We disagree with Ian Boyd's implication that bias may have influenced the commissioning and publication of research on pollinator declines (Nature 501, 159–160; 2013).

Our paper on falls in European bee-species richness (J. C. Biesmeijer et al. Science 313, 351–354; 2006), along with others on honeybee colony collapse and bumblebee declines, prompted widespread public concern. Subsequent decisions in continental Europe and the United Kingdom to commission further research in this area therefore seemed sensible and proportionate.

These calls for research used “pollinator declines” as a convenient shorthand, not to steer the work. This is borne out by results from the studies funded, including our own, indicating that past declines in some pollinator groups may have recently slowed or even partially reversed (L. G. Carvalheiro et al. Ecol. Lett. 16, 870–878; 2013).

Publication bias undoubtedly occurs, but it can be identified only by reviewing whole fields, not individual papers. This should be addressed as part of a systematic review when policy issues arise, as carried out by the UK's Parliamentary Office of Science and Technology or by independent research teams.

We agree that there are uncertainties in our conclusions, as Boyd suggests; indeed, our papers list strong caveats pertaining to our data sets and methodology, which were largely ignored by the media.

A national pollinator-monitoring programme, recommended recently in a parliamentary report, would provide much more robust estimates of pollinator trends in future.