Calculating the flower preference of pollinators

Many of conservation efforts regarding pollinators include planting the preferred flowers of pollinators in order to provide them with quality floral resources. Determining which flowers pollinators prefer is challenging because flower abundance can confound results as a pollinator is most likely to visit an abundant flower. Various researchers have developed methods for mathematically calculating which flowers pollinators prefer. I compared preference metrics by L. Sutter, D. H. Johnson, D. B. Robson, K. P. Maia, and P. P. A. Staniczenko. Each of these preference metrics make sense in theory and have been published in peer-reviewed literature. However, when I applied these metrics to a novel dataset, I found that they all produce markedly different lists of preferred flowers. For example, chickpea milkvetch (Astragalus cicer) was the second most preferred plant in the lists produced by Johnson and Staniczenko metrics, the fifth least preferred plant in the list produced by the Robson metric, and the least preferred plant in the lists produced by the Sutter and Maia metrics. All these metrics attempt to reconcile with the problem of abundance in a different way but fail to recognize other likely important factors such as flower community composition, fine-scale spatial distribution, and phenology. Therefore, those looking to use any of these metrics need to be aware of the bias inherent in each metric to make an informed decision regarding which one or ones to use.