Without the triplicate structure, you don't have a survey. You have an observation. The difference matters when you're submitting to a restoration program that requires statistical validity per plot.
The Cost of Collapsing What Shouldn't Be Collapsed
Every mangrove monitoring program that starts with a simple spreadsheet eventually breaks under the same pressure: triplicates get merged into averages before anyone meant to merge them, Plot A's Tree #3 gets confused with Plot B's Tree #3, and by the second field season, nobody can tell which coordinates are the base point versus the transect endpoint.
The architecture here — Triplicate No. (1, 2, 3) and Plot (A, B, C) as top-level fields before any tree data — isn't administrative overhead. It's the structural key that prevents two surveyors from creating ambiguous records when they work the same site on the same day. Coordinates (Base) captures the fixed reference point per plot, not the surveyor's starting position, not the nearest landmark — the actual GPS base from which the transect runs. That distinction compounds in importance when you have 36 records from the same corridor and need to check spatial consistency across visits.
Tide — Low or High — is the field that a lot of field teams record in their notebooks and forget to transcribe, and then can't remember six months later when the data analyst asks why Tree #4's DBH measurement in Plot B2 looks 12% smaller than the previous visit. It's not tree loss. It's a high-tide measurement taken when the base was moved 4 meters inland.
The Anatomy of a Per-Tree Record
Tree #1 through Tree #7 each carry the same four fields: ID, measure point (DBH or COD), Circumference, and Height, plus Health (Healthy, Sick, Dead). The choice between DBH — diameter at breast height — and COD — circumference over bark at a defined point — is not a preference. It is a protocol decision that must be consistent within a plot across time. Recording the measure point per tree, per visit, is what allows you to catch when a substitute surveyor defaulted to DBH on a plot established with COD without catching the discrepancy until back at the office.
Health status as a three-way choice rather than a numeric index is deliberate for this use case. Sick captures visible crown dieback, prop root bleaching, and early pneumatophore mortality without requiring the surveyor to make a quantitative judgment they aren't trained to make in the field. Dead is unambiguous. The field is fast to complete in knee-deep mud with a data recorder glove covered in sulfurous sediment.
Additional Notes or Measures as a list field — not a single text block — allows structured appending without overwriting. Multiple surveyors can add observations to the same record without one entry collapsing another.
The Transect Data Layer
Eight substrate types along the transect line, each with start and end points as separate integer fields, builds the line-intercept data that vegetation cover analysis depends on. The separation of type name (text) from start and end coordinates (integers) means you can calculate substrate lengths programmatically without parsing free text. When substrate #3 is "pneumatophore zone" and runs from 12 to 27 meters along a 50-meter transect, that 15-meter coverage figure is recoverable by arithmetic, not by reading notes.
The record ends with a Notes field — free text, positioned last — for the conditions that don't fit the checkboxes: the kingfisher perched at the plot boundary during the count, the outfall pipe that appeared since the last visit 80 meters downstream, the bench mark stake that shifted 20 centimeters and needs re-anchoring before the next session.