Skip to main content

Full text of "White-tailed deer along the lower Yellowstone"

See other formats


s 

f>39.P 
W-1.20- 
12-ts? 
BS-2. 0 



Dusek, Gary L 

White-tailed 
deer alon? the 
lower Yellowstone 



STATE DOCUMENTS COLLECTION 



I UN 



{MONTANA TATE. LIBRARY 
1515 £. 6th AVE. 

H& jhNA MONTANA 




WHITE-TAILED 
1, DEER 
ALONG THE 
LOWER 
YELLOWSTONE 




FINAL REPORT 



RESEARCH PROJECT 



STATE: 



PROJECT NO 



PROGRAM NO 



STUDY NO . : 



JOB NO . : 



Montana 

W-120-R-12-18 

I 

BG-2.0 



TITLE: Statwide Wildlife Research 

TITLE: Big Game Research 

TITLE: Statewide White-tailed Deer 
Ecology Studies 

TITLE: Population ecology and 

habitat relationships of 
white-tailed deer in river 
bottom habitat in eastern 
Montana 



Period Covered: July 1, 1980 - June 30, 1987 




Prepared by: Gary L. Dusek Approved by: Arnold Olsen 

. Richard J. Mackie John P. Weigand 



Date : 



March 31, 1988 



« mm il il eer a,ong ,he lower Y °"°" 




i i 

ACKNOWLEDGEMENTS 

We extend our gratitude to numerous employees of the Montana 
Department of Fish, Wildlife and Parks, students at Montana State 
University, and others for their assistance in data collection. 
We thank K. L. Hamlin, D. F. Pac, and A. K. Wood for their 
assistance with data collection and analysis and editorial 
reviews of the manuscript. The editorial comments provided by C. 
D. Eustace and weather severity indexes calculated by H. E. 
Jorgensen are also appreciated. Administrative support and 
assistance were provided by K. G. Seaburg and J. P. Weigand. We 
also extend our appreciation to the Ft. Keogh Livestock and Range 
Research Station, Miles City, MT, and the Department of Animal 
Science, University of Idaho, Moscow, for analysis of blood sera. 
Individual contributions of R. B. Campbell, M. Friedman, R. W. 
Hay, K. A. Kinder), R. B. Staigmiller, R. P. Stoneberg, and K. C. 
Walcheck, are gratefully acknowledged. We thank all private 
landowners for granting access to their lands and cooperation. 



Digitized by the Internet Archive 

in 2015 



https://archive.org/details/whitetaileddeera1988duse 



iii 



TABLE OF CONTENTS 



INTRODUCT ION 1 

STUDY AREA 2 

Land Uses . - 8 

Intensive Study Units 9 

METHODS 11 

POPULATION CHARACTERISTICS 17 

Historical Abundance and Trend, 1804-1950 17 

Recent Population Trends 17 

Spatial Variation in Trend on the River Bottom 20 

Population Structure 22 

General Patterns 22 

Spatial Va ria t i on 26 

Physical Characteristics 26 

Weight Relationship s 2 6 

KFI 2 7 

Diastema 1 and M a_ i_n_ Antler Beam Length 2 9 

REPRODUCTION 31 

Breeding and Fawning Seasons 31 

Pregnancy Rates 32 

Reproductive Potential 34 

Net Productivity 35 

Spatial Variation in Reproduction and Recruitment 40 

MORTALITY 41 

Neonatal/ Summer Fawn Mortality 41 

General Patterns of Mortality 42 

Hunting Mortality 43 

Traffic-related Morta 11 1 y 47 

Natural Mortality 49 

DISTRIBUTION 50 

Habitat Factors Influencing Distribution 50 

Influence of Livestock 52 

MOVEMENTS 54 

Diurnal Movements 55 

Diel Movements 58 

Home Range Characteristics 62 

Yearlong Home Ranges 62 

S_e_a_s_o_n_a_ 1_ Home Ranges 6 4 

Diel Home Ranges 64 

Home Range Stability 65 

Dispersal 67 

HABITAT SELECTION 68 

Diurnal Habitat Selection 68 

Spatial Relationships 71 

Sex and Age Relationships 73 

Nocturnal Habitat Selection 75 

FOOD HABITS „ 7 7 

Seasonal Trends 77 

Spatial Relationships 79 

Sex and Age Relationships 82 

Factors Influencing Seasonal Food Habits 82 



iv 

DEER-HABITAT RELATIONSHIPS AND POPULATION REGULATION 85 

Habitat Influences on Distribution and Abundance ....... 85 

Habitat Influences on Population Dynamics 90 

Population Regulation 93 

Influence o f Density 94 

In fluence o_f_ Dispersal 9 6 

Influence of other Factors 97 

Influence of Hunting 97 

MANAGEMENT IMPLICATIONS . 100 

LITERATURE CITED 103 



V 



ABSTRACT 

Population characteristics and habitat relationships of 
white-tailed deer ( Odocoi leus virginianus dakotensis ) were 
studied on bottomlands of the lower Yellowstone River in eastern 
Montana during 1980-86. Early autumn populations on the 224 km 
study area varied between 3,100 and 5,784 deer. Total numbers 
increased during 1980-83 and declined during 1984-86 in response 
to density influenced fawn recruitment and hunter harvests. The 
population was characterized by relatively high productivity and 
a predominantly young age structure. The average fawnrdoe ratio 
for autumn was 85:100 (64-11 2:100). Fawns comprised 32-43% of 
autumn populations, while deer 2 years and younger accounted for 
63% of all females and 93% of all males. Only 5% of the female 
segment was 8 years and older. Antlered males made up 16-23% of 
pre-hunt populations; less than 1% were 4 years and older. 

Fetal measurements indicated a median breeding date of 21 
November. The incidence of successful breeding among female 
fawns was low (< 2%); the pregnancy rate for older females was 
93%. Ovulation rates were 1.41 for yearlings and 2.09 for older 
females; fetal rates were 1.26 and 1.88, respectively. Age- 
specific reproduction and recruitment rates increased through age 
classes 5-6 (6-7 years at parturition) and declined among older 
females. Reproductive success varied inversely with density of 
adult females; females in age classes 2 and 3 were most affected. 
Average annual mortality rates were 50% for fawns, 22% for adult 
females, and 60% for adult males. Neonatal and summer mortality 
accounted for 32% of all fawns produced and varied directly with 
adult female density. Hunting accounted for 81% of all mortality 
documented among deer older than 6 months and 13-35% of autumn 
populations. Hunting-related mortality equalled or exceeded fawn 
recruitment during 1983-86, years of population decline. Intense 
hunter selection of antlered over antlerless deer contributed to 
a skewed sex ratio among adults favoring females (~2:1) in early 
autumn. Average mortality rates were low in winter among fawns 
(10%) and adults (6%) and were not significantly correlated with 
winter severity. Low overwinter mortality, high productivity, 
and seasonal trends in whole weights and kidney fat indexes (KFI) 
collectively indicated that the population wintered on a high 
nutritional plane. 

The relative abundance of woody cover accounted for 74% of 
the variation in density distribution of deer along the river 
bottom. The amount of island area, sloughs, and livestock 
distribution also influenced distribution of deer. Deer were 
generally nonm i g rat o r y , occupying relatively small home ranges 
that overlapped seasonally within the river bottom. Diurnal home 
ranges were typically smaller during summer than winter. Diel 
home ranges were twice as large as diurnal home ranges during 
summer and 3 times larger in winter. Females exhibited a higher 
fidelity to summer than winter home ranges. Diurnal movement 
indexes among adult deer were typically lowest during June-August 
and highest during November. The most distinct daily movements 
were from riparian cover used during daytime to agricultural 



vi 



fields at night. Nocturnally, deer generally selected alfalfa 
fields during summer and sugar beet or cereal grain fields during 
winter; however seasonal selection for specific crop types and 
riparian forest/shrub communities varied temporally and spatially 
with availability and sex and age of deer. Adult females used 
habitats having greater diversity of cover types, including 
agricultural fields and forest types with a shrub/herbaceous 
understory than those used by adult males. Agricultural crops 
accounted for 4U of the yearlong diet; spatially, increased use 
of crops or crop residues coincided with decreased use of browse. 

The current widespread distribution and high relative 
abundance of white-tailed deer along the lower Yellowstone River 
appeared to reflect habitat changes brought about by settlement 
elimination or reduction of other wild ungulates, and land use ' 
PFa A . Ce l' , An lntensivel y fa ™ed river bottom with a relatively 
undisturbed riparian forest/shrub complex provided a highly 

t\HV e> VSlh* P redictabl e habitat-environmental complex 

within a highly variable and unpredictable regional environment 
Observed patterns of habitat useage, that included dispersed 
distribution and limited movement within high quality cover- 
forage complexes during late spring and summer, were consistent 
with optimal opportunity to sustain high female densities and 
successfully rear fawns. Population characteristics and dynamics 
were not consistent with traditional concepts of regulation 
involving winter forage supplies. Alternatively, regulation 
involved intraspecific interactions that influenced use of 
autumn' d " rit ! g , lat \ e s P r r in § and summer and fawn survival to early 

habitat zlyJV , ° r mana § ement of ^e population and its 
naoitat are discussed. 



1 



I INTRODUCTION 

White-tailed deer are widely distributed across the northern 
Great Plains where they now occupy a wide spectrum of habitats-- 
from foothills on the east slope of the Rocky Mountains to 
bottomlands along rivers and streams, timbered uplands, and 
shortgrass prairie. Although commonly associated with 
agriculture, especially in portions of the region characterized 
by semiarid climate (Allen 1971, Swenson et al. 1983, Herriges 
1986), they can also be found in habitats where agriculture has 
only negligible influence (Dusek 1987). 

Historically, whitetails primarily inhabited river bottoms 
in the northern plains (Allen 1971). Though gradually extending 
their range to other habitats in relatively recent times, 
bottomlands along major rivers continue to be a stronghold of the 
species throughout the region. Nearly half of all white-tailed 
deer in southeastern Montana winter in riparian habitat along 
streams (Swenson et al. 1983). Similarly, riverine habitats 
typically support much higher densities of whitetails than 
adjacent prairies. 

Population characteristics and dynamics of deer may vary in 
relation to physical and biological characteristics of individual 
habitats (Cheatum and Morton 1946, Gill 1956, Dapson et al. 1979, 
Mackie 1983) but few comparative data are available. Long-term 
research on population ecology of white-tailed deer has been 
limited largely to penned or confined populations (Arnold and 
Verme 1963, McCullough 1979, Woolf and Harder 1979). Published 
studies of natural, free ranging populations are limited to 
chaparral shrublands of the Southwest (Teer et al. 1965), mixed 
conifer-deciduous forest of the upper Midwest (Hoskinson and Mech 
1976, Nelson and Mech 1981, 1984), and the tidelands of the lower 
Columbia River (Gavin et al. 1984). 

Our study on population ecology of whitetails along the 
lower Yellowstone River was initiated in 1980 as part of a broad, 
long-term research effort to determine and compare population 
ecology and habitat relationships of deer among the major 
habitats in which they occur in Montana. That effort, which 
began in 1975, was precipitated by the inability of existing 
knowledge to explain widespread declines in deer populations, 
especially mule deer (C). h em i onu s ) , across the state during the 
early 1 970's. Comparative studies offered the best opportunity 
to determine the relative influence of various potential limiting 
factors. In addition, the lower Yellowstone River offered an 
opportunity to study relationships between stream dynamics, 
riparian habitat, and white-tailed deer populations. These data 
were required in efforts to maintain an instream flow reservation 
of water in the Yellowstone River for the benefit of fish and 
wildlife. The Yellowstone River is unique in being one of the 
last large rivers in the lower 48 states that remains free- 
flowing in its entirety. 



2 



Data were collected from February 1980 through October 1986 
with emphasis on factors affecting population size and trend and 
habitat use relationships- Specific objectives were: (1) to 
determine ecological and biological parameters associated with a 
major river bottom population of white-tailed deer; (2) to relate 
those parameters to environmental characteristics and specific 
factors potentially limiting deer populations, including 
nutrition, land use practices, climate, and natural and man- 
caused mortality; and, (3) to develop methods, criteria, and 
guidelines for management of deer and their habitat in riverine 
environments. Our approach was two-fold: extensive studies of 
population characteristics and dynamics and habitat relationships 
along an entire segment of the lower Yellowstone River; and, 
intensive studies within specific units. 



STUDY AREA 

The lower Yellowstone River extends 350 km, from the mouth 
of the Bighorn River east of Billings, MT, to the confluence of 
the Yellowstone and Missouri Rivers in western North Dakota. The 
study area included 79 km (224 km ) of floodplain and islands of 
the lower Yellowstone between Inters tat e-94 at Glendive and 
Montana highway 23 near Sidney (Fig. 1). The river is paralleled 
by Montana highway 16 on the northwest side throughout the entire 
length of the study area and by the U.S. Bureau of Reclamation 
(USBR) Canal from Intake to Sidney. 



The lower Yellowstone River traverses both glaciated and 
unglaciated portions of the Missouri Plateau in the Great Plains 
Province (Fenneman 1931). The landscape of the region slopes 
gently eastward. Physiographic features consist of rolling 
uplands, alluvial deposits, and a terraced floodplain to the 
northwest of the river, and high benches and/or rugged "badlands" 
immediately adjacent to the river channel on the southeast side. 
Width of the floodplain varies from 2 km immediately upstream 
from Intake to 7 km south of highway 23 (Fig. 1). Sandstone, 
shale that includes some outcropping of clinker or "scoria", and 
coal seams of the Fort Union Formation are intermittently exposed 
along the Yellowstone Valley. Elevations on the river vary from 
625 m at Glendive to 577 m at Sidney. 

The length of the main channel in the study area is 101 km 
with an average sinuosity of 1.3 (Hinz 1 977). Mean stream 
discharge normally peaks in June (1,121 m^/second) and drops off 
from July (662 m ^/second) to August (223 m /second) (Moore and 
Shields 1980). Major flooding caused by ice jams occasionally 
occurs along the floodplain during February or March. 



Soils of 
loams on t he 



the 
low 



study area are deep, fine sandy loams 
terraces and floodplain and silty clay 



and silt 
loams on 



3 



MONTANA 



if SIDNEY 



1% 



INTAKE 
DIVERSION 
DAM, 



>3> 



w 



£8 



t 

[rl 



CREEK 



6.5 (km.) 



SCALE 



GLENDSVE 



\ 



Areas of Intensive Study 

1. Seven Sisters 

2. Elk Island 

3. Burns Creek 

4. Intake 

5. Glendive 



Figure 1. The lower Yellowstone River study area. 



4 



alluvial fans and upper terraces (Holder and Pescador 1976, 
Pescador and Brockrnan 1980). Soils of adjacent dissected 
sedimentary plains vary from deep silt loams to gravelly sands 
underlain by silty clay sedimentary beds or a tbick layer of 
sand. 

The regional climate is semiarid with marked variation in 
seasonal precipitation and temperature. Weather data for the 
study area (Table 1, Fig. 2) were from official records for 
Sidney, Savage, and Glendive (U.S. Dep. Comm. 1 979-86) with 
longterm average values based on data for the years 1951-80. 

Mean annual precipitation in the region is about 35 cm of 
which approximately 80% falls during April-September. Sidney and 
Savage receive slightly greater annual precipitation than 
Glendive, although monthly means are variable between locations 
and highly variable between years (Table 1). Mean annual 
precipitation between the 3 stations varied during 1979-85 from 
78 to 82% of the long-term average. Precipitation was above the 
long-term average only in 1982. 

The warmest and coldest mean daily temperatures are recorded 
in July and January, respectively (Fig. 3). Average monthly 
temperatures are increasingly cooler with distance north from 
Glendive. Monthly mean temperatures at Glendive average 2.3 C 
warmer than those at Sidney. The frost-free period along the 
valley floor at Glendive is 130 days. 

Winter weather, characterized by subfreezing mean daily 
temperatures and periodic snowfall, may begin during November and 
persist through March (Fig. 3). Average annual snowfall along 
the lower Yellowstone Valley is about 71 cm; significant 
accumulations are rare. Trends in winter severity, based on an 
index that considered the cumulative effects of winter weather on 
the condition of free-ranging deer using concepts of Leckenby and 
Adams (1986), for only Sidney and Savage appear in Table 2. By 
comparison, the winter of 1981-82 was the severest during the 
period of study, but less severe than 1978-79. The winters of 
1979-80, 1980-81, and 1982-83 were comparatively warm and 
snowfree. Winter 1983-84 was characterized by abnormally cold 
temperatures and complete snow cover during December through mid 
January followed by above average temperatures and snow-free 
conditions through February. Temperatures and precipitation 
approximated long-term averages during the winters of 1984-85 and 
1 985-86. 



Major riparian communities along the lower Yellowstone were 
identified and successional relationships described by Boggs 
(1984). Communities include young cottonwood, mature Cottonwood, 
shrub, grassland, willow-shrub, green ash ( F raxi nus 
pennsy lvanicus ) , and peach- leaved willow ( S a 1 i x a mygdaloides ). 
The grassland sere that dominates the floodplain originates with 
plains cottonwood ( Populus deltoides ) and sandbar willow ( S . 
f luviatilis ) seedlings colonizing recently formed sand and gravel 
bars. The willows disappear after approximately 20 years. The 



5 



Table 1, Monthly precipitation data for 3 reporting stations on 
the lower Yellowstone River including 30-year (1951-80) means 
(N), and 7-year means (X), and coefficients of variation (CV) 
for the calender years of 1979-85. 

Precipitation by Station (cm) 



Sidney Savage Glendive 



Month N X CV N X CV N X CV 



January 


1 


. 02 


1 . 


00 


91% 


0. 


94 


0. 


71 


109% 


1 . 


14 


0. 


74 


9 9% 


February 


1 


.02 


0. 


69 


108% 


0. 


79 


0. 


51 


130% 


1 . 


04 


0. 


46 


13 9% 


March 


1 


. 19 


1 . 


68 


87% 


1 . 


09 


1 . 


59 


102% 


1 . 


14 


1. 


68 


84% 


April 


3 


.05 


2. 


1 7 


68% 


3. 


27 


2. 


04 


63% 


2. 


97 


1. 


63 


73% 


Ma y 


5 


. 33 


2. 


74 


67% 


5. 


26 


3. 


85 


8 8% 


5. 


18 


3. 


39 


82% 


Jun e 


7 


.06 


5. 


53 


57% 


8. 


23 


5. 


77 


70% 


7. 


95 


4. 


97 


5 7% 


July 


4 


. 47 


4. 


38 


4 8% 


4. 


98 


3. 


27 


59% 


4. 


37 


2. 


78 


67% 


Aug us t 


4 


.14 


3. 


64 


75% 


3. 


86 


3. 


55 


72% 


3. 


40 


4. 


14 


108% 


Sept ember 


3 


. 28 


2. 


89 


4 3% 


3. 


25 


2. 


63 


4 7% 


2 . 


92 


2. 


22 


72% 


October 


1 


.98 


3. 


22 


91 % 


1 . 


78 


3. 


09 


89% 


1 . 


72 


2. 


81 


81% 


No vember 


1 


. 19 


0. 


88 


56% 


0. 


91 


0. 


62 


71% 


1 . 


16 


0. 


66 


76% 


December 


1 


.07 


1 . 


22 


71 % 


0. 


84 


1 . 


24 


45% 


1 . 


04 


1 . 


13 


59% 


Tota 1 


34. 


80 


30. 


04 




35. 


2 


2 8. 


87 




34. 


03 


26. 


61 





resulting young cottonwood community is characterized by a 
maximum canopy height of 20 m and a relatively sparse understory 
of grasses. Cottonwoods continue to mature and reach a maximum 
height of about 25 m. The mature cottonwood community is 
characterized by widely spaced trees and a dense understory of 
shrubs and herbage. Cottonwoods disappear after approximately 
100 years, leaving a shrub community dominated by western 
snowberry (Sym phoricarpos occidentalis ) and Woods rose ( Rosa 
w oods i i ). Herbaceous cover in the shrub community is less than 
in mature cottonwood stands. Receding shrublands are replaced by 
grasslands dominated by western wheatgrass ( Ag ropy ron smith! i ) 
and prairie sandreed ( Ca la m ovilfa longifolia ). Silver sagebrush 
( Arte m isia can a ) is locally abundant throughout the grassland 
c ommu n i t y . 

Green ash may first appear in either the young or mature 
cottonwood stands and ultimately replaces the relatively 
ephemeral cottonwoods and apparently constituting climax 
vegetatation on sites capable of continuing to support tree 
cover. Other tree species occurring with green ash include 
American elm ( IJ ]_m va s_ am ericana ) and boxelder ( Acer negund o ) . 
Boggs (1984) characterized regeneration of these species as 
"patchy . " 



< 10 - 

o 

0 -f — i ~ — i — — 1 " 1 — r~ 1 1 — 1 

79 80 81 82 83 84 85 86 

YEAR 




YEAR 

Figure 2. Annua 1 precipitation and departure from the longterm 
mean (1951-80) at 3 locations on the lower Yellowstone River, 
1 979-86. 



7 



30 -i 
25 - 
20 - 
15 - 
10 - 
5- 
0- 
-5 - 
-10 - 
-15 - 



-20 



Glendive 
Savage 
Sidney 




T 



JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC 

MONTH 



Figure 3. Mean longterra (1951-80) daily temperatures by month at 
3 locations on the lower Yellowstone River. 



The peach- leaved willow sere becomes established where 
willows invade newly-formed sand/gravel bars in the absence of 
plains Cottonwood. A green ash community may eventually replace 
the willows similar to succession in cottonwoods. Marshes occur 
in ephemeral stream channels, but successional relationships were 
no t determined. 

Plant communities in the uplands adjacent to the river 
bottom were described by Swenson (1982). Mixed grasslands, where 
western wheatgrass, blue grama ( Rou t e 1 oua graci lis ) and thread- 
leaf sedge ( C a r ex filifoli a) predominate, occur on flat to 
rolling uplands. Some of the lower side drainages are bordered 
by badlands characterized by steep, sparsely vegetated slopes. 
Stands of Rocky Mountain juniper ( Juniperus s copuloru m) occur 
along some of the deeply eroded drainages. The more mesic draws 
support deciduous trees and shrubs that include green ash, 
boxelder, American elm, and silver buffaloberry ( Shepherdia 
argentea ). Cultivated lands on the upland terraces and benches 
support primarily dryland cereal crops. 



8 



Table 2. Winter severity based on a modified Leckenby index for 
Savage and Sidney, MT, 1975-76 through 1985-86. The index 
includes 15 November through 31 March. 



Winter Severity Index 
Year Savage Sidney 



1975-76 


1095 


1500 


1 976-77 


1053 


13 93 


1977-78 


1612 


2191 


1978-79 


1 973 


3063 


1979-80 


958 


1115 


1 980-81 


710 


816 


1981-82 


1536 


2110 


1 982-83 


736 


879 


1983-84 


1090 


1238 


1 984-85 


1058 


1164 


1985-86 


1119 


1409 



Land Uses 

The lower Yellowstone served as a route of travel by 
Indians, trappers, and traders prior to settlement in the late 
1800's (Brown 1961). The area was first settled by cattlemen 
during the late 1870's. Extensive homesteading occurred between 
1900-30. The area includes several small rural communities 
established during the homestead era or construction of the Lower 
Yellowstone Project: Crane, Savage, Intake, and Stipek. Human 
population has dwindled since the 1930's in all but Savage. 

Irrigation of native meadows that involved spreading water 
with simple diversions also began during the early 1900's as a 
result of low production or failure of native hay during dry 
years. Construction of the Lower Yellowstone Project began in 
1905 with water available in 1909. The project was built to 
irrigate 22 4 km 2 (56,000 acres) of floodplain and terraces 
downstream from Intake; however, the potential for irrigated 
farming lay dormant until production of sugar beets became 
important in the local economy in 1925 (Mercier 1985). 
Agriculture is the principal land use along the lower Yellowstone 
Valley and includes production of livestock, forage, and cash 
crops. The floodplain on the northwest side of the river 
downstream from Intake that is flood irrigated from the Lower 
Yellowstone Project (US B R main canal, Fig. 1) is the most 
intensively farmed portion of the study area. Sugar beets, 
alfalfa, corn, and small grains are the principal crops. 
Untilied portions of the floodplain are grazed by livestock. 

Livestock forage, including corn and hay crops, is produced 
on a few bottomlands on the southeast side of the river. These 
occur primarily where ephemeral streams enter the Yellowstone. 
Water for irrigation is provided by private pumping systems. 



9 



Most of these bottoms are also grazed, at least during fall 
and / or winter. 

Production of livestock and forage crops are the principal 
land uses on the floodplain above Intake. Small grains are also 
produced. Hay crops include alfalfa and native and introduced 
grasses. Most areas are grazed during fall and/or winter, 
although they may also be grazed during abnormally dry summers 
when forage production in the uplands is poor. 

Although some oil production structures occur on the 
floodplain near Sidney, recreation, including hunting, fishing, 
boating, and searching for agates, is the only other major land 
use. Because of their isolation and relative inaccessibility, 
few islands on the river receive agricultural use. Their major 
use is by re c r e a t i o n i s t s . The Montana Department of Fish, 
Wildlife, and Parks (MDFWP) owns and manages 2 parcels of land 
that include both floodplain and islands. These areas are 
managed for wildlife habitat and recreation, including access to 
the river. 



In 1978, the MDFWP was granted an instream reservation of 
water in the Yellowstone River under the 1973 Montana Water Use 
Act that recognized fish and wildlife as beneficial uses of 
water. This reservation, 5.5 million acre/feet at Sidney, 
constitutes a flow that can be expected to be equalled or 
exceeded in 85 of 100 years. 



Intensive Study Units 



Initially, 3 areas that reflected variation in kinds and 
intensity of agricultural cropping were selected for intensive 
study. The Elk Island and Seven Sisters units represented 
segments of river bottom influenced by the lower Yellowstone 
Project downstream from Intake (Fig. 1); the Intake unit 
represented bottomlands above Intake where livestock production 
and dryland farming are the principal land uses. Intensive 
effort was extended to two additional areas in 1984, primarily to 
further quantify factors influencing deer distribution (Compton 
1986b). These were near the mouth of Burns Creek and immediately 
below Glendive (Fig. 1). Size, amount of riparian cover, and 
relative agricultural usage of intensive study areas are 
summarized in Table 3. 



The Elk Island unit is located in Richland County 
approximately 26 km downstream from the Intake diversion dam 
(Fig. 1). It includes the MDFWP-owned Elk Island Wildlife 
Management Area (EIWMA) and surrounding private lands northeast 
of the community of Savage. The EIWMA also includes some 
agricultural land. The floodplain is bounded on the northwest by 
a broad terrace that is also irrigated from the Lower Yellowstone 
Project, and on the southeast by steep bluffs and badlands. 
Principal crops include sugar beets, corn, small grains, pinto 
beans and hay. A portion of the area is grazed by sheep. The 



10 



Table 3. Total area, riparian cover (forest and shrubland), 
agricultural croplands, and grazing intensity in each of 5 study 
units on the lower Yellowstone River. 



Study 


To t a 1 


Ri pa r ia n Co ve r 


Agriculture 


Grazed 3 




Un i t 


Ar ea 


Area 


Ar ea 


Ar ea 




Seven Sis. 


23. 92 


9.47 (0.40) b 


5.82 (0.24) 


8.16 (0. 


34) 


Elk Isl. 


24.30 


8.69 (0.36) 


9.23 (0.38) 


8.45 (0. 


35 ) 


Burns Cr . 


1 3. 58 


4.93 (0.36) 


3.93 (0.29) 


7.23 (0. 


53) 


Intake 


36.21 


8.61 (0.24) 


7.87 (0.22) 


32.45 (0. 


90) 


Gl end ive 


7. 43 


2.25 (0.30) 


2.14 (0.29) 


4.44 (0. 


60) 



a The portion of each study unit that is grazed by livestock 
may include either or both the agricultural and riparian 
compo nen t s . 



Areas are km (proportion of total area). 



only floodplain area on the southeast side of the river in the 
unit occurs where Smith Creek enters the Yellowstone. This area 
includes irrigated hay meadows and is grazed by cattle during 
fall, winter, and spring. A relatively small proportion (35%) of 
the entire unit is actually grazed. 



The Seven Sisters unit that includes some floodplain and 
Seven Sisters Island lies west of the small community of Crane 
approximately 39 km downstream from the Intake diversion dam. 
The unit includes the Seven Sisters Wildlife Management Area 
(SSWMA) and surrounding private lands. Cropping practices are 
similar to Elk Island in that floodplain northwest of the river 
is flood irrigated from the Lower Yellowstone Project. The 
southeast side of the river also includes floodplain on which 
livestock grazing and hay production are the principal land uses. 
As at Elk Island, livestock grazing, which affects 34% of the 
area is a minor land use overall (Table 3). 

The Intake unit lies immediately upstream from the diversion 
dam in Dawson County (Fig. 1). The area is almost exclusively in 
private ownership with the exception of one island administered 
by the Bureau of Land Management (USDI). Nearly all of the unit 
is grazed by livestock (Table 3). Floodplain areas are grazed by 
cattle during fall and winter. Irrigated hay crops including 
alfalfa are also grown on the floodplain. Small grains including 
winter wheat and barley are grown under dryland farming practices 
on the floodplain and on adjacent terraces. The southeast side 
of the river floodplain is bounded by a terrace and rolling 
uplands dissected by several hardwood draws. Badlands occur on 
the northwest side of the floodplain. 



11 



The Burns Creek unit (Fig. 1) lies approximately 16 km 
downstream from Intake. Although irrigated crops were grown on 
more than one-third of the unit, over one-half of the area was 
grazed by livestock (Table 3). Bottomlands were grazed primarily 
during late fall and winter. 

The Glendive unit abutted the southern boundary of the study 
area approximately 21 km upstream from Intake. Sixty percent of 
the unit was grazed by livestock with most cultivated and 
irrigated lands used for production of hay crops. 



METHODS 

A total of 426 deer was captured and marked on the Elk 
Island, Seven Sisters, and Intake units during 1980-86; 39 deer 
were captured and marked at Burns Creek and Glendive during 1985. 
Most were captured from late December through early April with 
Clover traps (Clover 1954) or cannon nets (Hawkins et al. 1968). 
Six were taken with a hand-held net gun (Barrett et al. 1982) on 
the Intake unit during February 1984. We assume that trap 
placement using a combination of techniques over time provided 
representative samples of the population. 

Captured deer were manually restrained. Sex, age determined 
by tooth replacement and wear (Sever inghaus 1949), and general 
condition were noted for each. During 1984-86, a blood sample 
was taken and, when practicable, heart girth was measured. A 
numbered metal ear tag was attached to each ear of all deer, 141 
were equipped with transmitter collars, and the remainder were 
marked with 10-cm-wide individually recognizable vinyl neckbands. 

Relocation of radio-collared deer employed a PA-1 8 Super Cub 
aircraft. Each rad io- co 1 la red deer was relocated 2-3 times per 
month during Jun e -Nov e mbe r and 1-2 times per month during 
December-May. A visual search was made for each animal. The 
animal's location, cover type occupied, and time of day to the 
nearest 15 min. were recorded. Deer with individually 
recognizable neckbands were observed incidental to relocating 
radio-collared deer, during population census and herd 
composition surveys, or during other reconnaissance. The number 
of fawns observed at-heel with marked females was recorded during 
June-October. A sample of 368 marked deer provided 5,877 
relocations during diurnal and crepuscular periods from March 
1980 through May 1986. 

During summer 1982 and 1983 and winter 1984, hourly 
relocations of radio-collared deer were also obtained throughout 
the diel period at Elk Island and Intake by tr iangulat ion (Heezen 
and Tester 1967). This employed stationary or truck-mounted 
antenna arrays as described by Herriges (1986). A total of 5,648 
usable locations were obtained. 



12 



Population characteristics were determined from aerial 
surveys designed for census following rationale and procedures 
described by Mackie et a 1. (1981) and Rice and Harder (1 977). 
Basic surveys were conducted from a fixed-wing PA-1 8 Super Cub 
during October, late De cember-early January, and late March-early 
April of each year. All employed the same pilot and observer and 
were flown in a manner to provide complete coverage in areas 
within predictable survey units. Autumn surveys consisted of a 
single flight over study units where marked deer were present, 
except during 1984 and 1985 when the entire river bottom study 
area was flown. Early winter surveys consisted of a single 
flight covering the entire river bottom during each of the 6 
years. Early spring surveys consisted of 2-3 replicate counts of 
study units where marked deer were available. In addition, the 
Elk Island and Burns Creek study units were surveyed with a Bell 
47G helicopter during February 1984-86 and February 1985, 
respectively. Numbers, sex and age when practicable, and the 
number and identification of marked deer observed were recorded 
and locations plotted on aerial photographs. 

From aerial surveys during autumn and early winter, deer 
were classified as fawn ( <1 2 months of age), adult female, 
yearling (12-24 months of age) male and adult male. Yearling and 
older males were generally differentiated by gross differences in 
antler size. No attempt was made to differentiate yearling and 
older females. Abundance of females older than fawns and males 
older than yearlings in autumn populations was estimated from 
samples of hunter-killed and live-trapped deer by assigning 
individuals to their respective cohorts for each year that they 
occurred in the population. We assumed the 2 sexes of fawns to 
occur in equal proportion during autumn. Deer within individual 
cohorts were assigned to successive age classes on 1 June. 
During spring aerial surveys, individual deer were classified as 
adult or fawn. 

