Towards an energetic landscape: broad-scale accelerometry in woodland caribou

For Review O

nly

Towards an energetic landscape: broad-scale accelerometry

in woodland caribou

Journal: Journal of Animal Ecology

Manuscript ID: JAE-2013-00130.R1

Manuscript Type: Standard Paper

Date Submitted by the Author: n/a

Complete List of Authors: Mosser, Anna; University of Guelph, Integrative Biology Avgar, Tal; University of Guelph, Integrative Biology Brown, Glen; Ontario Ministry of Natural Resources, Ontario Terrestrial Assessment Program Walker, Spencer; University of Guelph, Integrative Biology Fryxell, John; University of Guelph,

Key-words: accelerometry, biotelemetry, energetics, landscape, Rangifer

Journal of Animal Ecology

For Review O

nly

363x241mm (300 x 300 DPI)

Page 1 of 27 Journal of Animal Ecology

For Review O

nly

1

Towards an energetic landscape: 1

broad-scale accelerometry in woodland caribou 2

3

4

5

6

Anna A. Mosser*a, Tal Avgara, Glen S. Brownb, C. Spencer Walkera, and John M. Fryxella 7

8

a Department of Integrative Biology, University of Guelph, Guelph, Canada 9

b Ontario Terrestrial Assessment Program, Ontario Ministry of Natural Resources, 10

Sault Ste. Marie, Canada 11

*Corresponding author: [email protected] 12

Page 2 of 27Journal of Animal Ecology

For Review O

nly

2

Summary (350 words) 13

1. Energetic balance is a central driver of individual survival and population change, yet 14

estimating energetic costs in free and wide ranging animals presents a significant 15

challenge. Animal-borne activity monitors (using accelerometer technology) present a 16

promising method of meeting this challenge and opens new avenues for exploring 17

energetics in natural settings. 18

2. To determine the behaviours and estimated energetic costs associated with a given activity 19

level, three captive reindeer (Rangifer tarandus tarandus) at the Toronto Zoo were fitted 20

with collars and observed for 53 hours. Activity patterns were then measured over thirteen 21

months for 131 free-ranging woodland caribou (R. t. caribou) spanning 450,000 km2 in 22

northern Ontario. The captive study revealed a positive but decelerating relationship 23

between activity level and energetic costs inferred from previous behavioural studies. 24

3. Field-based measures of activity were modelled against individual displacement, 25

vegetation abundance (NDVI), snow depth, and temperature, and the best fit model 26

included all parameters and explained over half of the variation in the data. Individual 27

displacement was positively related to activity levels, suggesting that broad differences in 28

energetic demands are influenced by variation in movement rates. After accounting for 29

displacement, activity was highest at intermediate levels of vegetation abundance, 30

presumably due to foraging behaviour. Snow depth, probably associated with digging for 31

winter forage, moderately increased activity. Activity levels increased significantly at the 32

coldest winter temperatures, suggesting the use of behavioural thermoregulation by 33

caribou. These interpretations of proximate causal factors should be regarded as hypotheses 34

subject to validation under normal field conditions. 35


For Review O

nly

3

4. These results illustrate the landscape characteristics that increase energetic demands for 36

caribou and confirm the great potential for the use of accelerometry in studies of animal 37

energetics. 38

39

40

Key-words 41

accelerometry, biotelemetry, energetics, landscape, Rangifer 42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59


For Review O

nly

4

Introduction 60

An organism’s balance between energy acquisition and expenditure affects its survival and 61

reproduction (Brown et al. 2004); thus ecologists are interested in quantifying patterns of energetic 62

gains and losses. Recent work suggests animal movement trajectories can be strongly influenced 63

by energetic tradeoffs (Shepard et al. 2013). Energetic constraints are expected to determine the 64

extent of a species’ range (Anderson & Jetz 2005; Buckley 2008) and energy deficits are likely 65

associated with population decline. Therefore, examining variation in energy expenditure across 66

environmental gradients, evocatively termed the energetic landscape (Wilson et al. 2011, Shepard 67

et al. 2013), is an important component of understanding and predicting patterns of space use, 68

range shifts or retractions as well as population dynamics. 69

A relatively new, yet promising, method of estimating energetic costs is body 70

accelerometry (Wilson et al. 2006; Halsey et al. 2009a; Gleiss, Wilson, & Shepard 2011; Shepard 71

et al. 2013). Accelerometers, typically integrated into tracking devices attached to an animal, 72

measure body movements along three axes and approximate the animal's locomotion, a primary 73

source of energy expenditure in animals (Alexander 2003). While laboratory research confirms the 74

utility of this method, practical use in field studies remains in development (e.g. Wilson, Quintana, 75

