Two Kinds of Dearth

Why feeding pollen in July might help, and feeding it in September might not

Back in March I wrote about honey bee nutrition, and I pontificated (yes, I do that) about the merits of feeding during a dearth. I said that I suspected my bees had suffered nutritional deficits in late summer, that this would have left them more exposed to varroa and deformed wing virus, and that it could have produced poor quality winter bees. My high (for me) winter losses of 20% could have been related to these things.

An article by Clare McGettigan in the BBKA News, August 2026, called “It wasn’t just the rain,” delved into high winter losses across much of the UK. The evidence shows, unsurprisingly, that beekeepers who treated for varroa suffered significantly lower losses than the non-treaters. Anyone even loosely familiar with this blog will know where I stand on that. As one who treats three times per year I am reasonably confident that I have mites under control in my colonies. I have also recently been raising queens from a breeder with hygienic traits, but time will tell what impact that has.

So, with the current season following a similar pattern to last year, I thought I’d visit the issue of honey bee nutrition once more, because it’s not straightforward. Last season most things flowered quite early, and the so-called ‘June gap’ was either short or non-existent. However, from the third week of July and into August, the natural flow dried up. The exceptions were colonies sitting on borage (current season), and those close to the SFI plots with their phacelia, clover and vetch. Some of my bees found spring-sown field beans, and at another apiary there’s a field of sunflowers currently in bloom. Basically, if local farmers were growing beneficial crops the summer flow continued, but bees in areas relying on bramble, lime, and sweet chestnut had nothing much to forage on in late July.

The obvious response is to feed. Pollen patties, a bit of syrup, tide them over. And that is more or less what I said in March: when the pollen stops, I feed a patty of some description. However, there is also the research on winter bee production. It turns out the obvious response is only obvious if you feed at the right time of year. Feed at the wrong time and you may be working against the colony rather than for it.

The Damage a Gap Does

I will not repeat the whole of the March post, but it’s worth reflecting on the key points. Larvae are only larvae for six days, so a pollen gap of even three or four days is a significant chunk of a larva’s entire developmental window. Scofield and Mattila found that workers raised under nutritional stress as larvae were lighter, shorter lived, less likely to forage, less likely to waggle dance, and more likely to die after a single day of foraging¹. Di Pasquale and colleagues showed the direct link between restricted pollen and poor hypopharyngeal gland development in the nurses themselves². These glands produce the protein fraction of the worker and royal jelly that is fed to developing larvae and queens.

An important feature is the delay between cause and effect. A gap in the third week of July produces damaged workers in the second week of August and shows up as reduced foraging efficiency in early September. By the time you notice the consequence, the cause is six weeks behind you and you have probably forgotten about it.

It Is Not Just the Larvae

Carroll, Brown and Huang published a study in 2025 that looked specifically at colonies in a summer dearth³. Their focus was E-beta-ocimene, a volatile pheromone produced by eggs, young larvae, prepupae and laying queens, which is part of how the colony regulates brood rearing.

In pollen-deprived colonies, ocimene emissions fell as the workers steadily cannibalised the very brood that produces the pheromone. Supplemented colonies reared more brood and emitted more ocimene. When they examined twelve day old workers, those that had completed their adult development in deprived colonies had less well developed hypopharyngeal glands and smaller fat bodies than those from supplemented colonies. Abdominal lipid content was significantly higher in the supplemented group.

Two things follow from that:

The first is that a July gap does not only cost you the brood that got eaten. It also produces a cohort of adult bees carrying a physiological deficit into their nursing careers. Those are the bees that will be asked to rear September’s winter bees. The damage travels forward a generation.

The second is more troubling. The researchers pointed out that colonies normally resume brood rearing once conditions improve, but that prolonged cannibalism strips out the brood signals that support brood rearing in the first place. In other words, a long enough gap may become partly self-sustaining. The colony does not simply spring back the moment the balsam comes out, because it has removed some of the machinery that would tell it to (open brood).

Schmickl and Crailsheim had already shown the front end of the problem, with five days of enforced non-foraging leading to cannibalism of the youngest larvae and earlier capping of the survivors⁴.

So far, so straightforward. Pollen gap bad, feeding good, and I should probably be feeding patties from mid-July if there’s no bee bread stored on, or near, brood frames.

Now the Awkward Bit

Except that the absence of pollen is also how a colony knows the year is over.

Heather Mattila and Gard Otis ran the study that established this⁵. They took colonies and either depleted them of pollen using traps, which cuts stored pollen by something like 60 to 90 per cent, or supplemented them with pollen patties at two parts pollen to one part sugar, delivering around 1.3 kg of extra pollen per colony. The manipulation ran from late August to late October. They then marked cohorts of emerging bees and tracked their survival, which lets you see the point at which the colony switches from producing short-lived summer bees to long-lived winter bees.

Their findings were:

• The longevity of a worker was inversely proportional to how much brood was left to rear when she emerged.
• Long-lived winter bees did not appear until brood rearing declined, and brood rearing declined when pollen ran out.
• Colonies given extra pollen kept rearing brood for longer and delayed the production of winter bees.
• Whether the transition was brought forward by depletion or pushed back by supplementation, the total number of winter bees produced was the same.
  

