Hive Simulator · The Alleman Apiary

Build a hive, set your queen, schedule mites & disease, and watch brood and honey rise and fall over a season.

📖 Frame Guide 🧰 The Bee Bench

Scenarios

Shares this exact hive setup as a link.

Challenges

Pick a goal, then use treatments, requeening and management to hit it. Re-run after each change.

Hive Boxes

Starting Colony

Queen

Real queens peak around 1,500–2,000 eggs/day in late spring.

Timing & Location

Push past a year to run multiple seasons (overwintering & spring rebuild).
Default: Harrisburg, PA. Affects modeled weather & season.

Weather & Forage

Using modeled seasonal weather.
Scales the seasonal nectar/pollen flow (spring build, summer dearth, fall flow).

Hive & Winter Protection

How much heat the hive holds and how it handles winter damp — the two biggest survival levers after mites. Compare setups by running into winter (push "Days to simulate" past a full year).

Starting Health

Diseases & Pests

Introduce a disease on a chosen day — watch the workforce and honey react.

Hazards

Pesticide exposure — a nearby spray drifts onto your bees. Kills foragers and lingers.

Mite Treatments

Schedule a treatment. Real-world kill depends on temperature, brood, and whether supers are on.

Feeding

Feed sugar (syrup or solid) or a pollen patty. Thin 1:1 syrup stimulates spring brood; thick 2:1 packs on winter stores; fondant & dry sugar work in the cold when bees won't take syrup.
lb of sugar / patty

Management

Scheduled Timeline

Compare Hives (A/B/C)

Run up to 3 hives head-to-head. They share every setting above — only the stock & treatment differ, so it's a fair experiment.

⏱ Treatment Timing Lab

The same colony and the same product, treated on three different dates plus an untreated control — so you can see why when you treat matters as much as whether you treat.

Bee Yard

A whole yard where mites drift between hives — one neglected colony can mite-bomb its neighbors.

