Mapping the functional connectivity of ecosystem services supply across a regional landscape

  1. Rachel D Field  Is a corresponding author
  2. Lael Parrott
  1. The Okanagan Institute for Biodiversity, Resilience and Ecosystem Services (BRAES) Irving K. Barber Faculty of Science, Department of Biology, University of British Columbia, Canada
10 figures, 8 tables and 1 additional file


Schematic and definitions for ES supply areas and functional connections in the case study landscape.
Maps showing the original, full-extent of distribution and weighting for ES supply areas in the case study landscape, including (a) plant food agriculture, (b) waterflow regulation, and (c) landscape aesthetics (Field et al., 2017).
Distribution and weighting of top-50%-valued ES supply areas and functional connections on the case study landscape.

Insets show (a) all top-value supply areas and links; top-value supply areas for (b) plant food agriculture (PA), (c) waterflow regulation (WF), and (d) landscape aesthetics (LA); overlapping connections from (e) LA to WF, (f) PA to LA, (g) PA to WF, (h) WF to LA, and (i) WF to PA; and topographic connections from (j) WF to WF, (k) WF to PA, and (l) WF to LA. Adjacent circular coxcomb charts represent the proportion of ES supply and link areas covered by major LULC types. LULC types are color-coded and include forests, agriculture, rocks/exposed areas, residential areas, shrubs, grasslands, urban areas, water, wetlands, and areas with unknown use and/or cover.

Distribution and weighting of link values amalgamated across all eight (8) overlapping and topographic link types across the case study landscape.
Location of the case study landscape in southern interior ‘Okanagan’ region of British Columbia, Canada.

Major waterbodies, elevation (masl), and the most populous cities and towns in Okanagan regional districts are indicated.

General approach guidelines for mapping the functional relationships between multiples ES within a comprehensive ES assessment framework (e.g. Value of Nature to Canadians Study Taskforce, 2017).
Appendix 2—figure 1
Map of the distribution of major LULC types across the case study region.

General LULC types in the region include, in decreasing order of area: forests (16,281 km2), grasslands (1482 km2), natural parks (2,403 km2; NB: contains several of the other listed LULC categories), shrubs (1349 km2), agricultural (842 km2), lakes (599 km2), urban residential (220 km2), rural residential (220 km2), wetlands (182 km2), rock/rubble (161 km2), exposed land (113 km2), manicured parks (45 km2), rivers (38 km2), commercial (23 km2), industrial (23 km2), urban institutional (16 km2), and reservoirs (6 km2; Field et al., 2017).

Appendix 5—figure 1
Boxplots summarizing percent overlap of buffered BC TRIM data with buffered topographic links between (a) WF; (b) WF and PA; and (c) WF and LA supply areas produced by functional connectivity and null models for Bellevue Creek validation analyses.
Appendix 7—figure 1
Map of major watersheds and sub-basins within and surrounding the case study landscape in southern interior British Columbia, Canada.
Appendix 7—figure 2
Schematic outlining steps for the creation of topographic ES corridors from each origin ES supply area to downslope supply areas.

(a) An initial line feature resulting from a least cost path (LCP) analysis, that is, from the origin to the goal point, amalgamated with a line from the goal point to a downstream influential landscape feature (ILF). (b) Types of LCP segment deletions addressed, including (red) segments overlapped by ES supply area polygons, (i) segments from origin points external to origin ES polygon, and (ii) segments flowing between two areas of an intersected (i.e. non-origin) ES supply area. (c) Resulting ES flow corridors after deletions, including feedbacks to origin ES supply area (4), flows to downslope ES supply areas (1) and (3), and flows to downstream ILF areas (2).


Table 1
Rationale behind functional connection mechanisms, directionality, and weighting between top-value ES supply areas.

