The Audit That Came Back Empty
Six weeks into a revegetation contract, the project ecologist requests a species breakdown by site. How many Allocasuarina verticillata went into the EGLN-3 corridor, and how many Banksia marginata are in the FEVL wetland fringe? The crew lead opens their records. There are totals — aggregate numbers from each planting day — but no species-level breakdown by location. The data exists, in someone's memory, but it's not in the log. The answer to a straightforward compliance question is now a reconstruction exercise that nobody has time for mid-contract.
This is the failure mode. Not the dramatic kind. The slow, grinding kind where every reporting obligation generates a data archaeology project, and the crew's professionalism is undermined by a recording system that didn't capture what the ecology actually requires.
What Survives a Wet Season
The structure here tracks at species-individual level — not "native acacias, approx. 45" but Acacia implexa: 12, Acacia mearnsii: 8, Acacia melanoxylon: 25. Each taxon gets its own integer field. This granularity is what allows survival-rate analysis after a season. When the monitoring team returns in eighteen months to do condition assessments, they need to know not just how many trees went in but which species, in which quantities, at which site. Without that, the condition report is a narrative. With it, you can run establishment rates against rainfall data, soil type, and planting date.
Fernlea Location ID, EGLN Sites, and FEVL Sites are separate multi-choice fields precisely because the site taxonomy matters to different stakeholders. The client might care about Fernlea-coded sites; the land trust monitoring agreement tracks EGLN designations; the state grant acquittal references FEVL boundaries. Running all three as independent fields means the same planting session can be filtered and reported against all three frameworks simultaneously without re-keying.
Crew as a multi-choice field rather than a free-text entry is another deliberate choice. Free-text crew fields fragment into spelling variants and nickname variations that make aggregation unreliable. When you need to calculate crew-hours-per-thousand-stems for a productivity assessment, a normalized crew list returns clean numbers. A free-text field returns noise.
The Session That Tells You Rate
Time Started and Time Finished per session give you planting duration, which combined with the species count totals gives you planting rate: stems per crew-hour. Over the life of a contract, this number becomes your benchmarking tool. If the EGLN sites are consistently running 30% slower than the FEVL sites, that's a terrain or access signal worth investigating before you price the next phase. If a particular crew configuration is consistently outperforming others, that's a staffing insight.
Single-day records look like administrative overhead. Thirty sessions of records look like operational data.
The Location field feeds GIS mapping. Each session record carries coordinates, which means at the end of the contract you have a spatial data layer showing actual planting density and species distribution across the restoration site — not a projected map based on the work plan, but a confirmed plant-by-plant record of what went in the ground and where.
Acacia verticillata in the southern wetland fringe, Eucalyptus ovata on the mid-slope, Banksia marginata on the rocky ridge. Species matched to microhabitat, confirmed by coordinates, with the crew and date attached. That's the record that holds up in a biodiversity offset acquittal.