& Hobson 2011; Nathan et al. 2012). In laboratory experiments for a range of species, these 76

acceleration patterns have been calibrated against the rate of oxygen consumption, a commonly 77

used measure of metabolic rate and energy expenditure (Elliott & Davison 1975). These 78

calibrations have reliably yielded positive linear relationships between body acceleration and 79

oxygen consumption (Halsey et al. 2009b). Accelerometry data, however, when measured at a 80

sub-second frequency, present complications for use in field studies due to limitations of on-board 81

data storage and the amount of information that can be transmitted via satellite. Manufacturers of 82

wildlife tracking collars have therefore implemented the use of activity sensors derived from 83


For Review O

nly

5

integrated accelerometry information (Body et al. 2012). Here we evaluate the potential for the 84

use of such activity sensors based on accelerometer scores as a proxy for metabolic costs 85

applicable to free-ranging animals in remote settings. 86

In order to explore the use of this type of accelerometer technology, we paired a field-87

based study of woodland caribou (Rangifer tarandus caribou) with a short captive study of 88

European reindeer (Rangifer tarandus tarandus). Prior to field deployment of accelerometer-89

equipped collars, we conducted an observational study of three captive reindeer fitted with the 90

same collars, to verify the relationships among the simple accelerometer statistic, behaviour and 91

activity level, as well as estimated energy expenditure. 160 woodland caribou in northern Ontario 92

were then fitted with collars for a multi-year study of caribou ranging ecology. The collar 93

accelerometer measurements are reported as total active seconds within a 5 or 20 minute time 94

interval (see methods), allowing for easy transmission via satellite and deployment of collars over 95

multiple years. Here we report on one full year of movement and accelerometry data for 131 96

female caribou ranging across 450,000 km2, a large fraction of the province of Ontario, Canada 97

This study is among the first to report such data for a large number of free-ranging 98

individuals inhabiting a large and varied area and to consider the environmental drivers of 99

variation in energetic costs at a landscape scale (Wilson, Quintana, & Hobson 2011). Our central 100

aim, after verifying the utility of a simple activity measurement in a captive study, is to 101

characterize the broad-scale patterns of variation in individual activity levels and to identify the 102

environmental parameters underlying these patterns. Identifying the factors that increase activity, 103

and thus energetic costs, will be a key component in determining whether energetic constraints are 104

an important cause of decline in Ontario caribou populations (Schaefer 2003; Callaghan, Virc, & 105

Duffe 2011; Festa-Bianchet et al. 2011). 106

107


For Review O

nly

6

Materials and methods 108

Captive study 109

Three individuals from a herd of seven female captive European reindeer at Metropolitan 110

Toronto Zoo (Ontario, Canada) were fitted with collars and observed during a 12-day period from 111

15-Dec-2009 to 4-Jan-2010. Use of collars was pre-approved by a zoo review committee and 112

animal handling was carried out by zoo staff. All observations took place roughly between 8:30 113

am and 4:30 pm. Although this subspecies is slightly smaller than that subsequently studied in the 114

wild, behavioural patterns are otherwise known to be similar. The reindeer were 2, 13, 15 years of 115

age; two of the three were substantially older than typically found in the wild. The observed 116

reindeer had access to ground level grass and hay as well as pellets provided in a trough. Cratering 117

to acquire food was unimportant for the zoo animals, although animals did dig through shallow 118

snow occasionally to forage on grasses. As a result foraging costs due to cratering are not included 119

here. 120

The collars (Telonics Inc., Mesa, Arizona, USA, Gen4 model number TGW-4680) use 121

uGPSi-20 accelerometers to detect movement. Accelerometer measurements were reported as 122

"active seconds", where an active second was recorded if the change in acceleration from one 123

second to the next in any one of three orthogonal dimensions exceeded a threshold as set by the 124

manufacturer (about 0.3 g, or 2.87 m/s2). The sensor is also sensitive to rotation, as this movement 125

also causes a change in the measured acceleration, and rotation angles of 17° to 45° thereby trigger 126

an "active second". This activity sensor falls somewhere between full dynamic acceleration and 127

static acceleration. The virtue of such a measure is that it allows data to be collected and stored 128

over appreciable periods, which could in principle enhance field applications in remote settings. 129


For Review O

nly

7

For this captive study of short duration, a continuous stream of second-by-second data was 130

recorded and stored on the device and then later downloaded to a computer. 131

Observations began in the morning before the reindeer were fed and continued during and 132

after feeding until either the reindeer was recumbent or until the zoo closed for the evening. The 133

following behavioural categories were recorded: lying down, standing, walking, grazing, rubbing 134

antlers, trotting, social interaction, and trough feeding. Data were logged with each change in 135

behavioural category and the total duration of each behavioural bout was recorded (to the second) 136

along with a time stamp. Observations focussed on a single individual at a time and observation 137

periods averaged approximately 20-minutes. We accumulated 21, 13, and 19 hours of observation 138

for the 15, 13, and 2 year-old reindeer, respectively. 139

Accelerometer data were matched, second-by-second, with the behavioural observations. 140

