home ranges, movement patterns, and habitat use of …
TRANSCRIPT
HOME RANGES, MOVEMENT PATTERNS, AND HABITAT USE OF A NEWLY COLONIZED POPULATION OF FLORIDA BLACK BEARS (URSUS AMERICANUS
FLORIDANUS) IN A FRAGMENTED FLORIDA LANDSCAPE
By
DANA L. KARELUS
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2017
© 2017 Dana L. Karelus
To my Mom and Dad
4
ACKNOWLEDGMENTS
I am immensely grateful to my advisor, Dr. Madan Oli, for taking a chance on
accepting a former engineer into his lab. Despite my somewhat unconventional
background, he insisted from my first day that I was an ecologist. With Madan’s
attentive guidance, generous support, and with the copious amount of time that he
spent to help me, I can now say without a doubt, that I am an ecologist. I also want to
express my extreme thanks to Walt McCown. He graciously welcomed me on to this
project, has been a wonderful mentor through these years, and taught me how to “be
the bear”. Likewise, I thank Brian Scheick for serving on my committee and for sharing
his wealth of knowledge about black bear ecology and issues related to their
management. Being out in the field with Walt and Brian or just in the office talking about
bears and other wildlife management issues was always exciting and insightful.
I would also like to thank the rest of my committee members, Dr. Ben Bolker, Dr.
Eric Hellgren, and Dr. David Reed. Ben has been a tremendous source of positivity and
analytical creativity. His statistical knowledge and R coding skills were invaluable to me
for completing my dissertation; his patience and understanding as a teacher/mentor
have been priceless for my growth as an ecologist, and our conversations in between
always spawned my analytical creativity. Eric and David provided much insight that
strengthened my dissertation and were both also great sources of support and
encouragement. Eric helped me to think about my research not only in terms of bears,
but also more broadly. David helped me to speak about my research more colloquially,
always reminding me to explain it “as if I’m talking to my grandmother”.
My dissertation work would not have been possible without the funding provided
by the following sources, for which I am extremely thankful: Camp Blanding Joint
5
Training Center, the School of Natural Resources and Environment, the Department of
Wildlife Ecology and Conservation, the Florida Fish and Wildlife Conservation
Commission, and to many individuals who donated to my crowd sourcing campaign in
support of my field work. I also very much appreciate the interest for the conservation
and management of the bears in the area expressed by the land managers and other
personnel from Camp Blanding and the Florida Forest Service, and by the private land
owners in the surrounding area; I appreciated their concern and interest in black bears
in the area. Everyone at my study site that I spoke with shared stories and pictures of
the bears on their property.
I would like to thank the administrative personnel in the School of Natural
Resources and Environment and in the Department of Wildlife Ecology. Karen Bray,
Cathy Ritchie, Kirsten Hecht, Dr. Tom Frazer, Elaine Culpepper, Caprice McRae,
Monica Lindberg, Claire Williams, Heather Bradley, Fiona Hogan, Gay Hale, and Dr.
Eric Hellgren, thank you all for your help with paperwork, travel grants, room
reservations, coffee provisions in the WEC lounge, and encouragement and support. I
could not have navigated through all the University’s requirements or survived being a
graduate student without you!
Aside from those on my committee, I also had the opportunity to learn from and
receive help from many other fantastic faculty members at the University of Florida,
especially those from the Department of Wildlife Ecology and Conservation. I express
my gratitude to all of them but I must specifically acknowledge a few individuals. Thank
you to Dr. Christina Romagosa for being so very supportive, kind, and helpful; I very
much enjoyed our one semester of combined lab meetings and TAing for you. Thank
6
you to Dr. Rob Fletcher, your help during my first year was especially instrumental for
getting my dissertation work started and your help since then has also been much
appreciated. I also must thank Dr. Mark Hostetler for organizing coffee hour in the WEC
lounge and for increasing my knowledge about coffee and how to make a proper latte.
The coffee helped get me through the day and the time in the lounge promoted many
good conversations with other faculty and students, with whom I may not have
otherwise had many interactions. Thank you, again, to all the other faculty as well. Also,
I’d like to thank the biology faculty from the University of Central Florida, specifically Dr.
Pedro Quintana-Ascencio, for guiding me as I entered the field of ecology from 2011-
2013 and also to Dr. Karen Holloway-Adkins and Daryl Adkins for their help in changing
fields from engineering to an ecologist.
I am very thankful for the help, stimulating discourse, and friendship I received
from my past and present labmates and pseudo-labmates in the Oli Population Ecology
Lab and the Fletcher Landscape Ecology Lab, particularly Dr. Varun Goswami, Dr.
Madelon van de Kerk, Jennifer Moore, Rashidah Farid, Thomas Selby, Arjun Srivathsa,
Dr. Elise Morton, Mahi Puri, Vratika Chaudhary, Marta Prat, Dr. Divya Vasudev, Dr.
Katie Haase, Dr. Mauricio Nuñez-Regueiro, Isabel Gottlieb, and Jessica Hightower.
Madelon’s help through the years was especially important for much of my dissertation
work. Additionally, I want to thank many other friends from WEC and FWC that helped
me with my work along the way, including Dr. Dan Greene, Brian Smith, Brittany
Bankovich, Johanna Freeman, Erin Leone, and Elina Garrison. Furthermore, I must
thank my awesome volunteers who graciously helped with the vegetation sampling
portion of my work, including Adriana Betancourt, Jeanelle Brisbane, Zachary Holmes,
7
Heidi Hetzel, Lethia Johnson, Gage LaPierre, Shelby LeClare, Roccio Manobanda,
Marina McCampbell, Noah Mueller, Meagan Muir, and Shelby Shiver. And I also want to
extend a general, but incredibly significant, thank you to all my friends, near and far,
throughout these years, especially to those who watched Howie for me when I went
away to conferences; I got by with more than a little help from my friends.
Finally, I want to thank my family for all their love and support throughout my PhD
and throughout all my endeavors over the years that have ultimately led me to this
point. From my dad, I learned to love adventure, science, nature, and wildlife, and from
my mom, I learned to put my heart into everything that I do and to dream big. I am
grateful to my family for always encouraging me to follow my dreams, to work hard, and
to never give up.
8
TABLE OF CONTENTS
page
ACKNOWLEDGMENTS .................................................................................................. 4
LIST OF TABLES .......................................................................................................... 10
LIST OF FIGURES ........................................................................................................ 12
ABSTRACT ................................................................................................................... 14
CHAPTER
1 INTRODUCTION .................................................................................................... 16
2 HOME RANGES AND HABITAT SELECTION BY BLACK BEARS IN A NEWLY COLONIZED POPULATION IN FLORIDA .............................................................. 18
Field-Site Description .............................................................................................. 21 Methods .................................................................................................................. 22
Land Cover Categories..................................................................................... 23 Home Ranges .................................................................................................. 23 Habitat Selection .............................................................................................. 25
Results .................................................................................................................... 27 Home Ranges .................................................................................................. 27
Habitat Selection .............................................................................................. 28 Discussion .............................................................................................................. 29
3 EFFECTS OF ENVIRONMENTAL FACTORS AND LANDSCAPE FEATURES ON MOVEMENT PATTERNS OF FLORIDA BLACK BEARS ................................ 45
Methods .................................................................................................................. 48 Study Site ......................................................................................................... 48 Field Methods and Data Collection ................................................................... 49 Movement Metrics ............................................................................................ 50
Statistical Analysis of Movement. ..................................................................... 53 Analysis of Road Crossing ............................................................................... 54
Results .................................................................................................................... 55 Discussion .............................................................................................................. 58
4 INCORPORATING MOVEMENT PATTERNS TO DISCERN HABITAT SELECTION: BLACK BEARS AS A CASE STUDY ................................................ 81
Methods .................................................................................................................. 83
Study Species and Site .................................................................................... 83 Field Methods ................................................................................................... 85 Habitat Covariates ............................................................................................ 85
9
Movement Metrics and Identification of Movement States ............................... 86
Step-Selection Functions.................................................................................. 87
Results .................................................................................................................... 89 Discussion .............................................................................................................. 91
5 MICROHABITAT FEATURES INFLUENCING HETEROGENOUS HABITAT-USE BY FLORIDA BLACK BEARS ...................................................................... 112
Methods ................................................................................................................ 113
Study Site ....................................................................................................... 113 Bear Captures and GPS Data Collection ....................................................... 114 Identification of High- and Low-Use Areas Within Home Ranges ................... 115 Vegetation Sampling ...................................................................................... 116
Statistical Analyses ........................................................................................ 117 Results .................................................................................................................. 118
Discussion ............................................................................................................ 122 Conclusions and Management Recommendations ............................................... 128
6 CONCLUSIONS AND MANAGEMENT IMPLICATIONS ...................................... 145
APPENDIX
A DATA PREPARATION AND LAND COVER MAP ................................................ 148
B HOME RANGE SIZES AMONG STUDIES ........................................................... 150
C FRAGMENTATION ANALYSIS ............................................................................ 152
D TABLES OF MOVEMENT METRIC AVERAGES AND MODEL SELECTION TABLES ................................................................................................................ 154
E FIGURES FROM MODELS OF MOVEMENT METRICS ...................................... 191
F TABLES OF PLANT SPECIES FOUND IN AREAS OF HIGH-USE BY BEARS .. 202
LIST OF REFERENCES ............................................................................................. 237
BIOGRAPHICAL SKETCH .......................................................................................... 258
10
LIST OF TABLES
Table page 2-1 Percentage of each land cover category composing the 99% minimum
convex polygon ................................................................................................... 34
2-2 Ranking matrix from compositional analysis for second-order habitat selection ............................................................................................................. 35
2-3 Ranking matrix from compositional analysis for third-order habitat selection ..... 38
2-4 Model selection results from mixed effects logistic regression testing for factors influencing habitat selection .................................................................... 41
2-5 Estimates (± SE) of slope parameters, as well as 95% confidence intervals, for the fixed effect variables included in the most parsimonious mixed effects logistic regression model .................................................................................... 42
3-1 Model selection statistics testing for the effect of various covariates on movement metrics .............................................................................................. 67
3-2 Average number of weeks that bears were monitored, average road length within individual home ranges (km), and the average number of road crossings ............................................................................................................ 72
4-1 Average step-lengths (±SE) and turning angles by season ................................ 98
4-2 Model selection results from hidden Markov models (HMMs) testing for the number of movement states and factors that influenced the transition probabilities among movement states ................................................................ 99
4-3 Results of model selection for step-selection functions from conditional logistic regression ............................................................................................. 100
4-4 Odds and 95% confidence intervals for the variables included in the most parsimonious conditional logistic models .......................................................... 103
5-1 Averages (± SE) of vegetation measures from high and low bear-use sites .... 130
5-2 Average distances (± SE) from the sampled bear locations to major roads, minor roads, and creeks ................................................................................... 131
5-3 Results of generalized linear mixed models testing for the effect of individual habitat covariates on the probability of high-use............................................... 132
5-4 Pairwise correlation between habitat variables ................................................. 133
5-5 Principal component (PC) loadings from microhabitat variables....................... 134
11
5-6 Model selection from generalized linear mixed models testing for the effect of principal components (PC) 1 through 4 on the probability of high-use ............. 135
5-7 Estimates (± SE) of slope parameters, as well as 95% confidence intervals, for the fixed effects of the principal component loadings. ................................. 138
B-1 Overall and annual average home range sizes ................................................ 150
C-1 Quantification of habitat fragmentation in our study site at Camp Blanding and the surrounding areas ................................................................................ 153
D-1 Average movement metrics for female and male Florida black bears at multiple temporal scales ................................................................................... 155
D-2 Model selection tables for responses of each weekly movement metric .......... 156
D-3 Model selection tables for each movement metric from Florida black bears in north-central Florida at the daily temporal scale ............................................... 171
F-1 Percent of high and low bear-use sites in which tree species were present ..... 202
F-2 Understory species found in plots 4x4 m plots at high- and low-use bear sites 207
12
LIST OF FIGURES
Figure page 2-1 Map showing the location of the Camp Blanding Joint Training Center and
the closest designated primary Florida Black Bear ranges ................................. 43
2-2 Average Florida Black Bear home range sizes in the Camp Blanding area ....... 44
3-1 Map of the study site at Camp Blanding Joint Training Center, Florida. Roads and creeks are also shown. ................................................................................ 73
3-2 Average bi-hourly step-length (± 95% CI) in meters throughout the diel period.. 74
3-3 Average bi-hourly step-length (± 95% CI) in meters throughout the diel period for female Florida black bears (Ursus americanus floridanus) in north-central Florida with and without cubs of the year ........................................................... 75
3-4 The mean squared displacement (MSD) ............................................................ 76
3-5 Weekly average directional persistence, E(c), and directional bias, E(q) ........... 77
3-6 Observed versus expected displacements for bi-hourly location data ................ 78
3-7 Effect of covariates on the weekly average bi-hourly step-length ....................... 79
3-8 Effect of covariates on the weekly observed displacement ................................ 80
4-1 Plots of step-length parameter distributions from 3-state HMMs ...................... 107
4-2 Plots of turning angle parameter distributions from 3-state HMMs ................... 108
4-3 Proportion of steps in each movement state across the diel period as assigned by the Viterbi algorithm ...................................................................... 109
4-4 Predictive odds of a bear choosing a land cover type based on the full conditional logistic models ................................................................................ 110
4-5 Predictive odds of a bear choosing a land cover type based on the full conditional logistic models ................................................................................ 111
5-1 Map of the dynamic Brownian Bridge movement model utilization distribution for a representative female Florida black bear.................................................. 139
5-2 Locations of vegetation sampling in high- and low-use areas. ......................... 140
5-3 Average (± SE) A) percent canopy cover and B) visual obstruction from sites of high and low black bear-use. ........................................................................ 141
13
5-4 The absolute density (± SE) of hardwood and pine trees in high- and low-bear-use sites. .................................................................................................. 142
5-5 The average (± SE) percent cover by 15 of the most common species found in the shrub layer in 4 x 4m plots ...................................................................... 143
5-6 The average percent cover (± SE) of A) food producing shrubs and B) non-food producing shrubs within 4 x 4 m plots. ...................................................... 144
E-1 Effect of covariates on the weekly average directional persistence .................. 192
E-2 Effect of covariates on the weekly average directional bias ............................. 193
E-3 Effect of covariates on the weekly expected displacement ............................... 194
E-4 Effect of covariates on the daily average bihourly step-length. ......................... 195
E-5 Effect of covariates on the daily average directional persistence ..................... 196
E-6 Effect of covariates on the daily average directional bias ................................. 197
E-7 Effect of covariates on the daily observed displacement .................................. 198
E-8 Effect of covariates on the daily expected displacement. ................................. 199
E-9 Effect of covariates on the bihourly step-length ................................................ 200
E-10 Effect of covariates on the bihourly step-length ................................................ 201
14
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
HOME RANGES, MOVEMENT PATTERNS, AND HABITAT USE OF A NEWLY
COLONIZED POPULATION OF FLORIDA BLACK BEARS (URSUS AMERICANUS FLORIDANUS) IN A FRAGMENTED FLORIDA LANDSCAPE
By
Dana L. Karelus
December 2017
Chair: Madan K. Oli Major: Interdisciplinary Ecology
Understanding how newly colonized populations of animals use the landscape in
anthropogenically fragmented habitats is important for species management because
animals in fragmented habitats may use the landscape differently than conspecifics in
contiguous habitats. Florida black bears (Ursus americanus floridanus) recently
recolonized Camp Blanding Joint Training Center and the surrounding private lands in
north central Florida. Using global positioning system (GPS) location data collected from
this newly established population, we investigated black bear space- and habitat-use,
movement patterns, and structural and compositional features of habitat influencing
heterogeneous use of areas.
Black bear home ranges in our study area were generally larger than those
reported for bears inhabiting the nearby contiguously forested habitat of Ocala National
Forest, indicating that habitat fragmentation may influence home range size. Bears
moved more slowly and less directedly when near creeks, in forested wetlands, and in
marsh habitats, possibly indicating foraging behavior. In urban areas, bears moved
15
more quickly and along more directed paths. Major roads tended to act as a semi-
permeable barrier to bear movement
Three movement states best described the bears’ movement pattern; a state with
short step-lengths and high turning angles (likely resting), a state with moderate step-
lengths and high turning angles (likely foraging), and a state with long step-lengths and
directed travel (likely traveling). Habitat selection analyses revealed that bears selected
most strongly for riparian forests and urban areas were generally avoided. Canopy
cover, visual obstruction, and hardwood density were important micro-habitat features
associated with areas that received high use by bears. The probability of high use was
positively associated with habitats closer to creeks and high canopy and shrub cover.
These results indicated that high bear-use sites were generally located in forested
wetlands.
Our results highlight the importance of forested wetlands for black bears
inhabiting human-dominated landscapes and reinforced the general importance of
riparian forests for bears in southeastern North America. Because forested wetlands
provide foraging and denning habitats and roads act as barriers to bear movement,
conservation planners should consider mitigating the impacts of future road
development on forested wetlands as a priority for bear conservation and for promoting
genetic connectivity.
16
CHAPTER 1 INTRODUCTION
Anthropogenic fragmentation and habitat loss remain among the greatest threats
to wildlife conservation worldwide (Fahrig 1997; Crooks 2002; Henle et al. 2004;
Melbourne et al. 2004; Bennett and Saunders 2010). Fitness of animals inhabiting these
altered environments could be negatively impacted if they need to devote extra time and
energy to acquire sufficient resources or to find mates (Andrén 1994; Andreassen et al.
1998). Animals inhabiting human-dominated landscape may also suffer an increased
mortality due to conflicts with humans and vehicular collision (Flather and Bevers 2002;
Fahrig and Rytwinski 2009; McCown et al. 2009; Benítez-López et al. 2010).
Additionally, large-scale habitat loss and fragmentation can cause population isolation,
thereby severely restricting or completely eliminating gene flow (Clobert et al. 2009;
Holderegger and Di Giulio 2010).
Addressing large-scale conservation issues in fragmented or otherwise human-
modified landscape requires an understanding of how animals inhabiting these altered
landscapes move, and utilize space and resources (Andreassen et al. 1998; Haddad
1999; Woodroffe 2011; Baguette et al. 2013). Knowledge about these aspects of a
species’ ecology can be helpful in implementing effective management, and
establishing wildlife corridors or landscape linkages to ensure the long-term persistence
of wildlife species of conservation concern (Beier and Noss 1998; Sawyer et al. 2011;
Saura et al. 2014). Due to the growing human population and the encroachment of
anthropogenic land-use in natural areas, species residing outside the boundaries of
protected lands often exist in human-dominated areas (Foley et al. 2005). Some
species readily adapt to human-dominated landscapes and exploit anthropogenic
17
features and food sources (Verdade et al. 2011; Bateman and Fleming 2012; Magle et
al. 2012). In such cases, the knowledge of spatial and movement ecology of animals in
fragmented landscapes can contribute to the mitigation of human-wildlife conflicts
(Wilton et al. 2014). Large carnivores in anthropogenically fragmented areas are of
special concern because of the relatively large spaces they require in the best of
habitats. They often travel long distances when dispersing and can thus incur higher
risks of road-related mortality (Spellerberg 1998; Hodgson et al. 2011; Costello et al.
2013).
Few studies have examined the pattern of space use, habitat selection, or
movement of carnivores in newly-settled, fragmented areas. Therefore, the overall goal
of my dissertation is to understand the ecology of a newly-colonized population of
Florida black bears (Ursus americanus floridanus) in a fragmented landscape in the
north peninsula of Florida. This dissertation is divided into four manuscript chapters,
addressing the following specific objectives:
1. Investigate the pattern of space and habitat use by black bears in the human-
dominated landscape of north Florida;
2. Examine movement patterns of black bears, and factors influencing these
patterns;
3. Investigate habitat selection by black bears in north Florida, while accounting
for movement patterns;
4. Discern the habitat features influencing black bears’ heterogeneous use of
habitat.
18
CHAPTER 2 HOME RANGES AND HABITAT SELECTION BY BLACK BEARS IN A NEWLY
COLONIZED POPULATION IN FLORIDA
Many carnivore populations have suffered precipitous declines due to habitat
loss and fragmentation (Woodroffe 2000; Crooks 2002; Ripple et al. 2014), but some
have responded positively to conservation efforts and have begun to recolonize portions
of their historic range (Linnell et al. 2001; Chapron et al. 2014; Gompper 2015).
Examples of such species include Pteronura brasiliensis Gmelin (Giant River Otter; dos
Santos Lima 2014), Canis lupus Linnaeus (Wolf; Pletscher et al. 1997), Gulo gulo L.
(Wolverine; Flagstad et al. 2004), Puma concolor L. (Cougar; Larue et al. 2012), Ursus
arctos L. (Brown Bear; Bjornlie et al. 2014, Hagen et al. 2015, Swenson et al. 1998),
and Ursus americanus Pallas (American Black Bear; Bales et al. 2005, Frary et al.
2011, Onorato et al. 2004, Unger et al. 2013). Animals colonizing new areas are
expected to distribute themselves among the best quality habitat available (Ideal Free
Distribution; Fretwell 1972). Therefore, understanding how species use space and
habitat as they naturally expand their range can help prioritize land management
practices and aid in corridor design for species of conservation concern (Beier et al.
2008; Marcelli et al. 2012; Bocedi et al. 2014), help identify suitable habitat for future
expansions of their population (Mladenoff et al. 1999), and possibly help reduce
human–wildlife conflicts (Wilton et al. 2014). But few studies have examined space- and
resource-use patterns of recently established populations of native carnivores.
Across the Black Bear’s range, many populations are thought to be growing
(Hristienko and McDonald 2007; Scheick and McCown 2014). As populations grow,
Black Bears settling into new areas should select the highest-quality habitats (Fretwell
1972). Habitat productivity and spatial arrangement of resources affect how Black Bears
19
use the landscape (Mitchell and Powell 2007). Black Bears living in particularly
productive habitats with rich nutritional resources should require a smaller home range
than what might otherwise be expected (Lindzey and Meslow 1977; Oli et al. 2002). The
patchy distribution of resources in anthropogenically or naturally fragmented landscapes
should require Black Bears to travel farther and thus to have larger home ranges than
those inhabiting unfragmented natural habitats (Hellgren and Maehr 1992; Mitchell and
Powell 2008). The increased travel needed to secure sufficient resources in
anthropogenically fragmented landscapes could also increase the risk of vehicular
mortality and conflict with humans (McCown et al. 2004; Baruch-Mordo et al. 2008;
Evans et al. 2014).
The size of a Black Bear home range varies seasonally due to the species’
annual physiological cycles and fluctuations in food availability (Hellgren et al. 1989;
Powell et al. 1997; Baruch-Mordo et al. 2014). Black Bears may use larger home ranges
in the fall while foraging more actively to prepare for winter denning (Hellgren et al.
1989; Moyer et al. 2007). Due to the high variability in space and resource use among
Black Bear populations, investigating the seasonal differences in home range size and
habitat selection can provide details that may otherwise be obscured.
Habitat for Black Bears must include 3 main resources; food, escape cover, and
sufficient vegetation or trees for denning sites (Powell et al. 1997; Reynolds-Hogland et
al. 2007). Diet of Black Bears consists mainly of plant matter (soft and hard masts); in
the Southeast, Serenoa repens (Bartram) Small (Saw Palmetto) is a particularly
important food source where available (Maehr and Brady 1984; Dobey et al. 2005).
Also, Black Bears in the Southeast generally prefer riparian forests and wetland habitats
20
(Hellgren et al. 1991; Wooding and Hardisky 1994; Stratman et al. 2001) over conifer
forests and open areas (Powell et al. 1997; Stratman et al. 2001; Moyer et al. 2008).
Intensively managed conifer forests often have relatively little understory and therefore
fewer sources of food than riparian and wetland habitats and do not provide adequate
cover for denning sites. Black Bears in Florida typically use ground nests for denning
and require thick understory for protection from disturbance (Garrison et al. 2012).
Roads may also influence space and habitat use by Black Bears, but responses vary
among populations and among individuals, depending on traffic volume, presence of
human activities, and habitat and vegetation along the road (Hellgren et al. 1991;
Gaines et al. 2005; Reynolds-Hogland and Mitchell 2007; Switalski and Nelson 2011;
Costello et al. 2013).
The subspecies of Black Bear in Florida, Ursus americanus floridanus Merriam
(Florida Black Bear; Black Bear hereafter), occurs in 7 relatively disconnected
populations across the state, but the overall population is growing and occupied range
is expanding (Florida Fish and Wildlife Conservation Commission 2012). The largest
population inhabits Ocala National Forest and surrounding areas in central Florida
(Florida Fish and Wildlife Conservation Commission 2012). A patchwork of public and
private lands, including the Camp Blanding Joint Training Center (hereafter Camp
Blanding; operated by the Florida National Guard), connects Ocala National Forest with
Osceola National Forest, which harbors another sizable Black Bear population (Figure
2-1; hereafter refered to as the corridor; Hoctor et al. 2000). Extensive sampling during
2002–2003 using hair snares revealed the presence of Black Bears in the corridor, but
there was no evidence for the presence of females with cubs and thus no evidence of a
21
population reproducing within the corridor (Dixon et al. 2006). However, based on
increased bear sightings as well as recovery of females killed on the road, a
reproductive population of Black Bears was suspected to have settled at Camp
Blanding and the adjacent corridor area (J. Walter McCown, Florida Fish and Wildlife
Conservation Commission, Gainesville, FL, unpubl. data).
Our objectives were to investigate space use and habitat selection by the
recently colonized population of Black Bears in the Camp Blanding area of north-central
Florida, USA. We hypothesized that Black Bears in our fragmented study site (1) would
have larger home ranges than those residing in nearby contiguous forests because they
would have to travel farther to acquire sufficient resources; (2) would have larger home
ranges in fall than in summer, similar to other Black Bear populations, because Black
Bears often forage more intensively before winter denning; (3) would select for riparian
forests, which provide the most cover and food sources; and (4) would avoid habitats
closer to major roads (but not necessarily minor roads) because of disturbance and the
risk of road-related mortality (McCown et al. 2009).
Field-Site Description
Our study was conducted at the 295-km2 Camp Blanding Joint Training Center
and adjacent private lands located in north-central Florida. Camp Blanding is located
near the center of the corridor between the Black Bear populations in Ocala National
Forest and Osceola National Forest (Figure 2-1). The area is fragmented by agricultural,
rural, and urban land uses and by several roads. The largest urban zones occur in the
cities of Starke and Keystone Heights and the unincorporated area of Middleburg. Pinus
L. spp. (Pine) plantations further fragment the natural vegetation communities and are
the dominant land cover at the study site. Natural habitats consist of mesic flatwoods,
22
sandhill uplands, and scrub, as well as hardwood swamps and hammocks that occur
near the creeks and drainages that traverse the area. Prevalent understory species
include Saw Palmetto, Myrica cerifera L. (Wax Myrtle), Ilex glabra (L.) Gray (Gallberry),
and Smilax L. spp. (Greenbriers).
Camp Blanding hosts military training activities multiple times per year that result
in an increased use of the training center property by several hundred to several
thousand troops. When training activities are not in progress, Camp Blanding is closed
to the public but allows controlled hunting and fishing by permit. Black Bear hunting in
Florida was illegal during our study.
Methods
We (B.K.S., J.W.M, and the field crew) captured Black Bears in the summers of
2011 and 2012 at baited sites using Aldrich spring-activated foot snares with a double-
anchor cable set (Scheick et al. 2009). The double anchor prevented Black Bears from
reaching either anchor tree, thus preventing injury to the animal from becoming
wrapped around a tree or limb while ensnared. We set traps during dawn and dusk
hours and attached a sentinel VHF collar to the anchor cable of each trap to monitor the
snares. We remained ≤2 km from trap sites and continuously monitored the VHF
signals; we responded within an hour of a Black Bear’s capture. We anesthetized each
captured Black Bear with Telazol® (3.5–5 mg/kg) and weighed, measured, and fit the
animal with a collar housing a global positioning system (GPS; Lotek WildCell MG)
tracking device, then released each individual at its capture site. We programmed each
collar to obtain locations every 2 hours and to drop off after 2 years, using a built-in
mechanism, but some collars fell off sooner. When female locations during winter
23
months indicated the possibility of denning, we visited the site to document
reproduction.
Animal handling was performed by biologists of the Florida Fish and Wildlife
Conservation Commission following agency policy and, as the wildlife regulatory
authority of the state, they need no permits.
Land Cover Categories
We used the raster format of the Florida Vegetation and Land Cover 2014
geographic information system (GIS) layer to classify land cover (Redner and
Srinivasan 2014); the layer had a resolution of 10 × 10 m. The study area contained 51
land-cover types; we grouped these into 6 categories (marsh/wetland, rural/agricultural,
urban, forested wetlands, wood/scrub, and tree plantations; Table 2-1) based on
similarity of landscape and vegetation (e.g., we combined all land cover categories of
marshes and wetlands that had open canopy cover) using the R package raster
(Hijmans 2015) (Appendix A). Urban areas consisted of medium to high density
residential, commercial, and industrial areas. Shapefiles for creeks and roads were
obtained from the Florida Geographic Data Library (http://www.fgdl.org/). Roads were
combined into 2 categories; major roads (Class 1: primary routes, including interstates
and U.S. highways; and Class 2: secondary routes, including state roads) and minor
roads (Class 3: larger roads or streets in residential areas; and Class 4: smaller roads
or streets in residential areas) using ArcMap (version 10.3; ESRI 2015).
Home Ranges
We prepared the Black Bear location data as described in Appendix A. We
estimated home-range size for each Black Bear based on the bihourly locations as 95%
utilization distribution using the kernel density estimator (KDE; Worton 1989) with
24
bivariate normal kernels. To determine appropriate bandwidth, we first estimated overall
KDE home ranges for each individual with the ad hoc bandwidth for the smoothing
parameter. We averaged the ad hoc bandwidth separately for females (0.389 km) and
males (1.39 km) because females have smaller home ranges (Hellgren and Vaughan
1990; Dobey et al. 2005) and then re-estimated KDE home ranges for each individual
using the sex-specific estimate of bandwidth. The bandwidths were biologically
reasonable (Powell et al. 1997) and larger than the estimated location error (20.3 m).
For comparison, we also estimated home ranges using 95% minimum convex polygon
(MCP; Mohr 1947).
We also estimated home ranges for 2 active seasons based on Black Bear
biology: summer (1 May–31 August) and fall (1 September–31 December). We
designated the beginning of the fall season as September because this corresponds to
the end of the breeding season as well as the beginning of acorn availability (Maehr and
Brady 1984; Moyer et al. 2007). An individual was included in a seasonal analysis if its
collar had been functional for at least 1 month during that season. Location data
collected during January–April were excluded because female Black Bears den during
that period (Moyer et al. 2007). Seasonal home ranges were estimated using the same
methods and average bandwidths as described previously.
We used the R package adehabitatHR (Calenge 2006) to estimate home ranges.
We used the nonparametric Wilcoxon rank sum test to compare home-range sizes
between females and males and between summer and fall (Conover 1999). All
statistical tests were performed in R (version 3.1.0; R Core Team 2013).
25
Habitat Selection
We performed compositional analysis of habitat selection (Aebischer et al. 1993)
at both second-order (selection of a home range within the study area) and third-order
(selection of land cover categories within a home range; Johnson 1980) scales. For
second-order habitat selection analysis, we estimated availability as the proportion of
area comprised by each land cover category in the study area, defined as the 99% MCP
calculated from all Black Bear locations. Use for second-order selection analysis was
estimated as the proportion of area comprised by each land cover category within the
99% MCP for each individual. For third-order selection analysis, the proportion of area
occupied by different land cover categories within each individual’s 99% MCP was
designated as available, and the proportion of each individual’s locations within each
land cover category was designated as usage. If a land cover category was not
available to an individual, we combined it with similar categories so that all were
available for all Black Bears. Any cases of 0 usage were replaced by 0.1 to avoid
problems associated with log transformation of 0, which is not defined (Aebischer et al.
1993).
We used Wilks’ Λ to test the null hypothesis that Black Bears used land cover
categories in proportion to the categories’ availability. If the null hypothesis was
rejected, the ranking matrix was computed and significance of preference of 1 land
cover category over another was determined using a randomization test (10,000
repetitions; Aebischer et al. 1993). We performed seasonal analyses in the same
manner. Compositional analysis of habitat selection was performed using the R
package adehabitatHS (Calenge 2006).
26
Habitat selection by animals is often influenced by measurable features on the
landscape, such as distance to nearest water source, road, or to an area of high human
activity. Compositional analysis does not permit the testing of how continuous
covariates might influence the pattern of habitat selection by animals. Thus, we used
mixed-effects logistic regression (MELR) with a binary response variable (1 = observed
GPS locations; 0 = random location; Gillies et al. 2006, Godvik et al. 2009, Klar et al.
2008, Nielsen et al. 2006). Random locations were represented by 5,000 randomly
generated locations within each Black Bear’s 99% MCP. Individual Black Bears were
treated as a random effect, which accounted for variation among individuals and the
nested structure of the data (Gillies et al. 2006). We considered fixed effects of land
cover category, distance to creek, major road, and minor road and the biologically
relevant additive effects of these covariates. We calculated the distances using the R
package rgeos (Bivand and Rundel 2016). Distances to creeks and roads were
standardized by subtracting the mean of the respective category from each value and
then dividing by the standard deviation; this method centered the mean on 0.
We fitted MELR models using the R package lme4 (Bates et al. 2015) with the
function glmer. For model comparison and statistical inference, we used an information-
theoretic approach using Akaike information criterion (AIC; Burnham and Anderson
2002, Klar et al. 2008) and considered models to have support if the difference in AIC
score was less than 2.0 from the highest ranked model. Fit of the MELR model was
assessed using the conditional coefficient of determination (R2GLMM(c); Nakagawa and
Schielzeth 2013); R2GLMM(c) was calculated using the R package MuMIn (Barton 2015).
27
Results
We fitted 16 Black Bears (6 females, ages 1 to 9 years; 10 males, ages 2 to 8
years) with a GPS collar and tracked them for a total of 5,362 bear-days, from June
2011 to August 2013. This yielded 46,922 bihourly, 3D-validated GPS locations (2932.6
± 88.4 per Black Bear; standard deviation, SD = 1415.1). All values reported indicate
mean ± SE unless otherwise indicated.
Home Ranges
Females had smaller home ranges than males (MCP: W = 2, P < 0.005, KDE: W
= 0, P < 0.005) (Figure 2-2). Overall 95% home range size for females estimated using
KDE (bandwidth h = 0.39 km) ranged from 12.53 km2 to 68.22 km2 and averaged 31.16
± 8.23 km2 (SD = 20.15 km2) and female home ranges from 95% MCP ranged from
10.07 km2 to 95.57 km2 and averaged 34.49 km2 ± 12.76 km2 (SD = 31.26 km2). Home
ranges for males estimated using 95% KDE (h = 1.39 km) ranged from 106.28 km2 to
387.65 km2 and averaged 220.93 ± 28.48 km2 (SD = 90.07 km2) and male home ranges
from 95% MCP ranged from 55.76 km2 to 528.06 km2 and averaged 226.04 km2 ± 45.32
km2 (SD = 143.33 km2). Annual KDE home range estimates from Camp Blanding and
from other studies of nearby Black Bear populations are presented in Appendix B.
Female home range sizes estimated using KDE for summer ranged from 8.45
km2 to 38.22 km2 with an average of 22.27 ± 3.57 km2 (SD = 11.28 km2); fall home
ranges were between 15.12 km2 to 59.31 km2 and averaged 27.78 ± 4.85 km2 (SD =
13.72 km2). Male home range sizes estimated using KDE for summer ranged from
59.02 km2 to 287.81 km2 with an average of 160.88 ± 20.96 km2 (SD = 59.29 km2); fall
home range sizes ranged from 89.30 km2 to 409.42 km2 with an average of 200.22 ±
28.60 km2 (SD = 90.45 km2). Summer and fall home range sizes were not significantly
28
different for females (KDE: W = 49, P = 0.46) or males (KDE: W = 53, P = 0.27) (Figure
2-2).
Habitat Selection
At the scale of second-order habitat selection, we concluded that selection
occurred over the entire study period (Wilks’ Λ = 0.414, P = 0.04) and for each season
(summer: Λ = 0.136, P = 0.003; fall: Λ = 0.326, P = 0.024). Forested wetlands were
significantly more preferred by Black Bears than marsh/wetland, rural/agricultural, or
urban for all 3 time periods. Urban areas were significantly less preferred by Black
Bears than all other land-cover categories except for rural/agricultural areas over the
entire study period and for fall. In summer, there was no significant difference among
preference for urban areas, wood/scrub, or rural/agricultural land cover categories
(Table 2-2).
Selection also occurred at the third-order scale over the entire study period
(Wilks’ Λ = 0.063, P < 0.001) and for summer (Λ = 0.050, P < 0.001) as well as fall (Λ =
0.032, P < 0.001) seasons. Black Bears preferred forested wetlands over all other land
cover categories for the entire study period and during the fall, but in the summer
forested wetlands and marsh/wetland were almost equally preferred. Generally, Black
Bears avoided habitat in rural/agricultural and urban land cover categories (Table 2-3).
