Boreal Avian Modelling Project: Database Structure

0d7863afe1ef2ad8a9229afffb54ab9a?s=47 Peter Solymos
February 02, 2016

Boreal Avian Modelling Project: Database Structure

This talk was presented remotely to the Audubon Society in February, 2016. The talk describes how the Boreal Avian Modelling (BAM) project data is structured and standardized for bird density modelling and prediction.

0d7863afe1ef2ad8a9229afffb54ab9a?s=128

Peter Solymos

February 02, 2016
Tweet

Transcript

  1. Boreal Avian Modelling Project

  2. What is BAM? u  A collaborative science program u  committed

    to improved understanding of the ecology of birds and their habitats u  in the boreal region of North America. u  Using quantitative modelling techniques and u  a comprehensive dataset assembled from projects across the continent, u  we derive information on abundance, distribution and habitats of boreal birds, u  and use this to evaluate and predict the effects of human activity. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 2
  3. Who is involved in BAM? Boreal Avian Modelling Project Ι

    Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca Steering Committee (scientific direction and relevance) Fiona Schmiegelow (U of A) Samantha Song (EC) Erin Bayne (U of A) Steve Cumming (U. Laval) BAM Team (project delivery) Nicole Barker (Coordinating Scientist), Trish Fontaine (Database Manager), Péter Sólymos (Statistician), Diana Stralberg (Ecologist), Alberto Suarez Esteban (Post-doc), Lionel Leston (Post-doc), Lisa Mahon (Contributing scientist - EC), Samuel Haché (Contributing scientist - EC), Steve Van Wilgenburg (Contributing scientist - EC), Steve Matsuoka (Contributing scientist - USGS affiliate), Tara Stehelin (Ph.D. student), Alana Westwood (Ph.D. student) Technical Committee (advice, data sharing, collaboration) Marcel Darveau, DUC/Université Laval André Desrochers, Université Laval Pierre Drapeau, Université Québec à Montréal Charles Francis, Environment Canada Colleen Handel, USGS - Alaska Keith Hobson, Environment Canada Craig Machtans, Environment Canada Julienne Morissette, Ducks Unlimited Gerald Niemi, University of Minnesota, Duluth Rob Rempel, Ontario MNR / Lakehead University Stuart Slattery, Ducks Unlimited Canada Phil Taylor, Acadia University / Bird Studies Canada Lisa Venier, Canadian Forest Service Pierre Vernier, University of British Columbia Marc-André Villard, Université de Moncton 3
  4. Who funds BAM? u  Grants u  Environment Canada (Canadian Wildlife

    Service, Migratory Birds Program) u  USFWS NMBCA grants u  Joint Oil Sands Monitoring u  Climate Change and Emissions Management Corporation u  Vanier Canada Graduate Scholarships u  Alberta Pacific Forest Industries Inc. u  Institutional/infrastructure support u  University of Alberta u  Université Laval u  Past Funders u  Alberta Biodiversity Monitoring Institute, Alberta Conservation Association, Government of Alberta, Canada Foundation for Innovation, Canada Research Chairs, Ducks Unlimited Canada, Fonds Québécois de la recherche sur la nature et les technologies, Canadian Boreal Initiative, Forest Products Association of Canada, EC Habitat Stewardship Funds, Killam Trusts, Ministere des ressources naturelles et de la faune, National Fish & Wildlife Foundation, NSERC, US Landscape Conservation Cooperatives Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 4
  5. Inception – 2004 u  Lack of comprehensive monitoring data for

    management and conservation of North American boreal bird populations. u  Recognized the potential in the many extant data sets: u  Small scale projects u  Long-term large scale: BBS u  Systematic effort to compile all such data from the boreal forest into a common data structure and repository. à Solitit data Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 5 Similar methodology
  6. Individual Data Partners K. Aitken, A. Ajmi, B. Andres, J.