Energetic costs associated with each behaviour category (Boertje 1985a; Fancy & White 1986) 141

were used to assign an energetic cost to each observed second. These data were then summarized 142

over 5 and 20 minute intervals, randomly selected with no overlap from the larger dataset. This 143

allowed comparison with the 5 and 20 minute recording periods of the field-based study. Inverse 144

negative exponential curves were then fitted to the data to describe the relationship between 145

activity level and the estimated energetic expenditure during that interval. 146

147

Free-ranging study 148

During February to April of 2009 and 2010, 160 female woodland caribou across northern 149

Ontario were net-gunned from a helicopter and fitted with GPS telemetry collars of the type 150

described above. Capturing and handling were performed according to Ontario Ministry of Natural 151

Resources animal care regulations. Telemetry collars were distributed in an effort to provide good 152


For Review O

nly

8

geographic representation across the range of prominent vegetation communities inhabited by 153

caribou in Ontario, including upland conifer forests in the west and lowland communities 154

dominated by fens and bogs in the northern and eastern portions of range adjacent to Hudson Bay 155

and James Bay. The study area, determined by the ranging patterns of all individuals, is roughly 156

450,000 km2 (Fig. 2). 157

Each collar recorded GPS coordinates and activity counts every 5 hours for 13 months and 158

transmitted the data via an Argos uplink every 5 days. Four activity intervals were recorded with 159

each location fix, two before and two after the fix, and the total number of active seconds was 160

reported for each. Activity intervals were 5 or 20 minutes, differing by collar, and thus the 161

maximum possible activity counts were 300 or 1200 active seconds. We accordingly normalized 162

the data by dividing by the maximum possible activity level, allowing us to merge data from both 163

collar types. Activity counts differed depending on the interval length, and this is treated 164

appropriately in the statistical analyses (see below). A subset of data including only consecutive 5-165

hour steps was used, and total displacement (in meters) was calculated for each step. Data from the 166

first 24 hours following collaring and the last 24 hours before any mortality were excluded from 167

the dataset, as were any points that were likely due to device error (those indicating an improbable 168

velocity >200 km/5hr and a mean activity count of 0) or fixes that were not associated with at least 169

3 activity intervals. This resulted in a final data set based on 103,000 fixes out of 155,000 possible. 170

Each GPS location was coupled with three environmental variables that are expected to be 171

associated with foraging and thermoregulatory behaviour. We expected foraging behaviour to be 172

associated with vegetation abundance and snow depth. Caribou often feed on lichen and ground 173

level plants in the winter, and must dig through the snow to reach them, a behaviour known as 174

cratering. The amount of available forage was estimated using maps of Normalized Difference 175

Vegetation Index (NDVI) at a 250 meter resolution and 16 day intervals (Land Processes 176


For Review O

nly

9

Distributed Active Archive Center, U.S. Geological Survey - Earth Resources Observation and 177

Science Center, lpdaac.usgs.gov). NDVI values range from -2000 to 10000, with higher values 178

indicating more vegetation. A mean NDVI value was calculated from the area within a 250 meter 179

radius of each point, half the median 5-hour displacement. Thermoregulatory behaviour was 180

expected to be associated with temperature. Maps of snow depth (meters) and temperature (°C), 181

based on real-time weather data and meteorological modelling, were obtained at a 40 km 182

resolution for each 3 hour period (North America Regional Reanalysis dataset DSI-6175, NOAA 183

Operational Model Archive and Distribution System, nomads.ncdc.noaa.gov). Each GPS point 184

was associated with the nearest location and time for these two meteorological variables. 185

Location, activity count, displacement, and environmental measures were averaged for 186

each individual and month. Means were only included if the individual had at least 10 days of 187

activity data within the month (leaving 131 individuals and 1231 individual-months, and between 188

1 and 13 months of data per individual). 189

Mean monthly activity counts (transformed to percent of possible maximum) were 190

modelled against displacement, NDVI, snow depth, and temperature, using a generalized least 191

squares model (gls function; The R Foundation for Statistical Computing, Vienna, Austria, version 192

2.13.1), which included individual and x-y coordinates to account for non-independence and 193

spatial autocorrelation structure in the residuals. All models also included interval type (5 or 20 194

minute) as a parameter, as the mean activity levels differed significantly, with a slightly higher 195

mean for the 5-minute activity intervals. Statistical models considered included a null model 196

