diff --git a/CHANGELOG.md b/CHANGELOG.md
index 7f821f6b..df0774e7 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,6 +12,7 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - ES-CL: the ITACyL server is https-only, the 2025 shapefiles sit in province subfolders, and C_REFREC is the identifier
 - A test refuses a fixture above 5 MB, committed or merely lying in the fixture folder, because a failing convert test downloads the real source there
 - ES-MD: find RECINTO.shp wherever the archive puts it
+- DE-NDS: give the collection an id (the row index), which it was published without
 - DE-BB: ref_ident holds the FLIK (field block reference), not a farmer, and the shapefile is cp1252
 - Update vecorel-cli to v0.2.17:
   - GeoJSON is read as UTF-8 as the format mandates, instead of the platform locale (cp1252 on Windows mangled umlauts)
diff --git a/fiboa_cli/datasets/de_nds.py b/fiboa_cli/datasets/de_nds.py
index 732a6d60..92f91e80 100644
--- a/fiboa_cli/datasets/de_nds.py
+++ b/fiboa_cli/datasets/de_nds.py
@@ -34,7 +34,11 @@ class Converter(AdminConverterMixin, AddHCATMixin, FiboaBaseConverter):
 
     # https://www.sla.niedersachsen.de/download/141235/Verzeichnis_Nutzungscodes.xlsx
     ec_mapping_csv = "de.csv"
+    # The shapefile has no unique key: 880,581 features share 874,222 distinct
+    # (FLIK, SCHLAGNR) pairs, so the row index is the only identifier available.
+    index_as_id = True
     columns = {
+        "id": "id",
         "geometry": "geometry",
         "FLIK": "flik",
         "SCHLAGNR": "subfield_id",