    Ball, E. Bayne, P. Belagus, S. Bennett, R. Berger, M. Betts, J. Bielech, A. Bismanis, R. Brown, M. Cadman, D. Collister, M. Cranny, S. Cumming, L. Darling, M. Darveau, C. De La Mare, A. Desrochers, T. Diamond, M. Donnelly, C. Downs, P. Drapeau, C. Duane, B. Dube, D. Dye, R. Eccles, P. Farrington, R. Fernandes, M. Flamme, D. Fortin, K. Foster, M. Gill, T. Gotthardt, N. Guldager, R. Hall, C. Handel, S. Hannon, B. Harrison, C. Harwood, J. Herbers, K. Hobson, M.-A. Hudson, L. Imbeau, P. Johnstone, V. Keenan, K. Koch, M. Laker, S. Lapointe, R. Latifovic, R. Lauzon, M. Leblanc, L. Ledrew, J. Lemaitre, D. Lepage, B. MacCallum, P. MacDonell, C. Machtans, C. McIntyre, M. McGovern, D. McKenney, S. Mason, L. Morgantini, L. Morton, G. Niemi, T. Nudds, P. Papadol, M. Phinney, D. Phoenix, D. Pinaud, D. Player, D. Price, R. Rempel, A. Rosaasen, S. Running, R. Russell, C. Savignac, J. Schieck, F. Schmiegelow, D. Shaw, P. Sinclair, A. Smith, S. Song, C. Spytz, D. Swanson, S. Swanson, P. Taylor, S. Van Wilgenburg, P. Vernier, M.-A. Villard, D. Whitaker, T. Wild, J. Witiw, S. Wyshynski, M. Yaremko, as well as the hundred of volunteers collecting Breeding Bird Survey (BBS) data. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 6
  7. Institutional Data Partners Acadia University; Alaska Bird Observatory; Alaska Natural

    Heritage Program; Alberta Biodiversity Monitoring Institute; Alberta Pacific Forest Industries Inc.; AMEC Earth & Environmental; AREVA Resources Canada Inc.; Avian Knowledge Network; AXYS Environmental Consulting Ltd.; Bighorn Wildlife Technologies Ltd.; Bird Studies Canada; Breeding Bird Survey (coordinated in Canada by Environment Canada); BC Breeding Bird Atlas; Canadian Natural Resources Ltd.; Canfor Corporation; Daishowa Marubeni International Ltd; Canada Centre for Remote Sensing and Canadian Forest Service, Natural Resources Canada; Canadian Wildlife Service and Science & Technology Branch, Environment Canada; Global Land Cover Facility; Golder Associates Ltd.; Government of British Columbia; Government of Yukon; Hinton Wood Products; Hydro-Québec Équipement; Kluane Ecosystem Monitoring Project; Komex International Ltd.; Louisiana Pacific Canada Ltd.; Manitoba Breeding Bird Atlas; Manitoba Hydro; Manitoba Model Forest Inc.; Manning Diversified Forest Products Ltd.; Maritimes Breeding Bird Atlas; Matrix Solutions Inc. Environment & Engineering; MEG Energy Corp.; Mirkwood Ecological Consultants Ltd.; NatureCounts; Nature Serve; Numerical Terradynamic Simulation Group; Ontario Breeding Bird Atlas; Ontario Ministry of Natural Resources; OPTI Canada Inc.; PanCanadian Petroleum Limited; Parks Canada (Mountain National Parks Avian Monitoring Database); Petro Canada; Principal Wildlife Resource Consulting; Regroupement QuébecOiseaux; Rio Alto Resources International Inc.; Saskatchewan Environment; Shell Canada Ltd.; Suncor Energy Inc.; Tembec Industries Inc.; Tolko Industries Ltd.; U.S. Army; U.S. Fish and Wildlife Service; U.S. Geological Survey, Alaska Science Center; U.S. National Park Service; Université de Moncton; Université du Québec à Montréal; Université du Québec en Abitibi- Témiscamingue; Université Laval; University of Alaska, Fairbanks; University of Alberta; University of British Columbia; University of Guelph; University of New Brunswick; University of Northern British Columbia; URSUS Ecosystem Management Ltd.; West Fraser Timber Co. Ltd.; Weyerhaeuser Company Ltd.; Wildlife Resource Consulting Services MB Inc. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 7
  8. Lesson #1: identify data requirements u  We initially accepted data

    in widely different formats: u  scanned data sheets u  relational databases. u  Required substantial effort to input and reformat the data into a common structure. u  Identify minimum data requirements to control the cost of the data collection effort: u  essential fields, u  usable file structures. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 8
  9. Project life cycle Boreal Avian Modelling Project Ι Projet de

    modélisation de l'avifaune boréale Ι www.borealbirds.ca 9 Barker et al. 2015 time capsule for each contributing project, facilitating quality assurance and version control
  10. u  Project ID u  Project lead u  Methodology (can change

    over the years à multiple fields) u  Data sharing flags Database relations Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 10 PCODE METHOD Projects
  11. u  Unique location ID (revisits are sometimes tricky) u  Latitude/longitude

    u  Coutry, jurisdiction, BCR region, Boreal boundary u  Nested structure indicators (‘site’ ~ nearby points) Database relations Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 11 PCODE METHOD PCODE SS Projects Locations
  12. u  Survey ID u  Date, time (between year revisits) u 