Directionality is represented by the top row as originating (or ‘source’) ES supply areas; and the left column as recipient (or ‘sink’) ES supply areas (PA, WF, or LA). Functional connection mechanisms are distinguished by Link Type (i.e., Overlap or Topographic), and their definitions and weighting rationale are provided in matrix cells. If no functional connection exists from one ES to another, the matrix cell indicates that there is ‘None’ in either direction, or that there is a connection in the ‘Other direction’. Superscripts denote the following references: (1) Crossman et al., 2013; (2) Power, 2010; (3) Daniel et al., 2012; (4) Zhang et al., 2007; (5) Bennett et al., 2009; (6) Poff et al., 1997; (7) Burmil et al., 1999; (8) Raudsepp-Hearne et al., 2010; (9) Nicholls and Altieri, 2012.

Supply areas
linked from →
linked to ↓
Link TypePlant Agriculture (PA)
supply area weight
: potential PA crop area (ha)
Waterflow Regulation (WF)
supply area weight:
summed WF model value
Landscape Aesthetics (LA)
supply area weight:
area (ha) x LA model value
Supply AreasPlantAgricultureOverlapNoneWF regulation on PA croplands4,5link weight: summed
WF model supply area values within PA supply area
Other direction
TopographicWF regulation downslope4,5link weight: summed WF model values along LCP pathway from WF to PA supply areaNone
Waterflow RegulationOverlapPA croplands providing WF regulation1,2link weight: all summed WF model values within entire PA supply areaNoneLA areas providing WF regulation8,9link weight: summed WF model values within LA supply area
TopographicOther directionWF regulation downslope6link weight: summed WF model values along LCP pathway from WF1 to WF2 supply areaOther direction
Landscape AestheticsOverlapPA cropland providing LA3link weight: summed LA model values within PA supply areaWF regulation on LA areas7link weight: summed WF model supply area values within LA supply areaNone
TopographicNoneWF regulation downslope7link weight: summed WF model values along LCP pathway from WF to LA supply area
Key resources table
Reagent type
or resource
DesignationSource or
IdentifiersAdditional information
Software, algorithmR (v.3.6.2)R Development Core Team, 2013RRID:SCR_001905
Software, algorithmArcGIS (v.10.7.1)ESRI, 2011RRID:SCR_011081
Appendix 3—table 1
Percent of LULC overlapped by each ES supply area.
High value supply areas (Top 50%)
Study AreaPAWFLA
LULC TypeTotal area (ha)215800112,606922,4251456241
% of study areaArea (ha)% of supply area% of total LULCArea (ha)% of supply area% of total LULCArea (ha)% of supply area% of total LULC
Forest (incl. parks)1,654,21576.6530.850.241.86e-03725852.3078.6943.881354434.0093.0181.88
Rock / exposed27,3731.272.78e-04005387.800.5819.68779.820.052.85
Appendix 3—table 2
Percent of LULC overlapped by each topographic link.
Topographic links
Study AreaWF —> PAWF —> WFWF —> LA
LULC TypeTotal area (ha)21580014,07944,4494,695
% of study areaArea (ha)% of link area% of total LULCArea (ha)% of link area% of total LULCArea (ha)% of link area% of total LULC
Forest (incl. parks)1,654,21576.652205.9454.080.1319069.7742.901.152688.5657.260.16
Rock / exposed27,3731.2746.701.140.17671.991.512.45112.702.400.41
Appendix 3—table 3
Percent of LULC overlapped by each overlapping links.
Overlapping links
Study AreaPA —> LAPA —> WFWF —> PAWF —> LALA —> WF
LULC TypeTotal area (ha)21580018.570244747.303004672.03900684,301685,377
% of study areaArea of overlap (ha)% of link area% of LULC areaArea (ha)% of link area% of total LULCArea (ha)% of link area% of total LULCArea (ha)% of link area% of total LULCArea (ha)% of link area% of total LULC
Forest (incl. parks)1,654,21576.652.4428.471.47e-0431.280.661.89e-0329.110.621.76e-03621652.4090.8437.58624572.0091.1337.76
Rock / exposed27,3731.270.050.531.64e-045.400.
Appendix 5—table 1
Summary and validation analyses data for topographic links produced by functional connectivity and null models for Bellevue Creek.
Functional connectivity modelNull model
Number of links336213563411
Total link area (km2)9.0518.235.1223.2120.063.58
Total overlap area (km2) with BC TRIM data5.0812.223.633.803.880.99
Median % overlap with BC TRIM data426769122529
Wilcoxon p-value1.59e-035.29e-092.09e-04---
Appendix 6—table 1
Summary of data sources and rationale for water flow regulation LULC mapping prioritization (from Field et al., 2017).
Map rankData sourceSourceFile name(s)Year(s)TypeFields/ClassesDetails
1Agricultural Land Use Inventory (ALUI)BC Ministry of Agriculture (data user agreement)2013 + 2014Cover = ATVC (Anthropogenic Terrestrial Vegetated Cultivated)Ground-truthed data, not strictly based on remotely sensed image interpretation. Used ATVC category (Agricultural Terrestrial Vegetated Cultivated) class to identify specific agricultural categories – Orchards, Vineyards, Field, Other (2013/14).