Annual population estimates for the river bottom study area 
were developed from counts during early winter surveys. 
Particular emphasis was given to data for 1981-82, 1982-83, and 
1985-86, years of optimal observability and numbers and 
distribution of individually marked deer. For each year, survey 
efficiency was calculated as the mean proportion of individually 
marked deer observed among intensive study units and applied to 
the base count derived from an initial population estimate. 
Population estimates for other seasons and years were 
extrapolated both forward and backward in time from estimates of 
mortality and recruitment using change- i n- ra t io procedures 
(Downing 1980), documented mortality of marked deer, and periodic 
aerial surveys. This resulted in 3 estimates for early autumn, 
early winter, and spring of each year that were then averaged. 

Base population estimates for the Elk Island and Intake 
units were calculated using a Lincoln index (Davis and Winstead 
1980) with an average of 12% and 6% of the respective populations 
marked during fixed-wing aerial surveys in March and April, 1981- 
86 and for the Burns Creek unit during February and March 1985. 



13 



Estimates of numbers and variance from 2-3 flights over each area 
followed the multiple sampling technique of Rice and Harder 
(1977). Lincoln index estimates were also made for the Elk 
Island unit from a single helicopter flight during February in 
each of the years, 1984-86, and in the Burns Creek unit in 1985. 
This allowed comparison of population trend and dynamics in areas 
of contrastingly different land use practices. 

Biological material was taken from deer carcasses obtained 
from several sources: special collections, depredation kill 
permits, highway mortalities, trap mortalities, hunter harvest, 
and occasionally, various other mortalities. Sex, age, location, 
and cause of death were noted for all carcasses. Ages were 
assigned to all animals by tooth replacement and wear 
(Se ve r inghaus 1949), and a middle incisor was extracted from 
those 2 years and older for cementum analysis (Gilbert 1966). 

Mean whole weights of white-tailed deer from the lower 
Yellowstone River were compared during autumn, winter, and late 
winter/early spring. Whole weights for autumn were estimated 
from f iel d- d re s s e d weights of 129 deer handled at checking 
stations and in the field during 1980-85 from a regression of 
whole to dressed weights of 26 individual deer. Weights for 
early winter and late winter/early spring were from heart girth 
measurements (Smart et al. 1973) of 138 live-trapped deer during 
winters of 1983-84, 1984-85, and 1985-86. 

Kidneys and attached perirenal fat were taken from carcasses 
that were not badly decomposed. The kidney fat index (KFI) was 
used to evaluate trends in accumulation and mobilization of body 
fat from 153 deer examined from 1980 to 1986. Procedures for 
determining the KFI followed Monson et al. (1 974). A factorial 
ANOVA was used to determine the effects of season, sex, age, and 
study unit (Elk Island and Intake) and the various interactions 
on the KFI. Summer data were analyzed in a separate ANOVA to 
compensate for the absence of data for fawns during this season. 
A least significant difference procedure was used to determine 
individual differences. 



Diastemal length and antler main beam length were recorded 
for yearling males examined during autumn hunting seasons and 
used as indexes of condition (French et al. 1 956, Reimers 1 972). 
Dressed weights were obtained from hunter-killed deer examined at 
check stations, while both whole and f ie 1 d- d re s sed weights were 
obtained from 26 deer during autumn, winter and spring. 

Breeding and fawning dates were estimated from assigned ages 
of fetuses from 38 females necropsied from January through April. 
Fetal ages assigned from forehead- rump measurements, based on 
growth rates of known age fetuses (Cheatum and Morton 1946), were 
inversely related to the Julian date of collection. This bias 
was also reported by Jackson and Hesselton (1973). We therefore 
considered fetal weights in addition to f orehead- r ump length to 
assign fetal ages following Armstrong (1950). Only the largest 
fetus in litters was used to assign an age to minimize 



14 



differences attributable to litter size. The midpoint in ranges 
of fetal ages was used to assign ages to individual litters. 

The incidence of ovulation (Cheatum 1949) was used to 
estimate the reproductive potential of the lower Yellowstone 
population, as also used by Te e r et a 1. (1 965) and others. 
Because most females had presumably completed their breeding 
cycle by late December, corpora lutea of pregnancy (CLP) were 
used to estimate ovulation rates. Reproductive tracts were not 
examined until mid-January. Corpora albicantia were not used 
because they offer a more subjective estimation of ovulation than 
the CLP (Golley 1957). Fertilization rates were determined from 
the ratio of viable embryos to CLP. Pregnancy was determined 
from 77 necropsied females during all years and by assay of sera 
from peripheral blood (Wood et al. 1986a) among 113 females 
captured during 1984-86. Ca p ture- rel at ed stress presumably had 
no raeasureable affect on pregnancy diagnosis (Wood et al. 1986b). 



Postpartum fawn production was determined from observed 
f a wn s-a t-hee 1 during July-October, Age specific reproductive 
success, or recruitment to early autumn, was evaluated from 
f a wn s-a t-hee 1 among individually marked females of known or 
assigned ages following Mund inger (1981). During 1 980-86, 155 
individually marked females provided 269 cases of estimated 
reproductive performance. The sample included 75 (48%) females 
that were fawns or yearlings at initial capture. Assignment to 
these age classes on the basis of tooth replacement allowed for 
minimal subjectivity. Of 80 females, whose ages were assigned on 
the basis of wear of cheek teeth, assigned ages of 19 individuals 
were corrected or confirmed by cementum analysis upon death of 
the animal. Because capture and marking have no known effect on 
reproductive success of females (Hamlin et al. 1982), age 
specific rates of reproduction were considered representative of 
individual female cohorts within a given year. 



Annual and seasonal mortality rates were based on 
differences between population estimates for the respective 
periods. Summer fawn mortality was interpreted as the percent 
difference between fetal rates and autumn fawn:adult female 
ratios. Seasonal and annual rates of mortality were calculated 
using exponential rates of population change and were reported as 
finite rates (Caughley 1977). Although an estimate of total 
annual mortality was calculated for fawns in this manner, annual 
turnover in the population only considered mortality of 
individuals older than 4 months of age. Factors influencing 
population dynamics were determined by correlation analysis using 
exponential rates of change. 

Distribution of deer in the study area was analyzed by 
multiple regression by dividing the river bottom into 1.6 km 
segments as described by Compton et al. (1 988). Habitat 
parameters including the amount of riparian cover and the amount 
of island area were the independent variables. The dependent 
variable was the number of deer observed per segment during early 
winter (December-January) aerial surveys from 1981-82 to 1985-86. 



15 



All locations of marked and unmarked white-tailed deer were 
plotted on 1:24,000 topographic quad maps and were assigned 
Universal Tranverse Mercator (UTM) coordinates. Annual and 
seasonal home ranges as expressed by minimum convex polygons 
(Mohr 1947) and average activity radii (AAR, Hayne 1949) were 
calculated using the TELDAY software (T. N. Lonner and D. E. 
Burkhalter, Users manual for the computer program TELDAY, Mont. 
Dep. Fish, Wildl., and Parks, unpubl.). 

Monthly movement patterns of deer on the river bottom were 
evaluated from 1,737 individual monthly AARs from 189 radioed and 
neckbanded deer observed 2-4 times in at least one month from 
June 1980 through May 1985 following procedures described by 
Dusek and Wood (1986). Differences in monthly AARs by month, 
sex, and age were determined from ANO VA. 

Diurnal home ranges were calculated from both fixed-wing 
aerial telemetry surveys and from tr iangulat i on. Annual home 
ranges included the area normally used by an animal from 1 June 
through 31 May. Yearlong and/or seasonal diurnal home ranges 
were computed for 92 individual rad io- c o 1 1 a r ed deer from June 
1980 through April 1985. The nonparamet ric Kruska 1-Wallis test 
was used to test for differences between sex and age groups based 
on AAR. Differences between groups were determined from LSD 
tests on the ranks of AARs (Conover and Iman 1981). Estimates of 
total area (polygon) were only reported. Data for individual 
deer, during years of dispersal, or periods of temporary movement 
„ outside of normal home ranges, were excluded from the analysis. 
Both diel and diurnal home ranges were similarly calculated from 
houly relocations obtained by t riangulat ion. A detailed analysis 
was described by Herriges (1986). 

Dispersal (Bunnell and Harestad 1983) direction was 
evaluated using a Ch i- square goodness of fit test for circular 
distributions . 



General trends in habitat selection or avoidance followed a 
use versus availability procedure (Byers et al. 1984) using 4,579 
point locations of rad io- col la red deer pooled from all 
intentensive study units. The amount of each cover type was also 
pooled for all units. All riparian forest communities, 
regardless of serai stage, were combined in the analysis. 
Availability of habitat components was determined from maps 
developed from o r t hopho t oquad s and color aerial photographs. 

A grid cell analysis following Porter and Church (1987) was 
used to determine which habitats were important to deer and the 
effects of habitat diversity and interspe rsion on habitat 
selection from 4,932 relocations of individually marked deer 
older than fawns from the Seven Sisters, Elk Island, and Intake 
areas. The 3 combined areas included 858 computer-generated 9-ha 
cells with dimensions of 300 m per side. The amount of each 
cover type, number of cover patches, number of cover types, and 
the presence or absence of domestic livestock and slough areas 



16 



were recorded for each cell- Use versus availability procedures 
as described above, were used to determine factors influencing 
habitat selection by deer. The n onpa ra m e t r i c Kruskal-Wa His and 
multiple comparison tests were used to determine differences in 
association with habitat complexes by season, and by sex and age 
Cells with no observed use during diurnal periods, 1-2 
observations, and 5+ observations were used to contrast habitat 
complexes with no observed use with those receiving occasional 
use and those used intensively. Response to grazing treatments 
by sex and age of deer were determined from 2 -way classification 
chi-square tests (Everitt 1977). Similarly, these techniques 
were used to evaluate the importance of slough areas and serai 
communities of riparian forest to habitat selection by deer. 

Habitat selection and use during diurnal and nocturnal 
periods of summer and winter at Elk Island and Intake were 
evaluated from relocation data for radio-collared deer obtained 
by triangulation. Location data were intersected with digitized 
habitat files using GEOCALC (W. Ho s kins , computer software 
written for the Interagency Grizzly Bear Study, Bozeman, MT ). 
The program measured area and length of edge of each cover type 
within a scanning circle of 125 m radius around each location. 
All serai stages of riparian forest and crop types were 
differentiated for this analysis. Random or grid points and 
scanning circles were generated within each unit to test 
hypotheses of use versus availability. These procedures are 
described in greater detail elsewhere (Herriges 1986). 



Approximately 1 liter of the contents of the rumen of 184 
deer that died from all causes except trapping from April 1980 
through March 1986 was collected for analysis of food habits 
(Wilkins 1957). Data were analyzed by month, season, land use 
patterns, and by sex and age. 

Statistical procedures were consistent with Za r (1984) 
unless stated otherwise. Analyses were conducted using the 
Statistical Analysis System (Ray 1982) and Montana State 
University computing services. 



POPULATION CHARACTERISTICS 



Historical Abundance and Trend, 1804-1950 

The journals of early visitors to the lower Yellowstone and 
other accounts suggest that white-tailed deer were less abundant 
historically than at the present time. White-tailed deer were 
commonly observed along the Missouri River and the lower portions 
of principal tributaries, including the mouth of the Yellowstone 
River, during the expedition of Lewis and Clark in 1804-1806 
(Koch 1946, Burroughs 1961). The occurrence of white-tailed deer 
was not documented during W. Clark's trip down the Yellowstone 
River, although accounts of numerous sightings of bison ( Bison 
bison ) and elk ( Ce r vus canadensis ) were reported (Brown 1961). 

Settlement of the northern Great Plains in the 1860's and 
1870's indirectly led to near extirpation of big game populations 
by the 1880's as a result of subsistence and market hunting 
(Petersen 1984). An occasional sighting of a whitetail, or a 
hunter taking an exceptionally large buck, on the mainstem or 
tributaries of the lower Yellowstone around 1880 were reported by 
Brown and Felton (1955). Yet, more numerously reported sightings 
of other big game suggested that white-tailed deer were still 
relatively scarce. Mule deer were the only big game animals 
observed on riparian bottomlands of the Yellowstone above the 
mouth of the Bighorn during the late 1940's (U.S. Fish and 
Wildlife Service 1952), and Allen (1971) indicated that the 
relatively limited distribution of white-tailed deer in eastern 
Montana included the extreme lower Yellowstone River in about 
1941. 



Recent Population Trends 

Whitetails increased in both number and distribution since 
the early 1940's and occupied several habitats in eastern Montana 
by the 1950's and 1 960*s (Allen 1971). Lacking accurate census 
data for white-tailed deer for years prior to the study, 
estimates of relative abundance and trend for the lower 
Yellowstone are highly subjective. The best available indicator 
may be provided by harvest and recruitment trends (Figs. 4 and 
5). 

Although harvest trends may not be directly correlated with 
population size, we believe they reflect gross temporal changes 
in abundance. For example, total harvest of white-tailed deer in 
administrative region 7, encompassing 83,200 km^ of southeastern 
Montana, increased from 1,000-3,000 per year in the mid to late 
1950's to 5,000-2 1,000 in the early-to-mid 1980's. Sharp 
increases and subsequent peaks in deer harvests occurred in the 
early to mid 1970's (11,000) and again in the early to mid 1980's 
(21,000) and were attributed to liberalized antlerless hunting 
regulations associated with periods of "peak" abundance. 
Antlerless regulations have a significant impact on annual 
harvests (J. E. Swenson. 197 8. Montana Dep. Fish, Wildl., and 



18 



q 6000 
m 

07 5000 
UJ 
> 
DC 
< 



3 
Z 



4000- 



3000 H 
g 2000 



1000 - 



0 



75 



Southeastern Montana 
Lower Yellowstone Basin 
Study Area 



s X. 



s. 



— r - 
76 




83 84 Si 



Figure 4. Annual harvests of antlered white-tailed deer in 

southeastern Montana as reported by the MDFWP (Fed. Aid Proj. W 

130 -R), 1976-85, and in the lower Yellowstone River study area, 
1980-85. 



Parks, Fed. Aid Proj. W-130-R, unpubl. rep.). 

Harvests in the portion of the region comprising the lower 
Yellowstone Basin and the divide separating the Missouri and 
Yellowstone drainages (12,300 km^) s including the study area, 
followed a similar pattern. The lower Yellowstone Basin 
accounted for about half of the regional white-tailed deer 
harvest from 1 960 to 1 985, while the study area (< 2% of the 
basin) accounted for approximately 20%. 

Trends in harvest of antlered white-tailed deer suggest that 
the population on the lower Yellowstone River bottom experienced 
the most recent low point in numbers in the late 1970's (Fig. 4) 
just prior to initiation of this study. Harvests in the lower 
Yellowstone Basin and throughout southeastern Montana were lowest 
during 1979 and highest during 1983. Harvest regulations for 
antlered whitetails were unchanged during the period of 1976-85, 
while season lengths and numbers of permits for antlerless deer 
varied and increased substantially from 1983 to 1985. 



Exceptionally high rates of fawn recruitment during 1978 and 
1979 (Fig. 5) led to increasing populations throughout the 
region- Although recruitment declined after 1979, populations 
continued to increase until 1983. 



19 



V) 

z 
< 

LL 



60 n 



55- 



50 



45- 



40- 



LU 

O 35 
DC 
LU 

CL 30 - 
25 
20 



Southeastern Montana 
Lower Yellowstone Basi 
Study Area 



If 
f 



* *•. 
\ 

\ 

\ 



// 

•7 
j 




74 



75 76 



— r 
77 



— I 1 — 

78 79 
YEAR 



i 1 1 1 1 

81 82 83 84 85 



Figure 5. The proportion of fawns in autumn whitetail 
populations in southeastern Montana as reported by MDFWP (Fed. 
Aid Proj. W-130-R), 1976-85, and in the lower Yellowstone River 
study area, 1 980-85. 



Deer numbers in eastern Montana during 1982-84 may have 
represented an all-time peak in distribution and abundance for 
the species in this region (R. J. Mackie et al. 1 985. Montana 
Dep. Fish, Wildl. and Parks, Fed. Aid Proj. W-120-R, unpubl. 
rep.). Some comparative data also suggested that population size 
on the lower Yellowstone was relatively high during this study. 
For example, Swenson (1979) reported a late summer density of 
approximately 16 deer/km^ in the Intake area during 1977. An 
epizootic of hemorrhagic disease (EHD) reduced that population by 
about 33% by autumn 1977. We estimated early autumn densities of 
18-20 deer/km for approximately the same area in 1980 and 50 
deer/km^ in September-October 1983 when population growth had 
leveled off. 

Total deer numbers on the river bottom study area in autumn 
increased from an estimated 3,100 individuals during 1980 to 
5,784 in 1983, then declined to 4,336 by 1986 (Fig. 6). Early 
winter and spring numbers peaked during 1982-83. Finite rates of 
change from year-to-year in autumn populations were 4 2, 25, 4, - 
5, -13, and -9% for individual years 1980-86, respectively. 

Trends in numbers of deer by sex and age (Fig. 7) appeared 
to reflect declining reproductive success and, in turn, numbers 
of individuals recruited into successive age classes from 1982 



20 



LU 

n 



O 3000 - 



g 

8— 
< 



CL 

o 

CL 



6000 n 



5000 - 



2000) 




FALL 



— --EARLY WINTER 
SPRING 



1 i — i 1 r~ 

80-81 81-82 82-83 83-84 84-85 

BIOLOGICAL YEAR 



85-88 



86-87 



Figure 6. Population trend of white-tailed deer on the lower 
Yellowstone River, 1 980-86. 



through 1985. There was a one-year time lag between peak numbers 
of fawns in autumn populations and peak population numbers. The 
number of fawns in autumn populations (1,343 in 1980) peaked 
during 1982 at 2,077 individuals and declined to 1,561 
individuals by 1 986 (Fig. 7). Numbers of yearlings peaked in 
1983 at 1,551. Numbers of adult (2 + ) females (798 in 1980) 
increased through 1983 (1,964), leveled off through 1984, and 
declined through 1986 (1,631). Adult males increased from 1980 
to 1982, leveled off through 1984, and increased slightly in 
1985, although the total antlered segment that included yearlings 
declined from 1,263 individuals in 1983 to 767 in 1986. 



Spatial Variation in Trend on the River Bottom 

Population trend over time differed between segments of the 
river bottom as exemplified by data for the Elk Island and Intake 
areas (Table 4). Populations were similar during spring 1981 
with a between-area difference of less than 100 individuals. 
Numbers of deer during spring increased from 1981 to 1984 and 
declined through 1986 at Elk Island. At Intake, spring 
population size increased relatively rapidly to 1983, levelled 
off to 1984, and declined sharply from 1984 to 1986. 



21 



8 n 



DC 
LU 
UJ 
Q 

LL 
O 

CO 
DC 
LU 
CD 



1 5 : : 
o T 




■ FAWNS 
° YEARLINGS 
o ADULT FEMALE 
• ADULT MALE 



T 



1980 1981 



1 ! 

1982 1983 
YEAR 



1984 1985 1986 



Figure 7. Numbers and trend of white-tailed deer on the lower 
Yellowstone River by age class, 1980-86. 



Population estimates of 654 +108 (X +SE) and 600 +96 animals 
from helicopter surveys at Elk Island in February of 1984 and 
1985, respectively, were consistent with estimates from multiple 
fixed wing surveys in March and April of those years (Table 4). 
Estimates of 175 +50 and 173 +3 5 deer from one helicopter survey 
and 2 fixed-wing surveys in the Burns Creek area during February 
and March 1985, respectively, also indicated consistency between 
a single helicopter survey and multiple fixed-wing surveys. The 
differences between estimates for the helicopter survey during 
February 1 986 (4 34 +7 3) and the mean of 3 fixed-wing surveys 
during March-April 1986 (332 +84) at Elk Island apparently 
resulted from redistribution of deer from flooding during an ice 
jam in late February. Two of 16 trans mittered deer had not 
returned to that area by the time of the early spring fixed-wing 
census . 



Population density was comparatively lower at Burns Creek 
(12.7/mi 2 than at Elk Island (25.0/mi 2 ) or Intake (19.9/mi 2 ). 
The Burns Creek, Elk Island, and Intake units accounted for 9%, 
10% and 14%, respectively, of the total main channel length in 
the study area. Burns Creek accounted for 5% of total spring 
numbers in 1985. Numbers of deer at Elk Island and Intake during 
spring for the years 1981-86 averaged 15% and 18%, respectively, 
of total numbers in the study area. 



22 



Table 4. Population estimates for white-tailed deer in the Elk 
Island and Intake units from spring 1981-86 using multiple 
Lincoln estimates and autumn estimates from recruitment of fawns 
over summer, 



Da t e 






Elk Island 






Intake 




Ad ults 


Fa wn s 


To ta 1 


Adults 


Fa wn s 


To t a 1 


Mar/Apr 


'8 1 


19 8 


1 90 (4 9% ) a 


3 8 8 


2 5 5 


209 (4 5% ) 


4 64 


Se p/Oc t 


• 81 


3 88 


2 59 (4 0%) 


647 


464 


3 09 (4 0%) 


773 


Ma r / Ap r 


'82 


26 9 


1 79 (40% ) 


44 8 


404 


2 92 (4 2% ) 


69 6 


Sep/Oct 


' 82 


4 48 


2 99 (4 0% ) 


747 


6 96 


375 (3 5%) 


1071 


Ma r / a p r 


«83 


312 


2 66 (4 6% ) 


57 8 


566 


3 62 (3 9% ) 


928 


Se p/Oc t 


' 83 


5 78 


2 85 (3 3%) 


863 


928 


417(3 1%) 


1345 


Ma r / Ap r 


'84 


43 5 


245 (36% ) 


68 0 


63 8 


27 3 (3 0% ) 


911 


Se p/Oc t 


'84 


680 


366 (3 5%) 


1046 


911 


354 (2 8%) 


1265 


Ma r / Ap r 


•85 


356 


1 83 (34%) 


539 


354 


183 (34% ) 


537 


Se p/Oc t 


'85 


539 


3 30(38%) 


869 


537 


315(3 7%) 


852 


Ma r / a p r 


'86 


23 6 


9 6 (2 9% ) 


33 2 


23 7 


102 (3 0%) 


339 



Percent of total population. 



At Elk Island, autumn densities were 26.6, 30. 7, 35. 5, 43.0, 

and 35.8 deer/km 2 during 1981-85, respectively. This compared to 

2 8. 6, 3 9. 6, 4 9.7 , 4 6. 8, and 31.5 deer/km 2 on the floodplain at 
Intake during the respective years. 

Greater between-year increases, and consequently higher 
animal density, at Intake than at Elk Island from 1982 to 1983 
probably reflected differences in accessibility of the two areas 
to hunters. Landowners at Intake allowed only limited access to 
that area until 1984 when more hunting was permitted because of 
depredation by deer on alfalfa. At Elk Island, the EIWMA 
provided public hunting during all years, reducing the influence 
of private landowners on hunter access that apparently minimized 
year-to-year variation in a cc e s s ab i 1 i t y as well as population 
density . 

Population Structure 

General Patterns .-~Ag e structure was pyramidal, comprised 
largely of younger age classes in both the female and male 
segments of the population (Fig. 8). An average of 75% (68-79%) 
of total autumn populations during 1980-85 was 2 years or 
younger. These age classes included 63% of the females and 93% 



23 




t 1 1 1 1 r — i— — r 1 1 

25 20 15 10 5 0 5 10 15 20 25 



I i 1 r 



25 20 15 10 5 0 5 1015 20 25 





8+ 




7 


(f) 


6 


(f) 


5 




4. 


o 


3 


HI 


2 


o 




< 


1 




0 



1982 ™ 
■ 


i 










i — i — i — i — i — 


" 1 

1 1 T "I 1 




25 20 15 10 5 0 5 1015 20 25 



25 20 15 10 5 0 5 10 15 20 25 




25 20 15 10 5 0 5 10 15 20 25 
% OF POPULATION 




25 20 15 10 5 0 5 1015 20 25 
% OF POPULATION 



Figure 8. Composition (% ) of the 
Yellowstone River during autumn, 



whi t e ta i 1 
1 980-85. 



population on lower 



f 



24 



of the males. Females 3-7 years old comprised 32% of the female 
segment, and those 8 years and older 5 % . Males older than 4 
years comprised less than 1% of the male segment of the 
population. 

Yearling and older females accounted for an average of 44% 
(39-50%) of pre-hunt populations of early autumn (Table 5). 
Relative abundance of females in the various cohorts (Fig. 8) 
suggested that the 1977 cohort (3-year-olds in 1980) was weaker 
than those of preceding or following years. Conversely, the 1979 
cohort (yearlings in 1980) was comparatively strong. 



Yearling and older males accounted for an average of 20% 
(16-2 3%) of pre-hunt populations (Table 5). Their relative 
abundance declined sharply from autumn when buck:doe ratios 
averaged 4 5:100 (range 37-58:100) to winter when the mean was 
2 5:1 00 (range 18-37: 100). 

Fawns comprised 32-43% (X = 37%) of autumn populations on 
the study area during 1980-86 (Table 5). The average autumn 
fawn:doe ratio was 85:100 and ranged from 64 to 112:100 (Table 
5). The relative abundance of fawns declined from 1980 to 1983 
and leveled off in 1984, though absolute numbers of fawns 
added/year increased through 1982 (Fig. 7). Fawns were 
relatively more abundant in the lower Yellowstone whitetail 
population than in nearby prairie habitat during the same time 
period (Wood 1 987). 



An increase in the proportion of fawns in classifications 
from autumn to winter during all years except 1984 and 1985 
(Table 5) probably reflected disproportionately higher mortality 
among adult deer during autumn- Trends to spring were confounded 
by difficulty in distinguishing fawns from adults during late 
winter and spring. During 1984-85, estimated weights of male 
fawns and yearling females, trapped during late winter and early 
spring, overlapped. However, data for both early winter and 
spring indicated a steadily declining trend in recruitment of 
young animals into the population from 1980-81 to 1984-85. 

Years of high relative abundance of fawns throughout the 
river bottom coincided with years of low total population 
numbers. The lower fawn: doe ratios in 1 98 3 and 1 98 4 (Table 5) 
coincided with years of relatively high population numbers (Fig. 
6), whereas the high ratio in 1980 coincided with relatively low 
numbers of deer. Fawn:adult female ratios of 68:100 and 140:100 
were reported for autumn 1977 and 1979, respectively (J. E. 
Swenson 1 97 8, 1 980, Mont. Dep. Fish, Wildl. and Parks, Pro j. W- 
130 -R, unpubl. rep). The low recruitment rate for 1977 may have 
been influenced by the EHD epizootic (Swenson 1979); that of 
1979, which was higher than any recorded during this study, 
coincided with the recent low point in deer abundance. 



Table 5. Seasonal 
tailed deer on the 



herd composition by sex and age of white- 
lower Yellowstone River, 1980-86. 



Number % % % % Fawns: Fawns: Bucks: 

Classi- Males Hales Fem. Fawns 100 Fern. 100 Ads. 100 
Year fied 2 + yrl. 1+ 1+ 1+ Does 



Autumn (Pre-hunt) 



1 98 0 


540 


6 


1 2 


39 


43 


112 


76 


46 


1981 


334 


5 


1 1 


43 


41 


96 


69 


39 


1982 


664 


8 


1 5 


4 0 


37 


95 


60 


58 


1983 


3 16 


8 


14 


46 


32 


68 


46 


47 


1 98 4 


484 


. 8 


10 


50 


32 


64 


47 


37 


1985 


485 


1 2 


1 0 


45 


33 


75 


50 


50 


1 986 


452 


10 


8 


45 


37 


82 


59 


39 



Early Winter 



1980-8 1 


687 


5 


1 0 


40 


45 


112 


81 


37 


1 981 -82 


1581 


3 


7 


46 


45 


99 


82 


20 


1982-83 


108 9 


2 


6 


47 


44 


92 


78 


18 


1 983-84 


1562 


3 


6 


52 


39 


75 


63 


18 


1984-85 


84 9 


4 


9 


54 


33 


60 


48 


25 


1 985-86 


531 


6 


10 


53 


31 


58 


44 


31 



Early Sp ring 



% Adults 

1 981 4 25 5 5 4 5 81 

1982 490 59 41 68 

1 983 5 63 5 8 4 2 7 3 

1984 803 67 33 48 

1985 490 73 27 37 

1986 414 71 29 42 



Longevity of females in the lower Yellowstone population was 
approximately twice that of males (Fig. 8). The oldest ages 
assigned to females were 13-15 years; the oldest male was 7. 
Expected ecological longevity of both sexes of deer, determined 
by wear on the cheek teeth, is about 13 years (McCullough 1979). 
Greater longevity of females than males may be related to 
physiological stress imposed on males by breeding (Woolf and 
Harder 1979, Gavin et al. 1984) and/or other factors favoring 
female survival such as hunter selectivity (Roseberry and 
Klimstra 1 974). 



2 6 



Spatial Variation . — Fawns were relatively more abundant at 
Elk Island than at Intake during autumn of all years except 1980 
(Table 6). Fawn:doe ratios at Elk Island averaged 99:100 
(CV=18%) and ranged from 7 6 to 118:100; the average for Intake 
was 78:100 (CV=23%) with a range from 61 to 115:100. Fawn 
numbers during autumn at Elk Island were less variable than at 
Intake over time, coinciding with comparatively lower and more 
stable numbers of adults at Elk Island (Table 4). Coefficients 
of variation in numbers of adults during autumn were 26% and 37% 
for the respective areas during 1981-86. For estimated numbers 
of fawns during the same years, the coefficient of variation was 
20% at Elk island and 25% at Intake. 



Buck:doe ratios were more variable than fawn:doe ratios. 
The respective coefficients for buck :doe ratios were 35% and 32%. 
Some of the variability in buck:doe ratios probably resulted from 
differential observability between adult males and females 
(Downing et a 1 . 19 7 7). 



Adult males were relatively, though not necessarily 
numerically, more abundant at Elk Island than at Intake (Table 
6). Average buck:doe ratios were 53:100 (range 26-73:100) and 
3 8:100 (range 20-49:100) in the respective segments of river 
bottom. A year-to-year trend in relative abundance of antlered 
deer within each unit was not apparent perhaps because of high 
variability. Some of the difference in mean ratios between areas 
probably reflected disproportionately different rates of removal 
of bucks by hunting. 



Physical Characteristics 



W eight Re lationships . — Dressed weights of fawns during 
autumn averaged 23.5 and 24.1 kg for females and males, 
respectively. Those of yearlings were 39.6 kg for females and 
49.8 kg for males. Dressed weights increased with age to 4 5.2 kg 
for females 4 years and older and 74.8 kg for males of similar 
age . 

Field- dressed carcass weights of lower Yellowstone 
whitetails during autumn were comparable to average weights from 
hunter check stations from throughout Montana during 1948-63 
(Mackie 1964). Whitetails from the lower Yellowstone were 
slightly heavier than those reported by Wood (1987) from nearby 
prairie habitat that probably reflected differing effects of 
environmental conditions of deer in the respective habitats. 
Mean age- specific field-dressed weights of males and females from 
the lower Yellowstone were 91% and 87%, respectively, of weights 
reported for the Swan Valley of northwestern Montana (Dusek et 
al. 1987). The overall trend may have reflected genetic 
differences between the 2 populations. However, comparatively 
lighter weights among young adult females on the Yellowstone 
suggested they delayed body growth in favor of reproductive 
success as influenced by a comparatively higher nutritional plane 
during winte r. 



2 7 



Table 6. Sex and age ratios of wh i t e- t a il ed deer on the Elk 
Island and Intake units during early fall, 1980-86. 



Elk Island Intake 



Fawns: Bucks: Fawns: Bucks: 

Year N 100 Does 100 Does N 100 Does 100 Does 



1 980 


253 


118: 


100 


37: 


100 


152 


115: 


100 


38: 


100 


1981 


77 


111: 


100 


64: 


100 


164 


86: 


100 


27: 


100 


1982 


273 


11 1: 


100 


65: 


100 


217 


78: 


100 


49: 


100 


1983 


112 


74: 


100 


50: 


100 


182 


67: 


100 


47: 


100 


1984 


77 


76: 


100 


26: 


100 


120 


61: 


100 


49: 


100 


1 985 


1 79 


1 06: 


100 


73: 


100 


122 


70: 


100 


20: 


100 


1 986 


136 


9 5: 


100 


53: 


100 


121 


68: 


100 


27: 


100 



Whole weights of fawns, estimated from field-dressed weights 
during autumn were only slightly less than those of late 
winter/early spring that were determined from heart girth (Table 
7). Male fawns were slightly heavier than females from late 
autumn through early spring. 

Females older than fawns maintained their autumn weights 
through January and experienced slight losses of weight during 
February and March. Females age classes 1, 2, and 3+ lost 11, 2, 
and 9% of their respective autumn weights by late winter-early 
soring. Mean autumn weights of 2-year-old females did not depart 
appreciably from those of yearlings (Table 7), whereas females 3 
years and older were heavier than younger adult females during 
all seasons. 