That last point is the one that ought to give every beekeeper pause. The feeding manipulation moved when winter bees are produced, not how many. In a companion paper the conclusion was put about as bluntly as research papers ever put things: manipulating pollen supply in the autumn does not affect the performance of winter bees⁶. Feeding pollen in the autumn did not make them heavier, or higher in protein, or more numerous by spring.

David Evans at The Apiarist has written the clearest accessible summary of this work that I know of (as usual), and it is worth reading.⁷

So the pollen patty some beekeepers feed to production colonies during September is, on this evidence, doing one of two things. Either nothing at all, or actively delaying the transition to winter bees, extending brood rearing into October, and giving the varroa a few more cycles of brood to breed in. Neither justifies the cost or effort involved.

Reconciling the Two

It seems at first sight that there are two bodies of research that contradict each other, but they don’t. It’s all in the timing. And timing is a big thing in beekeeping. Not the calendar watching kind, but the understanding of the colony, reading of the frames, and the conditions etc.

A dearth in mid-season is a nutritional problem. In late July there are still months of forage ahead. The absence of pollen does pure damage: cannibalised brood, undersized nurses, poor foragers three weeks later, and possibly a stalled brood nest that struggles to restart.

A dearth in autumn is a signal. The same lack of incoming pollen means something entirely different. It is the cue the colony uses to stop rearing brood and start building the long-lived population that will carry it to spring. Overriding that cue is not a kindness, to the bees or the beekeeper.

So, the situation observed on brood frames is the same in both cases, but the meaning is completely the opposite. The difference is entirely down to where you are in the season.

Which suggests a feeding sweet spot, at least in my part of the world, for full-sized colonies:

Period	What the pollen gap means	What I should do
Mid-July to mid-August	Damage	Feed protein if the signs are there
Balsam and ivy in flow	N/A	Nothing needed
September onwards	Signal to make winter bees	Leave it alone

The balsam is a useful trigger because it is unmissable. When it starts, stop feeding.

Mattila and Otis depleted or supplemented from late August and the season simply ended. Nobody, as far as I can find, has tested a dearth that dips in midsummer and then recovers, which is what a British summer gap followed by balsam and ivy looks like. My reasoning could be wrong, but I do believe that the widely reported higher than normal losses last winter had something to do with nutritional deficits (as well as beekeepers not properly caring for their livestock by omitting varroa treatments).

Standifer and colleagues concluded back in 1973 that protein feeding during a dearth can increase brood production but that there is no benefit during a pollen flow, and that pattern has been confirmed repeatedly since⁸. The most recent systematic review of pollen substitutes reached much the same place: they beat nothing at all, but they are far less useful when any natural pollen is available, and probably not economic then⁹.

Small Colonies Are Different

Everything above concerns full sized production colonies. Nucs and late splits are a separate problem and I do not think the same logic applies. They have to reach a viable cluster size in order to get through winter, and that hurdle must be overcome before anything else.

Döke and colleagues ran 60 colonies at three apiaries in central Pennsylvania and found that stock origin did not predict overwintering success, while the weight and population size reached before winter strongly did¹⁰. The paper is paywalled, so I am relying on the Entomological Society of America’s own write-up for the numbers¹¹. That reports a threshold effect, with colonies under 20 kg surviving poorly and those over 30 kg surviving at around 94 per cent, based on measurements taken in October.

There are two important caveats before anyone reaches for the luggage scales. First, that is net weight: adult bees, brood and stores, with the weight of boxes, frames, floor, crown board and roof subtracted. A National hive on the scales at 30 kg is not a 30 kg colony in Döke’s terms, it is more like 15 to 18 kg, which puts it below the bottom of the range. Second, this was Pennsylvania, where winter drops to around -4 ℃, which is a harder winter than most of Cheshire gets.

The National Bee Unit’s figures for the UK point the same way and are easier to apply, with around 20 kg of stores wanted in autumn, fewer than 9,000 bees giving a poor chance of survival, and five frames covered in bees as a working minimum¹².

A colony that fails that test is failing a headcount and anything that puts more bees in the box is worth considering. Mattila herself leaves the door open. She noted that a late season brood boost might be more advantageous in smaller colonies, but that there was no information available on the point. So “do not feed protein in autumn” is, for nucs, an extrapolation from an experiment that did not include them. I routinely make September nucs and feed plenty of syrup to them, but often there is pollen still coming in. If it wasn’t, I’d use patties too. We seem to have plenty of pollen in September in my area. Bad weather would mess that up, of course.

What It Costs

Randy Oliver says that the summer downsizing is genuinely useful to the colony, because it conserves honey stores for winter, but it also stresses an ageing, protein-deficient population and leaves it less able to cope with parasites and pathogens. Feed to prevent the downsizing and you have also produced unlimited fodder for varroa. It is a double-edged sword¹³.

