diff --git a/changes/3464.doc.md b/changes/3464.doc.md
deleted file mode 100644
index 155a4575f4..0000000000
--- a/changes/3464.doc.md
+++ /dev/null
@@ -1 +0,0 @@
-Add documentation example for creating uncompressed arrays in the Compression section of the user guide.
diff --git a/changes/3562.misc.md b/changes/3562.misc.md
deleted file mode 100644
index e164ab39f8..0000000000
--- a/changes/3562.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Add continuous performance benchmarking infrastructure.
\ No newline at end of file
diff --git a/changes/3603.bugfix.md b/changes/3603.bugfix.md
deleted file mode 100644
index 37e1da5cb1..0000000000
--- a/changes/3603.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Correct the target bytes number for auto-chunking when auto-sharding.
\ No newline at end of file
diff --git a/changes/3605.misc.md b/changes/3605.misc.md
deleted file mode 100644
index b8c0757b69..0000000000
--- a/changes/3605.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Fix a bug in the test suite that prevented stand-alone example scripts from being tested.
\ No newline at end of file
diff --git a/changes/3619.misc.md b/changes/3619.misc.md
deleted file mode 100644
index 8c36e473b5..0000000000
--- a/changes/3619.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Remove upper bounds on `pytest` and `pytest-asyncio` test dependencies.
\ No newline at end of file
diff --git a/changes/3623.misc.md b/changes/3623.misc.md
deleted file mode 100644
index 4060e55e5f..0000000000
--- a/changes/3623.misc.md
+++ /dev/null
@@ -1,5 +0,0 @@
-This PR contains minor, non-function-altering, changes to use `ZarrFormat` across the repo as opposed to duplicating is with `Literal[2,3]`.
-
-Additionally, it fixes broken linting by using a `Literal[True, False]` type hint for Numpy hypothesis testing, as opposed to `bool`.
-
-Basically improves the typehints and reduces fat-finger error surface area slightly.
diff --git a/changes/3636.misc.md b/changes/3636.misc.md
deleted file mode 100644
index a814160c8b..0000000000
--- a/changes/3636.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-The minimum required version of NumPy is now 2.0.
diff --git a/changes/3648.misc.md b/changes/3648.misc.md
deleted file mode 100644
index 156f8671de..0000000000
--- a/changes/3648.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Fix deprecation of setting a shape on an array directly in ``numpy`` 2.5+.
diff --git a/changes/3655.bugfix.md b/changes/3655.bugfix.md
deleted file mode 100644
index 67d384f00d..0000000000
--- a/changes/3655.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Fixed a bug in the sharding codec that prevented nested shard reads in certain cases.
\ No newline at end of file
diff --git a/changes/3656.misc.md b/changes/3656.misc.md
deleted file mode 100644
index 159f24d072..0000000000
--- a/changes/3656.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Removed *rich* and *mypy* from the `[test]` dependencies, and added a new `[dev]` dependency group that can be used to install all the development dependencies.
diff --git a/changes/3657.bugfix.md b/changes/3657.bugfix.md
deleted file mode 100644
index 1411704674..0000000000
--- a/changes/3657.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Fix obstore _transform_list_dir implementation to correctly relativize paths (removing lstrip usage).
\ No newline at end of file
diff --git a/changes/3658.misc.md b/changes/3658.misc.md
deleted file mode 100644
index f400d97473..0000000000
--- a/changes/3658.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Switch from `pre-commit` to [`prek`](https://github.com/j178/prek) for pre-commit checks.
\ No newline at end of file
diff --git a/changes/3668.feature.md b/changes/3668.feature.md
deleted file mode 100644
index def196ec8a..0000000000
--- a/changes/3668.feature.md
+++ /dev/null
@@ -1,4 +0,0 @@
-Exposes the array runtime configuration as an attribute called `config` on the `Array` and
-`AsyncArray` classes. The previous `AsyncArray._config` attribute is now a deprecated alias for `AsyncArray.config`.
-
-Adds a method for creating a new `Array` / `AsyncArray` instance with a new runtime configuration, and fixes inaccurate documentation about the `write_empty_chunks` configuration parameter.
\ No newline at end of file
diff --git a/changes/3673.misc.md b/changes/3673.misc.md
deleted file mode 100644
index 83643f5d3c..0000000000
--- a/changes/3673.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Benchmark CI now only runs for PRs with the `benchmark` label, reducing CodSpeed credit usage.
diff --git a/changes/3695.bugfix.md b/changes/3695.bugfix.md
deleted file mode 100644
index a7d847e4f1..0000000000
--- a/changes/3695.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Raise error when trying to encode :class:`numpy.dtypes.StringDType` with `na_object` set.
