Spatial Capture-Recapture Scenario Detectors Animal locations - PowerPoint PPT Presentation

Spatial Capture-Recapture

Scenario Detectors Animal locations

Effective area? – No Problem: bread & butter to Distance Samplers! • This is just MRDS Point Transect Survey Detector ID 1 2 3 4 … 1 0 1 0 0 2 0 0 1 1 3 1 1 0 1 4 0 1 1 1 5 1 1 0 0 etc. • … but with lots of points, (and sometimes counts instead of 0/1 data)

Detection function Distance Detector

Combining across detectors and occasions

Combining across detectors and occasions ò = a p ( X d ) X i s the area effectively sampled by detectors • (if animal density constant in space).

Problem: Don’t observe distance, only observe capture location • But locations of captures/non-captures contains information about detection probability and location/distance. • Like a multiple-point point transect survey with errors in distance estimation.

Where is the information on location? Color is elevation − 16 Unobserved activity centre 1400 − 18 ● ● 4 1200 − 20 ● ● 1000 3 y ● ● 800 − 22 ● 1 600 ● ● − 24 ● ● 5 2 400 − 26 200 − 15 − 10 − 5 0 5 x Camera traps

Where is the information on location? Observed capture history 3 4 2 5 1 8 0 ● 0 5 − 1

Where is the information on location? 3 4 2 5 1 8 0 ● 0 5 − 1 Observed capture history in 1D n i1 =8 n i3 =5 n i2 =5 n i4 =4 n i5 =0 ● ● ● ● ● 5 3 1 2 4 Distance

Where is the information on location? 3 4 2 5 1 8 0 ● 0 5 − 1 Observed capture history in 1D n i1 =8 Detection 8 function 6 n i3 =5 n i2 =5 n i4 =4 λ ( d ) T 4 2 n i5 =0 ● ● ● ● ● 5 3 ● 1 2 4 0 − 2 − 10 − 5 0 5 10 Distance

Where is the information on location? 3 4 2 5 1 8 0 ● 0 5 − 1 Detection function in 2D 2 2 1 1 3 4 8 5 ● ● 4 6 7 6 ● ● 8 3 Activity centre ● ● ● 1 4 ● ● ● ● 5 2 2 0

(Can also predict locations after the event)

Can fit Density Surface Models 7500000 7500000 0.50 0.0035 7450000 7450000 0.45 0.0030 7400000 7400000 0.40 0.0025 0.0020 0.35 7350000 7350000 y y 0.0015 0.30 7300000 7300000 0.0010 0.25 7250000 7250000 0.0005 0.20 7200000 7200000 0.0000 300000 340000 380000 300000 340000 380000 x x

Can Model Habitat Use 3 4 2 5 1 8 0 ● 0 5 − 1 − 16 6 5 2 4 8 − 18 8 ● ● 4 − 20 6 ● ● 1 1 3 y ● ● 2 − 22 ● 1 4 3 4 ● ● − 24 ● ● 5 2 2 3 − 26 0 5 2 6 7 − 15 − 10 − 5 0 5 x

Mist-netting red-eyed vireo

Genetically capturing stoats Photographs: Bruce Warburton Picture: Murray Efford

Acoustically Surveying Gibbons Human acoustic detectors 1570 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 1565 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● UTM.Y (km) ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 1560 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 1555 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 1550 ● ● ● 665 670 675 680 685 690 695 700

Acoustically Surveying Frogs

Summary • Mark-recapture with distance detection function (and unknown distance) • R library � secr � on CRAN (Murray Efford) • Also Bayesian software • R library “ascr” (Ben Stevenson) for acoustic SCR at https://github.com/b-steve/asecr/