1. Integrated Monitoring in Bird Conservation Regions
David J. Hanni
&
Jennifer A. Blakesley
Rocky Mountain Bird Observatory

2. Guidelines for Avian Monitoring
Goal 1: Integrate monitoring into bird management and conservation practices.
Goal 2: Coordinate monitoring programs among organizations and integrate them across spatial scales.
Goal 3: Increase the value of monitoring information by improving statistical design.
Goal 4: Maintain bird population monitoring data in modern data management systems.
In 2007 … 2

3. Our Objectives
Provide a design framework to integrate bird monitoring efforts
Precisely estimate distribution, density, site occupancy, population trends and species richness
Provide habitat association data relevant to landscape changes
Maintain a high-quality database, accessible online
Create decision support tools to help guide conservation efforts 3

4. Bird Conservation Regions

5. Current Monitoring Programs
All lands in BCR 17
All lands in CO, WY and MT
Portions of 9 additional states
All BLM lands in: CO, WY, MT, ND, SD
All USFS lands in Regions 1 & 2
(CO, WY, NE, KS, SD, ND, ID, MT)
3 National Forests in Region 3
(Kaibab, Coconino, Prescott) 5

6. 2010 Landbird Monitoring

7. Current Partnerships Colorado Division of Wildlife
Wyoming Game and Fish Department
Montana Fish, Wildlife and Parks
South Dakota Game, Fish and Parks
USFS: 27 National Forests, 9 National Grasslands, 4 regions, 12 states
BLM in 5 states
National Park Service
…(continued)

8. Current Partnerships (cont.)
Northern Great Plains Joint Venture
Audubon Wyoming
Wyoming National Diversity Database
Montana Natural Heritage Program
Idaho Bird Observatory
Great Plains Landscape Conservation Cooperative

9. Sampling Design and Methods
Sampling Frame
Stratification
Sampling Units
Sample Selection
Sample Allocation (effort)
Sampling Methods
Analytical Methods
Example Results

10. Sampling Frame: BCRs
Spatial scale should correspond to ecological process of interest.

11. Stratification
Stratification should be defined by areas to which we want to make inferences
Strata are based on fixed attributes
Federal/state land ownership
Elevation, latitude, soil type, ecoregion
All vegetation types available for sampling
Flexible: Each state within the BCR and each BCR within a state can be stratified differently (depending on local needs) 11

12. Colorado BCR 16 12

13. Colorado BCR 18 13

14. Montana BCR 11, 10 and 17

15. Wyoming BCRs 16, 17, 18

16. Wyoming BCR 9 and 10

17. Sampling Units
1 km2 cell Step 1: Overlay grid on sampling frame Step 2: Attribute each cell with pertinent data (unique ID, spatial location, land ownership, elevation, soil type, …)

18. Sample Selection
Spatially Balanced: Generalized Random Tessellation Stratification (GRTS)
Ensures a spatially balanced distribution of samples within each stratum
~ Random
Analyses can incorporate spatial information in estimation of sampling variance 18

19. Sample Allocation
Minimum of 2 samples required/stratum Recommend 10 Determined by funding partners May vary annually

20. Base Sample

21. Additional Samples in Colorado

22. Additional Samples - Comanche NG

23. Sampling Effort example: BCR 17
All of BCR 17: 264 samples
MT: 84 samples
WY: 46 samples
ND: 37 samples
SD: 90 samples
NE: samples

24. Sampling Effort example: BCR 17

25. Sampling methods 16 points per cell 5 minute point count
250 m spacing
125 m from edge
5 minute point count
1 minute intervals
Record distance to each bird seen or heard
Record species and sex for each observation
1 minute intervals allow for Removal modeling
Distances are required for Distance sampling 25

26. Analytical methods Estimate detection probability and density
Distance sampling (Buckland et al. 2001)
Removal sampling (Farnsworth et al. 2002)
Estimate detection probability and occupancy rate
Combined Removal and Occupancy modeling (Pavlacky et al. in review)
Habitat modeling
Multi-scale occupancy estimation 26

27. Scalable Design
Aggregation of stratum-wide estimates to BCR- or state-wide estimates is built into the design
Example: Black Hills National Forest in WY
Contributes to estimates for:
BHNF
Wyoming
BCR 17

28. Results: Density and Abundance
Example: Townsend’s Solitaire D N
%CV n
Black Hills NF
3.74
24,533 19 89
Wyoming BCR 17
0.36
23,406 77 7
BCR 17
0.24
84,817 40
109

29. Results: Site Occupancy
Example: Brown Creeper
Psi
%CV
n Tran
Black Hills NF
0.287 34 10
BCR 17
0.073 63 15

30. Results of Habitat Modeling
Example: Brewer’s Sparrow
CO, BCR 16, BLM
Large scale (1 km2 sampling cell)

31. Results of Habitat Modeling
Example: Brewer’s Sparrow
CO, BCR 16, BLM
Large scale (1 km2 sampling cell)
Occupancy rates predicted using sagebrush area.

32. Results of Habitat Modeling
Example: Brewer’s Sparrow
CO, BCR 16, BLM
Small scale (sampling point)

33. Benefits of IMBCR design
Collaboration
Shared costs among partners
Handles fluctuating funding
Ability to compare bird trend to habitat trend
Ability to compare local to regional results
Flexibility in stratification
All vegetation types available for sampling
Can be (is) used for other taxa
Aggregation of strata-wide estimates to BCR- or state-wide estimates is built into the design
Because each stratum has its own spatially-balanced, ordered sample, sampling effort can vary among strata and among years and still provide statistically valid estimates

34. Acknowledgements Original Collaborators in Project Design
Robert Skorkowsky, USFS
David Klute, CDOW
Paul Lukacs, CDOW
Greg Hayward, USFS
Christina Hargas, USFS 34

35. GPLCC Introduction Landscape Conservation Cooperatives (LCCs)
Great Plains LCC
Landscape Conservation Cooperatives (LCCs)
Science-based Partnerships
Help guide and coordinate conservation efforts at regional levels
Great Plains LCC (GPLCC): BCR 18 and 19

36. GPLCC Project Objectives
Create a standardized grid
USNG
Develop strata for GPLCC
Grid attributes
Select samples
Spatially balanced sampling
grts function in R

37. US National Grid (USNG)
USNG identified as a potential standard for monitoring populations at regional levels
Developed by FGDC for emergency response coordination
1-km square grid cells
Contains key elements for a proper monitoring grid (national coverage, commonly used datum/projection, scalable)

38. Benefits of a standardized grid
Same starting point
Spatial continuity across different projects
decrease project costs, necessary labor, and duplicate sampling efforts
No need to create multiple grids every time a project is funded
A sampling grid provides an organized spatial extent, partitioned into sampling units needed in spatially explicit models, and is integral to research and monitoring.

39. Standardized grid
Will improve partnership coordination at the landscape and local level to accomplish shared conservation goals
Can also be used for other monitoring programs
Adaptive management

40. GPLCC Stratification Strata based on grid attributes
Primarily NRCS ecoregions, federal land ownership, and strahler order

41. For more details… RMBO Avian Data Center:
e.g., Monitoring the Birds of the Badlands and Prairies Bird Conservation Region (BCR 17):
2009 Field Season Report