Among males, whole weight increased progressively with age 
(Table 7), and this trend was apparent during all seasons. 
Weight loss from late autumn to early winter was readily apparent 
among males older than fawns. Weight losses, experienced from 
late autumn to late winter/early spring were 22, 22, and 30% for 
males of ages of 1, 2, and 3+ years, respectively. 



KFI . — The KF I was characterized (Table 8) by greater 
variability than whole carcass weights (Table 7), although, as 
with whole weights, the KFI varied seasonally by sex and age. 
ANOVA indicated differences (P_ < 0.001) among seasons, age 
classes, and season/sex interaction, but not between study units 
and other interactions (P_ > 0.10). 

Indexes were highest during autumn and/or winter and lowest 
during spring and summer coinciding with yearlong maximum and 
minimum whole weights within age classes older than fawns. 
Although a decline in the KFI among both sexes of fawns was 
apparent from winter to spring, seasonal differences were not 
significant (P_ > 0.05). That fawns mobilized body fat while 



28 



Table 7. Estimated whole weights (kg) by season, sex and age of 
white-tailed deer on the lower Yellowstone River, 1980-86. 



Sex and Late Early Late Winter/ 

Age Class Autumn Winter Early Spring 



Females: 
















0 


33 + 1 (1 4 ) a 


35 


+ 


1(18) 


35 


+ 


1(13) 


1 


53 + 1 (18 ) 


49 


+ 


1(9) 


47 


+ 


3(4) 


2 


5 1 + 2(7) 


51 


+ 


2(7) 


50 


+ 


2(7) 


3 + 


58 + 1(22) 


60 


+ 


2(7) 


53 


+ 


1(10) 


Ma 1 e s : 
















0 


34 + 1(13) 


38 


+ 


1(19) 


38 


+ 


2(19) 


1 


65 + 1 (25 ) 


57 


+ 


2(4) 


51 


+ 


2(7) 


2 


76 + 2(13) 


62 


+ 


5(2) 


59 


+ 


2(10) 


3 + 


87 + 4 (17 ) 








61 


+ 


1(2) 


a Seasonal 


mean we i gh t 


(kg) + 


SE 


(sample 


size) . 







Table 8. Mean seasonal kidney fat indexes (KFI) of white-tailed 
deer by sex and age class from the lower Yellowstone River, 1980- 
1 986. 



Sex and 

Age Class Summer Autumn Winter Spring 



Fe ma les : 


















0 






59 


+ 9(5) a 


58 


+15(7) 


34 


+11(7) 


1 


46 


+16(6) 


12 9 


+30 (5 ) 


162 


+2 9 (7 ) 


48 


+ 9(7) 


2 + 


43 


+17(10) 


99 


+14(17) 


118 


+17(17) 


53 


+10(17) 


Males: 


















0 






53 


+ 13(10) 


59 


+22(4 ) 


13 


+ 4(4) 


1 


2 9 


+1 1 (4 ) 


14 1 


+28(10) 


52 


+ 1(2) 


30 


+ 8(3) 


2 + 


1 6 


(1 ) 


1 59 


+2 9(6 ) 


68 


+12(6) 


21 


+ K2) 





















Seasonal mean + S E ( s amp 1 e size). 



29 



maintaining their weight through early spring, suggested some 
body growth through winter. 

Differences attributable to age, when fawns were excluded 
from the analysis, were not significant (P_ > 0.10). Thus, 
yearling and older deer were combined for subsequent analyses. 
The KF Is of yearling and older males were characterized by 
increases (P_ < 0.05) from summer to autumn and significant 
declines from autumn to winter. Those of similar aged females 
were also characterized by significant increases from summer to 
autumn. Females apparently continued to accumulate perirenal fat 
into winter, though differences between autumn and winter were 
not significant. A decline (P < 0.05) in the mean KF I of females 
was observed from winter to spring. 

During autumn, highest mean KF Is occurred among yearling and 
older males (147%), followed by yearling and older females (106%) 
and fawns (59 and 53% for females and males, respectively). 
Differences between adult males and all other groups were 
significant (P < 0.05) as was the difference between adult 
females and male fawns. Yearling and older females exhibited the 
highest mean KF Is during winter (130%), followed by yearling and 
older males (64%) and fawns (58 and 59% for females and males, 
respectively). Differences between adult (1+) females and all 
other groups were significant (P_ < 0.05) as was the difference 
between adult males and female fawns. 

KF Is for deer on the lower Yellowstone were comparatively 
higher than reported for other areas of eastern Montana (Allen 
1968, Dusek 1987). The annual cycle by sex and age was generally 
consistent with that reported by Kie et al. (1983) and Johns et 
al. (1984). That yearling and older female white-tailed deer on 
the lower Yellowstone continued to accumulate perirenal fat 
through winter departed from data reported for mule deer by 
Anderson et al. (1972) that suggested the KF I generally peaks 
just prior to or at the peak of breeding. 

Diastemal and M ain Ant ler Bea m Length . — Di asternal length of 
yearling males during autumn was correlated with field-dressed 
carcass weight (r = 0.49, 18 d.f, P_ < 0.05). A nonsignificant (r 
= 0.21, 14 d.f., P_ > 0.05) relationship between main antler beam 
length and carcass weight suggested that factors affecting antler 
growth differed from those influencing body size. Coefficients 
of variation between years ranged from 5 to 7% for diastemal 
length and from 18 to 41% for antler main beam length. 

Diastemal length of yearling males during autumn was 
inversely related to numbers of adult (2+) females during summer 
and positively related to precipitation during both the immediate 
preceding September, October, April, and May and the same months 
of the previous year (preceding birth of the cohort) (P < 0.01). 
These 3 factors accounted for 99% of the variation in diastemal 
length. The highest and lowest mean diastemal lengths were 
observed in 1980 and 1984, respectively (Table 9); a decline 
(P < 0.05) was observed from 1 983 to 1 984. 



30 



CD 
t-H 

to 
0 

an 
c 

•H 
r-l 

co 
oj 
>> 



oo 



CM 



+ ! 
on 



on 

m 
i-i 
+ 1 

CM 
CM 
CM 



00 
ON 



CM 



+ 1 



ON 
r-H 
+ 1 



CO 




.C 








cm 


c 




CD 


• 


rH 






00 


0 




CO 


o 


OJ 


00 


X> 


CT\ 






u 




CD 




r-l 




4J 


OJ 


C 


> 


CO 






c 






ill 


CO 


e 


0 


o 




•u 




CO 


c 


5t 


CO 


o 




I-l 


r-i 


I— 1 


CO 


OJ 


0 


>■< 


OJ 




4J 


I-l 


10 


OJ 


CO 


& 


•H 


o 


*X3 


rH 


C 


a) 


CO 




0) 


ij 


s 






c 


c 


o 


•H 






J-l 


CO 


a) 


T3 


OJ 


c 


^3 


CD 




*j 


13 


H 


OJ 




f-l 




♦H 




CO 


o> 


AJ 

! 


(D 


OJ 


i-l 


J_) 


Xi 




«3 




H 





u 

CO 
OJ 









f=3 








vO 


ON 










rH 


















lO 


vw 










l — 4 




CO 






^> 




WW 




» ( 


00 




ON 






r—1 




i— ( 




T 1 


+ 1 








r— t 


CM 










lO 










CM 


OJ 










N 
















/— N 




CO 






CO 


CO 








CM 


CM 


CD 


VN 








i— 1 


00 




r—l 




Oj 


ON 






rH 


0 


p— t 




T 1 


+ I 


CO 










CO 






o 




\ — s 








vO 










CM 


w 




















+ 1 






vO 


in 








CM 


CM 




«— 1 




— ' 


V ' 


0 


00 




r-i 


CO 


0 


ON 






«— t 




r— 1 




+ ! 


+ 1 












U 






r-t 


vO 


C 








LO 


OJ 








CM 


0 






co 




CD 










U 






00 










r-l 


r-t 


CO 








■ — «< 


CO 


O 






r-» 


CD 


00 1 






r-i 


0 


On 




+ ! 


+ 1 




i— i 








C 






r-l 


m 


CO 






r-. 


vO 


CD 








CM 


0 










rH 








0 


CO 




CO 




CO ! 


3 




0 




OJ |, 


C 




at 




PQ X 


e 




jj 




u 


<: 




CO 


em 


C 61) 






CO 


a 


tH C j 






•H 


OJ 


CO OJ 


CO 




Q 









31 



The negative effects of adult female numbers and the 
positive effects of precipitation of the immediate past autumn- 
spring period also appeared to explain 85% of the yearly 
variation in average main antler beam length (P_ = 0.06). A 
density-antler growth relationship was also reported for 
supplementally fed deer (Ozoga and Verme 1982). Average beam 
length declined (P < 0.05) from 1983 to 1984, but years of 
maximum and minimum average beam length coincided with a year of 
maximum annual precipitation (1982) and the 2nd of 2 consecutive 
years of below normal precipitation (1984), respectively. 
Yearling males with antlers consisting of single tines or 
"spikes" accounted for 20, 17, 24, 50, 50, and 54% during the 
respective years, 1980 through 1985. The influence of 
precipitation only during the immediate past autumn-spring 
suggested that antler growth was more sensitive than diastemal 
length to yearly variation in environmental conditions. The 
lower jaw is a zone of priority growth (Reimers 1972). 

The cumulative influence of precipitation received in 
successive years appeared to have a more profound influence on 
body growth than on antler growth. This was indicated by less 
variation in diastemal length between years and a positive 
correlation between diastemal length and dressed carcass weight 
of yearling males. Interaction of population size and 
precipitation may have affected growth of yearling females 
similarly, though diastemae of young females were not measured 
during this study. 

Seasonal precipitation has been cited as a key variable 
affecting annual rangeland forage production, although the 
precise relationship may vary with geographical location, plant 
species, and other interacting environmentral variables 
(Blaisdell 1958, Shiflet and Dietz 1974). Precipitation over 
consecutive years may have cumulative effects on forage 
production (Hanson et al. 1 982, Wisiol 1 984). Precipitation for 
the combined months of September, October, April, and May was 
most highly correlated with antler or diastemal length measured 
the following autumn, but coefficients were not significant (P_ > 
0.05). Precipitation during those months presumably influenced 
the availability and abundance of natural herbaceous forage 
during the early growing season, a period when this type of 
forage was used most heavily by deer. 

REPRODUCTION 

Breeding and Fawning Seasons 

Back-dated ages of fetuses from 38 females collected between 
21 January and 24 April suggested that breeding occurred between 
12 November and 28 December with a median date of 21 November. 
Thirty (79%) bred between 14 and 28 November. Median breeding 
dates for mature and yearling females were 20 and 27 November, 
respectively. The estimated date of breeding for the only fawn 



32 



among the sample of pregnant females was 26 November. Assuming a 
period of gestation of about 200 days (Severinghaus and Cheatum 
1956), parturition occurred from early June to mid July with a 
median date of about 10 June. Although the earliest observation 
of a newborn fawn with a radio-collared female was 9 June, 1982, 
fawns were not readily observed until mid-June. 



Thirty-six females (95%) presumably bred during their first 
estrus cycle as interpreted from the clustering of breeding dates 
during the last 2 weeks of November. Two mature females 
conceived from raid to late December presumably during a second 
estrus cycle (Cheatum 1 949). 

Although the peak of breeding among adult (2+) females 
varied little between years, that for 1984 (2 3 Nov.) was a few 
days later than in 1981 (19 Nov.) and 1 983 (1 5 Nov.). Median 
breeding dates for all years combined were 22 November in the 
portion of river bottom above Intake and 19 November below 
Intake. 

The breeding cycle of deer is presumed to reflect 
photoperiodism and generally progresses from November in northern 
latitudes to January or February in southern latitudes in the 
U.S. (Verme and Ullrey 1984). The mid-to-late November peak in 
breeding on the lower Yellowstone River was similar to that 
reported for New York (Cheatum and Morton 1946, Jackson and 
Hesselton 1973), and the Midwest (Roseberry and Klimstra 1970, 
Haugen 1975). 



Pregnancy Rates 

Breeding was a rare occurrence among fawns. Only 1 of 56 
fawns, for which reproductive status was determined from 
reproductive tracts (n = 17) during 1980-85 or serum assay (n=39) 
during 1984-86, was pregnant. This contrasts with findings in 
the Midwest agricultural region where a high incidence of fawn 
pregnancy was attributed to availability of high quality forage 
(Roseberry and Klimstra 1 970, Haugen 1 975). 

Sexual maturity of female fawns has been predicated on the 
basis of autumn weights (Cheatum and Severinghaus 1950). Female 
fawns of 6-7 months of age were reported to reach sexual maturity 
at whole weights of 39-41 kg (Verme and Ullrey 1984) or 31-33 kg 
(Roseberry and Klimstra 1970). Growth rates, primarily as 
influenced by energy intake, may influence physiological 
development (Abler et al. 1976). 

Whole weights of female fawns on the lower Yellowstone 
(Table 7) did not change appreciably from November at a median 
age of 5 months to December and January at median ages of 6-7 
months. By November, 11 of 13 female fawns had reached a weight 
of 31 kg or greater, and only 1 attained a weight of 39-41 kg. 
By December-January, 16 of 18 female fawns had reached the 
threshold of 31 kg, while only 2 weighed more than 39 kg. The 



33 



variation in critical weights reported in the literature and the 
negligible change in weight from autumn to winter among fawns in 
this study indicated that weight alone is not a sound criteria 
for attainment of sexual maturity among fawns. 

Mueller and Sadlier (1979) indicated that relative 
physiological state, rather than body size, may be the major 
determinant of sexual maturity among black-tailed deer (0 . h . 
co lumbianus) . They documented a leveling off of whole weights 
among fawns from November through winter as found in our study 
(Table 7). This suggested that at the onset of winter-like 
conditions from mid-November to early December, attainment of 
sexual maturity may be sacrificed in favor of redirection of 
resources to growth and maintenance. Suppression of the final 
stages of puberty and ovulation in white-tailed deer fawns in 
northern Michigan was attributed to the combined effects of 
photoperiod and winter-like weather acting on the ne urohor m ona 1 
regulatory mechanisms of puberty (Budde 1983). Because of 
latitude and prevailing environmental factors in the northern 
Great Plains, a high incidence of fawn pregnancy probably rarely 
occurs even in spite of a comparatively high nutritional plane. 

The late November breeding date of the single pregnant fawn 
in the sample suggested that successful breeding among this age 
class may be limited to either the most precocious or earliest 
born females. Some adult males began shedding antlers during mid 
December suggesting a decline in breeding activity (Robinson et 
al. 1965). However, this does not account for the low incidence 
of pregnancy among fawns, because of mid-December breeding dates 
of some older females and observed breeding behavior among males 
with antlers. 

Among 134 yearling and older females examined (60 
reproductive tracts and 74 serum samples), 93% were pregnant. 
Pregnancy rates increased with age through 3-6 years: 32 (86%) of 
37 yearling females, 31 (94%) of 33 2-year-old females, 42 (98%) 
of 43 females of 3-6 years, and 20 (95%) of 21 females 7 years 
and older were pregnant. However, differences between age 
classes for all years combined were not significant (P > 0.10). 

Annual pregnancy rates appeared to vary more among yearling 
than older females (Table 10). Coefficients of variation between 
years for yearling and adult females were 16% and 5%, 
respectively. All 31 yearling and older females examined during 
winter/early spring of 1981 through 1983 were pregnant (Table 
10). Five (19%) of 27 yearlings and 4 (5%) of 76 adults were not 
pregnant among females examined during 1984-86 when pregnancy 
rate was dependent (P < 0.05) on age. 



34 



Table 10. Annual rates of pregnancy (%) of yearling and older 

female white-tailed deer on the lower Yellowstone River. 

Yearling Females Adult Females 

Year Bred a N % Preg. N % Preg. 

1 980, 1 982 3 1 00 3 100 

1981 7 100 18 100 

1983 8 88 19 89 

1984 16 81 43 95 

1985 3 67 14 100 

Total 37 86 97 96 



a Data were collected during winter or early spring 



following breeding season 



Reproductive Potential 

Potential productivity of the lower Yellowstone population 
in terms of ovulation and fetal rates, determined from 
examination of 60 reproductive tracts from yearling and older 
females during 1980-86 (Table 11), was comparable to that 
reported for other populations in the northern Great Plains and 
Rocky Mountains (Table 12). Ovulation and fetal rates were higher 
than that reported for the southern U.S. but slightly lower than 
that reported for the Midwestern agricultural and forest regions. 
The greatest difference between midwestern deer and those from 
the lower Yellowstone was observed among yearling females. 
Reproductive performance of the youngest breeding age class has 
been a classic indicator of overall herd productivity (Ransom 
1 967, Verme 1 969, Mansell 1 974). 

Ovulation and fetal rates increased with age suggesting that 
females of 3 to 6 years (4-7 yrs. at parturition) exhibited the 
greatest reproductive potential (Table 11). The pooled sample of 
females from all years suggested a potential of 180 fetuses:100 
pregnant females and 168 fetuses:100 total females older than 
fawns. Increasing reproductive potential with age was also 
reported by others (Roseberry and Klimstra 1970, Mansell 1974, 
Woolf and Harder 1979). 

Fetal rates, like pregnancy rates, varied between years for 
which comparative data were available: 193:100 total females in 
1982 (n = 24), 175:100 in 1984 (n = 13), and 153:100 in 1985 (n = 
16). The low rate in 1 98 5 followed 2 years of below normal 
precipitation (Fig. 2). It also followed the peak in total 
autumn numbers by one year (Fig. 6). 



3 5 



Table 11. Potential productivity of white-tailed deer on the 
lower Yellowstone from collection of reproductive tracts during 
winter and spring, 1 980-85. 

No . No . Fetuses: 100 Fetuse s: 100 Ovulation* 1 

Age Females Pregnant Preg. Fem. Females Rate 



0 17 1 100 6 1.00 

1 19 17 141 126 1.41 

2 13 12 1 92 177 1. 88 
3-6 1 6 16 213 213 2. 28 

7+ 12 11 182 167 2.11 



a Ovulation rates were based on 43 of the 77 total females 
from which reproductive tracts were examined. 

Of 81 corpora lutea of pregnancy from 42 females for which 
both ovaries were examined, 75 resulted in viable, implanted 
embryos. The overall fertilization rate (93%) was comparable to 
that reported for the Midwestern agricultural region (Haugen 
1975). Fertilization rates (85-100%) did not differ (P > 0.10) 
between age classes among data pooled for all years. Females in 
age classes 4, 6, and 7+ had the lowest fertilization rates, 92, 
88, and 85%, respectively, but also exhibited the highest 
ovulation rates (Table 11). 

The pregnant fawn carried a single fetus. Among older 
pregnant females, 7 (41%) of 17 yearlings, 10 (83%) of 12 2-year- 
olds, 14 (88%) of 16 aged 3-6 years, and 9 (82%) of 11 7 years 
and older carried twins or triplets. The 4 sets of triplets were 
all from females 3-6 years of age. 



The pooled sample of females for all years indicated that 
yearlings produced more male than female fetuses (425:100, n = 
21, P_ <G 0.05). The sex ratio among fetuses of older females 
(1 2 4:100) did not depart (P > 0.10, n= 65) from an expected 1:1 
ratio. The age-specific trend in fetal sex ratios generally 
agrees with that reported by Dapson et al. (1979) and others. 



Net Productivity 

Net productivity, or recruitment, is used here to express 
the number or proportion of young surviving to autumn or 
approximately 4 months of age. Overall productivity of white- 
tailed deer on the lower Yellowstone declined from 1980 to 1984 
and increased slightly in 1985. This trend was evident from herd 
composition and age ratios from autumn aerial surveys (Table 5) 
and from fawn-at-heel ratios during July through early October 
(Table 13). Ratios of fawns:100 producing females among pooled 
samples of females with 1 or more fawn(s) at heel during July- 
August and September-October for all years were 150:100 (n = 538) 



36 



Table 12. Reproductive potential of whi t e- t a i 1 ed deer across the 
inhabited range in North America. 



Ovulation Rate Fetal Rate 



Geographical 

Location Yrlg. Ad. Yrlg. Ad. Source 



Southeast 

Flor i da 



1.14 1.32 Richter and 

Labisky (1985) 



Southwes t 

South Texas 



Llano Basin, 
Texas 1.27 

Coas ta 1 Texa s — 



1.56 



1.32 1.59 Barron and 

Harwell (1973 ) 



0.75 1.08 Teer et al. 

(1965) 
1.57 White (1973) 



Mi dwe s t Ag r . 
Iowa 

Illinois 
Reg ion w id e 



2. 36 
1.9 3 



2. 23 
2.24 



Midwest Oak/Hickory Forest 
Regionwide 



Northern Great Plains 
L . Yell ow- 

s tone R . 1.41 
Long Pines — 

Northern Rocky Mts. 
Swan Valley 
NW MT 1.30 



Idaho 



1.5 0 



Southern Manitoba 

Whiteshell 1.22 
Turtle Mt . 2.00 
Duck Mt . 1.43 



2 . 09 
1 .80 



1.95 
2.09 



1 . 92 
2.18 

2 . 00 



2. 00 
1.76 

1.66 



1.66 



1. 26 



1.25 
1.50 



1 . 00 
1.7 1 
1 . 29 



2.23 Haugen (1975) 
1.94 Roseberry and 

Klimstra (1970) 
1.91 Gladf elter 

(1984) 



1.87 Torgerson and 
Porath (1984) 



1.88 This study 
1.80 Dusek (1980) 



1.64 Mundinger 

(1981 ) 
1.88 Will (1 973 ) 



89 
87 
92 



Ransom (1967) 



Ontario 



1.2 1 



1.74 



0.82 



1.31 Mansell (1974) 



37 



Table 13. Productivity of adult white-tailed deer females on the 
Yellowstone River from the number of f a wn s-a t-hee 1 during July- 
October 1 980-85. 3 



Sept. /Oct 
1980 

Jul . / Aug . 
1981 

Sept . /Oct 
1981 

Jul./ Aug . 
1982 

Sep t . /0c t 
1982 

Jul . / Aug . 
1983 

Sept . /Oct 
1983 

Jul . / Aug . 
1984 

Sept . / Oct 
1984 

Jul . / Au g . 
1985 

Sep t . 
1985 



Fawns: 100 
Producing 
Females 



154: 100 



162 : 100 



154: 100 



1 54 : 1 00 



143 : 100 



1 52 : 100 



143: 100 



137: 100 



1 27: 100 



1 43 : 1 00 



1 52: 100 



(1 33) b 

(101) 

(123) 

(96) 

(164) 

(135) 

(95) 

(105) 

(81) 

(101) 

(2 7) 



% of Ad. 
Fema les 
Pr od uc in g 



% of Producing 
Females w/ 



Singles Twins Triplets 



68 



60 



60 



49 



46 



47 



42 



41 



51 



49 



58 



52 



57 



64 



73 



58 



52 



51 



57 
43 
48 
41 
44 
43 
35 
27 
41 
44 



Sample obtained from Elk Island, Seven Sisters, and 
Intake study areas. 



Sample size of producing females in parentheses. 



38 



and 146:100 (n = 623), respectively. These ratios cannot be 
directly compared with the potential rate per pregnant female 
(180:100), because they did not account for females that lost all 
fawns. 

The trend in the ratio of f awns:s uccessf ul females during 
September and October (Table 13) suggested a decline in 
reproductive success from 1980 to 1984 and an increase from 1984 
to 1985. The proportion of successful females that reared twins 
or triplets also declined. Of 515 females observed with fawns- 
at-heel during September-October of all years, 215 (42%) were 
accompanied by twins and 11 (2%) by triplets. 

Age-specific reproductive rates, determined from 
individually marked females during 1980-86 (Table 14), indicated 
that reproductive success increased with age to 4 years, declined 
slightly among 5-year-olds, increased again among 6-7 year-olds, 
and declined among females 8 years and older. Only 2 (5%) of 44 
yearling females successfully reared fawns (1 each during 1980 
and 1983). This, together with the low incidence of pregnancy 
among fawns, indicated that females younger than 2 years did not 
contribute to reproduction (P_ > 0.10). Of 227 records of 
reproductive success among marked females older than yearlings, 
170 (75%) successfully reared 261 fawns. Expanding these data, 
on the basis of female age class structure and age-specific fetal 
and recruitment rates, resulted in a fawn:adult female ratio of 
82:100 that compared to a mean of 85:100 from pre-hunt aerial 
herd composition surveys. 

Reproductive status of 57 females 2 years and older during 2 
or more successive years provided 95 2-year reproductive 
histories. Females successfully reared fawns during both years 
in 58 (61%) cases, reared fawns in alternate years in 24 (25%) 
cases, and failed to rear fawn(s) during both years in only 13 
(18%) cases. Of 36 cases of 3-year reproductive histories among 
24 females, 22 (61%) were successful during all 3 years, and 3 
(8%) failed in 3 consecutive years. Nine females provided 11 
cases of 4-year histories of which 4 successfully reared fawns 
and none failed during 4 consecutive years. Two females had 5- 
year histories, and both successfully reared fawns during all 5 
years. Of 4 females with 2 or more successive years of 
reproductive failures, 3 initially failed as 2 -year- olds, and 1 
was 8 years old. All 4 were from the Intake unit. 

Of individually marked adult (2+) females, all of 5 in 1980, 
19 (8 6%) of 22 in 1981, 28 (82%) of 34 in 1982, 32 (7 8%) of 41 in 
1983, 31 (67%) of 46 in 1984, 30 (68%) of 44 in 1985, and 23 
(70%) of 33 in 1986 successfully reared fawns. The data from all 
females observed during September and early October also 
suggested a decline in the proportion of females successfully 
rearing fawns from 1 980 to 1 985 (Table 13). In spite of the fact 
that yearling females accounted for 34% of all adult females 
during summer 1980, only 32% of 133 observed females did not 
successfully rear fawns. This suggested that nearly all females 
older than yearlings successfully reared fawns that year. 



39 



Table 14. Productivity of wh i t e- t a i 1 ed d ee r on the lower 
Yellowstone from 155 individually marked females observed during 
June through October, 1980-86. 

No. Females Fawns: 100 Fawns: 10 0 

Successfully Producing Total 

Age No. Females Rearing Fawn(s) Females Females 



1 


46 


2 


100 


4 


2 


45 


30 


137 


91 


3 


48 


36 


153 


115 


4 


38 


31 


161 


132 


5 


29 


21 


152 


110 


6 


23 


23 


152 


152 


7 


15 


14 


157 


147 


8 + 


25 


15 


140 


84 



Assuming the contribution of yearling females to annual 
production was negligible during 1981-85, and that 26, 31, 27, 
25, and 21% of the females older than fawns were yearlings during 
the respective years, expanding the data in Table 13 resulted in 
fawn-rearing success of 86, 91, 76, 71, and 63% for all females 
older than yearlings during those years. This trend was similar 
to that among individually marked females during 1980-85. In 149 
cases of reproductive success among marked females, 69 (46%) were 
accompanied by twins and only 1 case was documented of a marked 
female successfully rearing triplets. 

Reproductive success among individually marked females 
varied inversely with population size, and 2 and 3 year-old 
females appeared more sensitive to these changes than older 
females (Fig. 9). A negative linear relationship existed between 
numbers of adult (2+) females during summer and reproductive 
success among both 2 and 3 year-old females (P_ < 0.01) and older 
females (P_ < 0.05). Numbers of yearling females and yearling and 
older males did not contribute to this relationship (P_ > 0.10). 
A paired t-test indicated lower (P_ = 0.05) reproductive success 
among females of 2 and 3 years of age than among older females. 

Although net productivity fell short of the potential for 
the lower Yellowstone population, nearly two-thirds of the 
females of fawn-rearing age successfully reared fawns in 
successive years. This was in contrast to northwestern Montana, 
where net productivity was characterized by alternate year 
success predicated by physiological stress of winter on females 
(Mundinger 1 981). 



40 



100 1 



LL 
(/) 

m 
O 
O 
ZD 
</) 



LU 

o 

DC 
LU 
Ol 



8© i 




females 
1 = 



▲ 4+ yr. females 
y = 115.17 -0.02X 



500 1000 1500 2000 2500 

NUMBER OF ADULT FEMALES 



300© 



Figure 9. The relationship of summer density of mature females 
to reproductive success of 2-3 year-old and older females on the 
lower Yellowstone River, 1980-86. 



Spatial Variation in Reproduction and Recruitment 

Fawnradult female ratios (Table 6) and age-specific 
reproductive rates for marked females (Table 15) suggested that 
net productivity was lower at Intake than at Elk Island. Age- 
specific pregnancy and fetal rates were also lower at Intake. 



Reproductive patterns of marked females indicated that 
reproductive success, age of females, and geographic location 
were mutually dependent (Chi-sq. = 4 1.50 3, 7 df , P < 0.001). The 
frequency of reproductive success among young adult (2-3 y r s .) 
and old (8+ yrs.) females at Intake was less than expected (P_ < 
0.05). In contrast the frequency of reproductive failures among 
prime-age females (4-7 yrs.) at Elk Island was less (P_ < 0.05) 
than expected. Differences between fetal rates and net 
productivity indicated that approximately one-half of the 
reproductive potential of yearling and older females was lost 
over summer at Intake; at Elk Island, the loss was only one-third 
of the potential. 



41 



Table 15. Spatial variation in potential productivity and 
recruitment of female white-tailed deer on the lower Yellowstone 
River, 1980-8 5. 

Age Above Intake Below Intake 



Pregnancy Rates 

Yrlg. 83% (12) 91% (23) 

2+ 91% (45) 98% (52) 

Fetal Rates 

Yrlg. 100:100 (7) a 155:100 (11) 

2+ 156:100 (16 ) 200:100 (26) 

Fawn-rearing Success 

2 43% (14) 78% (27) 

3 44% (16) 92% (25) 
4-7 79% (39) 90% (51) 
8+ 22% (9) 92% (12) 

Net Productivity ( Fa wn s : Fe ma les ) 

2 50:100 (1 4) 104:100 (2 7) 

3 63:100 (16) 144:100 (25) 
4-7 108:100 (39) 149:100 (51) 
8+ 2 2:100 (9 ) 13 3:100 (12) 

% Multiple Births 

2 17% (6) 33% (21) 

3 43% (7) 56% (23) 
4-7 3 5% (31) 6 7% (4 6) 
8+ -- 45% (11) 



Sample size is in parentheses 



MORTALITY 

Neonatal/Summer Fawn Mortality 

All 101 fetuses from 55 pregnant females examined from the 
end of the 2nd through the 5th month of gestation appeared viable 
and normally developed. Because of this, we assumed that the 
difference between fetal rates and autumn fawn:adult female 
ratios reflected mortality occurring from parturition through 4 
months postpartum. Others studies (eg. Teer et al. 1965) have 
documented only negligible fetal mortality in deer. 

Neonatal and summer fawn mortality was 1 0, 3 1, 28, 44, 48, 
and 3 3% for the years 1980-85, respectively. There was a 
positive linear relationship (P_ < 0.05) between summer fawn 
mortality rates and adult female density (Fig. 10). Numbers of 
females of 2 years of age and older in summer populations 
accounted for 98% of the variability in oversummer mortality of 



4 2 




500 1000 1500 2000 2500 

NUMBER OF ADULT FEMALES 



Figure 10. The relationship between summer fawn mortality and 
density of 2 year-old and older females, 1980-86. 

fawns. Densities of yearling females and yearling and older 
males did not significantly contribute to this relationship. 

Fawn mortality appeared to occur largely during the first 
month of life. Little change was observed in the fawn-at-heel 
ratios from July/August to September/October (Table 13). Of 48 
individually marked females of 2 years of age and older that 
failed to successfully rear fawns during 1981-85, only 6 were 
observed with fawns through late July. 

Rates of summer fawn mortality on the lower Yellowstone fell 
within the range reported for the species. Relatively high fawn 
mortality within 90 days of the peak of fawning was reported by 
Teer et al. (1 965) and Gavin et al. (1 984). Cook et al. (1971) 
and 0'Pezio (1978) also reported nearly all summer fawn mortality 
to occur within one month postpartum. 



General Patterns of Mortality 

Estimates of annual mortality were 17, 22, 29, 36, 42, and 
42% during the biological years ending in May 1981-86, 
respectively. Average annual mortality of yearling and older 
males was more than twice that of yearling and older females 
(Table 16). The average annual rates for individually marked 
females (31%) and males (59%) were slightly higher than those 



43 



Table 16. Annual and seasonal rates of mortality (%) for white- 
tailed deer on the lower Yellowstone River, 1980-86. 

_ Total Adul t Adul t 

Season Fawns Adult (1+) Female Male 



An nua 1 5 0 


(+ 7) a 


34 (+ 4) 


22 


(+ 6) 


60 


Jun . -Aug . 3 2 


(+ 5) 


0 (+ 1) 


0 


(+ 3) 


0 


Sep. -Nov. 19 


(+ 6) 


29 ( + 4) 


18 


(+ 4) 


55 


Dec. -May 10 


( + 2) 


6 ('+ 2) 


5 


( + 2) 


6 


a Percent 


mortality 


(+ SE) 








determined from 


seasona 1 


population 


estimates 


but still 





(± 
(+ 
(± 
(+ 



7) 
6) 
7) 
2) 



emphasized the higher rate among males than females. 