(intercept and interval type only), all linear and second-order combinations, and two-way 197

interactions between environmental variables. For simplicity and ease of interpretation, second-198

order terms and two-way interactions were not considered within the same model. Candidate 199

models were compared using Akaike's information criteria (AIC) with the best model having the 200


For Review O

nly

10

highest model weight (ranging from 0 to 1) (Burnham & Anderson 2002). True R2 values cannot 201

be calculated for these models, yet an estimate for the variance explained by the best model can be 202

calculated by comparing the residual variance of the null model to that of the best model (Zheng 203

2000). These values are reported where appropriate. 204

Results 205

Captive study 206

Activity measures for the captive reindeer ranged from 0 to 46% of seconds active within 207

an interval, with a mean and standard deviation of 7.6% ± 8.0. Of all observations, 43% of time 208

was spent lying down, 27% grazing, 15% standing, 8% feeding at trough, 5% walking, 0.4% 209

rubbing antlers, 0.4% in social interactions, and 0.1% of time was spent trotting. Direct 210

observations confirmed that intervals with low activity scores were associated with lower cost 211

behaviours such as lying down, while intervals with higher activity scores were associated with 212

higher cost behaviours such as grazing and walking (Fig. 1A). Energy expenditure, as estimated by 213

the energetic costs of the behaviours observed within each interval, was significantly correlated 214

with the activity measures (Fig. 1B). Non-linear models fit to the data explained 64% of the 215

variance for 5-minute intervals and 67% for 20-minute intervals, and all model coefficients were 216

significant with p≤ 0.0001. 217

218

Free-ranging study 219

For the study of free-ranging woodland caribou, measures of activity ranged from 2.9% to 220

94.8% of maximum possible activity level, with a mean and standard deviation of 17.4% ± 14.9. 221

That is, on average, individuals were active for 17.4% of the seconds within a 5 or 20 minute 222


For Review O

nly

11

activity interval. The mean for 5 minute intervals was 18.0% ±16.5, whereas the mean for 20 223

minute intervals was 17.1% ±13.8. 224

Mean annual activity levels ranged from 14% to 21% and varied across the study range 225

(Fig. 2), with a mean across all individuals of 17.6% ± 3.0. Mean monthly activity levels ranged 226

from 15.4% to 22.0%, with activity levels highest in December and January and variation greatest 227

in the summer months. The best model for monthly mean activity (Table 1) included displacement, 228

NDVI, snow depth, and temperature, with a model weight of 0.49 and an estimated 51% of the 229

variance explained (Table 2). This model included positive linear relationships with displacement 230

and snow depth and second order terms for NDVI and temperature (see Fig. 3 for the shapes of 231

these curves). A model with x- and y-coordinates did not provide a more plausible model, 232

indicating that displacement and local environmental variables captured a large fraction of 233

observed spatial variation. 234

All top ranked models include displacement and a displacement-only model explained 37% 235

of the data variance. A plot of mean monthly activity versus displacement (Fig. 3A) illustrates the 236

strong association between these variables. The scatterplot of monthly activity versus displacement 237

(Fig, 3A) suggests there may be a lower bound to the data distribution, characteristic of so-called 238

polygonal relationships (Scharf et al. 1998). Plotting the residuals after accounting for 239

displacement (Figs 3B-C), illustrates the remaining variance in activity patterns that are associated 240

with the environmental variables. 241

242

Discussion 243

The captive study suggested that activity scores derived from accelerometry were 244

associated with differences in reindeer behavior that have been linked in earlier laboratory studies 245


For Review O

nly

12

with predictable variation in energy expenditure. There was, however, appreciable variance left 246

unexplained, larger than that reported for other published correlates of metabolic rate, such as 247

heart rate (Nilssen et al. 1984; Fancy & White 1986) or O2 consumption (Halsey et al. 2009b). 248

Additional research on meaningful summary statistics for accelerometer data would be of great 249

value. The use of “overall dynamic body acceleration” (ODBA), for example, has proven to be 250

highly accurate in laboratory studies (Halsey, Shepard, & Wilson 2011). It would instructive in a 251

future study to measure both the derived activity metric used in the current study and overall 252

dynamic body acceleration. 253

For the study of free-ranging caribou, displacement was associated with activity 254

level. This result is not unexpected. Locomotion is a primary component of activity in a vagile 255

animal (Boertje 1985a; Duquette & Klein 1987), is associated with elevated heart rates and likely 256

metabolism in reindeer (Nilssen et al. 1984), and it is a behaviour that is accurately captured by 257

accelerometry (Gleiss et al. 2011). After accounting for displacement, our best explanatory model 258

suggested a parabolic relationship with vegetation abundance, where foraging activity is highest at 259

intermediate levels of NDVI. Separate analyses for colder and warmer months also suggested that 260

this reflects a seasonal difference in foraging patterns. At lower measures of vegetation abundance 261

seen during the winter, we found the expected positive correlation and caribou are more active 262

where there is more vegetation, presumably due to foraging behaviour. At higher levels of 263

vegetation abundance seen during the summer months, we found the opposite pattern. One 264

plausible explanation is that caribou living in these areas reach asymptotic levels of food intake 265

when vegetation is plentiful, and therefore actually spend less time foraging (Boertje 1985b). 266

