Randy's Varroa Model on the Web

Scientific Beekeeping
Excel Varroa Model

×

After two decades struggling with varroa, I realized that I needed to deeply understand varroa population dynamics, relative to climate, colony strength and amount of brood, mite influx from drift, and the ability of bees to affect varroa reproductive success. To that end I reviewed nearly every published study on varroa, and spent a year creating an accurate predictive model of varroa buildup in a colony, and the effects of treatments, swarming and other management techniques - confirmed by alcohol wash data from my own commercial operation, as well as feedback from beekeepers worldwide.
I created this model to be user-friendly to the beekeeper, yet of use to bee biologists. This simple model makes roughly 1500 calculations each time you enter a value, and is fully customizable. I continually revise and update this model, so always use the latest version.
Basic assumptions, adjustments, and calculations for r values
Since it is nearly impossible to accurately predict the precise effects of varroa and virus buildup upon colony population dynamics, the model assumes that mites will be managed to never exceed a level that affects the colony. I made no attempt to reflect the effect of mite buildup upon colony health, other than the eventual crash of the colony. My own field data indicate that a proportion of colonies will exhibit mite buildup rates far outside the mean (either way) - thus please assume that the simulations reflect median values.
The basic assumption of the model is that the rate of cell invasion by foundress mites is a matter of chance, dependent upon the ratio of 5-day larvae to adult bees. I used Willem Boot's data to arrive at a basic regression equation, then drew most of my other variables from the excellent model proposed by Stephen Martin (to whom I'm immensely appreciative), and added mite immigration and exodus. The root assumption is that foundress mites invade cells at the rate of y = -5*ln(x) + 5, with x being the brood:adult bee ratio, and y being the number of days of phoresy. I further used data from Lloyd Harris and Katie Lee to come up with additional values, and the tested the model against my own field data, as well as data from Jeff Harris, Bob Danka (and all) from the Baton Rouge lab, Eva Frey, and a number of other researches to whom I owe a great debt of gratitude.

I offer this model to the beekeeping and scientific community, and invite suggestions for improvement.
Randy Oliver.

×

Important: the alcohol wash reflects the proportion of mites on the adults bees, so early in the season it underestimates the degree of buildup of the actual mite population since most of the mites are in the brood (refer to "%mites in the brood" above). When the colony reduces the broodnest in autum, the mite infestation will appear to explode.
Warning: Simulations reflect median expected values--actual hive-to-hive mite counts typically vary greatly!
Once the alcohol wash exceeds 15 mites, there is typically a decrease in colony performance. Crash is indicated when the mite/virus load overwhelms the colony. It is easier and better to proactively keep mite levels low, rather than to later try to bring them down.

×
×

This number is often referred to as "fitness". A foundress mite in a worker cell has the potential of producing a little over 2 daughters, but various studies indicate that the observed number is typically lower.

See Dr. Stephen Martin's graph on the "Relative Rates of Increase" tab [not attached yet], plus:

Al Ghamdi, A and R Hoopingarner (2003) Reproductive Biology of Varroa jacobsoni Oud. in Worker and Drone Brood of the Honey Bee Apis mellifera L. under Michigan Conditions. Pakistan Journal of Biological Sciences 6 (8):756-761.

Kuster (2014) Immunogene and viral transcript dynamics during parasite Varroa destructor mite infection of developing honey bee (Apis mellifera) pupae. doi: 10.1242/jeb.097766)

A more recent review is Xie, X. et al. (2016) Why do Varroa mites prefer nurse bees? Sci. Rep. 6, 28228; doi: 10.1038/srep28228.

# daughter mites in a WORKER Brood Cell (avg 1.45):

1.11.7

# daughter mites in a DRONE Brood Cell (avg 3.5):

1.54

Mite mortality/day when broodless (avg 0.5%):

0.0%0.7%

Mite mortality/day during broodrearing (avg .6%):

0%0.8%

Percent Varroa Sensitive Hygiene (avg 5%):

0%50%

×

Default Winter: in temperate climate, managed to prevent swarming, slight fall buildup
Hard Winter (zone 1-6): High-latitude colony with 5 mo. brood break, managed to prevent swarming
Sub-tropical: dry climate, no winter break - Southern California
Package: 3 lbs, installed on drawn comb and fed well
Nucleus hive: 5 frames of bees and 3 frames of brood to start
Split or Swarm: Colony split or swarming on May 15.
Feral hive: Feral colony in 40-liter cavity, swarming twice (after Loftus 2016 doi:10.1371/ journal.pone.0150362).

×

Mites can be brought into the colony by the colony's own foragers. This can happen when bees which are compromised by mites return to the wrong colony, and when bees rob honey from a failing hive and bring back more than just honey. Foragers will find failing hives that are within a 1 mile radius from the hive. This will vary depending on the density of the hives in the area, and how successfully the mites in the apiary (and the area) are being managed.

  • No neighbors: no incoming mites on your own or lost foragers to supplement the home mite population.
  • Mite-managed apiary: mites coming in from nearby hives that have low mite populations.
  • Mite-troubled apiary: mites coming in from nearby hives that have on average a moderate mite population.
  • Few mite bombs: There are a small number of heavily infested hives which are being robbed by the hive.
  • Many mite bombs: There are many mite-infested hives within flight range which are being robbed by the hive.
×

[note from Trish] It is useful to quantify the daily growth rate (r-value) of the varroa mite. This can tell you how many days until the mite population doubles. It can also serve as a measure of how well mite reproduction was compromised, either by the bees' genetics or beekeeper manipulations. The mite populations from 4/15 to 9/15 are used for calculating the daily r-value here.
The equation for 'r' is: ln(mite pop on 9/15 - mite pop on 4/15)/(number of days between)
[note from Randy] Important note: the mite r value is density dependent--the more mites, the slower their growth. Make sure that you check your r value with a low starting mite level. The daily rate of mite increase during the colony's linear buildup phase, for non-resistant commercial stock, is typically in the 0.015-0.023 range (but can reach 0.030 with immigration).

