How it works
How we score a spot for honey bees
The forage check reads public maps of crops and vegetation around a point, gives every patch of land a bee value by season, adds it up the way a colony would experience it, and ranks the result against random land in the same state. This page explains each step and what it cannot do.
1. What is growing around the spot
Every 30 m square within about 3.7 miles (6 km) of the point is classified. Where it is farmland, the class comes from the USDA Cropland Data Layer, a yearly map of what crop grew in each square. Fields rotate, so we average the last five years: a field that grew canola in two of five years counts as two-fifths canola. Everywhere else, the class comes from the LANDFIRE vegetation map, which names the plant community: ponderosa pine woodland, sagebrush steppe, suburban yards, cottonwood river bottom, and so on.
2. What each kind of land is worth to a bee
Crops take their pollen and nectar ratings from the USDA's 2017 report on the attractiveness of agricultural crops to bees: "++" for crops bees work heavily, "+" for crops they use under some conditions, "−" for crops they ignore. Wild vegetation is harder. Each vegetation type is a mix of plants, so we estimate how much of its ground is flowering plants bees use and how much is grass, conifer or bare. In the Pacific Northwest those mixes are measured from Forest Service inventory plots and published plant-association guides; elsewhere they are our estimates from the type descriptions, and the report says so when that applies. Each wild plant carries its own honey bee rating from the beekeeping literature, with the source quoted in the model's tables.
3. When it blooms
Every plant has a bloom window. Where a window is given relative to spring, it is shifted by the USA National Phenology Network's 30-year first-bloom date for the spot, so the same plant blooms earlier in Georgia than in Montana. The result is a season of 24 half-months showing how much pollen and nectar the land offers in each, where the flow peaks, and where it fails.
4. How far bees fly
Nearer land counts more. The weighting follows published studies that decoded waggle dances to see where colonies actually foraged: in a New York forest the median trip was 1.6 km and 95% of foraging lay within 6 km. In spring, when colonies are small and the weather is cold, bees forage closer to home, so spring forage is weighted more tightly than summer forage.
5. The percentile
For each state, each ecoregion and the continental US we score 3,000 random land points the same way (open water and bare rock excluded) and keep them as a benchmark. Your spot's percentile is its place among those points. An 80th percentile in Washington means the spot scores better than 80% of random Washington land. The points are land in general, not beekeepers' yards, which tend to be chosen well, so a high percentile is a good sign rather than a promise.
6. The rest of the report
The honey estimate takes the nectar sources by season and uses published descriptions of each plant's honey for colour, setting and flavour. Water comes from the USGS hydrography map, flood zones from FEMA, and the "things to watch" list combines crops with little bee value, crops with heavy insecticide use, long gaps in the season and dense development. Every source is listed.
What the maps cannot see
- Garden plantings, flowering street trees and small plots below 30 m.
- How thick the understory is in a particular wood. The model uses the average for that forest type.
- Other beekeepers' hives, which share the same flowers.
- This year's weather, crop choices and spray schedules. The crop maps are last year's and earlier.
- Fallow fields full of weeds, which often feed bees but carry no rating.
Treat the check as a first look that tells you what to go and see. A walk in April and July tells you the rest.
Common questions
How far do honey bees fly to forage?
Most trips are within a mile or two, with a long tail out to about 6 km (3.7 miles). The forage check scores everything within 6 km and counts nearer land more, following the study of decoded waggle dances that found a median foraging distance of 1.6 km and 95% of foraging within 6 km.
What makes a good location for a beehive?
Forage that runs from early spring to autumn without long gaps, water within reach, and not too many hives already working the same ground. Bee crops such as canola, alfalfa grown for seed, sunflowers and clover give heavy flows; a mix of flowering trees, shrubs and weeds gives steadier forage. Grain fields, bare ground, conifer plantations and dense pavement give almost nothing. Early tree pollen, from willows and maples, matters for spring build-up.
Why does my forest score low (or high)?
A forest is scored by what grows under and among the trees. Open ponderosa pine with snowberry, ninebark and spring wildflowers scores well; a dense Douglas-fir plantation with a bare floor scores poorly. The model uses the measured or estimated understory for each forest type, not your particular wood, so a wood that is unusually rich or unusually bare will be off.
Why does corn and soybean country rank the way it does?
Corn offers pollen and nothing else; soybeans are rated "+", attractive under some conditions, and there are a great many acres of them. The model counts a "+" crop at a quarter of a "++" crop, so wide soybean country gets a moderate score. Beekeepers there make honey from clover, basswood and roadside weeds, which the model also credits where the maps show them.
What is a nectar flow, and a dearth?
A flow is a period when nectar comes in faster than the colony uses it, so the hive gains weight and stores honey. A dearth is a gap when little blooms, often in late summer in dry country, when colonies can lose weight and get defensive. The season timeline marks both.
What does the percentile mean?
We score 3,000 random land points in each state, each ecoregion and the continental US. A spot in the 80th percentile for Washington scores better than 80% of random Washington land: land in general, not other beekeepers' yards.
Can I compare two spots?
Yes. Check each spot and use the compare card at the bottom of the report; it keeps up to a few spots in your browser only.
Is my address stored?
No. The address becomes a point in your browser's request, the point is rounded to a 250 m grid before it is used, and only the rounded point is cached so the next visitor to the same area gets a fast answer. A link you copy from a report carries that rounded point and nothing else.
Does it work outside the United States?
Not yet. The crop and vegetation maps cover the continental US; Alaska, Hawaii and other countries are outside the mapped area.
How accurate is it?
It is an estimate from public maps, checked against hive-scale records and our own yards, and the model's sources and limits are published so you can judge them. It will tell you a canola basin from a wheat basin and a rich wood from a bare one. It cannot tell you about the neighbour's garden or next year's weather. The check is a first look, not a verdict.
Try it on your own spot
Enter an address or tap the map, and see what blooms within flight range and when.