Agriculture - OrchardsCovType = C200 (Tree Fruits)
Agriculture - FieldCovType = C320 (Berries) or CovType = C500 (Vegetables)
Agriculture - VineyardsCovGroupType = C311 (Grapes)
Agriculture - OtherCovType = C100 (Cereals, Grains, Oilseeds), C400 (Forage, Pasture), C600 (Floriculture), C710 (Specialty), C720 (Turf), C730 (Nut Trees), C810 (Nursery), C820 (Tree Plantations), C900 (Other)
2Okanagan Wetlands Strategy (OWS)Ecoscape Consulting Ltd.2009 + 2010 + 2011 + 2014WetlandsCompiled regional database. Includes data from the following sources: City of Kelowna WIM (2009);BC Freshwater Atlas (2014); MOE Wetland Inventory Project (2009); Alkali-Saltgrass Herbaceous Vegetation Community Assessment (2011); SEI/TEM for the study area; SHIM (BX Creek, NORD, Vaseux Creek and Oliver, Prairie Creek, Winfield Creek, various dates); FIM (Kalamalka, Wood, Mabel, Mara, Okanagan, Osoyoos, various dates); LRIM (Lower Shuswap River Inventory and Mapping, 2010); Ducks Unlimited (DU) data (various sources).
3BC Freshwater Atlas (FWA)BC Ministry of Environment (Data Distribution Service)2009LakesLakes
RiversRivers (polygons)
Man Made WaterbodiesReservoirs
4Municipal and Regional ZoningRDNO Open DataRDNO2017Urban - CommercialUrban - InstitutionalUrban - IndustrialUrban - ResidentialRural - ResidentialParksSimplified and compiled zoning data from all municipal jurisdictions in the study area. Used supporting by-law documentation to limit local codes to categories: Urban – Commercial, Urban – Industrial, Urban – Institutional, Urban – Residential, Rural – Residential. (Parks were also identified in the zoning data but not incorporated until later in the model).
AreaBZoning2017Zoning and Zoning_Des
AreaCZoning2017Zoning and Zoning_Des
AreaDZoning2017Zoning and Zoning_Des
AreaEZoning2017Zoning and Zoning_Des
AreaF_OCP2017Zoning and Zoning_Des
Coldstream_Zoning2017Zoning and Zoning_Des
Enderby_Zoning2016Zoning and Zoning_Des
Lumby_Zoning2012Zoning and Zoning_Des
City of Vernon Open DataVernon_Zoning2004Zoning_Val and Zoning_Des
RDCO (data user agreement)RDCO2017
City of Kelowna Open DataKelowna_Zoning2017ZoningCode
RDOS Open DataRDOS Zoning2017Designation
5Sensitive Ecosystems Inventory (SEI)RDCO Open DataRDCO_SEI2001–2015ForestBW (Broadleaf Woodlands); MF (Mature Forest); OF (Old Forest); WD (Coniferous Woodlands)RDCO SEI coverage for RDCO portion of study area, provincial SEI coverage for north and south regional districts. Only selected polygons that contained sensitive ecosystems in the primary ground cover field.
ShrubSV (Sparsely Vegetated)
GrasslandGR (Grasslands)
BC Ministry of Environment (Data Distribution Service)SEI2010ForestBW (Broadleaf Woodlands); MF (Mature Forest); OF (Old Forest); WD (Coniferous Woodlands)
ShrubSV (Sparsely Vegetated); AS (Antelope-Brush Steppe); SS (Sagebrush Steppe)
GrasslandGR (Grasslands); DG (Disturbed Grasslands)
6GrasslandsGrasslands Conservation Council (GCC)GCC Grasslands2017Compiled grasslands data.
7Vegetative Resource Inventory (VRI)BC Ministry of Forest, Lands and Natural Resource Operations (Data Distribution Service)2002–2017ForestBCLCS Level 1 (BCLCS_LEVE) = V (Vegetated); BCLCS Level 4 (BCLCS_LEVE3) – TC (Treed Coniferous); TB (Treed Broadleaf); TM (Treed Mixed)VRI categories: Vegetated – Treed, Vegetated – Shrub, or Vegetated – Grasslands
8ParksSee notes on zoning dataZoning Data (all jurisdictions)2004–2017Park_ManicuredSee notes on zoning dataParks were subdivided into two classes – Natural and Manicured. Natural parks included provincial parks and regional parks. Manicured parks included parks from urbanized areas (Penticton, Vernon, Armstrong, RDCO). This included all of RDCO since this zoning mainly covers urban and urban fringe areas. Manicured parks also included any zoned parks in the study area that were coincident with DMTI golf course points data, even if these fell outside of the urban centres.
Data Distribution ServiceProvincial Parks2017Park_Natural
Data Distribution ServiceConservation Lands2013Park_Natural
RDCO Open DataRDCO Parks2017Park_Natural
9Vegetative Resource Inventory (VRI)BC Ministry of Forest, Lands and Natural Resource Operations (Data Distribution Service)2002–2017Rock/RubbleBCLCS Level 4 = Rock/Rubble (1447)VRI categories: Rock/Rubble, Snow/Ice, Exposed Land, Water (Lakes, Rivers, Reservoirs)
RiversBCLCS Level 4 = Water (1069)
LakeBCLCS Level 4 = Water (1069)
Exposed LandBCLCS Level 4 = Exposed Land (865)
ReservoirBCLCS Level 4 = Water (1069)
Appendix 6—table 2
Calculated and assumed NDVI results, infiltration rates (0%–100%), and qualitative aesthetic valuations for each LULC type.