The most parsimonious MELR model included an additive effect of land cover category,
distance to creeks, distance to major roads, and distance to minor roads (Model 1,
Table 2-4). The conditional R2 (R2GLMM(c)) was 0.281, suggesting no evidence for the
lack of fit of the MELR model to data. The next closest model differed from the top
model by >250 ΔAICc (Model 2, Table 2-4), indicating a substantial decrease in model
fit. Based on the most parsimonious model (Model 1, Table 2-4), Black Bears most
29
preferred forested wetlands and least preferred urban areas (Table 2-5). The effect of
distance to creeks and distance to major roads indicated that Black Bears used areas
closer to these features than expected at random. The effect of distance to minor roads
indicated that the Black Bears selected areas farther from these roads than expected at
random (Table 2-5). The variance (± SD) of the random effect was 0.397 ± 0.630.
Discussion
Although the presence of males in the Camp Blanding area had been reported,
previous studies, including Dixon et al. (2006), found no evidence of the presence of
females or a locally breeding population of Black Bears in the area. The earliest
available map of the distribution of Black Bears in Florida does not designate the area
as Black Bear range (Brady and Maehr 1985). During our study, we radio-collared 6
female Black Bears and also documented the birth of 5 cubs from 3 litters; these
findings provide evidence that a locally breeding population of Black Bears currently
inhabits the Camp Blanding area and that female Black Bears recently colonized the
area. This provided us with the opportunity to investigate space and resource use by a
newly colonized population of Black Bears in a human-dominated landscape with
substantial anthropogenic habitat fragmentation. Compared with relatively
unfragmented habitats in Ocala and in Osceola National Forests, the Camp Blanding
area exhibited a lower proportion of suitable habitat that was less aggregated, more
dispersed, and more patchily distributed across the landscape (Appendix C). Therefore,
we expected the Black Bears in our study site to have larger home ranges than those
inhabiting relatively unfragmented habitats.
We could not statistically compare our estimates of home range sizes with those
reported from other studies. That would require a consistent bandwidth among the
30
studies that used KDE and the same or a comparable number of locations among
studies that used either KDE or MCP (Seaman and Powell 1996; Laver and Kelly 2008;
Kie 2013; Börger et al. 2014). The home range studies of nearby Black Bear
populations did not report bandwidths, and the number of locations varied widely among
studies. Qualitatively, overall and seasonal Black Bear home ranges in the Camp
Blanding area were larger than those for Black Bears in Ocala National Forest, except
for females in 2000 (Moyer et al. 2007). An extreme, prolonged drought occurred in
Florida from 1998 to 2001 that resulted in a forest-wide mast failure in Ocala National
Forest (McCown et al. 2004), likely causing the Black Bears to use substantially larger
home ranges in 2000 to meet their resource needs. Black Bear home ranges in Osceola
National Forest and Okefenokee National Wildlife Refuge (Dobey et al. 2005) were
comparable to or larger than those in the Camp Blanding area. But much of the data
used in Dobey et al.’s (2005) study were also collected during the drought years, which
could have led to larger home ranges. Like Camp Blanding, Eglin Air Force Base and
the surrounding landscape are fragmented and receive substantial military use (e.g., as
airfields, test ranges, and sewage spray fields), which likely causes resources in that
area to be more dispersed and thus may explain the fairly large Black Bear home
ranges reported by Stratman (1998). In addition to fragmentation, the quality of habitat
also influences home range size. The smallest reported American Black Bear home
ranges in the southeastern United States have been reported for highly productive
habitats in the Mississippi Delta region (Benson and Chamberlain 2007, Oli et al. 2002)
and the Black Bears in the Camp Blanding area used much larger home ranges.
Therefore, our results are generally consistent with the expectation that Black Bears
31
inhabiting less productive or fragmented habitats, or a combination of the two, would
use larger home ranges than those in unfragmented or more productive habitats.
Most of the Black Bears in our study site exhibited larger home ranges in fall than
in summer, although the differences were not significant. This tendency for larger fall
home ranges is attributed to the increased foraging area during fall hyperphagia
experienced by Black Bears in preparation for winter denning and is consistent with
findings in the Ocala Black Bear population and several other populations (Hellgren et
al. 1989; Powell et al. 1997; Moyer et al. 2007). Therefore, our failure to detect a
significant difference between summer and fall was most likely due to our small sample
sizes.
Black Bears in the Camp Blanding area consistently preferred forested wetlands
over all other types of land cover at both second- and third-order scales during the
entire study period as well as in the summer and fall. Black Bears also selected for
areas close to creeks. Together, these results suggest that riparian forests represent
the best-quality habitat for Black Bears in the area. This is not surprising because
forested wetlands include relatively abundant mast from oaks and palmettos, a thick
understory for ground den sites and cover, and connectivity with other habitats (Hellgren
et al. 1991; Wooding and Hardisky 1994; Stratman et al. 2001). Black Bears generally
avoided agricultural, rural, and urban land cover at both scales of selection in all
seasons, most likely due to the lack of cover and higher levels of human disturbance.
However, this finding does not indicate that Black Bears avoided, or will avoid,
agricultural landscapes or urban areas. Black Bears can become habituated to humans
and alter their behavior to exploit food sources found in neighborhoods, especially when
32
resources are scarce (Beckmann and Berger 2003; Bateman and Fleming 2012;
Johnson et al. 2015). Securing garbage and other food sources early in the Black
Bear’s recolonization could help mitigate potential human–bear conflict.
There are many challenges inherent in the use-availability design of habitat
selection studies (Garshelis 2000; Beyer et al. 2010). For example, criteria used to
partition habitat types usually are arbitrary, distinction between habitat and non-habitat
is often blurred, measuring habitat units that are available to study animals is difficult,
and unbiased and error-free quantification of habitat use is rarely possible (Garshelis
2000). While we cannot rule out the possibility that some of our results may have been
influenced by aforementioned challenges, the concurrence between the results of
compositional analyses and mixed effect logistic regression models lead us to believe
that that our results are robust.
Black Bears used habitats closer to major roads and farther away from minor
roads than would be expected at random. These results may be a consequence of 2
major roads that bisect large blocks of forested habitat in the Camp Blanding area,
rather than Black Bears showing preference for areas closer to a major road. Several
home ranges spanned both sides of those roads, and 3 radio-collared Black Bears were
killed while crossing major roads. Similar results have been reported by Reynolds-
Hogland and Mitchell (2007), and Coster and Kovach (2012). Black Bears may have
stayed farther away from minor roads than expected due to high levels of disturbance
during military training exercises, deer-hunting season, and land-management activities
(van Manen et al. 2012; Morrison et al. 2014), but more data on human use of the area
would be required to determine whether that was the case.
33
Our findings suggest that Black Bears occupying fragmented habitats generally
require larger home ranges to acquire sufficient resources and reinforced the
importance of riparian forests. Conservation planning that focuses on preserving and
restoring riparian habitats and on maintaining or increasing the distribution and
abundance of soft- and hard-mast-producing plants in adjacent uplands will help ensure
the availability of essential resources for Black Bears. These management actions
would help increase the odds of colonization and persistence of stepping-stone
populations and would facilitate greater connectivity among Black Bear populations.
34
Table 2-1. Percentage of each land cover category composing the 99% minimum convex polygon constructed using locations from all Black Bears in the study (% Composition) and the percentage of Florida Black Bear GPS locations found in each land cover category (% Black Bear locations) in the Camp Blanding area in north-central Florida. See Appendix A for details.
Land cover category % Composition % Bear locations
(N = 46,922)
Marsh/wetland 6.90 7.75
Rural/agricultural 7.46 1.69
Urban 14.08 0.85
Forested wetlands 16.36 56.08
Wood/scrub 24.94 15.41
Tree plantations 30.25 18.21
35
Table 2-2. Ranking matrix from compositional analysis for second-order habitat selection (selection of a home range within the study area) by Florida Black Bears in north-central Florida for (A) the entire study period (1 August 2011–31 July 2013), (B) fall seasons, and (C) summer seasons. Signs indicate preference, with a (+) indicating that the row land cover category is preferred over the column land cover category and a (−) indicating the opposite. Triple signs represent a significant preference for (+++) or avoidance (+++) (P < 0.05). Rank represents the order of preference for the land cover categories, in order of most strongly preferred (1) to least strongly preferred (6).
Forested
wetlands
Tree
plantations
Woods/
scrub
Marsh/
wetland
Rural/
agricultural Urban Rank
A. Overall
Forested wetlands 0 + + +++ +++ +++ 1
Tree plantations − 0 + + + +++ 2
Woods/scrub − − 0 + + +++ 3
Marsh/wetland − − − − − 0 + +++ 4
Rural/agricultural − − − − − − 0 + 5
Urban − − − − − − − − − − − − − 0 6
36
Table 2-2. Continued
Forested
wetlands
Tree
plantations
Woods/
scrub
Marsh/
wetland
Rural/
agricultural Urban Rank
B. Fall
Forested wetlands 0 + + +++ +++ +++ 1
Tree plantations − 0 + +++ +++ +++ 2
Woods/scrub − − 0 + +++ +++ 3
Marsh/wetlands − − − − − − − 0 + +++ 4
Rural/agricultural − − − − − − − − − − 0 + 5
Urban − − − − − − − − − − − − − 0 6
37
Table 2-2. Continued
Forested
wetlands
Tree
plantations
Woods/
scrub
Marsh/
wetland
Rural/
agricultural Urban Rank
C. Summer
Forested wetlands 0 + + +++ +++ +++ 1
Tree plantations − 0 + +++ +++ +++ 2
Woods/scrub − − 0 +++ + +++ 3
Marsh/wetland − − − − − − − − − 0 + + 4
Rural/agricultural − − − − − − − − 0 + 5
Urban − − − − − − − − − − − 0 6
38
Table 2-3. Ranking matrix from compositional analysis for third-order habitat selection (selection of land cover categories within a home range) by Florida Black Bears in north-central Florida for (A) the entire study period (1 August 2011–31 July 2013), (B) fall seasons, and (C) summer seasons. Signs indicate preference, with (+) indicating that the row land cover category is preferred over the column land cover category and (−) indicating the opposite. Triple signs represent significant preference for (+++) or avoidance (---) (P < 0.05). Rank represents the order of preference for the land cover categories, in order of most strongly preferred (1) to least strongly preferred (5 or 6). For (A) overall and (C) summer, at least 1 Bear lacked availability in rural/agricultural and urban areas. Therefore, the test was repeated by combining these 2 land cover categories into 1 category.
Forested
wetlands
Marsh/
wetland
Woods/
scrub
Tree
plantations Urban
Rural/
agricultural Rank
A. Overall
Forested wetlands 0 +++ +++ +++ +++ 1
Marsh/wetland − − − 0 - +++ +++ 2
Woods/scrub − − − − − − 0 + +++ 3
Tree plantations − − − − − − - 0 +++ 4
Rural/agricultural − − − − − − − − − − − − 0 5
& Urban
39
Table 2-3. Continued
Forested
wetlands
Marsh/
wetland
Woods/
scrub
Tree
plantations Urban
Rural/
agricultural Rank
B. Fall
Forested wetlands 0 + +++ +++ +++ +++ 1
Marsh/wetland − 0 + +++ +++ +++ 2
Woods/scrub − − − − 0 +++ +++ +++ 3
Tree plantations − − − − − − − − − 0 + + 4
Urban − − − − − − − − − − 0 + 5
Rural/agricultural − − − − − − − − − − − 0 6
40
Table 2-3. Continued
Forested
wetlands
Marsh/
wetland
Woods/
scrub
Tree
plantations Urban
Rural/
agricultural Rank
C. Summer
Forested wetlands 0 + +++ +++ +++ 1
Marsh/wetland − 0 + + +++ 2
Woods/scrub − − − − 0 + +++ 3
Tree plantations − − − − − 0 +++ 4
Rural/agricultural − − − − − − − − − − − − 0 5
and urban
41
Table 2-4. Model selection results from mixed effects logistic regression testing for factors influencing habitat selection by Florida Black Bears in north-central Florida from 2011 through 2013. Models are sorted based on the ΔAICc (Akaike information corrected for small sample size) values in an ascending order. Land cover category is represented by Land cover. Major roads, Minor roads, and Creeks all represent distances to the nearest respective feature. The number of parameters in each model is indicated by K. The weight indicates the Akaike weight or model probability. Only the top 10 models, out of 16 total, are shown.
Rank Candidate model K Log-
likelihood ΔAICc Weight
1 Land cover + Major roads + Minor roads + Creeks 10 -70155.96 0 1
2 Land cover + Minor roads + Creeks 9 -70303.89 293.85 0
3 Land cover + Major roads + Minor roads 9 -70462.44 610.95 0
4 Land cover + Minor roads 8 -70608.92 901.92 0
5 Land cover + Major roads + Creeks 9 -70645.31 976.69 0
6 Land cover + Creeks 8 -70752.50 1189.07 0
7 Land cover + Major roads 8 -70876.29 1436.67 0
8 Land cover 7 -70985.09 1652.25 0
9 Major roads + Minor roads + Creeks 5 -76413.07 12504.20 0
10 Minor roads + Creeks 4 -76573.32 12822.72 0
Land cover categories: Wood/scrub, Marsh wetlands, Rural/agricultural, Urban, Tree plantations, and Forested wetlands
42
Table 2-5. Estimates (± SE) of slope parameters, as well as 95% confidence intervals, for the fixed effect variables included in the most parsimonious mixed effects logistic regression model (Model 1, Table 2-4). All slope parameters are significantly different from zero at P≤ 0.001.
Variable Estimate ± SE Confidence Interval
Land cover categorya
Marsh/wetland −0.300 ± 0.027 (-0.354, -0.247)
Woods/scrub −1.231 ± 0.019 (-1.268, -1.193)
Tree plantation −1.559 ± 0.017 (-1.593, -1.526)
Rural/agriculture −1.879 ± 0.041 (-1.960, -1.798)
Urban −2.697 ± 0.055 (-2.804, -2.590)
Distance to creeks −0.199 ± 0.008 (-0.215, -0.183)
Distance to major roads −0.131 ± 0.008 (-0.146, -0.116)
Distance to minor roads 0.241 ± 0.008 (0.226, 0.257)
a Reference = forested wetlands. Negative coefficients indicate that the respective land cover category is less strongly preferred than the reference category, forested wetlands. Positive coefficients would indicate that the category is preferred over the reference category.
43
Figure 2-1. Map showing the location of the Camp Blanding Joint Training Center and the closest designated primary Florida Black Bear ranges, the Ocala Black Bear population to the south and the Osceola Black Bear population to the northwest. The area between the 2 populations has been thought to act as a bear corridor and coincides with what is designated as secondary Black Bear range in this area. Major roads are shown and the largest human settlements in the corridor are labeled.
44
Figure 2-2. Average Florida Black Bear home range sizes in the Camp Blanding area based on bihourly GPS telemetry data. Home ranges were estimated using the minimum convex polygon (MCP) and kernel density estimator (KDE) methods for: (A) the entire study period (females: n = 6; males: n = 10), (B) Summer (females: n = 8; males: n = 10), and (C) Fall (females: n = 10; males: n = 8). Vertical bars represent standard error.
45
CHAPTER 3 EFFECTS OF ENVIRONMENTAL FACTORS AND LANDSCAPE FEATURES ON
MOVEMENT PATTERNS OF FLORIDA BLACK BEARS
Movement is fundamental for animals to obtain resources, escape threats,
disperse, and find mates. Therefore, movement affects population dynamics and
genetic connectivity among populations as well as affecting an individual animal’s
fitness (Morales et al. 2010). Knowledge of movement patterns can be used to improve
our understanding of animals’ habitat requirements, to predict future range expansions,
and to plan potential wildlife corridors in a more informed way (Colchero et al. 2011;
Buchmann et al. 2012; Avgar et al. 2013; Clark et al. 2015; Allen and Singh 2016).
Additionally, as more land is altered for human use, understanding animal movement in
fragmented, human-dominated landscapes may provide important insights into the
potential impact of human disturbance on wildlife (Belotti et al. 2012; Martin et al. 2013)
and suggest ways to reduce conflict with humans (May et al. 2010; Jachowski et al.
2013; Russell et al. 2013; Vasudev and Fletcher 2015).
Many factors affect an animal’s movement patterns. Individuals within a species
may exhibit different movement patterns depending on their sex, age or life-history
stage, and reproductive status (Aschoff 1966; Nathan et al. 2008; Laidre et al. 2013;
Martin et al. 2013; van de Kerk et al. 2014). Extrinsic factors such as habitat quality,
resource availability and access, as well as anthropogenic features on the landscape,
also influence animal movement (McClennen et al. 2001; Ager et al. 2003; Fahrig 2007;
Kauhala et al. 2007; Belotti et al. 2012; Kozakai et al. 2013). For example, many
animals move more slowly in resource-rich habitats than in poor quality or fragmented
habitats (Fryxell et al. 2008; Avgar et al. 2013; Ehlers et al. 2014; van Moorter et al.
2016). Additionally, anthropogenic features may impede or facilitate animal movement.
46
Animals may avoid crossing roads or traversing through areas with high levels of human
activity or buildings (Tigas et al. 2002; Revilla and Wiegand 2008; Holderegger and Di
Giulio 2010; Beyer et al. 2016). Alternatively, animals may use roads or other human-
made linear features as travel paths and thus may exhibit increased movement rates
and higher directionality in these areas (Tigas et al. 2002; Dickson et al. 2005; Roever
et al. 2010). Animals also may exhibit different movement patterns if they are attracted
to anthropogenic areas due to the availability of human foods (Rogers 1987; Tigas et al.
2002; Merkle et al. 2013; Lewis et al. 2015). Because large carnivores are highly
mobile, require relatively large spaces and a large amount of resources, and because
they can potentially come into serious conflict with humans, knowledge of their
movement patterns may be particularly useful for species and land management.
The Florida black bear (Ursus americanus floridanus) is a subspecies of
American black bear that currently occurs in 7 relatively isolated populations across the
range (Larkin et al. 2004; Dixon et al. 2006; Florida Fish and Wildlife Conservation
Commission 2012). The statewide population is thought to be increasing and bears are
recolonizing portions of their former range (Pelton et al. 1999; Dobey et al. 2005;
Hostetler et al. 2009; Florida Fish and Wildlife Conservation Commission 2012).
However, the human population in Florida is also increasing (U.S. Census Bureau
2017). Black bears are omnivorous habitat generalists, and therefore may utilize a wide
variety of habitats, including residential areas, which may lead to human-bear conflicts.
Threats to Florida black bears include habitat loss (due to habitat fragmentation and
residential and commercial development), road-related mortalities, and, in smaller
47
populations, low genetic diversity (Larkin et al. 2004; Dixon et al. 2007; Florida Fish and
Wildlife Conservation Commission 2012).
Space use by black bears varies widely across their geographic range, with
home range size typically showing an inverse relationship with habitat productivity (Alt
et al. 1980; Smith and Pelton 1990; Oli et al. 2002). Sex also influences space use by
bears; males generally use larger home ranges than females (Alt et al. 1980; Hellgren
and Vaughan 1990; Dobey et al. 2005). Furthermore, space use by bears of both sexes
varies seasonally in response to the bears’ physiological cycles and food availability; for
example, bears generally use larger home ranges in the fall when hyperphagia causes
them to forage more actively in preparation for winter denning (Garshelis and Pelton
1981; Hellgren et al. 1989; Noyce and Garshelis 2011). In the southeastern U.S., bears
tend to select riparian forests over more open habitats (Hellgren et al. 1991; Wooding
and Hardisky 1994; Stratman et al. 2001; Dobey et al. 2005; Karelus et al. 2016).
Florida black bears rely heavily on saw palmetto (Serenoa repens) and other hard and
soft mast as food sources, but also eat insects (Maehr and Brady 1984; Stratman and
Pelton 1999; Dobey et al. 2005).
While patterns of space and habitat use of bears are generally well understood
(Garshelis and Pelton 1980b; Masters 2002; Lewis and Rachlow 2011; Guthrie 2012),
few studies have directly investigated movement patterns based on fine temporal
scales, such as hourly or bi-hourly locations. Fewer still have tested for additive and
interactive effects of spatial or temporal factors on movement patterns, and investigated
how these patterns vary across temporal scales. Habitat and other extrinsic
environmental factors can affect animal movement (Fryxell et al. 2008; Nathan et al.
48
2008) but the temporal scale at which data are collected may influence the results.
Avgar et al. (2013) developed a framework for quantifying the effect of environmental
factors on movement of woodland caribou (Rangifer tarandus caribou) at various
temporal scales using the expected squared displacement, E(R2), as a primary
movement metric. This summary statistic can be used to describe movement patterns
and to make predictions at different temporal scales (Kareiva and Shigesada 1983;
Morales and Ellner 2002; Nouvellet et al. 2009). How environmental factors affect
movement metrics such as step-length, turning angles, and expected squared
displacement can provide insights into habitat quality for black bears, or predict how
landscape features might affect space use and dispersal.
We studied the movement patterns of Florida black bears in north-central Florida
at various temporal scales and, using the analytical framework developed by Avgar et
al. (2013), tested for the effects of intrinsic and extrinsic environmental factors thought
to influence animal movement. We predicted that 1) males would travel at higher
speeds (i.e., with longer step-lengths) and exhibit more directed movements than
females; and 2) black bears would travel faster in fall than in summer or winter. In terms
of environmental factors, we expected that black bears would 3) travel more slowly in
forested habitats and when near creeks; and 4) move shorter distances near major
roads.
Methods
Study Site
Our study site was in north-central Florida at Camp Blanding Joint Training
Center (295 km2) and adjacent private lands (Figure 3-1). Several creeks and drainages
run through the area. Natural habitats consist of mesic flatwoods, sand hill uplands and
49
scrub, as well as mixed hardwood hammocks and cypress swamps (Karelus et al.
2016). The natural vegetation communities are fragmented by roads, tree plantations,
agriculture, and human communities. Human disturbance on Camp Blanding varies
throughout the year. Military training activities occur at the base multiple times per year.
When military operations are not occurring, hunting, fishing, wildlife viewing, and hiking
are allowed on a portion of the property. Bears were not hunted in Florida during our
study.
Field Methods and Data Collection
We captured bears using Aldrich spring-activated foot snares with a double
anchor cable set (Scheick et al. 2009) and with culvert traps. The double anchor set
reduced the potential of injury to captured bears resulting from wrapping the cable
around a tree. We anesthetized each captured bear with Telazol® (3.5 – 5 mg/kg),
removed a pre-molar for aging (Willey 1974), fitted them with global positioning system
(GPS; Lotek WildCell MG) transmitting collars, then released them at the capture sites.
The collars obtained GPS locations every 2 hours and were programmed to fall off after
2 years. The collars were accurate to a 20 m radius for 95% of the locations (Karelus et
al. 2016). We visited the sites of suspected denning females to document reproduction.
Animals were handled by Florida Fish and Wildlife Conservation Commission staff using
methods that match the American Society of Mammalogy guidelines (Sikes and the
Animal Care and Use Committee of the American Society of Mammalogists 2016).
We recorded the landcover type, and distance to creeks, major roads, and minor
roads for each bear location. We used the Florida Vegetation and Land Cover 2014
geographic information system (GIS) raster layer (Redner and Srinivasan 2014), which
has a resolution of 10 x 10 m. We grouped landcover types with similar vegetation and
50
combined minimally available land cover types into 6 land cover categories. We
obtained the shapefiles for both creeks and roads from the Florida Geographic Data
Library (http://www.fgdl.org/). We classified primary routes (i.e., interstates and U.S.
highways) and secondary routes (state highways and county roads) as major roads. We
classified all other roads as minor roads (e.g. neighborhood roads or private roads that
were either paved or native materials). We calculated the distances from each location
to the nearest creek, major road, and minor road using the package ‘rgeos’ (Bivand and
Rundel 2016) in R (R Core Team 2014).
We divided the day into dawn (05:01 – 09:00 h), midday (9:01 – 17:00 h), dusk
(17:01 – 21:00 h), and night (21:01 – 05:00 h) to investigate diurnal variation in
movement. Because bears undergo seasonal physiological shifts (Hellgren et al. 1989),
we defined 3 seasons based on bear biology in Florida: winter (January 1 – April 30;
when bears typically den), summer (May 1 – August 31; when breeding occurs), and fall
(September 1 – December 31; when hard mast becomes available). We compared
winter movements of females with newborn cubs to females without newborn cubs and
summer movements for mothers with and without cubs, redefining summer season as
May 1 – July 31 based on the known survival of the cubs from cub VHF collars, game
cameras, and sightings. Cub VHF collars were lightweight and expandable to
accommodate the cubs’ growth (Garrison et al. 2007).
Movement Metrics
We investigated bear movements at bi-hourly (every 2 hours), daily, weekly, and
monthly temporal scales. We defined a day as starting at 08:00 and ending at the same
time on the consecutive morning. If a location was missing at 08:00, we used the next
location closest in time from between 06:00 and 10:00. If no locations from an individual
51
were obtained in that time frame on a day, both that day and the previous day were
removed from the daily scale data. We defined the beginning and the ending of each
respective week and month for an individual as the bear’s first and last location within
the time scale. We removed a day, week, or month for an individual if more than half of
the expected number of bihourly locations were missing within that time frame (e.g.,
daily required at least 6 locations per day). We calculated the following movement
metrics for all bihourly locations: step-length, directional heading, and directional
persistence. Step-length, l, is the straight-line distance between 2 successive bi-hourly
locations. Directional heading, θ, is the angular difference between the direction of the
step and 0̊. Directional persistence, c, is the cosine of the difference between 2
consecutive θs (Avgar et al. 2013). At the daily, weekly, and monthly scales, we
calculated the displacement (straight-line distance between the first and last location for
each bear), and overall heading, γ (angle made from the trajectory of the displacement
and 0)̊. We calculated the directional bias, q, as cos(γ - θ).
We used the expected squared displacement, E(R2), to assess overall movement
patterns at daily, weekly, and monthly scales while incorporating the statistical
properties of the movements (Nouvellet et al. 2009; Avgar et al. 2013). The method for
calculating E(R2) depended on the type of movement the animal displayed at each
temporal scale: either a biased random walk (BRW) or a correlated random walk
(CRW). We determined the type of movement by calculating the correlation between c
and q over each time scale for each individual; intervals with a positive and significant
correlation were classified as BRW, all others were considered to be CRW (Benhamou
52
2006). We calculated E(R2) for BRW (Codling et al. 2008; Avgar et al. 2013) and for
CRW (Benhamou 2006; Avgar et al. 2013), respectively, as:
𝐸(𝑅2) = 𝑛𝐸(𝑙2) + 𝑛(𝑛 − 1)𝐸(𝑞)2𝐸(𝑙)2 (3-1)
𝐸(𝑅2) = 𝑛𝐸(𝑙2) + 𝐸(𝑙)22𝐸(𝑐)
(1 − 𝐸(𝑐)) (1 −1 − 𝐸(𝑐)𝑛
1 − 𝐸(𝑐))
(3-2)
where n is the expected number of time steps in the given interval (e.g., 12
bihourly locations in a day and 84 bihourly locations in a week); E(l) is the average
bihourly step-length over the respective temporal scale; E(c) is the weighted average
directional persistence; and E(q) is directional bias. The directional persistence and
directional bias were weighted by the average step-length to account for potential
correlation between travel speed and direction (Avgar et al. 2013).
For each bear, we calculated the mean squared displacement (MSD), a measure
of how far an animal moves within a certain time interval (Kareiva and Shigesada 1983;
Benhamou 2006; Codling et al. 2008), as:
𝑀𝑆𝐷(∆𝑡) = 1
𝑁 − ∆𝑡∑ (𝑥𝑡𝑘+∆𝑡
− 𝑥𝑡𝑘)
2+ (𝑦𝑡𝑘+∆𝑡
− 𝑦𝑡𝑘)
2𝑁−∆𝑡
𝑘=1
(3-3)
where N is the number of locations for each bear, x and y are Universal
Transverse Mercator (UTM) coordinates of each location, and Δt represents the time
interval. MSD provides information about distance moved over a certain window of time
(i.e. extent of movement), and about whether or not the animal exhibited bounded
movement (Singh et al. 2016). The MSDs of individuals exhibiting confined movement
53
(e.g. within a home range) will reach a plateau with longer time intervals, whereas
MSDs of individuals exhibiting unconstrained movement (e.g., dispersal) will continue to
increase monotonically.
Statistical Analysis of Movement.
We used linear mixed models with a random effect of individual bear to examine
the effect of environmental covariates on the bi-hourly step-lengths and other movement
metrics at the bihourly, daily, and weekly temporal scales (Avgar et al. 2013). We tested
for the additive fixed effects of sex, season, distance to creeks, distance to major roads,
and distance to minor roads on the following movement metrics: bi-hourly step-length l,
average step-length E(l), observed displacement E(R2) (hereafter, expected
displacement), directional persistence E(c), and directional bias E(q). All variables
except E(c) and E(q) were natural log-transformed. Additionally, we tested for 2- and 3-
way interactions between season, sex, and each of the distance covariates. For models
at the bi-hourly scale, we also included the fixed effects of land cover and an interactive
effect of time of day. We incorporated 1st-order autoregressive error in all models to
account for autocorrelation in the data. The distances to creeks and roads in all models
were standardized by subtracting the mean and dividing by the standard deviation, thus
centering the mean on 0 and reducing convergence problems (Bolker et al. 2009).
We selected the most parsimonious model for each response at the weekly and
daily scales, while respecting marginality in the interactive terms by including the
respective main effects in each model, based on an information-theoretic approach
using the Akaike Information Criterion (AIC; Burnham and Anderson 2002; Klar et al.
2008). We assessed the fit of the models using the conditional coefficient of
54
determination (R2GLMM(c); Nakagawa and Schielzeth 2013). The R2
GLMM(c) was calculated
using the R package MuMIn (Barton 2015).
Analysis of Road Crossing
Roads can act as semi-permeable barriers to animal movement. To assess
whether, or to what extent, major and minor roads within the bears’ home ranges
affected their movement patterns, we analyzed the frequency of road crossing by bears.
We recorded the number of times each bear crossed either a major or minor road
between each successive bi-hourly location using ArcGIS (ESRI 2015) and Geospatial
Modeling Environment (Beyer 2015). We used 95% kernel density estimates as derived
in Karelus et al. (2016) for home ranges and used ArcGIS (ESRI 2016) to calculate the
length of each the major and minor roads within each bear’s home range. We analyzed
the number of road crossings for each road type separately using generalized linear
mixed models with a negative binomial distribution with linear-variance parameterization
due to overdispersion in Poisson models. We first tested for the fixed effect of sex
alone; we then tested for the additive effect of sex and linear road length (log
transformed and scaled to a mean of 0) to test for the effect of sex while accounting for
the length of the roads. The number of weeks that each bear was tracked was used as
an offset; thus, in effect, our response variable was the number of road crossings per
week.
We used the R package adehabitatLT (Calenge 2006) to calculate the movement
path descriptors. We fitted the mixed effect models with function “lme” in the package
‘nlme’ (Pinheiro et al. 2015) and generalized linear models with a negative binomial
response with the function “glmmTMB” in the package ‘glmmTMB’ (Magnusson et al.
2017); all analyses were performed in program R (version 3.2.2; R Core Team 2014).
55
Results
We collared 16 bears (6 females, 10 males; ages 1.5 to 9.5, all of potential
breeding-age) and tracked them for 5,812 bear-days between 2011 and 2014, yielding
58,951 bi-hourly 3D-validated GPS locations (mean ± SE: 2,907.3 ± 1,033.2 per bear).
When only considering high-quality, 3D-validated bi-hourly locations, the average fix
rate at which the locations were obtained was 0.84 ± 0.01 (range: 0.74 to 0.93). After
aggregating data at longer temporal scales, there were 4,628 daily (289.3 ± 103.6 per
bear), 711 weekly (44.4 ± 5.61 per bear), and 169 monthly (10.6 ± 3.9 per bear)
locations.
Overall, females moved with shorter step-lengths than males, with average bi-
hourly step-lengths (mean ± 1 SE) of 228.45 ± 21.88 m for females compared to 346.77
± 41.46 m for males (Table D-1). Step-lengths of both sexes varied across the diel
period during all seasons, with females having the longest step-lengths between 20:00
– 24:00 h and the shortest step-lengths between 14:00 – 18:00 h. Males showed a
similar pattern but had a longer period of shortest step-lengths (10:00 -18:00 h), which
were more similar over the seasons. Bears of both sexes traveled the farthest distances
during dusk in the fall and the shortest in the winter (Figure 3-2). During our study, only
2 mothers had sufficient GPS data after leaving their dens for analysis of movement
during summer. In the winter, females with cubs had shorter step-lengths than females
without cubs across the diel period; however, in the summer their step-lengths were
similar (Figure 3-3).
Based on MSD, all but 2 bears exhibited confined movement over time (Figure 3-
4), with males reaching their maximum displacement after approximately 160 h and
females after approximately 80 h. The 2 bears with unbounded movements were 2-year
56
old males, suggesting that these bears were dispersing. Directional persistence and
bias for females and males across all temporal scales and seasons were close to zero
indicating frequent turns throughout movement trajectories (Table D-1; Figures 3-5A
and 5B).
A total of 95.2%, 94.4%, and 95.3% of bi-hourly locations were classified as
CRW at daily, weekly, and monthly scales, respectively, with the remaining locations
classified as BRW. Observed and expected displacements increased with temporal
scale for both males and females, and displacements were larger for males than
females across all temporal scales (Table D-1). The daily, weekly, and monthly
expected displacement explained 59.9%, 55.2% and 46.0% of the variation in the
respective observed displacement (Figure 3-6). The expected weekly displacements
were longer than observed weekly displacements for females in fall and summer but
were similar in winter; however, the expected and observed weekly displacements were
similar for males in all seasons (Figure 3-5C and 3-5D).
We further analyzed weekly locations using linear mixed models (model results
based on bi-hourly and daily locations are presented in Appendix D and Appendix E).
The small sample size of monthly locations precluded more detailed statistical analyses
at that temporal scale. The most parsimonious model for E(l) at the weekly scale
included an effect of distance to creeks and 3-way interactions among season, sex, and
major roads as well as season, sex, and distance to minor roads (Table D-2). The top
model for observed weekly displacement included the 3-way interaction between
season, sex, and distance to major roads, and the 2-way interaction of season with
minor roads, but did not include distance to creeks (Tables 2-1 and D-2). The top model
57
for the weekly expected displacement included 2-way interactions between season and
distance to creeks, and a 3-way interaction between season, sex, and distance to major
roads (Tables 3-1 and D-2). The same variables generally were the most influential at
the daily scale (Table D-3).
The most parsimonious models for movement metrics revealed the following
patterns: males had longer step-lengths compared to females; bears of both sexes had
shorter step-lengths in winter than in summer or fall; and bears exhibited shorter
average step-lengths and displacements when they were closer to creeks (Figures 3-7
and 3-8). Also, females generally traveled shorter bi-hourly distances near minor roads,
whereas males generally traveled longer bi-hourly distances near minor roads. Both
sexes tended to have shorter step-lengths near major roads but females generally
responded more drastically to major roads than males (Figures 3-7 and 3-8). The most
parsimonious model for directional persistence at the weekly scale indicated that bears
of both sexes turned more near creeks and females turned more near major roads.
Males, on the other hand, exhibited more directed travel, especially during the winter,
when near major roads (Appendix E). The most parsimonious model for directional bias
indicated that females traveled in a more directed manner over the entire week when
they were closer to creeks and farther from minor roads; whereas males traveled in
more directed manner over the week when closer to minor roads and when farther from
creeks. However, there were some seasonal differences. In the summer and winter,
bears traveled along more directed paths over the week when farther away from creeks
(Appendix E). Models for expected weekly displacement and other movement metrics at
daily and bi-hourly scales indicated similar trends (Figures E-3 – E-9). Analyses at the
58
bi-hourly scale also indicated that bears generally traveled with the shortest step-
lengths when in forested wetlands and marsh wetlands, and with the longest step-
lengths when in rural, agricultural, and urban areas (Figure E-9).
Collectively, females crossed major roads a total of 11 times and males crossed
major roads 120 times (Table 3-2). There were 4 females and 3 males that did not cross
major roads; however, 2 of those females did not have any major roads within their
home range and were removed from further analysis of major road crossings. All bears
crossed minor roads. Females crossed minor roads a total of 2,928 times (range: 59 to
793) and males crossed minor roads 4,033 times (Table 3-2). Generally, males crossed
roads more frequently than females (Table 3-2). However, when we included the length
of roads within each individual’s home range as a covariate, the sex-specific differences
in the frequency of road crossing disappeared for major roads (slope parameters:
β[Male] = -0.61, 95% CI: -2.72, 1.50, β for road length = 1.49, 95% CI: 0.56, 2.43), as
well as for minor roads: β[Male] = -0.17, 95% CI: -0.96, 0.63, β for road length = 0.39,
95% CI: 0.04, 0.75).
Discussion
Male bears in our study area travelled at higher speeds (i.e., with longer step-
lengths) than females (Figure 3-2); these findings agree with results of earlier studies of
black bear movement (Garshelis et al. 1983; Masters 2002; Lewis and Rachlow 2011;
Guthrie 2012) and home range dynamics (Hellgren and Vaughan 1990; Dobey et al.