NDVI is significantly and positively correlated with percent conifer tree cover, as estimated from a 267

LANDSAT satellite land cover classification (Spectranalysis Inc. 2004), in both warm and cool 268


For Review O

nly

13

months. Conifer cover is a preferred habitat for caribou and source of forage, particularly for 269

lichens in winter months (Brown, Rettie, & Mallory 2006; Brown et al. 2007). 270

Our best model also included a positive linear relationship between activity and snow 271

depth, suggesting that activity levels are higher in deeper snow, presumably due to cratering 272

behaviour. Note that higher snow depths depress movement across the landscape (Avgar et al. 273

2013) and yet increase overall activity levels. As this behaviour is potentially energetically costly 274

(Fancy & White 1985), especially during the lean winter months, this aspect warrants further 275

analysis at a finer scale. 276

The best model for monthly mean activity levels suggested an inverted polynomial 277

relationship with temperature. Considerably higher activity levels were only found at the lowest 278

temperatures. The lower critical temperature for reindeer is estimated to be about -30°C (Tyler & 279

Blix 1990), yet we see rises in activity at mean monthly temperatures much above this threshold. 280

The temperatures actually experienced by the caribou during the coldest months were obviously 281

colder than the mean and factors such as wind contribute considerably to heat loss as well. Caribou 282

are well adapted to cold temperatures and are known to rely upon shivering and metabolic (or non-283

shivering) thermogenesis (Soppela, Nieminen, & Timisjàrvi 1986), yet this pattern suggests that 284

they may also use behavioural thermal substitution. Little is known, however, about activity as a 285

form of thermoregulation in large mammals (McNab 2002; Humphries & Careau 2011), and we 286

cannot say what form of behaviour this thermoregulatory activity may take. Thermoregulatory 287

constraints are an important determinant of activity patterns (e.g. Armstrong & Robertson 2000) 288

and range limits (e.g. Porter et al. 2002) for many species.. 289

The modelled relationships summarize the primary components of energetic costs for 290

woodland caribou. Landscapes that increase movement rates or displacement distances, of low to 291

intermediate vegetation abundance, with greater snow depth, and lower temperatures are expected 292


For Review O

nly

14

to increase energetic costs. These parameters explain over half the variation in the monthly data. 293

The results of this study suggest that activity sensors based on accelerometery may be useful for 294

field estimation of activity patterns and energetic costs in free-ranging animals. Our results also 295

demonstrated that activity levels varied with both biological and environmental parameters. In 296

sum, this work supports the growing recognition that accelerometery offers a new and important 297

addition in remote biotelemetry (Cooke et al. 2004), allowing ecologists to probe otherwise 298

unobtainable patterns. 299

A number of practical limitations would still limit the utility of accelerometry-based 300

activity scores for precise estimation of absolute energetic costs in the field. The range of 301

behaviours in captive animals, especially those in zoos or holding pens, no doubt differs from that 302

seen in free-ranging caribou (Boertje 1985b; Fancy & White 1986). Our zoo animals exhibited 303

much lower frequencies of walking trotting and (particularly) running than typically found in free-304

living ungulates, as well as foraging infrequently from ground level lichen, grasses or forbs. 305

Moreover, the field estimates of energetic costs we have derived from the literature are possibly 306

clouded by complexities associated with spatial variation in topographic relief, day-to-day 307

variation in snow hardness, depth, and surface ablation, as well as variation amongst individuals in 308

their internal condition, motivational state, and accumulated energy deficits. Although we feel 309

confident in concluding from our study that the relative magnitude of energetic costs in free-living 310

caribou vary considerably with temporal and spatial variation in environmental conditions as well 311

as mobility levels, limited knowledge about caribou behavioral budgets as well as lack of a direct 312

calibration of metabolic costs still hamper our ability to precisely predict absolute costs across the 313

caribou energetic landscape (Shepard et al. 2013). 314

315

Acknowledgements 316


For Review O

nly

15

This work was funded by grants from the Forest Ecosystem Science Cooperative Inc, 317

Ontario Ministry of Natural Resources (OMNR), and Natural Sciences and Engineering Research 318

