Import White/Salinas cover crop + compost dataset into template

[dlebauer]

Curate the White/Salinas dataset into the template, including site metadata, treatment structure, observations, and management-event history as far as recoverable.

Dataset: https://doi.org/10.1016/j.dib.2020.106481

• Create site, plot, treatments
• Populate citation, site, treatment, variable, and methods
• Enter all usable SOC and flux observations in long format.
• Extract management events:
  • planting, cover crop, compost, fertilizer, tillage, harvest.
  • if no date provided, exclude dates
• Flag missing/inferred management explicitly.
• Document the data entry method as a markdown file; submit as PR to uncertainty repository google doc
  • include a section on challenges / assumptions made during ingestion, and suggest changes required to schema or method

[dlebauer]

Updated Guidance

Description

Curate the White / Salinas dataset into the CCMMF calibration / validation template.

Primary sources:

Data
White, K.E., Brennan, E.B., Cavigelli, M.A., 2020a. Soil carbon and nitrogen data during eight years of cover crop and compost treatments in organic vegetable production. Data in Brief 33, 106481. https://doi.org/10.1016/j.dib.2020.106481

• raw data are in supplementary materials file White_Data in Brief_Supplemental Tables.xlsx

Paper
White, K.E., Brennan, E.B., Cavigelli, M.A., Smith, R.F., 2020b. Winter cover crops increase readily decomposable soil carbon, but compost drives total soil carbon during eight years of intensive, organic vegetable production in California. PLOS One 15, e0228677. https://doi.org/10.1371/journal.pone.0228677

Additional information
This experiment has produced many publications; use them only where necessary to fill critical gaps. For MVP, stick to Data and Paper if possible.

Scope

Follow the calibration / validation data curation protocol.

• Use White_Data in Brief_Supplemental Tables.xlsx as the primary source
• Curate all 8 systems in XLSX file OR (constrain to 5 systems reported in paper AND document why other 3 were excluded; perhaps they are not relevant)
• Use the paper text for metadata, methods, and management context as needed; use other sources only where necessary and cite appropriately

Out of scope

• Statistical summaries presented in the paper when the same information is available as raw individual observations
• CSVs checked into the repository until the spreadsheet + README/doc have been reviewed and an MVP dataset has been collected

Specific guidance

• Site and treatment names should be recognizable from the original publication, e.g. site salinas_socs
• Do not encode treatments as sites
• Map treatments clearly to the original White system IDs (1–8)
• Treatment descriptions should be based on experimentally manipulated factors

Observations

• Enter target observations from this dataset in long format

• Include, where available:

  • SOC concentration
  • SOC stock
  • bulk density

• If summary values are retained, include n and statistic(s); only retain if replicate-level observations are not available.

• Focus only on required variables

• Do not collect cumulative values (e.g. cumulative root biomass)

• If other information seems useful, ask before collecting it

• Optionally keep a table listing additional information available for this study but not collected

Management events

Focus on what happened. The treatment / experimental design should be recoverable from the sequence of management events.

• What crop was planted and when

  • the presence of a cover crop should be encoded by the presence / absence of a cover crop planting event

• Fertilizer application rate

• Compost type and application

  • record C:N ratio or %C and %N if available

• Winter fallow vs. cover crop years

• Harvest

• If a management attribute is not reported, leave it missing; these will be gap-filled programmatically

• Ask if there is other management information that might be worth collecting

Documentation

• Update README

• Document:

  • data not collected
  • assumptions that could not be avoided
  • open questions
  • table / column definitions
  • units
  • unresolved gaps

• Document discrepancies or unresolved ambiguities

• Document data of interest from this study that are not collected

Definition of done

This issue is complete when:

• all 8 systems are curated, or exclusion is documented
• site / treatment / replicate are cleanly separated
• observations from the supplemental XLSX are entered in long format
• there are no separate rows for statistical summaries when individual observations are collected
• README / documentation reflects the current dataset accurately

[AritraDey-Dev]

@dlebauer Two decisions before I proceed:

1. Systems 3, 6, 8
   The Zotero XLSX (Data in Brief) has all 8 systems, but the PLoS ONE paper only reports 5 (Systems 1, 2, 4, 5, 7). Should I include all 8 in the template, or stick with the 5 from the paper and document why the other 3 are excluded?

2. Variable scope
   The guidance says focus on SOC concentration, SOC stock, and bulk density. I also have block-level data for:

• Total N concentration + stock (paired with SOC in Table 1)
• POXC stock (active C indicator, Table 3)
• Nitrate N concentration (also in Table 1)

Should I keep any of these, or strictly SOC + bulk density only?

[dlebauer]

1. It depends on why they were excluded. If they were excluded because they were wrong or testing treatments that are not directly relevant to our work, they can be excluded.
2. Keep SOC concentration, bulk density, and SOC stock. Put everything else in a separate worksheet called 'white-additional-data', where we can come back to it in the future. For this project, many would be directly useful for comparing to SIPNET outputs. Nitrate N is only useful if Ammonium N is also included. In general, we should focus on just the target variables. SIPNET doesn't keep track of POXC, MAOM, POM, but since you've already collected them, it is worth keeping around.

Please make sure to review the Calibration / Validation Data Collection Protocol. And comment / ask questions / edit anything that is missing or unclear.