My Hive Log

Log your real mite-wash counts to overlay them (white dots) on the charts vs the model's prediction.
🔮 Make a prediction before you run
Guess how the colony ends up, lock it in, then run — and see how close you were. Best used before you change one thing and re-run.
Nothing predicted yet.
Day 0 ＋ tap/drag on charts
One brood frame ▸ Capped brood Larvae & eggs Drone brood Pollen Capped honey Empty comb Dead/spotty brood
Whole brood box ▸ frames side by side — the brood nest fills the center, honey on the outside frames
Hive entrance ▸ live forager traffic — busy on warm flow days, quiet & clustered when it's cold
Adult bees Brood
Honey Mite load (%)
Brood Honey Empty comb
Daily high °F Forage flow Foraging days
📖 Beekeeping glossary — the terms used here
About the model & assumptions
This is an educational approximation, not a validated scientific model — great for "what if" exploration, not regulatory decisions. Key assumptions:
  • Worker development: egg→emergence in 21 days. Brood tracked as daily cohorts.
  • Adult lifespan ~42 days in the active season, but months for winter bees. Winter (diutinus) bees are induced by shortening daylength + low brood rearing in fall — the model computes true daylight hours from your latitude (Forsythe daylength model).
  • Laying is cued by photoperiod (lengthening days in spring, shortening in fall) as well as temperature, flow and pollen — which is why buildup and shutdown happen on schedule even in an odd-weather year.
  • Foraging only when modeled high temp ≥ 55°F and it's not a rain day.
  • Seasonal nectar flow: spring build, early-summer dearth, fall (goldenrod) flow, winter zero.
  • Comb: deep frame ≈ 6,000 brood cells or ≈5 lbs honey; medium ≈ 4,000 cells or ≈3.5 lbs. Brood only in deeps.
  • Varroa is modeled as a true mite population (count), not a fixed rate (VarroaPop-style): mites reproduce only inside capped brood, drone cells yield ~2× the offspring of worker cells, growth is density-dependent (saturates as mites catch up to available cells), and the reported mites/100-bees is derived from the live bee count — so a shrinking colony concentrates its mites (the "mite bomb"), a feedback the old rate model couldn't show. A phoretic fraction (mites riding on bees vs sealed in brood) sets how much of the population a contact treatment can actually reach. High loads amplify DWV, shorten lifespan & kill brood.
  • Queenless colonies stop laying; without requeening the colony dwindles.
  • Diseases each carry an infection level (0–1) that progresses, is modulated by weather/colony strength, and can self-clear (except AFB). Each hits some mix of brood survival, adult lifespan, foraging, and laying — so you see the workforce and honey react.
  • DWV is tied to your varroa load — knock down mites and it fades.
  • Mite treatments keep working over their whole label window — extended-release strips (Apivar, VarroxSan ~6–8 weeks; Apiguard, HopGuard ~4 weeks) go on killing mites as they emerge from brood, so one well-timed fall treatment can carry a colony through winter, while a single-dose product (oxalic dribble, drone-brood removal) acts once. Drone-brood removal only takes the mites sealed in capped drone comb, so it does nothing in fall or winter when there's no drone brood. Realized kill is degraded by being out of temperature range, by capped brood (for contact-only products like oxalic acid), and some (formic acid) can set back the queen/brood in heat. The summary reports the realized kill.
  • Small hive beetle & wax moth are modeled as weak-colony pests — they surge and bite harder the weaker the colony, and strong colonies clean them up.
  • Varroa treatment thresholds follow Cornell's Northeast numbers: 2 mites/100 bees April–July, 3/100 August–October (fall is set higher because brood is shrinking, so more mites ride on the bees and washes read higher). Mite load is also shown as a per-300-bee wash, the way you'd actually sample it.
  • Bee stock matters: each genetic line adjusts spring buildup, varroa growth, hygienic mite removal (VSH), winter hardiness, swarminess and thrift — so a mite-resistant stock can hold varroa down on its own.
  • Starting type (overwintered / nuc / package / swarm) sets your bees, brood, stores, queen age and how much comb is drawn. Undrawn comb caps laying & storage until the bees build it out — why a package is slow at first.
  • Drones are reared Apr–Aug (most in May–June; ~5% of bees at the midsummer peak), take 24 days, live ~3–4 weeks, and are evicted starting in September. Varroa strongly prefer drone brood, so drones speed up the mite build — and drone-comb removal is a real IPM tool.
  • Swarming fires when a strong colony gets crowded in May–Jul (it leaves with the old queen and ~55% of the bees, then has a ~3-week brood break with no eggs while the new queen emerges and mates); a worn-out old queen is superseded; and you can split a colony as a brood-break mite control (you keep the queenless half, which raises its own queen — ~3½ weeks with no eggs). A requeen takes ~6 days before the new queen is laying.
  • The brood-frame view is a representative central frame: a capped-brood nest with a dappled core of young larvae/eggs, a pollen ring, and a honey arc on top — the classic "rainbow." It turns spotty/shotgun (and shows mites) exactly when brood is actually dying, and goes queenless (no eggs) if the queen is lost.
  • The brood-box strip shows every frame side by side — the brood nest is weighted to the center frames and honey to the outer frames, the way a box looks when you fan it open. Hit ▶ Play season to watch it all change day by day.
  • Compare Hives runs up to three colonies on the same boxes, dates, weather and forage — only the stock & treatment differ — so you can put genetics and IPM head-to-head as a controlled experiment.
  • Pesticide hazards (spray drift / direct kill) kill foragers outright on the day they hit, then leave a sublethal load that fades over ~2 weeks — extra adult/brood loss and weaker foraging.
  • The honey-value estimate multiplies your harvestable surplus (anything over a 40 lb winter reserve) by your $/lb — shown in the summary and in the compare table.
  • The Bee Yard runs several hives together and lets mites drift between them: a high-mite or collapsing colony sheds mites (a dead-out gets robbed) that raise its neighbors' loads — a rough model of how one neglected hive mite-bombs the yard. Even a well-treated hive can pick up extra mites from a bad neighbor.
  • My Hive Log lets you record your real mite-wash counts (and bee estimates) by date; they overlay as white dots on the charts and show the model's predicted value on that date, so you can see how your actual hive compares to the simulation.
  • Multi-year: push the day slider past a year to run several seasons — the colony overwinters and rebuilds each spring. Tick "repeat events every year" to apply an annual management regimen, and the summary breaks results down year by year. Note: without consistent, well-timed mite control, varroa wins within a year or two — and an un-requeened colony slowly declines as its queen ages (requeen to counter it).
  • Challenges set a goal (a honey target, rescuing a mite-bomb hive, getting winter-ready, surviving a spray). Pick one, then use treatments, requeening and management to hit it — the result banner tells you if you passed and by how much.
  • The hive-entrance animation shows live forager traffic for the day you're viewing — busy when it's warm and the flow is on, quiet and clustered when it's cold or raining. Hit ▶ Play season and watch the traffic rise and fall with the seasons.
  • Uncertainty band (🎲): tick it and the model re-runs your exact plan 25× with randomized weather (temperature & rain), then shades the 10th–90th-percentile range on the population & honey charts and reports how many runs survived. A season is probabilistic — the same plan can win or lose on the weather, and a wide band tells you the plan is fragile.
  • Under the hood: the coach box shows which of three ceilings — the queen's own potential, available nurse bees, or open drawn comb — is actually limiting laying on the day you're viewing (the queen can only lay the lowest of the three), plus the day's daylight hours.
  • The story of this colony: after each run the model reads its own output back as a plain-English timeline — buildup, mite threshold crossings, DWV, swarms, treatments, the winter-bee switch and any collapse — so the cause-and-effect chain is spelled out, not left to interpret from the charts.
  • Treatment Timing Lab: runs the same colony with the same product treated on three different dates (plus an untreated control) so you can see that when you treat — especially hitting the Aug–Sep window before the winter bees are reared — matters as much as which product you use.
  • Predict-then-reveal (🔮): lock in a guess for peak honey, ending bees and survival before you run, and the model scores how close you were — a simple way to test and sharpen your intuition.
Sources: treatment efficacy/temperature figures from the Honey Bee Health Coalition Tools for Varroa Management 9th Edition (2026); the treatment thresholds, disease symptoms & the Northeast management calendar from the Cornell Master Beekeeping course materials. The varroa population dynamics follow the concepts of the USDA VarroaPop model (DeGrandi-Hoffman & Curry) — reproduction in capped brood with a drone-cell preference and a phoretic phase; daylength uses the Forsythe (1995) CBM model. Coefficients are tuned for educational realism, not laboratory precision.