fig. 28. Relative densities o f moose within the primary moose impact zone along the Susitna River in GMU 13, as determined from stratification and census flights, November 1983 . D over 4000 ft. til low density • medium density • hi gh density Fig. 29 . Relative densities of mo o se within the primary moose impact zone along the Susitna River in GMU 13, as determined from t ratification flights, March 1985. ....... o.e t a t t 0 t t t 15 N o.e .._, """'""' en-a :;)i 0.7 I! t f t f t ~ 0.8 t ~ ::I o.s - 0.~ 0 2 ~ 6 10 12 MONTH Fig. 30. Average month l y relative elevation occupied by radio-collared moose in GMU 13 of southcentral Alaska, 1976-1984 (standard deviation denoted by solid line). Lowest elevation occupied by each radio-col l ared moose was considered zero elevation . 20~--------------------~------------------------------. 1ta 18 17 18 15 14 13 12 11 10 8 8 7 8 s 4 .3 2 <18 us D 20 22 24 28 28 30 32 ELEVATlON IN FEET (X 1 00) WINTER + SU~~E!R 34 38 38 40 FALL. Fig. 31. Frequency occurrence of radio-collared moose relocations by elevation and season within GHU 13 of southcentral Alaska, 1976-1984. >40 ,g~--------------------------------------------------. 18 17 18 1~ 14 ~3 12 11 10 9 8 7 8 ~ 4 3 2 1 0 -LpiL~~.I'-fllL-".fli&-1"- <18 18 20 22 24 ~8 28 30 32 34 38 38 40 >40 ___.SLEVATIONAL STRATA (X .J.Q9) lZZJ AVAILABLE ~ USE Fig. 32. Comparison of year-round use of various elevations by radio-collared moose in relation to availability of elevations within the moose study area along the eusitna River of southcentral Alaska, 1976-1984. F\.AT I22J AVAILABLE CENn.E SLOPE CATEGORY ISS1 WINTER ~ SUMMER MODERATE Fig. 33. Use of slopes by radio-collared moose in comparison to slope availability along the Susitr.a River of southcentral Alaska, 1976-1984. 28 28 24 22 20 18 :s 18 ' 14 12 10 8 8 4 2 0 F\.AT N NE E SE s sw w NW ~ WINTER COMPAQ.ASPECT ~ SUMMER ~ FAU. liB DlPECTED Fig. 34. Annual use of compass aspects by radio-collared moose in relation to aspect availability along the Susitna River in GMU 13, 1976-1984. 800 eo8 eoo 100 87 en 48 z 0 fi eoo ~ 37 500 !57 0 ~ 400 "ffi 300 e ID 2 24 :;, z 200 1!57 110 100 83 0 0 0 0 0 0 0 4 0 3 s 7 e 11 13 1!5 17 18 21 23 TIME OF DAY Fig. 35. Distribution of radio-collared moose relocations by time of day in GMU 13 of southcentral Alaska, 1976-1985. 4 eo Ul !50 z 0 ~ ~ 40 IS 0 a 30 a Ill ~ 20 :5 0 15 10 11. 0 e 8 10 11 12 13 14 1!5 18 17 18 18 20 21 TIME OF DAY Fig. 36. Percent of observations, by time of day radio-collared moose were observed bedded, in GMU 13 of southcentral Alaska , 1976-1985. 800 en z 500 ~ ~ .ao R ~ .300 15 m :a 200 ::) z 100 0 2 3 4 IS MONTH 7 a 10 11 12 Fig. 37. Number of monthly observations of radio-collared moose in GHU 13 of southcentral Alaska. 1976-1985. 70 eo eo ~ eo ~ ~ 50 5 40 a 0 ~ 30 ~ 8 20 ffi ~ 10 0 MONTH Fig. 38 . Perc ent of total observa tions per month that radi o-collared moose were observed bedded i n GMU 13 o f southcentral Alaska, 1976-1985. 28 24 Cll 22 ~ ~ 20 ! 18 I 18 0 " 14 I 12 10 ~ IS ffi • ~ 4 2 0 2 3 s t.AONTH 7 • 8 10 11 Fig. 39 . Percent of t o tal observations that radio-collared moose were observed f eeding in GHU 13 of southcentral Alaska, 1976-1985 . 12 32 31 30 28 28 27 2& 25 24 23 22 21 20 18 18 17 1& 15 14 &3-&4 ee-e7 &8-70 72-73 YEAR 75-78 78-78 81-82 Fig. 40. Comparison of annual winter severity indices in the middle Susitna Ri ver Basin of southcentral Alaska, 1964-1985. Larger indices correspond to wi nters of increasing severity. 32~--------------------------------------------------,. 31 :so 28 28 27 28 25 2-4 23 22 21 20 18 • 18 20 22 • Y•-31.74 +18 .361 nX r-<1.95. pco.os 2-4 28 .JANUNf'f WINTER SEVERITY INDEX • 28 30 32 Fi g. 41. Relation sip between annual and January winter severity indices each year within the middle Susitna Rive Basin of southcentral Alaska, 1974-1985. 32 31 30 28 lj 28 0 27 ~ 28 ~ 2S • ~ 2<4 Ill 23 ~ 22 21 i 20 ~ 18 ~ 18 r • 0.98; P < 0.05 Y • 0.96 + l.OOX 17 18 1S 1<4 1<4 18 18 20 22 28 28 32 JAN-FEB WINTER SEVERITY INDEX Fig. 42. Relationship between annual and January-February winter severity indices each year within the middlP. Susitna River Basin of southcentral Alaska, 1964-1985. 27 28 a a 25 a 24 a ~ a a 23 ~ ~ 22 D a ~ 21 Ill D m 20 D D z 18 ~ 18 17 Y•7. 29+4 .801 nX ra0.74. P&O.OS 18 15 4 a 12 18 20 24 28 32 38 ,; WOOSE OBSERVAnONS < 2200" ELEVAnON Fig. 43. Relationship between monthly winter severity indices and percent of 1110ose relocations per month at elevations less than 2,200 ft during 198 1-84. 100 ~ 80 II) ~ 70 0 iii so 1&1 II) ~ !50 I 1&1 40 II) 8 30 2 ~ 20 10 0 2000 2200 2400 2800 2800 3000 3200 3400 ELEVATION (FEET) Fig. 44. Cumulative percent of moose browse biomass by elevation along the middle Susitna River Basin of GMU 13 in 1984 and 1985. 0 .72 0.7 o.ee o.ss 0.~ a O.S2 ~ o.s 0.58 i 0 .5S ~ 0.5-4 ~ 0 .52 0 0.5 ~ 0.48 It: G. 0.4S 0 ...... 0.42 0 .4 0 .38 0 .3S 2000-2200 2401-2SOO ELEVATION 2801-3000 3201-340< Fig . 45 . Proportion of browse used by moose at different elevations al~ng the middle Susitna River in GMU 13 in 1985. 0.9 o.a 0 0.7 w en ~ a: o.s ID z 0 ~ o.s a: 0 Cl. 0 a: Cl. 0.4 0.3 0 .2 PLANT DENSITY Fig. 46 . Comparison of browse used with plant density by area for the midd le Susitna River Basin of GMU 13 in 19 85. 100 80 en eo z 0 ~ ~ 70 en \10 CD 0 1&1 !50 en 8 :::1 40 ~ 30 jt ~ 20 10 0 <1800 1800 2200 2GOO 3000 ~00 3800 >4000 ELEVATION (FEET) Fig. 47. Percent of moose relocations by elevation for radio-collared moose along the middle Susitna River of southcentral Alaska. 1976-1985.