2005; Karelus et al. 2016). Both female and male black bears traveled the least during
the day; females tended to exhibit more crepuscular movements, whereas movements
by males were more nocturnal. Activity patterns of black bears in natural areas (using
active–non-active signals on VHF collars) generally follow a crepuscular pattern, with
59
bears being less active during the day or night (Amstrup and Beecham 1976; Garshelis
and Pelton 1980b; Masters 2002). A predominantly nocturnal activity pattern is
generally thought to be a strategy to avoid human disturbance (Lariviere et al. 1994;
Beckmann and Berger 2003; Lyons 2005; Matthews et al. 2006; Ordiz et al. 2012). In
our fragmented study area, male bears may have exhibited greater nocturnal
movements than females because males travel more widely and therefore are likely to
experience a greater degree of anthropogenic disturbance.
We expected that black bears would move faster in the fall than in summer or
winter because bears may have to travel greater distances to acquire sufficient food
resources during fall hyperphagia (Garshelis et al. 1983). Whereas females exhibited
the expected movement patterns, differences in travel speed of males during summer
and fall were less dramatic. Bears in the southern Appalachians traveled greater
distances in the fall than in the summer in only 1 of the 3 study sites (Garshelis et al.
1983). Females in the Okefenokee-Osceola area of Florida did not exhibit variation in
their movement speeds among seasons; however, some made notable long-distance
moves outside their home range in the fall (Masters 2002). Mate-seeking behavior by
males during the summer breeding season may lead to faster travel speeds (Alt et al.
1980; Smith and Pelton 1990; Lewis and Rachlow 2011), and could potentially have
masked the expected seasonal movement patterns.
Our results indicated a substantial reduction in travel speed during winter for both
females and males. We confirmed the birth of cubs to 3 collared females during our
study and females with newborn cubs consistently exhibited restricted movements in
the winter. In contrast, males and females without newborn cubs typically remained
60
more mobile in the winter throughout our study (Figure 3-3). This pattern was expected
because pregnant females must den in the winter, but all other cohorts of Florida bears
do not necessarily den (Wooding and Hardisky 1992b; Garrison et al. 2012).
For animals moving within the bounds of an established home range, the MSD
plot reaches an asymptote as the time interval (Δt) is increased. The MSD plot indicated
that the movements of female bears reached a maximum distance after approximately
80 h (3 days) and those of males after about 160 h (or 6 days; Figure 3-3), suggesting
that females cover their home range in about half the time taken by males to complete
the same action. The MSD of dispersing individuals tends to continue to increase after
that of other individuals has plateaued, reflecting the fact that their movements are not
restricted within confined areas. While two 2-year old males appeared to be dispersing
(Costello 2010), we could not confirm that these movements represented dispersal
rather than exploratory movements because both bears lost their collars after
approximately 3 months. More studies of dispersal, such as that conducted by Lee and
Vaughan (2003), are needed.
The expected squared displacement incorporates several movement descriptors
into a single value (Kareiva and Shigesada 1983; Morales and Ellner 2002; Nouvellet et
al. 2009; Avgar et al. 2013), providing information that is not contained in any other
single movement descriptor; furthermore, it allows predictions of an animal’s movement
pattern. The framework developed by Avgar et al. (2013) allows for the expected
squared displacement to be calculated assuming either a CRW or BRW, thus
accounting for periods where an animal may be exhibiting markedly different movement
patterns. Bears were unlikely to travel in a fixed direction or exhibit a strong directional
61
bias in their movement paths; consequently, only a small proportion of trajectories were
classified as BRW. This matches search strategy theory (Zollner and Lima 1999) in that
movements while foraging generally correspond to CRW (e.g., woodland caribou; Avgar
et al. 2013).
For both CRW- and BRW-designated bear movement paths, the expected
displacement tended to be larger than the respective observed displacement. Avgar et
al. (2013) attributed the overestimates of expected displacements of woodland caribous
to high primary productivity, which likely led the caribou to increase their foraging
activities. The average difference between expected and observed displacements for
bears was the highest in the fall, when Florida black bears forage intensely (Wooding
and Hardisky 1994; Moyer et al. 2007). However, the strength of the relationship
between observed and expected displacements also weakened with an increasing
temporal scale, highlighting the importance of temporal scales in the study of animal
behavior (McCann et al. 2017). The mismatch between the observed and expected
displacements, especially at high temporal scales, is likely due to the fact that CRW
models do not account for confined movements (Bergman et al. 2000; Fryxell et al.
2008; Auger-Méthé et al. 2016).
Results of mixed model analyses suggested that most movement metrics differed
between sexes and varied across seasons either in an additive or interactive fashion.
Distance to landscape features such as creeks and roads varied in their effect on the
movement metrics. Male bears traveled at higher speeds than females, and both sexes
traveled faster in fall than in winter. Movements of both sexes were least directed in the
winter and most directed in the summer. Directed travel may be advantageous to an
62
animal while searching for new resources or mates (Bailey and Thompson 2006;
Bartumeus et al. 2008; Gurarie and Ovaskainen 2013; Laidre et al. 2013; Wilson et al.
2013); thus, our results may indicate that the bears may be searching for highly
dispersed food resources in our fragmented study area or for mates in the summer.
We expected that bears would travel with slower speeds in high-quality habitats
because they should spend more time in these areas while foraging or resting, and
should travel faster through poor-quality habitat (Franke et al. 2004; Shepard et al.
2013). In general, bears moved slower and turned more frequently in forested wetlands
and marsh habitats, and areas near creeks. Forested wetlands generally are resource-
rich and provide cover and food (Hellgren et al. 1991; Wooding and Hardisky 1994;
Karelus et al. 2016), and creeks are an important source of water. In contrast, bears
moved with faster speeds in urban and rural or agricultural areas (Figure E-9). Faster
travel speeds with longer step-lengths in human-dominated landscapes may be a
strategy to minimize anthropogenic disturbance. Bears increase movement speeds and
have heightened stress responses when in open agricultural areas without edible crops
(Ditmer et al. 2015a). However, with experience, bears can learn to use anthropogenic
food sources (Ditmer et al. 2015b), which may change how they move through urban
areas.
Animals should reduce their travel speed when encountering landscape features
that increase resistance to movement and animals should move faster if some feature
of the landscape facilitates movement (Zeller et al. 2012; Avgar et al. 2013; Beyer et al.
2016). Black bears in our study generally exhibited slower movement-rates and
displacements when near major roads, except for males in the winter. In winter, males
63
exhibited faster movements and displacement near major roads. Shorter step-lengths
near major roads were either because of edge effects (Hellgren et al. 1991) or due to a
semi-permeable barrier effect of the roads (Whittington et al. 2005; Beyer et al. 2016).
In our study area, field observations of bear tracks persisting along stretches of
dirt roads suggested that bears used minor roads as travel pathways (J. W. McCown,
Florida Fish and Wildlife Conservation Commission, personal observation). In fact,
many large mammals including cougars (Puma concolor), bison (Bison bison), and
wolves (Canis lupis) use low-traffic roads as travel pathways (Hellgren et al. 1991;
Dickson et al. 2005; Bruggeman et al. 2007; Gurarie and Ovaskainen 2011;
Zimmermann et al. 2014). Bears may exhibit seasonal differences in how they use
areas near roads due to increased use by humans, such as hunting (Stillfried et al.
2015). When near minor roads, males traveled faster and more directedly, especially in
the summer.
Spatial heterogeneity can profoundly influence many aspects of an animal’s
ecology, including movement, with individuals occupying small habitat fragments being
most adversely affected. For example, using an experimental approach, Diffendorfer et
al. (1995) found that several species of small mammals travelled substantially longer
distances in small fragments; MSDs for individuals inhabiting small fragments were ~2-
fold greater than individuals inhabiting large fragments. Conversely, propensity to move
decreased as the patch size decreased (Diffendorfer et al. 1995). Roads present an
additional challenge to wildlife inhabiting urban areas because they often act as barriers
to animal movement and animals inhabiting such habitats tend to avoid roads (National
Research Council 2005; Leblond et al. 2013). Our analysis of road crossings indicated
64
that males were more likely to cross roads than females. This is likely because males
have larger home ranges and travel greater distances than females and are therefore
more prone to encounter roads in a fragmented habitat. However, when we included the
length of roads within individual home ranges as a covariate in our analysis, the
differences in frequency of road crossings between males and females disappeared.
Indeed, males tended to have more major roads within their home ranges than did
females; 2 of the females included in our study did not have major roads within their
home ranges (Table 2). Thus, females avoided crossing roads by establishing home
ranges away from major roads, which was not necessarily the case for males.
Therefore, the major roads in the Camp Blanding area likely acted as a semi-permeable
barrier to bear movement, an observation also reported by other studies of bear
movement (Brody and Pelton 1989; Beringer et al. 1990; McCown et al. 2009; van
Manen et al. 2012). In contrast, minor roads appear to have no effect on bear
movement. Bears of both sexes crossed minor roads with much greater frequency than
major roads, and there was no difference in this frequency between males and females.
Unlike major roads, the length of minor roads within each bear’s home range had no
effect on the road crossing rates, suggesting that minor roads present no resistance to
movement of bears of either sex.
Knowledge of movement ecology has direct implications for wildlife conservation
and management, because movement determines species’ geographic range and
animals’ ability to survive and reproduce (Nathan et al. 2008; Owen-Smith et al. 2010;
Barton et al. 2015). Allen and Singh (2016) suggested that linking animal movement
ecology with conservation requires knowing movement attributes of the animals and
65
how this knowledge can be used for the planning and implementation of management
actions. Understanding movement patterns may be even more important for the
management of large mammals inhabiting anthropogenically fragmented landscapes
where they may come into conflict with humans (Tigas et al. 2002; Kertson et al. 2011;
Goswami et al. 2015) or die from vehicular collisions (Hostetler et al. 2009; McCown et
al. 2009; Benson et al. 2011; Basille et al. 2013). In this study, we provided data on
movement patterns of Florida black bears inhabiting a highly fragmented landscape,
and showed that male and female black bears exhibit fundamentally different movement
patterns, that reproductive status strongly influences movement of females, and that
bears may alter their movement patterns depending on habitat quality and
anthropogenic disturbance. Bears of both sexes travelled at slower speeds and
exhibited less directed movements when near creeks, marshes, or forested wetland
habitats, highlighting the importance of forested wetlands for black bears inhabiting
human-dominated landscapes. Because forested wetlands provide foraging and
denning habitats, conservation planners should consider mitigating the impacts of future
road development on forested wetlands as a priority for bear conservation and for
promoting genetic connectivity.
Conservation of species that travel long distances, such as migratory species,
often requires managing the entire ecological network, which includes breeding and
wintering grounds as well as stopover areas (Faaborg et al. 2010; Brower et al. 2012;
Allen and Singh 2016). We suggest that management of highly mobile species
inhabiting fragmented landscapes necessitates a similar approach, whereby
management efforts should not only focus on habitat patches, but also the linkages
66
among those patches. Such an approach would not only benefit the target species but
also promote ecological connectivity and contribute to a broader goal of biodiversity
conservation in anthropologically fragmented landscapes (LaPoint et al. 2015).
67
Table 3-1. Model selection statistics testing for the effect of various covariates on movement metrics of black bears (Ursus americanus floridanus) in north-central Florida: A) weekly average bi-hourly step-length, E(l), B) weekly average directional persistence, E(c), C) weekly average directional bias, E(q), D) weekly observed displacement and E) weekly expected
displacement, √(E(𝑅2)). Models appear in order of the difference in the
Akaike Information Criterion corrected for small sample sizes (ΔAICc). The difference in the log-likelihood from the top model (ΔLL), model probability (Weight), conditional coefficient of determination (R2
GLMM(c)), and the number of parameters (K) are also given. A plus sign (+) indicates an additive effect, whereas a colon (:) indicates an interactive effect. Covariates are: Creeks, Minor roads, Major roads (distances from each location to the respective feature); Sex (females or males), and Season (summer, fall, and winter). See Appendix D for a complete list of models.
Rank Model K ΔLL ΔAICc Weight R2GLMM(c)
A) Weekly average step-length, E(l)
1 Creeks + Major roads +
Minor roads + Season +
Sex + Major roads:Season
+ Major roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
22 0.00 0.00 0.15 0.41
2 Creeks + Major roads +
Minor roads + Season +
Sex + Major roads:Season
+ Major roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 -4.26 0.03 0.15 0.40
68
Table 3-1. Continued Rank Model K ΔLL ΔAICc Weight R2
GLMM(c)
B) Average Weekly directional persistence, E(c)
1 Creeks + Major roads +
Minor roads + Season +
Sex + Major roads:Season
+ Major roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 0.00 0.00 0.15 0.31
2 Creeks + Major roads +
Season + Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
16 -4.87 1.31 0.08 0.29
69
Table 3-1. Continued Rank Model K ΔLL ΔAICc Weight R2
GLMM(c)
C) Average Weekly directional bias, E(q)
1 Creeks + Minor roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Minor
roads:Sex
13 0.00 0.00 0.09 0.09
2 Creeks + Minor roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Minor
roads:Sex + Season:Sex
15 2.03 0.11 0.08 0.09
70
Table 3-1. Continued Rank Model K ΔLL ΔAICc Weight R2
GLMM(c)
D) Weekly Observed displacement
1 Major roads + Minor roads
+ Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 0.00 0.00 0.17 0.31
2 Creeks + Major roads +
Minor roads + Season +
Sex + Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 2.00 0.22 0.16 0.31
71
Table 3-1. Continued Rank Model K ΔLL ΔAICc Weight R2
GLMM(c)
E) Weekly Expected Displacement
1 Creeks + Major roads +
Minor roads + Season +
Sex + Creeks:Season +
Major roads:Season +
Major roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 0.00 0.00 0.14 0.47
2 Creeks + Major roads +
Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 -2.53 0.83 0.09 0.46
72
Table 3-2. Average number of weeks that bears were monitored, average road length within individual home ranges (km), and the average number of road crossings by female and male Florida black bears (Ursus americanus floridanus) in north-central Florida. All values are shown ± SE (minimum to maximum).
Weeks
monitored
Road length (km) Number road crossings
Major roads Minor roads Major roads Minor roads
Females 74.4 ± 5.8
(51.9, 88.9)
3.2 ± 1.3
(0, 8.4)
58.3 ± 19.8
(24.4, 154.9)
1.8 ± 1.3
(0, 8)
488 ± 110.2
(59, 793)
Males 36.4 ± 5.7
(6.3, 59)
35.5 ± 7.8
(10.4, 88)
436.5 ± 57.1
(170.7, 790.3)
12 ± 6.1
(0, 65)
403.3 ± 77.3
(44, 790)
Combined 50.6 ± 6.2
(6.3, 8.2)
23.4 ± 6.3
(0, 88)
294.7 ± 59.2
(24.4, 790.3)
8.2 ± 4
(0, 65)
435.1 ± 62.2
(44, 793)
73
Figure 3-1. Map of the study site at Camp Blanding Joint Training Center, Florida. Roads and creeks are also shown.
74
Figure 3-2. Average bi-hourly step-length (± 95% CI) in meters throughout the diel period for Florida black bears (Ursus americanus floridanus) in north-central Florida for: A) females by season (summer [7,450 locations], fall [8,124 locations], and winter [8,456 locations]; n = 6 for all seasons), and B) males by season (n = 10 in fall [5,820 locations] and summer [7,495 locations], n = 7 in winter [7,126 locations]). Winter was defined as January 1 – April 30, summer as May 1 – August 31, and fall as September 1 – December 31 based on the biology of Florida black bears (see text).
75
Figure 3-3. Average bi-hourly step-length (± 95% CI) in meters throughout the diel period for female Florida black bears (Ursus americanus floridanus) in north-central Florida with and without cubs of the year during: A) winter (denning mothers: n = 3 [1,666 locations]; females without newborn cubs: n = 6 [6,492 locations]), and B) summer (mothers with new cubs: n = 2 [1,589 locations]; females without new cubs: n = 4 [3,772 locations]). Winter was defined as January 1 – April 30, summer as May 1 – August 31, and fall as September 1 – December 31 based on the biology of Florida black bears (see text).
76
Figure 3-4. The mean squared displacement (MSD) over different time intervals for: A) individual female (n = 6; 24,030 locations) and B) male (n = 10; 20,441 locations) Florida black bears (Ursus americanus floridanus) in north-central Florida. The MSDs reaching asymptotes indicate confined movement (i.e., within a home range).
77
Figure 3-5. Weekly average directional persistence, E(c), and directional bias, E(q), for: A) female and B) male Florida black bears (Ursus americanus floridanus). Also presented are weekly average observed displacement and average expected displacement for: C) Female and D) Male Florida black bears.
78
Figure 3-6. Observed versus expected displacements for bi-hourly location data from Florida black bears (Ursus americanus floridanus) in north-central Florida at: A) daily, B) weekly, and C) monthly temporal scales for both biased random walks (BRW) and correlated random walks (CRW). The line on each graph indicates a perfect correlation between the observed and expected values.
79
Figure 3-7. Effect of covariates on the weekly average bi-hourly step-length (± 95% CI) for Florida black bears (Ursus americanus floridanus) in north-central Florida: A) main effect of distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, and C) 3-way interaction among sex, season, and distance to minor roads. All weekly average step-lengths are on the log scale and all distances are standardized.
80
Figure 3-8. Effect of covariates on the weekly observed displacement (± 95% CI) for Florida black bears (Ursus americanus floridanus) in north-central Florida: A) 3-way interaction among sex, season, and distance to major roads, and B) 2-way interaction between sex and distance to minor roads. All weekly displacements are on the log scale and all distances are standardized.
81
CHAPTER 4 INCORPORATING MOVEMENT PATTERNS TO DISCERN HABITAT SELECTION:
BLACK BEARS AS A CASE STUDY
As animals traverse the landscape, they make choices about where and how to
move as they search for resources and mates, and seek safety from predators or
disturbance. Their choices regarding where to move, the pattern of movement within
their home ranges, and the amount of time spent in each habitat component defines
space- and habitat-use patterns. In essence, space-use and habitat selection by
animals are emergent properties of their movement patterns. Therefore, a thorough
understanding of space- and habitat-use patterns requires knowledge of animals’
movement patterns (Moorcroft and Barnett 2008; Forester et al. 2009; Moorcroft 2012;
van Moorter et al. 2016). This understanding is particularly important in human-
dominated landscapes where higher quality habitats are typically separated by
degraded habitats, thus requiring animals to move more extensively which can increase
mortality risks from vehicular collision and other anthropogenic factors (Forman and
Alexander 1998; Tigas et al. 2002; Buchmann et al. 2012; Beyer et al. 2016; Karelus et
al. 2017).
Internal physiological and behavioral states of animals are important
determinants of animal movement patterns (Fryxell et al. 2008; Nathan et al. 2008;
Schick et al. 2008; Gurarie et al. 2015). For example, finding mates may be the primary
motivation for movement during breeding season, whereas finding foods or avoiding
predators may be the primary drivers of movement during non-breeding seasons.
Likewise, behavioral states such as traveling from one portion of the home range to
another, hunting for prey, or resting will leave signatures in movement patterns that may
be indicative of the relevant movement states. The observed movement patterns,
82
therefore, are determined by the interactive effects of animals’ internal physiological or
behavioral states, and extrinsic factors such as time of the year, habitat quality, or
barriers to movements (Jonsen et al. 2003; Nathan et al. 2008; Martin et al. 2013).
Although many of the extrinsic factors can be measured, animal’s behavioral states are
often difficult to quantify especially for species that are nocturnal, travel widely, occupy
forested or other dense habitats or are otherwise difficult to observe. A practical solution
to this problem is offered by hidden Markov movement models (HMMs), which have
been recently applied to animal movement studies because they permit identification of
hidden behavioral states based on observed movement trajectories (Langrock et al.
2012; Schliehe-Diecks et al. 2012; van de Kerk et al. 2014).
The HMMs are discrete-time, discrete-state, state-space models that use serially
observed data to explore the underlying, unobservable states causing the observed
patterns and the probabilities of transitioning among the states (Schick et al. 2008;
Patterson et al. 2009; Langrock et al. 2012; Zucchini and Macdonald 2016). When
applying HMMs to animal movement, the observed data typically consist of the animals’
step-lengths (distance between successive locations) and turning-angles (angle made
by three successive locations). The underlying unobservable states represent the
behavioral states. Using HMMs, it is possible to discern the number of states and
estimate the probability of transitioning between states and associated parameters
which, in turn, permits inferences regarding corresponding to behavioral states (Schick
et al. 2008; Visser 2011; Langrock et al. 2012). HMMs have been applied to animal
movement studies with biologically insightful results (Franke et al. 2004, 2006; Boyd et
al. 2014; McKellar et al. 2014; van de Kerk et al. 2014). However, there have been
83
limited studies attempting to integrate state-specific movement patterns to habitat
selection.
Our objective was to investigate the pattern of habitat selection by a large
carnivore (American black bear [Ursus americanus]) inhabiting a highly fragmented
landscape while explicitly incorporating state-specific movement. Although black bear
habitat-use patterns have been studied in Florida and elsewhere in North America
(Onorato et al. 2003; Dobey et al. 2005; Benson and Chamberlain 2007; Moyer et al.
2008; Karelus et al. 2016), none have considered state-specific movement patterns
while making inferences about habitat selection. We fitted HMMs to high-resolution
Global Positioning System (GPS) location data to identify the appropriate behavioral
states underlying observed black bear movement patterns and to estimate the relevant
model parameters. Then, we applied Viterbi algorithms to assign the most likely
behavioral state to each step of individual bears’ movement paths (Zucchini and
Macdonald 2016). Finally, we used step-selection functions (SSFs) to compare each
movement step with simulated alternative steps a bear might have taken to make
inferences regarding black bear habitat selection (Fortin et al. 2005; Thurfjell et al.
2014). Thus, our analysis of habitat selection adequately incorporates state-specific
movement patterns.
Methods
Study Species and Site
Historically, black bears lived throughout most of North America, but they have
been extirpated from a large portion of their former range (Servheen et al. 1999; Pelton
2003; Scheick and McCown 2014). However, black bear populations are growing in
many parts of their range (Garshelis and Hristienko 2006; Hristienko and McDonald
84
2007; Scheick and McCown 2014), including in Florida, U.S.A. (Karelus et al. 2016). In
Florida, there are 7 relatively isolated bear populations (Florida Fish and Wildlife
Conservation Commission 2012). Two of these populations occur in Ocala National
Forest and in Osceola National Forest and bears recently colonized the area between
them (Karelus et al. 2016). Female bears typically have smaller home ranges and tend
to move with shorter step-lengths than male bears (Alt et al. 1980; Hellgren and
Vaughan 1990; Dobey et al. 2005; Karelus et al. 2017). Bears also change their
movements among seasons and by time of day (Garshelis and Pelton 1980b; Garshelis
et al. 1983; Bridges et al. 2004; Karelus et al. 2017). Black bears often have larger
home ranges in the fall than in other seasons because they increase their caloric intake
and thus increase foraging to prepare for winter (Garshelis and Pelton 1980a; Hellgren
et al. 1989; Noyce and Garshelis 2011). In Florida, male bears and non-pregnant
females are not obligated to den in the winter (Wooding and Hardisky 1992b; Garrison
et al. 2012).
Our study site was at the Camp Blanding Joint Training Center (295 km2) and
surrounding private lands in north-central Florida. The area lies in what is referred to as
the Ocala to Osceola corridor (Hoctor et al. 2000), although the area is not officially
designated as a wildlife corridor. Natural habitats in the area are dominated by mesic
flatwoods, sandhill uplands and scrub, as well as mixed hardwood hammocks and
cypress swamps (Karelus et al. 2016). Several creeks and drainages traverse the site.
Anthropogenic land uses, including tree plantations, agriculture, and rural and urban
developments, fragment the habitat (Karelus et al. 2016). The area is likely to become
85
more fragmented in the coming years due to the anticipated human population growth in
the state and the sprawling suburbs of the Jacksonville area (Zwick and Carr 2006).
Field Methods
We captured bears using Aldrich spring-activated foot snares with a double
anchor cable set (Scheick et al. 2009) and with culvert traps. The double anchor set
reduced the potential of injury to captured bears resulting from wrapping the cable
around a tree. We anesthetized each captured bear with Telazol® (3.5–5 mg/kg),
removed a premolar for age estimation (Willey 1974), fitted them with global positioning
system (GPS; Lotek WildCell MG) transmitting collars, then released them at the
capture sites. The collars obtained GPS locations every 2 hours and were programmed
to fall off after 2 years. The collars were accurate to a 20-m radius for 95% of the
locations (Karelus et al. 2016). We visited the sites of suspected denning females to
document reproduction. Animals were handled by Florida Fish and Wildlife
Conservation Commission staff following approved protocols.
Habitat Covariates
We extracted the landcover type for each location from the Florida Vegetation
and Land Cover 2014 geographic information system (GIS) raster layer (10x10 m
resolution; Redner and Srinivasan 2014). We grouped landcover types with similar
vegetation and combined minimally available land cover types into 6 landcover
categories: forested wetlands, marsh-wetland, rural-agricultural areas, tree plantations,
urban areas, and woodland-scrub (see details in Karelus et al. 2016).
We calculated the distances from locations to the nearest creek, major road, and
minor road using shapefiles for both creeks and roads from the Florida Geographic Data
Library (http://www.fgdl.org/). We classified primary routes (i.e., interstates and U.S.
86
highways) and secondary routes (state highways and county roads) as major roads. We
classified all other roads as minor roads (e.g. neighborhood roads or private roads that
were either paved or native materials). We calculated the distances from each bear
location to the nearest creek, major road, and minor road using the package ‘rgeos’
(Bivand and Rundel 2016) in R (R Core Team 2016). Because bears undergo seasonal
physiological shifts (Hellgren et al. 1989), we defined 3 seasons based on bear biology
in Florida: winter (1 Jan–30 Apr; when bears typically den), summer (1 May 1–31 Aug;
when breeding occurs), and fall (1 Sep–31 Dec; when hard mast becomes available).
Movement Metrics and Identification of Movement States
Using only successive bihourly locations, we calculated the step-lengths
(distance between successive locations) and turning-angles (the difference in angle
between the step from the previous location to the current location and the step from the
current location to the next location; where the turning angle of an animal continuing in
the same direction would equal 0 radians (i.e., 0°), and the turning angle of an animal
going back where it came from would equal π radians (i.e., 180°) along each bear’s
trajectory. We then used the step-lengths and turning angles as bivariate input data in
HMMs with various distributions for step-length (gamma, Weibull, log-normal, and
exponential distributions), and for turning angle (von Mises and wrapped Cauchy
distributions). To ensure that our models were numerically stable, we tested 30 different
sets of randomly chosen starting values for each distribution and number of states. We
expected that the movement patterns would include the signature of a resting state
(very short step-lengths and high degrees of turning), a traveling state (long step-
lengths and more directed travel), and potentially also include a state with moderate
87
step-lengths, which could represent foraging. Thus, we tested HMMs with 2 and 3
biologically meaningful movement states.
Black bear home range sizes and movements typically vary between the sexes
and among seasons (Alt et al. 1980; Garshelis and Pelton 1981; Hellgren et al. 1989;
Powell et al. 1997; Moyer et al. 2007; Karelus et al. 2017); their movements also vary
across the diel period (Garshelis et al. 1983; Lewis and Rachlow 2011; Karelus et al.
2017). Therefore, we expected sex-specific differences and variation among seasons
and throughout the diel period. We tested for additive effects of sex, season, and time of
day on the probability of transitioning among states. We used an information-theoretic
approach with Akiake’s Information Criterion (AIC) to select the most parsimonious
model and to make statistical inferences (Burnham and Anderson 2002). Then, based
on the most parsimonious model, we used the Viterbi algorithm to assign the most likely
state to each step in the trajectories for all animals (Langrock et al. 2012; Zucchini and
Macdonald 2016).
Step-Selection Functions
Animal habitat selection is often studied using resource selection functions
(RSFs; Boyce et al. 2002; Manly et al. 2002). Despite many advantages, RSFs do not
explicitly incorporate movement patterns and also have been criticized for the manner in
which habitat availability is defined (Johnson and Nielsen 2006; Martin et al. 2008;
Fieberg et al. 2010; Thurfjell et al. 2014). A solution to these issues is provided by the
step-selection functions (SSFs), which take into account the animal’s movements and
the serial structure of GPS location data (Fortin et al. 2005; Forester et al. 2009; Squires
et al. 2013; Latombe et al. 2014; Thurfjell et al. 2014). SSFs use a Cox Proportional
Hazards modeling framework, in which each step in an animal’s trajectory (“used”) is
88
compared to multiple “unused” steps an animal might have taken during that time step
(Forester et al. 2009; Thurfjell et al. 2014). Thus, SSFs specifically consider where the
animal chose to move from available options. We incorporated HMMs into the SSF
framework by selecting unused steps based on the sex- and season-specific movement
parameters defining the respective behavioral state that was assigned to the step by the
Viterbi algorithm (Langrock et al. 2012; Zucchini and Macdonald 2016).
We selected 6 unused steps (Thurfjell et al. 2014) corresponding to each used
step by randomly choosing turning angles and step-lengths from within the distributions
of those parameters of the observed step’s respective assigned movement state. We
designated a unique step ID to each observed step and its corresponding unused steps.
We then performed conditional logistic regression (Hosmer Jr. et al. 2013) with the step
ID as the strata and a binary response (used versus unused). We tested singular and
additive effects of landcover type, distance to creeks, distance to major roads, and
distance to minor roads. We selected the most parsimonious model based on AIC score
(Burnham and Anderson 2002). We then repeated conditional logistic regression for
each season and for each movement state to investigate differences in seasonal habitat
selection or differences in selection based on behavioral state.
We used the R package moveHMM (Michelot et al. 2016) to calculate the
movement path descriptors, fit HMMs, and to assign the most likely states to the steps
with the Viterbi algorithm and the package CircStats (S-plus original by Ulric Lund and R
port by Claudio Agostinelli 2012) to calculate the average turning angles. We fitted the
conditional logistic models with the function “clogit” in the package survival (Therneau
89
and Grambsch 2000; Therneau 2015). All analyses were performed in program R
(version 3.3.1; R Core Team 2016).
Results
We tracked 16 bears (6 females, 10 males) for 5,812 bear-days from July 2011
to March 2014, which resulted in a total of 58,951 bi-hourly 3D-validated GPS locations
(mean ± SE: 2,907.3 ± 1,033.2 per bear). From these data, we estimated movement
descriptors for 28,485 locations for females and 23,773 locations for males.
Average step-lengths and turning angles differed between females and males
and both varied among seasons (Table 4-1). Females generally traveled shorter bi-
hourly distances and turned more frequently than males across all seasons. Females
and males traveled with the longest bi-hourly step-lengths and most directedly in the
fall, and the slowest bi-hourly speeds and least directedly in the winter. Bears of both
sexes exhibited differences in movement patterns throughout the diurnal cycle, with the
longest step-lengths at dawn and dusk for females, and at night for males, and the
shortest step-lengths at midday for both sexes (see Karelus et. al 2017).
We found that HMM models with 3 states, a Gamma distribution for step-length,
a wrapped Cauchy distribution for turning angle, and additive effects of sex, season,
and hour of day provided an adequate fit to our data (Table 4-2). However, due to the
large differences in movement patterns between females and males and among
seasons, we then fit 3-state HMMs separately for each sex in each season; all models
included a covariate of hour of day and the Gamma distribution for step-length and the
wrapped Cauchy distribution for turning angle. We identified the following 3 general
movement states: a state with short step-lengths and turning angles around 3.14
radians (180°; likely behavioral state: resting), one with moderate step-lengths and
90
turning angles around 3.14 radians (180°; likely behavioral state: foraging; though
turning angles for this state for males in the summer instead averaged around 0
radians), and a state with long step-lengths and turning angles around 0 radians (0°;
likely behavioral state: traveling; Figures 4-1 and 4-2). However, for females in the
winter and for both sexes in the summer, the distributions for both step-lengths and
turning angles in state 2 overlapped highly with those in either state 1 or state 3
(Figures 4-1 and 4-2). In the fall for both sexes, the resting, foraging, and traveling
states were the most distinct from each other (Figures 4-1 and 4-2). Bears typically
traveled during morning and evening hours but males also traveled during the night in
the winter and fall (Figure 4-3). During winter when the distributions of the resting and
foraging states were similar, females exhibited less of a diurnal pattern (Figure 4-3).
Ignoring seasonal variation, the most parsimonious SSF model included an
additive effect of landcover type and distance to major roads. Considered seasonally,
the most parsimonious models also included an additive effect of distance to creeks in
winter and summer, and of distance to creeks and minor roads in the fall (Table 4-3).
Overall and among all seasons, bears chose forested wetlands significantly more than
any other landcover type, except in summer, when bears also chose marsh wetlands
(Figure 4-4). Distance to major roads had a significant, positive effect in all models,
whereas the effects of distances to minor roads and to creeks were negligible in all
models that where they appeared (Table 4-4).
It is possible that bears can potentially select for different habitats while in
different behavioral states. Thus, we also tested for state-specific habitat selection
patterns. The most parsimonious model for bears in state 1 (resting) included only
91
habitat type and distance to creeks; whereas the most supported model for states 2
(foraging) and 3 (traveling) also included additive effects of distances to major roads,
minor roads, and creeks. Bears preferentially chose forested wetlands significantly more
than other land cover types while in foraging and traveling states; while in the resting
state they chose forested wetlands, rural and agricultural areas, and marsh wetlands
(Figure 4-5). Distance to major roads had a positive effect in the foraging and resting
states and the effects of distance to creeks and roads were negligible (Table 4-4).
There existed some degree of model selection uncertainty because some of the
SSF models were separated by ΔAICc <2 (Table 4-3). Thus, we also performed model
averaging for the top 5 models using AICc weight as weights. Model-averaged
estimates of odds did not substantially differ those from the most supported model for
each respective season or state (Table 4-4).
Discussion
Many wildlife species exhibit annual cycles in their behaviors and physiology, and
exhibit different movement patterns in different seasons. For example, species that
hibernate often must gather resources in the fall to prepare and then reduce their
movements in the winter. Also, males and females can exhibit different movement
patterns, especially during the breeding season, due to sex-specific differences in
reproductive strategies. Therefore, sex and season must be considered when
investigating animal movement and habitat selection. Because black bears exhibit
strong seasonal patterns in physiology, and females and males have different
reproductive strategies (Garshelis et al. 1983; Hellgren et al. 1989), we sought to
understand how movement patterns differ between sexes and vary among seasons and
how these translate into variation in habitat selection. To achieve these goals, we first
92
applied HMMs to identify behavioral states and tested for differences in those states by
sex and season. Then we linked state-specific movement patterns to their habitat
selection by incorporating our HMM results into SSFs.
We found that male bears tended to move farther bi-hourly distances and turn
around less than female bears. Both sexes moved at faster bi-hourly speeds during the
fall and moved the slowest in the winter. This pattern was expected because, in the fall,
bears must eat more to survive through winter and so typically travel more extensively
in search of food (Garshelis et al. 1983; Hellgren et al. 1989; Moyer et al. 2007; Lewis
and Rachlow 2011; Karelus et al. 2017).
An HMM with 3 movement states and covariates of sex, season, and hour of day
was best supported by data. In our models, covariates only affected the probability of
transitioning among states; therefore, we fit sex- and season-specific HMMs with 3
states and a covariate of hour of day. However, the distribution of step-lengths and
turning angles were essentially the same for 2 of the 3 states for both sexes in the
winter and summer. In the fall, all 3 states were more distinct from each other. The 2
clear movement states in the winter and summer likely represented the behavioral
states of resting (a movement state with very short step-lengths and sharp turns) and
traveling (a movement state with long step-lengths and directed travel). In the fall, we
interpreted the additional state as foraging (a movement state with moderate step-
lengths and sharp turns). Similar patterns have been found for other species (Franke et
al. 2004; Pohle et al. 2017), though the interpretation of a movement state with
moderate step-lengths (what we defined as foraging) may differ among species. For
Florida panthers (Puma concolor coryi), this state was simply considered “moderately
93
active” (van de Kerk et al. 2014).This type of state was considered “locally active at a kill
site” for wolves (Canis lupus; Franke et al. 2006). For black bears, “foraging” is a fitting
interpretation for this moderate state because their diet is composed of mostly plant
matter and they spend much of their time eating, especially in the fall (Maehr and Brady
1984; Stratman and Pelton 1999; Dobey et al. 2005).
In the winter, females were more likely to be in the resting/foraging states
throughout the entire diel period whereas males tended to be in the resting/foraging
states in midday and in the traveling state through the evening and nighttime hours. In
Florida, only pregnant females must den in the winter, whereas other bears do not
necessarily den (Wooding and Hardisky 1992a; Garrison et al. 2012); however, all
bears reduce their movements (Karelus et al. 2017). Only 3 females gave birth during
our study, so our results provide evidence that non-pregnant females may have been
using day beds (Rayl et al. 2014) for extended bouts throughout the winter and spent
little time traveling.
In the summer, females changed their movements to a crepuscular pattern of
activity whereby they spent the night and midday in resting/foraging states and were in
traveling states in the morning and evening. The step-lengths in each of their states
were similar to those in the winter. Conversely, males were more likely to be traveling
throughout the night and had longer distance movements in their traveling state
compared to those in the winter. Males may have increased their movements in the
summer as they were looking for mates, whereas females may have only been
responding to increased food availability after winter denning (Powell et al. 1997).
