Thomas' Plant-Related Blog

On plant science. Mostly.

Do bees benefit from a balanced diet?

with 2 comments

We discussed this paper at a journal club in our department yesterday (Monday 1st February). Some of our thoughts are below.

Although the media coverage of this study played heavily on the link to colony collapse disorder (which is causing honeybee colonies to die off around the world), the authors only allude to it in one sentence near the end. The study tries to show that bees fed a mixture of pollen from different plants are healthier than those fed only one type of pollen, which is interesting because modern farming tends to create large areas of just one species. It’s not entirely convincing, though: of the four things they measure, only one (glucose oxidase) convincingly shows that pattern, and even that’s not a terribly dramatic difference (see the graph). The researchers deserve kudos for showing all four results, even though I’ve just picked out one here:

Bar chart of glucose oxidase levels in bees fed different pollen diets.

Glucose oxidase levels in bees fed different diets. The highlighted two bars on the right are from the mixed diets, the 'control' on the left had a diet without pollen, and the other four were fed one type of pollen each. Numbers underneath are the proportion of protein in the pollen. From Alaux et al. 2010.

Interestingly, glucose oxidase is used as an antiseptic, to kill off bacteria in honey and food for the larvae. So it’s colony defence against disease that is improved, not just the individual bees’ resistance.

Much of our discussion focussed on the methods, which leave something to be desired. The two pollen mixtures both included two types of pollen (willow and maple) which weren’t in any of the single-pollen diets, so the apparent effect of blending pollen could be down to something in one of those types. We also questioned whether the pollen sources are relevant to what bees naturally eat. Some of them are wind-pollinated trees, which sounded counterintuitive, but in fact I can find most of the plants in a list of bee pollen sources, and the French company that supplied the pollen actually collects it from beehives.

The experiment didn’t run for very long, at just ten days, and was only done on adult bees, while a bee’s diet as a larva could have important effects. This might go some way to explain why the differences between groups are fairly small. Ideally, it would be interesting to limit the pollen available to entire hives for a full generation and study the effects, although that would take much more time and funding. It would also be good to look at how the differences in ‘immunocompetence’ measured here relate to the bees’ response to a disease. The study mentions a couple of papers which have looked at diet and immune responses, although neither of them are in bees.

Moving on to the results of the study, we mentioned that the different parts of the immune system may be involved in trade-offs; the more of one thing a bee makes, the less it can produce of another. So it might be interesting to combine the four different variables measured into a general pattern, and then look at how diet affects that; although of course that shouldn’t get in the way of publishing the simple data.

Among their results, one measure, the count of haemocytes (which are kind of like white blood cells), did something unexpected: it went up in the control bees, which were fed no pollen. There are different types of haemocytes, and not all of them are involved in immunity, but it’s still an odd result. The study suggests that perhaps the greater number is compensating for something, for example if each cell is less active, but it didn’t strike me as convincing.

Although the researchers focussed on the protein in the pollen, we reckoned that the effect of mixed-pollen diets was more likely to hinge on micronutrients such as vitamins, since they’re numerous, and perhaps no one type of pollen would contain them all.

Finally, a few of our more speculative thoughts:

  • How does this affect solitary bees, which by definition don’t have a hive to protect? Solitary bees are probably important pollinators for many wild species, whereas honey bees are used commercially to pollinate many crop species, especially fruits.
  • Does the ‘hygiene hypothesis’ (that the immune system goes wrong if it’s not exposed to diseases and can’t ‘learn’ about them) apply to bees? Microbial symbionts might help about against disease, but could spraying of hives with antibiotics have killed off the symbionts, leaving the bees more vulnerable?
  • Glucose oxidase and phenol oxidase (another thing which was measured) probably increase as bees get older; could the different pollen diets have been affecting how quickly the bees aged?
  • Do worker bees have different castes, and would their diets differ? A bit of searching suggests that they change tasks over their lifespan of a few weeks (‘temporal castes’ to use the technical term; see this paper), starting out inside the hive and ending up as foragers. It’s possible that foragers (older bees) have a less varied diet, if they only eat what they collect, while the bees in the hive dine on pollen from many foragers, but this is entirely guesswork.


Alaux, C., Ducloz, F., Crauser, D., & Le Conte, Y. (2010). Diet effects on honeybee immunocompetence Biology Letters DOI: 10.1098/rsbl.2009.0986


Written by Thomas Kluyver

2 February, 2010 at 11:06 pm

Posted in Papers

Tagged with , ,

2 Responses

Subscribe to comments with RSS.

  1. Social comments and analytics for this post…

    This post was mentioned on Twitter by ResearchBlogs: Do bees benefit from a balanced diet?

    uberVU - social comments

    3 February, 2010 at 2:06 am

  2. There was also the point about how the need to make the pollen diets more realistic in order to make better inferences about what would happen in the wild. If honeybees generally forage within a kilometre from the hive, then the number of different types of pollen they will be able to collect will be limited by the types of habitat within range. For example, the authors use a Cistus sp. and an Erica sp. amongst their pollen types, in England it would be unlikely (not impossible, given limestone heaths and geological transitions) to find members of these families growing in close association, except in a horticultural setting.

    An associated point relates to phenology, trees in England generally flower in spring, whereas members of the Cistaceae (Helianthemum, mainly) would be in the summer, and Erica slightly later. So, a honeybee would be unlikely to have access to all these resources all at once, even if they were local (though I suppose there might be a cumulative effect of variety through the year).

    I actually found it very interesting to discover that honeybees collect pollen from anemophilous tree genera such as Quercus (according to a study in the journal Grana, this pollen can make up a significant part of a hive’s pollen collection), it certainly makes one think a bit more about the strength of flower syndromes for attracting pollinators.

    It would be interesting to take some hives into the middle of at least a kilometre or so of homogeneous habitat, choosing vegetation communities across a range of species richnesses (Kew > limestone grassland > heathland, for example), then see what happens to those GOX levels,,,


    3 February, 2010 at 10:07 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: