Power analysis for a Mark recapture project

1
Power analysis for a Mark recapture project

• Dan Bachen
• Department of Ecology
• Montana State University

2
Project overview

• Data collection
• Blacktail WMA
• 24 plots
• 4 weeks

3
Mark Recapture

• Data collection Methods
• Trap grids 8x8
• Sherman live traps
• 5 nights per plot
• Unique marks
• 4 species
• Deer mouse dominant gt 90 of captures

4
Data example
Site Capture Location N1 Capture Location N2 Capture Location N3 Capture Location N4 Capture Location N5
SC1 0 4G 0 0 0
SC1 0 4E 0 0 0
SC1 0 6F 4G 0 0
5
Power analysis

• Goals
• Optimize sample size
• Treatment effect -.1 to -1
• 20 sites used (10 treatment, 10 control)
• 40 sites used (20 treatment, 20 control)
• Effect of -.84 recorded by Ostoja 2009
• Assumptions
• Abundance estimates normal
• Capture probability
• .2 un-captured individuals
• .9 captured individuals

6
Power analysis

• Methods
• Generate abundances using likelihood analysis
  (Rcapture)
• Quantify statistics
• Use statistics to set parameters
• Iterate experiment
• Score success

7
Results
Mean 7.44 SE 5.84
8
Results
9
Results
10
Discussion

• Experiment might have worked
• Did I use the right tools?
• No
• Aggregation of samples
• Example
• Abundance 158.2
• SE 2.3

11
Future Work

• Re-run power analysis aggregating samples
• New tools
• RMark
• Compare results to other techniques