Blank (grey) NDVI values were not calculated; therefore, and infiltration rate was assumed (from Field et al., 2017).

NDVI values
LULCArea (ha)MinMaxRangeMeanStdvInfiltration %Aesthetic Value
Urban - Industrial2,302–0.23290.60990.84280.16140.104019No Data
Urban - Commercial2,319–0.30940.61940.92870.20000.136932No Data
Urban - Residential21,952–0.15940.61420.77360.24100.086845318
Urban - Institutional1,566–0.20740.63490.84230.25240.130948No Data
Exposed Land11,272–0.23320.67920.91240.25280.119748No Data
Rural - Residential21,952–0.15310.63260.78570.30250.085164318
Agriculture - Field510.04940.62970.58030.31870.101769
Unknown1,2910.00000.60840.60840.32640.090872No Data
Agriculture - Vineyards4,3350.00610.61890.61280.33110.074273275
Agriculture - Other33,930–0.06360.67040.73400.36370.123183
Agriculture - Orchards5,802–0.03640.63490.67130.41640.0943100
Park - Manicured4,534100205
  1. * = assumed NDVI value for paved roads based on impervious threshold (

  2. No Data = no value applied in mapping due to lack of data.

  3. Gray cells = not included in analyses.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Rachel D Field
  2. Lael Parrott
Mapping the functional connectivity of ecosystem services supply across a regional landscape
eLife 11:e69395.