94
In the fall, both sexes exhibited faster bi-hourly movements in the traveling state
than in the other seasons and the foraging state was distinct from the resting and
traveling state. Again, both sexes were most likely to rest during midday and males
were more likely than females to be traveling at night. Bears in the southeast U.S. tend
to have the largest home ranges in the fall when they increase their caloric intake to
prepare for the winter (Hellgren et al. 1989; Powell et al. 1997; Moyer et al. 2007). Our
results indicate that bears in our study area not only increase their bi-hourly speeds of
travel, but also were more likely to forage (i.e. move moderate distances and turn
around) at night in fall.
Determining the appropriate number of states for HMMs is challenging because
Information Theoretic approaches for model selection tend to favor HMMs with more
states (van de Kerk et al. 2014; Li and Bolker 2017; Pohle et al. 2017). However,
additional states may not relate to an actual underlying biological process, but instead
may be a product of noise in the data, temporal autocorrelation, correlation within a
state, or individual heterogeneity (Li and Bolker 2017; Pohle et al. 2017). Black bears
exhibited variation in their movements throughout the day; therefore, we accounted for
temporal autocorrelation by including hour of the day as a covariate in our HMMs (Li
and Bolker 2017) and we attempted to account for correlation within a state by running
season-specific models. However, considerable overlap in step-length and turning
angles between the resting and foraging states in the winter for both sexes and in the
summer for females, and overlap between the foraging and traveling state for males in
the summer may indicate possible overfitting.
95
Our next step was to discern habitat selection patterns while explicitly
incorporating movement patterns using SSFs. We found that bears consistently chose
to move to forested wetlands significantly more than to human-modified land cover
types, across all seasons and generally in any behavioral state. These results suggest
that the bears actively avoid human-dominated, highly modified areas within the
landscape. An exception to the trend was that bears also used rural and agricultural
areas for resting. Many of the rural and agricultural areas where the bears rested were
within the Camp Blanding Joint Training Center, not on the adjacent agricultural lands.
These areas on the Camp Blanding had sparse human structures or open fields and
were therefore likely to have received limited human-use throughout the day most of the
year. Bears likely chose to move to forested wetland habitats because these areas
provide them with food, water, and cover (Powell et al. 1997). These results are
consistent with previous findings that black bears in southeastern populations also
select for forested wetlands (Hellgren et al. 1991; Wooding and Hardisky 1994;
Stratman et al. 2001), and generally travel at a slower speed (i.e., shorter step-lengths)
while in the forested wetlands (Karelus et al. 2017).
We also found that bears chose to move away from major roads. This result was
consistent across all seasons and when foraging and traveling. However, when bears
were in the resting state, roads did not have a significant effect on their movements. Our
previous analyses indicated that the bears exhibited shorter step-lengths when near
major roads (Karelus et al. 2017) and that bears selected areas closer to major roads
than at random (Karelus et al. 2016). Taken together, our previous results regarding
bear selection of areas closer to roads may have been influenced by bears in the
96
resting state, or may have been due to the configuration of the bears’ home ranges
within the landscape. Other studies have found that bears select for areas away from
roads (Costello et al. 2013), and also for areas closer to major roads (Reynolds-
Hogland and Mitchell 2007; Coster and Kovach 2012; Karelus et al. 2016). We also
found that minor roads and creeks had little to no effect on the bears’ choices. However,
it is likely that bear movement is more strongly influenced by major roads, minor roads,
and creeks at finer temporal scales.
In summary, our study provided a general framework for explicitly incorporating
sex-specific, seasonal, and diurnal variation in movement patterns while testing for
habitat selection. Our results showed that bear movement varied throughout the
seasons and they chose to move to forested wetlands. These findings were not
substantially different from those from past studies; however, we showed that bears
exhibited variation in their behavioral states throughout the year and that their
behavioral state had some influence on their choice of where to move. These results
highlighted the importance of forested wetlands.
Because movement patterns determine animal’s pattern of habitat selection, we
suggest that studies of habitat-selection analyses should include movement behaviors
whenever feasible (Thurfjell et al. 2014; Avgar et al. 2016). The HMM analyses and the
Viterbi algorithm allowed us to investigate the behavioral mechanisms that cause the
observed differences in space-use by sex and by season and to investigate how the
animals were using landcover types in the area. Though previous studies have explored
movement states by season and land cover (Patterson et al. 2009; Schliehe-Diecks et
al. 2012; McKellar et al. 2014) and several studies have used SSFs to incorporate
97
movement in habitat selection (Latham et al. 2011; Clark et al. 2015; McGreer et al.
2015; Zeller et al. 2016), we suggest that an integration of these two approaches (i.e.,
HMMs to investigate state-specific movement patterns, and SSF to subsequently
discerns habitat selection) offers a more complete and arguably more accurate
understanding of space and resource selection by animals.
98
Table 4-1. Average step-lengths (±SE) and turning angles by season for (A) female and (B) male bears in the Camp Blanding area in Florida based on GPS locations between 2011 and 2014.
Season Number
of bears
Number of
locations Average step-length ± SE (m)
Average turning
angle (radians)
A. Females Winter 9 9519 90.21 ± 36.73 1.84
Summer 13 8877 248.95 ± 24.51 0.87
Fall 10 10089 352.96 ± 21.98 0.53
B. Males Winter 7 7933 212.92 ± 44.48 1.67
Summer 15 6778 385.73 ± 41.71 0.45
Fall 11 9062 425.7 ± 35.19 0.33
99
Table 4-2. Model selection results from hidden Markov models (HMMs) testing for the number of movement states and factors that influenced the transition probabilities among movement states by Florida Black Bears in north-central Florida from 2011 through 2014. Only 2 and 3 states were tested because at the bihourly scale of the data without direct observations of behavior, discerning the biological relevance of 4 or more states would be difficult. The top 10 models are sorted based on the ΔAIC (Akaike Information Criterion) values in an ascending order. The weight indicates the Akaike weight or model probability.
Rank Model ΔLogLik ΔAIC Weight
1 3 State: Sex + Season + Hour 0.00 0.00 1.00
2 3 State: Season + Hour 372.65 733.30 0.00
3 3 State: Sex + Hour 1183.74 2343.47 0.00
4 3 State: Hour 1577.16 3118.31 0.00
5 3 State: Sex + Season 2115.89 4207.78 0.00
6 3 State: Season 2325.33 4614.65 0.00
7 3 State: Sex 3024.15 6000.29 0.00
8 2 State: Sex + Season + Hour 3124.18 6190.36 0.00
9 2 State: Season + Hour 3172.07 6282.14 0.00
10 3 Sate: Null 3332.07 6604.13 0.00
100
Table 4-3. Results of model selection for step-selection functions from conditional logistic regression from (A) all seasons and all behavioral states, (B) winter, (C) summer, (D) fall, (E) state 1, (F) state 2, (G) state 3, testing for factors influencing habitat selection by Florida Black Bears in north-central Florida from 2011 through 2014. Models are sorted based on the ΔAICc (Akaike information criterion corrected for small sample size) values in an ascending order. Landcover categories include: wood/scrub, marsh wetlands, rural/agricultural, urban, tree plantations, and forested wetlands. Major roads, minor roads, and creeks all represent distances to the nearest respective feature in km. The number of parameters in each model is indicated by K. The weight indicates the Akaike weight or model probability.
Rank Model K ΔLogLik ΔAICc Weight
A. Overall 1 Land cover + Major roads 6 0.00 0.00 0.45
2 Land cover + Major roads + Minor roads 7 0.43 1.15 0.25
3 Land cover + Major roads + Creeks 7 0.15 1.71 0.19
4
Land cover + Major roads + Minor roads
+ Creeks 8 0.58 2.84 0.11
5 Land cover 5 -41.74 81.47 0.00
B. Winter 1 Land cover + Major roads + Creeks 7 0.00 0.00 0.32
2
Land cover + Major roads + Minor roads
+ Creeks 8 0.85 0.29 0.28
3 Land cover + Creeks 6 -1.39 0.78 0.22
4 Land cover + Minor roads + Creeks 7 -0.57 1.15 0.18
5 Land cover + Major roads 6 -6.39 10.77 0.00
C. Summer 1 Land cover + Major roads + Creeks 7 0.00 0.00 0.34
2 Land cover + Major roads 6 -1.02 0.04 0.34
3
Land cover + Major roads + Minor roads
+ Creeks 8 0.25 1.51 0.16
101
Table 4-3. Continued Rank Model K ΔLogLik ΔAICc Weight
4 Land cover + Major roads + Minor roads 7 -0.80 1.60 0.16
5 Land cover 5 -15.00 26.00 0.00
D. Fall 1
Land cover + Major roads + Minor roads
+ Creeks 8 0.00 0.00 0.63
2 Land cover + Major roads + Creeks 7 -1.66 1.32 0.32
3 Land cover + Major roads + Minor roads 7 -3.98 5.96 0.03
4 Land cover + Major roads 6 -5.59 7.18 0.02
5 Land cover + Minor roads + Creeks 7 -28.68 55.36 0.00
E. State 1 1 Land cover + Creeks 6 0.00 0.00 0.50
2 Land cover + Major roads + Creeks 7 0.14 1.73 0.21
3 Land cover + Minor roads + Creeks 7 0.04 1.92 0.19
4
Land cover + Major roads + Minor roads
+ Creeks 8 0.18 3.65 0.08
5 Land cover 5 -5.10 8.19 0.01
F. State 2 1
Land cover + Major roads + Minor roads
+ Creeks 6 0.00 0.00 0.42
2 Land cover + Major roads + Minor roads 7 -1.04 0.09 0.40
3 Land cover + Major roads + Creeks 7 -2.56 3.12 0.09
4 Land cover + Major roads 8 -3.59 3.19 0.09
5 Land cover + Minor roads + Creeks 5 -12.33 22.65 0.00
102
Table 4-3. Continued Rank Model K ΔLogLik ΔAICc Weight
G. State 3 1
Land cover + Major roads + Minor roads
+ Creeks 8 0.00 0.00 0.78
2 Land cover + Major roads + Creeks 7 -2.36 2.72 0.20
3 Land cover + Major roads + Minor roads 7 -4.83 7.65 0.02
4 Land cover + Major roads 6 -7.08 10.17 0.00
5 Land cover + Minor roads + Creeks 7 -31.82 61.64 0.00
103
Table 4-4. Odds and 95% confidence intervals for the variables included in the most parsimonious conditional logistic models from (A) all seasons and all behavioral states, (B) all states in the winter, (C) all states in the summer, (D), all states in the fall, (E) state 1 (resting) in all seasons, (F) state 2 (foraging) in all seasons, and (G) state 3 (traveling) in all seasons (Table 4-3). Land cover types with odds less than 1 and confidence intervals that do not include 1, indicate that the landcover type was less likely to be chosen than the reference category, forested wetlands. Odds greater than 1 and confidence intervals that do not include 1, would indicate that the category is preferred over forested wetlands. For distance variables, odds greater than 1 indicates the bears chose to move away from the feature and odds less than 1 indicates the bears chose to move closer to the feature.
Variable Odds 95% Confidence Interval
A. All Marsh wetland 0.85 (0.806, 0.903)
Pine/Scrub/Flatwoods 0.70 (0.671, 0.726)
Rural/Agricultural 0.55 (0.502, 0.597)
Tree plantations 0.53 (0.512, 0.55)
Urban 0.39 (0.354, 0.437)
Major roads (km) 1.13 (1.104, 1.166)
B. Winter Marsh wetland 0.81 (0.732, 0.896)
Pine/Scrub/Flatwoods 0.63 (0.582, 0.691)
Rural/Agricultural 0.60 (0.503, 0.714)
Tree plantations 0.44 (0.411, 0.48)
Urban 0.33 (0.256, 0.417)
Major roads (km) 1.06 (0.9898, 1.1359)
Creeks (km) 1.0004 (1.0002, 1.0006)
104
Table 4-4. Continued Variable Odds 95% Confidence Interval
C. Summer Marsh wetland 0.91 (0.811, 1.011)
Pine/Scrub/Flatwoods 0.78 (0.725, 0.832)
Tree plantations 0.61 (0.574, 0.646)
Rural/Agricultural 0.36 (0.299, 0.443)
Urban 0.22 (0.174, 0.287)
Major roads (km) 1.14 (1.089, 1.203)
Creeks (km) 0.9998 (0.9996, 1.0001)
D. Fall Marsh wetland 0.68 (0.641, 0.721)
Pine/Scrub/Flatwoods 0.85 (0.778, 0.926)
Rural/Agricultural 0.61 (0.544, 0.688)
Tree plantations 0.52 (0.492, 0.547)
Urban 0.52 (0.454, 0.595)
Major roads (km) 1.15 (1.111, 1.195)
Minor roads (km) 0.9998 (0.9996, 1.00001)
Creeks (km) 0.9997 (0.9995, 0.9999)
105
Table 4-4. Continued Variable Odds 95% Confidence Interval
E. State 1 Rural/Agricultural 1.06 (0.681, 1.654)
Marsh wetland 0.98 (0.857, 1.11)
Pine/Scrub/Flatwoods 0.72 (0.614, 0.846)
Tree plantations 0.67 (0.605, 0.75)
Urban 0.53 (0.307, 0.925)
Creeks (km) 1.00 (1.0002, 1.0007)
F. State 2 Marsh wetland 0.95 (0.839, 1.07)
Pine/Scrub/Flatwoods 0.60 (0.547, 0.66)
Rural/Agricultural 0.52 (0.433, 0.632)
Tree plantations 0.49 (0.45, 0.529)
Urban 0.41 (0.32, 0.522)
Major roads (km) 1.27 (1.156, 1.398)
Minor roads (km) 1.0003 (1.00004, 1.0006)
Creeks (km) 1.0002 (0.9999, 1.0005)
106
Table 4-4. Continued Variable Odds 95% Confidence Interval
G. State 3 Marsh wetland 0.79 (0.737, 0.855)
Pine/Scrub/Flatwoods 0.71 (0.676, 0.741)
Tree plantations 0.53 (0.478, 0.585)
Rural/Agricultural 0.52 (0.502, 0.546)
Urban 0.38 (0.339, 0.43)
Major roads (km) 1.12 (1.091, 1.155)
Minor roads (km) 0.9998 (0.9997, 0.99998)
Creeks (km) 0.9998 (0.9996, 0.9999)
107
Figure 4-1. Plots of step-length parameter distributions from 3-state HMMs for Florida black bears in north-central Florida for A) females in winter, B) males in winter, C) females in summer, D) males in summer, E) females in fall, and F) males in fall. All distributions of step-lengths were from the Gamma distribution.
108
Figure 4-2. Plots of turning angle parameter distributions from 3-state HMMs from Florida black bears in north-central Florida for A) females in winter, B) males in winter, C) females in summer, D) males in summer, E) females in fall, and F) males in fall. All distributions of turning angles were from the wrapped Cauchy distribution.
109
Figure 4-3. Proportion of steps in each movement state across the diel period as assigned by the Viterbi algorithm for Florida black bears in north-central Florida by sex and season: A) females in winter, B) males in winter, C) females in summer, D) males in summer, E) females in fall, F) males in fall.
110
Figure 4-4. Predictive odds of a bear choosing a land cover type based on the full
conditional logistic models for the A) overall data and B) by season.
111
Figure 4-5. Predictive odds of a bear choosing a land cover type based on the full
conditional logistic models for each state; State 1 is resting (short step-lengths and sharp turning angles), State 2 is foraging (moderate step-lengths and sharp turning angles), and State 3 is traveling state (long step-lengths and directed travel).
112
CHAPTER 5 MICROHABITAT FEATURES INFLUENCING HETEROGENOUS HABITAT-USE BY
FLORIDA BLACK BEARS
Animals use landscapes heterogeneously because resources are not distributed
evenly (Benhamou 1992; Morris 1992; Mitchell and Powell 2007). Habitat selection
studies often focus on selection of home ranges from within the larger geographic area
or selection of habitat types within a home range (second- and third-order selection,
respectively; Johnson 1980). However, even within those habitat types, an animal may
select certain areas more than others based on microhabitat features. Investigating a
species’ habitat requirements at multiple scales may have important management
implications because animals may select for certain habitat features at fine scale that
could be obscured at larger spatial scales (George and Zack 2001). Therefore, an
understanding of the fine-scale structural features that influence animal use of habitat
could provide information pertinent to management of forested habitats for the
conservation of forest-dependent wildlife. Indeed, management plans that provide
specific information about habitat requirements at multiple scales are likely to be
successful (Freemark et al. 1995). With technological advances (e.g., Global Positioning
System [GPS] tracking), we now have the capability to investigate how animals use the
landscapes in more detail (e.g., Rayl et al., 2014).
Heterogenous use of habitats within home ranges is a common feature of
animals’ space-use patterns across taxa, including reptiles (Adolph 1990; Brewster and
Brewster 1991), birds (Saab 1999; Fletcher and Hutto 2008; Hansen et al. 2016), and
mammals (Price 1978; Longland and Price 1991; Leblond et al. 2010; Mitchell and
Powell 2012). Preferential use of certain sites within home ranges likely is motivated by
selection of specific sites for foraging, resting, evading predators or finding mates (Tew
113
et al. 2000; George and Zack 2001). Indeed, an animal’s memory of and selection for
microhabitat features may explain the formation of their home ranges (Spencer 2012).
Because large carnivores often occur at low densities and require relatively large home
ranges, an understanding of microhabitat features that influence the areas that they use
most often may have important management implications, especially in fragmented
landscapes where resources tend to be scarce and patchily distributed (Gittleman and
Harvey 1982; Noss et al. 1996).
It is well known that black bears (Ursus americanus) use habitat within their
home-ranges heterogeneously (Horner and Powell 1990; Hellgren et al. 1991; Benson
and Chamberlain 2007; Moyer et al. 2008). However, studies examining the
microhabitat features that influence selective use of certain areas within their home
ranges are rare. Our objective was to discern the structural and compositional features
of the microhabitat associated with high- and low-use areas by a recently colonized
population of Florida black bears (U. a. floridanus) in a fragmented landscape of north-
central Florida (Florida Fish and Wildlife Conservation Commission 2012; Karelus et al.
2016). We expected that areas of high bear-use would be characterized by greater
canopy cover, more visual obstruction, greater density of hardwoods, and more shrub
cover of food plants than low-use sites. We also expected that high-use sites would
typically be closer to creeks and farther from major roads than low-use sites.
Methods
Study Site
Our study site was located in north-central Florida at Camp Blanding Joint
Training Center and the surrounding private lands, encompassing a total area of
approximately 315 km2. The area contains a diverse array of natural land cover types,
114
including forested wetlands, mesic flatwoods, mixed hardwood hammocks, sand hill
uplands, and scrub. The north and south forks of Black Creek begin in the area. The
north fork is an outflow from Kingsley Lake on Camp Blanding, and the south fork is
formed from several small streams from sandhill seepages and wetlands on Camp
Blanding and the surrounding private lands. Black Creek is a tributary to the St. Johns
River. Natural land-cover types are fragmented by roads, tree plantations, agriculture,
and rural and urban land-use. Land-cover types in the area are described in detail by
Karelus et al. (2016).
Camp Blanding lands are jointly managed by the Florida Department of Military
Affairs and Florida Fish and Wildlife Conservation Commission. The area is managed
for several species, including gopher tortoises (Gopherus Polyphemus), the federally
endangered red-cockaded woodpecker (Picoides borealis) (Gregory et al. 2006), and,
more recently, for Florida black bears (Karelus et al. 2016). Camp Blanding also
supports several other wildlife species, including the endemic Black Creek crayfish
(Procambarus pictus), the federally threatened eastern indigo snake (Drymarchon
couperi), Sherman’s fox squirrel (Sciurus niger shermani), and the federally threatened
wood stork (Mycteria americana) (Gregory et al. 2006). Additionally, white-tailed deer
(Odocoileus virginianus) and turkey (Meleagris gallopavo) are popular game species
throughout the study site.
Bear Captures and GPS Data Collection
We trapped bears with Aldrich spring-activated foot snares using a double
anchor cable set (Scheick et al. 2009) in the summers of 2011 and 2012. We
anesthetized bears with Telazol® (3.5 – 5 mg/kg), fit them with Lotek WildCell MG GPS
collars and released them on site. All animals were captured and handled by Florida
115
Fish and Wildlife Conservation Commission biologists, consistent with agency protocols
and with guidelines of the American Society of Mammalogists (Sikes et al. 2016). We
set the collars to collect GPS coordinates of the bears’ location every 2 hours. The
collars were accurate to a 20.3-m radius for 95% of the locations (Karelus et al. 2016).
Identification of High- and Low-Use Areas Within Home Ranges
To identify the areas of high and low use by bears, we used each bear’s bihourly
GPS locations to estimate their respective utilization distributions (UD) using a dynamic
Brownian bridge movement model (DBBMM; Kranstauber et al., 2012). We used
DBBMM because it incorporates the temporal structure of GPS data and estimates
utilization distributions based on the animal’s movement trajectory (Dray et al. 2010;
Kranstauber et al. 2012; Byrne et al. 2014). We defined the 50% contour of each bear’s
DBBMM as that bear's high-use area and the area between the 75% and 99% contours
of each bear’s DBBMM as their low-use area (Figure 5-1). We then randomly chose 10–
13 of the respective bear’s GPS locations from within their high-use area and 10–13 of
their GPS locations from within their low-use area. If two randomly chosen locations
were within 50 m from each other, either from the same bear or two different bears, we
randomly selected a different location to avoid sampling the same area twice. If the area
around the initial bear location had been burned or cleared since the time the GPS
location was obtained, we selected a new bear location randomly from the respective
bear’s locations and use-category.
We estimated the DBBMMs with the R package ‘move’ (Kranstauber and Smolla
2016) and converted the UD contours to GIS shapefiles for visualization in ArcMap
(Version 10.4; ESRI 2016) using the R package ‘maptools’ (Bivand and Lewin-Koh
116
2016). Additionally, randomly selected bear locations were chosen using R software
(Version 3.3.1; R Core Team. <http://www.r-project.org/>, 2016).
Vegetation Sampling
At each selected location, we collected detailed data on canopy cover, visual
obstruction, tree density, and understory vegetation. Sampling occurred during March-
July 2016. We measured canopy cover in each cardinal direction with a spherical
densiometer (Lemmon 1956). We measured the visual obstruction in each cardinal
direction using a 2-m tall cover board that was divided into 4 sections, each 0.5-m tall
(Coulloudon et al. 1999). The cover board was vertically held 15 m away from the
central point and the proportion of visible space on the board was estimated while
kneeling to approximate bear-eye level. To quantify tree density, we used the point-
center-quarter method (PCQ; Cottam and Curtis, 1956; Mueller-Dombois and Ellenberg,
1974) and measured the distance from the center point to the nearest pine and
hardwood in each of the 4 cardinal quadrants (northeast, southeast, southwest,
northwest). We recorded the tree species and measured its diameter at breast height
(DBH at 130 cm from the ground) if ≥4 cm DBH. If no tree was located within 65 m of
the point within a quadrant, we treated it as a missing value for that quadrant. To
investigate understory vegetation composition, we made a 4 x 4 m plot at each site,
using the point as the southwest corner of the plot. We recorded each plant species
present inside the plot and estimated the percent cover and maximum height (excluding
height for vines) of each respective species in the plot. We did not distinguish among
species within the families of Cyperaceae (sedges), Poaceae (grasses), or Juncaceae
(rushes), and instead grouped these into one category of Poales.
117
We calculated the distance from each sampling location to the nearest major
road, minor road, and creek using the R package rgeos (Bivand and Rundel 2016) and
shapefiles obtained from the Florida Geographic Data Library (http://www.fgdl.org/).
Major roads included Class 1 roads (interstates and U.S. highways) and Class 2 roads
(state roads). Minor roads included Class 3 roads (larger roads or streets in residential
areas) and Class 4 roads (smaller roads or streets in residential areas).
Statistical Analyses
For the measures of canopy cover and visual obstruction, we averaged the
estimates from each cardinal direction for each site. Using the DBH data, we calculated
the absolute tree density per hectare for conifers and hardwoods using the Cottam et al.
(1953) method and the correction factor proposed by Warde and Petranka (1981)
(Mitchell 2015). We transformed percent canopy cover, percent visual obstruction, and
the percent cover for each species in the plot with an arcsine square root transformation
(Sokal and Rohlf 2011). We classified shrub species into two categories, food and non-
food shrubs, based on species that have been found in bear diets (Maehr and Brady
1984; Dobey et al. 2005) or those with fleshy fruits likely eaten by bears (e.g., Asimina
spp. [paw paw], Diospyros virginiana [persimmon], Prunus umbellate [wild plums]).
We calculated summary statistics for each covariate and compared them
between high- and low-use sites with nonparametric Wilcoxon rank-sum tests (Conover
1999). Using generalized linear mixed effect models (GLMM; Zuur et al., 2009) with
binomial distribution, we tested for the effect of individual habitat covariates on the
probability of high-use. Because of multi-collinearity among several covariates, we could
not use the original habitat variables to test for the additive and interactive effects of
habitat variables. Thus, we used principal components analysis (PCA) to reduce data
118
dimensionality and to combat problems associated with correlated explanatory variables
(Graham 2003; Manly and Alberto 2017). We used the first four principal components
(PCs) as explanatory variables in GLMM with a binomial response (high-use = 1, low-
use = 0), a random effect of individual bear, and fixed effect of PC1 – PC4, and their
additive effects. We selected the most parsimonious model using an Information
Theoretic approach with Akaike’s information criterion (AIC; Burnham and Anderson,
2002). We used the conditional coefficient of determination (R2GLMM(c); Nakagawa and
Schielzeth, 2013) for assessing the fit of the resulting most parsimonious model. To test
for seasonal differences, we repeated the GLMMs with PCs as covariates for bear
locations obtained in summer (1 May – 31 August) and fall (1 September – 31
December) separately; data were insufficient to perform similar analysis using winter
locations. We used the R package ‘lme4’ for fitting GLMMs (Bates et al. 2015) and
R2GLMM(c) was calculated using the R package MuMIn (Barton 2015). All statistical
analyses were performed in R (Version 3.3.1; R Core Team. <http://www.r-project.org/>,
2016).
Results
We used GPS location data collected from 10 bears (5 females; 5 males) during
2011-2014 for our analysis. The average size of an individual’s high-use area (50%
DBBMM contour ± SE) spanning a timeframe of ca. 10-22 months was 5.10 ± 1.63 km2
(range: 0.93 – 16.26) and contained an average of 1976.60 ± 229.52 bear-locations
(range: 1119.00 – 3442.00). Together, high-use areas covered 43.45 km2 of our study
area (Figure 5-2). The average high-use area for females was 2.08 ± 0.71 km2 (range:
0.93 – 4.70) and contained an average of 2449.80 ± 297.78 bear-locations (range:
1839.00 – 3442.00), whereas the average high-use area for males was 8.12 ± 2.64 km2
119
(range: 2.32 – 16.26) and contained an average of 1503.40 ± 190.85 bear-locations
(range: 1119.00 – 2205.00). The average overall low-use area (99% – 75% DBBMM
contour) was 82.87 ± 22.02 km2 (range: 11.11 – 202.20) and contained an average of
604.30 ± 42.10 locations (range: 439.00 – 797.00). Together, low-use areas covered
489.33 km2 of our study area (Figure 5-2). The average low-use area for females was
29.72 ± 8.87 km2 (range: 11.11 – 62.98) and contained an average of 577.20 ± 72.03
locations (range: 439.00 – 764.00); the average low-use area for males was 136.02 ±
26.27 km2 (range: 63.22 – 202.20) and contained an average of 631.40 ± 49.18
locations (range: 519.00 – 797.00).
We conducted vegetation sampling at 213 sites; 108 of these sites were
considered high bear-use (39 in fall, 27 in summer, 42 in winter) and 105 were low bear-
use (36 in fall, 52 in summer, 17 in winter; Figure 5-2). Canopy cover, visual obstruction,
and density of hardwoods were greater in areas of high bear-use than in areas of low
bear-use (p ≤ 0.05 for all; Table 5-1; Figures 5-3 and 5-4). Additionally, high-use sites
were farther from minor roads and closer to creeks than low-use sites (p ≤ 0.05; Table
5-2) but there was no significant difference in distance to major roads between high-
and low-use sites (p ≥ 0.05; Table 5-2).
We recorded 41 hardwood species and seven conifer species, almost all of
which were pines (Appendix F). The most common conifer across all sites was Pinus
elliotti (slash pine; present in 49.77% of all sites, 44.44% of high-use sites, and 55.24%
of low-use sites) but the most common conifer in high-use sites was Pinus taeda
(loblolly pine; present in 35.21% of all sites, 49.07% of high-use sites, and 20.95% of
low-use sites); 23.00% of all sites had no pine trees within 60 m (Table F-1). The most
120
common hardwood species in high-use sites were Quercus nigra (water oak; present in
34.74% of all sites, 33.33% of high-use sites, and 36.19% of low-use sites), Gordonia
lasianthus (loblolly bay; present in 26.29% of all sites, 32.14% of high-use sites, and
20.00% of low-use sites), and Nyssa sp. (tupelo; present in 19.72% of all sites, 25.93%
of high-use sites, and 13.33% of low-use sites). A complete list of tree species found in
high-and low-use plots is presented in Table F-1.
We documented 155 (and another 15-unidentified) understory plant species (not
including the unknown number of species within the Poales group; Table F-2). The most
common understory species found across all sites were Smilax spp. (greenbrier;
present in 73.24% of all sites, 70.37% of high-use sites, and 73.24% of low-use sites),
Vitis rotundafolia (muscadine grape; present in 50.2% of all sites, 60.2% of high-use
sites, and 40.0% of low-use sites), and Serenoa repens (saw palmetto; present in
46.5% of all sites; 52.8% of high-use sites, and 40% of low-use sites; Table F-2). Of the
most commonly occurring understory species, those that were more abundant in high-
use areas were saw palmetto, Lyonia lucida (fetterbush), Ilex coriacea (sweet gallberry),
muscadine grape, Persea spp. (bays), and Vaccinium corymbosum (highbush
blueberry; Figure 5-5; Table F-2). Understory plant species that were more abundant in
low-use sites included the group of species in the order Poales and Ilex glabra (bitter
gallberry; Figure 5-5: Table F-2). Cover of food plants was greater in high-use sites than
in low-use sites (W = 7034.5, p = 0.002; Figure 5-6; Table 5-1), whereas cover of non-
food plants did not differ between high- and low-use sites (W = 6376, p = 0.12; Figure 5-
6; Table 5-1).
121
Examining the effect of the individual habitat variables, the probability of high-use
was positively influenced by canopy cover, visual obstruction, food shrub cover, and
minor roads; it was negatively influenced by distance to creek. Other variables did not
significantly influence the probability of high-use black bears (Table 5-3).
Several of the habitat covariates were significantly correlated with each other
(Table 5-4); thus, we used principal component analysis to reduce the data
dimensionality and address the problem of collinearity. Principal components (PC)
analysis (Table 5-5) indicated that: (1) PC1 was positively correlated with canopy cover,
visual obstruction, shrub cover (both food and non-food plants), hardwood density, and
negatively correlated with pine density and distance to creeks; (2) PC2 was positively
correlated with visual obstruction, pine density, cover of food and non-food plants, and
distance to creeks and was negatively correlated with canopy cover and hardwood
density; (3) PC3 was positively correlated with visual obstruction and was negatively
correlated with canopy cover, hardwood density, and pine density; (4) PC4 was
positively correlated with canopy cover and pine density, and was negatively correlated
with visual obstruction and hardwood density. PC1-PC4 cumulatively explained 70% of
variance in the data (Table 5-5).
The GLMMs using one PC at a time as a predictor variable indicated that the
probability of high-use of an area by bears was positively affected by PC1, PC2, and
PC4, and negatively affected by PC3 (Tables 5-6 and 5-7). The most parsimonious
GLMM testing for the singular and additive effects of PCs included an additive effect of
PC1 and PC4 and had an R2GLMM(c) of 0.226. Based on this model, both PC1 and PC4
positively affected the probability of high-use. The second ranked model included an
122
additive effect of PC1, PC3, and PC4. Based on coefficients from this model, the effect
of PC1 and PC4 remained the same but PC3 negatively influenced the probability of
high-use (Tables 5-6 and 5-7). Repeating the GLMM models seasonally, we found that
the most parsimonious model for summer included an additive effect of PC1 and PC3,
whereas that for fall included an additive effect of PC1 and PC4 (Table 5-6). The pattern
of influence of the PCs was similar to that reported for the overall analyses.
Discussion
Several studies have reported that American black bears use habitat within their
home range heterogeneously (Sadeghpour and Ginnett 2011; Sollmann et al. 2016).
Bears are expected to forage optimally for resources and to use some patches within
their home ranges more than others because resources are temporally and spatially
isolated across the landscape (Mitchell and Powell 2012). However, why bears use
some areas within their home ranges more than others of the same habitat type
remains unclear. The structural and compositional aspects of the habitat driving these
patterns may help identify limiting resources for bears. We expected that high-use sites
would have more canopy cover, more visual obstruction, and a higher density of
hardwood trees because of the importance of escape cover, saw palmetto across all
seasons, and mast-producing tree species for bears. Our results supported these
expectations. Consistently, high-use sites had greater than 90% closed canopies and
greater than 80% visual obstruction; both much greater than in low-use sites. Cover is
an important component of bear habitat (Brady and Maehr 1985; Wooding and Hardisky
1994; Stratman et al. 2001; van Why 2003; Mitchell and Powell 2012) and is likely
important to reduce disturbance from other bears or humans (Mitchell and Powell 2012)
123
and possibly to aid in thermoregulation (Stratman et al. 2001; Fecske et al. 2002; Lyons
et al. 2003); although our data cannot separate habitat use for foraging from resting.
High-use sites were closer to creeks than low-use sites; this result was
consistent with our expectation because bears move shorter distances when closer to
creeks (Karelus et al. 2016), and often select for forested wetland habitats within their
home ranges (Weaver and Pelton 1994; Stratman et al. 2001; Dobey et al. 2005;
Takahata et al. 2014; Lewis et al. 2015; Duquette et al. 2017). Hardwood densities were
greater in high-use sites than in low-use sites, indicating that the bears preferentially
used forested habitats over other land-cover types. Additionally, several of the
hardwood species that we recorded, such as oaks and tupelos, likely provide the bears
with important sources of food from soft or hard mast. Tupelo drupes (more specifically,
black gum, N. sylvatica) were important early fall foods for bears in Osceola and
Okeefenokee National Forests (Dobey et al. 2005), just north of the Camp Blanding
area, and for bears in other populations as well (Hellgren and Vaughan 1988). In our
study area, tupelos were more common at high-use sites than in low-use sites, so it is
likely that bears in this area also rely on the tupelo drupes for sustenance. Acorns are
an important food source for bears across their range (Dobey et al. 2005; Benson and
Chamberlain 2006; Reynolds-Hogland et al. 2007), but density of oaks did not differ
substantially between high- and low-use sites. Whereas tupelos occur in forested
wetlands, oaks occur in several different habitat types, including those with more open
canopies and less visual obstruction, such as scrub. Several other hard- and soft-
masting hardwood species occurred in our study area (e.g., Carya spp. [hickory
species], Cornus spp. [dogwood species], persimmons, plums). Although bears eat
124
fruits and nuts when available (Hellgren 1993; Stratman and Pelton 1999; Dobey et al.
2005; Inman and Pelton 2011; Simek et al. 2012), these trees occurred infrequently in
our study site. Densities of conifers, mostly loblolly pines, were marginally greater in
high-use sites. Loblolly pines occur in a variety of habitats, including forested wetlands.
However, a large portion of our study area is managed as slash pine plantation and
thus, overall, slash pine was the most common species. Depending on how they are
managed, stands of pine plantation typically provide only scattered food sources, but
they do provide ample cover.
Several understory plant species that likely provide bears with food occurred
more frequently in high-use sites, including saw palmetto, sweet gallberry, muscadine
grape vine, and high bush blueberry. Bears in Florida rely heavily on saw palmetto for
food throughout the year; they eat the fruits in the fall and winter and the hearts of the
palm in the spring (Maehr and Brady 1984; Stratman and Pelton 1999; Dobey et al.
2005). Indeed, we often found evidence of bears eating the hearts of the palmettos
throughout our study site (D. Karelus, personal observation). Although saw palmettos
occurred frequently in both high- and low-use sites in our study area, they were more
common in high-use sites. Sweet gallberry, blueberries, and [to a lesser extent]
muscadine grapes, were important foods for bears in the Okefenokee-Osceola system
(Dobey et al. 2005) and at Eglin Air Force Base in the Florida panhandle (Stratman and
Pelton 1999). Given the frequency of sweet gallberries, high bush blueberries, and
muscadine grapes in our high-use sites, they are likely important for bears in the Camp
Blanding area as well. Throughout our study site, we also found several other species of
berry-producing shrubs, including several other species of blueberry (V. arboreum, V.
125
myrsinites, V. stamineum), huckleberries (Gaylussacia dumosa and G. frondosa), and
blackberries (Rubus sp.) and also berry-producing vines of greenbrier species. These
also species may have provided important food sources for bears (Maehr and Brady
1984; Stratman and Pelton 1999; Dobey et al. 2005; Belant et al. 2006; Mosnier et al.
