DOI: 10.1016/j.ophoto.2024.100063
Take-home message: Pollen colour is useful for automated bee monitoring and for assessing pollen colour diversity, but colour alone is not reliable enough to identify the botanical origin of corbicular pollen loads.
Why pollen colour matters for bee monitoring
Pollen analysis is an important tool for studying plant–pollinator interactions, bee nutrition, biodiversity and landscape-level resource availability. For honey bees, corbicular pollen loads provide direct information about the floral resources collected by foragers.
Traditional pollen identification is usually based on microscopy or other laboratory methods. These approaches can be accurate, but they are time-consuming and require specialist expertise. In contrast, colour-based pollen assessment is simpler and can be combined with camera-based or automated monitoring systems. This makes pollen colour attractive for precision beekeeping, ecological monitoring and large-scale biomonitoring.
The central question
The key question of this study was whether pollen colour alone can be used to classify the botanical origin of corbicular pollen loads. This is particularly relevant for automated bee monitoring systems, where camera images of pollen loads could potentially be used to infer which floral resources bees are using.
What we analysed
We analysed more than 85,000 corbicular pollen loads representing 30 major pollen types. The aim was to quantify natural colour variation within pollen types and to test how strongly different pollen types overlap in colour space.
Main findings
The results show that natural colour variation within individual pollen types is often large. In many cases, pollen from the same botanical origin can appear in clearly distinguishable colours. This within-type colour dispersion limits the reliability of colour-based botanical classification.
When using colour alone, the overall correct pollen type classification rate was 67%. However, classification accuracy strongly depended on the pollen type. Some pollen types with distinct colours could be classified with high accuracy, while rare pollen types with common colours could not be classified reliably.
Implications for automated pollen monitoring
These findings do not mean that pollen colour is useless. Colour remains valuable for estimating pollen colour diversity, comparing resource patterns and supporting automated pollen monitoring. However, colour should not be interpreted as a stand-alone identifier of botanical origin.
Future monitoring systems will likely need to combine pollen colour with additional information, such as pollen morphology, multispectral imaging, flowering phenology, spatial context or laboratory-based reference data. This is especially important when pollen data are used to draw conclusions about plant diversity, bee nutrition or landscape quality.
Conclusion
This study highlights both the potential and the limitations of colour-based pollen analysis. Pollen colour can support automated bee and landscape monitoring, but its natural variation limits reliable botanical classification. For robust biomonitoring, colour should be treated as one informative feature among several, not as a precise stand-alone marker of plant identity.
Publication
Borlinghaus, Parzival; Tausch, Frederic; Odemer, Richard
Natural color dispersion of corbicular pollen limits color-based classification Journal Article
In: ISPRS Open Journal of Photogrammetry and Remote Sensing, 2024.
@article{nokey,
title = {Natural color dispersion of corbicular pollen limits color-based classification},
author = {Parzival Borlinghaus and Frederic Tausch and Richard Odemer},
url = {http://vibee-project.net/wp-content/uploads/2024/04/Borlinghaus_et_al_pollen_color_dispersion.pdf},
doi = {10.1016/j.ophoto.2024.100063},
year = {2024},
date = {2024-04-16},
urldate = {2024-04-16},
journal = {ISPRS Open Journal of Photogrammetry and Remote Sensing},
abstract = {Various 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.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Various 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.
Impact