Mortality among deer older than 4 months, as determined from 
population estimates (Table 16), was highest during autumn and 
lowest during summer. Nearly all adult mortality (88%) occurred 
during autumn. The average fawn mortality rate from September 
through May was 26%, and similar to adults, most deaths (67%) 
occurred during autumn. Of 179 confirmed deaths of marked deer 
(Table 17), 154 (86%) died from September through November, 13 
(7%) died during winter, 9 (5%) during spring, and 3 (2%) during 
summe r . 



An increase in annual mortality rates from 1980-81 to 1984- 
85 reflected increases in hunting-related mortality (Table 18). 
Hunting accounted for an average of 81% of annual mortality of 
individually marked deer. Mortality from other causes varied 
slightly, but was generally low, claiming 5, 9, 5, 6, 11, and 9% 
of autumn populations during the 6 years, respectively. Among 
deaths of yearling and older deer during autumn, 5 resulted from 
accidents and 3 from other or undetermined causes. Only 25 
deaths of marked deer were documented during winter, spring, and 
summer. Three resulted from injury or stress induced by capture 
and handling, accidents accounted for 7, 3 were related to 
hunting, 2 were confirmed overwinter mortalities, 1 was a 
predator kill, and 9 were from undetermined causes. 

Hunting M ortality . — In Montana, killing of deer was first 
restricted by season length in 1872, by bag limit in 1907, and by 
buck-only hunting in 1921 (Mussehl and Howell 1971). Hunting of 
white-tailed deer was discontinued in eastern Montana during the 
early 1940's because of their relative scarcity. As deer numbers 
increased statewide during the late 1940's and early 1950's buck- 
only harvest regulations were replaced by one deer, either sex 
seasons from 1 952 to 1 954 (Cole 1 959). From 1 955 to 1 957, these 
were replaced with 2 deer, either sex seasons. From 1960 to 
1975, harvest regulations included 2 -deer bag limits in 
southeastern Montana that encouraged the harvest of white-tailed 
deer in hunting districts where they were the predominant species 



44 



Table 17. Number of marked yearling and older white-tailed 
deer dying from various causes by season and age class, 1980- 
1986. Causes: A=accidents, H=hunting related, P=pr ed a t i on , 
T = t ra p- r e 1 a t e d , W = winter malnutrition, M =o t he r known causes, and 
U=unde t e rm i ned causes. 

Female Deaths 



Age 
Class 



Numbe r 
at Start 



Tot a 1 
Deaths 



Summe r 



Season 



Au t umn 



Winter 



Spring 



1 

2 

3-7 
8 + 



82 
65 

200 

29 



1 9 
1 9 

56 

8 



2A 
1U 



1 5H 
1 4H 



2U 

1A, 1H, IT 
1U 



4 3H, 2A, 2M , 2H,1A,2U 

1U 

4H, 1A 



1U 



1A,1P, 
1U 

1W , 1 u 



To ta 1 



376 



102 



82 



12 



Age 
Class 



Numbe r 
at Start 



To ta 1 
Deaths 



Male Deaths 



Season 



Summer Autumn 



Winter 



Spring 



1 

2 

3-7 



97 

54 
28 



3 1 
31 
1 5 



2A , 2 7H 
2 9H 
14H 



IT 



2A 
IT 
1W 



To tal 



179 



77 



72 



or where it was desirable to reduce hunting pressure on mule deer 
(C. D. Eustace and J. E. Swenson. 1 9 7 7. Montana Dep. Fish, 
Wildl. and Parks, Fed. Aid Proj. W-130-R, Unpubl. rep.). 
Restricting one or both tags to white-tailed deer was necessary 
to increase annual harvests in some areas. 



In 1976, following a widespread decline in mule deer 
populations, hunters were limited to either 1 antlered mule deer 
or a whitetail of either sex. Increased populations of both 
species during the late 1970's and early 1980's were followed by 
increasingly liberalized harvest regulations. The "A" tag 
allowed taking a deer of either species/either sex, whereas the 
"B" tag was largely limited to antlerless deer. Multiple "B" 
tags were offered from 1 984 to 1 985. In 1984, a hunter could 
legally harvest 6 white-tailed deer from the lower Yellowstone of 
which only one could be an antlered deer. 



45 



Table 18. Hunting efficiency on the lower Yellowstone River by 
sex and age class of deer from population and harvest statistics, 
hunting mortality of marked adult deer, and change- in- ratio 
procedures. 

Percent Removal by Sex and Age Class 



Total 3 Total Adult 0 Adult 

Population Fawns* 5 Adults Females 1.5+ Males 1.5 + 

Year N=3 d N=3 



1980 13 3 

1981 14 5 

1982 25 10 

1983 32 14 

1984 35 17 

1985 34 22 



21 11 38 

21 6 60 

34 14 68 

40 26 69 
36 30 53 

41 33 57 



Calculated from percentage of adults removed with the 
proportion of fawns and adults in fall populations and 
annual harvests known. 

Calculated with the same knowns given above. 

Calculated from the percentage of total adults and males 
removed with the proportion of adult males and females in 
fall populations and annual harvests known. 

N - the number of independent estimates which give the 
mean total adults and adult males removed per year. 



Hunting-related deaths documented during this study included 
both legal harvest and deaths resulting from crippling. Some 
deaths attributed to crippling may have resulted from hunters 
failing to retrieve their kills or deliberate abandonment, but 
they were not differentiated. 

Crippling losses accounted for 20% of all hun t i ng- rel a t ed 
deaths of marked deer, or 24% of the legal harvest, from 1980 to 
1985. Archery hunting accounted for 10 (7%) of 148 hunting- 
related deaths. Two of the 10 died as a result of crippling. 

Hunting mortality generally reflected disproportionately 
heavy harvest of antlered over antlerless deer. Harvest of adult 
males (2+ yrs.) over represented (P < 0.05) their abundance in 
early autumn populations during all years (Table 19). Yearling 
males were harvested in greater proportion than their abundance 
(P < 0.05) only during 1 980 and 1981. The decline in the 
proportion of yearling males in annual harvests from 1982 to 1985 



46 



Table 19. Composition of annual harvests of white- tailed deer 
from the lower Yellowstone River from deer examined at check 
stations and in the field during 1980-85. 

Percent of Total Harvest 



Year 



Fwn Male Fwn Fern 



Yr Male 



Y r Fem 



Ad Male 



Ad Fem 



1 980 
1981 
1982 
1983 
1 984 
1985 



4 
7 
8 
5 
12 
13 



_a 



7 
6 
7 
9 
1 7 
1 1 



30 
28 
18 
14 
14 
13 



+ 
+ 



12 
5 

10 
9 

10 
8 



22 
38 
37 
26 
18 
22 



+ 
+ 
+ 
+ 
+ 
+ 



25 
15 
20 
37 
28 
32 



A (+) indicates that a sex and age class was harvested 
in greater proportion (P_ < 0.05) than its representation 
in the population; a (-) indicates a harvest less than 
availability (P_ < 0.05); and blanks indicate no 
difference. 



coincided with a significant decline in antler main beam lengths 
from 1983 to 1984 (Table 9) and liberalized harvest regulations 
for antlerless deer during 1983-85. The number of hunters 
reporting difficulty in distinguishing yearling males from 
antlerless deer was particularly noticeable during 1984 and 1985. 
The decline in the proportion of all antlered deer in annual 
harvests from 1 981 (66%) to 1 984 (3 2%) (Table 19) also coincided 
with liberalized regulations and increasing proportions of autumn 
populations being harvested (Table 18). 

The proportion of fawns in annual harvests under represented 
(P_ < 0.05) their abundance in early autumn populations during all 
years except 1984 and 1985 (Table 18). This reflected decreased 
hunter selection against fawns when multiple antlerless tags were 
available. The proportion of yearling and older females in 
annual harvests never exceeded (P > 0.05) their availability 
(Table 18). However, yearling females were under represented ( P 
< 0.05) in the 1981 harvest and adult females were under 
represented in 1981 and 1984. 

The sex ratio of fawns in samples of hunter- harvested deer 
did not depart from 50:50 (P_ > 0.05) during any year. Although 
males consistently outnumbered females among yearling deer in 
annual harvests, the ratio departed from the expected 50:50 
(P < 0.05) only during 1 980 and 1981. 

Harvest of antlered deer was correlated (r = 0.98, P_ < 
0.001) with their abundance in early autumn populations from 1980 
to 1985. This relationship was more difficult to ascertain for 
antlerless deer because season structure, bag limits, and season 
length changed from year to year. The increased proportion of 



47 



antlerless deer in annual harvests (Table 19) from 1982 to 1983 
coincided with increasingly liberalized antlerless regulations. 

Others have also indicated that vulnerability to hunting was 
dependent upon the sex and age of deer. Roseberry and Klimstra 
(1974) and Coe et al. (1980) reported a tendency of hunters to 
take adult deer over fawns. The latter study indicated that 
adult females were most vulnerable among antlerless deer because 
they were often the first deer seen, and fawns were less 
vulnerable when in groups than when alone. Roseberry and 
Klimstra (1974) suggested that yearling and 2-year-old males were 
harvested at a higher rate than other sex or age classes. They 
also reported a decline in vulnerability with cumulative hunting 
pressure from increased wariness as a result of harassment. In 
the absence of selectivity, adult males may be less vulnerable to 
hunting than adult females and fawns (Van Etten et al. 1965). 

The rate of crippling on the lower Yellowstone was 
comparable to that reported for the Midwest by Roseberry et al. 
(1969) and Stormer et al. (1979) but somewhat higher than 
reported in Texas by Teer et al. (1 96 5). Stormer et al. (1 97 9 ) 
attributed a decline in the rate of crippling to a decline in 
deer density and hunter success. We were not able to detect 
changes in rates of crippling with changes in population size. 

Traffic-related M ortality . --Automobile- deer collisions 
claimed an average of 2% of autumn populations annually. They 
accounted for 11 (6%) confirmed mortalities of individually 
marked deer. Mortality determined from marked deer may have 
overestimated highway mortality because marked deer dying along 
roads may have been reported in greater proportion than those 
dying from natural causes. 

The distribution of road-killed deer was nonrandom (P_ < 
0.001) in both time and space. Of 170 carcasses examined along 
roads from 1 980 to 1 986, 147 (8 5%) died during October-April 
(Table 20). October-December and March were peak periods of 
highway-related deaths when 49% and 15% of this mortality 
occurred. Most vehicle-deer collisions presumably occurred 
during crepuscular and nocturnal periods, because many fresh 
road-kills were first sighted during these periods or shortly 
after daylight. This also closely agreed with daily periods of 
maximum activity and movement reported by Herriges (1986). 

Over the study, 157 (92%) deer were killed along Highway 16; 
the remainder died on secondary Highway 23 and on county roads. 
Seventy-two percent of the kills along Highway 16 occurred within 
approximately 16% of the distance from Glendive to Sidney; no 
kills were observed on 20% of that distance. 



Topography, distance from the river, and land use did not 
influence the number of deer killed (P_ > 0.10) per segment (1.6 
km) of highway, although there was a tendency for more accidents 
to occur near the river where agricultural crops were present 
(P_ = 0.12). Approximately 43% of the highway mortality occurred 



48 



Table 20. Monthly auto- related mortality of white- tailed 
deer along the lower Yellowstone River from April 1980 to May 
1 986. 



Femal e s 



Month Fawn Yr 1 g . Adult 



June 




1 








1 


2 


July 


1 


3 






1 


1 


6 


Augus t 




1 


1 








2 


September 




3 


4 


3 


2 




13 a 


Oc tobe r 


7 


3 


1 1 


4 


1 


2 


28 


No vember 


9 


5 


6 


8 


8 




36 


December 


4 


2 


7 


4 


2 


1 


20 


Ja nua r y 


4 


2 


3 


3 






12 


February 


3 




3 


2 


3 




1 1 


Ma r ch 


3 


4 


5 


5 


8 


1 


26 


Apri 1 




2 


6 


2 




1 


11 


Ma y 


1 








2 




3 


Total 


32 


26 


46 


31 


27 


7 


1 70 



Males 



Fawn Yrlg. Adult Total 



Total includes fawn of unknown sex. 



less than 2 km from the river channel in areas where agricultual 
crops were grown (30% of the total distance). The lowest 
mortality (2% of the total) occured near housing developments or 
other commercial buildings on the floodplain at distances 
exceeding 2 km from the river (8% of the total distance). 

Composition of highway mortality by sex and age during June- 
November reflected herd composition of early autumn. Relative 
numbers of deer killed by sex and age class did not depart ( P_ > 
0.05) from expected frequencies, nor did sex ratios of fawns and 
yearlings depart from 50:50. For December-May, the composition 
of road -killed deer was compared with early winter herd 
composition. Yearling males were killed in greater proportion 
than their abundance (P_ < 0.05) in the winter population while 
mortality of other sex and age classes did not depart from 
expected frequencies (P < 0.05). 

Several studies have suggested that trends in traffic 
mortality reflected gross changes in population size (Jahn 1959, 
Case 1978). McCaffery (1973) reported that traffic- related 
mortality most accurately reflected these changes when corrected 
for changes in traffic volume. Numbers of road-kills examined 
along the lower Yellowstone per year was not correlated with 
changes in population size (r = 0.25, 4 d.f. P > 0.10). This may 
have reflected unequal rates of recovery between years as well as 
a change in traffic volume over time. 



4 9 



The aggregated distribution of traffic-related mortality in 
time and space generally agrees with findings reported by Bellis 
and Graves (1971), Puglisi et al. (1 97 4), and Bashore et al. 
(1985). Allen and McCullough (1976) attributed a relatively 
random spatial distribution of road-kills to similar habitat 
conditions in an agricultural region of southern Michigan. 
However, they reported that most car-deer collisions occurred 
during crepuscular and nocturnal periods. The nonrandom pattern 
of mortality on the lower Yellowstone in both time and space 
probably reflected distribution of agricultural fields and daily 
patterns of activity associated with their use (Herriges 1986). 

Movement of individually marked deer into areas they did not 
routinely use except during periods of comparatively heavy 
hunting resulted in traffic-related deaths of 3 animals in the 
Intake unit during late October-early November 1984. Deer moving 
to feed in winter wheat fields in upland terraces west of Highway 
16 during March and comparatively early green-up of roadside 
vegetation during March and early April may have contributed to 
the high frequency of highway-related mortality during the early 
spring period. 

Natural M or ta 1 i ty . — Natural mortality claimed an average of 
5% of autumn populations during the study. Deaths related to 
malnutrition during late winter and spring accounted for 3% of 
confirmed mortalities of individually marked deer for all years 
combined. All occurred during February -April; 3 of 6 were fawns, 
2 were females of 9 years and older, and one was a 4 year-old 
male. Winter severity and estimates of overwinter mortality of 
fawns were not correlated (r = 0.67, 4 d.f., P > 0.10), although 
highest estimated fawn mortality (17%) occurred during the 
severest winter, 1981-82 (Table 2). The relatively low incidence 
of overwinter mortality is consistent with the comparatively good 
condition in which animals enter the winter period as reflected 
by the KF I (Table 8). Substantial overwinter mortality was 
documented during severe winters in conifer habitat in 
southeastern Montana where deer were in relatively poorer 
condition than those on the lower Yellowstone (Dusek 1987). 

Predation, accidents (not hum a n- r el a t ed) , and undetermined 
causes collectively accounted for 9% of all confirmed deaths of 
marked deer. Two marked deer were suspected victims of EHD 
during late summer 1984, and several sightings or reports of dead 
deer along the river bottom were obtained. One radio-collared 
deer died during the same period in 1985. There was one 
confirmed report of a white-tailed deer with symptoms of 
hemorrhagic disease from elsewhere in eastern Montana in 1984. 



50 



DISTRIBUTION 

Wh i t e- t a i 1 e d deer occurred throughout the entire length of 
river bottom, though densities varied. The occurrence of small 
numbers of whitetails in adjacent uplands and along ephemeral 
tributary streams appeared to reflect seasonal shifts of some 
individuals off the river and/or local distribution, movement, 
and habitat selection relative to specific habitat 
characteristics . 

Diurnal distribution of deer during summer generally 
centered on large blocks of riparian tree and shrub cover, but 
also included small, isolated patches of trees and/or shrubs 
along drainage ditches, the USER main canal, and field borders. 
Telemetry and spotlight surveys (Herriges 1986, Compton 1986a) 
indicated that nocturnal distribution also included agricultural 
fields. Use of uplands, which included wooded draws, was most 
prevalent in the Intake area, where the river bottom was narrow 
and upland riparian cover occurred in close proximity to the 
river bottom. 



Diurnally in winter, deer were distributed almost 
exclusively in large stands of riparian cover near the main river 
channel and on some of the larger islands. The only notable 
exception occurred at Intake during 1983-84 when, coincident with 
high deer densities, some individuals used upland draws. At Elk 
Island, many ra d i o- co 1 1 a r e d deer that summered locally throughout 
the floodplain shifted their daytime distribution to the large 
island in winter when ice provided easy access across the river 
channel. Nighttime distribution in winter extended from the 
riparian cover used during daytime to agricultural fields 
throughout the river bottom and, locally, to adjacent uplands. 
As during summer, use of agricultural fields in uplands was most 
prevalent on the Intake area. 



Habitat Factors Influencing Distribution 

Riparian cover, which included shrub and forest cover types, 
was the single most important factor affecting density 
distribution of deer within the study area (Fig. 11). A linear 
relationship (P < 0.0001) existed between the amount of riparian 
cover and numbers of deer observed in 1.6 km segments of river 
bottom during early winter surveys each year from 1981 to 1985 
and for all years combined (Table 21). This factor alone 
explained 74% of the overall variation in relative abundance of 
deer observed along the river bottom from the pooled data for all 
years. These findings generally agree with those of Compton et 
aL (1 988) for a portion of the study area during winter, summer, 
and autumn of 1985. 



The amount of island area may also have influenced 
distributional patterns. Multiple regression models indicated 
that island area combined with the amount of riparian cover 
influenced distribution and relative density of deer (P_ < 0.05) 



51 



C 




0 100 200 300 400 

Amount Riparian Cover / Section (ha) 



Figure 11. Relationship between abundance of riparian 
forest/shrub cover (ha) and numbers of deer observed per 1.6 km 
segments of river bottom during early winter surveys, 1981-82 
through 1985-86. 



Table 21. Multiple regression models of deer density (Den) per 
segment of river bottom (dependent) as influenced by the amount 
of riparian cover (RC) and island area from early winter aerial 
surveys of the entire lower Yellowstone study area, 1981-82 to 
1985-86. 



Year Regression Model P - value R 



1981 


-82 


Den = 


2 .63 + 0.2 8RC 






< 


0. 


0001 


0. 


49 


1 982 


-83 


De n = 


-11.99 + 0. 31RC 






< 


0 . 


0001 


0. 


68 


1983 


-84 


De n = 


1 .27 + 0.2 9RC 






< 


0. 


0001 


0. 


56 


1 984 


-85 


De n = 


-2.55 + 0. 18RC 






< 


0. 


0001 


0. 


55 


1985 


-86 


De n = 


-3.92 + 0.07RC + 


0. 


1 2IA 


< 


0. 


02 


0. 


44 


Al 1 


Ye ars 


Den = 


-3.56 + 0. 21RC + 


0. 


061 A 


< 


0. 


05 


0. 


77 



52 

over all years as well as during winter 1985-86. 

Aerial surveys of the Elk Island area indicated increased 
deer use of the large island f_rom early autumn (X = 22%) to a 
y e a r 1 on g_h ig h during winter (X = 40%) and decreased use during 
spring (X = 31%). This pattern was also observed among radioed 
deer which made minimal use of the. island during summer and 
maximum use during late autumn and winter. In contrast, some 
deer at Intake used a large island during autumn-spring of all 
years, but individually marked deer rarely were found there 
except during the autumn hunting season. The amount of island 
area was also directly related to density distribution of deer 
along the river bottom during autumn 1985 (Compton et al. 1988). 

Sloughs occurred in 4% of 858 9-ha cells covering the 3 
principal study units. Nine percent of 4,932 relocations of 
marked deer occurred in cells with sloughs, indicating selection 
(P_ < 0.01) for such areas. Sex and age classes responded 
differently to the presence of sloughs (P_ < 0.01). Thirteen 
percent of all (2,476) relocations of mature (+3) females 
occurred in habitat with sloughs; they were 2.5 times more likely 
to be found in such areas than other deer. 



Influence of Livestock 

Studies by Compton et al. (1988) indicated that livestock 
grazing had a negative effect on local distribution and density 
of deer along the river bottom. In general, deer avoided areas 
where cattle were present. Data concerning deer-livestock 
relationships on the intensive Elk Island and Intake study units, 
which received markedly different livestock usage, were more 
speci f ic . 

During 1980-86, approximately 70% of the Elk Island area was 
not grazed, 21% was grazed by cattle from autumn through early 
spring, and 9% was grazed by sheep from summer through early 
winter. Under this regime, deer distribution was dependent 
(P_ < 0.01) upon grazing treatment. During all seasons, marked 
deer selected (P < 0.05) areas that were not grazed and avoided 
(P_ < 0.05) areas grazed by cattle (Table 22). During summer and 
winter, deer avoided pastures traditionally grazed by sheep, but 
use of those pastures by deer did not depart from availability 
(P_ > 0.05) during autumn and spring. 

Seasonal differences in responses of deer to grazing 
treatment were highly significant ( P_ < 0.01). The relative use 
of nongrazed areas by deer was higher during summer than during 
all other seasons (Table 22). Sheep pastures received their 
greatest yearlong use by deer from autumn through spring, whereas 
relative use of portions of the f loodplain grazed by cattle did 
not appreciably change between seasons. 



5 3 



Table 22. Response of whi t e- t ai led deer to nongrazed areas (N), 
areas grazed by cattle (C), and areas grazed by sheep (S) on the 
Elk Island area from 2273 relocations of individually marked 
deer, 1980-86. 



Sex 
Age 


and 
Class 


Ye a r 1 on g 
N C S 


S umme r 

N C S 


Aut umn 

N C S 


Winter 

N C S 


Spring 
N C S 


All 


De er 


V (2273) a 
qi + o _ 7 - b 


V (659) 
9 5+ 1 - 4 - 


V (524 ) 
90+ ?- 8 


V (561) 
on+ a _ 6_ 


V (529) 
89+ 2- 9 


Ad . 


Fem . 


V (1249) 
95+ 1- 4- 


V (359) 
96+ 1- 3- 


V (311) 
95+ 2- 3- 


V (313) 
96+ 1- 4- 


V (266) 
94+ 1- 5- 


Sub ad . Fem . 


V (632 ) 
88+ 3- 9 


V (158) 
96+ 0- 4- 


V (124) 
86+ 2-12 


V (172) 
84+ 8- 8 


V (178) 
84+ 2-14 


Ad. 


Ma 1 e 


V (171) 
84+ 4-12 


V (75) 
95+ 5- 0- 


V (52) 
69 2-29+ 


(22) 
82 5 14 


(22) 
86 5 9 


Yrl. 


Male 


V (221) 
84+ 5-11 


V (67) 
88+ 0-12 


V (37) 
86+ 0-14 


(54) 
80 7 13 


(63) 
81 11 8 



V - Differences between observed and expected use of 
grazing treatments by deer in appropriate group and season 
resulted in significant (P < 0.01) chi-sq. values (sample 
size) . 

Percent of annual or seasonal relocations in the 3 
grazing treatments. Bonferroni Z tests were used to 
determine which treatments were selected (+) or avoided (- ) 
(P < 0.05). 



The yearlong influence of grazing treatment on distribution 
of deer was dependent ( P_ < 0.01) on sex and/or age. Both 
subadult and mature females selected (P_ < 0.05) nongrazed 
portions of the Elk Island area. Mature (3+) females avoided 
(P < 0.05) areas grazed by both cattle and sheep during all 
seasons. Yearling and 2-year-old females also avoided areas used 
by cattle during all seasons but avoided areas grazed by sheep 
only during summer (Table 22). 

Males exhibited no selection or avoidance of habitats with 
respect to grazing treatment during winter and spring. During 
summer and autumn, both yearling and older males avoided areas 
that were traditionally grazed by cattle from autumn to early 
spring. In summer, adult (2+) males avoided pastures grazed by 
sheep; in autumn, those areas were selected (Table 22). This may 
have been largely influenced by hunting pressure; sheep pastures 
were on privately-owned lands where hunting pressure was much 
lower than on the adjoining EIWMA and large island. Yearling 



54 



males selected (F < 0.05) nongrazed areas, but their use of areas 
grazed by sheep did not depart from availability (P_ > 0.05) 
during either summer or autumn. 

The differential response of sex and age classes to grazing 
treatment was most apparent during autumn, winter, and spring 
when the area was most intensively used by livestock, 
particularly cattle. During autumn and winter, mature (3+) 
females were 4 times more likely than other deer to occur in 
nongrazed areas; in spring this probability was 3 X. They were 7 
times more likely than subadult females or males to occur in 
areas not grazed by sheep during autumn. 

In contrast to Elk Island, nearly all of the Intake area was 
grazed by cattle from mid autumn through winter (Table 3). There 
were no apparent differences in deer distribution from spring 
through mid autumn that could be directly related to winter 
grazing. Similarly, because of the extensive occurrence of 
cattle from mid autumn through winter, we were unable to detect 
any measureable changes in distribution of deer along the river 
bottom during that period. However, deer use of adjacent, 
ungrazed uplands for feeding increased after cattle were placed 
on the bottom during all years. Some used upland areas 
throughout the diel period during winter 1983-84, when deer 
densities were highest and cattle were fed on the bottom through 
most of the winter. Lacking ungrazed areas to which they could 
move, deer that remained on bottomlands during winter apparently 
accomodated the presence of cattle through more subtle movements 
and changes in habitat use. 



MOVEMENTS 

White-tailed deer on the lower Yellowstone River were 
generally nonm i g ra t o r y . Seasonal ranges overlapped among 27 
(79%) of 34 radioed adults and 13 (68%) of 19 yearlings. Mean 
distances between summer and winter geographic centers of 
activity (GACs) for deer with overlapping seasonal ranges were 
less (P_ < 0.05) than for those with discrete seasonal ranges 
(Table 23). Of 13 deer (7 adults, 6 yearlings) with discrete 
seasonal ranges, yearlong movements of 10 were confined to the 
main Yellowstone Valley which included the floodplain, islands, 
adjacent terraces, alluvial plains, and lower reaches of 
ephemeral streams. Discrete seasonal ranges of 2 yearlings were 
related to dispersal behavior; as adults seasonal ranges of each 
animal overlapped. 

Of 52 radioed deer monitored for at least one biological 
year, 49 (94%) had home ranges entirely within the river valley; 
3 left the main drainage for periods of 6-7 months during the 
year. The latter included one female which, as a 3-year-old, 
spent summer and autumn in upland prairie habitat 9.1 km from her 
winter home range, but subsequently established a permanent home 
range on the river bottom approximately 1.5 km from the winter 
range. A yearling male left the river bottom during June, spent 



5 5 



Table 23. Distance between summer and winter GAC's by sex, age, 
and seasonal ranges based on analysis of 78 deer years of data 
from 1 980-85. 



Distance Between GACs (km) 



Type of 

Seasonal Ranges Sex and Age 



N 



SE 



Overlapping 



Females 

Yr Igs 
Ads . 



1 1 
44 



0. 40 
0.34 



0. 09 
0. 04 



Males 

Yrlgs 
Ads . 



1 . 20 
0.93 



0. 90 
0.20 



Di scret e 



Females 

Yrlgs 
Ad s . 



6. 80 
2.87 



3. 80 
0.88 



Males 



Yrlgs 



8. 73 



3. 73 



the summer and autumn in prairie habitat 14.6 km from his 
ancestral home range, and returned to the bottom during December. 
He repeated this movement pattern the following year until killed 
in November. Another male, a 2-year-old, moved to uplands 
approximately 12.1 km from the river bottom during July, returned 
to the bottom in January, and remained there until killed the 
following November. Other movements by marked deer into uplands 
were related to daily movement, were of short duration (2 wks. or 
less), or were dispersals. Fidelity of whitetails to bottomland 
habitats, with little or only occasional use of uplands, was also 
reported elsewhere in eastern Montana (Allen 1968), Missouri 
(Zwank et al. 1979), and Washington (Gavin et al. 1984). 



Diurnal Movements 

Movement patterns of deer varied (P < 0.05) by month, sex 
and age (Fig. 12). There were no differences (P_ > 0.10) in 
diurnal movements related to study unit, either alone or in 
combination with the 3 independent variables. 

Movements of deer while in uplands (2% of the total animal 
months) were not included in the analysis. Monthly AARs of an 
adult female that spent late spring-autumn in upland prairie 
habitat (1.7 0, 0.3 0, 0.5 0, and 2.20 km during May, June, July, 
and September, respectively) were much larger than those recorded 



56 




W 
ill 

< 

+ 
DC 

S 

>■ 

CM 




CM 



CM 



CO 



CD 



QJ 



CO 

u 
o 
u 
u 

\ — I QJ 
•H 

nj -d 

4-> u 

I a} 

0) TJ 

4-1 C 

•H cd 

rCl 4-> 

£ CO 

o 

m 4-) 

qj a) 
bfl co 

u 

t3 a. 
C 0) 
03 U 




(0 

8 

0) 

CO 




€M 



CM 



O 



in 



— r 
o 



in 
d 



to 



o 

■ 

© 



UJ 



u. 

+ 

DC 

< 
UJ 

>• 

CO 




UJ 

< 
2 

UJ 

LL 

z 




CM 



CM 



. o 



CO 



CO 



r 
O 



—r 
io 
o 



o 
o 



X 
Z 

o 



X co 
aj a) 
co a 

•H 

rH 

r-H 

pcj a 

<j -rl 
<JJ 4-1 
^ 5-1 

CD 

CO > 

d 

•rl 

aj m 
5-i oo 
i 

>> o 

4-1 00 

•H On 

> ' ' 
•H 

4-) ~ 

a 5-i 

aj aj 
> 

a) -h 

bO pi 



5-4 

a) 
> 



a) 
C3 
o 

4- ) 
CO 

i — I o 

X r-i 
4-1 i — I 

d aj 

O >H 

e 

5- 1 

QJ 

O 



a 

CD 



(iu>0 yvv 



CN 4-1 



a 

<D O 
5-J 

d 5-i 
00 <U 
•H QJ 
En Td 



57 



for females of all ages on the river bottom. The movements of 
two males, while in uplands, were also generally greater than 
those of males of the same ages on the bottom. These movement 
patterns closely fit those reported for whitetails in prairie 
habitat (Wood 1987), where monthly movement indexes for females 
were nearly 3 times greater than on the lower Yellowstone. 

Monthly movements of fawns on the river bottom during 
September-December were assumed to approximate those of females 
of fawn-rearing age. Movements of fawns marked in winter 
corresponded to those of adult females (3+) from January through 
April (Fig. 12). The increased movement during May was 
attributed to the breakup of family groups. Among 4 family 
groups in which both doe and offspring were radio-collared, the 
pairs separated between 13 May and 7 July. All offspring (2 
females and 2 males) left their ancestral home range, although 
one female later rejoined its mother. 

Comparatively high movement indexes among yearling and 2- 
year-old females during May and June (Fig. 12) also reflected 
breakup of family groups and included dispersal of some females 
from ancestral home ranges. This was consistent with the 
hypothesized socially subordinate status of these age classes 
(Hawkins and Klirastra 1970) and was similar to movement patterns 
reported by Downing and McGinnes (1975). 

Ozoga and Verme (1982) and Nelson and Mech (1984) indicated 
that females may remain with their mothers until about 3 years of 
age, when some may establish a new home range some distance from 
the ancestral home range. This could explain the increased 
movement during May of females nearing their third birthday (Fig. 
12). Movement of a radio-collared female from the river bottom 
just prior to her third birthday may have reflected such 
movement, but she apparently did not establish a traditional home 
range until 4 years old. During June, one rad io- co 1 la red 2 year- 
old female left the river bottom permanently, another temporarily 
left her summer home range. Monthly AARs of yearling and 2-year- 
old females during October-April, although erratic and variable 
between individual animals, resembled those of older females. 

AARs for females 3 years and older were generally less 
variable between months and between individuals than those of 
other deer (Fig. 12). They were lowest from May to August, the 
late gestation and fawning period. Movements increased slightly 
in September, coincident with weaning of fawns or regrouping of 
maternally-related females. 

The marked increase in the AAR among adult females during 
October-November coincided with the hunting and breeding seasons. 
Females commonly responded to hunting pressure by moving across 
the river channel to islands or to nearby floodplain areas where 
hunting pressure was comparatively light or absent. Radio- 
collared females on the Elk Island and Seven Sisters areas either 
moved from or extensively within public hunting areas from the 
opening of pheasant season in mid-October through the end of the 



58 



deer season in late November or early December. One moved 16.1 
km from her home range In the Elk Island area to an island near 
the mouth of Burns Creek but returned to her home range within 10 
days. Marked females on the Intake unit commonly moved into 
cover along tributary streams during periods of harrassment. 
AARs for all females declined after November but remained higher 
during December-April than during May-August (Fig. 12). Michael 
(1965) observed little or no seasonal variation in movements of 
adult females in an unhunted population in Texas. 