2008; Baldwin and Bender 2009; Hertel et al. 2016), but they were generally equally as
common in both high- and low-use sites. We expected that the amount of shrub cover
by potential food plants would be important for characterizing high-use sites (e.g.,
Mitchell and Powell, 2012). Indeed, we found that high-use sites had a greater amount
of shrub cover by food-producing plants but that cover of non-food producing plants did
not differ between high- and low-use sites.
Results of GLMMs testing for the effect of each habitat variable indicated that
high-use sites were generally characterized by high canopy cover, high visual
obstruction, high cover of food-producing shrubs, and closer to the creeks but farther
away from roads. These results are generally consistent with those obtained from
univariate analyses, and thus reinforce the importance of these habitat features.
Testing for the effect of each microhabitat feature on the probability of high-use
provides useful insight into which microhabitat features are important for bears;
however, all microhabitat variables must be considered together to fully understand the
bears’ heterogeneous use of the landscape within their home ranges. When there are
many explanatory variables, multicollinearity can cause computational problems and
yield results that may be difficult to interpret (Graham 2003). To avoid these issues, we
performed PCA to investigate how the suite of microhabitat features jointly influenced
the probability of high-use by bears. Based on the PC loadings, the first PC identified
126
sites with high canopy cover and high visual obstruction, dense hardwoods, proximity to
creeks; these are characteristics of forested wetlands. The second PC identified sites
with low canopy cover, high visual obstruction, high pine density, low hardwood density,
and greater distance from creeks, typified by pine flatwoods. The third PC characterized
sites similar to those in PC2, but had an even more open canopy, lower visual
obstruction, and lower pine density; this characterizes prairies and open fields. The
fourth PC characterized sites with high canopy cover and high pine density, and low
shrub cover; this is indicative pine plantation and sand pine forest. Together PC1 – PC4
explained 70 % of the variation in the microhabitat data.
Using GLMMs with PC1 – PC4 as explanatory variables, we modeled the
probability of high-use by bears. Considered individually, PC1 and PC4 positively
influenced the probability of high-use; the effect of PC2 and PC3 was negligible, but
generally negative. The most parsimonious GLMM included PC1 and PC4, both of
which positively affected high-use; indicating that, within their home ranges, bears
preferentially use forested wetlands, targeting sites with high understory cover and
abundance of potential food sources. The fact that PC1 was included in the top model
for overall as well as seasonal analyses highlight the fact that black bears in our study
site intensively use forested wetlands during all seasons.
Our findings regarding forested wetlands are consistent with conclusions from
black bear habitat selection studies at our study site and elsewhere in the southeastern
United States (Stratman et al. 2001; Dobey et al. 2005; Karelus et al. 2016). A previous
study showed that bears in this area did not select for pine plantations at the second- or
third-order scales (Karelus et al. 2016). Bears in Ocala National Forest, located south of
127
our study site, selected for sand pine forests in the summer (Moyer et al. 2008). Sand
pine forest and pine plantations may afford bears with escape cover; however, we note
that all high bear-use sites that were in pine plantations were close to edges of the
plantation adjacent to forested wetlands. Therefore, bears’ use of pine plantations may
potentially reflect the fact that they are the largest land cover type in our study site and
are generally juxtaposed to forested wetlands (Karelus et al. 2016). Our results show
bears in our study site show strong preference for forested wetlands with high canopy
and shrub cover and with an abundance of food-producing understory plant species;
they use open habitats (e.g., savannas) rather infrequently.
Animals are likely to use different habitats depending on the activity in which they
are engaging (Lima and Zollner 1996; Morales and Ellner 2002; Nathan et al. 2008;
Ordiz et al. 2011; Abrahms et al. 2016). Bears may use low-use areas primarily for
traveling and high-use areas for foraging or resting. We tested this idea by comparing
step-lengths of bears while in low- vs. high-use areas. The mean (±SE) bi-hourly step
length while in the high-use sites was 142 ± 34 m, compared to step length of 659 ± 113
m, while in the low-use sites. High-use sites were generally associated with shorter-
distance movements and the low-use sites were generally associated with longer-
distance movements (Karelus et al. 2017). Therefore, we would expect that high-use
sites were generally indicative of resting or foraging sites, whereas the low-use sites
were generally indicative of sites obtained while the bear was traveling. Thus, our
results can also be considered in terms of the structural and vegetative features that
influenced bears to use a site for resting or foraging versus traveling.
128
Conclusions and Management Recommendations
Demographic parameters of large carnivore populations often reflect habitat
quality (Beschta and Ripple 2012). It is, therefore, important to identify key habitat
components such that management practices can be directed towards promoting the
components of habitat that are critical to the target species. Our study showed that the
best predictors of high-use of an area by black bears are proximity to creeks, high
canopy and escape cover, high densities of hardwoods, and high abundance of
potential food plants. Based on these results, we suggest that managers may want to
incorporate this description of bear habitat in choosing recipient sites for reintroductions
of bears to depauperate areas or for translocations done to address negative human-
bear interactions. Conservation planners should conserve forested wetlands, and
enhance the diversity and abundance of hard- and soft-mast producing trees, and berry
producing shrubs and vines across all habitat types (Hellgren et al. 1991; Hellgren and
Vaughan 1994). When managing forests for black bears in Florida, managers should be
particularly mindful to take actions to promote growth and berry production of saw
palmettos, especially in or near forested wetlands (Maehr et al. 2001). Additionally,
conservation efforts should focus on preserving the linkages among forested wetlands
afforded by creeks and riparian forests. Creeks and riparian forests generally have high
conservation value in fragmented landscapes because they offer critical resources for
many wildlife species, and provide connectivity across the landscape. (Naiman et al.
1993; Taulman and Smith 2004; Keuroghlian and Eaton 2008; Lees and Peres 2008;
Kniowski and Gehrt 2014).
Supporting healthy populations of large carnivores in fragmented, human-
dominated landscape is challenging and will require cooperation among multiple
129
agencies and private land holders. Nonetheless, managing forests for black bears is a
worthy conservation goal, given their status as the largest terrestrial carnivore of Florida
and their role as an umbrella species for biodiversity conservation (Simberloff, 1999;
Karelus et al. 2017).
130
Table 5-1. Averages (± SE) of vegetation measures from high and low bear-use sites in the Camp Blanding area. N represents the number of sampled locations. Canopy cover is the percent canopy cover as measured with a spherical densitometer and averaged for each site among the four cardinal directions. Visual obstruction represents the percent of a 2-m tall coverboard obstructed by vegetation and averaged at each site among the four cardinal directions. Absolute density of pines and hardwoods represents the average density of trees as measured by the Point-Center Quartered Method. Wilcoxon Rank Sum tests were used to test for differences between high- and low- sites; W indicates the rank-sum and p < 0.05 was considered significant.
Averages ± SE
N Canopy
cover
Visual
Obstruction
Absolute Density
Pines (trees/hectare)
Absolute Density
Hardwoods (trees/hectare)
Shrub cover
food species
Shrub cover
non-foods
All
High 108 0.95 ± 0.01 0.81 ± 0.02 355.72 ± 91.49 1023.75 ± 91.83 107.86 ± 5.24 56.82 ± 3.76
Low 105 0.79 ± 0.03 0.66 ± 0.03 346.59 ± 53.90 872.09 ± 154.11 82.17 ± 4.73 58.17 ± 4.10
W 7951 7434 4804 6962 7034.5 6376
p < 0.0001 < 0.0001 0.05 0.004 0.002 0.12
Female
High 53 0.94 ± 0.02 0.83 ± 0.02 180.59 ± 59.70 1128.71 ± 134.89 112.94 ± 7.62 59.58 ± 4.90
Low 50 0.84 ± 0.02 0.67 ± 0.04 334.40 ± 64.87 1132.72 ± 301.86 84.34 ± 6.67 65.20 ± 5.72
Male
High 55 0.96 ± 0.01 0.80 ± 0.03 524.48 ± 167.89 922.61 ± 124.65 102.96 ± 7.22 54.16 ± 5.69
Low 55 0.74 ± 0.05 0.65 ± 0.04 357.68 ± 84.89 635.15 ± 99.85 80.20 ± 6.74 51.78 ± 5.76
131
Table 5-2. Average distances (± SE) from the sampled bear locations to major roads, minor roads, and creeks for high and low bear-use sites in the Camp Blanding area. N represents the number of locations sampled. Wilcoxon Rank Sum tests were used to test for differences between high- and low- sites for the overall measures; W indicates the-rank-sum and p < 0.05 was considered significant.
Average distances ± SE (m)
N Major roads Minor roads Creeks
All High 108 2481.68 ± 154.53 306.93 ± 16.84 222.56 ± 22.17
Low 105 2327.16 ± 141.42 232.87 ± 18.89 346.47 ± 29.90
W 6128 7253 4143
p 0.31 < 0.001 < 0.001
Female High 53 2036.87 ± 138.04 326.80 ± 21.35 150.35 ± 18.75
Low 50 2245.70 ± 199.99 216.39 ± 22.02 307.70 ± 33.73
Male High 55 2910.31 ± 230.64 287.77 ± 25.83 292.15 ± 37.45
Low 55 2401.21 ± 233.65 247.86 ± 30.05 381.72 ± 47.97
132
Table 5-3. Results of generalized linear mixed models testing for the effect of individual habitat covariates on the probability of high-use by black bears (Ursus americanus floridanus) in north-central Florida. Regression coefficients testing for the effect of each habitat covariate on the probability of high-use are presented along with the associated significant level (P).
Variable Coefficient ± SE P
Canopy cover 3.37 ± 0.68 <0.001
Visual obstruction 1.73 ± 0.44 <0.001
Creek -0.002 ± 0.001 0.003
Food shrub cover 0.490 ± 0.16 0.002
Minor roads 0.002 ± 0.001 0.005
Non-food shrub cover 0.29 ± 0.18 0.09
Hardwood density 0.96 ± 1.15 0.405
Major roads 0.001 ± 0.001 0.459
Pine density 0.15 ± 1.77 0.932
133
Table 5-4. Pairwise correlation between habitat variables. Significance level (p values) are given in parenthesis. Bold type face indicates significant correlations. Bold type faces indicate significant correlation.
Canopy
cover
Visual
obstruction
Hardwood
density Pine density
Food shrub
cover
Non-food
shrub cover
Major
roads
Minor
roads Creeks
Canopy cover
Visual
obstruction
0.29
(< 0.001)
Hardwood
density
0.34
(< 0.001)
0.17
(0.0122)
Pine density
-0.01
(0.92)
-0.08
(0.255)
-0.13
(0.051)
Food shrub
cover
0.36
(< 0.001)
0.47
(< 0.001)
0.19
(0.006)
-0.11
(0.098)
Non-food
shrub cover
0.16
(0.021) 0.14 (0.035)
0.08
(0.224)
-0.08
(0.257)
0.17
(0.013)
Major roads
0.03
(0.62) 0.09 (0.196)
-0.03
(0.646)
-0.09
(0.188)
0.06
(0.405)
0.12
(0.089)
Minor roads
0.29
(< 0.001)
0.28
(< 0.001)
0.11
(0.107)
-0.11
(0.106)
0.13
(0.052)
0.16
(0.023)
0.16
(0.020)
Creeks
-0.35
(< 0.001)
-0.30
(< 0.001)
-0.17
(0.011)
0.25
(<0.001)
-0.19
(0.004)
-0.16
(0.022)
-0.15
(0.026)
-0.33
(< 0.001)
134
Table 5-5. Principal component (PC) loadings from microhabitat variables measured at 213 Florida black bear locations in the Camp Blanding area. Only the first 4 principal components are shown here from 9 total PCs. Positive values indicate a positive loading of the variable on the PC and negative values indicate a negative loading on the PC. The amount of variance in the data explained by each PC is given as the Proportion of variance. The combined total of variance explained by the PCs in order is given by the Cumulative proportion.
Variables PC1 PC2 PC3 PC4
Canopy cover 0.34 -0.21 -0.42 0.36
Visual obstruction 0.48 0.41 0.11 -0.10
Hardwood density 0.23 -0.28 -0.55 -0.13
Pine density -0.15 0.47 -0.21 0.75
Distance to major roads 0.12 -0.22 0.60 0.27
Distance to minor roads 0.30 -0.28 0.19 0.42
Distance to creeks -0.34 0.41 -0.10 -0.05
Cover of food shrubs 0.37 0.21 -0.17 -0.15
Cover of other shrubs 0.47 0.39 0.17 -0.11
Proportion of variance 0.33 0.13 0.13 0.11
Cumulative proportion 0.33 0.47 0.59 0.70
135
Table 5-6. Model selection from generalized linear mixed models testing for the effect of principal components (PC) 1 through 4 on the probability of high-use by black bears (Ursus americanus floridanus) in north-central Florida for the A) the entire study period, B) and results using bear locations obtained during summer, and C) results using bear locations obtained in the fall. PC loadings are provided in Table 5. Models appear in order of the difference in the Akaike Information Criterion corrected for small sample sizes (ΔAICc). The difference in the log-likelihood from the top model (ΔLL), model probability (Weight), and the number of parameters (K) are also given.
Rank Model K ΔLL ΔAICc Weight
A) Overall 1 PC1 + PC4 4 0.00 0 0.43
2 PC1 + PC3 + PC4 5 0.46 1.17 0.24
3 PC1 + PC2 + PC4 5 0 2.09 0.15
4 PC1 + PC2 + PC3 + PC4 6 0.47 3.29 0.08
5 PC1 3 -3.39 4.7 0.04
6 PC1 + PC3 4 -2.86 5.72 0.02
7 PC1 + PC2 4 -3.39 6.77 0.01
8 PC1 + PC2 + PC3 5 -2.86 7.82 0.01
9 PC4 3 -15.44 28.8 0.00
10 PC3 + PC4 4 -14.99 29.97 0.00
11 PC2 + PC4 4 -15.41 30.82 0.00
12 PC2 + PC3 + PC4 5 -14.94 31.97 0.00
13 PC3 3 -17.66 33.25 0.00
14 PC2 3 -18.04 34.01 0.00
15 PC2 + PC3 4 -17.66 35.32 0.00
136
Table 5-6. Continued Rank Model K ΔLL ΔAICc Weight
B) Summer 1 PC1 + PC3 4 0.00 0.00 0.24
2 PC1 3 -1.15 0.08 0.23
3 PC1 + PC2 + PC3 5 0.21 1.86 0.09
4 PC1 + PC4 4 -1.01 2.02 0.09
5 PC1 + PC3 + PC4 5 0.04 2.20 0.08
6 PC1 + PC2 4 -1.11 2.22 0.08
7 PC3 3 -2.75 3.28 0.05
8 PC1 + PC2 + PC3 + PC4 6 0.29 4.05 0.03
9 PC1 + PC2 + PC4 5 -0.92 4.12 0.03
10 PC4 3 -3.74 5.25 0.02
11 PC2 + PC3 4 -2.63 5.26 0.02
12 PC2 + PC3 3 -3.79 5.36 0.02
13 PC3 + PC4 4 -2.74 5.49 0.02
14 PC12+ PC4 4 -3.71 7.42 0.01
15 PC2 + PC3 + PC4 5 -2.63 7.54 0.01
137
Table 5-6. Continued Rank Model K ΔLL ΔAICc Weight
C) Fall 1 PC1 + PC4 4 0.00 0.00 0.42
2 PC1 + PC3 + PC4 5 0.38 1.54 0.20
3 PC1 + PC2 + PC4 5 0.21 1.88 0.17
4 PC1 + PC2 + PC3 + PC4 6 0.77 3.11 0.09
5 PC1 3 -3.46 4.68 0.04
6 PC4 3 -4.24 6.24 0.02
7 PC2 + PC4 4 -3.27 6.55 0.02
8 PC1 + PC2 4 -3.30 6.61 0.02
9 PC1 + PC3 4 -3.44 6.88 0.01
10 PC2 + PC3 + PC4 5 -2.96 8.23 0.01
11 PC3 + PC4 4 -4.16 8.33 0.01
12 PC1 + PC2 + PC3 5 -3.28 8.86 0.01
13 PC3 3 -8.47 14.71 0.00
14 PC2 3 -8.52 14.81 0.00
15 PC2 + PC3 4 -8.47 16.94 0.00
138
Table 5-7. Estimates (± SE) of slope parameters, as well as 95% confidence intervals, for the fixed effects of the principal component loadings that were included in select generalized linear mixed models as indicated by their rank (Table 4-6).
Rank Model PC1 PC2 PC3 PC4
1 PC1 + PC4 0.52 ± 0.10
(0.31, 0.72) - -
0.41 ± 0.17
(0.09, 0.74)
2 PC1 + PC3 + PC4 0.52 ± 0.10
(0.31, 0.72) -
-0.14 ± 0.15
(-0.44, 0.15)
0.41 ± 0.17
(0.08, 0.75)
3 PC1 + PC2 + PC4 0.52 ± 0.10
(0.31, 0.72)
-0.01 ± 0.14
(-0.28, 0.26) -
0.41 ± 0.17
(0.08, 0.74)
4 PC1 + PC2 + PC3 +
PC4
0.52 ± 0.11
(0.31, 0.73)
-0.01 ± 0.14
(-0.28, 0.27)
-0.14 ± 0.15
(-0.44, 0.15)
0.41 ± 0.17
(0.08, 0.74)
5 PC1 0.48 ± 0.10
(0.29, 0.67) - -
9 PC4 - - - 0.35 ± 0.16
(0.03, 0.66)
13 PC3 - - -0.11 ± 0.13
(-0.37, 0.14) -
14 PC2 - 0.01 ± 0.12 (-
0.23, 0.26) - -
139
Figure 5-1. Map of the dynamic Brownian Bridge movement model utilization distribution
for a representative female Florida black bear (Ursus americanus floridanus) at different levels of utilization, indicated by percent contours. Bear GPS locations within the 50% contours were designated as high-use and the locations in the area between the 75% and 99% contours were designated as low-use.
140
Figure 5-2. Locations of vegetation sampling in high- and low-use areas for Florida black bears (5F, 5M) at Camp Blanding Joint Training Center and the surrounding private and state forest lands. White lines indicate major roads. The dots and triangles represent the sampled locations, and the shaded areas represent the combined contours of the bears’ utilization distributions estimated using dynamic Brownian Bridge movement models.
141
Figure 5-3. Average (± SE) A) percent canopy cover and B) visual obstruction from sites of high and low black bear-use in the Camp Blanding area. A value of 1 represents 100% cover and a value of 0% indicates no cover.
142
Figure 5-4. The absolute density (± SE) of hardwood and pine trees in high- and low-bear-use sites in the Camp Blanding area. Density was calculated from measures of diameter at breast height of trees collected using the Point Center Quartered Method.
143
Figure 5-5. The average (± SE) percent cover by 15 of the most common species found in the shrub layer in 4 x 4m plots in high- and low-use sites used by bears in the Camp Blanding area. If a species was not present in the shrub layer in a plot, we inserted a value of 0 for the cover within the plot for that species; therefore, the averages were calculating including a large number of zeros and do not represent a measure of the actual amount of cover found when a species was present. A list of all species, total percent cover, percent cover when present, and percent of plots where they were present is provided in Table F-2.
144
Figure 5-6. The average percent cover (± SE) of A) food producing shrubs and B) non-
food producing shrubs within 4 x 4 m plots at high- and low-use bear sites in the Camp Blanding area of north central Florida. Values amounted to more than 100% because cover of individual species was estimated within the plot and these were then added together when concatenating data into food-producing shrubs and non-food producing shrubs.
145
CHAPTER 6 CONCLUSIONS AND MANAGEMENT IMPLICATIONS
The bears in the Camp Blanding area provided us with a unique opportunity to
investigate the space-use, resource-use, and movement patterns of animals in a newly
colonized population of large carnivores in a fragmented landscape. Our study
population is also of management interest because it likely acts as a stepping-stone
population (Baguette et al. 2013; Saura et al. 2014) within the proposed Ocala-Osceola
wildlife corridor and thereby may help maintain or increase genetic and demographic
connectivity between the bear populations in Ocala and Osceola National Forests.
We found that black bears in our study area used generally larger home ranges
than bears inhabiting Ocala National Forest, just south of our study site, suggesting that
resources may be scarcer or more spread out in the Camp Blanding area. The bears
exhibited strong selection for forested wetlands and areas near creeks. They also
moved shorter distances in forested wetlands and were more likely to move to these
habitats than to other habitats, regardless of season or behavioral state, indicating that
riparian forests are critical habitat for the bears in this area. These results reinforced the
general importance of riparian forests for black bears in the southeast (Hellgren et al.
1991; Wooding and Hardisky 1994; Dobey et al. 2005). We also found that major roads
acted as a semi-permeable barrier to the bears’ movements and bears were selecting
least for, and traveling fastest through, urban areas. Additionally, males tended to
exhibit more nocturnal movements and females exhibited more crepuscular
movements. Areas that received high levels of bear use had dense canopy cover, thick
understory vegetation, and higher densities of trees than in other portions of their home
ranges; these features likely provide bears with escape cover and protection from
146
thermal stress. Saw palmetto and sweet gallberry were more common in areas with
high bear-use as well, which are both likely important food sources for the bears.
Management plans aimed at conservation of black bears in human-dominated
landscapes should focus on preserving and restoring forested wetlands and riparian
forests because they provide bears with foraging and denning habitats (Hellgren and
Vaughan 1989; Wooding and Hardisky 1994; Stratman et al. 2001; Dobey et al. 2005).
Also, maintaining or increasing the distribution and abundance of soft- and hard-mast-
producing plants within forested wetlands and adjacent uplands will help ensure the
availability of essential resources for bears. Managing for a variety of food sources
within these areas may help reduce bear road mortalities or human-bear conflicts
(Moyer et al. 2007; Ryan et al. 2007; Baruch-Mordo et al. 2014; Johnson et al. 2015).
Additionally, conservation planners should consider mitigating the impacts of future road
development on forested wetlands as a priority for bear conservation and for promoting
genetic connectivity (Dixon et al. 2007; McCown et al. 2009).
Management of black bears and other large carnivores inhabiting fragmented
landscapes also requires a landscape-based approach, including not only the primary
habitat in forested wetlands, but also the linkages among those areas (Hilty and
Merenlender 2004; Chetkiewicz et al. 2006; Beier et al. 2008; Lacher and Wilkerson
2014). The linkages allow for connectivity thru the human-dominated matrix (Noss et al.
1996; Beier and Noss 1998; Clark et al. 2015). Additionally, landscape linkages help
afford access to other habitats which may provide important seasonal resources; with
appropriate access to natural foods, large carnivores, such as bears, may then be less
likely to come into conflict with humans (Baruch-Mordo et al. 2014; Lewis et al. 2015).
147
Because of the relatively large areas required by large carnivores, sustaining healthy
populations of these animals in fragmented, human-dominated landscapes also
requires cooperation among multiple agencies and private land holders (Hoctor et al.
2000; Larkin et al. 2004). These management actions would help increase the odds of
colonization and persistence of stepping-stone populations, and would facilitate greater
connectivity among bear populations (Florida Fish and Wildlife Conservation
Commission 2012; Baguette et al. 2013; Saura et al. 2014).
148
APPENDIX A DATA PREPARATION AND LAND COVER MAP
We removed all but the highest quality fixes obtained with 4 or more satellites
(3D validated) from the GPS location data to ensure that we used only accurate bear
locations. We manually removed a small number of 3D-validated locations from 2
collars that were clearly incorrect based on geography (i.e., those in the Atlantic Ocean)
and biologically unreasonable distances between successive locations (i.e., more than
60 km from both the previous and the subsequent location within 30 minutes).
We also removed duplicate fixes that occurred in rapid succession (within
minutes) after false mortality signals. The collars were programmed to collect bursts of
locations to force an out-of-schedule submission of data. We rounded the time of each
remaining location to the nearest hour but used only bihourly locations for the analysis
because half-hour fixes were obtained only during dawn and dusk.
Florida Fish and Wildlife Conservation Commission personnel performed a test to
determine the location error for the GPS collars. The collars were found to be accurate
to a 20.3-m radius for 95% of the locations. Collars continued to transmit after dropping
off of the Black Bear, creating a cluster of locations that made the last true bear location
difficult to discern. To exclude bias by analyzing locations from a dropped collar and
ensure that all data represented true bear locations, we used ArcMap to create a 20.3 m
buffer around the last location of all Black Bear collars that had final locations in a tight
cluster. Then we removed each sequentially preceding location inside the buffer until
more than 5% of the locations in the final cluster were outside the buffer. The first
location during that time period inside the buffer and those inside the buffer from a
previous time period remained in the data set.
149
We used the raster format of the Florida Vegetation and Land Cover 2014 GIS
layer for land cover classification (Redner and Srinivasan 2014). The layer covered the
entire state at a resolution of 10 m. We used ArcMap (version 10.3; ESRI 2015) to clip
the statewide layer to our study area, the 99% MCP of all Black Bear locations. The
resulting layer contained 51 land cover categories, which we then grouped into six land
cover categories based on similarity of landscape and vegetation and combining
minimally available categories using the R package raster (Hijmans 2015). The
remaining six land cover categories are:
• Forested wetlands – freshwater forested wetlands, cypress, isolated freshwater swamp, floodplain swamp, other coniferous wetlands, wet flatwoods, baygall, cultural-palustrine, dome swamp, and basin swamp.
• Marsh/Wetland – coastal uplands, freshwater non-forested wetlands, prairies and bogs, marshes, isolated freshwater marsh, non-vegetated wetland, lacustrine, natural lakes and ponds, cultural - lacustrine, riverine, natural rivers and streams, and cultural – riverine.
• Rural/Agricultural - barren and outcrop communities, rural, cropland/pasture, orchards/groves, vineyards and nurseries, other agriculture, and improved pasture.
• Tree plantations - not combined with any other category.
• Urban – cultural and terrestrial, low intensity urban, high intensity urban, and transportation, communication, extractive.
• Woods/Scrub – upland hardwood forest, mesic hammock, xeric hammock, high pine and scrub, scrub, sand pine scrub, sandhill, pine flatwoods and dry prairie, dry flatwoods, dry prairie, palmetto prairie, shrub and brushland, and utilities.
150
APPENDIX B HOME RANGE SIZES AMONG STUDIES
Table B-1. Overall and annual average home range sizes for A) Female and B) Male Florida Black Bears estimated using 95% Kernel Density Estimates (KDE) in the Camp Blanding study site and nearby locations. The number of bears in the estimate are represented by n. A dash (-) indicates missing information.
Study Location Years n Mean number of
locations per bear ± SE
95% KDE ± SE
(km2)
A. Females
Current
study
Camp Blanding,
FL
2011 – 2013 6 4094.50 ± 397.89 31.16 ± 8.23
2011 2 2915.00 ± 314.00 30.68 ± 11.47
2012 6 2682.50 ± 180.08 28.06 ± 7.22
Moyer
et al.
2007
Ocala National
Forest, FL 2000 – 2003 24 49.38 ± 2.88 25.89 ± 4.44
Lynne, FL 2000 – 2003 7 60.64 ± 5.59 18.54 ± 3.86
Lynne and Ocala 2000 14 35.57 ± 1.51 42.58 ± 9.96
2001 11 39.09 ± 1.69 22.54 ± 3.04
2002 15 62.13 ± 2.10 15.52 ± 2.90
2003 8 79.25 ± 3.75 10.62 ± 1.76
Dobey
et al.
2005
Osceola National
Forest, FL 1996 – 1999 53 - 30.3 ± 4.0
1996 5 - 16.5 ± 2.5
1997 9 - 21.8 ± 3.1
1998 22 - 33.9 ± 7.4
1999 17 - 34.3 ± 7.7
151
Table B-1. Continued
Study Location Years n
Mean number of
locations per bear ±
SE
95% KDE ± SE
(km2)
Dobey
et al.
2005
Okefenokee
National Wildlife
Refuge, GA
1996 - 1999 69 - 55.9 ± 6.9
1996 18 - 51.8 ± 14.0
1997 16 - 51.8 ± 11.8
1998 17 - 46.7 ± 10.3
1999 18 - 72.2 ± 17.8
B. Males
Current
study
Camp Blanding,
FL 2011 - 2013 10 2235.50 ± 370.10 249.91 ± 18.77
2011/12 4 26424.00 ± 211.71 213.44 ± 46.90
2012/13 3 2301.33 ± 369.52 191.04 ± 52.31
McCown
et al.
2004
Ocala National
Forest, FL
1999-2001 7 - 94.3A
Dobey
et al.
2005
Okefenokee
National Wildlife
Refuge, GA
1996 - 1999 10 - 342.8 ± 71.5
1996 1 - 208.3 ± 0.00
1997 4 - 294.0 ± 101.0
1998 3 - 422.4 ± 210.9
1999 2 - 388.0 ± 99.6
AStandard Error was not reported
152
APPENDIX C FRAGMENTATION ANALYSIS
We quantified the degree of habitat fragmentation within our study site in Florida
at the Camp Blanding Joint Training Center and the surrounding public and private
lands. We delineated our study site (hereafter, Camp Blanding area) using the 99%
minimum convex polygon estimated from the GPS locations of all the Black Bears in our
study. We also quantified fragmentation within the outer boundaries of nearby Ocala
National Forest and nearby Osceola National Forest for comparison with our site
because each supports a major population of Florida Black Bears. We grouped the land
cover types in each area into 2 categories: suitable Black Bear habitat (marsh/wetland,
wood/scrub, tree plantations, and forested wetlands) and unsuitable Black Bear habitat
(rural/agricultural and urban). We defined open water as background to exclude those
areas from analysis. We then calculated contagion, percent land cover, and patch
density for each of the 3 areas, following Hostetler et al. (2009). We performed these
calculations in the program Fragstats (version 4.2.1; McGarigal et al., 2015).
Out of 2,280 suitable habitat patches within the 99% MCP of all Black Bear home
ranges in the Camp Blanding area, 81% were ≤ 0.01 km2, whereas only 52% of suitable
patches in Ocala were ≤0.01 km2 (N = 792) and 43% of suitable patches in Osceola
were ≤0.01 km2 (N = 92). The Camp Blanding area exhibited a lower proportion of
suitable habitat, with a more strongly dispersed, less strongly aggregated, patchy
distribution, and smaller average patch sizes than in Ocala National Forest or Osceola
National Forest (Table C-1). This indicates that the Camp Blanding area is more
fragmented than the other 2 areas.
153
Table C-1. Quantification of habitat fragmentation in our study site at Camp Blanding and the surrounding areas (delineated by the 99% minimum convex polygon from all bear locations in our study), compared to that at Ocala National Forest and Osceola National Forest. Contagion (CONTAG) was calculated for the entire landscape and represents the aggregation and interspersion of habitat in each area. The remaining Fragstats values describe only suitable Black Bear habitat. Percent land cover (PLAND) is the sum of the suitable habitat area divided by the total landscape area. Patch density (PD) is the number of suitable habitat patches divided by the landscape area (a higher value indicates more patchiness). The mean patch area (AREA_MN) is the sum of the suitable habitat area divided by the number of all patches. All metrics were calculated in Fragstats.
Study site
Total
landscape
area (km2)
CONTAG
Suitable habitat
PLAND PD AREA_MN
Camp
Blanding and
surrounding
areas
1,525.8 54.8 76.84 1.49 51.51
Ocala National
Forest 1,774.4 76.6 90.2 0.85 105.59
Osceola
National
Forest
654.08 83.8 94.9 0.33 287.5
154
APPENDIX D TABLES OF MOVEMENT METRIC AVERAGES AND MODEL SELECTION TABLES
Averages for each movement metric and the model selection tables for
movement metrics at daily and weekly scales for Florida black bears (Ursus americanus
floridanus) in north-central Florida using GPS data collected from 2011 to 2014.
155
Table D-1. Average movement metrics for female and male Florida black bears at multiple temporal scales; A) over the entire study period, B) daily, C) weekly, and D) monthly. The average step-length, E(l), is the average distance between 2 successive locations. The average directional persistence is denoted by E(c), and E(q) is the average directional bias. The observed displacement is the distance between the first and last bear location for the
respective temporal scale. The expected displacement, √(E(𝑅2)), is based on
the biased random walk and correlated random walk equations. All values are shown with ± standard error.
E(l)
(m)
E(c)
E(q)
Observed
displacement
(m)
Expected
displacement
(m)
A) Overall
Female 224 ± 21.2 -0.11 ± 0.03 - - -
Male 338.81 ± 39.82 -0.04 ± 0.02 - - -
B) Daily
Female 168.19 ± 15.18 0.01 ± 0.05 0.26 ± 0.04 1141.76 ± 134.85 1139.01 ± 112.3
Male 248.33 ± 27.04 0.12 ± 0.03 0.27 ± 0.03 2014.53 ± 247.93 1761.52 ± 167.94
C) Weekly
Female 153.38 ± 11.62 0.04 ± 0.05 0.12 ± 0.01 1862.9 ± 215.24 3173.62 ± 250.78
Male 235.91 ± 25.62 0.2 ± 0.02 0.18 ± 0.01 4798.57 ± 529.06 5215.97 ± 429.51
D) Monthly
Female 153.39 ± 9.35 0.07 ± 0.05 0.04 ± 0.01 2227.02 ± 376.24 6703.79 ± 431.83
Male 236.58 ± 26.28 0.22 ± 0.02 0.05 ± 0.01 7982.11 ± 1475.96 10575.2 ± 828.47
156
Table D-2. Model selection tables for responses of each weekly movement metric from Florida black bears in north-central Florida; A) Weekly average of bihourly step-length, E(l), B) Average weekly directional persistence, E(c), C) Average weekly directional bias, E(q), D) Average observed weekly displacement and
E) Average weekly expected displacement, √(E(𝑅2)). Models appear in order
of the difference in the Akaike Information Criterion score (ΔAICc, corrected for small sample sizes). Creeks, Major roads, and Minor roads all represent distances to the nearest respective feature. The number of parameters in each model is indicated by K. The difference in the log-likelihood from the top model is indicated by ΔLL. The weight indicates the ‘Akaike weight’ or model probability. Only the top 10 models are shown out of 32 for each weekly movement metric.
Rank Model K Deviance ΔLL ΔAICc Weight R2GLMM(c)
A. Average weekly step-length, E(l)
1 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
22 1112.32 0.00 0.00 0.17 0.46
2 Creeks + Major roads + Season
+ Sex + Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 1121.79 -4.73 0.98 0.10 0.45
157
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
3 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
20 1117.64 -2.66 1.07 0.10 0.46
4 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
23 1112.15 0.09 1.96 0.06 0.46
158
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
5 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
24 1110.36 0.98 2.31 0.05 0.46
6 Creeks + Major roads + Season
+ Sex + Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 1121.26 -4.47 2.56 0.05 0.46
7 Creeks + Major roads + Season
+ Sex + Major roads:Season +
Major roads:Sex + Season:Sex
+ Major roads:Season:Sex
16 1127.69 -7.68 2.68 0.04 0.44
159
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
8 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 1121.52 -4.60 2.83 0.04 0.45
9 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
18 1123.66 -5.67 2.85 0.04 0.45
10 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 1119.54 -3.61 2.97 0.04 0.46
160
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
B. Average Weekly directional persistence, E(c)
1 Creeks + Major roads + Season
+ Sex + Major roads:Season +
Major roads:Sex + Season:Sex
+ Major roads:Season:Sex
16 -95.70 0.00 0.00 0.23 0.30
2 Creeks + Major roads + Season
+ Sex + Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 -95.77 0.04 2.02 0.08 0.30
3 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 -95.71 0.01 2.08 0.08 0.30
4 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 -101.96 3.13 2.17 0.08 0.32
161
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
5 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
18 -97.44 0.87 2.46 0.07 0.31
6 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Season:Sex +
Major roads:Season:Sex
19 -99.45 1.88 2.56 0.06 0.31
7 Creeks + Major roads + Season
+ Sex + Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 -96.85 0.58 3.05 0.05 0.30
162
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
8 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
22 -104.35 4.33 4.04 0.03 0.32
9 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 -95.77 0.04 4.13 0.03 0.30
10 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 -102.03 3.17 4.22 0.03 0.32
163
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
C. Average Weekly directional bias, E(q)
1 Creeks + Minor roads + Season
+ Sex + Creeks:Season +
Creeks:Sex + Minor roads:Sex
13 -527.65 0.00 0.00 0.03 0.09
2 Creeks + Minor roads + Sex +
Creeks:Sex + Minor roads:Sex
9 -519.36 -4.14 0.02 0.03 0.07
3 Creeks + Minor roads + Season
+ Sex + Creeks:Sex + Minor
roads:Sex
11 -523.01 -2.32 0.49 0.02 0.08
4 Minor roads + Season + Sex +
Minor roads:Sex
9 -518.76 -4.45 0.63 0.02 0.07
5 Creeks + Minor roads + Season
+ Sex + Creeks:Season + Minor
roads:Sex
12 -524.93 -1.36 0.65 0.02 0.07
6 Minor roads + Season + Sex +
Minor roads:Season + Minor
roads:Sex
11 -522.64 -2.50 0.86 0.02 0.07
7 Creeks + Minor roads + Season
+ Sex + Creeks:Season +
Creeks:Sex + Minor roads:Sex
+ Season:Sex
15 -530.85 1.60 0.96 0.02 0.09
164
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
8 Creeks + Minor roads + Sex +
Minor roads:Sex
8 -516.18 -5.74 1.15 0.01 0.06
9 Creeks + Minor roads + Season
+ Sex + Minor roads:Sex
10 -520.28 -3.69 1.16 0.01 0.07
10 Minor roads + Sex + Minor
roads:Sex
7 -514.07 -6.79 1.22 0.01 0.06
D. Weekly Observed displacement
1 Major roads + Minor roads +
Season + Sex + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
17 2532.32 0.00 0.00 0.10 0.29
2 Creeks + Major roads + Season
+ Sex + Major roads:Season +
Major roads:Sex + Season:Sex
+ Major roads:Season:Sex
16 2534.67 -1.17 0.25 0.08 0.29
165
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
3 Major roads + Season + Sex +
Major roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
15 2536.95 -2.31 0.43 0.08 0.28
4 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
18 2531.19 0.57 0.97 0.06 0.30
5 Creeks + Major roads + Season
+ Sex + Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 2531.86 0.23 1.64 0.04 0.29
6 Major roads + Minor roads +
Season + Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
16 2536.29 -1.98 1.86 0.04 0.29
166
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
7 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
19 2530.26 1.03 2.15 0.03 0.30
8 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
20 2528.21 2.06 2.22 0.03 0.30
9 Creeks + Major roads + Season
+ Sex + Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 2534.63 -1.15 2.30 0.03 0.29
167
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
10 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 2534.65 -1.16 2.33 0.03 0.29
E. Weekly Expected Displacement
1 Creeks + Major roads + Season
+ Sex + Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 1383.93 0.00 0.00 0.20 0.48
2 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
20 1381.52 1.20 1.83 0.08 0.48
168
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
3 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 1383.68 0.12 1.87 0.08 0.48
4 Creeks + Major roads + Season
+ Sex + Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 1383.73 0.10 1.92 0.08 0.48
5 Creeks + Major roads + Season
+ Sex + Major roads:Season +
Major roads:Sex + Season:Sex
+ Major roads:Season:Sex
16 1391.05 -3.56 2.92 0.05 0.47
169
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
6 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
22 1378.50 2.71 3.06 0.04 0.48
7 Creeks + Major roads + Season
+ Sex + Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
21 1380.68 1.62 3.11 0.04 0.48
8 Creeks + Major roads + Minor
roads + Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 1382.91 0.51 3.22 0.04 0.48
170
Table D-2. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
9 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Creeks:Sex +
Major roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 1383.58 0.17 3.88 0.03 0.48
10 Creeks + Major roads + Minor
roads + Season + Sex +
Creeks:Season + Creeks:Sex +
Major roads:Season + Major
roads:Sex + Minor roads:Sex +
Season:Sex + Major
roads:Season:Sex
21 1381.51 1.21 3.94 0.03 0.48
171
Table D-3. Model selection tables for each movement metric from Florida black bears in north-central Florida at the daily temporal scale. Models appear in order of the difference in the Akaike Information Criterion score (ΔAICc, corrected for small sample sizes). Creeks, Major roads, and Minor roads all represent distances to the nearest respective feature. At the bihourly scale, models with and without Land Cover as a covariate were tested. The number of parameters in each model is indicated by K. The difference in the log-likelihood from the top model is indicated by ΔLL. The weight indicates the ‘Akaike weight’ or model probability. Only the top 10 models for each movement metric are shown.