Council (NSERC) of Canada as well as well as a Vanier Fellowship to T.A. We thank the staff of 319

the OMNR, notably Lyle Walton, for support in the field operations, the Toronto Zoo and staff, 320

and Luba Broitman for software development. We also thank Douglas Morris, Daniel Fortin, and 2 321

anonymous reviewers for comments on the manuscript. None of the authors has a conflict of 322

interest. 323

324

References 325

Alexander, R.M.N. (2003) Principles of Animal Locomotion. Princeton University Press, 326

Princeton and Oxford. 327

Anderson, K.J. & Jetz, W. (2005) The broad-scale ecology of energy expenditure of endotherms. 328

Ecology Letters, 8, 310–318. 329

Armstrong, H.M. & Robertson, A. (2000) Energetics of free-ranging large herbivores: when 330

should costs affect foraging behaviour? Canadian Journal of Zoology, 78, 1604–1615. 331

Avgar, T., Mosser, A., Brown, G.S. & Fryxell, J.M. (2013) Environmental and individual drivers 332

of animal movement patterns across a wide geographical gradient (ed A Mysterud). 333

Journal of Animal Ecology, 82, 96–106. 334

Body, G., Weladji, R.B., & Holand, Ø. (2012). The recursive model as a new approach to validate 335

and monitor activity sensors. Behavioral Ecology and Sociobiology, 66, 1531-1541. 336


For Review O

nly

16

Boertje, R.D. (1985a) An Energy Model for Adult Female Caribou of the Denali Herd, Alaska. 337

Journal of Range Management, 38, 468–473. 338

Boertje, R.D. (1985b) Seasonal activity of the Denali caribou herd, Alaska. Rangifer, 5, 32–42. 339

Brown, J.H., Gillooly, J.F., Allen, A.P., Savage, V.M. & West, G.B. (2004) Toward a Metabolic 340

Theory of Ecology. Ecology, 85, 1771–1789. 341

Brown, G.S., Rettie, W.J., Brooks, R.J. & Mallory, F.F. (2007) Predicting the impacts of forest 342

management on woodland caribou habitat suitability in black spruce boreal forest. Forest 343

Ecology and Management, 245, 137–147. 344

Brown, G.S., Rettie, W.J. & Mallory, F.F. (2006) Application of a variance decomposition method 345

to compare satellite and aerial inventory data: a tool for evaluating wildlife-habitat 346

relationships. Journal of Applied Ecology, 43, 173–184. 347

Buckley, L.B. (2008) Linking Traits to Energetics and Population Dynamics to Predict Lizard 348

Ranges in Changing Environments. The American Naturalist, 171, E1–E19. 349

Burnham, K.P. & Anderson, D.R. (2002) Model Selection and Multimodel Inference: A Practical 350

Information-Theoretic Approach. Springer Scientific, New York. 351

Callaghan, C., Virc, S., & Duffe, J. 2011. Woodland caribou, boreal population, trends in Canada. 352

Canadian Councils of Resource Ministers. Canadian Biodiversity: Ecosystem Status and 353

Trends 2010, Technical Thematic Report 11. 354


For Review O

nly

17

Cooke, S.J., Hinch, S.G., Wikelski, M., Andrews, R.D., Kuchel, L.J., Wolcott, T.G. & Butler, P.J. 355

(2004) Biotelemetry: a mechanistic approach to ecology. Trends in Ecology & Evolution, 356

19, 334–343. 357

Duquette, L.S. & Klein, D.R. (1987) Activity budgets and group size of caribou during spring 358

migration. Canadian Journal of Zoology, 65, 164–168. 359

Elliott, J.M. & Davison, W. (1975) Energy equivalents of oxygen consumption in animal 360

energetics. Oecologia, 19, 195–201. 361

Fancy, S.G. & White, R.G. (1985) Energy Expenditures by Caribou while Cratering in Snow. The 362

Journal of Wildlife Management, 49, 987–993. 363

Fancy, S.G. & White, R.G. (1986) Predicting energy expenditures for activities of caribou from 364

heart rates. Rangifer, Special Issue No. 1, 123–130. 365

Festa-Bianchet, M., Ray, J.C., Boutin, S., Côté, S. & Gunn, A. 2011. Conservation of caribou 366

(Rangifer tarandus) in Canada: an uncertain future. Canadian Journal of Zoology 89, 419-367

434. 368

Gleiss, A.C., Wilson, R.P. & Shepard, E.L.C. (2011) Making overall dynamic body acceleration 369

work: on the theory of acceleration as a proxy for energy expenditure. Methods in Ecology 370

and Evolution, 2, 23–33. 371

Halsey, L.G., Green, J.A., Wilson, R.P. & Frappell, P.B. (2009a) Accelerometry to Estimate 372

Energy Expenditure during Activity: Best Practice with Data Loggers. Physiological and 373

Biochemical Zoology, 82, 396–404. 374


For Review O

nly

18

Halsey, L., Shepard, E., Quintana, F., Gomezlaich, A., Green, J. & Wilson, R. (2009b) The 375

relationship between oxygen consumption and body acceleration in a range of species. 376

