Data From: Fall Diet Supplements Have Delayed And Context-dependent Effects On Honey Bee (Hymenoptera: Apidae) Colony Size And Profit

Colonies, apiaries, and diet treatments

Colonies used in this study were part of an existing commercial beekeeping operation. The colonies were on pallets containing 4 colonies per pallet. The colonies were not normalized or equalized prior to the start of the experiment. All the colonies had an adult population, open and sealed brood, and no obvious signs of disease at the start of the experiment and were housed in two Langstroth deep boxes with 16-18 frames and 1-2 syrup feeders. Prior to the start of the experiment, the colonies were transported from Fannin County, Texas, USA to North Dakota, USA for honey production on May 20th (Table 1). The colonies were split evenly between two apiaries (“Edmore” and “Loma”; Table 1) and were treated with oxalic acid miticide in early August after the honey was harvested and before the experiment started.

Colonies were fed one of five commercial diets (MegaBee (megabee.com; 60 colonies), BeePro (Mann Lake; 60 colonies), UltraBee (Mann Lake; 56 colonies), AP23 (Dadant; 57 colonies), or Global 4% (base diet + 4% irradiated natural pollen; www.globalpatties.com; 60 colonies)) or a positive control diet (Global base diet + 75% irradiated natural pollen (globalpatties.com; 59 colonies)) in trial 1. The Global 75% diet contained natural pollen that comprised 75% of the dry weight, the maximum amount of dry material that could be used. In trial 2, colonies were fed one of the same five commercial diets, the same positive control diet, or no diet as a negative control (50 colonies per treatment split evenly between the apiaries). All diets came from the manufacturer in patty form. The diet patties weighed 0.454 kg (1lb) except for the MegaBee diet, which came as one solid block of patty that the field crew cut to approximately 0.454 kg patties. The diets were first applied immediately after the August pre-feed colony assessments (Figure 1). Patties were placed under the hive cover on the top bars of the frames weekly for 5 weeks.

In both trials, the colonies were moved in early October from each apiary site to the cold storage facility in Filer, Idaho, USA to overwinter (“post-feed”). In January, the colonies were transported from the cold storage facility to almond orchards in Tulare County, California, USA for pollination (“pre-almonds”). It was logistically unrealistic to assign colonies to specific orchards or spots in the cold storage facility, but the colonies were randomized as much as possible.

Colony assessments

Colonies were assessed at four time points in both trials: a pre-trial assessment before the diet was added (“pre-feed”), after weekly feedings and just prior to cold storage (“post-feed”), after cold storage in the almond orchard before bloom (“pre-almonds”), and during almond bloom (“post-almonds”). Colony assessments took 2 days and were performed using methods described in (DeGrandi-Hoffman et al. 2019). Briefly, the frames of bees (FoBs) in all colonies were measured using a variation of the “California count” that is used to grade colonies for almond pollination (Goodrich and Goodhue 2016). This method enabled the rapid measurement of >300 colonies in 2 days. The top and bottom boxes were separated, tilting the top box against the bottom box. The number of frame tops covered by bees in the bottom box (looking down) and on the frame bottoms in the top box (looking up) were counted. The California count colony size estimates are represented in frames of bees (FoBs).

In trial 1 (2021-2022), at each time point, a more detailed colony assessment was performed in a subset (Nsubset=121) of the experimental colonies located in each apiary (18 BeePro, 19 UltraBee, 17 MegaBee, 15 AP23, 20 Global 4%,16 Global 75%). These data were collected after the California counts were performed. In this more detailed colony assessment, the area of a comb occupied by adult bees was measured as described in Delaplane et al. (2013). Briefly, each frame was divided into five sections per side/frame face so that 10 sections were measured on each frame. In addition, the bees on the hive lid were measured by dividing the lid into 10 sections. Areas covered with adult bees were summed for each colony to provide an estimate of comb coverage. This value can reach 21 in colonies with two hive bodies (20 frames and 1 lid). At the pre-feed time point before the experiment began, we also assessed the amount of comb occupied by honey and pollen.

Mite counts __

In trial 1 (2021-2022), at each time point, mites were collected from the same subset (Nsubset=121) of the experimental colonies used for the comb coverage assessments. In trial 2 (2023-2024), all the colonies (N=350) were sampled at each time point. In all cases, ~100-200 bees were collected from the brood nest area into a 120 ml polypropylene specimen container (Qorpack, Clinton, Pennsylvania, USA) with 100% ethanol (Dietemann et al. 2013). The sample cups containing the bees and mites were topped off with additional 100% ethanol and sent to USDA-ARS Carl Hayden Bee Research Center in Tucson, Arizona, USA for mite counts. Mite counts were performed using the Varroa easy check system (Véto-Pharma, Palaiseau, France). The number of bees in the sample was also counted. Mite level was the number of mites divided by the number of bees multiplied by 100 (i.e., the percent of bees with mites).

Data sets

"FoBs and mites both years.csv" contains (from left to right) the colony number ("colony"), the apiary ("Yard"), trial ("Year"; 2021 = Trial 1, 2023 = Trial 2), diet ("Diet"), time point ("Time_point"), mite levels ("Mites_per_100_bees"), and frames of bees ("FoBs"). In trial 1, we collected mite data for the Nsubset=121 colonies and all of the colonies in trial 2.

"FoBs no mites both years.csv" contains (from left to right) the colony number ("colony"), the apiary ("Yard"), diet ("Diet"), trial ("Year"; 2021 = Trial 1, 2023 = Trial 2), time point ("Time_point"), and frames of bees ("FoBs"). All of the colony size data from both trials are included.

SAS code

"GLMMIX analysis of colony size data with mites.docx" is the GLMMIX analysis of the colony size and mite count data. For use with input file "FoBs and mites both years.csv".

"GLMMIX analysis of colony size data without mites.docx" is the GLMMIX analysis of the colony size data without the mite count data. For use with input file "FoBs no mites both years.csv".
Organization: Department of Agriculture
Last updated: 2025-10-02T22:13:59.867478
Tags: almond-pollination, apis-mellifera, cold-storage, diet, fall, honey-bee, nutrition, supplement