Yearling males were comparatively more mobile than other sex 
and age classes from June through November but less mobile than 
adult males from December through May. Mean monthly AARs were 
characterized by sharp peaks during June and November (Fig. 12). 
Breakup of family groups in June usually resulted in the male 
offspring leaving the ancestral home range. Such movement was 
observed among 4 of 5 radioed yearlings and included 3 dispersals 
and one temporary shift to adjacent uplands. The large AAR 
during November was probably influenced by small sample size (n = 
4) and dispersal of one radio-collared individual. 

Adult males were more mobile than adult females during all 
months except October and November (Fig. 12). However, like 
adult (3+) females but unlike both sexes of yearlings, monthly 
AARs of adult (2+) males were relatively stable throughout the 
year. Monthly means gradually increased from September to 
January, leveled off through April, and declined through August. 

The greater mobility of males than females was consistent 
with findings of Carlsen and Farmes (1957), Thomas et al. (1964), 
and Michael (1 965). Nelson and Mech (1 984) also reported greater 
mobility of yearlings than adults in both sexes of deer. 



Die! Patterns of Movement 



The most distinct feature of movements throughout the diel 
period consisted of travel between wooded cover and agricultural 
fields. Deer used the same general areas of riparian habitat 
during diurnal periods and moved to the same fields during 
crepuscular and nocturnal periods (Herriges 1986). Those that 
shifted areas of diurnal use within seasons used the same fields 
as other deer in the new location. 



During summer, diurnal movements of deer were minimal (Fig. 
13), and crepuscular and nocturnal movements were typically to 
the nearest alfalfa field, usually within 0.5 km of riparian 
cover. Deer at Intake appeared to move to agricultural fields 
more consistently than deer in the Elk Island area. The 
proportion of nocturnal periods in which deer travelled to fields 
averaged 86% and 62% in the respective areas. 



59 



Z 

o 

< 

O 

o 



DC 

z> 
o 

X 

z 

LU 



LU 
CQ 



600" 
500- 
400- 
300- 
200. 
100 
0 



SUMMER 



ELK ISLAND 
INTAKE 





— i — i — i — ! — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i i i — r~~i — r— t — i 
-8 -7-6-5 -4-3 -2 -1 1 2 3 4 5 6 7 8-9 -8 -7-6-5 -4 -3 -2-1 1 2 3 4 5 6 7 8 



HOURS FROM SUNRISE 



HOURS FROM SUNSET 



500 i 

LU 

Z 400 " 

< 

H 

(/) 300. 
Q 

200. 
100 



WINTER 





i — i — i — r~T~ 



— i — i — r~T — i — i — g — i 1 1 i — i — i — t~t — i — i — i — i — i — i — r 
-8 -7-6-5 -4-3 -2 -1 1 2 3 4 5 6 7 8-9 -8 -7-6 -5 -4 -3-2-1 1 2 3 4 5 6 7 8 



HOURS FROM SUNRISE 



HOURS FROM SUNSET 



Figure 13. Mean distances moved between successive locations of 
deer during summer and winter on the Elk Island and Intake study 
units. 



60 



The proportion of deer moving to agricultural fields and the 
time spent in fields increased from early to late summer, 
especially for females with fawns. During mid-June, 7% of 
nighttime relocations of radio-collared adult females with fawns 
occurred in agricultural areas compared to 2 6% of those without 
fawns. These proportions increased to 22 and 44%, respectively, 
by early August. By early September, over half of all nighttime 
relocations of females were in agricultural fields. Fawns were 
first observed accompanying females to fields during late July- 
early August (Herriges 1986). Yearlings of both sexes and older 
nonproducing females visited fields more frequently and stayed 
longer per trip during early summer. Deer that spent diurnal 
periods in upland draws in the Intake area regularly traveled 2 
km or more to the river bottom at dusk and returned before 
sunrise, whereas those that restricted their movements to the 
floodplain typically moved less than 0.5 km from bedding sites to 
fields. 

The river appeared to impede movement of females with fawns 
until early August, and at least partially hindered movements of 
other deer. Minimal movement across the river channel during 
early summer may have been influenced by river flow which 
normally peaks in June and drops rapidly from July to August. 
Deer were often observed swimming across the river at dawn and 
dusk during late summer, although hourly rad io- t r ack i ng indicated 
that individual deer did not make regular nocturnal movements 
across the main channel of the river. 



A radioed adult female, that summered on a small area of Elk 
Island, did not move across the channel to agricultural fields 
when monitored over 24-hr periods during the summers of 1982 and 
1983. Another, occupying an adjacent home range on the island 
and monitored only during 1982, also restricted her summer 
movements to the island. A third female, a 2-year-old with fawn, 
shifted her daytime activity off the island during late July 
1 983. During this same period an adult female on the main 
floodplain moved to and spent at least 1 hour in nearby fields 
during each nighttime tracking session in both summers. A 
radioed adult male remained on the island until 24 August and 
then moved to the mainland. 



Hourly relocations at Intake during summer 1983 gave no 
evidence of deer crossing the river. However, relocations from 
periodic aerial flights indicated that 2 radioed adult females 
frequently crossed the river channel during summer and/or early 
fall. One, that did not successfully rear fawns during either 
1984 or 1985, was located on either side of the river throughout 
the summer of both years. The other successfully reared fawns 
during the 3 summers, 1984-86, and did not cross the river until 
after mid August in any year. 



61 



During winter, movements of deer at Elk Island were 
generally more extensive than those of summer (Fig. 13) 
reflecting the greater average distance between diurnal bedding 
cover and fields at night. Deer at Intake did not exhibit the 
extensive movements during crepuscular periods during winter that 
were observed at Elk Island (Fig. 13). During winter as in 
summer, deer exhibited high fidelity to specific fields, the use 
of which appeared dependent on the location of diurnal bedding 
sites (Herriges 1986). 

Radio-collared deer were located in agricultural fields 
during an average of 79% (40-100%) of all tracking sessions 
spanning 3 or more hours between 1800 and 0600 hrs. The time 
spent in fields ranged from less than 1 hour to an entire 12-hr 
period. Radioed deer commonly moved into fields at dusk and 
returned to riparian cover before daylight, consistent with large 
numbers of unmarked deer moving at those times. Deer using 
fields within 0.5 km of riparian cover were less likely to spend 
the entire night in fields than those travelling greater 
distances . 

Daily movements at Intake during winter 1983-84 differed 
from those at Elk Island where all radioed deer spent the entire 
diel period on the river bottom. At Intake, 3 of 11 radio- 
collared deer spent the diurnal period on bottomlands and moved 
to fields on upland terraces in late afternoon; 3 others spent 
the diurnal period in upland draws and either traveled to the 
same fields or moved up the draws to feed in fields on the 
adjacent plateau; and the remaining 5 spent the entire diel 
period on the river bottom. Deer travelling to fields in uplands 
tended to move greater distances and remained in fields through 
most of the night, whereas those that remained on the river 
bottom usually traveled shorter distances and fed for shorter 
periods . 

At Intake, deer moved from riparian cover to agricultural 
fields from mid afternoon through evening, while deer at Elk 
Island usually did not move from woody cover until dusk (Fig. 
13). These differences may have been influenced by human 
disturbance. The Elk Island area had a higher density of 
farmsteads and other occupied dwellings than the Intake area, and 
also included an area open to public hunting. These factors, 
singly or in combination, may have modified daily 
movement/activity patterns of deer at Elk Island. Vogel (1983) 
reported a shift in activity patterns of deer from diurnal to 
nocturnal as a result of disturbance associated with homesite 
development. Although intensive farming and associated human 
activity did not preclude deer use of developed portions of the 
floodplain, it apparently influenced the time of day that 
nonforested areas were used. 



62 



Home Range Characteristics 

Yearlong Ho me Ranges .-- Life home ranges, calculated from 
daytime relocations, for deer 2 years of age and older were 2.52 
and 5.24 km^ for females and males, respectively (AARs: 0.47, 
0.91 km). There was no difference between the adult life AAR and 
the mean annual AAR for all years in which individuals were 
monitored (P > 0.1 0). 

Mean annual home range size for deer with overlapping 
seasonal ranges was smallest for females 2 years and older and 
largest for yearling and older males (Table 24). Differences in 
home range by sex and age of deer were highly significant 
(P_ < 0.01): adult females < yearling females < all males ( 1 +). 
Annual home ranges of yearling and adult males were not different 
(P > 0.10). Population size had no discernable affect on annual 
home range size for any sex and age class. 



Nine rad io- co 1 la red yearling and older females occupied 
discrete summer and winter home ranges; 6 were monitored for 
periods exceeding one year; 4 occupied the same discrete summer 
and winter ranges each year. One, that maintained the same 
winter home range during each of 3 years, summered in upland 
prairie as a 3-year-old before establishing a permanent summer 
home range on the river bottom. Another, that occupied the same 
summer range for 4 years, wintered in uplands as a yearling 
before establishing a permanent winter home range on the river 
bottom. 



Among females, portions of annual home ranges occupied 
during autumn and spring were not used as consistently as areas 
occupied during summer and winter. Some adult females remained 
on s umme r / wint e r home ranges during those seasons, while others 
used areas common to neither seasonal home range. Hunting 
undoubtedly influenced portions of annual home ranges occupied 
during autumn. A r a d io- c o 1 la red female, whose home range 
included a portion of the EIWMA, occasionally moved to the island 
in autumn during 1 of the 4 years she was monitored. Other 
marked deer exhibited similar variability in autumn movements. 
One female with discrete summer and winter home ranges made 
several exploratory movements to her summer home range during 
April-May in one year but remained on her winter home range 
through May of the next year. 

Although some dichotomy in seasonal use of annual home 
ranges was apparent among adult (2+) males, most ranged widely 
and could be found in any portion of the annual range at any time 
of the year. The high mobility and variability in movements of 
yearling males from June-November indicated that most did not 
establish traditional summer home ranges immediately following 
breakup of family groups. By December, movments of yearlings 
were similar those of adult males (Fig. 12). Relocations from 
radio-collared yearling males in subsequent years indicated that 
most established permanent home ranges by their second winter. 



6 3 



Table 24. Characteristics of diurnal home ranges of white-tailed 
deer on the lower Yellowstone River bottom by sex and age from 
aerial surveys, 1980-85. 



AAR (km) Polygon (km 2 ) 



Season Sex and Age N a X SE X SE 



Annua 1^ 



Summe r 



Winter 



Fema les 

Yrlgs. 12 0.72 0.16 4.81 2.20 

Ads. 43 0.47 0.03 1.14 0.11 

Males 

Yrlgs. 2 1.04 0.09 

Ads. 5 0.91 0.11 5.24 1.48 

Fema les 

Yrlgs. 18 0.41 0.03 0.44 0.07 

Ad. (2+)101 0.29 0.02 0.19 0.03 

Males 

Yrlgs 6 1.18 0.29 1.11 0.57 

Ads. 27 0.56 0.05 0.79 0.12 

Females 

Yrlgs. 14 0.40 0.06 0.62 0.18 

Ads. 53 0.35 0.02 0.36 0.05 

Ma le s 

Yrlgs. 5 0.41 0.08 0.39 0.16 

Ads. 6 0.91 0.20 2.00 0.86 



N = number of animal seasons or animal years for each sex 
and age class (does not represent total number of animals 
involved) . 

Annual diurnal home ranges includes only those individuals 
with overlapping summer and winter home ranges. 



Annual diurnal home ranges of adult deer (Table 24), 
generally fell within the size range reported for nonmigratory 
white-tailed deer (Marchington and Hirth 1984). They closely 
approximated those reported by Gavin et al. (1984) for the lower 
Columbia River in southern Washington. Lower Yellowstone 
whitetails had much smaller annual and seasonal home ranges than 
reported for other eastern Montana areas including prairie (Wood 



64 



1987) and upland pine (Dusek 1987) habitats. 

consistently occupied 
August and winter home 
February, though dates 
individuals and years. 



Seasonal Home Ranges . — Adult females 
summer home ranges from 1 June through 31 
ranges from 1 December through the end of 
of movement between ranges varied between 



Mean summer home ranges of yearling and older females were 
smaller than those of yearling and older males (Table 24). Adult 
males had smaller home ranges than yearlings. Differences 
between adult females and yearling females, yearling males, and 
adult males were highly significant (P < 0.01). 



Summer home range size among females of age classes 1 
through 3 may have been influenced by relative female density 
(Fig. 14), although differences within age groups were not 
significant (P > 0.10). Controlling for the effects of 
population size, there were differences (P_ < 0.05) in the AARs 
among ages, although age-specific responses to changes in female 
density varied and were not linear (P_ > 0.10, Fig. 14). 

At low female density, AARs suggested decreasing home range 
size with increasing age, but the only difference (P_ < 0.05) was 
between yearling females and those 4 years and older. Home range 
size appeared to increase among all age groups from low to medium 
female density. At medium female density, home ranges of 
yearlings appeared larger than those of all older females, but 
differences were not significant. From medium to high density, 
home range size continued to increase among yearling and 2-3 
year- old females, but a slight decrease was observed among mature 
and old females (Fig. 14). At high female density, differences 
among all age groups were significant (P_ < 0.05). 

Winter home ranges were larger than summer home ranges for 
females and males 2 years and older, approximately the same size 
for yearling females, and smaller for yearling males (Table 24). 
Winter home ranges were larger (P_ < 0.05) than summer home ranges 
only for females 4 years of age and older. 

As during summer, adult females had smaller diurnal home 
ranges than other deer during winter (Table 24). Differences 
between adult males and yearling and older females were highly 
significant (P < 0.01). 

D i e 1 Home Range . --Data from triangulation over 24-hr periods 
at Elk Island and Intake during 1982-84 indicated that nighttime 
movements substantially increased seasonal home range size. The 
mean diel polygon was twice as large as the diurnal polygon 
during summer and nearly 4 times larger during winter (Table 25). 
Despite this, the relationship between sex and age classes in 
diel home range size was similar to that of diurnal home ranges 
from both triangulation and aerial surveys (Tables 24 and 25): 
adult (2+) female home ranges were smaller (P < 0.05) than those 
of adult males, while ranges of yearling females appeared smaller 
than those of yearling males (P = 0.07). Among yearling and 



6 5 



DC 

DC 
LU 



0.6 -i 
0.5 - 
0.4 
0.3 



5 0.2- 
0) 

0.1 H 




s (16) 



low 



MED 
DENSITY 



HIGH 



Figure 14. The relationship between summer home range (AAR) and 
density, and age of female white-tailed deer on the lower 
Yellowstone. Vertical lines represent + SE of the mean. 



older females, diel home ranges differed in size by 
age (P < 0.001). 



season and 



Diel home ranges were markedly larger in winter than in 
summer among all sex/age groups (Table 25). Data from 
t r iangulation during summer, 1982 and 1983, suggested no apparent 
difference in diel home range size between 2-year-old and older 
females between the 2 years. However, population size also 
changed little between those years. Differences in diel home 
ranges of summer for females with and without fawns were not 
apparent . 

Home Range Stability . --The straight- line distance (km) 
between consecutive GACs demonstrated the spatial relationship 
of, or fidelity to, seasonal home ranges for each of 35 females 
monitored for more than one year (Table 26). A distance <C 
seasonal age-specific AAR was interpreted to indicate little or 
no between-year change; 1-2 seasonal AARs represented a minor 
between-year shift; and, > 2 AARs reflected a major shift in 
seasonal home range or dispersal. Median distances for each age 
class illustrated trends during summer and winter (Table 26). 



66 



Table 25. Characteristics of diel and diurnal home ranges of 
white-tailed deer by season, sex, and age on the lower 
Yellowstone from hourly relcations by t r ian gul a t i on , 1982-84. 



Season Sex and Age 



N 



AAR (km) 
X SE 



Diel Home Ranges 



Polygon (km^) 
X SE 



S umme r 



Fema les 

Yrlgs. 4 
Ads. 2 7 



0. 55 
0.3 1 



0. 03 
0.02 



1. 74 
0.87 



0. 22 
0.08 



Males 



Yrlgs 
Ads . 



0. 87 
0.6 3 



0.03 



3. 28 
2.75 



0.40 



Winter 



Fema les 

Yrlgs. 7 
Ads. 11 



0.69 0.11 
0.62 0.08 



2. 76 
1.82 



0. 79 
0.26 



Summe r 



Males 



Yrlgs . 
Ads . 



Fema les 

Yrlgs 
Ad s . 



0. 97 0. 14 
1.18 0.40 



Diurnal Home Ranges 



3 
1 8 



0. 51 
0.24 



0. 09 
0.02 



6. 72 
6.22 



0. 91 
0.37 



1 . 49 
2.45 



0. 24 
0.05 



Ma le s 



Yrlgs 
Ads . 



0. 66 
0.6 5 



0.06 



1. 63 
0.86 



0.18 



Winte r 



Females 

Yrlgs 
Ads . 



0. 52 
0.3 2 



0. 08 
0.05 



0. 84 
0.40 



0. 21 
0.12 



Ma le s 



Yrlgs 
Ads . 



0. 58 
1.14 



0. 29 
0.39 



1 . 38 
0.94 



0. 98 
0.33 



6 7 



Table 26. Fidelity of yearling and older marked female white- 
tailed deer to summer and winter home ranges on the lower 
Yellowstone, 1980-85. 

Median Fidelity by Age Class 
1 yr. 2 yrs . 3 yrs . 4+ yrs . 

Season N Median N Median N Median N Median 



Summer 11 0.36(88) a 10 0.27(86) 11 0.22(79) 16 0.20(71) 
Winter 8 0.73(182) 6 0.29(81) 8 0.47(134) 11 0.32(91) 



Median distance between successive GACs in km (fidelity 
index expressed as a percentage of the AAR for each age 
class) . 



Fidelity indexes (median distances expressed as percent of 
the age-specific AAR) suggested higher fidelity to summer than 
winter home ranges among females of all age classes (Table 26). 
This together with the relatively smaller home range size of 
summer (Table 24) indicated more intensive use of summer than 
winter home ranges. Similar findings were reported by Tierson et 
al. (1 98 5). 

Females in age classes 4 and older generally exhibited 
higher fidelity to seasonal home ranges than younger females 
(Table 26). This was particularly apparent in summer when home 
range fidelity stabilized as females approached 4 years of age. 
Interactions between age, distances between GACs, and season, 
however, were not significant (Chi-sq. = 8.75, 7 d.f., P_ > 0.10). 

Di spersal 

We considered dispersal to be individual movements beyond 
the limits of a home range that exhibited no predictable return 
(Bunnel and Harestad 1983). Our interpretation included 
establishment of a new home range discrete from the original, or 
ancestral, home range. 

Movements were monitored at least through 18 months of age 
for 103 (63%) of 163 fawns captured and individually marked 
during 1980-84. Movements by 9 (17%) of 53 females and and 24 
(4 6%) of 50 males suggested dispersal. The rate of dispersal was 
dependent upon sex (Chi-sq. = 10.12, 1 df, P_ < 0.01). Twenty-six 
deer were known to have dispersed as yearlings. Another 7 were 
not observed on their ancestral home ranges beyond autummn 
following their first birthday but presumably dispersed as 
yearlings . 



68 



Twenty-seven (82%) of 33 dispersers established new GACs on 
the river bottom, 6 moved into adjacent uplands. The tendency to 
remain on the river bottom did not vary between the 2 sexes of 
yearling deer (Chi-sq. = 0.42, 1 df, P_ > 0.10). The directions 
of movement by both dispersers and nond i s pe r se r s from the GAC of 
ancestral home ranges were nonrandom (Chi-sq. = 26.96, 2 8.78, 1 1 
d . f . , P < 0.01). The resulting frequency distributions were 
bimodal on roughly a northeast-southwest axis which conformed to 
that of the river valley. Among yearlings which dispersed along 
the river valley, 13 moved upriver and 14 moved downriver. 

Mean distances between ancestral and new GACs of dispersing 
yearlings were 19.5 km for females and 18.5 km for males; the 
difference was not significant (Chi-sq. - 2.77, 1 df, P_> 0.10). 
Medians were 8.2 and 12.9 km, respectively. The longest yearling 
dispersal was 109.8 km by a female. 

Yearling females appeared to disperse primarily during 
summer and males during late autumn. Among 14 yearlings for 
which month of dispersal was determined, 3 of 4 females dispersed 
during May-August, and 6 of 10 males dispersed during October- 
November. Of the other 19 yearling dispersals, 13 were taken by 
hunters outside ancestral home ranges. Some may have moved in 
response to hunting pressure, although the last observations of 
these deer on ancestral home ranges were made during May- 
September suggesting that some had dispersed prior to hunting 
s easons . 

Dispersal of deer older than yearlings was uncommon. Two 
marked females were known to have dispersed as 2-year-olds. Both 
left the river bottom during May or June. A radio-collared 
female shifted her annual home range 3 km within the Elk Island 
unit when approximately 7 years-old. 

The higher rate of dispersal among yearling males than 
females (46 vs. 17%) agreed with findings of Hawkins et al. 
(1971), Kamme rnie ye r and Marchington (1976), Nelson and Mech 
(1984). Disproportionately high rates of dispersal among 
yearlings, particularly males, has also been documented for mule 
and black-tailed deer (Robinette 1966, Bunnell and Harestad 
198 3 ) . 



HABITAT SELECTION 
Diurnal Habitat Selection 



Deer selected (P < 0.05) riparian forest and shrub cover 
types and avoided (P < 0.05) agricultural and grassland types 

during daytime in all seasons (Figs. 15 and 16). Use of other 
types did not depart from availability (P_ > 0.05) or varied 
seasonally in analysis of use vs. availability (Fig. 15) and/or 
in the relative occurrence of types within 9 -ha cells (Fig. 16). 



69 



100 



Lit 




> 
< 
> 

DO 



0 i ■ • ' i * i ■ • ■ i ■ i 
JUN JUL AUG SEPT OCT NOV DEC JAN FEB MAR APR MAY 

MONTHS 

■ AGRICULTURE 
H UPLAND DRAW 
H SHRUB 

SLOUGH 

□ GRASSLAND 

■ GRAVEL BAR 

□ RIPARIAN FOREST 

Figure 15. Use among cover types (%) relative to availability by 
radio-collared white-tailed deer from point locations from aerial 
surveys along the lower yellowstone River, 1980-86. 



Seasonal variation in association of deer with riparian 
forest and agricultural types during daytime (Figs. 15 and 16) 
reflected the shift in distribution of deer from throughout the 
floodplain during summer to large patches of riparian forest near 
the river in winter. Riparian forest accounted for 75% of 1,394 
relocations during summer and 80% of 887 relocations during 
winter. Agricultural fields accounted for 5% and 3% of all 
relocations during those seasons. All differences were highly 
significant (P < 0.01). 

Habitat cells used by deer during summer contained less 
(P_ < 0.05) riparian forest (43%) than those used during winter 
(51%) and spring (47%). Also, cells occupied during summer, 
autumn and spring included more (P < 0.05) agriculture than those 
occupied during winter (20, 20, and 17% vs. 11%). 

The s umme r- 1 o-wi n t e r shift to riparian forest near the river 
was also reflected in the increased association of deer with 
streambed (water and sand/gravel bars) and with interspersed, 
s uc ces s ional ly-r el a t ed shrublands (Fig 16). Cells used by deer 
during winter and spring contained more (P_ < 0.05) shrubland than 
those used during summer; they also included more (P_ < 0.05) 
streambed than cells occupied during all other seasons. 



70 

60 

50 

40 

30 

2© ' 

10 

0 
70 
60 
50 
40 
30 
20 
10 

0 

70 
60 
50 
40 

30 
20 

10 
0 
70 
60 
50 
40 
30 

20 
10 

0 
70 
60 
50 
40 
30 
20 
10 

0 

re 

s ^ 
a r 



7 0 



ALL DEER ™ EXPECTED 

tm YEARLONG 




AGR RSPFOR GRASS WATER SHRUB UPLDRAW OTHER 

HABITAT 



6. Relative abundance of cover types within 9 
th which deer were associated compared to all 
partitioned by sex and age. 



-ha grid 
eel Is. 



71 



Of the 858 habitat cells on the 3 intensive study units, 
only 29% were occupied by individually marked deer in summer; 
while 27, 22, and 25% were occupied during autumn, winter, and 
spring, respectively. Unused cells contained more agricultural 
fields than those with documented use during all seasons (Table 
27), but the only differences (IP < 0.05) were in summer and 
autumn between cells with no documented use and those with 1-2 
observations. Cells used intensively (5+ relocations) contained 
more (P_ < 0.05) riparian forest and shrublands than cells with 
little or no use during all seasons. They also contained less ( P_ 
< 0.05) grassland than those with occasional or no use during all 
seasons but spring. 

Although deer selected habitats with relatively large 
amounts of riparian cover, habitat diversity and inter spersion of 
riparian with other cover types were important factors 
influencing habitat selection. The number of patches of cover 
and number of cover types were greater (P_ < 0.05) in cells 
occupied by deer than in unoccupied cells during all seasons 
(Table 27). 

Spatial Relationships .- — Although deer exhibited similar 
selection for riparian forest, shrubland, and agricultural types 
at Elk Island and Intake, diurnal relocations from triangulat ion 
indicated important differences in the use of specific 
communities (Table 28). 

Among serai forest communities, deer preferred (P_ < 0.01) 
mature cottonwood at Elk Island during summer; use of younger 
forest communities either did not depart from availability or 
they were avoided. Use of mature willow and green ash did not 
depart (P > 0.05) from availability. Deer avoided all 
agricultural crop types during daytime. 

At Intake, green ash was very highly selected (P_ < 0.0001) 
during summer. All other forest cover types were either used in 
proportion to availability or were avoided (Table 28). The 
absence of a well developed shrub understory in mature cottonwood 
stands may have discouraged use of that community. The apparent 
avoidance of mature willow was an artifact of the distribution of 
radioed deer away from the river where most willow occurred. 
During summer, deer at Intake also selected (P_ < 0.05) alfalfa 
fields during daylight periods, although most use of this type 
occurred shortly before sunset. 

Analysis of grid cells indicated that deer at Intake were 
significantly ( P_ < 0.05) associated with upland hardwood draws 
during summer. These draws occurred in close proximity to the 
river bottom and provided additional mesic sites that were 
attractive to deer and optimized spatial segregation of deer in 
summer, especially during years of high population density. 



0) 








s 








o 








•u 








CO 
















o 








r-l 








r-l 








CU 


0 






r* 


O 




c 








■H 


u 






J-i 








CL 


& 






CO 


O 


+ 






i-l 


m 














(1) 








43 


>. 






4J 


rH 








CU 






a 


> 






o 


•H 








CO 








S3 






r* 


CU 






o 


4-1 




Jj 


o 


a 




CU 


I-l 


•H 




4-J 


43 






c 




73 




•H 


CO 


CD 




ES 


4- 


CO 






1 








ON 










M 






00 


O 






in 








00 
















ng 








CN| 
I 






o 








0 














CO 




CO 


C 






• 


0 


CO 


1-1 




3 


o 


r-l 


CO 


4-J 


•H 


CO 


1 


3 


4J 


c 


o 


<3 


CO 


o 


00 




T-l 


•H 


ON 




u 


CO 


1— 1 




cu 


CO 






u 


o 






o 


o 


u 




CO 


o 


cu 




u 




cu 




CO 


73 


73 






4) 






o 


CO 


73 






3 


CU 




4_) 




M 


t-i 


CO 




u 


cu 


J-I 


/ — N 


CO 


0 


T-l 


o 


0 


0 


43 






3 


CO 






CO 


43 


73 


T-l 






CU 


I— 1 




«4-l 


1-4 


CO 




o 


a 


3 






3 


73 




c 


o 


•H 




o 


o 


> 




CO 


o 


•H 




•H 


c 


73 




t-l 


3 


c 




CO 








a. 


CO 






0 




<4-l 




o 


cu 


o 




o 


•3 










CO 






4J 


c 




• 


CO 


o 






43 


1-t 




CM 


4-) 


4-1 








CO 




0) 


14 


o 




i-i 


CU 


o 




43 


> 


r-i 




CO 


•H 


CU 




H 









I m 



CN 



1 + 



CM 



O 



+ 

m 



CN 
I 



! + 
m 



|CN 
I 



O 



CO 

T-l 

u 

CO 

4-1 4-1 

CO o 

4J CO 

•r-l t-l 

43 CO 

CO 43 

33 u 















PQ 


PQ 




PQ 


PQ 


PQ 


U 


PQ 


< 


O 


O 


00 


O 


ON 


r<- 


co 


O 


r-l 


• 


• 




CN 






I—l 




r-l 


CO 


















PQ 


PQ 


I—l 


PQ 


<3 


PQ 


PQ 


< 


<d 


On 




CO 




00 


O 


oo 


CM 


m 






1—1 




I— 1 








i— i 


CN 


co 
















< 


<3 


o 


< 


<3 


< 


< 


<5 


< 


vO 


CM 










CO 


CN 








nO 


co 


CN 


CN 






r-l 


CN 


co 




pa 


PQ 


U 


PQ 


PQ 


-S3 


PQ 


PQ 




ON 


00 


r-» 


CM 


O 


xt 


O 


ON 


co 






in 


r-l 




i—l 


• 


• 


r<» 














CO 


CO 




pa 


<3 


PQ 


PQ 


< 


< 


PQ 


PQ 


r»» 






CO 


00 


CM 


CO 


O 


On 


r-i 


i— ( 


r-l 








r-4 


. 


• 


r-i 














CO 


CO 




<c 


< 


<3 


< 


< 


<J 


< 


< 


CO 


CO 




r— • 




CN 


\ ) 






vD 


CO 


CN 


CN 






r-l 


• 


• 


NO 














CM 


co 




PQ 


















< 


PQ 


U 


CJ 


PQ 


< 


PQ 


PQ 


ON 


ON 


O 




CO 


O 


I—l 


O 


ON 


00 


r- 1 


r-i 


<fr 


r-l 




r-l 


• 


















CO 


co 




PQ 


< 


PQ 


PQ 


< 


< 


PQ 


PQ 


CO 


ON 


CN 




r». 


CM 


co 


O 


00 




i— 1 


CN 


CO 






r-4 


« 


• 


1-1 














CO 


co 




< 


< 


< 


<3 


< 




< 




NO 


CN 




co 


CO 


CN 


<r 


CO 


CM 


CN 


CO 


CN 


CN 






r-l 


• 


• 


vO 














CM 


CO 






















PQ 


















<J 


PQ 


CJ 


PQ 


PQ 


< 


PQ 


PQ 


00 


co 


r-t 


CN 


CM 


r-l 


o 


r-l 


O 


o 


CM 


i—l 


<f 


r-l 




r-i 


• 


• 


r-l 














CO 






PQ 


<3 


PQ 


PQ 


< 


<J 


PQ 


PQ 


r-l 


00 


I—l 


CO 


ON 


CN 


r> 


ON 00 


<1- 


r-l 


CN 


co 






r-l 




• 


I— 1 














CM 


co 




< 


< 


<J 


< 


< 


< 


< 


< 


o> 


r-l 






CO 


CM 


CO 


m 


CM 


o 


CO 


CN 


CM 






r-l 


• 




vO 














CM 


co 





4-) 

CO 









cu 




r-l 










u 




CU 






CO 




o 




C 






I-l 




En 


CO 


3 






3 


73 




73 


5 CO 




• • 


4-1 


C 


c 


c 


CO 43 




c 


r-l 


CO 


CO 


CO 


tJ CJ 






3 


r-l 


i-i 


rH 


Q 






O 


CO 




43 


rJ 




CO 


i-« 


CO 


CO 


3 


t CU 


• • 


CO 


t-i 


CO 


a 


U 


rH > 


*4-l 


t-l 


bti 


u 


•H 


4= 


a, *h 


o 


< 


< 


o 


pd 


CO 


tO 





CO 
OJ 
43 
O 

4J CO 
CO cu 

0- i Cu 

>s 

4J H 

CO 

4-J tJ 

1- 4 CU 
43 > 
CO O 



14-1 

o 



o 
IX 



CO 

4*5 

CJ 

O 
rH 
PQ 



O 



c 

•H 
43 



• 


w 




CO 






rH 


d 




rH 


V 




CU 


04 I 




O 






U 


4-1 




O 


c 






cu 




CO 


r-l 




M 


CU 




o 


14-4 




o 


*4-4 




f™l 


•H 




43 


73 




& 






•H 


rH 






4J 




CO 


c 




JU 


CO 




it- 1 


CJ 




•H 


•H 




o 


U-l 




a 


•H 






3 




co 


c»n 




4-1 


•H 




CO 


CO 




73 








CU 




4— l 


t 1 




o 


CO 




CO 


CO 




i-i 


S-i 




CU 


CU 




43 


4J 




0 


4-1 




3 


CU 




C 


iH 




4-1 


4-1 




c 






cu 


CU 




CO 


H 




CO 


cu 




t-l 


U-l 




a 


IH 




cu 


•H 




u 


73 




CO 


>% 




cu 


43 




CO 






cu 


73 




43 


CU 




4J 






c 


O 




CO 


rH 




n 


r-4 




CO 


o 




a 


«4-l 




c 


co 






cu 






3 


• 


CO 


rH 


CO 


u 


CO 


c 


cu 


> 


o 


43 




CO 


0 




CO 


3 


o 


OJ 


53 




CO 



73 



Table 2 8. Diurnal and nocturnal selection of bottomland cover 
types by white-tailed deer at Elk Island and Intake determined 
from relocating t ran s m i t t e r ed deer by t r ian gula t i on. 