    Roadside or off-road survey u  Visits (time gap in between, within year) Database relations Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 12 PCODE METHOD PCODE SS SS PKEY YEAR DATE TIME ROAD VISIT Projects Locations Surveys
  13. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 13
  14. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 14
  15. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 15
  16. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 16
  17. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 17
  18. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 18
  19. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 19
  20. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 20
  21. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 21
  22. Avian database Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 22
  23. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune

    boréale Ι www.borealbirds.ca 23 Barker et al. 2015 Points are clustered Points are close to linear features Points are not random
  24. Lesson #2: maintain design/metadata u  Sampling design & access: u 

    survey points are spatially structured. u  This information is useful in controlling for lack of statistical independence: u  spatial weighting, subsampling; u  random effects in hierarchical models. u  Maintain comprehensive information about: u  spatial sampling design/hierarchy, u  Experimental design (treatment, control-impact, etc.) Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 24
  25. u  Track individuals u  Removal design u  Distance bands u 

    Species (AOU codes) u  Behaviour codes Database relations Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 25 PCODE METHOD PCODE SS PKEY METHOD DURMETH DISMETH SPECIES BEH ABUND SS PKEY YEAR DATE TIME ROAD VISIT Projects Locations Surveys Count intervals SPECIES lookup table BEH lookup table DURMETH lookup table DISMETH lookup table
  26. Lesson #3: track protocol changes u  Survey protocol can change

    within the lifetime of projects: u  sampling effort stays same, more/less info collected, u  sampling effort changes, u  new technologies (e.g. ARUs). u  Methodology not always linked to projects u  Do not assume that protocols remain the same u  document protocols during each update. u  This also affects data harmonization when sampling effort changes. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 26
  27. With more data comes more protocols u  Surveys are not

    standardized: u  time intervals vary, u  distance intervals vary, u  53 protocols. u  Time and distance intervals: u  more information about the observation process. u  Sampling effort affects counts u  Standardization is required 27 # time int. # distance int. surveys % 1 1 >1 >1 1 >1 1 >1 75% 1% 12% 12% Matsuoka et al. 2014
  28. 28 The observation process 0–50 50–100 >100 m 0–3 3–5

    5–10 min time interval distance band count
  29. 29 The observation process 0–50 50–100 >100 m 0–3 3–5

    5–10 min time interval distance band count 2 1 1
  30. 30 The observation process 0–50 50–100 >100 m 0–3 3–5

    5–10 min time interval distance band count 2 3 1 3 1 1
  31. 31 The observation process 0–50 50–100 >100 m 0–3 3–5

    5–10 min time interval distance band count 2 3 2 1 3 3 1 1 1 5–10 min
  32. 32 The observation process q 0 1 q(r=50) q(r=100) q(r=∞)

    q(r): probability of detecting an individual that sung within a circle of radius r. 0–3 3–5 p 0 1 p(t=3) p(t=5) p(t=10) p(t): probability of an individual singing within t time interval. Time (minutes) Distance (m) 5–10 min 0–50 50–100 >100 m
  33. 33 Parameter estimation p 0 1 „singing” rate Time (minutes)

    q 0 1 Distance (m) Effective Detection Radius E[Y]=NC=(AD)(pq) Removal sampling Distance sampling
  34. 34 Constant singing rate Clay-coloured Sparrow (0.63/min) Blue Jay (0.15/min)

    Cape May Warbler (39 m) American Crow (171 m)
  35. Variable importance 35 species (40%) duration (21%) JDAY (13%) TSSR

    (<1%) p
  36. Variable importance 36 species (40%) duration (21%) JDAY (13%) TSSR

    (<1%) species (65%) radius (23%) TREE & LCC (4%) q p
  37. Variable importance 37 species (66%) radius (21%) duration (4%) JDAY

    (3%) TREE & LCC (1%) TSSR (<0.01%) species (40%) duration (21%) JDAY (13%) TSSR (<1%) species (65%) radius (23%) TREE & LCC (4%) pq q p
  38. u  Location specific variables u  Land cover u  Climate u 

