Color dispersion of corbicular pollen loads - Supplementary Material

Dataset DescriptionThis dataset belongs to paper [1] where Gaussian Mixture Models were used to establish a mapping between 62,343 corbicular pollen colors and their respective pollen types. The data set contains 30 different pollen types, including 14 of the 31 most important European pollen types.We provide the following resources in this repository:Folder pollen samples contains data for 86 samples / 62,343 corbicular pollenCSV filesColumn 1: Dataset identifierColumn 2: Sample identifierColumn 3-5: L*A*B* pollen colorsColumn 6-End: pollen type probabilitiesJSON filesFor each dataset and sample: The estimated mean and covariance parameters of the Gaussian Mixture Models are reported along with the relative and absolute pollen type abundance.PDF/PNG filesFor each dataset and sample: Scatter plots show the pollen colors (left) morphological pollen type analysis (middle) and GMM reconstruction (right).Folder pollen typesPDF/PNG filesFor each pollen type: A scatter plot of all corbicular pollen colors that were attributed to the type.File Method_Pollen_Determination_Mayen_en.pdfA protocol describing the method of morphological pollen type determination carried out at the Expert Center for Bees and Beekeeping, Mayen, GermanyDetailed information on the sampling locations, periods and intervals can be found in the paper.AbstractVarious methods have been developed to assign pollen to its botanical origin. They range from technically complex approaches to the less precise but sophisticated chromatic assessment, in which the pollen colors are used for identification. However, a common challenge lies in the similarity of colors of pollen from different plant species. The advent of camera-based bee monitoring systems has sparked renewed interest in classifying pollen based on color and offers potential advances for honey bee biomonitoring. Despite the promise of improved sensor accuracy, a critical examination of whether color diversity within a single species may be the primary limiting factor has been lacking. Our comprehensive analysis, which includes over 85,000 corbicular pollen from 30 major pollen species, shows that the average color variation within each species is distinguishable to a human observer, similar to the difference between two dissimilar colors. From today's perspective, the considerable color variation within a single pollen source makes the use of color alone to classify pollen impractical. When picking a single pollen color from the entire dataset, we report a correct pollen type classification rate of 67 %. The accuracy was highly dependent on the type and ranged from 0 % for rare types with common colors to 99 % for distinct colors. The large color dispersion within species highlights the need for complementary methods to improve the accuracy and reliability of color-based pollen identification in biomonitoring applications.References[1] P. Borlinghaus, R. Odemer and F. Tausch, ‘Natural color dispersion of corbicular pollen limits color-based classification’, Open Journal of Photogrammetry and Remote Sensing, 2024 (accepted).