Spatial-temporal dynamics of f grizzly bear movement presented by - - PowerPoint PPT Presentation

▶
spatial temporal dynamics of f grizzly bear movement
SMART_READER_LITE
LIVE PREVIEW

Spatial-temporal dynamics of f grizzly bear movement presented by - - PowerPoint PPT Presentation

Oct 07, 2022 214 likes •435 views

Spatial-temporal dynamics of f grizzly bear movement presented by Mathieu Bourbonnais Research question Q3A.1 Can grizzly bear responses to disturbance, as represented by movement and health, be modeled to allow probabilistic prediction of

slide-1
SLIDE 1

Spatial-temporal dynamics of f grizzly bear movement

presented by Mathieu Bourbonnais

slide-2
SLIDE 2

Research question

Q3A.1 Can grizzly bear responses to disturbance, as represented by movement and health, be modeled to allow probabilistic prediction of future movement associated with changes in landscape disturbance and habitat fragmentation?

2

slide-3
SLIDE 3

BEHAVIOUR

Movement

3

Source: Nielsen et al. (2009)

  • Link individuals/populations/landscape
  • Variation in space and time at multiple scales
  • Shifts in movement parameters/behaviours?
  • Time: 2001 – 2014
  • Management areas
  • 121 individuals, 97 with multiple captures (n =

218)

  • ~400 000 locations
slide-4
SLIDE 4

BEHAVIOUR

Movement parameters

4

β

SL = Step Lengths β = Turning Angles

slide-5
SLIDE 5

BEHAVIOUR

5

Movement parameters

slide-6
SLIDE 6

BEHAVIOUR

Space use

6

slide-7
SLIDE 7

BEHAVIOUR

Variables

7

  • 1. External
  • Disturbance (IRSS)
  • Roads, harvest, well-sites, mines,

infrastructure, and fires

  • Food (ACE Lab)
  • Seasonal food availability

(vegetation/meat)

  • Topography
  • Vegetation indices (NDVI & DHI)
  • Land cover
  • 2. Internal
  • Cohort (Gender x Age; Mortalities)
  • 3. Space
  • Conservation areas
  • 4. Time
  • Hourly, seasonal
  • Inter-annual
  • Time since hunt
slide-8
SLIDE 8

BEHAVIOUR

Selection ratio

8

  • Ratio of hourly disturbance f and

the seasonal disturbance f

  • SR < 1: select less than available
  • SR > 1: select greater than available
slide-9
SLIDE 9

BEHAVIOUR

General modelling approach

9

  • Multilevel Bayesian framework (STAN)
  • Group-level intercepts by bear
  • Response
  • Movement parameters (rates; space use)
  • Multiple combinations of variables
  • Time (diel; seasonal; inter-annual) + Space (CA’s) + Disturbance (SR’s;

Density) + Food + Landscape Productivity (veg indices) + Topography + Land cover + Internal factors (cohort)

  • Compare using WAIC
slide-10
SLIDE 10

BEHAVIOUR

Results

10

Hourly movements (∆WAIC = 319.85)

  • Time of day (TOD)
  • Cohort
  • SR: Road, Harvest, Wells
  • Density: Road, Harvest, Wells
  • Food
  • Conservation areas
  • Years since hunt
  • TOD * SR * Density
slide-11
SLIDE 11

BEHAVIOUR

Hourly movements

11

slide-12
SLIDE 12

BEHAVIOUR

Hourly movements

12

slide-13
SLIDE 13

BEHAVIOUR

Hourly movements

13

slide-14
SLIDE 14

BEHAVIOUR

Results

14

Daily movements (∆WAIC = 21913)

  • Season
  • Cohort
  • Density: Road, Harvest, Wells
  • Food
  • Conservation areas
  • Years since hunt
  • Season * Density
slide-15
SLIDE 15

BEHAVIOUR

Daily movements

15

slide-16
SLIDE 16

BEHAVIOUR

Daily movements

16

slide-17
SLIDE 17

BEHAVIOUR

Results

17

Seasonal movements (∆WAIC = 62.77)

  • Season
  • Cohort
  • Proportion home range: harvest, roads, well-sites
  • Food
  • Conservation areas
  • Years since hunt
  • Season * PHR
slide-18
SLIDE 18

BEHAVIOUR

Seasonal movements

18

slide-19
SLIDE 19

BEHAVIOUR

Seasonal movements

19

slide-20
SLIDE 20

BEHAVIOUR

Summary

20

  • Evidence that movement rates have changed over space and time
  • Dependent on temporal scale
  • Decreasing hourly/daily and increasing seasonal
  • Core area generally lowest movement rates (food, security?)
  • Hourly scale dependent on density and “strength” of selection
  • High density * high selection = greater movement rates
  • Increasing hourly SRs represent “novel” habitat compared to HR
  • Link to mortality well established
  • Energetic requirements
  • Daily distance travelled / seasonal range size more dependent on

density rather than selection

slide-21
SLIDE 21

BEHAVIOUR

Next steps

21

  • Hourly, daily, and seasonal models easily used for prediction
  • f movement rates / space use
  • Ideally, want to predict the probability of space use and movement
  • Chapter 3: Third order selection over time and linkages among

areas using movement probability (step selection function).

  • Do bears in poorer condition exhibit “risky” behaviour?
  • Chapter 4: Behavioural path segmentation (HMM) combined with

BCI data

  • Could predict changes in body condition using

new energetic food landscape models.