Comparative Biochemistry and Physiology - Part A: Molecular & Integrative Physiology, 377

152, 197–202. 378

Halsey, L.G., Shepard, E.L.C. & Wilson, R.P. (2011) Assessing the development and application 379

of the accelerometry technique for estimating energy expenditure. Comparative 380

Biochemistry and Physiology Part A: Molecular & Integrative Physiology, 158, 305–314. 381

Humphries, M.M. & Careau, V. (2011) Heat for Nothing or Activity for Free? Evidence and 382

Implications of Activity-Thermoregulatory Heat Substitution. Integrative and Comparative 383

Biology, 51, 419–431. 384

Legislative Assembly of Ontario. (2007) Endangered Species Act. 385

McNab, B.K. (2002) The Physiological Ecology of Vertebrates: A View from Energetics. 386

Comstock Publishing Assoc./Cornell University Press, Ithica, New York. 387

Nathan, R., Spiegel, O., Fortmann-Roe, S., Harel, R., Wikelski, M. & Getz, W.M. (2012) Using 388

tri-axial acceleration data to identify behavioral modes of free-ranging animals: general 389

concepts and tools illustrated for griffon vultures. The Journal of Experimental Biology, 390

215, 986–996. 391

Nilssen, K.J., Johnson, H.K., Rognmo, A., & Blix, A.S. 1984. Heart rrate and energy expenditure 392

in resting and running Svalbard and Norwegian reindeer. American Journal of Physiology 393

Regulation and Integrative Comparative Physiology, 246, R963-R967. 394


For Review O

nly

19

Ontario Woodland Caribou Recovery Team (OWCRT). (2008) Woodland caribou (Rangifer 395

tarandus caribou) (forest-dwelling, boreal population) in Ontario. 396

Porter, W.P., Sabo, J.L., Tracy, C.R., Reichman, O.J. & Ramankutty, N. (2002) Physiology on a 397

Landscape Scale: Plant-Animal Interactions. Integrative and Comparative Biology, 42, 431 398

–453. 399

Schaefer, J.A. (2003) Long‐Term Range Recession and the Persistence of Caribou in the Taiga. 400

Conservation Biology, 17, 1435–1439. 401

Scharf, F.S., Juanes, F., & Sutherland, M. (1998) Inferring ecological relationships from the edges 402

of scatter diagrams: comparison of regression techniques. Ecology 79, 448-460. 403

Shepard, E.L.C., Wilson, R.P., Rees, W.G., Grundy, E., Lambertucci, S.A., & Vosper, S.B. 2013. 404

Energy landscapes shape animal movement ecology. American Naturalist 182, 298-312. 405

Soppela, P., Nieminen, M. & Timisjàrvi, J. (1986) Thermoregulation in reindeer. Rangifer, 273 – 406

278. 407

Spectranalysis Inc. (2004) Introduction to the Ontario Land Cover Data Base, Second Edition 408

(2000): Outline of Production Methodology and Description of 27 Land Cover Classes. 409

report to the Ontario Ministry of Natural Resources. 410

Statutes of Canada. (2002) Species at Risk Act. 411

Tyler, N.J.C. & Blix, A.S. (1990) Survival strategies in arctic ungulates. Rangifer, Special Issue 412

No. 3, 211–230. 413


For Review O

nly

20

Weiner, J. (1992) Physiological limits to sustainable energy budgets in birds and mammals: 414

Ecological implications. Trends in Ecology & Evolution, 7, 384–388. 415

Wilson, R.P., Quintana, F. & Hobson, V.J. (2011) Construction of energy landscapes can clarify 416

the movement and distribution of foraging animals. Proceedings of the Royal Society B: 417

Biological Sciences, 279, 975–980. 418

Wilson, R.P., White, C.R., Quintana, F., Halsey, L.G., Liebsch, N., Martin, G.R. & Butler, P.J. 419

(2006) Moving towards acceleration for estimates of activity-specific metabolic rate in 420

free-living animals: the case of the cormorant. Journal of Animal Ecology, 75, 1081–1090. 421

Zheng, B. (2000) Summarizing the goodness of fit of generalized linear models for longitudinal 422

data. Statistics in Medicine, 19, 1265–1275. 423

424


For Review O

nly

21

Figure legends 425

426

Figure 1: Results of captive reindeer study. A) Observed behaviours for different levels of activity 427

as measured over 5 minute intervals. The final category includes activity levels of ≥0.34. The 428

energetic costs associated with each behaviour, from published research, are listed on the right. B) 429