\ No newline at end of file
diff --git a/changes/3700.bugfix.md b/changes/3700.bugfix.md
deleted file mode 100644
index 86acb71d0e..0000000000
--- a/changes/3700.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-CacheStore, LoggingStore and LatencyStore now support with_read_only.
\ No newline at end of file
diff --git a/changes/3702.bugfix.md b/changes/3702.bugfix.md
deleted file mode 100644
index 94a2902567..0000000000
--- a/changes/3702.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Skip chunk coordinate enumeration in resize when the array is only growing, avoiding unbounded memory usage for large arrays.
\ No newline at end of file
diff --git a/changes/3704.misc.md b/changes/3704.misc.md
deleted file mode 100644
index d15d4924e0..0000000000
--- a/changes/3704.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Remove an expensive `isinstance` check from the bytes codec decoding routine.
\ No newline at end of file
diff --git a/changes/3705.bugfix.md b/changes/3705.bugfix.md
deleted file mode 100644
index 2abcb4ee7c..0000000000
--- a/changes/3705.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Fix a performance bug in morton curve generation.
\ No newline at end of file
diff --git a/changes/3706.misc.md b/changes/3706.misc.md
deleted file mode 100644
index 70a0e44c58..0000000000
--- a/changes/3706.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Allow NumPy ints as input when declaring a shape.
\ No newline at end of file
diff --git a/changes/3708.misc.md b/changes/3708.misc.md
deleted file mode 100644
index dce7546c97..0000000000
--- a/changes/3708.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Optimize Morton order computation with hypercube optimization, vectorized decoding, and singleton dimension removal, providing 10-45x speedup for typical chunk shapes.
diff --git a/changes/3710.bugfix.md b/changes/3710.bugfix.md
deleted file mode 100644
index a40ddcee23..0000000000
--- a/changes/3710.bugfix.md
+++ /dev/null
@@ -1 +0,0 @@
-Add a dedicated in-memory cache for byte-range requests to the experimental `CacheStore`.
\ No newline at end of file
diff --git a/changes/3712.misc.md b/changes/3712.misc.md
deleted file mode 100644
index 8fa2f2d2f7..0000000000
--- a/changes/3712.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Added benchmarks for Morton order computation in sharded arrays.
diff --git a/changes/3713.misc.md b/changes/3713.misc.md
deleted file mode 100644
index 9b0680dfc0..0000000000
--- a/changes/3713.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Vectorize get_chunk_slice for faster sharded array writes.
diff --git a/changes/3717.misc.md b/changes/3717.misc.md
deleted file mode 100644
index 5fed76b2b7..0000000000
--- a/changes/3717.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Add benchmarks for Morton order computation with non-power-of-2 and near-miss shard shapes, covering both pure computation and end-to-end read/write performance.
diff --git a/changes/3721.misc.md b/changes/3721.misc.md
deleted file mode 100644
index c170712882..0000000000
--- a/changes/3721.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Adds synchronous (non-async) encoding and decoding methods to CPU-bound codecs. This is necessary for performance optimizations based on avoiding `asyncio` overhead. These new methods are described by a new protocol: `SupportsSyncCodec`.
\ No newline at end of file
diff --git a/changes/3728.misc.md b/changes/3728.misc.md
deleted file mode 100644
index a3cbb8d3f0..0000000000
--- a/changes/3728.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-Move development dependencies (`test`, `remote_tests`, `docs`, `dev`) from optional dependencies to [dependency groups](https://packaging.python.org/en/latest/specifications/dependency-groups/). This may cause breakage for anyone who used e.g. `pip install zarr[test]` to get access to test dependencies. To install these dependency groups from a local checkout, use `pip install --group <name>` (pip 25.1+) or `uv run --group <name>`.
\ No newline at end of file
diff --git a/changes/3769.doc.md b/changes/3769.doc.md
deleted file mode 100644
index f622ac525f..0000000000
--- a/changes/3769.doc.md
+++ /dev/null
@@ -1 +0,0 @@
-Add AI-assisted code policy to the contributing guide.
diff --git a/changes/3778.misc.md b/changes/3778.misc.md
deleted file mode 100644
index 17f26666ed..0000000000
--- a/changes/3778.misc.md
+++ /dev/null
@@ -1 +0,0 @@
-`Group.tree()` no longer requires the `rich` dependency. Tree rendering now uses built-in ANSI bold for terminals and HTML bold for Jupyter. New parameters: `plain=True` for unstyled output, and `max_nodes` (default 500) to truncate large hierarchies with early bailout.
