Contrasting with the great alarm on bee declines, it is astonishing how little basic biology we know about bees, including on abundant and widespread species that are becoming model species. Plant-pollinator relationships are one of the cornerstones of bee ecology, and researchers are increasingly documenting bees' diets. However, we rarely know which effects feeding on different flowers has on bees' health. This paper (Tourbez et al. 2023) uses an elegant experimental setting to test the effect of heather pollen on bumblebees' (Bombus terrestris) reproductive success. This is a timely question as heather is frequently used by bumblebees, and its nectar has been reported to reduce parasite infections. In fact, it has been suggested that bumblebees can medicate themselves when infected (Richardson et al. 2014), and the pollen of some Asteraceae has been shown to help them fight parasites (Gekière et al. 2022). The starting hypothesis is that heather pollen contains flavonoids that might have a similar effect. Unfortunately, Tourbez and collaborators do not support this hypothesis, showing a negative effect of heather pollen, in particular its flavonoids, in bumblebees offspring, and an increase in parasite loads when fed on flavonoids. This is important because it challenges the idea that many pollen and nectar chemical compounds might have a medicinal use, and force us to critically analyze the effect of chemical compounds in each particular case. The results open several questions, such as why bumblebees collect heather pollen, or in which concentrations or pollen mixes it is deleterious. A limitation of the study is that it uses micro-colonies, and extrapolating this to real-world conditions is always complex. Understanding bee declines require a holistic approach starting with bee physiology and scaling up to multispecies population dynamics.
Gekière, A., Semay, I., Gérard, M., Michez, D., Gerbaux, P., & Vanderplanck, M. 2022. Poison or Potion: Effects of Sunflower Phenolamides on Bumble Bees and Their Gut Parasite. Biology, 11(4), 545. https://doi.org/10.3390/biology11040545
Richardson, L.L., Adler, L.S., Leonard, A.S., Andicoechea, J., Regan, K.H., Anthony, W.E., Manson, J.S., & Irwin, R.E. 2015. Secondary metabolites in floral nectar reduce parasite infections in bumblebees. Proceedings of the Royal Society of London B: Biological Sciences 282 (1803), 20142471. https://doi.org/10.1098/rspb.2014.2471
Tourbez, C., Semay, I., Michel, A., Michez, D., Gerbaux, P., Gekière A. & Vanderplanck, M. 2023. Heather pollen is not necessarily a healthy diet for bumble bees. Zenodo, ver 3, reviewed and recommended by PCI Ecology. https://doi.org/10.5281/zenodo.8192036
DOI or URL of the preprint: https://doi.org/10.5281/zenodo.8116028
Version of the preprint: 2
Thanks for implementing the required changes. The paper now is very clear and provides reproducible evidence of how heather pollen affects bumble bees. I have a couple of very minor requests before final acceptance.
First, I missed in the text if Bombus terrestris naturally forages on heather (and in which proportions if known). This is very relevant for interpreting the results, and can be briefly stated in the introduction, or maybe in discussion. Also, the Cox models ignore random effects, right? Just state this, and justify why this is ok in methods.
Second, there are a few grammatical issues. e.g. line 144: "parasiteS … fed WITH … WERE parasite free"; line 356 needs a space in emergenceresulting.
DOI or URL of the preprint: https://doi.org/10.1101/2023.04.06.535809
Version of the preprint: 1
This is a very nice experimental test to assess the effect of diet in different aspects of bumblebees fitness. Given the current environmental challenges, understanding the physiological mechanisms by which floral resources (both pollen and nectar) affect bumblebees capacity to reproduce and withstand parasites is timely.
The manuscript is clear and the experiments robust. Both reviewers apreciated the manuscript and have some minor comments to enhance it. As you will see, Cristina Botías has interesting caveats on some experimental aspects. I think an honest discusion on the limitations or potential blind spots of the experimental dessign will strengthen the manuscript.
I would also recommend focusing the results around effect sizes, rather than on p-values. There are several ways to do so. First, estimates (for cathegorical factors those are usually mean and SE differences between factors) can be provided in a table in the appendix. This is important specially for researchers doing further meta-analysis. Extracting those from figures is painful. Second, this same estimates can be addedd when relevant in text, for example, instead of stating a significant increase in parameter X, you can mention which is the raw of percentual mean increase in this parameter. I think this will reinforce some of your results, as highlighting the effect sizes (e.g. a 20% increase in parameter X) will help the readers.
I am looking forward to read the revised version.