Four species per quadrat, each with a percentage cover estimate and a field photograph. Twenty-one fields total. That's the discipline this kind of data demands — not because it's elegant, but because percentage cover estimates decay into noise the moment a surveyor has to reconstruct them from memory rather than log them on-site.
The Quadrat Record That Actually Closes
Vegetation quadrat surveys have a persistent failure mode: the cover percentages get recorded on a datasheet, the photos get taken, and then they get separated. The datasheet gets transcribed. The photos get uploaded. Six months later, the photo labeled "Q4_species2" might be Indigofera sp. or it might be a Poaceae clump that got misidentified in the field and never checked against the percentage record.
This template solves the linkage problem directly. Species 1, Species 1 (%), and Species 1 (photo) are three consecutive fields in the same record. The photograph and the identification and the cover estimate are bound together at the point of entry, in the field, before the surveyor moves to the next quadrat.
The Ecologist multichoice field — SN, ST, RN — is the auditing mechanism. In multi-surveyor monitoring programs, inter-observer variability in cover estimation is real and documented. When you filter your dataset by ecologist and discover that one surveyor consistently logs Zygophyllum qatarense cover 8–12% higher than the others across the same habitat type, you have a calibration issue you can address. Without the field, that variability looks like ecological signal.
Species Cover as the Unit of Analysis
The species choices in this template — Zygophyllum qatarense, Indigofera sp., Poaceae, and additional options — reflect a specific floristic context. The choice field constrains entry to validated species names, which means no "Zygo" abbreviations, no "grass spp.," no transliterations that don't match the reference list. In dryland vegetation monitoring where total cover rarely exceeds 40%, the difference between logging a Zygophyllum clump as 8% versus 12% has real consequences for change detection analysis over a three-year monitoring cycle.
Cover percentage as an integer — not a decimal, not a range — is a deliberate precision calibration. A surveyor estimating 7.5% is overconfident in a 1m² quadrat in a variable system. Integers anchored to a Braun-Blanquet-derived scale or a simple 5% class system translate to real ground truth better than spurious decimal precision.
The Record in Context
Coordinates at the quadrat level — not the plot level, not the site level — means each quadrat is independently geolocated. When the vegetation boundary shifts 3 meters between your 2024 and 2025 surveys, you can detect whether the shift is real movement or positional error in relocating the quadrat. If you're working from a single site coordinate for the whole plot, you can't.
The quadrat photo field sits in the record before the species entries, which is the right order. You take the overview photo that documents the quadrat boundary and ground condition before you start estimating cover. The record forces that sequence.
Notes sits at the end — for the Haloxylon seedling that doesn't fit any of the four species slots, for the erosion channel that bisected the quadrat between visits, for the rabbit warren that explains why the Poaceae cover dropped from 22% to 4% since the last survey.