Pass zarr encoding through StoreToZarr()

[thodson-usgs]

This PR will allow passing ds.to_zarr() encodings to StoreToZarr(), which would allow custom compressors in recipes.
However, I'm hitting an error with my test recipe, which has me stumped for the moment...

line 49, in _store_data
    raise ValueError(
ValueError: Region (slice(0, 1, None),) does not align with Zarr chunks (2,). [while running 'Create|OpenURLWithFSSpec|OpenWithXarray|StoreToZarr/StoreToZarr/StoreDatasetFragments/Map(store_dataset_fragment)']

Here's my test recipe.
On the recommendation of @rabernat, I was attempting to use pcodec in StoreToZarr(), but for now, I'm testing with
zarr.Blosc(cname="zstd", clevel=3).

#!/bin/bash
REPO=https://github.com/hytest-feedstocks/gpcp-pcodec-feedstock.git
CONFIG_FILE=config.py
RECIPE_ID=gpcp-from-gcs
REF=recipe-dev
JOB_NAME=test

pangeo-forge-runner bake --repo=$REPO -f=$CONFIG_FILE --Bake.recipe_id=$RECIPE_ID --Bake.ref=$REF --Bake.job_name=$JOB_NAME --prune

Any ideas?

[thodson-usgs]

... I'm an idiot. My recipe was trying to compress "time" not "precip". No wonder.

I've tested this on local runner and the compression works. Now to write a test.

[cisaacstern]

Thanks so much @thodson-usgs, just approved the tests! Generally LGTM!

[cisaacstern]

All tests passing, so I'll merge, thanks so much @thodson-usgs !