Habitat network — Recipe (QGIS)

Tool-bound implementation of the Habitat network (Task) Task Schema in QGIS 4.2. Worked example: Habitat network at Ledreborg (Example).

Data you need

Role	Form	Level of measurement	Notes
Habitat patches	polygons	nominal type (optional)	e.g. from Nature content layer (Task)
Connection distance	a number	ratio (metres)	from the species or species group in question
Patches with cluster (output)	polygons	nominal cluster_id; ratio counts and areas

Build the clusters (Task Schema steps 3–5)

Buffer (Tool - QGIS) on the patches:

• Distance: half the connection distance
• Dissolve result: ticked
• Keep disjoint results separate: ticked

This one dialog realises three Operators — dilation, dissolve and explode — and gives one feature per cluster. Then:

1. Renumber feature ids (Tool - QGIS) on the clusters. Without it, saving to a GeoPackage fails (UNIQUE constraint failed), because the clusters share fid values.
2. Field calculator (Tool - QGIS): an integer field cluster_id = $id.

Give each patch its cluster (step 6)

Join attributes by location (Tool - QGIS): Join to features in the patches, By comparing to the clusters, predicate intersect, field cluster_id, join type Take attributes of the first matching feature only (one-to-one).

Measure and classify (steps 7–8)

Field calculator (Tool - QGIS) on the joined patches:

• n_in_cluster (integer) = count($id, group_by:="cluster_id")
• cluster_area (decimal) = sum($area, group_by:="cluster_id")

Style the layer with Categorized on the expression

CASE WHEN "cluster_area" = maximum("cluster_area") THEN 'Main network'
     WHEN "n_in_cluster" = 1 THEN 'Isolated patch'
     ELSE 'Smaller cluster' END

Repeat with at least one other connection distance, as the Task’s choices recommend.

Checked in QGIS 4.2.2, October 2026.