Quercus Score method

Native plants

Five dataset-based factors, each ranked against every native genus in North Carolina:

Caterpillars (45%)	Caterpillar species that feed on the genus in NC (Shropshire & Tallamy 2025). Caterpillars are the main food for nesting birds, and a few keystone plants feed most of them (Narango et al. 2020).
Bees (20%)	Bee species recorded on the genus in the southeastern US (curated GloBI bee–plant dataset).
Specialist bees (15%)	Pollen-specialist bees that collect pollen only from this genus or its family, limited to bees recorded in eastern North America (host table from the Big-Bee specialization study, compiled from J. Fowler's lists). Family specialists are credited to the genera they were recorded on. Wind-pollinated plants such as oaks score low here by design.
Birds (10%)	Bird species recorded eating the genus's fruit, seed or other parts in North America (Avian Diet Database).
Structure (10%)	Growth habit of the species (USDA PLANTS): tree 1, shrub 0.75, vine, grass or forb 0.5. A coarse stand-in for cover and nesting sites.

The baseline weighted mean is ranked against all native species in the state; the top 14% receive baseline scores of 80–100, the rest scale from 5 to 79. Subsequent adjustments can change band membership. Birds and structure carry less weight on purpose: bird diet studies over-represent crops and game-bird foods, and growth habit is a rough proxy. Bee and bird counts also follow how often a plant has been studied, so a low number can mean "not studied" rather than "not used".

Within a genus, species are separated by how common they are in North Carolina (NC Natural Heritage state rank): common (S5) no change, S4 −2, S3 −5, S2 −8, S1 or historical −10. Caterpillar data is recorded by genus, and a widespread species supports local caterpillar populations across more of the state than a rare one. This is an inferred adjustment, not a measurement, and it is replaced when species-level data is loaded. Rarity is flagged separately on each page.

Still to add: soil holding (erosion control) and species-level data in place of genus-level. Each page identifies the datasets used. A dataset with zero records is not evidence of zero ecological use.

Insect resource support (beta)

The baseline gives 80% of its weight to insect resources: larval hosts 45%, bee visitors 20%, specialist bee hosts 15%. These are model priorities, not measured shares of food-web importance. Specialist bees may also appear in general bee counts; the specialist factor intentionally emphasizes dependence, rather than counting unique insects across datasets.

Species-matched additional evidence earns one point per documented group: ants, beetles, flies, adult butterflies/moths and other wasps; maximum +5, with the score capped at 100 before the existing rarity adjustment. Other insects are displayed but not credited until resolved into comparable groups. Each group earns at most one point regardless of repeated observations, insect counts or number of studies. Adult flower visits and larval host use are separate resources, not unique-species totals.

This small credit is a provisional policy choice, not a scientifically calibrated weight. It preserves the baseline rankings and does not re-rank or lower plants lacking new evidence. Additional coverage still favors studied plants. Unknown relationships remain unknown, with no additional credit; that does not imply no ecological value.

Robertson records describe Illinois flower visits, a weak resource-use proxy rather than confirmed feeding or pollination. Penn & Crist records retain observed USA/Canada seed dispersal, seed consumption and appendage consumption separately. Seed consumption can support insects without benefiting the plant. These are relationships elsewhere in a species’ range, not proof of local NC insect occurrence. We do not infer specialization from a single observation or assign it to an entire insect group. No numerical confidence or survival probability is claimed.

County-first Fit Score (beta)

Three separate outputs: Quercus measures ecological value; site suitability describes the checked growing conditions; combined Fit blends that partial suitability with ecological value. County selection shows local occurrence evidence. At least one supported growing condition is required for suitability and Fit; an empty profile is not a perfect match.

Optional light, moisture and pH use published condition proxies. A value within range contributes 1.0; one scale step outside (0.5 for pH) contributes 0.6; further outside contributes 0.2. Suitability = 100 × lowest checked factor. Fit = suitability × (0.60 + 0.40 × Quercus/100) × county factor, rounded. County factor is 1.0 for an occurrence in your county and 0.95 for a neighboring county; otherwise combined Fit remains unresolved. The 60/40 choice is provisional policy, not calibrated survival evidence. Invasives retain their negative score and do-not-plant warning.

Every suitability number is explicitly a partial assessment: 100 means no mismatch in checked conditions, not proven overall suitability. Heat, winter cold, drainage, compaction and space are not evaluated. Missing heat evidence has no invented penalty. Published hardiness guidance is provided where reviewed, with disagreements retained; it is not yet matched against county winter data. Winter hardiness does not establish summer heat tolerance, and a cold mismatch would not automatically establish suitability as an annual. County occurrence is not proof of indigenous range. The calculator runs in your browser without sending your entries to Quercus.

Agents use the identical versioned calculator through the agent packet. A hosted agent may receive information you share with it; Quercus does not receive that profile.

Growing conditions (separate from the score)

Each page lists what the plant needs, from public data. These feed the yard-fit check, not the Quercus Score.

Moisture	Federal wetland rating (USACE, Eastern Mountains & Piedmont), widened or narrowed by USDA drought and saturated-soil tolerance.
Light	Shade tolerance of trees and shrubs (Niinemets & Valladares 2006). Otherwise USDA's shade class, used inverted and marked coarse: checked against the tree data, the stored USDA High class averages 2.01/5 in the independent shade dataset, Medium 2.72 and Low 3.47 (193 shared woody species). This is a provisional correction to this snapshot, not a confirmed diagnosis of the live USDA service; it is not independently validated for all herbaceous plants.
Soil pH	USDA PLANTS pH range.
Recorded in NC	USDA county records, grouped by province. The yard check flags natives with no records in or next to your county.

Coverage is partial. Pages say "no data" rather than guess. Soil compaction and disturbance tolerance are not evaluated.

Invasive plants

Fixed by the NC invasive list rank: Rank 1 −85, Rank 2 −55, Rank 3 −30, Watch List A −15, Watch List B −10, minus documented harm beyond spreading (for example, toxic berries).

Other non-native plants

Not scored. Marked Caution when the NC flora records them escaping into the wild.

Bands

Keystone ≥ 80 · Strong ≥ 50 · Good ≥ 20 · Modest ≥ 5 · Neutral ≥ −5 · Harmful ≥ −40 · Severe harm below.

The score measures value to wildlife, not whether a plant will grow in your yard. Growing conditions are listed separately on each page.

Sources and limitations

This is an experimental, versioned index, not an independently validated scientific standard. Source scope, genus-level estimates, study effort and the inferred rarity adjustment limit interpretation. Wetland ratings describe occurrence, not exact garden moisture requirements.

Source attribution and reuse status. Caterpillar host-count data derived from Shropshire & Tallamy are made available under the Open Database License 1.0; other source terms remain separate.

How the Quercus Score works (v0.6.4)

Insect resource support (beta)