There is also robbing. Feeding in a dearth, in August, in an apiary full of hungry colonies and wasps, is close to a worst case. Theoretically, you should feed late in the day, feed all colonies at once, reduce entrances, and not spill anything. I do that apart from the first point, as I’m travelling between different sites and can’t wait around until the evening.

We now also know that the patties themselves are imperfect. Nurse bees digest a much smaller fraction of the protein in artificial diets than they do from pollen⁹, and the sterol work I wrote about separately suggests that what is missing from a conventional substitute may matter more than the headline protein figure¹⁴.

Here’s A Plan

• Watch for the signs rather than the date. No band of fresh bee bread, a spotty pattern, eggs present but no young larvae, and larvae sitting in puddles rather than lakes.
• Feed protein through the gap if those signs appear, and stop when the balsam starts.
• Feed syrup in September to a weight target, because that is a stores problem and a well evidenced one, not an attempt to stimulate anything.
• Get the varroa treatment on early, and follow up in November/December with oxalic
• Judge nucs on size, and accept that some either need feeding or combining.
  

I am going to check my own loss records against site forage. Some bees were on borage or near to SFI crops in July, whereas others were not. I need to see what difference, if any, that makes. If anyone has done the same, it would be interesting to hear from you.

References

1. Scofield HN, Mattila HR (2015) Honey Bee Workers That Are Pollen Stressed as Larvae Become Poor Foragers and Waggle Dancers as Adults. PLoS ONE 10(4): e0121731. https://doi.org/10.1371/journal.pone.0121731
   
2. Di Pasquale G, Alaux C, Le Conte Y, et al. (2016) Variations in the Availability of Pollen Resources Affect Honey Bee Health. PLoS ONE 11(9): e0162818. https://doi.org/10.1371/journal.pone.0162818
   
3. Carroll MJ, Brown N, Huang E (2025) E-beta-ocimene and brood cannibalism: Interplay between a honey bee larval pheromone and brood regulation in summer dearth colonies. PLoS ONE 20(2): e0317668. https://doi.org/10.1371/journal.pone.0317668
   
4. Schmickl T, Crailsheim K (2001) Cannibalism and early capping: strategy of honeybee colonies in times of experimental pollen shortages. Journal of Comparative Physiology A 187: 541–547. https://doi.org/10.1007/s003590100226
   
5. Mattila HR, Otis GW (2007) Dwindling pollen resources trigger the transition to broodless populations of long-lived honeybees each autumn. Ecological Entomology 32: 496–505. https://doi.org/10.1111/j.1365-2311.2007.00904.x
   
6. Mattila HR, Otis GW (2007) Manipulating pollen supply in honey bee colonies during the fall does not affect the performance of winter bees. The Canadian Entomologist 139: 554–563. https://www.cambridge.org/core/journals/canadian-entomologist/article/abs/manipulating-pollen-supply-in-honey-bee-colonies-during-the-fall-does-not-affect-the-performance-of-winter-bees/D138EFB4649850F6902A73D930FCCB47
   
7. The Apiarist, Pollen and winter bees. https://theapiarist.org/pollen-and-winter-bees/
   
8. Topitzhofer E, Breece C, Wyns D, et al. (2022) Consumption of Supplemental Spring Protein Feeds by Western Honey Bee Colonies: Effects on Colony Growth and Pollination Potential. Journal of Economic Entomology 115(2): 417–429, citing Standifer et al. (1973). https://academic.oup.com/jee/article/115/2/417/6532102
   
9. Noordyke ER, Ellis JD (2021) Reviewing the Efficacy of Pollen Substitutes as a Management Tool for Improving the Health and Productivity of Western Honey Bee Colonies. Frontiers in Sustainable Food Systems 5: 772897. https://doi.org/10.3389/fsufs.2021.772897
   
10. Döke MA, McGrady CM, Otieno M, Grozinger CM, Frazier M (2019) Colony Size, Rather Than Geographic Origin of Stocks, Predicts Overwintering Success in Honey Bees in the Northeastern United States. Journal of Economic Entomology 112(2): 525–533. https://doi.org/10.1093/jee/toy377
    
11. Swett Walker M (2019) Colony Size Drives Honey Bees’ Overwinter Survival. Entomology Today, Entomological Society of America, 9 January 2019. https://entomologytoday.org/2019/01/09/colony-size-drives-honey-bees-overwinter-survival/ (This is the source of the 20 kg and 30 kg figures quoted above; they do not appear in the paper’s abstract, which is all that is publicly accessible.)
    
12. National Bee Unit, Preparing Your Hives for Winter (Fact Sheet 19). https://www.nationalbeeunit.com/assets/PDFs/3_Resources_for_beekeepers/Fact_Sheets/Fact_19_Preparing_your_hives_for_winter.pdf
    
13. Oliver R, Understanding Colony Buildup and Decline: Part 10, Summer Downsizing and Varroa. Scientific Beekeeping. https://scientificbeekeeping.com/understanding-colony-buildup-and-decline-part-10/
    
14. See my earlier post, Better Bee Food. https://thewalrusandthehoneybee.com/better-bee-food/