Rank Model K Deviance ΔLL ΔAICc Weight R2GLMM(c)
A. Average Daily Step-length, E(l)
1 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
23 11759.94 0.00 0.00 0.30 0.42
2 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
25 11757.77 1.08 1.88 0.12 0.42
172
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
3 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
22 11764.31 -2.18 2.35 0.09 0.42
4 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
24 11760.60 -0.33 2.68 0.08 0.42
173
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
5 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 11769.07 -4.56 3.07 0.07 0.41
6 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex + Minor
roads:Season:Sex
27 11755.04 2.45 3.19 0.06 0.42
174
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
7 Creeks + Major roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
21 11767.18 -3.62 3.20 0.06 0.42
8 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
23 11763.24 -1.65 3.30 0.06 0.42
175
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
9 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
25 11759.49 0.22 3.60 0.05 0.42
10 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 11767.88 -3.97 3.90 0.04 0.42
176
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
B. Averaged Daily Directional Persistence, E(c)
1 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 4991.03 0.00 0.00 0.14 0.14
2 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Season:Sex
+ Major roads:Season:Sex
21 4987.44 1.79 0.45 0.11 0.14
3 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
22 4985.96 2.53 0.99 0.09 0.14
177
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
4 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 4990.13 0.45 1.12 0.08 0.14
5 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 4990.60 0.21 1.59 0.06 0.14
6 Creeks + Major roads +
Minor roads + Season + Sex
+ Major roads:Season +
Major roads:Sex +
Season:Sex + Major
roads:Season:Sex
17 4997.05 -3.01 1.98 0.05 0.13
178
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
7 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Season:Sex
+ Major roads:Season:Sex
22 4987.18 1.93 2.20 0.05 0.14
8 Creeks + Major roads +
Minor roads + Season + Sex
+ Major roads:Season +
Major roads:Sex + Minor
roads:Season + Season:Sex
+ Major roads:Season:Sex
19 4993.39 -1.18 2.36 0.04 0.14
9 Creeks + Major roads +
Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 4995.76 -2.36 2.71 0.04 0.13
179
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
10 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 4989.89 0.57 2.89 0.03 0.14
C. Averaged Daily Directional Bias, E(q)
1 Creeks + Major roads +
Minor roads + Season + Sex
+ Major roads:Season +
Major roads:Sex +
Season:Sex + Major
roads:Season:Sex
17 4145.90 0.00 0.00 0.09 0.07
2 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 4141.98 1.96 0.11 0.08 0.08
180
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
3 Creeks + Major roads +
Minor roads + Season + Sex
+ Major roads:Season +
Major roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 4144.09 0.91 0.20 0.08 0.08
4 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 4144.50 0.70 0.61 0.06 0.08
5 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 4138.90 3.50 1.07 0.05 0.08
181
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
6 Major roads + Minor roads +
Season + Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
17 4147.08 -0.59 1.17 0.05 0.07
7 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 4143.22 1.34 1.35 0.04 0.08
8 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 4141.30 2.30 1.44 0.04 0.08
182
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
9 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 4141.59 2.15 1.74 0.04 0.08
10 Major roads + Minor roads +
Season + Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
16 4149.69 -1.89 1.77 0.04 0.07
D. Observed Daily Displacement
1 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 17258.59 0.00 0.00 0.19 0.30
183
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
2 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
23 17252.80 2.89 0.27 0.17 0.30
3 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 17257.09 0.75 0.52 0.15 0.30
184
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
4 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
22 17256.27 1.16 1.72 0.08 0.30
5 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
23 17254.89 1.85 2.36 0.06 0.30
6 Creeks + Major roads +
Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 17265.06 -3.24 2.44 0.06 0.30
185
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
7 Creeks + Major roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
21 17259.34 -0.38 2.78 0.05 0.30
8 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
25 17251.96 3.31 3.48 0.03 0.30
186
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
9 Creeks + Major roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 17264.28 -2.85 3.68 0.03 0.30
10 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
19 17265.00 -3.21 4.39 0.02 0.30
E. Expected Daily Displacement
1 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
23 12414.78 0.00 0.00 0.22 0.41
187
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
2 Creeks + Major roads +
Season + Sex +
Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
21 12419.67 -2.45 0.86 0.14 0.41
3 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
20 12422.57 -3.89 1.73 0.09 0.40
4 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
22 12418.59 -1.91 1.79 0.09 0.41
188
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
5 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
25 12412.80 0.99 2.07 0.08 0.41
6 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex +
Minor roads:Season:Sex
24 12414.96 -0.09 2.21 0.07 0.41
189
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
7 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Season:Sex +
Creeks:Season:Sex + Major
roads:Season:Sex
22 12419.58 -2.40 2.79 0.05 0.41
8 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
21 12422.13 -3.68 3.31 0.04 0.41
9 Creeks + Major roads +
Season + Sex +
Creeks:Season + Major
roads:Season + Major
roads:Sex + Season:Sex +
Major roads:Season:Sex
18 12428.29 -6.76 3.42 0.04 0.40
190
Table D-3. Continued Rank Model K Deviance ΔLL ΔAICc Weight R2
GLMM(c)
10 Creeks + Major roads +
Minor roads + Season + Sex
+ Creeks:Season +
Creeks:Sex + Major
roads:Season + Major
roads:Sex + Minor
roads:Season + Minor
roads:Sex + Season:Sex +
Major roads:Season:Sex
23 12418.23 -1.73 3.45 0.04 0.41
191
APPENDIX E FIGURES FROM MODELS OF MOVEMENT METRICS
Figures showing the effects of covariates for each movement metric not shown in
the main body of the manuscript at weekly, daily, and bihourly scales. Based on Florida
black bear biology, winter was defined as January 1 – April 30, summer as May 1 –
August 31, and fall as September 1 – December 31.
192
Figure E-1. Effect of covariates on the weekly average directional persistence, E(c), (± 95% CI) for Florida black bears: A) main effect of distance to creeks and B) 3-way interaction among sex, season, and distance to major roads. All distances are standardized.
193
Figure E-2. Effect of covariates on the weekly average directional bias, E(q), (± 95% CI) for Florida black bears: A) 2-way interaction between season and distance to creeks, B) 2-way interaction between sex and distance to creeks, and C) 2-way interaction between sex and distance to minor roads. All distances are standardized.
194
Figure E-3. Effect of covariates on the weekly expected displacement (± 95% CI) for Florida black bears: A) 2-way interaction between season and distance to creeks, B) 3-way interaction among sex, season, and distance to major roads. All weekly displacements are on the log scale and all distances are standardized.
195
Figure E-4. Effect of covariates on the daily average bihourly step-length (± 95% CI) for Florida black bears: A) 3-way interaction among sex, season, and distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, and C) 2-way interaction between sex and distance to minor roads. All step-lengths are on the log scale and all distances are standardized.
196
Figure E-5. Effect of covariates on the daily average directional persistence, E(c), (± 95% CI) for Florida black bears: A) 2-way interaction between season and distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, C) main effect of distance to minor roads. All distances are standardized.
197
Figure E-6. Effect of covariates on the daily average directional bias, E(q), (± 95% CI) for Florida black bears: A) main effect of distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, C) main effect of distance to minor roads. All distances are standardized.
198
Figure E-7. Effect of covariates on the daily observed displacement (± 95% CI) for Florida black bears: A) 2-way interaction between season and distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, C) 2-way interaction between sex and distance to minor roads. All distances are standardized.
199
Figure E-8. Effect of covariates on the daily expected displacement (± 95% CI) for Florida black bears: A) 3-way interaction among season, sex, and distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, C) 2-way interaction between sex and distance to minor roads. All distances are standardized.
200
Figure E-9. Effect of covariates on the bihourly step-length (± 95% CI) for Florida black bears: A) main effect of distance to creeks, B) 3-way interaction among sex, season, and distance to major roads, and C) 3-way interaction among sex, time of day, and distance to minor roads. All step-lengths are on the log scale and all distances are standardized. The model form was:
Sex × Season × (Land cover + Creeks + Major roads + Minor roads + Time of day) +
Sex × Time of day × (Land cover + Creeks + Major roads + Minor roads) +
Season × Time of day × (Land cover + Creeks + Major roads + Minor roads)
201
Figure E-10. Effect of covariates on the bihourly step-length (± 95% CI) for Florida black bears: A) 3-way interaction of sex, season, and time of day, B) 3-way interaction of sex, time of day, and land cover type. All step-lengths are on the log scale. The model form was:
Sex × Season × (Land cover + Creeks + Major roads + Minor roads + Time of day) +
Sex × Time of day × (Land cover + Creeks + Major roads + Minor roads) +
Season × Time of day × (Land cover + Creeks + Major roads + Minor roads)
202
APPENDIX F TABLES OF PLANT SPECIES FOUND IN AREAS OF HIGH-USE BY BEARS
Table F-1. Percent of high and low bear-use sites in which tree species were present in the Camp Blanding area of north-central Florida from A) hardwood trees and B) conifers, both from each sampled bear location measured using the Point Center Quarter method and trees ≥4cm diameter at breast height.
A) Hardwood species Percent
high-use
Percent
low-use Percent Percent Percent
Percent
high-use
Percent
low-use
Family Genus species Female Male Female Male High-use Low-use Total
Fagaceae Quercus nigra 35.85 30.91 44.00 29.09 33.33 36.19 34.74
Theaceae Gordonia lasianthus 24.53 40.00 18.00 21.82 32.41 20.00 26.29
Cornaceae Nyssa sp. 24.53 27.27 16.00 10.91 25.93 13.33 19.72
Lauraceae Persea sp. 18.87 18.18 2.00 20.00 18.52 11.43 15.02
Fagaceae Quercus laurifolia 26.42 10.91 26.00 30.91 18.52 28.57 23.47
Magnoliaceae Magnolia virginiana 13.21 14.55 12.00 3.64 13.89 7.62 10.80
Ericaceae Agarista populifolia 20.75 5.45 6.00 1.82 12.96 3.81 8.45
Altingiaceae Liquidambar styraciflua 11.32 9.09 14.00 9.09 10.19 11.43 10.80
Cyrillaceae Cyrilla racemiflora 11.32 3.64 4.00 3.64 7.41 3.81 5.63
Ericaceae Lyonia ferruginea 9.43 5.45 4.00 0.00 7.41 1.90 4.69
Fagaceae Quercus geminata 7.55 7.27 16.00 3.64 7.41 9.52 8.45
203
Table F-1. Continued
A) Hardwood species Percent
high-use
Percent
low-use Percent Percent Percent
Percent
high-
use
Percent
low-use
Family Genus species Female Male Female Male High-use Low-
use Total
Myricaceae Myrica cerifera 1.89 9.09 2.00 7.27 5.56 4.76 5.16
Fagaceae Quercus laevis 1.89 9.09 18.00 21.82 5.56 20.00 12.68
Fagaceae Quercus virginiana 7.55 3.64 10.00 5.45 5.56 7.62 6.57
Sapindaceae Acer rubrum 5.66 3.64 10.00 1.82 4.63 5.71 5.16
Magnoliaceae Liriodendron tulipifera 7.55 1.82 0.00 3.64 4.63 1.90 3.29
Ericaceae Vaccinium arboreum 9.43 0.00 4.00 1.82 4.63 2.86 3.76
Aquifoliaceae Ilex opaca 3.77 3.64 2.00 1.82 3.70 1.90 2.82
Juglandaceae Carya glabra 5.66 0.00 0.00 1.82 2.78 0.95 1.88
Aquifoliaceae Ilex cassine 1.89 3.64 2.00 1.82 2.78 1.90 2.35
Oleaceae Cartrema americana 3.77 0.00 2.00 0.00 1.85 0.95 1.41
Ebenacea Diospyros virginiana 1.89 1.82 2.00 1.82 1.85 1.90 1.88
Ericaceae Lyonia lucida 1.89 1.82 0.00 1.82 1.85 0.95 1.41
Ericaceae Vaccinium corymbosum 1.89 1.82 0.00 0.00 1.85 0.00 0.94
204
Table F-1. Continued
A) Hardwood species Percent
high-use
Percent
low-use Percent Percent Percent
Percent
high-
use
Percent
low-use
Family Genus species Female Male Female Male High-use Low-
use Total
Betulaceae Alnus serrulata 0.00 1.82 0.00 0.00 0.93 0.00 0.47
Rubiaceae Cephalanthus
occidentalis 0.00 1.82 0.00 0.00 0.93 0.00 0.47
Oleaceae Chionanthus virginicus 1.89 0.00 0.00 0.00 0.93 0.00 0.47
Cornaceae Cornus florida 0.00 1.82 0.00 0.00 0.93 0.00 0.47
Oleaceae Fraxinus caroliniana 1.89 0.00 2.00 0.00 0.93 0.95 0.94
Aquifoliaceae Ilex vomitoria 0.00 1.82 0.00 0.00 0.93 0.00 0.47
Fagaceae Quercus incana 1.89 0.00 0.00 0.00 0.93 0.00 0.47
Fagaceae Quercus margarettae 0.00 1.82 14.00 1.82 0.93 7.62 4.23
Fagaceae Quercus myrtifolia 0.00 1.82 0.00 1.82 0.93 0.95 0.94
Ericaceae Rhododendron
viscosum 0.00 1.82 0.00 0.00 0.93 0.00 0.47
205
Table F-1. Continued
A) Hardwood species Percent
high-use
Percent
low-use Percent Percent Percent
Percent
high-
use
Percent
low-use
Family Genus species Female Male Female Male High-use Low-
use Total
Adoxaceae Viburnum nudum 1.89 0.00 0.00 0.00 0.93 0.00 0.47
Aquifoliaceae Ilex cassine var.
myrtifolia 0.00 0.00 2.00 0.00 0.00 0.95 0.47
Magnoliaceae Magnolia grandiflora 0.00 0.00 0.00 1.82 0.00 0.95 0.47
Rosaceae Prunus serotina 0.00 0.00 2.00 0.00 0.00 0.95 0.47
Fagaceae Quercus michauxii 0.00 0.00 0.00 1.82 0.00 0.95 0.47
Anacardiaceae Rhus copallinum 0.00 0.00 2.00 0.00 0.00 0.95 0.47
Ericaceae Vaccinium stamineum 0.00 0.00 2.00 0.00 0.00 0.95 0.47
Unknown 1 0.00 0.00 0.00 1.82 0.00 0.95 0.47
No tree ≥ 4 cm DBH within 60 m 1.89 0.00 0.00 7.27 0.93 3.81 2.35
206
Table F-1. Continued
B) Conifers Percent
high-use
Percent
low-use Percent Percent Percent
Percent
high-
use
Percent
low-use
Family Genus species Female Male Female Male High-use Low-
use Total
Pinacea Pinus taeda 64.15 34.55 32.00 10.91 49.07 20.95 35.21
Pinacea Pinus elliottii 32.08 56.36 48.00 61.82 44.44 55.24 49.77
Pinacea Pinus clausa 1.89 14.55 16.00 10.91 8.33 13.33 10.80
Pinacea Pinus palustris 7.55 1.82 14.00 21.82 4.63 18.10 11.27
Cupressaceae Taxodium sp. 1.89 7.27 2.00 1.82 4.63 1.90 3.29
Pinacea Pinus serotina 0.00 3.64 4.00 3.64 1.85 3.81 2.82
No tree ≥ 4 cm DBH within 60 m 43.40 14.55 18.00 16.36 28.70 17.14 23.00
207
Table F-2. Understory species found in plots 4x4 m plots at high- and low-use bear sites in the Camp Blanding area of north-central Florida. An asterix (*) next to the row indicates that there was a difference between high- and low-use plots for that species. Species with a 1 in their superscript were considered potential food resources.
Genus
species
Percent present in
plots Average cover when present
Average maximum height when
present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Smilax spp.1 70.37 76.19 73.24 11.62 ± 1.06 8.99 ± 0.89 0 ± 0 0 ± 0 8.18 ± 1.06 6.85 ± 0.89
Vitis
rotundifolia1 60.19 40.00 50.23 17.97 ± 1.41 16.31 ± 1.29 0 ± 0 0 ± 0 10.81 ± 1.41 6.52 ± 1.29 *
Serenoa
repens1 52.78 40.00 46.48 39.25 ± 2.67 31.4 ± 2.32 176.67 ± 9.7 145.5 ± 8.08 20.71 ± 2.67 12.56 ± 2.32 *
Osmunda
cinnamomea 50.00 24.76 37.56 11.78 ± 0.89 15.23 ± 0.8 59.99 ± 3.65 66.73 ± 3.16 5.89 ± 0.89 3.77 ± 0.8 *
Persea sp. 1 43.52 29.52 36.62 15 ± 0.94 11.13 ± 0.64 153.13 ± 10.42 145.16 ± 8.72 6.53 ± 0.94 3.29 ± 0.64 *
Lyonia
lucida1 40.74 14.29 27.70 38.64 ± 2.67 26.33 ± 1.33 181.32 ± 11.33 164.93 ± 6.05 15.74 ± 2.67 3.76 ± 1.33 *
208
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Ilex coriacea1 39.81 19.05 29.58 37.72 ± 2.42 35.25 ± 1.66 219.58 ± 13.2 261.1 ± 11.29 15.02 ± 2.42 6.71 ± 1.66 *
Order:
Poales 30.56 61.90 46.01 15.94 ± 1.26 26.98 ± 2.39 67.27 ± 4.14 59.23 ± 4.68 4.87 ± 1.26 16.7 ± 2.39 *
Vaccinium
corymbosum1 30.56 16.19 23.47 18.18 ± 0.96 18.82 ± 0.85 204.45 ± 10.82 161.65 ± 7.12 5.56 ± 0.96 3.05 ± 0.85 *
Quercus
nigra1 29.63 19.05 24.41 9.97 ± 0.57 19.55 ± 0.99 183.41 ± 12.63 209.5 ± 10.32 2.95 ± 0.57 3.72 ± 0.99
209
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Gordonia
lasianthus 26.85 16.19 21.60 28.14 ± 1.67 26.18 ± 1.11 233.28 ± 12.76 195.24 ± 7.8 7.56 ± 1.67 4.24 ± 1.11
Agarista
populifolia 26.85 12.38 19.72 21.41 ± 1.13 22.77 ± 0.94 250.21 ± 12.68 236.31 ± 8.7 5.75 ± 1.13 2.82 ± 0.94 *
Gelsemium
sempervirens 24.07 39.05 31.46 8.81 ± 0.53 13.68 ± 1.23 0 ± 0 0 ± 0 2.12 ± 0.53 5.34 ± 1.23 *
Woodwardia
areolata 20.37 8.57 14.55 7 ± 0.41 6.89 ± 0.24 19.55 ± 0.89 18 ± 0.53 1.43 ± 0.41 0.59 ± 0.24 *
Quercus spp. 1 18.52 16.19 17.37 9.85 ± 0.54 7.65 ± 0.4 75.65 ± 5.16 41 ± 2.15 1.82 ± 0.54 1.24 ± 0.4
210
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Itea virginica 18.52 3.81 11.27 10.15 ± 0.53 5.5 ± 0.15 82.9 ± 4.48 26.25 ± 0.65 1.88 ± 0.53 0.21 ± 0.15 *
Myrica cerifera 17.59 10.48 14.08 19.53 ± 0.99 17.73 ± 0.58 144.95 ± 6.37 155.55 ± 5.28 3.44 ± 0.99 1.86 ± 0.58
Pinus spp. 16.67 17.14 16.90 4.56 ± 0.32 2.67 ± 0.16 34.67 ± 2.83 27.33 ± 2.36 0.76 ± 0.32 0.46 ± 0.16
Rhododendron
viscosum 16.67 10.48 13.62 13.94 ± 0.68 20.45 ± 0.92 165.28 ± 7.24 181.73 ± 6.89 2.32 ± 0.68 2.14 ± 0.92
Vaccinium
arboreum1 14.81 17.14 15.96 20.63 ± 0.95 16.11 ± 0.73 212.19 ± 9.36 136.28 ± 6.46 3.06 ± 0.95 2.76 ± 0.73
211
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Acer rubrum 14.81 6.67 10.80 7.13 ± 0.58 4 ± 0.12 97.38 ± 7.44 59.71 ± 2.22 1.06 ± 0.58 0.27 ± 0.12
Ilex glabra1 13.89 26.67 20.19 30.67 ± 1.27 36.61 ± 2.04 195.47 ± 7.2 163.86 ± 7.79 4.26 ± 1.27 9.76 ± 2.04 *
Lyonia
ferruginea 12.96 3.81 8.45 20 ± 0.76 15 ± 0.3 213.64 ± 7.3 191.25 ± 3.88 2.59 ± 0.76 0.57 ± 0.3 *
Leucothoe
axillaris 12.04 5.71 8.92 19.77 ± 0.95 10.83 ± 0.33 68.69 ± 2.37 53.08 ± 1.57 2.38 ± 0.95 0.62 ± 0.33
Vaccinium
stamineum1 11.11 19.05 15.02 17.5 ± 0.65 21.3 ± 1.01 149.88 ± 5.08 157.95 ± 7.27 1.94 ± 0.65 4.06 ± 1.01
Gaylussacia
frondosa1 11.11 6.67 8.92 22.08 ± 0.82 18.57 ± 0.57 138.08 ± 5.61 153.14 ± 4.29 2.45 ± 0.82 1.24 ± 0.57
Nyssa sp. 1 10.19 3.81 7.04 15.09 ± 0.53 12.5 ± 0.31 275.82 ± 9.43 231.25 ± 6.26 1.54 ± 0.53 0.48 ± 0.31
212
Table F-2. Continued
Genus species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Mitchella
repens 10.19 2.86 6.57 9.27 ± 0.34 8.67 ± 0.2 4 ± 0.13 3.67 ± 0.06 0.94 ± 0.34 0.25 ± 0.2 *
Pteridium
aquilinum 9.26 13.33 11.27 7.6 ± 0.28 14.71 ± 0.6 34.4 ± 1.03 43.71 ± 1.62 0.7 ± 0.28 1.96 ± 0.6
Magnolia
virginiana 9.26 4.76 7.04 9.7 ± 0.37 22.2 ± 0.54 158 ± 6.07 284.4 ± 6.83 0.9 ± 0.37 1.06 ± 0.54
Cyrilla
racemiflora 8.33 7.62 7.98 17.22 ± 0.55 20.63 ± 0.67 275.56 ± 8.07 203.13 ± 6.21 1.44 ± 0.55 1.57 ± 0.67
Toxicodendron
radicans 8.33 3.81 6.10 11.67 ± 0.53 2 ± 0.05 32 ± 0.61 24 ± 0.23 0.97 ± 0.53 0.08 ± 0.05
Rubus sp.1 7.41 18.10 12.68 10.88 ± 0.43 18.79 ± 1.08 32.67 ± 0.8 55.19 ± 2.69 0.81 ± 0.43 3.4 ± 1.08 *
Quercus
laurifolia1 7.41 13.33 10.33 22.63 ± 0.68 23.29 ± 1.08 252.38 ± 7.63 216.07 ± 8.52 1.68 ± 0.68 3.1 ± 1.08
213
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Bignonia
capreolata 6.48 4.76 5.63 7.86 ± 0.2 6.2 ± 0.15 0 ± 0 0 ± 0 0.51 ± 0.2 0.3 ± 0.15
Parthenocissus
quinquefolia 6.48 2.86 4.69 2.29 ± 0.09 7 ± 0.15 5 ± 0.05 0 ± 0 0.15 ± 0.09 0.2 ± 0.15
Hamamelis
virginiana 6.48 0.95 3.76 19.29 ± 0.51 30 ± 0.29 242.14 ± 6.69 251 ± 2.39 1.25 ± 0.51 0.29 ± 0.29 *
Quercus
virginiana1 6.48 0.95 3.76 15 ± 0.48 10 ± 0.1 158.21 ± 5.76 230 ± 2.19 0.97 ± 0.48 0.1 ± 0.1 *
Galactia elliottii 5.56 17.14 11.27 3.83 ± 0.11 10.78 ± 0.74 39 ± 0.36 19 ± 0.18 0.21 ± 0.11 1.85 ± 0.74 *
Callicarpa
americana1 5.56 5.71 5.63 5.83 ± 0.14 16.67 ± 0.52 72.83 ± 2.09 96 ± 2.83 0.32 ± 0.14 0.95 ± 0.52
Liquidambar
styraciflua 5.56 4.76 5.16 13.5 ± 0.37 12 ± 0.32 178.83 ± 5.07 135.8 ± 3.27 0.75 ± 0.37 0.57 ± 0.32
Ilex vomitoria1 4.63 2.86 3.76 18 ± 0.4 23.33 ± 0.43 213.8 ± 4.84 70.67 ± 1.31 0.83 ± 0.4 0.67 ± 0.43
214
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Asimina
incana1 3.70 17.14 10.33 4 ± 0.08 7.44 ± 0.39 43.25 ± 0.86 48.22 ± 2 0.15 ± 0.08 1.28 ± 0.39 *
Quercus
laevis1 3.70 11.43 7.51 7.75 ± 0.2 10.08 ± 0.37 86 ± 2.14 65.92 ± 2.68 0.29 ± 0.2 1.15 ± 0.37 *
Tragia urens 3.70 11.43 7.51 1 ± 0.02 2.08 ± 0.11 26.25 ± 0.49 23.83 ± 0.79 0.04 ± 0.02 0.24 ± 0.11 *
Ilex opaca1 3.70 9.52 6.57 6.5 ± 0.15 5.8 ± 0.21 72.25 ± 1.89 60.8 ± 2.37 0.24 ± 0.15 0.55 ± 0.21
Cartrema
americana 3.70 0.00 1.88
21.25 ±
0.43 0 ± 0 260.25 ± 5.47 0 ± 0 0.79 ± 0.43 0 ± 0 *
Vaccinium
myrsinites1 2.78 10.48 6.57 10 ± 0.17 8.73 ± 0.32 39.33 ± 0.65 35.09 ± 1.15 0.28 ± 0.17 0.91 ± 0.32 *
215
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Opuntia
humifusa1 2.78 6.67 4.69 1 ± 0.02 3.43 ± 0.12 14.67 ± 0.25 13.26 ± 0.36 0.03 ± 0.02 0.23 ± 0.12
Quercus
geminata1 2.78 2.86 2.82 18.33 ± 0.38 30 ± 0.61 162 ± 3.43 137.67 ± 2.49 0.51 ± 0.38 0.86 ± 0.61
Carya glabra1 2.78 1.90 2.35 8.67 ± 0.17 5 ± 0.07 77 ± 1.42 23.5 ± 0.32 0.24 ± 0.17 0.1 ± 0.07
Ilex cassine1 2.78 1.90 2.35 13.33 ± 0.23 7.5 ± 0.11 261.33 ± 4.85 125 ± 1.82 0.37 ± 0.23 0.14 ± 0.11
Rhapidophyllum
hystrix 2.78 0.95 1.88 20 ± 0.4 10 ± 0.1 93.33 ± 1.49 157 ± 1.5 0.56 ± 0.4 0.1 ± 0.1
216
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Chionanthus
virginicus 2.78 0.00 1.41 13.33 ± 0.23 0 ± 0 196.33 ± 3.15 0 ± 0 0.37 ± 0.23 0 ± 0
Ditrysinia
fruticosa 2.78 0.00 1.41 10 ± 0.2 0 ± 0 112.67 ± 2.08 0 ± 0 0.28 ± 0.2 0 ± 0
Eupatorium sp. 1.85 16.19 8.92 12.5 ± 0.19 10.53 ± 0.57 151 ± 2 72.53 ± 2.87 0.23 ± 0.19 1.7 ± 0.57 *
Tephrosia sp. 1.85 2.86 2.35 3 ± 0.05 4 ± 0.1 20.5 ± 0.27 20 ± 0.19 0.06 ± 0.05 0.11 ± 0.1
Toxicodendron
vernix 1.85 2.86 2.35 22.5 ± 0.31 13.33 ± 0.26 298 ± 4.71 144.33 ± 2.48 0.42 ± 0.31 0.38 ± 0.26
217
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Lyonia
ligustrina 1.85 1.90 1.88 7.5 ± 0.1 32.5 ± 0.45 182.5 ± 2.5 246 ± 3.39 0.14 ± 0.1 0.62 ± 0.45
Solidago sp. 1.85 0.95 1.41 3 ± 0.05 1 ± 0.01 58.5 ± 0.76 86 ± 0.82 0.06 ± 0.05 0.01 ± 0.01
Thelypteris
palustris 1.85 0.95 1.41 10 ± 0.13 1 ± 0.01 31.5 ± 0.41 26 ± 0.25 0.19 ± 0.13 0.01 ± 0.01
Alnus
serrulata 1.85 0.00 0.94 17.5 ± 0.28 0 ± 0 310 ± 4.75 0 ± 0 0.32 ± 0.28 0 ± 0
Asimina
parviflora1 1.85 0.00 0.94 5.5 ± 0.09 0 ± 0 107.5 ± 1.66 0 ± 0 0.1 ± 0.09 0 ± 0
Carpinus
caroliniana 1.85 0.00 0.94 22.5 ± 0.37 0 ± 0 163 ± 2.85 0 ± 0 0.42 ± 0.37 0 ± 0
218
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Cephalanthus
occidentalis 1.85 0.00 0.94 12.5 ± 0.17 0 ± 0 221 ± 2.9 0 ± 0 0.23 ± 0.17 0 ± 0
Saururus
cernuus 1.85 0.00 0.94 8 ± 0.14 0 ± 0 22.5 ± 0.32 0 ± 0 0.15 ± 0.14 0 ± 0
Sideroxylon
tenax 1.85 0.00 0.94 7.5 ± 0.1 0 ± 0 142 ± 1.97 0 ± 0 0.14 ± 0.1 0 ± 0
Symplocos
tinctoria 1.85 0.00 0.94 22.5 ± 0.31 0 ± 0 276.5 ± 3.71 0 ± 0 0.42 ± 0.31 0 ± 0
Viburnum
dentatum 1.85 0.00 0.94 10 ± 0.13 0 ± 0 166.5 ± 2.17 0 ± 0 0.19 ± 0.13 0 ± 0
Licania
michauxii 0.93 6.67 3.76 20 ± 0.19 13 ± 0.38 25 ± 0.23
21.86 ±
0.55 0.19 ± 0.19 0.87 ± 0.38 *
219
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Rhus
copallinum 0.93 6.67 3.76 10 ± 0.09 5.86 ± 0.16 42 ± 0.39 42.14 ± 1.18 0.09 ± 0.09 0.39 ± 0.16 *
Quercus
margarettae1 0.93 5.71 3.29 5 ± 0.05 10.83 ± 0.25 28 ± 0.26 96.17 ± 2.76 0.05 ± 0.05 0.62 ± 0.25 *
Stillingia
sylvatica 0.93 5.71 3.29 1 ± 0.01 3.83 ± 0.12 0 ± 0 26.67 ± 0.61 0.01 ± 0.01 0.22 ± 0.12 *
Stylisma
patens 0.93 4.76 2.82 10 ± 0.09 1.8 ± 0.05 0 ± 0 0 ± 0 0.09 ± 0.09 0.09 ± 0.05
Diospyros
virginiana1 0.93 3.81 2.35 10 ± 0.09 11.25 ± 0.24 68 ± 0.63 79.5 ± 1.58 0.09 ± 0.09 0.43 ± 0.24
220
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Houstonia
procumbens 0.93 2.86 1.88 1 ± 0.01 2.33 ± 0.05 6 ± 0.06 2.5 ± 0.04 0.01 ± 0.01 0.07 ± 0.05
Pinus clausa 0.93 2.86 1.88 15 ± 0.14 8.67 ± 0.17 550 ± 5.09 118.67 ± 2.04 0.14 ± 0.14 0.25 ± 0.17
Arnoglossum
floridanum 0.93 1.90 1.41 5 ± 0.05 5 ± 0.07 88 ± 0.81 8 ± 0.11 0.05 ± 0.05 0.1 ± 0.07
Diodia teres 0.93 1.90 1.41 5 ± 0.05 1 ± 0.01 18 ± 0.17 49 ± 0.8 0.05 ± 0.05 0.02 ± 0.01
Magnolia
grandiflora 0.93 1.90 1.41 5 ± 0.05 7.5 ± 0.11 25 ± 0.23 47 ± 0.74 0.05 ± 0.05 0.14 ± 0.11
221
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Osmunda
regalis 0.93 1.90 1.41 10 ± 0.09 7.5 ± 0.11 89 ± 0.82 36.5 ± 0.49 0.09 ± 0.09 0.14 ± 0.11
Urtica dioica 0.93 1.90 1.41 1 ± 0.01 3 ± 0.05 3 ± 0.03 14.5 ± 0.21 0.01 ± 0.01 0.06 ± 0.05
Viburnum
nudum 0.93 1.90 1.41 1 ± 0.01 1 ± 0.01 8.5 ± 0.08 12.5 ± 0.17 0.01 ± 0.01 0.02 ± 0.01
Aesculus
pavia 0.93 0.95 0.94 5 ± 0.05 5 ± 0.05 27 ± 0.25 40 ± 0.38 0.05 ± 0.05 0.05 ± 0.05
Chrysopsis
scabrella 0.93 0.95 0.94 5 ± 0.05 5 ± 0.05 12 ± 0.11 11 ± 0.1 0.05 ± 0.05 0.05 ± 0.05
Cornus
foemina1 0.93 0.95 0.94 20 ± 0.19 1 ± 0.01 294 ± 2.72 17 ± 0.16 0.19 ± 0.19 0.01 ± 0.01
222
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Pinckneya
bracteata 0.93 0.95 0.94 5 ± 0.05 30 ± 0.29 65 ± 0.6 340 ± 3.24 0.05 ± 0.05 0.29 ± 0.29
Quercus
incana1 0.93 0.95 0.94 1 ± 0.01 20 ± 0.19 15 ± 0.14 75 ± 0.71 0.01 ± 0.01 0.19 ± 0.19
Rhexia
petiolata 0.93 0.95 0.94 5 ± 0.05 5 ± 0.05 23 ± 0.21 47 ± 0.45 0.05 ± 0.05 0.05 ± 0.05
Woodwardia
virginica 0.93 0.95 0.94 5 ± 0.05 20 ± 0.19 33 ± 0.31 100 ± 0.95 0.05 ± 0.05 0.19 ± 0.19
Unknown X 0.93 0.95 0.94 10 ± 0.09 20 ± 0.19 47 ± 0.44 3 ± 0.03 0.09 ± 0.09 0.19 ± 0.19
Cercis
canadensis 0.93 0.00 0.47 30 ± 0.28 0 ± 0 81 ± 0.75 0 ± 0 0.28 ± 0.28 0 ± 0
223
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Clethra
alnifolia 0.93 0.00 0.47 5 ± 0.05 0 ± 0 90 ± 0.83 0 ± 0 0.05 ± 0.05 0 ± 0
Cornus
florida1 0.93 0.00 0.47 5 ± 0.05 0 ± 0 16.5 ± 0.15 0 ± 0 0.05 ± 0.05 0 ± 0
Desmodium
sp. 0.93 0.00 0.47 5 ± 0.05 0 ± 0 89 ± 0.82 0 ± 0 0.05 ± 0.05 0 ± 0
Erythrina
herbacea 0.93 0.00 0.47 5 ± 0.05 0 ± 0 42 ± 0.39 0 ± 0 0.05 ± 0.05 0 ± 0
Eupatorium
rotundifolium 0.93 0.00 0.47 1 ± 0.01 0 ± 0 31 ± 0.29 0 ± 0 0.01 ± 0.01 0 ± 0
224
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Prunus
serotina1 0.93 0.00 0.47 10 ± 0.09 0 ± 0 300 ± 2.78 0 ± 0 0.09 ± 0.09 0 ± 0
Sericocarpus
tortifolius 0.93 0.00 0.47 1 ± 0.01 0 ± 0 15 ± 0.14 0 ± 0 0.01 ± 0.01 0 ± 0
Taxodium sp. 0.93 0.00 0.47 1 ± 0.01 0 ± 0 33 ± 0.31 0 ± 0 0.01 ± 0.01 0 ± 0
Tilia
americana 0.93 0.00 0.47 20 ± 0.19 0 ± 0 303 ± 2.81 0 ± 0 0.19 ± 0.19 0 ± 0
Unknown X 0.93 0.00 0.47 20 ± 0.19 0 ± 0 0 ± 0 0 ± 0 0.19 ± 0.19 0 ± 0
Unknown X 0.93 0.00 0.47 1 ± 0.01 0 ± 0 28 ± 0.26 0 ± 0 0.01 ± 0.01 0 ± 0
225
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Unknown X 0.93 0.00 0.47 5 ± 0.05 0 ± 0 16 ± 0.15 0 ± 0 0.05 ± 0.05 0 ± 0
Unknown X 0.93 0.00 0.47 1 ± 0.01 0 ± 0 0 ± 0 0 ± 0 0.01 ± 0.01 0 ± 0
Unknown X 0.93 0.00 0.47 1 ± 0.01 0 ± 0 15 ± 0.14 0 ± 0 0.01 ± 0.01 0 ± 0
Pterocaulon
pycnostachyum 0.00 8.57 4.23 0 ± 0 4.44 ± 0.21 0 ± 0 37.33 ± 1.11 0 ± 0 0.38 ± 0.21 *
Cnidoscolus
stimulosus 0.00 6.67 3.29 0 ± 0 2.86 ± 0.11 0 ± 0 19.14 ± 0.57 0 ± 0 0.19 ± 0.11 *
226
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Croton
argyranthemus 0.00 4.76 2.35 0 ± 0 3.4 ± 0.08 0 ± 0 18.6 ± 0.41 0 ± 0 0.16 ± 0.08 *
Prunus
umbellata1 0.00 4.76 2.35 0 ± 0 15 ± 0.37 0 ± 0 160.6 ± 3.7 0 ± 0 0.71 ± 0.37 *
Carphephorus
sp. 0.00 3.81 1.88 0 ± 0 4 ± 0.08 0 ± 0 7.75 ± 0.15 0 ± 0 0.15 ± 0.08 *
Eryngium
aromaticum 0.00 3.81 1.88 0 ± 0 2 ± 0.05 0 ± 0 20.63 ± 0.43 0 ± 0 0.08 ± 0.05 *
Tephrosia
chrysophylla 0.00 3.81 1.88 0 ± 0 7.5 ± 0.15 0 ± 0 2 ± 0.02 0 ± 0 0.29 ± 0.15 *
227
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Rhexia sp. 0.00 2.86 1.41 0 ± 0 3.67 ± 0.07 0 ± 0 38.33 ± 0.64 0 ± 0 0.1 ± 0.07
Asimina
obovata1 0.00 1.90 0.94 0 ± 0 7.5 ± 0.11 0 ± 0 63 ± 0.86 0 ± 0 0.14 ± 0.11
Centella
asiatica 0.00 1.90 0.94 0 ± 0 3 ± 0.05 0 ± 0 15.5 ± 0.25 0 ± 0 0.06 ± 0.05
Chapmannia
floridana 0.00 1.90 0.94 0 ± 0 5 ± 0.07 0 ± 0 52 ± 0.71 0 ± 0 0.1 ± 0.07
Croton
michauxii 0.00 1.90 0.94 0 ± 0 1 ± 0.01 0 ± 0 41.5 ± 0.57 0 ± 0 0.02 ± 0.01
Eryngium
baldwinii 0.00 1.90 0.94 0 ± 0 1 ± 0.01 0 ± 0 17 ± 0.23 0 ± 0 0.02 ± 0.01
228
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Euthamia
caroliniana 0.00 1.90 0.94 0 ± 0 1 ± 0.01 0 ± 0 13.5 ± 0.18 0 ± 0 0.02 ± 0.01
