The significant decline in terrestrial insect pollinators and the resulting impact on pollination services has gained considerable media attention. Habitat loss is a major driver of these declines, leading to the promotion of pollinator-friendly gardening as a mitigation strategy. This involves recommending specific flowering plants believed to benefit pollinators. Various resources, including books, websites, seed mixtures, and labels, provide such recommendations. However, the accuracy, specificity, and timeliness of these recommendations remain unclear, with 'pollinator-friendly' labels often being generic and not considering the diverse needs of various pollinator species (bumblebees, honeybees, butterflies, hoverflies, beetles). This study aims to investigate the concordance between existing pollinator-friendly plant recommendations and actual bumblebee plant use observed through citizen science data.

Existing literature highlights the decline in pollinator populations and the importance of pollinator-friendly gardening. Studies have shown the effectiveness of pollinator-friendly plant schemes in specific settings, like garden centers. However, concerns have been raised about the lack of evidence-based specificity and the potential for outdated or unoriginal recommendations. Previous research has also pointed to the limitations of generic 'pollinator-friendly' labels, which fail to account for the diverse needs of different pollinator species. This study builds on this literature by directly comparing existing recommendations with real-world data on bumblebee plant use obtained via citizen science.

The study utilized data from BeeWatch, a UK-wide citizen science program where members of the public submit photographs of bumblebees visiting plants, along with location and date information. The researchers compiled pollinator- and bumblebee-friendly plant lists from 23 different UK sources. These lists contained 465 plants recommended as pollinator-friendly and 376 plants specifically for bumblebees. The researchers then compared the ranked lists from the practitioner sources to BeeWatch's 6429 records of bumblebee plant visitation (334 plant species). Pearson's correlations were calculated to assess the relationships between practitioner recommendations and BeeWatch observations. Rarefaction analysis was used to estimate the mean number of plant species used by different bumblebee species, accounting for varying sample sizes. The data was also analyzed to determine the relative abundance of specific plants used by different bumblebee species. Analyses were performed using R.

The analysis revealed a low correlation between practitioner recommendations and BeeWatch observations. The correlation between practitioner's bumblebee-friendly plant list and BeeWatch records was only r=0.57, and only marginally higher than the correlation between BeeWatch records and the practitioners' pollinator-friendly plant list (r = 0.52). This indicates a significant mismatch between the plants recommended as beneficial for pollinators and those actually used by bumblebees. The study further showed a high degree of similarity between pollinator-friendly and bumblebee-friendly plant lists (r=0.75), highlighting the lack of specificity in recommendations. Different bumblebee species exhibited distinct food plant preferences; there was no single list of plants universally beneficial to all bumblebee species. Although some plants like lavender were popular, they constituted only a small proportion (at most 11%) of the records for any single bumblebee species. Most bumblebee species used a wide range of plants, contrary to the high level of agreement among practitioner recommendations. Rarefaction analysis indicated that the number of plant species used by each bumblebee species increased with the number of observations. The study found that many of the plants recommended by practitioners were not observed in BeeWatch records, with taller plants and some wildflowers under-represented. Conversely, some common garden plants frequently used by bumblebees were absent from practitioner recommendations.

The findings highlight a significant disconnect between current pollinator-friendly plant recommendations and actual bumblebee foraging behavior. The lack of specificity and the limited predictive power of existing recommendations underscore the need for more dynamic and data-driven approaches. The reliance on generic labels and outdated information fails to capture the complexity of pollinator-plant interactions. While the 'pollinator-friendly' label serves as an important initial step in engaging the public, continued use of such generic recommendations without updated, species-specific information can lead to confusion and limited effectiveness. The limitations of BeeWatch data – primarily that it reflects plant use rather than attractiveness – need to be considered, but still present valuable insights.

This study demonstrates the urgent need to update current garden plant recommendations for pollinators. Existing recommendations lack specificity and are not always aligned with real-world bumblebee foraging behavior. The authors strongly suggest incorporating live citizen science data to create dynamic, species-specific planting guides. This approach can greatly improve the effectiveness of pollinator conservation efforts and support more effective public engagement with pollinator conservation.

The BeeWatch data primarily reflect plant use and not plant attractiveness. The opportunistic nature of the data collection might introduce biases, as it does not incorporate factors such as recording effort, weather conditions, or the abundance of floral resources. The study focused on bumblebee species in the UK; the findings might not be directly generalizable to other regions or pollinator groups.