Elk Island Intake 



Summer Winter Summer Winter 



Cover Type 


A a 


N 


D 


A 


N 


D 


A 


N 


D 


A 


N 


D 


Seedling- 


0 b 




- 


0 


0 




— 


-- 


0 




- 


0 


Wi 11 ./Cott . 


























Y. Cottonwood 


0 


0 




0 


0 










0 


0 


0 


M. Cottonwood 


++ 


++ 


++ 


++ 


+ 


+ 


0 




0 


0 


0 


+ 


M. Willow 


+ 


0 


0 


+ 


+ 


0 












0 


Gr. Ash 


0 


0 


0 


0 


0 


+ 


++ 


++ 


++ 


++ 


++ 


+ 


Shrub 


++ 


++ 


+ 


+ 


0 


0 


0 


0 


+ 


0 


0 


0 


F. PI. Grassl. 


0 




0 


0 


0 


0 


0 


0 


0 


+ 


+ 


0 


Cereal Grain 






0 


0 


0 




0 


+ 


0 


0 


0 


0 


Sugar Beets 








+ 


++ 
















Alfalfa 


0 


+ 


0 




0 


0 


++ 


++ 


+ 


+ 


+ 


0 


a A - diel 


period; 


N 


- nocturnal; 


a nd , 


D - 


diurnal 


* 





(0) indicates no selection (P > 0.05); - indicates 
avoidance (P_<0.05); (+) indicates preference (P_<0.05); 
and ( — ) or (++) indicates highly avoidance or preference 
(P<0 . 01 ) . 



During winter, deer at Elk Island selected (P_ < 0.05) mature 
cottonwood and shrublands (Table 28). The preference (P_ < 0.05) 
for green ash probably reflected increased use of the large 
island where the most extensive stands occurred. As during 
summer, deer avoided agricultural fields during daytime. 

Deer also selected mature cottonwood and green ash 
communities at Intake during winter (Table 28). The occasional 
use of upland hardwood draws during daytime occurred while cattle 
were on bottomlands. Daytime use of agricultural fields occurred 
primarily in late afternoon on sites where cattle had recently 
been supplementally fed and in winter wheat fields on upland 
terraces adjacent to the bottomlands. 



Sex and Age Relationships . — Selection for riparian forest, 
shrubland, agricultural, and streambed types differed (P < 0.01) 
with sex and age of the deer (Fig. 16). There also were sex-age 
related differences in deer use of various serai communities 
within the riparian forest (P < 0.01) (Table 29). 



74 



Table 29. Percent of seasonal relocations of transmittered deer 
by sex and age in individual cover types of riparian forest from 
aerial surveys, 1980-86. 



Riparian Forest Cover Type 



Season, Sex Willow/Cott. Young Mature Mature Green 

and Age Thicket Cott. Cott. Willow Ash 



Summe r : 

Fema les : 

1-2 yrs 
3+ yrs . 

Ma 1 e s : 

1 Yr. 
2+ yrs . 

Aut umn : 

Females: 

1-2 yrs 

3+ yrs . 

Male s : 

1 y r . 
2+ yrs . 

Winter: 

Fema les : 

1-2 yrs 
3 + yrs. 

Ma le s : 

1 y r. 

2 + yrs . 

Spring: 

Females : 

1-2 yrs 
3+ yrs . 

Males : 

1 y r . 

2 + y rs . 



(276) a 4 
(527) 3 



(38) 8 
(180) 8 



(212) 7 

(426 ) 9 

(35) 3 

(132) 8 



(239) 4 
(348 ) 3 



(75) 5 
(60 ) 5 



(222) 2 
(314) 4 



(89 ) 3 
(54 ) 6 



48 
45 



33 
38 



0 
16 



42 
38 



0 
4 



50 
33 



9 
8 



41 
44 



34 
31 



35 
34 



9 

20 



37 
31 



1 6 
12 



46 
55 



12 
11 



23 

19 



33 
28 



36 
47 



1 6 
13 



1 7 

7 



40 
50 



11 
9 



30 
30 



26 
28 



33 
35 



12 
11 



26 
20 



a 



Sample size for each sex and age by season. 



7 5 



Overall, habitats occupied by adult males during daytime 
contained more riparian forest and streambed than those occupied 
by other deer (Fig. 16). Within the riparian forest, adult males 
used the young cottonwood community to a greater extent than did 
females (Table 29). Cells occupied by mature females contained 
more riparian shrublands and agriculture than those occupied by 
other deer. Within the riparian forest, females were more 
associated with mid-to-late serai communities. Tests for sex- 
age differences in association with various habitat components 
indicated that mature females also used habitats with 
comparatively greater diversity and int erspers ion (more types and 
patches of each, P < 0.05) than those use by other deer. 

Patterns of habitat selection among subadult females 
paralleled and were not different (P_ > 0.05) from those of mature 
females, overall or seasonally (Fig. 16). Those of yearling 
males paralleled females during summer and autumn, but more 
closely resembled adult males in winter and spring. 

In summer, habitats occupied by yearling males contained 
more (P_ < 0.05) agricultural fields than those occupied by adult 
males, though differences in use of fields between yearling 
males, subadult females, and mature females were not significant 
(P > 0.05). The close association of these groups with 
agriculture was also apparent in their earlier appearance in 
fields; adult males rarely appeared in fields before dusk. 

During summer, habitat cells occupied by yearling males 
included less (P < 0.05) riparian forest than cells occupied by 
all other deer. In winter, they contained more (P_ < 0.05) 
riparian forest and less shrublands than cells occupied by mature 
females. By spring, there were no differences (P_ > 0.05) in 
habitat use between yearling and adult males. During spring, 
cells occupied by yearling and older males contained less (P_ < 
0.05) agriculture and shrubland types than those occupied by 
mature females. 



Nocturnal Habitat Selection 

Habitat use patterns and preferences after dark differed 
from those of daytime primarily in markedly greater use of 
agricultural fields and selection for specific crop and riparian 
forest cover types (Table 28). At Elk Island, where the most 
complete data were obtained, 27% of all nocturnal point locations 
from t r ian gul a t i o n during summer and 64% of those in winter were 
in agricultural fields. Only 3 and 5% of all daytime relocations 
of radioed deer on the river bottom during 1980-85 occurred in 
fields in summer and winter, respectively. 

Nocturnal patterns in selection for specific agricultural 
field or crop types differed seasonally and between the Elk 
Island and Intake areas (Table 28). During summer, deer selected 
alfalfa fields (P < 0.05) at both Elk Island and Intake. At Elk 
Island, 30% of all scanning circles around nocturnal point 



7 6 



locations contained alfalfa compared to 9% during diurnal 
periods. Spotlight surveys ( Co mp ton 1986a, Herriges 1 986) 
indicated selection of alfalfa fields by deer among crop types. 

At Elk Island, irrigated grain fields were avoided 
(P < 0.05) during summer. The preference for wheat fields at 
Intake (Table 28) appeared largely an artifact of selection by 
deer for alfalfa and riparian cover adjacent to wheat fields. 
The relative amount of edge per unit of wheat field area in 
scanning circles around deer locations was almost twice that of 
random scanning circles suggesting that most of these locations 
were not centered in the wheat fields. Spotlight surveys showed 
these wheat fields to be avoided by deer during summer (Herriges 
19 8 6). 

During winter, alfalfa was selected at Intake (Table 28), 
while its use at Elk Island did not depart from availability 
(P > 0.05). At Elk Island, deer exhibited a preference (P_ < 
0.01) for sugar beet fields. As in summer, the apparent 
selection for grain fields at Intake may have been an artifact of 
selection of nearby alfalfa fields as indicated by spotlight 
surveys (Herriges 1 986). 

Selection of fields during winter appeared to reflect 
availability of crop types and residues. At Elk Island, sugar 
beet fields were highly selected by deer at night even though 
fields used were often more than 1 km from riparian cover used 
during diurnal periods. Few alfalfa fields retained sufficient 
residual growth to attract deer following the harvest of 3 crops 
the previous growing season. Irrigated grain fields were plowed 
during autumn of most years also reducing availabilty of crop 
residues to deer. 

Sugar beets were not grown at Intake, where available crop 
types were limited to alfalfa and grass hay and cereal grains. 
Use of grain fields, including stubble and seeded winter wheat 
fields on upland terraces, did not depart (P_ > 0.05) from 
availability. Selection of floodplain grassland at Intake (Table 
28) was primarily attributable to deer using sites where cattle 
had been fed grain or alfalfa hay. 

Deer continued to use riparian forest during nighttime at 
both study units, especially on bottomlands at Intake where 
fields were lacking or were heavily grazed by cattle (Table 28). 
At Elk Island, mature cottonwood and mature willow were selected, 
while use of other riparian communities did not depart from 
availability. Only green ash was selected during nighttime at 
Intake; use of other riparian communities did not exceed 
availability . 



77 



FOOD HABITS 

Browse and agricultural crops dominated the yearlong diet of 
deer on the lower Yellowstone, accounting for 42% and 41% of the 
total volume of rumens, respectively. Browse received its 
highest use during January (69% by volume) and its lowest use 
during April (10%). Deer used agricultural crops heavily during 
August (41%), October (51%), November (52%), and April (75%) and 
minimally during May-June (8%). Forbs accounted for 7% of the 
yearlong diet; their relative volume in rumens increased 
dramatically from April (3%) to May (19%). Grasses, which 
accounted for 10% of the yearlong diet, were used in appreciable 
quantity only during March (19%). 

Seasonal Trends 

Leaves and the current year's stem growth of western 
snowberry, Woods rose, and willows accounted for nearly one-third 
of the summer diet (Table 30). Alfalfa was the most abundant 
single item contributing nearly one-third of the summer diet by 
volume and occurring in 12 of 15 rumens. No single naturally- 
occurring forb species was identified in more than 3 of the 15 
summer samples, and aster ( Aster spp.) was the only species 
contributing 1% or more of the seasonal diet. 

Neither the average volume of browse nor individual species 
used changed appreciably from summer to autumn. Autumn use of 
western snowberry also included ripened fruit. The only major 
change in the diet from summer to autumn was decreased use of 
forbs with increased use of agricultural crops (Table 30). 
Alfalfa, which was heavily used during summer, was replaced by 
cereal grains and sugar beets during autumn. Yearlong use of 
beets was highest in October, coincidental to harvest, when they 
accounted for 30% of the aggregate average volume of 23 samples. 

Browse dominated the winter diet and consisted of leaf 
detritus, buds, and woody stem tissue of plains cottonwood and 
western snowberry. Use of agricultural crops generally declined 
from autumn with sugar beets still predominating among crop 
residues. Alfalfa received its lowest seasonal use during winter 
and was present in only 16 of 45 rumens. 

Browse received its lowest seasonal use during spring (Table 
30). Species composition of browse changed from woody stem 
tissue and detritus from plains cottonwood during March to green 
leaves and succulent stem tissue of Woods rose and western 
snowberry during May. The average percent by volume of alfalfa 
increased from 3% in March to 44% in April. 



78 



Table 30. Annual and seasonal food habits of white-tailed deer 
on the lower Yellowstone River determined from rumen contents of 
184 deer, 1980-86. 





Annua 1 


Summe r 


Aut umn 


Winter 


Spring 




(184) b 


(15) 


(76) 


(4 5) 


(48) 


Taxa a 


FR/VOL 


FR/VOL 


FR/VOL 


FR/VOL 


FR/VOL 


BROWSE : 

Populus deltoides 
Prunus virginiana 
Rosa woods i i 
Salix spp. 
Symphoricarpos 

occidentalis 
Unidentified Browse 


68/17 c 
12 / 1 
52/ 4 
30/ 3 

66/1 1 
51/5 


47/ 1 
40/ 2 
100/ 8 
53/ 4 

87/18 
60/ 3 


51/ 7 
13/ 1 
50/ 4 
34/ 3 

78/15 
49/ 5 


87/36 

7/tr d 
64/ 4 
24/ 2 

60/1 1 
49/ 5 


83/1 9 
67 tr 
27/ 2 
19/ 2 

46/ 3 
52/ 5 


Total Browse 


99/42 


100/40 


100/38 


98/60 


96/33 


FORBS : 

Unidentified Forbs 


56/ 6 


87/1 9 


58/ 6 


58/ 5 


40/ 4 


Total Forbs 


67/ 7 


9 3/22 


72/ 7 


62/ 6 


52/ 5 


GRASSES : 

Unidentified Grass 


70/1 0 


33/ 2 


68/ 8 


67/ 8 


88/16 


lO I a 1 yl aS oc S 


71 /i n 


^ ^ / 9 
j j l £. 


o o / o 


67/8 


8 8/16 


AGRICULTURAL CROPS: 
Alfalfa 
Cereal Grains 
Sugar Beets 


48/14 
31/10 
36/1 5 


80/3 1 
13/ 2 


51/1 2 
38/12 
41/20 


36/ 8 
18/ 3 
42/13 


42/16 
38/ 4 
33/1 5 


Total Agriculture 


8 0/41 


80/36 


8 6/47 


68/26 


81/46 



a Species comprising less than 3% volume or occurring in 

less than 25% of seasonal samples were omitted but used in 
calculating seasonal totals among forage classes. 

k Seasonal sample sizes are in parentheses. 

c % of seasonal samples that the item occurred in/ % of 
seasonal diet. 



tr = < 1%. 



7 9 



Spatial Relationships 

Specific forage preferences and/or use reflected 
availability as influenced by local land use practices among 3 
segments of river bottom within which rumens were collected: 
Glend ive-Intake; Intake-Savage; and Savage-Sidney. Relative use 
among forage classes between the 3 segments during growing 
(spring, summer, and early autumn) and dormant (late autumn and 
winter) seasons suggested that agricultural crops replaced browse 
in the diet where crop residues were readily available (Table 
31). 

During the growing season, alfalfa was the major item used 
in all 3 segments of river bottom and occurred in at least two- 
thirds of the rumens from each segment (Table 31). Comparatively 
high use of agricultual crops and low use of browse occurred in 
the segment between Savage and Sidney. Five different crop types 
accounted for 4% or more of the diet from Savage to Sidney that 
compared to 2 crop types from Glendive to Intake and 1 from 
Intake to Savage. 

Cereal grains accounted for an appreciably larger 
proportion of the diet during the dormant season in the segment 
of the study area between Glendive and Intake than elsewhere 
(Table 31). Stubble from dryland cereal crops, which occurred on 
the floodplain as well as adjacent upland terraces, was generally 
not plowed during autumn allowing deer access to waste grain. 
These, as well as hay lands, were also grazed by cattle 
throughout much of the period of vegetative dormancy. 

The segment of river bottom between the diversion dam at 
Intake and Savage was characterized by a transition in land use; 
land use practices near Intake were devoted almost exclusively to 
livestock production, but intensive cropping of cash crops was 
progressively integrated into ranching operations downstream 
toward Savage. Many of the fields were occupied by cattle from 
late autumn through early spring. Although both sugar beets and 
alfalfa were grown, their combined use during the dormant season 
was approximately half that in the segment downstream from Savage 
(Table 31). A comparatively larger proportion of the diet 
consisted of browse compared to the intensively cultivated 
floodplain downstream from Savage (53 vs. 34%). 

Agricultural crops dominated the diet of deer throughout the 
year in the segment of river bottom between Savage and Sidney 
which was characterized by intensive cultivation and minimal 
livestock use (Table 31). Sugar beets accounted for nearly one- 
third of the diet of deer during the dormant season and declined 
to half that amount during the growing season. The occurrence of 
cereal grains among rumens from portions of the study area flood 
irrigated from the Lower Yellowstone Project was minimal because 
a u t umn- p 1 ow i ng of stubble minimized the amount of crop residue 
available to deer. Pinto beans were grown only north of Savage 
and, thus, occurred only in rumens from that segment of river 
bottom. 



80 



Table 31. Food habits of white-tailed deer on the lower 
Yellowstone River by river segment during the dormant and growing 
seasons from rumen contents of 185 deer, 1980-86. 



Dormant Season 



River Segment 



1*2 3 
(45) b (3 2) (62) 

Taxa c FR/VOL FR/VOL FR/VOL 



BROWSE : 

Populus deltoides 73/25 d 

Rhus trilobata 11/ 1 

Rosa woods i i 36/ 2 

S a lix spp . 31/4 
Symphoricarpos occi denta lis 5 6/ 9 

Unidentified Browse 47/ 5 

Total Browse 98/50 

FORBS : 

Unidentified Forbs 24/ 2 

Total Forbs 44/ 3 



63/24 
22/1 
63/ 6 
25/ 1 
63/14 
50/ 6 

100/53 



63/ 7 
72/ 8 



73/1 6 
16/tr d 
45/ 2 
27/2 
69/ 7 
60/ 6 

100/34 



61/ 4 
68/ 5 



GRASSES : 

Unidentified Grass 

Total Grasses 

AGRICULTURAL CROPS 
Alfalfa 
Cereal Grains 
Corn 

Pint o Beans 
Sugar Beets 



73/ 9 
73/ 9 



22/ 9 
60/27 
4/tr 

2/tr 



75/14 
75/14 



34/ 5 
22 / 1 
9/ 4 

38/1 5 



66/10 
66/10 



53/1 0 
29/ 5 
11/ 2 
7/ 2 
77/3 1 



Total Agriculture 



71/36 



66/25 



92 /4 9 



81 



Table 31 continued. 



Growing Season 



River Segment 



12 3 
(2 3) (3 ) (20) 

FR/VOL FR/VOL FR/VOL 



BROWSE : 



PoduIus deltoides 


57/ 4 


33/ 2 


65/ 2 


Prunus virginiana 


1 A 1 o 

JO/ z 




OA/ 1 

Z U / I 


Rosa woods i i 


74/ 6 


100/1 9 


55/ 6 


Salix spp. 


35/ 3 




35/ 5 


Sheoherdia argentea 


9/ 1 


67/ 5" 


10/ 1 


S ymohor i ca rpos occidentalis 


83/21 


100/18 


55/ 8 


Unidentified Browse 


JD / J 


1 UU / / 


/. A / 1 

4 0/ 3 


Total Browse 


I (J U / 4 L 


1UU / jU 


Q A / 1 £ 
7 U / Z 0 


F ORB S : 








Le gum inosae 





33/ 1 


_ _ _ 


Meli lotus officinalis 




67/ 1 




Unidentified Forbs 


O 1 1 1 1 


6 / / 1 z 


6 0/1 Z 


Total Forbs 


83/15 


100/13 


75/12 


GRASSES : 








Unidentified Grass 


61/ 7 


100/ 2 


70/ 7 


Total Grasses 


61/7 


100/ 2 


70/ 7 


AGRICULTURAL CROPS: 








Alfalfa 


70/30 


67/34 


75/26 


Cereal Grains 


13/4 




10/ 4 


Corn 


4/ 2 




10/ 4 


Pint o Beans 






5/ 4 


Sugar Be e ts 






25/1 7 


Total Agriculture 


74/36 


67/34 


85/55 



River segments: 1). Glend iv e-Intake ; 2). Intake-Savage; 
and 3). Savage- Sidney 

Sample sizes are in parentheses. 

Species comprising less than 3% by volume or occurring in 
less than 25% of samples were omitted but used in 
calculating totals amonng forage classes. 



d 



% of samples that the item occurred in/ % of diet. 



82 



Sex and Age Relationships 

Forage preference and/or use differed seasonally between 
adult males and females. Forage use by fawns of both sexes 
approximated that of adult females (Fig. 17); data were not 
available for fawns during the growing season. Agricultural 
crops, primarily alfalfa, dominated the diet of adult females 
during Apri 1-September, while adult males used more browse. 
Relative use of f orbs and grasses did not differ between sexes. 
During October-March, males used more crop residues and less 
browse than females. 



Factors Influencing Seasonal Food Habits 

Yearlong food habits of white-tailed deer on the lower 
Yellowstone River apparently were influenced primarily by 
phenological changes in natural forage on the river bottom and 
relative abundance and variety of agricultural crops available to 
deer (Table 30). The relative amount of browse used during 
winter was comparatively less than that reported for bottomlands 
of the Missouri River (Allen 1968) or in mountain foothills 
(Marti nka 1968) where alfalfa was the only major agricultural 
crop grown. Kamps (1969) also reported relatively low use of 
browse in mountain foothills where cereal grains were grown and 
ut il i zed by dee r. 

Seasonal changes in crop types in rumens generally 
corresponded to shifts in relative use of fields (Compton 1986a). 
With the exception of alfalfa during the growing season, use of 
crops was largely of residues remaining after harvesting. 
Agricultural crops primarily replaced browse in the diet and were 
used most heavily where agricultural lands were flood irrigated 
and where livestock use of fields and natural floodplain areas 
was minimal or absent (Table 31). Compton (1986a) reported that 
regular use of individual fields during winter was contingent 
upon availability of crop residues and absence of livestock 
grazing following harvesting of crops. 

Summer food habits of white-tailed deer on the lower 
Yellowstone were similar to those reported by Allen (1968) for 
Missouri River bottomlands in central Montana. Use of natural 
occurring forbs in both riverine environments was similar to that 
in ponderosa pine uplands of southeastern Montana (Dusek 1987) 
but considerably less than in mountain foothills of central 
Montana (Martinka 1968, Kamps 1969). The comparatively low use 
of natural herbaceous material in riverine and pine habitats in 
eastern Montana during summer, along with correspondingly high 
use of succulent browse, may not only reflect influence of 
abnormally dry years during this study but also differences in 
longterra precipitation patterns between mountain foothills and 
other habitats in eastern Montana. Use of alfalfa was high in 
all habitats where it was readily available regardless of the 
relative use of forbs. 



83 



CO 

z 
< 

LL 




CO 
CO 

II 



Oh 



D 
Q 
< 



■b 

H 

Q 
< 



< O 
DC ^ 

n m 



LLI 
CC 
3 

O 

E 
o 
< 



i 

QJ 
4-J 



CO CO 

CO OQ 

< DC 

DC O 

O LL 



CO 

o 






ii 




o< 



CO 
CO 

II 



CO 



03 O 

e n 

QJ bO 

m 

QJ 

4-J > 

rH -H 

3 4-1 

T3 03 

CO 4-1 



c 

03 



OJ 
bO 

> 



QJ o 
1 — I 

03 CO 
0 nd 
O 

4-1 -H 

rH r-l 

D <D 

T3 & 
03 

bO 

- C 

C -H 

15 5-1 

03 3 

<H T3 

>, QJ 

^ a 
o 

CO 4-J 

QJ CO 

CO £ 

CO o 

03 i — I 

rH rH 

O 0) 

a) 

bO 5-1 

o3 qj 

M & 

o o 

M-l rH 

M-| <D . 
O M 

4-) 00 

CU I 

CO C O 

P O 00 

!-l i-H 

• a) 

<-H X) t>> 

o 

QJ xl C 
M QJ 03 

3 rH S 

0.0 >H H 
•H 03 O 
h 4J -o 



84 



Woody stem tissue and leaf detritus have comparatively lower 
digestibility than green or succulent forage with greater 
proportions of readily digestible carbohydrates (Segelquist et 
al. 1 9 72, Harlow et a 1. 1 97 5). Coblentz (1970) and Suring and 
Vohs (1979) suggested a preference for green herbaceous forage by 
deer during winter and that heavy use of browse may occur only in 
the absence of green herbaceous forage. 

The observed pattern in forage use between sexes of adults 
may have reflected energetic costs during late gestation and 
lactation among females and the period following breeding among 
males. Because fawns accompanied adult females through the 
period of vegetative dormancy, their foraging habits would be 
expected to be similar. 



85 



DEER -HABITAT RELATIONSHIPS AND POPULATION REGULATION 

W h i t e- t a i 1 ed deer-habitat interactions and relationships 
along the lower Yellowstone River can be examined from two 
broadly different perspectives: 1) their role in determining 
general patterns of distribution and abundance, and 2) their 
influence on deer population dynamics. The former addresses how 
populations have come to be what they are and their general 
status today; the latter how they operate. 

Habitat Influences on Distribution and Abundance 

The current widespread distribution and abundance of white- 
tailed deer on the lower Yellowstone and other riverine habitats 
of the northern Great Plains apparently reflect significant 
changes in the characteristics of these habitats. Historical 
records suggested that whitetails occurred only sporadically in 
low numbers within a habitat-environmental complex that was very 
different from today. 

Mid-to-late 19th century descriptions and photographs (e.g., 
Brown and Felton 1955) indicate a sparsely wooded floodplain with 
little or no shrub understory similar to that described by Crouch 
(1979) for the South Platte River. Such conditions apparently 
resulted from heavy use by large numbers of bison and other 
native ungulates, highly variable subsurface water flow, frequent 
fires, and natural succession toward climax grasslands. 
Intensive grazing of riparian bottomlands has been shown to alter 
stream morphology (Platts 1979) and drastically reduce or 
eliminate understory cover (Severson and Boldt 1978, Crouch 1982, 
Smith and Flake 1983). Severson and Boldt (1978) also indicate 
that growth of woody vegetation on tree-dominated floodplains 
"was almost certainly limited to some extent by fire." 

Within that vegetation complex, the riparian forest and 
shrub habitats favored by whitetails were likely very limited. 
Without irrigation in a region characterized by variation in 
climatic and other environmental conditions, resource outputs and 
availability would have varied extremely. In addition, the 
presence of large numbers of bison, elk, and other wild ungulates 
probably imposed competitive resource partitioning among species 
and further limitation on resource availability and use by 
whitetails. Similarly, the presence of predators, including 
wolves, could have excluded deer from areas lacking suitable 
cover. Under such conditions white-tailed deer could have been 
relegated to a relatively narrow niche with wide spatial and 
temporal fluctuations in distribution and abundance. 

Elimination of bison and extirpation or reduction in the 
numbers of other wild ungulates in the region, settlement and 
development of agriculture along rivers, and construction of 
irrigation systems apparently all contributed to development of 
more extensive, diverse habitat along the floodplain. Crouch 
(1979) attributed development of a 1-km-wide belt of riparian 



86 



forest with a dense shrub/herb understory on the South Platte by 
the mid 20th century to similar physical and biological factors 
associated with settlement and subsequent trends in land use. 
Water management that provided yearlong surface and shallow 
subsurface water flow was especially important. The construction 
of large-scale irrigation systems along the lower Yellowstone may 
have resulted in shallower and more stable subsurface water flows 
that enhanced cottonwood survival and shrub/herb growth on 
bottomlands. However, agricultural practices that precluded or 
limited livestock grazing primarily to autumn and winter also 
abetted development of the cottonwood sere. 

These changes provided more extensive habitat and a broader 
ecological niche for whitetails. Together with predator control 
and restricted hunting, they set the stage for rapid expansion of 
populations by the 1940s and 1950s. Today, deer occur more or 
less continuously throughout the length of the Yellowstone and 
other major rivers and tributaries in eastern Montana. Although 
populations vary spatially and fluctuate somewhat over time, 
densities are generally high. 

Mackie (1983) related high density and stablity in mule deer 
populations in the northern Rocky Mountains and Great Plains to 
complex, diverse habitats with stable resource outputs; 
progressively lower densities and greater variability were 
associated with progressively simpler and more variable 
environments. These relationships may also apply to white-tailed 
deer in the northern plains and explain temporal and spatial 
variation in population densities along the lower Yellowstone. 

Riverine environments in the northern plains today represent 
microcosms of relatively high complexity, diversity, and 
stability within a region characterized by relative simplicity 
and high variability in environmental conditions. High 
complexity and diversity are provided by i nt e r s pe r s io n of 
relatively small units of many different vegetation types or 
communities, land uses, and agricultural practices. Stability is 
inherent in the sense that, on the average, resource availablity 
and outputs remain high and predictable over time within small, 
local areas. Gavin et al. (1984) imply similar interpretation in 
referring to stability or constancy of deer habitats along the 
lower Columbia River. 

White-tailed deer densities on our study area, that ranged 
from 1.4 to 26/km^ in autumn, were substantially higher than 
estimates for nearby upland habitats during 1980-86. Autumn 
densities for a population resident to a prairie habitat never 
exceeded 1 /km^ (Wood 1987). Density of another population on 
p r a i r i e- ag r i c ul t u ra 1 habitat was estimated as 5/km during spring 
1983, when whitetails were generally reaching a recent, if not an 
all-time peak in numbers throughout eastern Montana (Dusek et 
al., unpub 1.) . Deer densities also varied spatially within our 
study area — from less than 15 to approximately 50 deer/km 
among segments of river bottom during 1 985 (Compton et al. 1 988). 
Differences in the relative abundance of deer along the river 



87 



bottom were directly related to the abundance of riparian forest 
and shrub habitat. The relative abundance of riparian cover may 
also have been an important factor influencing differences in 
deer densities between the river bottom and prairie/prairie- 
agricultural habitats. Riparian draws comprised only 3% of the 
prairie habitat (Wood 1987) and 7% of the prairie- agricultural 
area (Dusek et al., unpubl.), while riparian communities covered 
24-40% of our 5 intensive study units (Table 3). However, these 
relationships between amount of riparian cover and deer density 
may be somewhat misleading. They could also be indicative of 
broader habitat/environmental differences that collectively 
influenced the occurrence and abundance of deer within and 
between the river bottom and uplands. 

Development of riparian vegetation along the river followed 
the cottonwood, willow, and green ash seres that originated on 
small sand-gravel bars. As the amount of riparian cover 
increased, complexity and diversity also increased as a result of 
increased numbers and i n t e r s pe r s ion of small stands of all serai 
stages. Bottomlands characterized by extensive riparian forest 
and shrublands were typically interspersed by irrigated 
agricultural fields varying in size, crop type, and cropping 
practice. Riparian cover in the uplands consisted largely of the 
green ash type (Nelson 1961) in narrow stands along draws and 
ephemeral stream courses. These stands were only broadly 
interspersed by adjacent mixed- or bun ch g ra s s- pr a i r i e , dryland 
grain, or, occasionally, dryland or s t rearabo t torn alfalfa fields. 
Increased amounts of riparian cover between areas or habitats 
largely reflected increased t op og ra ph i c / phy s i og ra ph i c diversity. 

The importance of diversity and stablility in the occurrence 
and abundance of whitetails along the lower Yellowstone was also 
attested by high fidelity to bottomlands as compared with only 
sporadic, seasonal use of more raonotypic uplands and ephemeral 
tributary drainages. It was also evident in limited daily and 
seasonal movements and relatively small home ranges on the river 
bottom. Uplands and drainages on our study area apparently did 
not fully meet diel or yearlong habitat needs. Their greatest 
and most consistent use occurred on the Intake area, where wooded 
draws and croplands were interspersed on benchland or terraces 
immediately adjacent to patches of riparian forest on the river 
bottom. Here, some deer bedded in draws during daytime and moved 
short distances to feed on bottomlands at night; during winter, 
others moved from bottomland riparian cover onto terraces to feed 
in grain fields. 



Gavin et al. (1984) suggested that the sedentary nature and 
small home ranges of adult females were tactics reflecting 
adaptation to a stable habitat-environmental regime under which 
the most efficient behavior would be to remain on a familiar 
piece of range large enough to provide adequate food and cover. 



Our findings generally agreed with that interpretation -- high 
fidelity of adult females to small home ranges on bottomlands was 
clearly related to diversity and intersper sion of riparian and 



88 



agricultural communities providing high quality forage and cover 
on small areas. Habitats with high plant species diversity, 
including dicotyledonous forbs and shrubs, provides the greatest 
option for selective feeding by largely concentrate selectors 
such as the white tail (Klein 1 985). A pattern of habitat usage 
that included relatively dispersed distribution and limited 
movement within high quality forage-cover complexes during late 
spring and summer could optimize opportunity to sus tain high 
female densities and successfully rear fawns. As with roe deer 
( Capreolus cap reo lu s ) (Klein and Strandgard 1972), the quality of 
local food-cover complexes available determines the minimum size 
of home range needed by white-tailed deer; the amount and quality 
of habitat overall determines how many home ranges an area may 
support. 

Behavioral spacing of females in a complex and diverse 
environment during a period of optimal resource availability and 
output may allow white-tailed deer to attain comparatively high 
population density as postulated for waterfowl (Patterson 1976). 
Agonistic interaction among adult females prior to and during 
fawning may serve as a spacing mechanism (Gavin et al. 1984) and 
lead to dispersed distribution of females throughout available 
habitats. Interaction between mature females and other deer may 
also facilitate habitat or resource partitioning resulting in 
occupancy of the choicest, most productive habitat by mature 
f eraa les . 