Galactia sp. 0.00 1.90 0.94 0 ± 0 1 ± 0.01 0 ± 0 0 ± 0 0 ± 0 0.02 ± 0.01
Galium sp. 0.00 1.90 0.94 0 ± 0 3 ± 0.05 0 ± 0 24 ± 0.35 0 ± 0 0.06 ± 0.05
Gaylussacia
dumosa 0.00 1.90 0.94 0 ± 0 22.5 ± 0.38 0 ± 0 21.5 ± 0.29 0 ± 0 0.43 ± 0.38
Lachnanthes
caroliana 0.00 1.90 0.94 0 ± 0 5 ± 0.07 0 ± 0 48.5 ± 0.65 0 ± 0 0.1 ± 0.07
Matelea
floridana 0.00 1.90 0.94 0 ± 0 3 ± 0.05 0 ± 0 0 ± 0 0 ± 0 0.06 ± 0.05
229
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Photinia
pyrifolia 0.00 1.90 0.94 0 ± 0 3 ± 0.05 0 ± 0 60 ± 0.9 0 ± 0 0.06 ± 0.05
Pleopeltis
polypodioides 0.00 1.90 0.94 0 ± 0 3 ± 0.05 0 ± 0 4 ± 0.06 0 ± 0 0.06 ± 0.05
Aeschynomene
viscidula 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 0 ± 0 0 ± 0 0.01 ± 0.01
Arisaema
triphyllum 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 9 ± 0.09 0 ± 0 0.05 ± 0.05
Aristolochia
serpentaria 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 12 ± 0.11 0 ± 0 0.1 ± 0.1
230
Table F-2. Continued
Genus species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Asimina
pygmea1 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 44 ± 0.42 0 ± 0 0.01 ± 0.01
Balduina
angustifolia 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 16 ± 0.15 0 ± 0 0.01 ± 0.01
Baptisia
lanceolata 0.00 0.95 0.47 0 ± 0 15 ± 0.14 0 ± 0 48 ± 0.46 0 ± 0 0.14 ± 0.14
Baptisia
lecontei 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 50 ± 0.48 0 ± 0 0.1 ± 0.1
Bidens
frondosa 0.00 0.95 0.47 0 ± 0 40 ± 0.38 0 ± 0 117 ± 1.11 0 ± 0 0.38 ± 0.38
Carya
tomentosa1 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 17 ± 0.16 0 ± 0 0.05 ± 0.05
231
Table F-2. Continued
Genus species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Castanea
pumila1 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 67 ± 0.64 0 ± 0 0.1 ± 0.1
Centrosema
arenicola 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 0 ± 0 0 ± 0 0.05 ± 0.05
Cirsium sp. 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 34 ± 0.32 0 ± 0 0.05 ± 0.05
Erechtites
hieraciifolius 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 171 ± 1.63 0 ± 0 0.05 ± 0.05
Froelichia
floridana 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 27 ± 0.26 0 ± 0 0.01 ± 0.01
232
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height when
present (cm)
Average cover over all
plots (includes zeros)
High Low All High Low High Low High Low
Galium
tinctorium 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 74 ± 0.7 0 ± 0 0.05 ± 0.05
Hieracium
gronovii 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 4 ± 0.04 0 ± 0 0.01 ± 0.01
Hypericum
sp. 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 60.2 ± 0.57 0 ± 0 0.01 ± 0.01
Juniperus
virginiana 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 62 ± 0.59 0 ± 0 0.1 ± 0.1
Kalmia
hirsuta 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 42 ± 0.4 0 ± 0 0.1 ± 0.1
Lechea
mucronata 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 40 ± 0.38 0 ± 0 0.05 ± 0.05
Liriodendron
tulipifera 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 8 ± 0.08 0 ± 0 0.05 ± 0.05
233
Table F-2. Continued
Genus
species
Percent present in
plots
Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Ludwigia sp. 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 69 ± 0.66 0 ± 0 0.01 ± 0.01
Mimosa
quadrivalvis 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 11 ± 0.1 0 ± 0 0.01 ± 0.01
Pediomelum
canescens 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 26 ± 0.25 0 ± 0 0.01 ± 0.01
Physalis
arenicola 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 25 ± 0.24 0 ± 0 0.01 ± 0.01
Pinus elliottii 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 164 ± 1.56 0 ± 0 0.01 ± 0.01
234
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum
height when present
(cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Pinus glabra 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 210 ± 2 0 ± 0 0.05 ± 0.05
Pinus
palustris 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 39 ± 0.37 0 ± 0 0.1 ± 0.1
Plantago
sparsiflora 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 115 ± 1.1 0 ± 0 0.1 ± 0.1
Pluchea
foetida 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 110 ± 1.05 0 ± 0 0.01 ± 0.01
Polygala
lutea 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 17 ± 0.16 0 ± 0 0.05 ± 0.05
Polygala sp. 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 4 ± 0.04 0 ± 0 0.01 ± 0.01
235
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Proserpinaca
pectinata 0.00 0.95 0.47 0 ± 0 15 ± 0.14 0 ± 0 30 ± 0.29 0 ± 0 0.14 ± 0.14
Quercus
michauxii1 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 76 ± 0.72 0 ± 0 0.1 ± 0.1
Quercus
myrtifolia1 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 75 ± 0.71 0 ± 0 0.05 ± 0.05
Rhexia
cubensis 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 57 ± 0.54 0 ± 0 0.01 ± 0.01
Sabatia
difformis 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 36 ± 0.34 0 ± 0 0.01 ± 0.01
236
Table F-2. Continued
Genus
species
Percent present in plots Average cover when
present
Average maximum height
when present (cm)
Average cover over all plots
(includes zeros)
High Low All High Low High Low High Low
Unknown X 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 25 ± 0.24 0 ± 0 0.1 ± 0.1
Unknown X 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 7 ± 0.07 0 ± 0 0.01 ± 0.01
Unknown X 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 29 ± 0.28 0 ± 0 0.05 ± 0.05
Unknown X 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 24 ± 0.23 0 ± 0 0.01 ± 0.01
Unknown X 0.00 0.95 0.47 0 ± 0 10 ± 0.1 0 ± 0 4 ± 0.04 0 ± 0 0.1 ± 0.1
Unknown X 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 12 ± 0.11 0 ± 0 0.01 ± 0.01
Unknown X 0.00 0.95 0.47 0 ± 0 5 ± 0.05 0 ± 0 5 ± 0.05 0 ± 0 0.05 ± 0.05
Unknown X 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 12 ± 0.11 0 ± 0 0.01 ± 0.01
Unknown X 0.00 0.95 0.47 0 ± 0 1 ± 0.01 0 ± 0 40 ± 0.38 0 ± 0 0.01 ± 0.01
237
LIST OF REFERENCES
ABRAHMS, B., N. R. JORDAN, K. A. GOLABEK, J. W. MCNUTT, A. M. WILSON AND J. S. BRASHARES. 2016. Lessons from integrating behaviour and resource selection: activity-specific responses of African wild dogs. Animal Conservation 19:247–255.
ADOLPH, S. C. 1990. Influence of behavioral thermoregulation on microhabitat use by two Scelopourus lizards. Ecology 71:315–327.
AEBISCHER, N. J., P. A. ROBERTSON AND R. E. KENWARD. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74:1313–1325.
AGER, A. A., B. K. JOHNSON, J. W. KERN AND J. G. KIE. 2003. Daily and seasonal movements and habitat use by female rocky mountain elk and mule deer. Journal of Mammalogy 84:1076–1088.
ALLEN, A. M. AND N. J. SINGH. 2016. Linking movement ecology with wildlife management and conservation. Frontiers in Ecology and Evolution 3:1–13.
ALT, G. L., G. J. MATULA JR., F. W. ALT AND J. S. LINDZEY. 1980. Dynamics of home range and movements of adult black bears in northeastern Pennsylvania. International Conference on Bear Research and Management 4:131–136.
AMSTRUP, S. C. AND J. BEECHAM. 1976. Activity patterns of radio-collared black bears in Idaho. Journal of Wildlife Management 40:340–348.
ANDREASSEN, H. P., K. HERTZBERG AND R. A. IMS. 1998. Space-use responses to habitat fragmentation and connectivity in the root vole Microtus oeconomus. Ecology 79:1223–1235.
ANDRÉN, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: A review. Oikos 71:355.
ASCHOFF, J. 1966. Circadian activity pattern with two peaks. Ecology 47:657–662.
AUGER-MÉTHÉ, M., A. E. DEROCHER, C. A. DEMARS, M. J. PLANK, E. A. CODLING AND M. A. LEWIS. 2016. Evaluating random search strategies in three mammals from distinct feeding guilds. Journal of Animal Ecology 85:1411–1421.
AVGAR, T., A. MOSSER, G. S. BROWN AND J. M. FRYXELL. 2013. Environmental and individual drivers of animal movement patterns across a wide geographical gradient. Journal of Animal Ecology 82:96–106.
AVGAR, T., J. R. POTTS, M. A. LEWIS AND M. S. BOYCE. 2016. Integrated step selection analysis: bridging the gap between resource selection and animal movement. Methods in Ecology and Evolution 7:619–630.
BAGUETTE, M., S. BLANCHET, D. LEGRAND, V. M. STEVENS AND C. TURLURE. 2013. Individual dispersal, landscape connectivity and ecological networks. Biological Reviews 88:310–326.
BAILEY, H. AND P. THOMPSON. 2006. Quantitative analysis of bottlenose dolphin movement patterns and their relationship with foraging. Journal of Animal Ecology 75:456–465.
238
BALDWIN, R. A. AND L. C. BENDER. 2009. Foods and nutritional components of diets of black bear in Rocky Mountain National Park, Colorado. Canadian Journal of Zoology 87:1000–1008.
BALES, S. L., E. C. HELLGREN, D. M. LESLIE AND J. HEMPHILL. 2005. Dynamics of a recolonizing population of black bears in the Ouachita Mountains of Oklahoma. Wildlife Society Bulletin 33:1342–1351.
BARTON, K. 2015. MuMIn: Multi-model inference. R package version 1.15.1. <https://CRAN.R-project.org/package=MuMIn>.
BARTON, P. S. ET AL. 2015. Guidelines for using movement science to inform biodiversity policy. Environmental Management 56:791–801.
BARTUMEUS, F., J. CATALAN, G. M. VISWANATHAN, E. P. RAPOSO AND M. G. E. DA LUZ. 2008. The influence of turning angles on the success of non-oriented animal searches. Journal of Theoretical Biology 252:43–55.
BARUCH-MORDO, S., S. W. BRECK, K. R. WILSON AND D. M. THEOBALD. 2008. Spatiotemporal distribution of black bear–human conflicts in Colorado, USA. Journal of Wildlife Management 72:1853–1862.
BARUCH-MORDO, S., K. R. WILSON, D. L. LEWIS, J. BRODERICK, J. S. MAO AND S. W. BRECK. 2014. Stochasticity in natural forage production affects use of urban areas by black bears: Implications to management of human–bear conflicts. PLoS ONE 9: e85122:1–10.
BASILLE, M. ET AL. 2013. Selecting habitat to survive: the impact of road density on survival in a large carnivore. PLoS ONE 8.
BATEMAN, P. W. AND P. A. FLEMING. 2012. Big city life: Carnivores in urban environments. Journal of Zoology 287:1–23.
BATES, D., M. MÄCHLER, B. M. BOLKER AND S. C. WALKER. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1–48.
BECKMANN, J. P. AND J. BERGER. 2003. Rapid ecological and behavioural changes in carnivores: the responses of black bears (Ursus americanus) to altered food. Journal of Zoology 261:207–212.
BEIER, P., D. R. MAJKA AND W. D. SPENCER. 2008. Forks in the road: Choices in procedures for designing wildland linkages. Conservation Biology 22:836–51.
BEIER, P. AND R. F. NOSS. 1998. Do habitat corridors provide connectivity? Conservation Biology 12:1241–1252.
BELANT, J. L., K. KIELLAND, E. H. FOLLMANN AND L. G. ADAMS. 2006. Interspecific resource partitioning in sympatric ursids. Ecological Applications 16:2333–2343.
BELOTTI, E., M. HEURICH, J. KREISINGER, P. SUSTR AND L. BUFKA. 2012. Influence of tourism and traffic on the Eurasian lynx hunting activity and daily movements. Animal Biodiversity and Conservation 35:235–246.
BENHAMOU, S. 1992. Efficiency of area-concentrated searching behaviour in a continuous patchy environment. Journal of Theoretical Biology 159:67–81.
239
BENHAMOU, S. 2006. Detecting an orientation component in animal paths when the preferred direction is individual-dependent. Ecology 87:518–528.
BENÍTEZ-LÓPEZ, A., R. ALKEMADE AND P. A. VERWEIJ. 2010. The impacts of roads and other infrastructure on mammal and bird populations: A meta-analysis. Biological Conservation 143:1307–1316.
BENNETT, A. F. AND D. A. SAUNDERS. 2010. Habitat fragmentation and landscape change. Pp. 88–106inConservation Biology for All(N. S. Sodhi & P. R. Ehrlich, eds.). Oxfod University Press Inc., New York, NY.
BENSON, J. F. ET AL. 2011. Intentional genetic introgression influences survival of adults and subadults in a small, inbred felid population. Journal of Animal Ecology 80:958–967.
BENSON, J. F. AND M. J. CHAMBERLAIN. 2006. Food habits of Louisiana black bears (Ursus americanus luteolus) in two subpopulations of the Tensas River Basin. The American Midland Naturalist 156:118–127.
BENSON, J. F. AND M. J. CHAMBERLAIN. 2007. Space use and habitat selection by female Louisiana black bears in the Tensas River Basin of Louisiana. Journal of Wildlife Management 71:117–126.
BERGMAN, C. M., J. A. SCHAEFER AND S. N. LUTTICH. 2000. Caribou movement as a correlated random walk. Oecologia 123:364–374.
BERINGER, J. J., S. G. SEIBERT AND M. R. PELTON. 1990. Incidence of road crossing by black bears on Pisgah National Forest, North Carolina. Bears: Their Biology and Management 8:85–92.
BESCHTA, R. L. AND W. J. RIPPLE. 2012. The role of large predators in maintaining riparian plant communities and river morphology. Geomorphology 157–158:88–98.
BEYER, H. L. ET AL. 2010. The interpretation of habitat preference metrics under use-availability designs. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365:2245–2254.
BEYER, H. L. 2015. Geospatial Modelling Environment (Version 0.7.4.0).
BEYER, H. L. ET AL. 2016. “You shall not pass!”: quantifying barrier permeability and proximity avoidance by animals. Journal of Animal Ecology 85:43–53.
BIVAND, R. AND N. LEWIN-KOH. 2016. maptools: tools for reading and handling spatial objects. R package version 0.8-39. https://CRAN.R-project.org/package=maptools.
BIVAND, R. AND C. RUNDEL. 2016. rgeos: interface to Geometry Engine - Open Source (GEOS). R package version 0.3-20. https://CRAN.R-project.org/package=rgeos.
BJORNLIE, D. D. ET AL. 2014. Methods to estimate distribution and range extent of grizzly bears in the greater Yellowstone ecosystem. Wildlife Society Bulletin 38:182–187.
BOCEDI, G., D. ZURELL, B. REINEKING AND J. M. J. TRAVIS. 2014. Mechanistic modelling of animal dispersal offers new insights into range expansion dynamics across fragmented landscapes. Ecography 37:1240–1253.
BOLKER, B. M. ET AL. 2009. Generalized linear mixed models: a practical guide for
240
ecology and evolution. Trends in Ecology and Evolution 24:127–135.
BÖRGER, L. ET AL. 2014. Effects of sampling regime on the mean and variance of home range size estimates 75:1393–1405.
BOYCE, M. S., P. R. VERNIER, S. E. NIELSEN AND F. K. A. SCHMIEGELOW. 2002. Evaluating resource selection functions. Ecological Modelling 157:281–300.
BOYD, C., A. E. PUNT, H. WEIMERSKIRCH AND S. BERTRAND. 2014. Movement models provide insights into variation in the foraging effort of central place foragers. Ecological Modelling 286:13–25.
BRADY, J. R. AND D. S. MAEHR. 1985. Distribution of black bears in Florida. Florida Field Naturalist 13:1–24.
BREWSTER, K. N. AND C. M. BREWSTER. 1991. Movement and microhabitat use by juvenile wood turtles introduced into a riparian habitat. Journal of Herpetology 25:379–382.
BRIDGES, A S., M. R. VAUGHAN AND S. KLENZENDORF. 2004. Seasonal variation in American black bear Ursus americanus activity patterns: quantification via remote photography. Wildlife Biology 10:277–284.
BRODY, A. J. AND M. R. PELTON. 1989. Effects of roads on black bear movements in western North Carolina. Wildlife Society Bulletin 17:5–10.
BROWER, L. P., O. R. TAYLOR, E. H. WILLIAMS, D. A. SLAYBACK, R. R. ZUBIETA AND I. M. RAMIREZ. 2012. Decline of monarch butterflies overwintering in Mexico: is the migratory phenomenon at risk? Insect Conservation and Diversity 5:95–100.
BRUGGEMAN, J. E., R. A. GARROTT, P. J. WHITE, F. G. R. WATSON AND R. WALLEN. 2007. Covariates affecting spatial variability in bison travel behavior in Yellowstone National Park. Ecological Applications 17:1411–1423.
BUCHMANN, C. M., F. M. SCHURR, R. NATHAN AND F. JELTSCH. 2012. Movement upscaled - the importance of individual foraging movement for community response to habitat loss. Ecography 35:436–445.
BURNHAM, K. P. AND D. R. ANDERSON. 2002. Model selection and multimodel inference: a practical information–theoretic approach. P. in. Second. Springer, New York, NY.
BYRNE, M. E., J. CLINT MCCOY, J. W. HINTON, M. J. CHAMBERLAIN AND B. A. COLLIER. 2014. Using dynamic Brownian bridge movement modelling to measure temporal patterns of habitat selection. Journal of Animal Ecology 83:1234–1243.
CALENGE, C. 2006. The package adehabitat for the R software: a tool for the analysis of space and habitat use by animals. Ecological Modelling 197:516–519.
CHAPRON, G. ET AL. 2014. Recovery of large carnivores in Europe’s modern human-dominated landscapes. Science 346:1517–1519.
CHETKIEWICZ, C. L. B., C. C. ST. CLAIR AND M. S. BOYCE. 2006. Corridors for conservation: Pattern and integrating process. Annual Review of Ecology, Evolution, and Systematics 37:317–342.
CLARK, J. D., J. S. LAUFENBERG, M. DAVIDSON AND J. L. MURROW. 2015. Connectivity
241
among subpopulations of Louisiana black bears as estimated by a step selection function. Journal of Wildlife Management 79:1347–1360.
CLOBERT, J., J. F. LE GALLIARD, J. COTE, S. MEYLAN AND M. MASSOT. 2009. Informed dispersal, heterogeneity in animal dispersal syndromes and the dynamics of spatially structured populations. Ecology Letters 12:197–209.
CODLING, E. A., M. J. PLANK AND S. BENHAMOU. 2008. Random walk models in biology. Journal of the Royal Society, Interface 5:813–834.
COLCHERO, F., D. A. CONDE, C. MANTEROLA, C. CHÁVEZ, A. RIVERA AND G. CEBALLOS. 2011. Jaguars on the move: modeling movement to mitigate fragmentation from road expansion in the Mayan Forest. Animal Conservation 14:158–166.
CONOVER, W. J. 1999. Practical nonparametric statistics. P. in. Third. John Wiley & Sons, New York, NY.
COSTELLO, C. M. 2010. Estimates of dispersal and home-range fidelity in American black bears. Journal of Mammalogy 91:116–121.
COSTELLO, C. M., S. I. CAIN, R. M. NIELSON, C. SERVHEEN AND C. C. SCHWARTZ. 2013. Response of American black bears to the non-motorized expansion of a road corridor in Grand Teton National Park. Ursus 24:54–69.
COSTER, S. S. AND A. I. KOVACH. 2012. Anthropogenic influences on the spatial genetic structure of black bears. Conservation Genetics 13:1247–1257.
COTTAM, G. AND J. T. CURTIS. 1956. The use of distance measures in phytosociological sampling. Ecology 37:451–460.
COTTAM, G., J. T. CURTIS AND W. HALE. 1953. Some sampling characteristics of a population of randomly dispersed individuals. Ecology 34:741–757.
COULLOUDON, B. ET AL. 1999. Sampling vegetation attributes. Denver, CO.
CROOKS, K. R. 2002. Relative sensitivities of mammalian carnivores to habitat fragmentation. Conservation Biology 16:488–502.
DICKSON, B. G., J. S. JENNESS AND P. BEIER. 2005. Influence of vegetation, topography, and roads on cougar movement in southern California. Journal of Wildlife Management 69:264–276.
DIFFENDORFER, J. E., M. S. GAINES AND R. D. HOLT. 1995. Habitat fragmentation and movements of three small mammals (Sigmodon, Microtus, and Peromyscus). Ecology 76:827–839.
DITMER, M. A. ET AL. 2015a. Behavioral and physiological responses of American black bears to landscape features within an agricultural region. Ecosphere 6:art28.
DITMER, M. A., T. E. BURK AND D. L. GARSHELIS. 2015b. Do innate food preferences and learning affect crop raiding by American black bears? Ursus 26:40–52.
DIXON, J. D., M. K. OLI, M. C. WOOTEN, T. H. EASON, J. W. MCCOWN AND M. W. CUNNINGHAM. 2007. Genetic consequences of habitat fragmentation and loss: the case of the Florida black bear (Ursus americanus floridanus). Conservation Genetics 8:455–464.
242
DIXON, J. D., M. K. OLI, M. C. WOOTEN, T. H. EASON, J. W. MCCOWN AND D. PAETKAU. 2006. Effectiveness of a regional corridor in connecting two Florida black bear populations. Conservation Biology 20:155–162.
DOBEY, S., D. V. MASTERS, B. K. SCHEICK, J. D. CLARK, M. R. PELTON AND M. E. SUNQUIST. 2005. Ecology of Florida black bears in the Okefenokee–Osceola ecosystem. Wildlife Monographs 158:1–41.
DRAY, S., M. ROYER-CARENZI AND C. CALENGE. 2010. The exploratory analysis of autocorrelation in animal-movement studies. Ecological Research 25:673–681.
DUQUETTE, J. F. ET AL. 2017. Black bear (Ursus americanus) functional resource selection relative to intraspecific competition and human risk. Canadian Journal of Zoology 95:203–212.
EHLERS, L. P. W., C. J. JOHNSON AND D. R. SEIP. 2014. Movement ecology of wolves across an industrial landscape supporting threatened populations of woodland caribou. Landscape Ecology 29:451–465.
ESRI. 2015. ArcGIS Desktop. Version 10.3. Environmental Systems Research Institute, Redlands, California.
ESRI. 2016. ArcGIS Desktop. Version 10.4. Environmental Systems Research Institute, Redlands, California.
EVANS, M. J., J. E. HAWLEY, P. W. REGO AND T. A. G. RITTENHOUSE. 2014. Exurban land use facilitates human-black bear conflicts. Journal of Wildlife Management 78:1477–1485.
FAABORG, J. ET AL. 2010. Conserving migratory land birds in the New World : do we know enough? Ecological Applications 20:398–418.
FAHRIG, L. 1997. Relative effects of habitat loss and fragmentation on population extinction. The Journal of Wildlife Management 61:603–610.
FAHRIG, L. 2007. Non-optimal animal movement in human-altered landscapes. Functional Ecology 21:1003–1015.
FAHRIG, L. AND T. RYTWINSKI. 2009. Effects of roads on animal abundance: An emperical review and synthesis. Ecology and Society 14:21–41.
FECSKE, D., R. BARRY AND F. PRECHT. 2002. Habitat use by female black bears in western Maryland. Southeastern Naturalist 1:77–92.
FIEBERG, J., J. MATTHIOPOULOS, M. HEBBLEWHITE, M. S. BOYCE AND J. L. FRAIR. 2010. Correlation and studies of habitat selection: problem, red herring or opportunity? Philosophical transactions of the Royal Society of London. Series B, Biological sciences 365:2233–2244.
FLAGSTAD, Ø. ET AL. 2004. Colonization history and noninvasive monitoring of a reestablished wolverine population. Conservation Biology 18:676–688.
FLATHER, C. H. AND M. BEVERS. 2002. Patchy reaction-diffusion and population abundance: The relative importance of habitat amount and arrangement. The American Naturalist 159:40–56.
243
FLETCHER, R. J. AND R. L. HUTTO. 2008. Partitioning the multi-scale effects of human activity on the occurrence of riparian forest birds. Landscape Ecology 23:727–739.
FLORIDA FISH AND WILDLIFE CONSERVATION COMMISSION. 2012. Florida black bear management plan. Florida Fish and Wildlife Conservation Commission, Tallahassee, FL.
FOLEY, J. A. ET AL. 2005. Global consequences of land use. Science 309:570–574.
FORESTER, J. D., H. K. IM AND P. J. RATHOUZ. 2009. Accounting for animal movement in estimation of resource selection functions: sampling and data analysis. Ecology 90:3554–3565.
FORMAN, R. T. T. AND L. E. ALEXANDER. 1998. Roads and Their Major Ecological Effects. Annual Review of Ecology and Systematics 29:207–231.
FORTIN, D., J. M. MORALES AND M. S. BOYCE. 2005. Elk winter foraging at fine scale in Yellowstone National Park. Oecologia 145:335–343.
FRANKE, A., T. CAELLI AND R. J. HUDSON. 2004. Analysis of movements and behavior of caribou (Rangifer tarandus) using hidden Markov models. Ecological Modelling 173:259–270.
FRANKE, A., T. CAELLI, G. KUZYK AND R. J. HUDSON. 2006. Prediction of wolf (Canis lupus) kill-sites using hidden Markov models. Ecological Modelling 197:237–246.
FRARY, V. J., J. DUCHAMP, D. S. MAEHR AND J. L. LARKIN. 2011. Density and distribution of a colonizing front of the American black bear Ursus americanus. Wildlife Biology 17:404–416.
FREEMARK, K. E., J. B. DUNNING, S. J. HEJL AND J. R. PROBST. 1995. A landscape ecology perspective for research, conservation, and management. Pp. 381–427inPopulation ecology and conservation of neotropical migrant birds(T. E. Martin & D. M. Finch, eds.). Oxford University Press, New York, New York.
FRETWELL, S. D. 1972. Populations in a seasonal environment. Monographs in Population Biology 5.
FRYXELL, J. M. ET AL. 2008. Multiple movement modes by large herbivores at multiple spatiotemporal scales. Proceedings of the National Academy of Sciences of the United States of America 105:19114–9.
GAINES, W. L., A. L. LYONS, J. F. LEHMKUHL AND K. J. RAEDEKE. 2005. Landscape evaluation of female black bear habitat effectiveness and capability in the North Cascades, Washington. Biological Conservation 125:411–425.
GARRISON, E. P., J. W. MCCOWN, M. A. BARRETT AND M. K. OLI. 2012. Denning ecology of Florida black bears in north-central Florida. Southeastern Naturalist 11:517–528.
GARRISON, E. P., J. W. MCCOWN AND M. K. OLI. 2007. Reproductive Ecology and Cub Survival of Florida Black Bears. Journal of Wildlife Management 71:720–727.
GARSHELIS, D. L. 2000. Delusions in habitat evaluation: Measuring use, selection, and importance. Pp. 112–164inResearch techniques in animal ecology: Controversies and consequences.(L. Boitani & T. K. Fuller, eds.). Columbia University Press, New
244
York, NY, USA.
GARSHELIS, D. L. AND H. HRISTIENKO. 2006. State and provincial estimates of American black bear numbers versus assessments of population trend State and provincial estimates of American black bear numbers versus assessments of population trend. Ursus 17:1–7.
GARSHELIS, D. L. AND M. PELTON. 1980a. Garshelis thesis: movement ecology and activity behavior of black bears in the Great Smokey Moutains National Park.
GARSHELIS, D. L. AND M. R. PELTON. 1980b. Activity of black bears in the Great Smoky Mountains National Park. Journal of Mammalogy 61:8–19.
GARSHELIS, D. L. AND M. R. PELTON. 1981. Movements of black bears in the Great Smoky Mountains. Journal of Wildlife Management 45:912–925.
GARSHELIS, D. L., H. B. QUIGLEY, C. R. VILLARRUBIA AND M. R. PELTON. 1983. Diel movements of black bears in the southern Appalachians. International Conference on Bear Research and Management 5:11–19.
GEORGE, T. L. AND S. ZACK. 2001. Spatial and temporal considerations in restoring habitat for wildlife. Resoration Ecology 9:272–279.
GILLIES, C. S. ET AL. 2006. Application of random effects to the study of resource selection by animals. Journal of Animal Ecology 75:887–898.
GITTLEMAN, J. L. AND P. H. HARVEY. 1982. Carnivore home-range size, metabolic needs and ecology. Behavioral Ecology and Sociobiology 10:57–63.
GODVIK, I. M. R., L. E. LOE, J. O. VIK, V. VEIBERG, R. LANGVATN AND A. MYSTERUD. 2009. Temporal scales, trade-offs, and functional responses in red deer habitat selection. Ecology 90:699–710.
GOMPPER, M. E. 2015. Carnivore coexistence: America’s recovery. Science 347:382–383.
GOSWAMI, V. R., K. MEDHI, J. D. NICHOLS AND M. K. OLI. 2015. Mechanistic understanding of human-wildlife conflict through a novel application of dynamic occupancy models. Conservation Biology 29:1100–1110.
GRAHAM, M. H. 2003. Confronting multicollinearity in ecological multiple regression. Ecology 84:2809–2815.
GREGORY, C. J., R. R. CARTHY AND L. G. PEARLSTINE. 2006. Survey and monitoring of species at risk at Camp Blanding Training Site, Northeastern Florida. Southeastern Naturalist 5:473–498.
GURARIE, E., C. BRACIS, M. DELGADO, T. D. MECKLEY, I. KOJOLA AND C. M. WAGNER. 2015. What is the animal doing? Tools for exploring behavioral structure in animal movements. Journal of Animal Ecology:n/a-n/a.
GURARIE, E. AND O. OVASKAINEN. 2011. Characteristic spatial and temporal scales unify models of animal movement. The American Naturalist 178:113–123.
GURARIE, E. AND O. OVASKAINEN. 2013. Towards a general formalization of encounter rates in ecology. Theoretical Ecology 6:189–202.
245
GUTHRIE, J. M. 2012. Modeling movement behavior and road crossing in the black bear of south Central Florida. M.S. thesis, University of Kentucky. Lexington, Kentucky.
HADDAD, N. M. 1999. Corridor use predicted from behaviors at habitat boundaries. The American Naturalist 153:215–227.
HAGEN, S. B., A. KOPATZ, J. ASPI, I. KOJOLA AND H. G. EIKEN. 2015. Evidence of rapid change in genetic structure and diversity during range expansion in a recovering large terrestrial carnivore. Proceedings of the Royal Society B: Biological Sciences 282:20150092.
HANSEN, C. P. ET AL. 2016. Microsite selection and survival of greater sage-grouse nests in South-Central Wyoming. Journal of Wildlife Management 80:862–876.
HELLGREN, E. C. 1993. Status, distribution, and summer food habits of black bears in Big Bend National Park. The Southwestern Naturalist 38:77–80.
HELLGREN, E. C. AND D. S. MAEHR. 1992. Habitat fragmentation and black bears in the eastern United States. Proceedings of the Eastern Black Bear Workshop 11:154–165.
HELLGREN, E. C. AND M. R. VAUGHAN. 1988. Seasonal food habits of black bears in Great Dismal Swamp, Virginia – North Carolina. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 42:295–305.
HELLGREN, E. C. AND M. R. VAUGHAN. 1989. Demographic Analysis of a Black Bear Population in the Great Dismal Swamp. The Journal of Wildlife Management 53:969–977.
