Polyscape

Tim Pagella, Fergus Sinclair, Eric Opiya

Spatial dimensions of ecosystem services

Ecosystem services often involve stocks and flows of material or individuals across landscapes: water, soil, carbon, organisms

Generation Reception

Haines-Young, 2009

Haines-Young, R. and Potschin, M., (2009). The links between biodiversity, ecosystem services and human well-being.

Mapping requirements for assessing ecosystem service requirements (Pagella and Sinclair, in review. Landscape Ecology)

50% of studies in last two years

National scale Landscape scaleover 1000 km2 but sub-national

Local scale.immediate landscape scales (10-1000 km2)

Snapshots rather than changes to ecosystem service provision?

Little evidence of participatory map development

What methodologies are available now?

Flow Pathways?

Fitness of mapping tools for managing ecosystem service provision

• Large scale, coarse resolution– Can’t link field and farm decisions to ES

• Few ecosystem services mapped, little explicit treatment of interactions amongst them

• Arbitrary geographical boundaries that focus on generation– Ignore reception– Different ES may need different boundaries

• Stakeholders rarely participating in map generation– ES providers– Intermediaries– ES receivers

Boundaries

Polyscape - a multiple criteria GIS toolbox

• Designed as a negotiation tool not as a prescriptive model

• Works at local scales with resolution appropriate for field decisions considering small (10 km2) to medium (1000 km2) landscape contexts

• Embraces the reality of ‘data sparse’ environments, using national scale digital elevation, land use/cover and soil data in the first instance

Participation and Knowledge Exchange

Sources of dataData set Type Resolution Notes

CCW 1980s Phase 1 Land use 10m2Data drawn from

field survey 1980sCCW 2009

Phase 1Land use 5m2

Remote sensed data 2009.

NSRI Soilscapes Soil 1 km2 Farewell et al., 2011

OS Land PROFILE DTM 10m2

EA Flood risk Flood risk Im2-10m2Uses DTM and

LIDARCore and Focal

Habitat NetworkHabitatnetwork

20m2 Watts et al., 2008

What single layer colours mean?

Areas with priority for maintaining current land use

Areas with moderate or unknown potential for land use change

Areas with high priority for land use change

High

Moderate

Moderate

High

Farm productivity layer (Pontbren)

• The base layer (represents farmer’s livelihood)• Difficult to represent all decisions

(idiosyncratic behaviour) • Inputs are digital elevation, soil type, and

critical slope values• The algorithm categorises land value

according to its degree of waterlogging, fertility and slope

Farm productivity layer – Marginal land identified in green – make interventions on wet and sloping areas not flat and dry (red); much of the catchment negotiable (orange).

Woodland habitat connectivity at Pontbren

Plant trees to enlarge existing woodland networks (green); not where there are trees or other key habitats already (red); large area where farmers may wish to plant trees that have low habitat value (orange)

© CCW

Water regulation map for Pontbren

Opportunities for tree planting because high flow (grassland with > 500 m2 contribution, green); Moderate Flow 100 – 500 m2; negligible flow, with <100 m2 contribution (orange); already has trees or other flow sinks (red).

Combining layers in Polyscape-2 -1 0 1 2

-12+ = 1

Layer A

Layer BCombined layer

Numerical score allocated to each zone

Additive approach taken to combining layers

Example

Trade-off layer

Combining layers in Polyscape

-12+ = -1

Layer A

Layer BCombined layer

1. A ‘Conservative’ approach:

-12+ = 2

Layer A

Layer BCombined layer

2. A ‘Opportunistic’ approach:

What trade-off layer colours mean?

Agriculture Surface runoff Habitat connectivity

Trade off maps

Pontbren (1000 ha)

Trade offs - Pontbren

Basic soil fertility map – based on NSRI Soilscape data

The Elwy Catchment - 230 km2

Issues with water quality/ sediment loads

Agricultural impact – farmers bravado

Slope threshold 15o

Agricultural impact – farmer reality?

Slope threshold 12o

Potential utilisation of marginal land?

Marginal land not agriculture8%

Woodland48%

Marginal land in

agriculture44%

MeirchionMar-ginal land not

agri-culture

22%

Woodland16%

Marginal land in agriculture62%

Gallen

Based on 10m2 DTM and 1980s land use data.

Red areas indicate sinks (woodland, wetland or depressions)

light green indicates high flow areas

Water regulation - Elwy

Agri-environment schemes - Elwy

Using Google Earth to display Polyscape layers

Jeldu

Land use data does not capture current tree cover

Google Earth

Old forested area on steep slopes converted to fields – High erosionRiparian planting high

value for timber

25-40% of the tree material leaves the system for sale as fuel or fibre

Eucalyptus planted near road infrastructure

No cultural services initially

Flat Plateau

Steep valley sides

Road to Gojo

Pathways

River system

Road to Gojo

Originally forested, now rapidly degrading wheat fields (high erosion)

Mosaic of Eucalyptus (especially near roads and rivers), wheat (poorer farmers) and Potato (wealthier farmers)

Road to Gojo

Agriculture

Road to Gojo

Timber

Road to Gojo

Water

Road to Gojo

Trade offs

Neutral

Trade off

Minor Trade off

Opportunity for change

Scaling UpSystem boundaries vary with ecosystem service

Potato farmer removing eucalyptus

Remnant tree cover

Reduced base flows in streams with Eucalyptus riparian areas

Key points

• The mapped output needed to integrate across scales from field to ‘landscape’.

• The output needed to be spatially explicit • Multiple services need to be mapped together• To be useful in any landscape the tool must be able

to utilise generally available data in the first instance. • Integrate scientific evidence with local knowledge. • The output should support the implementation of

policy at landscape scales.