The aggregate estimated cost of captive reindeer behaviours over 5 or 20 minute intervals in 430

relation to the activity score. 431

432

Figure 2: An illustration of broad spatial variation of activity levels in wild caribou. Monthly mean 433

activity values were averaged for each year and individual, and those with fewer than 10 months 434

of data were excluded (leaving 81 individuals represented here). Mean annual spatial locations are 435

indicated by white dots. Values are the percent of the possible maximum activity level. Inset 436

indicates region of study area. 437

438

Figure 3: A) Mean monthly activity of wild caribou versus mean monthly 5-hour displacement in 439

metres (N=1231). Residual mean monthly activity (after accounting for displacement and interval 440

type) versus mean monthly B) NDVI value, C) snow depth, D) and temperature (N=1231). Grey 441

lines (B and D) show the polynomial relationship as estimated by the best fit model. Dark shading 442

corresponds to summer months (May-October) whereas light shading corresponds to winter 443

months (November-April). 444

445

446

447

448


For Review O

nly

22

Figure 1 449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

1 (Boertje 1985a) 466

2 (Fancy & White 1986) 467

3 Assumed to be similar to “grazing” 468

4 Energetic cost=8.41+2.54(1-e-14.79(activity)) 469

5 Energetic cost=8.18+3.01(1-e-11.74(activity)� 470

471

0%

20%

40%

60%

80%

100%

0 0.1 0.2 0.3

Perc

ent

of

tim

e in b

ehavio

ur

Trotting

Walking

Rubbing

Grazing

Social interaction

Trough feeding

Standing

Laying

8

9

10

11

12

13

0 0.1 0.2 0.3 0.4 0.5

Energ

etic c

ost (k

J/k

g/h

)

Proportion of interval active

5 min intervals

5 min fitted curve

20 min intervals

20 min fitted curve

A

B

Behaviour Energetic cost

(kJ/kg/h)

Trotting 30.00 1

Walking 13.42 2

Rubbing 10.76 3

Grazing 10.76 2

Social interaction 10.76 3

Trough feeding 10.30 2

Standing 9.19 2

Lying down 8.37 2

4

5


For Review O

nly

23

Figure 2 472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488


For Review O

nly

24

Figure 3 489

490

491

0%

5%

10%

15%

20%

25%

30%

35%

0 1000 2000 3000 4000 5000

Mea

n a

cti

vit

y l

evel

(% o

f m

axim

um

)

Mean 5-hour displacement

-10%

-5%

0%

5%

10%

15%

-1500 0 1500 3000 4500 6000 7500

Ac

tivit

y l

eve

l (r

esid

ual)

NDVI

-10%

-5%

0%

5%

10%

15%

0 0.2 0.4 0.6

Ac

tivit

y l

evel (r

es

idu

al)

Snowdepth (m)

-10%

-5%

0%

5%

10%

15%

-30 -15 0 15 30

Acti

vit

y l

evel (r

es

idu

al)

Temperature (°°°°C)


For Review O

nly

25

Tables 492

493

Table 1: Top ten models (ordered by strength of model fit) for analysis of mean monthly activity, 494

including degrees of freedom (DF), ∆AIC relative to the best fit model, and model weight (ω). 495

Explanatory variables are displacement (disp), vegetation abundance (NDVI), snow depth (snow) 496

and temperature (temp). When second order terms were included in the model, the univariate 497

terms were necessarily included as well. The best explanatory model, with a model weight of 0.49, 498

explained an estimated 51% of the variance. 499

500

Model DF ∆AIC ω

disp NDVI2 snow temp

2 10 0.00 0.49

disp NDVI2 temp

2 9 0.83 0.33

disp NDVI2 snow

2 temp

2 11 1.99 0.18

disp temp2 7 24.41 0.00

disp snow temp2 8 25.86 0.00

disp NDVI temp2 8 26.40 0.00

disp snow temp snow:temp 8 26.87 0.00

disp temp snow:temp 7 27.10 0.00

disp snow2 temp

2 9 27.14 0.00

disp NDVI snow temp2 9 27.76 0.00

501


For Review O

nly

26

Table 2: Best model fit for mean monthly activity vs. candidate model including second order 502

terms. 503

504

505 Variable Value Std.Error t-value p-value

interval type (20 vs. 5min) -0.5954 0.2065 -2.88 0.004

displacement 0.0028 0.0001 23.80 <0.001

NDVI 0.0010 0.0002 5.18 <0.001

NDVI2 -1.422

-7 0.0000 -5.47 <0.001

snow depth 2.6989 1.6071 1.68 0.093

temperature -0.0930 0.0185 -5.02 <0.001

temperature2 0.0051 0.0009 5.52 <0.001


Towards an energetic landscape: broad-scale accelerometry in woodland caribou

Documents

Transcript of Towards an energetic landscape: broad-scale accelerometry in woodland caribou