HELLGREN, E. C. AND M. R. VAUGHAN. 1990. Range dynamics of black bears in Great Dismal Swamp, Virginia–North Carolina. Proceedings of the Annual Conference Southeastern Association of Fish and Wildlife Agencies 44:268–278.
HELLGREN, E. C. AND M. R. VAUGHAN. 1994. Conservation and management of isolated black bear populations in the southeastern coastal plain of the United States. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 48.
HELLGREN, E. C., M. R. VAUGHAN AND R. L. KIRKPATRICK. 1989. Seasonal patterns in physiology and nutrition of black bears in Great Dismal Swamp, Virginia - North Carolina. Canadian Journal of Zoology 67:1837–1850.
HELLGREN, E. C., M. R. VAUGHAN AND D. F. STAUFFER. 1991. Macrohabitat use by black bears in a southeastern wetland. Journal of Wildlife Management 55:442–448.
HENLE, K., D. B. LINDENMAYER, C. R. MARGULES, D. A. SAUNDERS AND C. WISSEL. 2004. Species survival in fragmented landscapes: Where are we now? Biodiversity and Conservation 13:1–8.
HERTEL, A. G. ET AL. 2016. Bears and berries: species-specific selective foraging on a patchily distributed food resource in a human-altered landscape. Behavioral Ecology and Sociobiology:831–842.
HIJMANS, R. J. 2015. raster: Geographic data analysis and modeling. R package version
246
2.4-20. <https://CRAN.R-project.org/package=raster>.
HILTY, J. A. AND A. M. MERENLENDER. 2004. Use of riparian corridors and vineyards by mammalian predators in Northern California. Conservation Biology 18:126–135.
HOCTOR, T. S., M. H. CARR AND P. D. ZWICK. 2000. Identifying a linked reserve system using a regional landscape approach: The Florida Ecological Network. Conservation Biology 14:984–1000.
HODGSON, J. A., A. MOILANEN, B. A. WINTLE AND C. D. THOMAS. 2011. Habitat area, quality and connectivity: Striking the balance for efficient conservation. Journal of Applied Ecology 48:148–152.
HOLDEREGGER, R. AND M. DI GIULIO. 2010. The genetic effects of roads: a review of empirical evidence. Basic and Applied Ecology 11:522–531.
HORNER, M. A. AND R. A. POWELL. 1990. Internal structure of home ranges of black bears and analyses of home-range overlap. Journal of Mammalogy 71:402–410.
HOSMER JR., D. W., S. LEMESHOW AND R. X. STURDIVANT. 2013. Applied logistic regression, 3rd edition. P. in. Wiley & Sons, Hoboken, New Jersey.
HOSTETLER, J. A. ET AL. 2009. Demographic consequences of anthropogenic influences: Florida black bears in north-central Florida. Biological Conservation 142:2456–2463.
HRISTIENKO, H. AND J. E. MCDONALD. 2007. Going into the 21st century: A perspective on trends and controversies in the management of the American black bear. Ursus 18:72–88.
INMAN, R. M. AND M. R. PELTON. 2011. Energetic production by soft and hard mast foods of American black bears in the Smoky Mountains. Ursus 13:57–68.
JACHOWSKI, D. S., R. A. MONTGOMERY, R. SLOTOW AND J. J. MILLSPAUGH. 2013. Unravelling complex associations between physiological state and movement of African elephants. Functional Ecology 27:1166–1175.
JOHNSON, C. AND S. NIELSEN. 2006. Resource selection functions based on use-availability data: theoretical motivation and evaluation methods. Journal of Wildlife Management 70:347–357.
JOHNSON, D. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65–71.
JOHNSON, H. E. ET AL. 2015. Shifting perceptions of risk and reward: Dynamic selection for human development by black bears in the western United States. Biological Conservation 187:164–172.
JONSEN, I. D., A. M. RANSOM AND J. M. FLEMMING. 2003. Meta-analysis of animal movement using state-space models. Ecology 84:3055–3063.
KAREIVA, P. M. AND N. SHIGESADA. 1983. Analyzing insect movement as a correlated random walk. Oecologia 56:234–238.
KARELUS, D. L., J. W. MCCOWN, B. K. SCHEICK, M. VAN DE KERK, B. M. BOLKER AND M. K. OLI. 2017. Effects of environmental factors and landscape features on movement
247
patterns of Florida black bears. Journal of Mammalogy 98:1463–1478.
KARELUS, D. L., J. W. MCCOWN, B. K. SCHEICK, M. VAN DE KERK AND M. K. OLI. 2016. Home ranges and habitat selection by black bears in a newly colonized population in Florida. Southeastern Naturalist 15:346–364.
KAUHALA, K., K. HOLMALA AND J. SCHREGEL. 2007. Seasonal activity patterns and movements of the raccoon dog, a vector of diseases and parasites, in southern Finland. Mammalian Biology 72:342–353.
VAN DE KERK, M. ET AL. 2014. Hidden semi-Markov models reveal multiphasic movement of the endangered Florida panther. The Journal of Animal Ecology 84:576–585.
KERTSON, B. N., R. D. SPENCER, J. M. MARZLUFF, J. HEPINSTALL-CYMERMAN AND C. E. GRUE. 2011. Cougar space use and movements in the wildland – urban landscape of western Washington. Ecological Applications 21:2866–2881.
KEUROGHLIAN, A. AND D. P. EATON. 2008. Importance of rare habitats and riparian zones in a tropical forest fragment: preferential use by Tayassu pecari, a wide-ranging frugivore. Journal of Zoology 275:283–293.
KIE, J. G. 2013. A rule-based ad hoc method for selecting a bandwidth in kernel home-range analyses. Animal Biotelemetry 1:13.
KLAR, N. ET AL. 2008. Habitat selection models for European wildcat conservation. Biological Conservation 141:308–319.
KNIOWSKI, A. B. AND S. D. GEHRT. 2014. Home range and habitat selection of the Indiana bat in an agricultural landscape. Journal of Wildlife Management 78:503–512.
KOZAKAI, C. ET AL. 2013. Fluctuation of daily activity time budgets of Japanese black bears: relationship to sex, reproductive status, and hard-mast availability. Journal of Mammalogy 94:351–360.
KRANSTAUBER, B., R. KAYS, S. D. LAPOINT, M. WIKELSKI AND K. SAFI. 2012. A dynamic Brownian bridge movement model to estimate utilization distributions for heterogeneous animal movement. The Journal of animal ecology 81:738–46.
KRANSTAUBER, B. AND M. SMOLLA. 2016. move: visualizing and analyzing animal track data. R package version 2.1.0. https://CRAN.R-project.org/package=move.
LACHER, I. AND M. L. WILKERSON. 2014. Wildlife connectivity approaches and best practices in U.S. State wildlife action plans. Conservation Biology 28:13–21.
LAIDRE, K. L., E. W. BORN, E. GURARIE, Ø. WIIG, R. DIETZ AND H. STERN. 2013. Females roam while males patrol: divergence in breeding season movements of pack-ice polar bears (Ursus maritimus). Proceedings of the Royal Society B: Biological Sciences 280: 20122.
LANGROCK, R., R. KING, J. MATTHIOPOULOS, L. THOMAS, D. FORTIN AND J. M. MORALES. 2012. Flexible and practical modeling of animal telemetry data: hidden Markov models and extensions. Ecology 93:2336–42.
LAPOINT, S., N. BALKENHOL, J. HALE, J. SADLER AND R. VAN DER REE. 2015. Ecological connectivity research in urban areas. Functional Ecology 29:868–878.
248
LARIVIERE, S., J. HUOT AND C. SAMSON. 1994. Daily activity patterns of female black bears in a northern mixed-forest environment. Journal of Mammalogy 75:613–620.
LARKIN, J. L., D. S. MAEHR, T. S. HOCTOR, M. A. ORLANDO AND K. WHITNEY. 2004. Landscape linkages and conservation planning for the black bear in west-central Florida. Animal Conservation 7:23–34.
LARUE, M. A. ET AL. 2012. Cougars are recolonizing the Midwest: Analysis of cougar confirmations during 1990-2008. Journal of Wildlife Management 76:1364–1369.
LATHAM, A. D. M., M. C. LATHAM, M. S. BOYCE AND S. BOUTIN. 2011. Movement responses by wolves to industrial linear features and their effect on woodland caribou in northeastern Alberta. Ecological Applications 21:2854–2865.
LATOMBE, G., L. PARROTT, M. BASILLE AND D. FORTIN. 2014. Uniting statistical and individual-based approaches for animal movement modelling. PLoS ONE 9.
LAVER, P. N. AND M. J. KELLY. 2008. A critical review of home range studies. Journal of Wildlife Management 72:290–298.
LEBLOND, M., C. DUSSAULT AND J. OUELLET. 2010. What drives fine-scale movements of large herbivores? A case study using moose. Ecography 33:1102–1112.
LEBLOND, M., C. DUSSAULT AND J. P. OUELLET. 2013. Avoidance of roads by large herbivores and its relation to disturbance intensity. Journal of Zoology 289:32–40.
LEE, D. J. AND M. R. VAUGHAN. 2003. Dispersal movements by subadult American black bears in Virginia. Ursus 14:162–170.
LEES, A. C. AND C. A. PERES. 2008. Conservation value of remnant riparian forest corridors of varying quality for Amazonian birds and mammals. Conservation Biology 22:439–449.
LEMMON, P. E. 1956. A spherical densiometer for estimating forest overstory density. Forest Science 2:314–320.
LEWIS, D., S. BARUCH-MORDO, K. WILSON, S. BRECK, J. MAO AND J. BRODERICK. 2015. Foraging ecology of black bears in urban environments: guidance for human-bear conflict mitigation. Ecosphere 6:1–18.
LEWIS, J. S. AND J. L. RACHLOW. 2011. Activity patterns of black bears in relation to sex, season, and daily movement rates. Western North American Naturalist 71:388–395.
LI, M. AND B. M. BOLKER. 2017. Incorporating periodic variability in hidden Markov models for animal movement. Movement Ecology 5:1–12.
LIMA, S. L. AND P. A. ZOLLNER. 1996. Towards a behavioral ecology of ecological landscapes. Trends in Ecology and Evolution 11:131–135.
LINDZEY, F. G. AND C. E. MESLOW. 1977. Home range and habitat use by black bears in southwestern Washington. Journal of Wildlife Management 41:413–425.
LINNELL, J., J. E. SWENSON AND R. ANDERSEN. 2001. Predators and people: Conservation of large carnivores is possible at high human densities if management policy is favourable. Animal Conservation 4:345–349.
249
LONGLAND, W. S. AND M. V PRICE. 1991. Direct observations of owls and Heteromyid rodents: can predation risk explain microhabitat use? Ecology 72:2261–2273.
LYONS, A. J. 2005. Activity patterns of urban American black bears in the San Gabriel Mountains of southern California. Ursus 16:255–262.
LYONS, A. L., W. L. GAINES AND C. SERVHEEN. 2003. Black bear resource selection in the northeast Cascades, Washington. Biological Conservation 113:55–62.
MAEHR, D. S. AND J. R. BRADY. 1984. Food habits of Florida black bears. Journal of Wildlife Management 48:230–235.
MAEHR, D. S., T. S. HOCTOR, L. J. QUINN AND J. S. SMITH. 2001. Black bear habitat management guidelines for Florida. Technical report no. 17. Florida Fish and Wildlife Conservation Commission. Tallahassee, FL, USA. 83 pp.
MAGLE, S. B., V. M. HUNT, M. VERNON AND K. R. CROOKS. 2012. Urban wildlife research: Past, present, and future. Biological Conservation 155:23–32.
MAGNUSSON, A. ET AL. 2017. glmmTMB: generalized linear mixed models using Template Model Builder. R package version 0.0.3.
VAN MANEN, F. T., M. F. MCCOLLISTER, J. M. NICHOLSON, L. M. THOMPSON, J. L. KINDALL
AND M. D. JONES. 2012. Short-term impacts of a 4-lane highway on American black bears in Eastern North Carolina. Wildlife Monographs 181:1–35.
MANLY, B. F. J. AND J. A. N. ALBERTO. 2017. Multivariate statistical methods: a primer. P. in. Fourth. Taylor and Francis Group, Boca Raton, Florida.
MANLY, B. F. J., L. L. MCDONALD, D. L. THOMAS, T. L. MCDONALD AND W. P. ERICKSON. 2002. Resource selection by animals: statistical design and analysis for field studies. P. in. 2nd edition. Kluwer Academic Publishers, Boston.
MARCELLI, M., L. POLEDNÍK, K. POLEDNÍKOVÁ AND R. FUSILLO. 2012. Land use drivers of species re-expansion: Inferring colonization dynamics in Eurasian otters. Diversity and Distributions 18:1001–1012.
MARTIN, J. ET AL. 2013. Reciprocal modulation of internal and external factors determines individual movements. Journal of Animal Ecology 82:290–300.
MARTIN, J., C. CALENGE, P.-Y. QUENETTE AND D. ALLAINÉ. 2008. Importance of movement constraints in habitat selection studies. Ecological Modelling 213:257–262.
MASTERS, D. V. 2002. Fine-scale habitat use, activity, and movements of female black bears in north-central Florida. M.S. thesis, University of Florida. Gainesville, Florida.
MATTHEWS, S. M., J. J. BEECHAM, H. QUIGLEY, S. S. GREENLEAF AND H. M. LEITHEAD. 2006. Activity patterns of American black bears in Yosemite National Park. Ursus 17:30–40.
MAY, R., J. VAN DIJK, A. LANDA, R. ANDERSEN AND R. ANDERSEN. 2010. Spatio-temporal ranging behaviour and its relevance to foraging strategies in wide-ranging wolverines. Ecological Modelling 221:936–943.
MCCANN, N. P., P. A. ZOLLNER AND J. H. GILBERT. 2017. Temporal scaling in analysis of
250
animal activity. Ecography 40:1–9.
MCCLENNEN, N., R. R. WIGGLESWORTH, S. H. ANDERSON AND D. G. WACHOB. 2001. The effect of suburban and agricultural development on the activity patterns of coyotes (Canis latrans). American Midland Naturalist 146:27–36.
MCCOWN, J. W., P. KUBILIS, T. H. EASON AND B. K. SCHEICK. 2009. Effect of traffic volume on American black bears in central Florida, USA. Ursus 20:39–46.
MCCOWN, J. W., P. KUBILIS, T. EASON AND B. K. SCHEICK. 2004. Black bear movements and habitat use relative to roads in Ocala National Forest. Final report. Florida Department of Transportation contract BD-016. Florida Fish and Wildlife Conservation Commission, Tallahassee, FL.
MCGARIGAL, K., S. A. CUSHMAN AND E. ENE. 2015. Fragstats 4.2.1: Spatial pattern analysis program for categorical and continous maps. University of Massachusetts, Amherst, Massachusetts.
MCGREER, M. T. ET AL. 2015. Selection for forage and avoidance of risk by woodland caribou (Rangifer tarandus caribou) at coarse and local scales. Ecosphere 6:1–11.
MCKELLAR, A. E., R. LANGROCK, J. R. WALTERS AND D. C. KESLER. 2014. Using mixed hidden Markov models to examine behavioral states in a cooperatively breeding bird. Behavioral Ecology 26:148–157.
MELBOURNE, B. A. ET AL. 2004. Species survival in fragmented landscapes: Where to from here? Biodiversity and Conservation 13:275–284.
MERKLE, J. A., H. S. ROBINSON, P. R. KRAUSMAN AND P. ALABACK. 2013. Food availability and foraging near human developments by black bears. Journal of Mammalogy 94:378–385.
MICHELOT, T., R. LANGROCK AND T. A. PATTERSON. 2016. moveHMM: an R package for the statistical modelling of animal movement data using hidden Markov models. Methods in Ecology and Evolution 7:1308–1315.
MITCHELL, K. 2015. Quantitative analysis by the point-centered quarter method. arXiv:1010.3303v2 [q-bio.QM]. Hobart and William Smith Colleges, Geneva, NY.
MITCHELL, M. S. AND R. A. POWELL. 2007. Optimal use of resources structures home ranges and spatial distribution of black bears. Animal Behaviour 74:219–230.
MITCHELL, M. S. AND R. A. POWELL. 2008. Estimated home ranges can misrepresent habitat relationships on patchy landscapes. Ecological Modelling 216:409–414.
MITCHELL, M. S. AND R. A. POWELL. 2012. Foraging optimally for home ranges. Journal of Mammalogy 93:917–928.
MLADENOFF, D. J., T. A. SICKLEY AND A. P. WYDEVEN. 1999. Predicting gray wolf landscape recolonization: Logistic regression models vs. new field data. Ecological Applications 9:37–44.
MOHR, C. 1947. Table of equivalent populations of North American small mammals. American Midland Naturalist 37:223–249.
MOORCROFT, P. R. 2012. Mechanistic approaches to understanding and predicting
251
mammalian space use: recent advances, future directions. Journal of Mammalogy 93:903–916.
MOORCROFT, P. R. AND A. BARNETT. 2008. Mechanistic Home Range Models and Resource Selection Analysis : A Reconciliation and Unification. Ecology 89:1112–1119.
VAN MOORTER, B., C. M. ROLANDSEN, M. BASILLE AND J.-M. GAILLARD. 2016. Movement is the glue connecting home ranges and habitat selection. Journal of Animal Ecology:21–31.
MORALES, J. M. ET AL. 2010. Building the bridge between animal movement and population dynamics. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences 365:2289–2301.
MORALES, J. M. AND S. P. ELLNER. 2002. Scaling up animal movements in heterogeneous landscapes: the importance of behavior. Ecology 83:2240–2247.
MORRIS, D. W. 1992. Scales and costs of habitat selection in heterogenous landscapes. Evolutionary Ecology 6:412–432.
MORRISON, C. D., M. S. BOYCE, S. E. NIELSEN AND M. M. BACON. 2014. Habitat Selection of a Re-Colonized Cougar Population in Response to Seasonal Fluctuations of Human Activity. Journal of Wildlife Management 78:1394–1403.
MOSNIER, A., J.-P. OUELLET AND R. COURTOIS. 2008. Black bear adaptation to low productivity in the boreal forest. Ecoscience 15:485–497.
MOYER, M. A., J. W. MCCOWN AND M. K. OLI. 2007. Factors influencing home range size of female Florida black bears. Journal of Mammalogy 88:468–476.
MOYER, M. A., J. W. MCCOWN AND M. K. OLI. 2008. Scale-dependent habitat selection by female Florida black bears in Ocala National Forest, Florida. Southeastern Naturalist 7:111–124.
MUELLER-DOMBOIS, D. AND H. ELLENBERG. 1974. The count-plot method and plotless sampling techniques. Pp. 158–159inAims and methods of vegetation ecology. Wiley & Sons, New York.
NAIMAN, R. J., H. DECAMPS AND M. POLLOCK. 1993. The role of riparian corridors in maintaining regional biodiversity. Ecological Applications 3:209–212.
NAKAGAWA, S. AND H. SCHIELZETH. 2013. A general and simple method for obtaining R2
from generalized linear mixed-effects models. Methods in Ecology and Evolution 4:133–142.
NATHAN, R. ET AL. 2008. A movement ecology paradigm for unifying organismal movement research. Proceedings of the National Academy of Sciences of the United States of America 105:19052–19059.
NATIONAL RESEARCH COUNCIL. 2005. Assessing and managing the ecological impacts of paved roads. P. in. National Academies Press, Washington D.C.
NIELSEN, S. E., G. B. STENHOUSE AND M. S. BOYCE. 2006. A habitat-based framework for grizzly bear conservation in Alberta. Biological Conservation 130:217–229.
252
NOSS, R., H. QUIGLEY, M. G. HORNOCKER, T. MERRILL AND P. C. PAQUET. 1996. Conservation biology and carnivore conservation in the Rocky Mountains. Conservation Biology 10:949–963.
NOUVELLET, P., J. P. BACON AND D. WAXMAN. 2009. Fundamental insights into the random movement of animals from a single distance-related statistic. American Naturalist 174:506–514.
NOYCE, K. V. AND D. L. GARSHELIS. 2011. Seasonal migrations of black bears (Ursus americanus): causes and consequences. Behavioral Ecology and Sociobiology 65:823–835.
OLI, M. K., H. A. JACOBSON AND B. D. LEOPOLD. 2002. Pattern of space use by female black bears in the White River National Wildlife Refuge, Arkansas, USA. Journal for Nature Conservation 10:87–93.
ONORATO, D. P., E. C. HELLGREN, R. A. VAN DEN BUSSCHE AND D. L. DOAN-CRIDER. 2004. Phylogeographic patterns within a metapopulation of black bears (Ursus americanus) in the American Southwest. Journal of Mammalogy 85:140–147.
ONORATO, D. P., E. C. HELLGREN, F. S. MITCHELL AND J. R. SKILES JR. 2003. Home range and habitat use of American black bears on a desert montane island in Texas. Ursus 14:120–129.
ORDIZ, A., O. G. STØEN, S. SÆBØ, J. KINDBERG, M. DELIBES AND J. E. SWENSON. 2012. Do bears know they are being hunted? Biological Conservation 152:21–28.
ORDIZ, A., J. E. SWENSON, A. ORDIZ, O. S. MIGUEL AND J. E. SWENSON. 2011. Predators or prey? Spatio-temporal discrimination of human-derived risk by brown bears. Oecologia 166:59–67.
OWEN-SMITH, N., J. M. FRYXELL AND E. H. MERRILL. 2010. Foraging theory upscaled: the behavioural ecology of herbivore movement. Philosophical Transactions of the Royal Society B: Biological Sciences 365:2267–2278.
PATTERSON, T. A., M. BASSON, M. V. BRAVINGTON AND J. S. GUNN. 2009. Classifying movement behaviour in relation to environmental conditions using hidden Markov models. Journal of Animal Ecology 78:1113–23.
PELTON, M. ET AL. 1999. American black bear conservation action plan. Pp. 144–156inBears: Status Survey and Conservation Action Plan(C. Servheen, S. Herrero & B. Peyton, eds.). 4th edition. IUCN, Gland, Switzerland & Cambridge, UK.
PELTON, M. R. 2003. Black Bears (Ursus Americanus). Pp. 547–556inWild mammals of North America(G. A. Feldhamer, B. C. Thompson & J. A. Chapman, eds.). 2nd edition. Johns Hopkins University Press, Baltimore, MD.
PINHEIRO, J., D. BATES, S. DEBROY, D. SARKAR AND R CORE TEAM. 2015. nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-122.
PLETSCHER, D. H., R. R. REAM, D. K. BOYD, M. W. FAIRCHILD AND K. E. KUNKEL. 1997. Population dynamics of a recolonizing wolf population. Journal of Wildlife Management 61:459–465.
253
POHLE, J., R. LANGROCK, F. M. VAN BEEST AND N. M. SCHMIDT. 2017. Selecting the number of states in hidden Markov models: pragmatic solutions illustrated using animal movement. Journal of Agricultural, Biological and Environmental Statistics:https://doi.org/10.1007/s13253-017-0283-8.
POWELL, R. A., J. W. ZIMMERMAN AND D. E. SEAMAN. 1997. Ecology and behaviour of North American black bears: Home ranges, habitat and social organization. P. in. Chapman and Hall, London, UK.
PRICE, M. V. 1978. The role of microhabitat in structuring desert rodent communities. Ecology 59:910–921.
R CORE TEAM. 2013. R: A language and environment for statistical computing. Available online at <https://www.R-project.org/>. R Foundation for Statistical Computing, Vienna, Austria.
R CORE TEAM. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
R CORE TEAM. <HTTP://WWW.R-PROJECT.ORG/>. 2016. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
RAYL, N. D., T. K. FULLER, J. F. ORGAN, J. E. MCDONALD, R. D. OTTO AND S. P. MAHONEY. 2014. Den abandonment and transitional day bed use by black bears (Ursus americanus) in Newfoundland. Wildlife Biology 20:222–228.
REDNER, J. AND S. SRINIVASAN. 2014. Florida Vegetation and Land Cover 2014. Final Report. Fish and Wildlife Research Institute grant #6207. Florida Fish and Wildlife Conservation Commission, Tallahassee, FL.
REVILLA, E. AND T. WIEGAND. 2008. Individual movement behavior, matrix heterogeneity, and the dynamics of spatially structured populations. Proceedings of the National Academy of Sciences of the United States of America 105:19120–19125.
REYNOLDS-HOGLAND, M. J. AND M. S. MITCHELL. 2007. Effects of roads on habitat quality for bears in the southern Appalachians: A long-term study. Journal of Mammalogy 88:1050–1061.
REYNOLDS-HOGLAND, M. J., L. B. PACIFICI AND M. S. MITCHELL. 2007. Linking resources with demography to understand resource limitation for bears. Journal of Applied Ecology 44:1166–1175.
RIPPLE, W. J. ET AL. 2014. Status and ecological effects of the world ’ s largest carnivores. Science (New York, N.Y.) 343:1–11.
ROEVER, C. L., M. S. BOYCE AND G. B. STENHOUSE. 2010. Grizzly bear movements relative to roads: application of step selection functions. Ecography 33:1113–1122.
ROGERS, L. L. 1987. Effects of food supply and kinship on social behavior, movements, and population growth of black bears in northeastern Minnesota. Wildlife Monographs 97:1–72.
RUSSELL, D. J. F. ET AL. 2013. Uncovering the links between foraging and breeding regions in a highly mobile mammal. Journal of Applied Ecology 50:499–509.
254
RYAN, C. W., J. C. PACK, W. K. IGO AND A. BILLINGS. 2007. Influence of mast production on black bear non-hunting mortalities in West Virginia. Ursus 18:46–53.
S-PLUS ORIGINAL BY ULRIC LUND AND R PORT BY CLAUDIO AGOSTINELLI. 2012. CircStats: Circular Statistics, from “Topics in circular Statistics” (2001).
SAAB, V. 1999. Importance of spatial scale to habitat use by breeding birds in riparian forests: a hierarchical analysis. Ecological Applications 9:135–151.
SADEGHPOUR, M. H. AND T. F. GINNETT. 2011. Habitat selection by female American black bears in northern Wisconsin. Ursus 22:159–166.
DOS SANTOS LIMA, D. 2014. Reoccupation of historical areas by the endangered giant river otter Pteronura brasiliensis (Carnivora: Mustelidae) in central Amazonia, Brazil. Mammalia 78:177–184.
SAURA, S., Ö. BODIN AND M. J. FORTIN. 2014. Editor’s choice: Stepping stones are crucial for species’ long-distance dispersal and range expansion through habitat networks. Journal of Applied Ecology 51:171–182.
SAWYER, S. C., C. W. EPPS AND J. S. BRASHARES. 2011. Placing linkages among fragmented habitats: Do least-cost models reflect how animals use landscapes? Journal of Applied Ecology 48:668–678.
SCHEICK, B. K., M. W. CUNNINGHAM, J. W. MCCOWN AND M. A. ORLANDO. 2009. Anchor modification for a foot-hold snare to capture American black bears. Ursus 20:47–49.
SCHEICK, B. K. AND W. MCCOWN. 2014. Geographic distribution of American black bears in North America. Ursus 25:24–33.
SCHICK, R. S. ET AL. 2008. Understanding movement data and movement processes: current and emerging directions. Ecology letters 11:1338–50.
SCHLIEHE-DIECKS, S., P. M. KAPPELER AND R. LANGROCK. 2012. On the application of mixed hidden Markov models to multiple behavioural time series. Interface Focus 2:180–9.
SEAMAN, D. E. AND R. A. POWELL. 1996. An evaluation of the accuracy of kernel density estimators for home range analysis. Ecology 77:2075–2085.
SERVHEEN, C., S. HERRERO AND B. PEYTON. 1999. Status, survey and conservation action plan: Bears. IUCN/SSC Bear Specialist Group. 321pp.
SHEPARD, E. L. C., R. P. WILSON, W. G. REES, E. GRUNDY, S. A. LAMBERTUCCI AND S. B. VOSPER. 2013. Energy landscapes shape animal movement ecology. American Naturalist 182:298–312.
SIKES, R. S. AND THE ANIMAL CARE AND USE COMMITTEE OF THE AMERICAN SOCIETY OF
MAMMALOGISTS. 2016. 2016 guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy 97:663–688.
SIMBERLOFF, D. 1999. Biodiversity and bears: a conservation paradigm shift. Ursus 11:21–27.
SIMEK, S. L., J. L. BELANT, B. W. YOUNG, C. SHROPSHIRE AND B. D. LEOPOLD. 2012. History
255
and status of the American black bear in Mississippi. Ursus 23:159–167.
SINGH, N. J., A. M. ALLEN AND G. ERICSSON. 2016. Quantifying migration behavior using net squared displacement approach: clarifications and caveats. PloS one 11:e0149594.
SMITH, T. R. AND M. R. PELTON. 1990. Home ranges and movements of black bears in a bottomland hardwood forest in Arkansas. International Conference on Bear Research and Management 8:213–218.
SOKAL, R. R. AND F. J. ROHLF. 2011. Biometry: the principles and practice of statistics in biological research. P. in. Fourth. W. H. Freeman and Company, New York, New York.
SOLLMANN, R., B. GARDNER, J. L. BELANT, C. M. WILTON AND J. BERINGER. 2016. Habitat associations in a recolonizing, low-density black bear population. Ecosphere 7:e01406.
SPELLERBERG, I. F. 1998. Ecological effects of roads and traffic: A literature review. Global Ecology and Biogeography Letters 7:317–333.
SPENCER, W. D. 2012. Home ranges and the value of spatial information. Journal of Mammalogy 93:929–947.
SQUIRES, J. R., N. J. DECESARE, L. E. OLSON, J. A. KOLBE, M. HEBBLEWHITE AND S. A. PARKS. 2013. Combining resource selection and movement behavior to predict corridors for Canada lynx at their southern range periphery. Biological Conservation 157:187–195.
STILLFRIED, M., J. L. BELANT, N. J. SVOBODA, D. E. BEYER AND S. KRAMER-SCHADT. 2015. When top predators become prey: black bears alter movement behaviour in response to hunting pressure. Behavioural Processes 120:30–39.
STRATMAN, M., C. ALDEN, M. PELTON AND M. SUNQUIST. 2001. Habitat use by American black bears in the sandhills of Florida. Ursus 12:109–114.
STRATMAN, M. R. 1998. Habitat use and effects of prescribed fire on black bears in northwestern Florida. Master’s Thesis, University of Tennessee, Knoxville, TN. 87 pp.
STRATMAN, M. R. AND M. R. PELTON. 1999. Feeding ecology of black bears in northwest Florida. Florida Field Naturalist 27:95–102.
SWENSON, J. E., F. SANDEGREN AND A. SÖDERBERG. 1998. Geographic expansion of an increasing brown bear population: Evidence for presaturation dispersal. Journal of Animal Ecology 67:819–826.
SWITALSKI, T. A. AND C. R. NELSON. 2011. Efficacy of road removal for restoring wildlife habitat: Black bear in the northern Rocky Mountains, USA. Biological Conservation 144:2666–2673.
TAKAHATA, C., S. E. NIELSEN, A. TAKII AND S. IZUMIYAMA. 2014. Habitat selection of a large carnivore along human-wildlife boundaries in a highly modified landscape. PLoS ONE 9:e86181.
256
TAULMAN, J. F. AND K. G. SMITH. 2004. Home range and habitat selection of southern flying squirrels in fragmented forests. Mammalian Biology 69:11–27.
TEW, T. E., I. A. TODD AND D. W. MACDONALD. 2000. Arable habitat use by wood mice (Apodemus sylvaticus). 2. Microhabitat. Journal of Zoology 250:305–311.
THERNEAU, T. M. 2015. A Package for Survival Analysis in S. R package version 2.38. https://cran.r-project.org/package=survival.
THERNEAU, T. M. AND P. M. GRAMBSCH. 2000. Modeling survival data: extending the Cox model. P. in. Springer, New York, NY, USA.
THURFJELL, H., S. CIUTI AND M. S. BOYCE. 2014. Applications of step-selection functions in ecology and conservation. Movement Ecology 2:26.
TIGAS, L. A., D. H. VAN VUREN AND R. M. SAUVAJOT. 2002. Behavioral responses of bobcats and coyotes to habitat fragmentation and corridors in an urban environment. Biological Conservation 108:299–306.
U.S. CENSUS BUREAU. 2017. United States Census Bureau. <https://www.census.gov/> (21 January 2017).
UNGER, D. E. ET AL. 2013. History and current status of the black bear in Kentucky. Northeastern Naturalist 20:289–308.
VASUDEV, D. AND R. J. FLETCHER. 2015. Incorporating movement behavior into conservation prioritization in fragmented landscapes: an example of western hoolock gibbons in Garo Hills, India. Biological Conservation 181:124–132.
VERDADE, L. M., L. M. ROSALINO, C. GHELER-COSTA, N. M. PEDROSO AND M. C. LYRA-JORGE. 2011. Adaptation of mesocarnivores (Mammalia: Carnivora) to agricultural landscapes in Mediterranean Europe and southeastern Brazil: A trophic perspective. P. in Middle-Sized Carnivores in Agricultural Landscapes.
VISSER, I. 2011. Seven things to remember about hidden Markov models: A tutorial on Markovian models for time series. Journal of Mathematical Psychology 55:403–415.
WARDE, W. AND J. W. PETRANKA. 1981. A correction factor table for missing point-center quarter data. Ecology 62:491–494.
WEAVER, K. M. AND M. R. PELTON. 1994. Denning ecology of black bears in the Tensas River Basin of Louisiana. International Conference on Bear Research and Management 9:427–433.
WHITTINGTON, J., C. C. S. CLAIR AND G. MERCER. 2005. Spatial responses of wolves to roads and trails in mountain valleys. Ecological Applications 15:543–553.
VAN WHY, K. R. 2003. Feasibility of restoring the Louisiana black bear (Ursus americanus luteolus) to portions of their historic range. Masters thesis. Louisiana State University.
WILLEY, C. H. 1974. Aging black bears from first premolar tooth sections. Journal of Wildlife Management 38:97–100.
WILSON, R. P. ET AL. 2013. Turn costs change the value of animal search paths. Ecology
257
Letters 16:1145–1150.
WILTON, C. M., J. L. BELANT AND J. BERINGER. 2014. Distribution of American black bear occurrences and human – bear incidents in Missouri. Ursus 25:53–60.
WOODING, J. B. AND T. S. HARDISKY. 1992a. Denning by black bears in Northcentral Florida. Journal of Mammalogy 73:895–898.
WOODING, J. B. AND T. S. HARDISKY. 1992b. Denning by black bears in northcentral Florida. Journal of Mammalogy 73:895–898.
WOODING, J. B. AND T. S. HARDISKY. 1994. Home range, habitat use, and mortality of black bears in north-central Florida. International Conference on Bear Research and Management 9:349–356.
WOODROFFE, R. 2000. Predators and people: Using human densities to interpret declines of large carnivores. Animal Conservation 3:165–173.
WOODROFFE, R. 2011. Ranging behaviour of African wild dog packs in a human-dominated landscape. Journal of Zoology 283:88–97.
WORTON, B. 1989. Kernel methods for estimating the utilization distribution in home-range studies. Ecology 70:164–168.
ZELLER, K. A., K. MCGARIGAL, S. A. CUSHMAN, P. BEIER, T. W. VICKERS AND W. M. BOYCE. 2016. Using step and path selection functions for estimating resistance to movement: pumas as a case study. Landscape Ecology 31:1319–1335.
ZELLER, K. A., K. MCGARIGAL AND A. R. WHITELEY. 2012. Estimating landscape resistance to movement: a review. Landscape Ecology 27:777–797.
ZIMMERMANN, B., L. NELSON, P. WABAKKEN, H. SAND AND O. LIBERG. 2014. Behavioral responses of wolves to roads: scale-dependent ambivalence. Behavioral Ecology 25:1353–1364.
ZOLLNER, P. A. AND S. L. LIMA. 1999. Search strategies for landscape-level interpatch movements. Ecology 80:1819–1830.
ZUCCHINI, W. AND I. L. MACDONALD. 2016. Hidden markov models for time series: an introduction using R, second edition. P. in. Chapman and Hall/CRC Press, Boca Raton, FL.
ZUUR, A., E. N. IENO, N. WALKER, A. A. SAVELIEV AND G. M. SMITH. 2009. Mixed effects models and extensions in ecology with R. P. in. Springer, New York, New York.
ZWICK, P. D. AND M. H. CARR. 2006. Florida 2060. A population distribution scenario for the State of Florida. Gainesville, Florida.
258
BIOGRAPHICAL SKETCH
Dana Karelus grew up in Florida and received her Bachelor of Science degree in
mechanical engineering from the University of South Florida. She went on to be an
engineer with Space Shuttle Program Operations at Kennedy Space Center.
Throughout this time, she volunteered with several different organizations in the areas
of ecological research, habitat restoration, wildlife management, wildlife rehabilitation,
and eco-education. When the Space Shuttle was retired, she took the opportunity to
change career paths and took classes as a post-baccalaureate at Brevard Community
College and at the University of Central Florida in preparation for applying to graduate
school to study wildlife ecology. She began her PhD in interdisciplinary ecology in the
School of Natural Resources and Environment with a concentration in wildlife ecology
and conservation under the advisement of Dr. Madan Oli at the University of Florida in
August of 2013 and completed her degree in December of 2017.