The relatively dispersed distribution of deer on our study 
area during summer reflected a dichotomy in habitat selection, 
and perhaps different habitat requirments between adult females 
and males. Differences in food habits probably reflected 
different physiological cycles between male and female deer (Fig. 
17) as reported for other cervids (Watson and Staines 1978). 
Dense overhead cover was selected yearlong by all deer, but 
comparatively greater use of mature cottonwood stands by females 
and younger serai communities by males suggested that a dense 
understory was less important to males. The ecosystem nutrient 
content increased through succession, peaked in the mature 
cottonwood sere, and declined with the senescence and 
d i sa p pea re nee of the cottonwoods (Boggs 1984). Females were more 
closely associated with agriculture as indicated by habitat 
selection and food habits during the growing season. The 
vegetative and structural diversity, that characterized habitats 
selected by adult females, emphasized the importance of edge. 
Williamson and Hi r t h (1985) correlated edge with an optimal 
foraging strategy that was consistent with relatively limited 
movements and use of small seasonal home ranges by adult females 
on the lower Yellowstone. 

Relatively lower fidelity to home ranges (Table 26) and 
greater mobility of subadult than mature females suggested the 
former may have been excluded from optimal habitat complexes. 
Yearling males may also have occupied suboptimal habitats during 
early summer, as suggested by patterns of habitat use, high and 
variable movement indexes, and variability in body and antler 



89 



growth. While summer movements of yearling males were generally 
greater than those of subadult females, habitat preferences were 
more similar to young females than to adult males. Forage-cover 
complexes undoubtedly varied in quality both spatially and 
temporally as influenced by environmental variability and/ or use 
by livestock. Compton (1986b) indicated that heavy grazing 
reduced shrub cover, primarily that of snowberry and rose, 
regardless of season, though summer grazing apparently reduced 
shrub cover to a greater extent than winter grazing. 

Assuming that longterm seasonal use of pastures by livestock 
physically and biologically altered habitats preferred by deer 
(Severson and Boldt 1978, Smith and Flake 1983), nongrazed areas 
may have favored establishment of home ranges that socially 
dominant females could successfully occupy with predictability 
over time. In segments of river bottom receiving extensive 
autumn-winter livestock use, adult females attempted to 
compensate for physical alteration of mid- serai riparian 
communities by using sites near agricultural fields during summer 
that structurally met requirements for successful fawn rearing. 
Huegel et al. (1986) characterized fawn bedsites in the Midwest 
to include a tall, dense herb/ shrub understory and dense tree 
overstory, irrespective of species composition, that provides 
visual isolation and comparatively cool microenvironments. The 
greater mobility of adult males than females and a more general 
pattern of habitat selection perhaps rendered them less sensitive 
to the effects of livestock. 

The relationship between deer and agriculture, that included 
temporal and spatial variation in seasonal use of crops and 
movement between cover and fields, reflected relative 
availability of forage crops as influenced by cropping practices 
as well as seasonal changes in the behavior of deer. Alfalfa, 
the principal crop type used in summer, occurred throughout the 
river bottom often in the same fields for several years in 
succession. Because of this, and the dispersed distribution of 
deer during summer, alfalfa fields tended to be used more or less 
"traditionally" as an integral part of the summer home ranges of 
deer associated with adjoining riparian cover. Food habits of 
deer during the growing season (Table 31) indicated that alfalfa 
comprised approximately one-third of the summer diet throughout 
the river bottom study area. Diel movement patterns at Elk 
Island and Intake were most similar during summer (Fig. 13); both 
suggested minimal spatial variation in the relationship between 
deer and agriculture during that season. Intensities of use of 
individual fields varied with the amount and location of adjacent 
riparian cover. Generally, the relatively dispersed and stable 
use of alfalfa fields by individual and small family groups of 
deer associated with adjacent cover minimized opportunity for 
intensive use of individual fields except where isolated fields 
occurred in proximity to high quality riparian cover in which 
deer densities were locally high. 



90 



The relationship between deer and agriculture changed 
greatly from summer to winter when deer were distributed 
primarily in large stands of riparian forest near the river. 
Harvest and post-harvest treatment (plowing, grazing, etc.) of 
fields greatly affected the kinds, amounts, and distribution of 
crop residues available to deer through the winter as also 
reported for agricultural habitats in the Midwest (Murphy et al. 
1965). Distribution of crop types used by deer during winter, 
that included sugar beets, small grains, and corn, varied 
spatially along the river bottom and was reflected in seasonal 
food habits (Table 31). Because of this, and the fact that these 
crop types were rotated among fields between years, winter use of 
fields tended to be more o p p o r t uni t i s t i c, intensive, and variable 
than summer use. 



Habitat Influences on Population Dynamics 

During winter, diel movements, habitat use, and food habits 
suggested a strategy of survival involving high energy input 
through selective foraging. Morphological attributes of deer 
allow them to increase selectivity as long as benefits outweigh 
energetic costs (Hanley 1982). Thus, such a strategy implied 
abundant and/or diverse forage resources (Klein 1985) and may be 
typical where deer have access to agricultural crop residues 
during winter. 

Seasonal trends in condition, mortality, and reproduction 
indicated that white-tailed deer on the lower Yellowstone 
generally existed on a high nutritional plane, especially during 
winter. Measurements of diasteraal and antler growth suggested 
that yearling males from the Yellowstone Basin existed on a 
higher nutritional plane than those in pine habitat in 
southeastern Montana (Swenson and Stewart 1982). 

Females and fawns on the lower Yellowstone River maintained 
their respective body weights of late autumn through mid- to-late 
winter, while yearling and older males experienced weight loses 
corresponding with fasting during the rut. These trends were 
similar to those reported for supplementally fed deer (Ozoga and 
Verme 1982). Nutritionally stressed fawns experience substantial 
overwinter weight loss (Severinghaus 1981). The seasonal pattern 
of accumulation and mobilization of body fat, as expressed by the 
KFI (Table 8), closely paralleled that of whole weights (Table 
7). The generally low rates of mortality among all sex and age 
classes during winter-spring (Table 16), and the lack of a direct 
correlation between winter severity and mortality, also indicated 
that deer wintered on a relatively high nutritional plane. 

Strong fidelity to summer home ranges and resource 
partitioning that provided stable and diverse cover/forage 
complexes was probably important in minimizing variation in 
reproductive performance among adult females. During late 
spring-early summer, when social interaction among females was 
most intense, availability and use of naturally occurring 



91 



herbaceous forage was at a yearlong high. The availability of 
irrigated alfalfa on the river bottom may buffer shortages of 
natural forage during dry years on reproduction that may 
contribute to higher reproductive output on the river bottom than 
observed on prairie rangeland by Wood (1987). The comparatively 
close association of females with agriculture during summer 
optimized opportunity for selective foraging when the 
physiological demand on lacating females was highest (Short et 
al. 1969). 

Socially influenced resource partitioning may allow dominant 
females use of habitats during spring-summer that assure optimal 
reproductive output often at the expense of younger females being 
forced into habitats with marginal forage/cover quality. The 
high association of yearling males and subadult females with 
agriculture during early summer may reflect interacting 
influences of resource partitioning, yearly variation of forage 
quality/abundance, and adult female density. Although all 
females were closely associated with agriculture during summer, a 
stategy favoring exclusive use of the choicest forage/cover 
complexes by socially dominant females, when physiological 
demands were highest, was consistent with high reproductive 
performance among mature females (3 through 7 yrs.), particularly 
age classes 6 and 7 (Table 14). 

The comparatively high ovulation and fetal rates of adult 
(2+) females from the lower Yellowstone overall (Tables 11 and 
12) compared favorably with those of supplementally fed females 
that were on a high nutritional plane (Verme 1969, Ozoga and 
Verme 1982). The high ova loss among females of 4 years and 
older was consistent with the "flushing" phenomenon (Ransom 1967, 
Roseberry and Klimstra 1970, Mansell 1974), suggesting that 
nutritional quality prior to breeding was sufficient to insure 
optimal reproductive potential among mature females. 

Increasing social competition among females with density may 
explain yearly variation in recruitment and the greater 
sensitivity to variation in female density among ages classes 2 
and 3 with respect to their reproductive performance. The 
interacting influence of density and environmental variability 
may also explain the more variable pregnancy rates among yearling 
females. Despite comparatively low reproductive potential among 
young adult females, about two-thirds of the 2 year-old females 
successfully reared fawns. Winter forage resources were 
obviously adequate to allow young females to successfully rear 
fawns, often at the expense of immediate body growth (Dusek et 
al. 1 98 7 ). 



Differences between pregnancy rates of yearling and older 
females were probably nutritionally influenced. More variable 
pregnancy rates among yearling than older females (Table 10) may 
have reflected yearly variation in availability of native 
herbaceous forage during summer. Years of relatively low 
pregnancy rates among yearling females generally followed years 
with abnormally dry growing seasons (Fig. 2). Lower pregnancy, 



92 



ovulation, and fetal rates suggested nutritional stress prior to 
breeding following growing seasons characterized by markedly 
below normal precipitation. Such pronounced effects of reduced 
nutritional plane have been documented for both free-ranging and 
confined populations of white-tailed deer (Morton and Cheatum 
1946, McGinnes and Downing 197 7, Kie et al. 1980). A greater 
sensitivity of the youngest breeding age class of females to 
influences of nutrition on reproduction was reported for deer 
(Julander 1961, Ransom 1967, Verme 1969) and other ungulates 
(Bergerud 1971, Skogland 1985). 

The slight decline in pregnancy rates among subadult females 
from 1983 to 1984 (Table 10) coincided with a decline in body 
size among yearling males as expressed by diastemal length (Table 
9). Body growth among yearling males appeared influenced by 
interaction of adult female density with environmental factors 
influencing forage production. If growth among yearling females 
was similarly affected, such interaction may also explain lower 
and more variable pregnancy rates among this age class. 
Nutritional stress may have inhibited physiological development 
among some females whereby they failed to conceive during the 
following autumn breeding season. A relationship between body 
weight and fertility was reported for young red deer ( C . e laphu s ) 
hinds reflecting increased forage competition at high densities 
( Albon et al. 1 98 3). 

Lower longterm rates of recruitment at Intake than at Elk 
Island may have been influenced by habitat conditions related to 
grazing by livestock, diversity and availability of agricultural 
crops, and female density. Reproductive performance at Intake 
was characterized by lower fetal rates and a larger proportion of 
summer fawn mortality than at Elk Island (Table 15). There may 
have been fewer sites favoring successful recruitment of fawns to 
autumn at Intake, a situation severely aggravated at high female 
density. Deer density was higher at Intake than at Elk island 
from 1981 to 1984 despite the more widespread presence of cattle 
and lower consumption of agricultural crops by deer at Intake. 
However, when female density was markedly reduced in both areas 
in 1984 and 1985, recruitment rates at Intake did not increase to 
the extent of those at Elk Island. 



The mechanics of density dependence on reproductive output 
apparently differed between the respective s ubpop ul a t ions . 
Comparatively high ovulation and fetal rates for both yearling 
and older females at Elk Island (Tables 12 and 15) suggested 
minimal nutritional deficiencies with little influence on 
potential productivity. Thus, net productivity was almost 
entirely influenced by adult female density as it influenced 
survival of fawns to 4 months of age. 

Conversely, the lower pregnancy rates (Table 4), as well as 
lower fertility and fecundity (Table 12), at Intake were probably 
nutritionally influenced as a result of interaction between total 
numbers of adult females and environmental variables influencing 
quantity and quality of forage resources. Markedly lower rates 



93 



of recruitment among all age classes of producing females during 
all years combined may have reflected higher female density at 
Intake. Demographic differences in reproductive potential or 
recruitment of fawns to autumn between subgroups of a population 
may reflect environmental variation or differences in density, 
neither of which is mutually exclusive (Dapson et al. 1979). 



Population Regulation 

Traditional theory concerning deer-habitat relationships and 
population regulation has centered on the concept of a more or 
less determinate carrying capacity (K) linked to the standing 
crop of forage through density-dependent interaction (Caughley 
1976, McCullough 1979). A general interpretation related deer 
numbers and dynamics to supply and optimum utilization of winter 
forage in northern environments (Dasraann 1971, Severinghaus 
1979), although summer forage quality may similarly affect 
numbers and dynamics in hot, dry environments (Anthony 1976, Kie 
et al. 1980). McCullough's (1984) "black box" (stock 
recruitment) models for white-tailed deer prescribe a direct 
density-recruitment feedback relationship through resource 
availability influenced by deer use. Implicit in this 
relationship are decreasing animal size and condition and 
increasing mortality with increasing population size. Our 
findings provide basis for evaluating the application of these 
concepts to habitat relationships and regulation in white-tailed 
deer populations along the lower Yellowstone River and in other 
riverine environments in the northern Great Plains. 



The general applicability of traditional theories has been 
questioned with regard to the role of predation (Peek 1980, 
Gasaway et al. 1983) and environmental variation (Wehausen et al. 
1987, Wood 1987). Several alternative hypotheses integrate 
population regulation with social behavior. A "social fence" 
hypothesis (Hestbeck 1982) involves regulation through emigration 
at low densities and forage exhaustion at high densities. Gavin 
et al. (1984) implicated availability of suitable home ranges as 
a limiting factor, but once filled, females would be subject to 
the effects of habitat deterioration. Christian and Davis (1964) 
documented substantial mortality of Sitka deer (C_^ nippon ) 
indirectly resulting from an endocrine response to density- 
related social stress, a phenomenon that has not been documented 
for indigenous North American cervids. 

Our findings were not consistent with a hypothesis of 
regulation through winter forage carrying capacity. Strategies 
of habitat use, biological parameters, and population 
characteristics and dynamics all indicated that neither quality 
nor quantity of winter habitat limited population size. A 
strategy of selective foraging during winter, attended by 
extensive daily movements from riparian cover to agricultural 
fields and negligible overwinter mortality indicated that the 
population wintered on a high nutritional plane. Availability 
and use of agricultural crops by deer during autumn-winter did 



94 



not support the traditional concept of "limitation" through 
winter forage resources. Browse species used by deer during 
winter, primarily cottonwood and snowberry, were not limited in 
distribution and/or abundance. Deer appeared to be relative 
specialists in their use of browse species and agricultural crops 
during winter. This was a departure from the traditonal "forage 
generalist" strategy of deer in northern climates during winter 
(Nud d s 1 980). 



Inf luence o f Density . — Density dependence on the lower 
Yellowstone appeared to manifest itself primarily in the rate of 
recruitment of fawns to autumn that varied inversely with adult 
(2+) female density. In contrast to Gavin et al. (1984), 
comparatively low fawn survival at high female density resulted 
from social interaction between females rather than from a 
deteriorating influence of deer on their habitat. Social 
interaction may have been a predisposing factor in summer fawn 
mortality. A similar hypothesis resulted from experimentation on 
an enclosed population of white-tailed deer maintained on a high 
nutritional plane (Ozoga et al. 1982, Ozoga and Verme 1982). 
However, they did not relate observed phenomena to physical and 
biological characteristics of the habitat. 

The matriarchal social organization of white-tailed deer 
(Hawkins and Klimstra 1970, Hirth 1977) was apparent from the 
comparatively small movement indexes and comparatively high home 
range stability among mature females during summer. Ozoga et al. 
(1982) hypothesized that parturition territoriality enhances 
reproductive success of matriarchal does presumably at the 
expense of younger socially-subordinate parous does. Ozoga and 
Verme (1982) also attributed density dependent mortality of fawns 
during summer to failure of proper imprinting and/or abandonment 
by socially subordinate females. Reproductive success on the 
lower Yellowstone varied inversely with abundance of adult (2+) 
females regardless of age, though changes in density apparently 
had greater influence on fawn rearing success among 2- and 3-year 
old females. Ozoga et al. (1 982) and Ozoga and Verme (1 982) 
indicated only younger females were affected, whereas our data 
indicated that ability of prime age and older females to 
successfully rear fawns was also affected at high density. 

It is highly probable that 2- and 3-year old females, by 
virtue of their subordinate social status, may experience 
comparatively more difficulty rearing fawns than older females 
(Ozoga and Verme 1986). This may have contributed to 
reproductive failure in successive years among some younger 
females, a phenomenon not observed among prime aged (4-7 yrs.) 
females. 



We have no evidence to suggest that predation directly 
limited population growth or that its role varied with population 
density. Studies in Texas (Cook et al. 1971, Be a som 1 974, 
Carroll and Brown 1 977 ) singled out d ens i t y- r e 1 a t ed predation on 
newborn fawns as a factor that may stabilize or limit population 
size in the seraiarid southwestern region of the U.S. More 



95 



favorable cover conditions at bedding sites contributed to 
somewhat lower fawn mortality in the agricultural Midwest, though 
predation was also a factor (Huegel et al. 1985). In New York, a 
rate of fawn mortality, similar to that in the Midwest, was 
reported by O'Pezio (1978) that was not influenced by predation 
or human interference. 

We also have no evidence to indicate that density dependence 
manifested itself in adult mortality or an endocrine feedback 
between social behavior and population size as hypothesized by 
Christian and Davis (1964). Experimentation with a confined herd 
of white-tailed deer provided with excellent nutrition indicated 
no density-dependent response in adrenal weight or secretory 
activity up to densities of 63 deer/ km 2 (Seal et al. 1983) that 
exceeded densities observed during this study. 

Our data suggest that feedback between social behavior and 
population size was related to allocation of forage/cover 
resources during summer as reflected in reproductive potential 
and recruitment. Such a strategy favors minimal fawn recruitment 
at habitat "fill" and a "compensatory" response in recruitment 
following a reduction in female density by either natural or 
artificial means as reported by Swenson (1985) for mountain goats 
( Orea m nos a mericanus ). Habitat fill is defined as the relative 
numbers or density of deer a habitat may sustain over time and 
space including all biological and physical attributes of a given 
habitat that influence deer numbers (Mackie 1983). 



The absence of behavioral spacing of females during summer 
could result in less than optimal recruitment of fawns to autumn. 
Adams (1960) speculated that a breakdown in spatial separation, 
attributable to overcrowding and subsequent displacement of deer 
into marginal habitats, may be a density dependent factor 
limiting population growth. At low density, all females on the 
lower Yellowstone, despite behavioral spacing, were able to 
occupy optimal habitats, and fawn recruitment to autumn was high. 
At medium densities, many young adult females occupied marginal 
habitats, and thus, lost one or both fawns. At high densities, 
not only did the rate of fawn recruitment decline among young 
adult females, but increased social interaction resulting from 
greater mobility of these younger females probably inhibited 
reproductive success of prime-age and old females. 

The role of maternal investment in female offspring through 
the first 3 years of life in maintaining population stability was 
consistent with hypotheses of Gavin et al. (1984) and Clutton- 
Brock et al. (1983) in its implication that regulation occurs at 
the level of the matrilineal group. The selection process 
perhaps favored females that successfully occupied small home 
ranges in vegetatively and structurally diverse habitats. 
Consequently, they reared more female fawns as a function of 
total fawn survival. Subsequent generations of females in those 
matrilineal groups, with increasing age and experience, also used 
those sites. Exceptions were those females making extensive home 
range shifts as yearlings or as 2 year-olds that included 



96 



dispersal from the river bottom environment. Tolerance for 
subadult female offspring around the periphery of their territory 
by female black bears ( Ursus am ericanus ) was offered by Rogers 
(1987) as an explanation for a comparatively lower rate of 
dispersal among young females than among males. Female 
philopatry may also explain an apparent decline in summer home 
range size among mature female white-tailed deer from medium to 
high female density. 

Influence o f Dispersal . — Although yearling males were more 
prone to disperse from ancestral home ranges than females, the 
low rate of dispersal of individuals from the river bottom study 
area did not measurably influence population size or the sex 
ratio of yearlings. Over 80% of yearling dispersals on the lower 
Yellowstone were confined to the river bottom despite higher deer 
densities than on surrounding uplands. This occurred during a 
period in which the river bottom population nearly doubled in 
size. Egress and ingress of yearling deer was documented in each 
of the intensive study areas suggesting that dispersal had a 
minimal effect on population dynamics among subpopulat ions. 



Sexual competition between males, prior to and during the 
breeding season, has been considered a major factor influencing 
the comparatively high rate of dispersal among yearling males 
(Downing and McGinnes 1975, Kammerraeyer and Marchington 1976, 
Bunnell and Harestad 1983). The lower affinity of yearling males 
than females to their matriarchal groups (Hawkins and Klimstra 
1970) may explain differential rates of dispersal between 
yearling males and females. However, Ozoga and Verne (1985) 
reported that dominant yearling males were less inclined to 
remain in ancestral home ranges than their subordinate cohorts at 
the onset of breeding; this behavior was observed even in the 
absence of older males. They also reported the most intense 
social interaction between yearling males and related older 
females occurred prior to and during the breeding season when we 
observed comparatively high rates of dispersal of yearling males. 
Most dispersal of yearling females occurred during May -August 
when social interaction with older females was probably most 
intense. Thus, dispersal of both sexes of yearlings from 
ancestral home ranges may have resulted from social interaction 
with adult females within family groups. 

Dispersal of yearling males may optimize inbreeding from its 
capacity to faithfully transmit adaptive parental gene 
combinations (Nelson and Mech 1984). Perhaps of greater 
consequence, is the role of proportionately high male dispersal, 
along with greater parental investment in female offspring, or 
female philopatry, in allowing optimal resource partitioning in 
polygamous species as implied by Rogers (1987). This may reflect 
a more plausible implication of dispersal to a strategy of social 
population regulation than one involving a direct reduction in 
population size through dispersal as proposed by Hawkins et a 1 . 
(1971). 



9 7 



I nf luence o f Other Factors .— Ca tastrophic mortality among 
adult whitetails during late summer and early autumn that 
strongly implicated EHD as the causative agent was not density- 
related. The outbreak in eastern Montana in 1977 was most 
significant on the lower Yellowstone River, claiming 
approximately one-third of the population in the Intake area 
(Swenson 1979). This presumably contributed to the population 
decline in the late 1970s. 



The EHD virus was first isolated in eastern Montana 
following a 1978 outbreak and was identified as serotype 2 
implying that cattle and mule deer may be passively involved in 
transmission of the virus (Feldner and Smith 1981). Their data 
confirmed that high mortality was limited to white-tailed deer. 
Other outbreaks of hemorrhagic disease occurred in eastern 
Montana in 1961, 1976, 1977, and 1978 with a few scattered 
reports of dead deer in 1970 and 1971 (Walcheck 1978). A major 
epizootic was reported for southwestern North Dakota in 1970 and 
a lesser one for 1971 (Hoff et al. 1973). All occurred from late 
July through October. Other diseases affecting white-tailed deer 
are of little consequence in the Great Plains region (Petersen 
1984) . 

Inf luence o f Hunting .-- Hunting was the single largest 
mortality factor during all years. A time lag model (McCullough 
1979) indicated that the annual rate of population change and 
population size were in equilibrium from 1980 to 1983 when the 
population was increasing (Fig. 18). As pre-hunt populations 
increased from 1980 to 1983, proportionately larger segments of 
the population were removed by hunting (Table 18), with some 
possible time-lag delay coinciding with liberalized hunting 
seasons from 1983 through 1985. Adult mortality, attributable to 
causes other than hunting, was generally very low regardless of 
population size. 

A sex ratio among yearling and older deer favoring females 
at a rate of approximately 2:1 appeared solely attributable to 
hunting. Our data suggested that hunters selected antlered over 
antlerless deer (Tables 19 and 20), although numbers of antlered 
deer taken by hunters varied directly with their abundance in the 
population. The relatively young age structure of males also 
reflected disproportionately greater removal of males by hunting, 
because mortality from other factors, except highway mortality 
during December-May, did not suggest any differential mortality. 
However, in the absence of hunting, natural mortality among 
adults may also claim disproportionately more males than females 
as reported for the lower Columbia River where adult females 
outnumbered males by a ratio of approximately 3:1 (Gavin et al. 
1984) . 

Among yearlings, disproportionately higher dispersal of 
males than females might suggest that males were more vulnerable 
to hunting by virtue of occupying unfamiliar habitat. The 
coincidence of a decline in the relative proportion of yearling 
males in annual harvests and declining antler growth suggested 



98 




•0.6 — t 1 1 ™| 1 -i — i — I 

2000 3000 4000 5000 6000 

FALL POPULATION SIZE 




< 
H 

CO 

z 



0.5 H — i i — ! — r — ~ 1 — — 

1000 2000 3000 4000 

SPRING POPULATION SIZE 



Figure 18. Time lag model of rate of change versus population 
size considering the effects of mortality and natality. 



99 



that disproportionate higher mortality was an artifact of hunter 
selectivity for antlered deer and probably larger over smaller 
antlers. Selectivity also appeared dependent upon hunters 
differentiating yearling males from antlerless deer. 

The increased antlerless harvest from 1982 to 1985 (Table 
19) reflected liberalized bag limits for antlerless deer that 
included multiple tags per hunter in 1984 and 1985. Hunters 
tended to take adult females over fawns during 1980-83. A 
greater proportion of fawns in annual harvests in 1984 and 1985 
probably reflected less hunter-selectivity between fawns and 
adult females when more than one antlerless deer could be taken 
per individual hunter. 

The effect of exploitation of antlerless deer was the 
overridding factor influencing population size and trend from 
1983 through 1985 when harvest rates equalled or exceeded 
recruitment rates (Tables 5 and 18). The river bottom population 
declined from 1983 to 1986 despite a compensatory increase in 
fawn survival in response to declining female density (Fig. 7). 

Comparatively greater stability among numbers of adults and 
higher, more stable recruitment rates at Elk Island than at 
Intake (Tables 4 and 6) may have reflected the relative influence 
of hunting in the respective areas. Because of greater hunter 
access during all years, continuous harvesting of antlerless deer 
at Elk Island resulted in lowered female density that was 
compensated by higher recuitment of fawns than observed at 
Intake. Average yearly gains through recruitment of fawns to 
autumn, expressed as a percent of total numbers of yearling and 
older deer during summer, were 60% (49-67%) at Elk Island and 53% 
(39-67%) at Intake. Because of relatively little exploitation 
from 1980 through 1983 at Intake, numbers of adults increased and 
recruitment sharply dropped off (Table 4). 

Gavin et al. (1 984) and Woolf and Harder (1979) attributed 
disproportionately high natural mortality among adult males to 
stress associated with breeding activity. Our data suggest that 
hunting mortality among males compensated for natural mortality, 
because males older than 4 years, that would be predisposed to 
natural mortality from the cumulative energetic cost of breeding, 
accounted for less than 1% of all males. 



From a hypothetical 1,000 yearling males in a pre-hunt 
population, and assuming an average annual mortality rate of 60% 
for yearling and older males determined from annual population 
estimates (Table 16), only 26 (3%) can be expected to live to an 
age of 5 years. Thus, few males reach an age where they might be 
predisposed to overwinter mortality from energy-related stress of 
breeding behavior. Similarly for females, 370 (3 7%) of 1,000 
yearlings can be expected to live to an age of 5 years, and 175 
(18%) can be expected to reach an age of 8 years. The average 
annual mortality rate of yearling and older females for all years 
was 22% (Table 16). The minimum age where females were 
predisposed to overwinter mortality appeared to be 8-9 years, 



100 



whereas the only documented case among individually marked adult 
males was a 4-year-old. 

During 1983-85, when harvest of antlerless deer was highest 
(Table 19), average annual mortality of all adult (1+) females 
from all causes was 34%. From a hypothetical 1,000 yearling 
females, 190 could be expected to reach an age of 5 years if that 
mortality rate were constant over time; 50 could be expected to 
reach an age of 8 years. Continuous annual turnover of one-third 
or more of the females would result in a much younger age class 
structure. Hunting mortality appeared to offer less opportunity 
for compensatory mortality among females than among males. 



Because female mortality from causes other than hunting was 
minimal, natural regulation of female numbers would be achieved 
through density-dependent rates of recruitment and attrition of 
females older than 8 years. 

MANAGEMENT IMPLICATIONS 



Both population density and habitat use during our study 
were indicative of the manner in which deer encountered, 
exploited, and responded to varying habitat characteristics. 
Major habitat components, vegetation, topography, land use, and 
climate, interacted with behavioral attributes of the deer to 
influence distribution, movements, use of specific habitats, and 
food habi ts . 

In part, our findings departed from traditional models of 
population regulation for white-tailed deer. However, they 
attest to the wide ecological amplitude previously documented for 
the species. Each environment occupied involves an adaptive 
strategy by which the particular population may successfully 
exploit its particular habitat and persist over time. These in 
turn are reflected in different population characteristics and 
dynamics between habitats. Management of white-tailed deer in 
riverine environments in the northern Great Plains should 
consider: 

1. White-tailed deer on the lower Yellowstone River were 
closely associated with mid to late serai communities of the 
grassland sere. Relative density of deer varied directly 
with abundance of riparian forest and shrub cover: a strong 
linear relationship suggested a continuum of density 
distribution of deer with that of riparian cover. Limited 
water management and land use practices on the lower 
Yellowstone, that includes minimal competitive partitioning 
of the riparian habitat by ungulates, have probably 
contributed to a relatively stable habitat favoring 
comparatively high densities of white-tailed deer. Under 
more intensive water management that pre-empts periodic 
flooding, including channelization or mainstem impoundment, 
the regional climax vegetation (grassland) would ultimately 
become more prevalent at the expense of mid to late serai 



101 



communities (Boggs 1984). Any significant alteration of 
river flow and dynamics that resulted in a reduction of 
riparian tree/shrub cover would ultimately result in a 
reduction of base densities of white-tailed deer. 

2. Land clearing for field development, or logging, that 
reduced the amount of mature stands of cottonwood with a 
tree/shrub understory on the floodplain may reduce deer 
numbers locally. However, clearing small stands of decadent 
cottonwood that contained a grass understory and planting 
them to legume or cereal crops may benefit deer by creating 
additional diversity and interspers ion of the 2 important 
habitat components. If employed on wildlife management 
areas, this may also provide an attractive alternative to 
other means of abating depredation by deer on agricultural 
crops. 

3. Seasonal shifts in distribution reflected a preference 
during summer, at least among adult females, for diverse 
habitats, or the amount of edge, created by the 

in t e r spe r s ion of riparian communities and irrigated 
agricultural fields. During winter, all deer used 
relatively large blocks of riparian cover during daytime 
periods, while moving up to 1-2 km at night to feed in 
agricultural fields. Agricultural crops provided deer with 
readily digestible, high energy forage yearlong. A temporal 
dichotomy in relative use of crops by adult males and 
females reflected a period of high energy demand in the 
respective sexes: following the breeding season among males 
and during late gestation and lactation among females. The 
presence of this source of forage enhanced productivity and 
minimized natural mortality of deer older than 4 months, 
thus pre-empting a relationship between population dynamics 
and density with condition and trend of key browse species. 
Whereas the abundance of riparian tree/shrub cover may be 
overriding in its influence on the overall density 
distribution of deer on the river bottom, crop type and 
availability of crop residues influence which fields are 
used. 

4. The traditional "limiting" role of winter forage was not 
applicable to white-tailed deer on the lower Yellowstone 
River. Such a consideration perhaps may be extended to 
white-tailed deer in other riverine habitats in the northern 
Great Plains where climatic variability is buffered by 
irrigated agriculture. In riverine environments, natural 
population regulation may be achieved through allocation of 
forage and cover resources by social spacing among adult 
females. Variability in reproductive potential with changes 
in population size and/or environmental variability may 
reflect relegation of socially subordinate females to less 
than optimal summer habitats. 



1 02 



5. Exploitation of the population by hunting was an 
overriding factor contributing to a decline in population 
numbers from 1983 through 1986. The decline occurred 
despite a compensatory increase in fawn survival over summer 
in response to reduced female density because hunting- 
related mortality equalled or exceeded recruitment. Our 
data indicated that other factors contributing to mortality 
of adult females were additive to that of hunting and did 
not change in response to changes in animal abundance. 
Population declines, precipitated by heavy exploitation of 
antlerless deer, were not limited to periods of 
comparatively low fawn recruitment as generally implied from 
the trend for the entire river bottom study area or at 
Intake specifically. A population decline also occurred in 
the Elk Island s ubpo p ul a t ion from 1984 to 1985 despite an 
increase in recruitment from 76 to 106 f a w n s : 1 00 adult 
females between the 2 years. This type of compensatory 
increase in recruitment may be limited to riverine habitats 
where availability of irrigated agricultural crops to deer 
buffer the effects of variable climatic conditions on 
availability of natural forage. 

6. The level of exploitation of antlered deer by sport 
hunting (58%) during our study that resulted in a relatively 
young age class structure among males had no apparent effect 
on population dynamics or trend. However, removal of 30-33% 
of the females older than fawns by hunting resulted in 
population declines; continuous harvesting at these levels 
would result in a younger age class structure than existed 
during this study. Removal of 26% of yearling and older 
females roughly maintained female numbers at the same level 
between years, and harvests of less than 20% were followed 
by population increases. Therefore, removal of 20-25% of 
those age classes of females should be sustainable in this 
habitat, though special antlerless whitetail regulations 
were necessary to achieve that level of harvest during this 
study. 



103 



LITERATURE CITED 

Abler, W. A., D. E. Buckland, R. L. Kirkpatrick, and P. F. 

Scanlon. 1976. Plasma progestins and puberty in fawns as 
influenced by energy and protein. J. Wildl. Manage. 40:442- 
446. 

Adams, W. H., Jr. 1960. Population ecology of white-tailed deer 
in northeastern Alabama. Ecology 41:706-715. 