    Survey specific variables u  Years since disturbance Biophysical data Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 38 PCODE METHOD PCODE SS PKEY METHOD DURMETH DISMETH SPECIES BEH ABUND SS PKEY YEAR DATE TIME ROAD VISIT Projects Locations Surveys Count intervals Biophysical data SS PKEY
  39. Biophysical data Boreal Avian Modelling Project Ι Projet de modélisation

    de l'avifaune boréale Ι www.borealbirds.ca 39 •  101 climate covariates •  30 vegetation covariates (Cumming et al. 2014) •  Cross-walked database of Forest Resource Inventory data, including stand age and type (Common Attribute Scheme for Forest Resource Inventory: CASFRI) •  Topographic indices (BAM) •  Wetland mapping (DU) CAS-FRI coverage Downscaled North America climate- change projections: tinyurl.com/ ClimateNA (Stralberg et al. 2015)
  40. Data “packages” for analysis Boreal Avian Modelling Project Ι Projet

    de modélisation de l'avifaune boréale Ι www.borealbirds.ca 40 PCODE METHOD PCODE SS PKEY METHOD DURMETH DISMETH SPECIES BEH ABUND SS PKEY YEAR DATE TIME ROAD VISIT Projects Locations Surveys Count intervals Biophisical data SS (location specific) PKEY (survey/year specific) OVEN JDAY LCC MAT Offset … 0 155 Decid 9.0 0.1 1 155 Mixed 8.8 -0.2 0 155 Conif 9.5 0.3 2 155 Decid 9.1 -0.1 0 155 Agr 10.5 0.2 …
  41. u  Basic species-habitat associations u  Species distribution maps u  Population

    size estimates u  Responses to climate change u  Responses to land use change u  Conservation planning Applications Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 41 Hache et al. in preparation Stralbeg et al. 2015
  42. Integrating Breeding Bird Survey data u  Roadside surveys: u  Edge

    effect à behavioural response u  Edge effect à numerical response u  Vegetation related sound attenuation patterns u  Road surface is 0-density strata à distance sampling assumption violated u  These effects are often species specific u  Data limitations à estimation is difficult u  Solutions: u  Model contrasts (ROAD x Land cover) u  Filter “worst offenders” (wide roads) Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 42
  43. Automated recording units (ARUs) u  Often smaller detection radius à

    smaller sampling area u  Needs calibration (another offset) u  Variation among ARU types, continuous QA/QC required u  Similar to human observers u  Mimicking human observer based counts is expensive: u  Multiple time intervals, counts (based on listening) u  Distance information is hard to get. u  Extended coverage for time-of-year, time-of-day: u  Better describe daily/seasonal behaviour u  Monitor changes in phenology Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 43 Lesson #4: flexibility for change Adapt data harmonization to new technologies
  44. Software tools: R extension packages u  mefa4 u  high performance

    data manipulation tools (sparse matrices) u  detect u  Single visit based occupancy and N-mixture models u  Removal and distance sampling models u  QPAD u  Facilitating offset calculations (and BAM based estimates) u  dclone u  Frequentist inference with Bayesian MCMC, parallel computations u  Wrappers for JAGS, Open/WinBUGS, Stan u  Distributed computing: u  Compute Canada / Westgrid used for running the models u  Computing time: days instead of months. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 44
  45. http://www.borealbirds.ca u  Archive u  Avian database u  Biophysical database u 

    Software tools Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 45 Short course @ NAOC2016 “Introduction to the analysis of messy data”
  46. Further reading Database u  Barker, et al. 2015. Ecological monitoring

    through harmonizing existing data […]. Wildlife Society Bulletin 39:480–487. u  Cumming et al. 2010. Toward conservation of Canada’s boreal forest avifauna […]. Avian Conservation and Ecology 5:8. Methods u  Sólymos, & Lele, in press. Revisiting resource selection probability functions and single-visit methods […]. Methods in Ecology and Evolution u  Matsuoka et al. 2014. Reviving common standards in point-count surveys for broad inference across studies. Condor 116:599–608. u  Sólymos et al. 2013. Calibrating indices of avian density from non-standardized survey data […]. Methods in Ecology and Evolution 4:1047–1058. u  Sólymos et al. 2012. Conditional likelihood approach for analyzing single visit abundance survey data […]. Environmetrics 23:197–205. u  Matsuoka et al. 2012. Using binomial distance-sampling models to estimate the effective detection radius of point-counts surveys […]. Auk 129:268–282. u  Sólymos, P., 2010. dclone: Data Cloning in R. R Journal 2(2):29–37. Applications u  Stralberg et al. 2015. Conservation of future boreal forest bird communities […]. Diversity and Distributions, 21:1112–1128. u  Stralberg et al. 2015. Projecting boreal bird responses to climate change: the signal exceeds the noise. Ecological Applications, 25:52–69. u  Mahon et al. 2014. Does expected future landscape condition support proposed population objectives […]? Forest Ecology and Management 312:28–39. u  Cumming et al. 2014. Climate and vegetation hierarchically structure patterns of songbird distribution […]. Ecography 36:137–151. Boreal Avian Modelling Project Ι Projet de modélisation de l'avifaune boréale Ι www.borealbirds.ca 46