Winter losses – not quite what they seem

Each Spring, we survey our treatment-free members to ascertain winter colony losses, and compare with the national surveys of conventional beekeepers’ losses. Over 8 years we’ve built up an interesting comparison graph, but though our treatment-free results (blue line) can be readily compared to conventional losses (red) at first glance, there’s more to it when you burrow into the data.

The really interesting finding

I was suspicious about our high losses this year so drilled down into the data more, creating a map of losses by postcode –

Do you see the pattern? Almost all the losses were in the Oxford valley. (The city covers postcodes OX1 – Ox4 and OX33.) Losses outside Oxford were only 15%!

Bear in mind that the sample numbers are low, so you can’t really draw firm conclusions here. But it alerts us to watch for consistent geographical clustering issues. One day I may get round to sifting through all our 878 reports over the last 8 years, to see if this geographical bias is consistent.

Anecdotally, we heard several reports of spectacular losses from nearby conventional apiaries, like 37/40 hives at one location, which do not seem to be reflected in the BBKA survey (perhaps those beekeepers were embarrassed to admit to huge losses?).

Cause of losses

The obvious unusual environmental factors are:

• Last Autumn was exceptionally warm for several weeks longer than usual, resulting in a long flow of ivy, followed by a sudden sharp drop to winter temperatures.
• Oxford is in a flood plain surrounded by hills, and has minor flooding every few years.
• Oxford has a significant heat island effect. Events like cherry trees blossoming and swarms occur about 2 weeks before north Oxfordshire.

The hives that died did not seem to have a disease. They didn’t have varroa. They had plenty of capped honey, much of it ivy. Typically they had been strong colonies going into winter and seemed very damp inside. Gareth suggests they laid normal foragers too late into Autumn, tricked by a late Ivy nectar flow, and left raising winter bees too late (i.e. they died of old age with no young winter bees, who are metabolically different and live several months, surviving to Spring).

It was noticeable that they had accumulated a lot of ivy honey – much uncapped. This seems to have been an incidental issue. They couldn’t finish processing and capping it when the temperature dropped below ~12C. Then this open nectar, plus the condensation from large colonies, created a damp environment, which tends to promote mould. So unusually, it was the large colonies that were at risk last winter.

Deeper analysis for data nerds

Our own results and analysis were complete in May, but we held off publishing this until we had the BBKA survey results to compare with. BBKA publication was delayed by more urgent issues – Asian hornets! This year 56 nests were destroyed in the UK. That’s very worrying.

The BBKA results were published in November’s BBKA News and mention the interesting statistic that “74% of beekeepers reported using at least one varroa treatment over Autumn/Winter”, which could be interpreted as 26% of their 4140 respondents being treatment-free.

The dotted orange line on the graph is the results for our local BBKA region, which is occasionally reshuffled. This year, Oxfordshire is in the Central England area.

When comparing TF and conventional losses, keep in mind they’re not like-for-like results in summer. Most of our respondents aren’t just treatment-free, we’re pretty much leave-alone beekeepers which means we don’t requeen. We let our colonies run through their natural cycle from swarm through supersedures and eventual queen failure after a few years. Commercially oriented beekeepers requeen every 1-2 years, at which point they effectively restart a new colony, but they only record a colony death if the entire colony dies. Of course, it’s rarely possible to requeen in winter so our loss figures are probably directly comparable at this time, but any summer losses really can’t be compared.