r value Days to mite pop doubling Cases where r was measured
0.03122Harbo, Villa, Danka 2003
0.02726DeGuzman 2007, Baton Rouge Italians
0.02330Harris, Harbo, Danka 2004, also Harris, Harbo 1999
0.02133Harris, Harbo, Villa, Danka 2004, and 2008, also DeGuzman 2007 Baton Rouge Italians
0.02035Harris, Harbo, Villa, Danka 2007, also DeGuzman 2007 Baton Rouge Russians
0.01936Harris, Harbo, Villa, Danka 2005
0.01741DeGuzeman 2007 Baton Rouge Italians, also Vandame Mexico EHB
0.01450Harris, Harbo, Villa, Danka 2015
0.01163Harris, Harbo, Villa, Danka, 2006
0.01069Harbo, Harris 1999 average colony, also Vandame 2000 EHB
0.00977DeGuzeman 2007 Baton Rouge Russians
0.00787Harris, Harbo, Villa, Danka 2013
0.005139Harris, Harbo, Villa, Danka 2009
0.004173Harbo, Harris 1999 average of resistant colonies
0.003231Harris, Harbo, Villa, Danka 2010 and 2012, also DeGuzman 2007 Baton Rouge Russians
0.002347Harris, Harbo, Villa, Danka 2011
×

Expected mite reductions by treatment type: Always perform a mite wash 12 days after treatment (after mites have emerged from brood) to confirm efficacy! If the mite count is above 30, it may take a long time for a colony to recover from the in-hive DWV epidemic, even after an effective treatment.
HOW TO USE IN MODEL: Enter the percent mite kill due to your treatment into yellow boxes for the DATE the treatment is started - ex. 20% for 1 drone removal, enter 20.

WHAT PERCENT MITE KILL TO USE FOR YOUR TREATMENT TYPE:
95%: High-efficacy synthetic miticide, such as amitraz applied over several weeks. Not quick acting, so best applied proactively, early in the season.
90%: Two 50-g Apiguard (or three Apilife Var) treatments at 10 days, in a rim. Best used in August, as it restricts brood rearing.
70-90% (sometimes less): 2 MAQS or Formic Pro pads, or strong formic acid treatment (time release or hard flash).Best treatment for quick mite knockdown in warm weather.Can be used during honey flow.
50% (approx): 1 MAQS or Formic Pro pad, or single weak formic acid "knockback" treatment. Quick knockback for buying time, easier on the queen.
25%: A single powdered sugar dusting, assuming 50% drop of phoretic mites. Requires multiple and continued applications.
15-20%: Complete removal of drone brood, if done early in the season, each removal. Provides some reduction, but labor intensive.
80-95%: Oxalic dribble or vapor when broodless. Excellent reduction when colony is broodless, due to natural or induced brood break.
80-90%: Hopguard 3 if colony is broodless, would require repeated applications if brood present.
15-45%: One Oxalic dribble or vapor in hive containing brood, proportional to the % of mites that are phoretic. Limited reduction when colonies contain brood.
85%, 3x: Extended-release OA/glycerin sponges; for end result at 45 days, enter 85% for 6 weeks, starting at time of application. Not yet approved.
80-90%: Hopguard 3 if colony is broodless; would require repeated applications if brood present. Can be used during honey flow.
~30%: The issuance of a prime spring swarm. Assume 80% of mites in the brood, 70% of workers leave, 14 day brood break.
~50%-55%: An induced brood break by removing the queen or a "walkaway split." Assume a 26 day brood break.

×

Note that starting mite population can be specified in the "starting mite" yellow square above.
Starting mite population must be between 1-999.
Also note that the graph displays mite population at a 1:10 scale; so 10,000 on the graph is 1,000 mites.
100 mites in the hive is a good place to start. A 3-lb package often contains 100-500 mites. For established/overwintered hives, tweak your starting mite count until your early season alcohol wash count matches that of the simulation on the same date. Note that a hive which will successfully survive a second year MUST NOT have the Dec 31 mite count be much higher than its starting mite count.

×

Phoretic mites are on the body of the bee, sometimes on the back or near the wing attachments, or tucked halfway inside the abdomen, but ALWAYS nearly invisible. If you see a mite on a bee, then the good spots for sucking hemolymph or on the fat bodies were already taken, and it is bad news for your mite levels. The phoretic mites are the only ones removed by most of the available mite treatments - except those based on formic acid, like MAQS or FormicPro, which penetrates under the wax cappings.
Most of the phoretic mites which were killed will fall off the bees onto the sticky board about 24-48 hours after the mite treatment begins (see here for more). This is not true for for continuous release treatments like Apivar (see here for more info about mite drop post treatment).
The phoretic mite drop onto a sticky board post treatment is another way to measure the mite infestation of the hive. Just remember that for each mite dead on the sticky board, another 3-5 were hiding in the brood. This does not apply to mites that fall on the sticky board when NO treatment was administered, see here for more info about why using the ongoing mite drop is an inaccurate way of measuring mite populations in the hive.

Alcohol Wash
Mites x10
Capped Brood w/o and w/mites
All Brood
Adult Bees
Starting Mites:
Apr1 80% mite kill
Aug1 80% mite kill
Nov1 80% mite kill
Ending Mites:

Turn sideways to "landscape" mode for more detailed options.

See any bugs? Or have an Excel file you'd like turned into a webpage? Email trish @chickabuzz.com
Check out Trish at chickabuzz.com for more info.