Albon, S. D, B . Mitchell, and B. W. Staines. 1 983. Fertility 
and body weight in female red deer: a density-dependent 
relationship. J. Anira. Ecol. 52: 969-980. 

Allen, E. 0. 1 96 8. Range use, foods, condition, and 

productivity of white-tailed deer in Montana. J. Wildl. 
Manage. 32: 130-141. 

. 1971. White-tailed deer. Pp. 69-79 i_n Game management 

in Montana (T. W. Mussehl and F. W. Howell, ed s .) . Montana 
Dep. Fish and Game, Fed. Aid Proj., Helena. 

Allen, E., and D. R. McCullough. 1976. Deer-car accidents in 
southern Michigan. J. Wildl. Manage. 40: 317-325. 

Anderson, A. E., D. E. Medin, and D. C. Bowden. 1 97 2. Indices 
of carcass fat in a Colorado mule deer population. J. 
Wildl. Manage. 3 6: 579-594. 

Anthony, R. G. 1976. Influence of drought on diets and numbers 
of desert deer. J. Wildl. Manage. 40:140-144. 

Armstrong, R. A. 1950. Fetal development of the northern white- 
tailed deer. Am. Midi. Nat. 4 3:650-66 6. 

Arnold, D. A., and L. J. Verme. 1963. Ten years of observation 
of an enclosed deer herd in northern Michigan. Trans. N.A. 
Wildl. Conf. 28:422-430. 

Barrett, M. W., J. W. Nolan, and L. D. Roy. 1982. Evaluation of 
a hand held net-gun to capture large mammals. Wildl. Soc 
Bull. 10:1 08-114. 

Barron, J. C, and W. F. Harwell. 1973. Fertilization rates of 
south Texas deer. J. Wildl. Manage. 3 7:179-18 2. 

Bashore, T. L. , W. M. Tzilkowski, and E. D. Bellis. 1 985. 

Analysis of deer-vehicle collision sites in Pennsylvania. 
J. Wildl. Manage. 49: 769-774. 

Beasom, S. L. 1974. Relationships between predator removal and 
white-tailed deer net productivity. J. Wildl. Manage. 
3 8: 854-85 9. 



1 04 



Bellis, E. D. , and H. B. Graves. 1971. Deer mortality on a 

Pennsylvania interstate highway. J. Wildl. Manage. 3 5: 2 3 2- 
2 3 7 . 

Bergerud, A.. T. 1971. The population dynamics of New Foundland 
caribou. Wildl. Monogr. No. 25. 55 pp. 

Blaisdell, J. P. 1958. Seasonal development and yield of native 
plants on the upper Snake River plains and their relation to 
certain climatic factors. USDA, Tech. Bull. 11 90. 68 pp. 

Boggs, K. W. 1984. Succession in riparian communities of the 
lower Yellowstone River. Unpubl. M.S. Thesis, Montana St. 
Univ., Bozeman. 107 pp. 

Brown, M. H. 1961. The plainsmen of the Yellowstone. Univ. 
Nebraska Press, Lincoln. 480 pp. 



, and W. R. Felton. 1955. The frontier years. Bramhall 

House, New York. 272 pp. 

Budde, W. S. 1983. Effects of photoperiod on puberty attainment 
of female white-tailed deer. J. Wildl. Manage. 4 7:595-604. 

Bunnell, F. L. , and A. S. Harestad. 1 983. Dispersal and 

dispersion of black-tailed deer: models and observations. 
J. Mamm. 64: 201-209. 

Burroughs, R. D. 1961. The natural history of the Lewis and 

Clark expedition. Michigan St. Univ. Press, Lansing. 340 
pp . 

Byers, C. R . , R. K. Steinhorst, and P. R. Krausraan. 1 984. 

Clarification of a technique for analysis of utilization- 
availability data. J. Wildl. Manage. 4 8:1050-105 3. 

Carlsen, J. C . , and R. E. Faroes. 1 9 5 7. Movements of white- 
tailed deer tagged in Minnesota. J. Wilcl. Manage. 2 1:397- 
40 1. 

Carroll, B. K. , and D. L. Brown. 1 97 7. Factors affecting 

neonatal fawn survival in southcentral Texas. J. Wildl. 
Manage. 4 1:63-69. 

Case, R. M . 1978. Interstate highway road-killed animals: a 
data source for biologists. Wildl. Soc. Bull. 6:8-13. 

Caughley, G. 1976. Wildlife management and the dynamics of 

ungulate populations. Pp. 399-421 i_n Applied Biology, Vol. 
I. Academic Press, London, UK. 



1977. Analysis of vertebrate populations, 
and Sons, New York. 234 pp. 



John Wiley 



1 05 



Cheatum, E- L. 1949. The use of corpora lutea for determining 
ovulation incidence and variations in fertility of the 
white-tailed deer. Cornell Vet. 39: 282-291. 

, and G. H. Morton. 1946. Breeding season of white-tailed 
deer in New York. J. Wildl. Manage. 10:249-263. 

, and C. W. Severinghaus. 1950. Variations in fertility of 

white-tailed deer related to range conditions. Trans. N. 
Am. Wildl. Conf. 1 5: 1 70-189. 

Christian, J. J., and D. E. Davis. 1964. Endocrines, behavior, 
and population. Science 14 6:1550-1560. 

Clover, M. R. 1954. A portable deer trap and catch-net. Calif. 
Fish and Game 40:367-37 3. 

C 1 u t t o n-B r ock , T. H. , F. E. Guinness, and S. D. Albon. 1 983. 

The costs of reproduction to red deer hinds. J. Anim. Ecol. 
52:367-383. 

Coblentz, B. E. 1 97 0. Food habits of George Reserve deer. J. 
Wildl. Manage. 34:535-540. 

Coe, R. J., R. L. Downing, and B. S. McGinnes. 1 980. Sex and 
age bias in hunter-killed white-tailed deer. J. Wildl. 
Manage. 44:245-249. 

Cole, G. F. 1959. Game management based on facts. Montana 
Wildlife, June:22-30. 

Compton, B. B. 1986a. Use of agricultural crop types by white- 
tailed deer. Proc Mont. Acad. Sci. 46:5-17. 

. 1 986b. Distribution of white-tailed deer along the lower 

Yellowstone River. M.S. Thesis, Montana State University, 
Bozeman. 74 pp. 

_, R. J. Mackie, and G. L. Dusek. 1 988. Factors influencing 
density distribution of white-tailed deer in riparian 
habitats. J. Wildl. Manage. 52:I_n Press . 

Conover, W. J., and R. L. Iraan. 1981. Rank transformations as a 
bridge between parametric and nonpa ram e t r i c statistics. The 
Amer. Stat. 3 5: 124-12 9. 

Cook, R. S., M. White, D. 0. Trainer, and W. C. Glazner. 1971. 
Mortality of young white-tailed deer in south Texas. J. 
Wildl. Manage. 35: 47-56. 



Crouch, G. L. 1 979. Changes in the vegetation complex of a 

cottonwood ecosystem on the South Platte River. Proc. Ann. 
Great Plains Ag r. Council 31:19-22. 



106 



. 1982. Wildlife on grazed and ungrazed bottomlands on the 

South Platte River, northeastern Colorado. pp. 186-197 in 
Proc. of the wildlife-livestock relations symp. (J. M. Peek 
and P. D. Dalke, eds.), Coeur d'Alene, ID. 

Dapson, R. W. , P. R. Ramsey, M. H. Smith, and D. F. Urbston. 

1979. Demographic differences in contiguous populations of 
white-tailed deer. J. Wildl. Manage. 43:889-898. 

Dasmann, W. 1971. If deer are to survive. The Wildlife 

Management Institute and Stackpole Books, Harrisburg, PA. 
12 8 pp. 

Davis, D. E., and R. L. Winstead. 1980. Estimating the numbers 
of wildlife populations. Pp. 221-246 jji Wildlife management 
techniques manual, 4th ed. (S. D. Schemnitz, ed.). The 
Wildl. Soc, Washington, D.C. 

Downing, R. L. 1980. Vital statistics of animal populations. 
Pp. 247-267 i_n_ Wildlife management techniques manual, 4th 
ed. (S. D. Schemnitz, ed.). The Wildl. Soc, Washington, 
D.C . 

, and B. S. McGinnes. 1975. Movement patterns of white- 
tailed deer in a Virginia enclosure. Proc. SE Fish and Game 
Comm. pp. 454-459. 

, E. D. Michael, and R. J. Poux. 1 97 7. Accuracy of sex and 

age ratio counts of white-tailed deer. J. Wildl. Manage. 
4 1: 709-714. 

Dusek, G. L. 1980. An inventory of vegetation, wildlife, and 

recreational resources of the Long Pines, Montana. Unpubl. 
Final Rep., Montana Dep. Fish, Wildl., and Parks, Helena. 
15 8 pp . 

. 1 987. Ecology of white-tailed deer in upland ponderosa 

pine habitat in southeastern Montana. Prairie Nat. 19:1-17. 

, and A. K. Wood. 1986. An evaluation of the average 

activity radius as an estimator of monthly movements of 
deer. Proc. Mont., Acad. Sci. 46:1 9-26. 

, J. G. Mundinger, S. J. Riley, and R. J. Mackie. 1 98 7. 

The implication of body weight to reproductive strategy 
among young adult female white-tailed deer. Proc. Congr. 
Int. Union Game Biol. 18: I__n Press. 

Everitt, B. S. 1977. The analysis of contingency tables. 

Monogr. on Appl. Prob. and Stat., Chapman and Hall, London. 
12 5 pp. 



Feldner, T. J., and M. H. Smith. 1981. Epizootic hemorrhagic 
disease virus in Montana: isolation and serologic survey. 
Am. J. Vet, Res. 4 2:1 1 98-1 202. 



1 07 



Fenneman, N. M. 1931. Physiography of the western United 
States. McGraw-Hill Book Co., New York. 534 pp. 

French, C. E., L. C. McEwen, N. D. Magruder, R. H. Ingram, and R. 
W. Swift. 1956. Nutrient requirements for growth and 
antler development in the white-tailed deer. J. Wildl. 
Manage. 20:221-232. 

Gasaway, W. C., R. 0. Stephenson, J. L. Davis, P. K. Shepherd, 

and 0. E. Burris. 1983. Interrelationships between wolves, 
prey, and man in interior Alaska. Wildl. Monogr. No. 84. 50 
PP • 

Gavin, T. A., L. H. Suring, P. A. Vohs, and E. C. Meslow. 1 984. 
Population characteristics, spatial organization, and 
natural mortality in the Columbian white-tailed deer. 
Wildl. Monogr. 91. 41 pp. 

Gilbert, F. F. 1966. Aging white-tailed deer by annul! in the 
cementum of the first incisor. J. Wildl. Manage. 30:200- 
202. 

Gill, J. 1956. Regional differences in size and productivity of 
deer In West Virginia. J. Wildl. Manage. 20:286-29 2. 

Gladfelter, H. L. 1984. Midwest agricultural region. Pp. 427- 
440 l_n (L. K. Halls, ed.) White-tailed deer ecology and 
management. The Wildlife Management Institute and Stackpole 
Books, Harrisburg, PA. 

Golley, F. B. 1957. An appraisal of ovarian analysis in 

determining reproductive performance of black-tailed deer. 
J. Wildl. Manage. 2 1: 62-65. 

Hamlin, K. L. , R. J. Mackie, J. G. Mundinger, and D. F. Pac. 

1982. Effect of capture and marking on fawn production in 
deer. J. Wildl. Manage. 46:1086-1089. 

Hanley, T. A. 1982. The nutritional basis for food selection by 
ungulates. J. Range Manage. 35: 146-1 51. 

Hanson, C. L. , J. R. Wight, J. P. Smith, and S. Smoliak. 1 98 2. 
Use of historical yield data to forcast range herbage 
production. J. Range Manage. 35:614-616. 

Harlow, R. F., J. B. Whelan, H. S. Crawford, and J. E. Skeen. 
1975. Deer foods during years of oak mast abundance and 
scarcity. J. Wildl. Manage. 39:330-336. 

Haugen, A. 0. 1975. Reproductive performance of white-tailed 
deer in Iowa. J. Mamm. 56: 151-1 59. 



Hawkins, R. E., L. D. Mortoglio, and G. G. Montgomery. 196 8. 
Cannon-netting deer. J. Wildl. Manage. 32: 191-195. 



108 



, and W. D. Klirastra. 1970. A preliminary study of the 
social organization of white-tailed deer. J. Wildl. Manage. 
34:407-419. 

, , and D. C. Autry. 1971. Dispersal of deer from 



the Crab Orchard National Wildlife Refuge. J. Wildll. 
Manage. 35:21 6-220. 

Hayne, D. W. 1 949. Calculation of size of home range. J. Mamm. 
3 0:1-17. 

Heezen, K. L. and J. R. Tester. 1 967. Evaluation of radio 

tracking by t r iangula t ion with special reference to deer 
movements. J. Wildl. Manage. 31:124-141. 

Herriges, J. D. 1986. Movement, activity, and habitat use of 
white-tailed deer along the lower Yellowstone River. 
Unpubl. M.S. Thesis, Montana St. Univ., Bozeman. 131 pp. 

Hestbeck, J. B. 1982. Population regulation of cyclic mammals: 
the social fence hypothesis. Oikos 3 9:157-163. 

Hinz , T. 1 977. The effect of altered streamflow on migratory 

birds of the Yellowstone River Basin, Montana. Montana Dep. 
Nat. Res. and Cons., Yellowstone impact study, Tech. Rep. 
No. 7, Helena. 107 pp. 

Hirth, D. H. 1977. Social behavior of white-tailed deer in 
relation to habitat. Wildl. Monogr. 5 3. 5 5 pp. 

Hoff, G. L. , S. H. Richards, and D. 0. Trainer. 1 973. Epizootic 
of hemorrhagic disease in North Dakota deer. J. Wildl. 
Manage. 3 7: 33 1-335. 

Holder, T. J., and P. Pescador, Jr. 1976. Soil survey of Dawson 
County, Montana. USDA Soil Cons. Se r v., and Montana Agr. 
Expt. Sta. 72 pp. 

Hoskinson, R. L. , and L. D. Mech. 1 976. White-tailed deer 

migration and its role in wolf predation. J. Wildl. Manage. 
40:429-441. 

Huegel, C. N. , R. B. Dahlgren, and H. L. Gladfelter. 1 985. 

Mortality of white-tailed deer fawns in southcentral Iowa. 

J. Wildl. Manage. 49: 377-380. , , and 

198 6. Bedsite selection by white-tailed deer fawns in Iowa. 
J. Wildl. Manage. 50:474-480. 

Jackson, L. W. s and W. T. Hesselton. 1 97 3. Breeding and 

parturition dates of white-tailed deer in New York. NY Fish 

Game J. 2 0:40-4 7. 

Jahn, L. R. 1959. Highway mortality as an index of deer 
population change. J. Wildl. Manage. 2 3:187-197. 



109 



Johns, P. E . , M . H. Smith, and R. K. Chesser. 1 984. Annual 

cycles of the KF I in a southeastern white-tailed deer herd. 
J. Wildl. Manage. 4 8:969-973. 

Julander, 0. W., W. L. Robinette, and D. A. Jones. 1961. 
Relation of summer range condition to mule deer herd 
productivity. J. Wildl. Manage. 

Kammermeyer, K. E., and R. L. Marchington. 197 6. Notes on 
dispersal of white-tailed deer. J. Mamm. 57: 776-778. 

Kamps, G. F. 1969. White-tailed and mule deer relationships in 
the Snowy Mountains of central Montana. M.S. Thesis, 
Montana State University, Bozeman. 59 pp. 

Kie, J. G. , D. L. Drawe, and G. Scott. 1 980. Changes in diet 

and nutrition with increased herd size in Texas white-tailed 
deer. J. Range Manage. 33:28-34. 

., M. White, and D. L. Drawe. 1983. Condition parameters 

of white-tailed deer in Texas. J. Wildl. Manage. 47: 583- 
594. 

interaction between 
Prod., The Royal 



, and H. Strandgard. 1972. Factors affecting growth and 

body size of roe deer. J. Wildl. Manage. 36:64-79. 

Koch, E. 1946. Big game in Montana from early historical 
records. J. Wildl. Manage. 5:357-37 0. 

Leckenby, D. A., and A. W. Adams. 1 986. A weather severity 

index on a mule deer winter range. J. Range Manage. 39:244- 
248. 

Mackie, R. J. 1964. Montana deer weights. Montana Wildl. 
Fall : 9-14. 

. 1 983. Natural regulation of mule deer populations. 

Pp. 112-125 i_n Symposium on natural regulation of wildlife 
populations (F. L. Bunnell, D. S. Eastman, and J. M. Peek, 
eds.). For., Wildl., and Range Expt. Sta., Univ. Idaho, 
Mo s cow . 

, K. L. Hamlin, and D. F. Pac 1981. Census methods for 

mule deer. Pp. 97-106 i_nS ympo s i um on census and inventory 
methods (F. L. Miller and A. Gu n n , eds.). For., Wildl., and 
Range Expt. Sta., Univ. Idaho, Moscow. 



Klein, D. R. 1985. Population ecology: the 
deer and their food supply. Biol. Deer 
Soc. N. Z. Bui 1. 22: 13-22. 



Mansell, W. D. 1 974. Productivity of white-tailed deer on the 
Bruce Peninsula, Ontario. J. Wildl. Manage. 3 8:808-81 4. 



110 



Ma rch ing ton , R. L. , and I). H. Hirth. 1 984. Behavior. Pp. 129- 
168 in (L. K. Halls ed.) White-tailed deer ecology and 
management. Wildl. Manage. Inst., Washington D.C. 

Martinka, C. J. 1968. Habitat relationships of white-tailed 
and mule deer in northern Montana. J. Wildl. Manage. 
32:558-565. 

McCaffery, K. R. 1 973. Road-kills show trends in Wisconsin deer 
populations. J. Wildl. Manage. 37:212-216. 

McCullough, D. R. 1 979. The George Reserve deer herd, 

population ecology of a K-selected species. Univ. Michigan 
Press, Ann Arbor. 271 pp. 

1984. Lessons from the George Reserve. Pp. 211-242 in 
(L. K. Halls, ed.) White-tailed deer ecology and management. 
The Wildlife Management Institute snf Stackpole Books, 
Harrisburg, PA. 

McGinnes, B. S., and R. L. Downing. 1 97 7. Factors affecting the 
peak of deer fawning in Virginia. J. Wildl. Manage. 41:715- 
719. 

Mercier, L. K. 1985. Montana episodes - memories of Sidney, 
Montana and the lower Yellowstone Valley, 1919-1939. 
Montana - the magazine of western history. 35:78-83. 

Michael, E. D. 1965. Movements of white-tailed deer on the 
Welder Wildlife Refuge. J. Wildl. Manage. 29: 44-52. 

Mohr, C. 0. 1947. Table of equivalent populations of North 
American small mammals. Am. Midi. Nat. 37: 223-249. 

Monson, R. A., W . B. Stone, B. L. Weber, and F. J. Spadaro. 

1974. Comparison of Riney and total kidney fat techniques 
for evaluating the physical condition of white-tailed deer. 
N.Y. Fish and Game J. 2 1: 6 7-72. 

Moore, L. G. , and R. R. Shields. 1 980. St ream flow 

characteristics of the Yellowstone River Basin, Montana, 
through 1 97 6. U. S. Geol. Su r v. , Water Res. Invest. 80-41. 

Morton, G. H. , and E. L. Cheatum. 1 946. Regional differences in 
breeding potential of white-tailed deer in New York. J. 
Wildl. Manage. 10:242-248. 

Mueller, C. C. , and R. M. F. S. Sadleir. 1 979. Age at first 

conception In black-tailed deer. Biol. Reprod. 2 1:1099-1104. 



Mundinger, J. G. 1981. White-tailed deer reproductive biology 

in the Swan Valley, Montana. J. Wildl. Manage. 45:132-139. 



Ill 



Murphy, R. K. , N. F. Payne, and R. K. Anderson. 1 985. White- 
tailed deer use of an irrigated agriculture-grassland 
complex in central Wisconsin. J. Wildl. Manage. 49: 125-128. 

Mussehl, T. W., and F. W. Howell, eds. 1971. Game management in 
Montana. Montana Dep. Fish and Game, Fed. Aid Proj. W-3-C, 
Helena. 238 pp. 

Nelson, J. R. 1961. Compositiona dn structure of the principal 
woody vegetation types in the North Dakota Badlands. M.S. 
Thesis. North Dakota State University, Fargo. 

Nelson, M. E., and L. D. Mech. 1981. Deer social organization 
and wolf predation in northeastern Minnesota. Wildl. 
Monogr. 77. 53 pp. 

, and . 1984. Home range formation and dispersal of 

deer in northeastern Minnesota. J. Mamm. 65:567-575. 



Nudds, T. D. 1980. Forage "preference": theoretical 

considerations of diet selection by deer. J. Wildl. Manage. 
44:735-740. 

O'Pezio, J. P. 1978. Mortality of young white-tailed deer fawns 
on the Seneca Army Depot. NY Fish and Game J. 25:1-15. 

Ozoga, J. J., L. J. Verme, and C. S. Bienz. 1 982. Parturition 
behavior and territoriality in white-tailed deer; impact on 
neonatal mortality. J. Wildl. Manage. 46:1-11. 

, and . 1982. Physical and reproductive 

characteristics of a supp 1 em e n t a 1 ly-f e d white-tailed deer 
herd. J. Wildl. Manage. 46:281-301. 

, and . 1 985. Comparative breeding behavior and 

performance of yearling vs. prime-age white-tailed bucks. 
J. Wildl. Manage. 49:364-372. 

, and . 1 986. Initial and subsequent maternal success 

of white-tailed deer. J. Wildl. Manage. 50:122-124. 

Patterson, J. H. 1976. The role of environmental heterogeneity 
in the regulation of duck populations. J. Wildl. Manage. 
40:22-32/ 

Peek, J. M. 1980. Natural regulation of ungulates (What 

constitutes a real wilderness). Wildl. Soc. Bull. 8:217- 
2 27. 



Pescador, P., Jr., and L. C. Brockman. 1 980. Soil survey of 
Richland county, Montana. USDA Soil Cons. Serv., and 
Montana Ag r. Expt. Sta. 72 pp. 



112 



Petersen, L. E. 1 984. Northern plains. Pp. 441-448 in (L. K. 
Halls, ed.) White-tailed deer ecology and management. The 
Wildlife Management Institute, and Stackpole Books, 
Harrisburg, PA. 

Platts, W. S. 1979. Livestock grazing and riparian/stream 
ecosystems - an overview, pp. 39-45 i_n Proc, Forum-- 
grazing and riparian/stream ecosystems (0. B. Cope, ed.), 
Trout Unlimited, Inc. 

Porter, W. F ., and K. E. Church. 1 987. Effects of environmental 
pattern on habitat preference analysis. J. Wildl. Manage. 
51:681-685. 

Puglisi, M. J., J. S. Lindzey, and E. D. Bellis. 1 974. Factors 
associated with highway mortality of white-tailed deer. J. 
Wildl. Manage. 38:799-807. 

Ransom, A. B. 1967. Reproductive biology of white-tailed deer 
in Manitoba. J. Wildl. Manage. 3 4: 114-12 3. 

Ray, A. A. 1 982. SAS user's guide. SAS Inst., Inc., Gary, N.C. 
9 2 1 pp . 

Reimers, E. 1972. Growth in domestic and wild reindeer in 
Norway. J. Wildl. Manage. 36:612-619. 

Rice, W. R., and J. D. Harder. 1 97 7. Application of multiple 
aerial sampling to a mark-recapture census of white-tailed 
deer. J. Wildl. Manage. 41:197-206. 

Richter, A. R., and R. F. Labisky. 1985. Reproductive dynamics 
among disjunct white-tailed deer herds in Florida. J. 
Wildl. Manage. 4 9:964-97 1. 

Robinette, W. L. 1 966. Mule deer home range and dispersal in 
Utah. J. Wildl. Manage. 30:335-34 9. 

Robinson, R. M . , J. W. Thomas, and R. G. Marburger. 1 965. The 
reproductive cycle of male white-tailed deer. J. Wildl. 
Manage. 29:53-59. 

Rogers, L. L. 1987. Effects of food supply and kinship on 

social behavior, movements, and population growth of black 
bears in northeastern Minnesota. Wildl. Monogr. No. 97. 72 
PP • 

Roseberry, J. L. , and W. D. Klimstra. 1 970. Productivity of 

white-tailed deer on Crab Orchard National Wildlife Refuge. 
J. Wildl. Manage. 34:2 3-28. 



, and 

contro lied 



1974. Differential vulnerability during a 
deer harvest. J. Wildl. Manage. 38:499-507. 



113 



, J. L., D. C. Autry , W. D. Klimstra, and L. A. 
Mehrhoff. 1969. A controlled deer hunt on Crab Orchard 
National Wildlife Refuge. J. Wildl. Manage. 3 3:791-795. 

Sanderson, G. C. 1 966. The study of mammal movements ~ a 
review. J. Wildl. Manage. 30:215-23 5. 

Seal, U. S., L. J. Verme, J. J. Ozoga, and E. D. Plotka. 1 983. 
Metabolic and endocrine responses of white-tailed deer to 
increasing population density. J. Wildl. Manage. 47:451- 
462. 

Segelquist, C. A., H. L. Short, F. D. Ward, and R. G. Leonard. 
1972. Quality of some winter forages in the Arkansas 
Ozarks. J. Wildl. Manage. 36: 174-177. 

Sever inghaus , C. W. 1949. Tooth development and wear as 

criteria of age in white-tailed deer. J. Wildl. Manage. 
1 3: 195-21 6. 

. 1 979. Weights of white-tailed deer in relation to range 

conditions in New York. NY Fish and Game J. 26:162-187. 

. 1981. Overwinter weight loss in white-tailed deer in Ne 

York. NY Fish and Game J. 28:61-67. 

., and E. L. Cheatum. 1956. Life and times of the white- 
tailed deer. Pp. 57-186 i_n (W. P. Taylor, ed.) The deer of 
North America. The Wildlife Management Institute, Stackpol 
Co., Harrisburg, PA. 

Severson, K. E. , and C. E. Boldt. 1 978. Cattle, wildlife, and 
riparian habitats in the western Dakotas. pp. 91-103 i_n 
Management and use of northern plains rangeland. Reg. 
Rangeland Symp., Bismarck, ND. 

Shlflet, T. N. , and H. E. Dietz. 1 974. Relationship between 
precipitation and annual rangeland herbage production in 
southeastern Kansas. J. Range Manage. 27: 272-276. 

Short, H. L. , J. D. Newsom, G. L. McCoy, and J. F. Fowler. 1969 
Effects of nutrition and climate on southern deer. Trans. 
N. Am. Wildl. Conf. 3 4: 137-14 5. 

Skogland, T. 1985. The effects of density-dependent resource 
limitations on the demography of wild reindeer. J. Anim. 
Ecol. 54: 3 59-374. 

Smart, C. W. , R. H. Giles, Jr, and D. C. Guynn. 1973. Weight 
tape for white-tailed deer in Virginia. J. Wildl. Manage. 
37:553-555. 

Smith, R. L. , and L. D. Flake. 1 983. The effects of grazing on 
forest regeneration along a prairie river. Prairie Nat. 
15: 41-44. ~ 



1 14 



Stormer, F. A., G. M. Kirkpatrick, and T. W. Hoekstra. 1 979. 

Hunter-inflicted wounding of white-tailed deer. Wildl. Soc. 
Bull. 7: 10-16. 

Suring, L. H. , and P. A. Vohs. 1 979. Habitat use by Columbian 
white-tailed deer. J. Wildl. Manage. 43:610-619. 

Swenson, J.E. 1979. Effects of a hemorrhagic disease epizootic 
on a white-tailed deer population in eastern Montana. Mont. 
Ac ad. Sci. 38:25-32. 

1982. Effects of hunting on habitat use by mule deer on 
mixed-grass prairie in Montana. Wildl. Soc. Bull. 10:115- 
12 0. 

1985. Compensatory reproduction in an introduced 
mountain goat population in the Absaroka Mountains, Montana. 
J. Wildl. Manage. 49:837-843. 

, and S. T. Stewart. 1982. on the use of population 
condition indices in deer management. Pp. 28-35 in The 
application of recent research in Montana (C. D. Eustace, 
ed.). Proc. Montana Chapt., The Wildl. Soc, Billings. 



S. J. Knapp, and H. J. Wentland. 1 98 3. Winter 
distribution and habitat use by mule deer and white-tailed 
deer in southeastern Montana. Prairie Nat. 15:97-112. 

Teer, J. G. , J. W. Thomas, and E. A. Walker. 1 965. Ecology and 
management of white-tailed deer in the Llano Basin of Texas. 
Wildl. Monogr. 15. 62 pp. 

Thomas, J. W . , J. G. Teer, and E . A. Walker. 1 964. Mobility and 
home range of white-tailed deer of the Edwards Plateau in 
Texas. J. Wildl. Manage. 28:463-472. 

Tierson, W. C. , G. F. Mattfield, R. W. Sage, and D. F. Behrend. 
1985. Seasonal movements and home ranges of white-tailed 
deer in the Adirondacks. J. Wildl. Manage. 4 9:760-769. 

Torgerson, 0., and W. R. Porath. 1 984. Midwest oak/hickory 

forest. Pp. 4 1 1-42 6 j_n (L. K. Halls, ed.) White-tailed deer 
ecology and management. The Wildlife Management Institute 
and Stackpole Books, Harrisburg, PA. 

U. S. Dep. Comm. 1 979-1 986. C 1 i ma to log ical data, Montana. 
National Climatic Data Center, Ashevi lie, NC. 



U. S. 



Fish and Wildlife Service, 
study unit, Montana. Office 
Bi 1 lings . 



1952. Yellowstone wildlife 
of River Basin Studies, 



115 



Van Etten, R. C, D. F. Switzenberg, and L. Eberhardt. 1 965. 
Controlled deer hunting in a square mile enclosure. J. 
Wildl. Manage. 29: 59-73. 

Verine, L. J. 1969. Reproductive patterns of white-tailed deer 
related to nutritional plane. J. Wildl. Manage. 33:881-887 

, and D. E. Ullrey. 1 984. Physiology and nutrition. Pp. 



91-118 in (L. K. Halls, ed.) White-tailed deer ecology and 
management. The Wildlife Management Institute and Stackpole 
Books, Harrisburg, PA. 

Vogel, W. 0. 1983. The relationship of white-tailed deer and 
mule deer to agriculture in the Gallatin Valley. M.S. 
Thesis, Montana St. Univ., Bozeman. 86 pp. 

Walcheck, K. 1978. Epizootics: mystery deer diseases. Montana 
Outdoors . 9:30-35. Nov. /Dec. 

Watson, A., and B. W. Staines. 1978. Differences in the quality 
of wintering areas used by male and female red deer ( C e r vus 
elaphus ) in Aberdeenshire. J. Zool., London 186:544-550. 

Wehausen, J. D., V. C. Bleich, B. Blong, and T. L. Russi. 1987 

Recruitment dynamics in a southern California mountian sheep 
population. J. Wildl. Manage. 51:86-98. 

White, M. 1973. The white-tailed deer of the Aransas National 
Wildlife Refuge. Texas J. Sci. 24:457-489. 

Wilkins, B. T. 1957. Range use, food habits, and agricultural 
relationships of mule deer, Bridger Mountains, Montana. J. 
Wildl. Manage. 21: 159-169. 

Will, G. C. 1973. Population characteristics of northern Idaho 
white-tailed deer. NW Sci. 47:114-122. 



Williamson, S. J., and D. H. Hirth. 1 985. An evaluation of edge 
use by white-tailed deer. Wildl. Soc. Bull. 1 3:252-257. 

Wisiol, K. 1984. Estimating grazingland yield from commonly 
available data. J. Range Manage. 3 7:471-47 5. 

Wood, A. K. 1987. Ecology of a prairie mule deer population. 
PhD. Thesis, Montana St. Univ., Bozeman. 20 5 pp. 

_, R. E. Short, A. E. Darling, G. L. Dusek, R. G. Sasser, and 
C. A. Ruder. 1986a. Serum assays for detecting pregnancy 
in mule and white-tailed deer. J. Wildl. Manage. 50: 684-687. 

_, G. L. Dusek, and R. E. Short. 1 986b. Effects of capture 
method on stress, survival, and pregnancy diagnosis in deer. 
Proc. Mont. Acad. Sci. 4 6:75-8 0. 



116 



Woolf, A., and J. D. Harder. 1979. Population dynamics of a 

captive white-tailed deer herd with emphasis on reproduction 
and mortality. Wildl. Monogr. 67. 53 pp. 

Zar, J. H. 1984. B io s t a t i s t ica 1 analysis, second edition. 
Prentice-Hall, Inc. 718 pp. 

Zwank, P. J., R. D. Sparrowe, W . R. Porath, and 0. Torgerson. 

1979. Utilization of threatened habitats by white-tailed 
deer. Wildl. Soc. Bull. 7: 2 26-232. 



If 



^ cMoi\tai\a Departn\eixt of 
} t Fisl^, e Wildhfe (Si TurH§