diff --git a/docs/release-notes.md b/docs/release-notes.md
index 25ebdb8edf..b974ae10c7 100644
--- a/docs/release-notes.md
+++ b/docs/release-notes.md
@@ -2,6 +2,42 @@
 
 <!-- towncrier release notes start -->
 
+## 3.1.6 (2026-03-19)
+
+### Features
+
+- Exposes the array runtime configuration as an attribute called `config` on the `Array` and
+  `AsyncArray` classes. The previous `AsyncArray._config` attribute is now a deprecated alias for `AsyncArray.config`. ([#3668](https://github.com/zarr-developers/zarr-python/issues/3668))
+- Adds a method for creating a new `Array` / `AsyncArray` instance with a new runtime configuration, and fixes inaccurate documentation about the `write_empty_chunks` configuration parameter. ([#3668](https://github.com/zarr-developers/zarr-python/issues/3668))
+- Adds synchronous methods to stores that do not benefit from an async event loop. The shape of these methods is defined by protocol classes to support structural subtyping. ([#3725](https://github.com/zarr-developers/zarr-python/pull/3725))
+- Fix near-miss penalty in `_morton_order` with hybrid ceiling+argsort strategy. ([#3718](https://github.com/zarr-developers/zarr-python/pull/3718))
+
+### Bugfixes
+
+- Correct the target bytes number for auto-chunking when auto-sharding. ([#3603](https://github.com/zarr-developers/zarr-python/issues/3603))
+- Fixed a bug in the sharding codec that prevented nested shard reads in certain cases. ([#3655](https://github.com/zarr-developers/zarr-python/issues/3655))
+- Fix obstore `_transform_list_dir` implementation to correctly relativize paths (removing `lstrip` usage). ([#3657](https://github.com/zarr-developers/zarr-python/issues/3657))
+- Raise error when trying to encode :class:`numpy.dtypes.StringDType` with `na_object` set. ([#3695](https://github.com/zarr-developers/zarr-python/issues/3695))
+- `CacheStore`, `LoggingStore` and `LatencyStore` now support with_read_only. ([#3700](https://github.com/zarr-developers/zarr-python/issues/3700))
+- Skip chunk coordinate enumeration in resize when the array is only growing, avoiding unbounded memory usage for large arrays. ([#3702](https://github.com/zarr-developers/zarr-python/issues/3702))
+- Fix a performance bug in morton curve generation. ([#3705](https://github.com/zarr-developers/zarr-python/issues/3705))
+- Add a dedicated in-memory cache for byte-range requests to the experimental `CacheStore`. ([#3710](https://github.com/zarr-developers/zarr-python/issues/3710))
+- `BaseFloat._check_scalar` rejects invalid string values. ([#3586](https://github.com/zarr-developers/zarr-python/issues/3586))
+- Apply drop_axes squeeze in partial decode path for sharding. ([#3763](https://github.com/zarr-developers/zarr-python/issues/3763))
+- Set `copy=False` in reshape operation. ([#3649](https://github.com/zarr-developers/zarr-python/issues/3649))
+- Validate that dask-style chunks have regular shapes. ([#3779](https://github.com/zarr-developers/zarr-python/issues/3779))
+
+### Improved Documentation
+
+- Add documentation example for creating uncompressed arrays in the Compression section of the user guide. ([#3464](https://github.com/zarr-developers/zarr-python/issues/3464))
+- Add AI-assisted code policy to the contributing guide. ([#3769](https://github.com/zarr-developers/zarr-python/issues/3769))
+- Added a glossary. ([#3767](https://github.com/zarr-developers/zarr-python/issues/3767))
+
+### Misc
+
+- [#3562](https://github.com/zarr-developers/zarr-python/issues/3562), [#3605](https://github.com/zarr-developers/zarr-python/issues/3605), [#3619](https://github.com/zarr-developers/zarr-python/issues/3619), [#3623](https://github.com/zarr-developers/zarr-python/issues/3623), [#3636](https://github.com/zarr-developers/zarr-python/issues/3636), [#3648](https://github.com/zarr-developers/zarr-python/issues/3648), [#3656](https://github.com/zarr-developers/zarr-python/issues/3656), [#3658](https://github.com/zarr-developers/zarr-python/issues/3658), [#3673](https://github.com/zarr-developers/zarr-python/issues/3673), [#3704](https://github.com/zarr-developers/zarr-python/issues/3704), [#3706](https://github.com/zarr-developers/zarr-python/issues/3706), [#3708](https://github.com/zarr-developers/zarr-python/issues/3708), [#3712](https://github.com/zarr-developers/zarr-python/issues/3712), [#3713](https://github.com/zarr-developers/zarr-python/issues/3713), [#3717](https://github.com/zarr-developers/zarr-python/issues/3717), [#3721](https://github.com/zarr-developers/zarr-python/issues/3721), [#3728](https://github.com/zarr-developers/zarr-python/issues/3728), [#3778](https://github.com/zarr-developers/zarr-python/issues/3778)
+
+
 ## zarr 3.1.5 (2025-11-21)
 
 ## Bugfixes