Kubelka-Munk theory applied to Flower Coloration Interpretation

Kubelka-Munk theory applied to Flower Coloration Interpretation
|| ||

Researchers from the University of Groningen, Netherlands are uniting the disciplines of biology and physics in the quantitative analysis of the reflectance spectra of flowers. The recent work on the Coloration of the Chilean Bellflower has important implications in the study of evolutionary biology and the relationship between plants and pollinators. Furthermore, this work contributes to our understanding of pigmentation and light scattering, as well as adding valuable reflectance spectra data for the Nolana paradoxa to the world’s catalog of reference spectra data.

 Applying Biology and Physics

The application of the Kubelka-Munk layer stack theory to plant colorization analysis also has commercial and industrial implications related to materials, coatings, dyes, and paints; especially in the world of semiconductor fabrication and automobile and auto parts manufacturing.


flower microFlower Petal Anatomy and the Stack Model

A great deal of effort has gone into studying the physical structure of plants, and in particular the colorization of flowers, in large part due to the relationship in signaling pollinators. Dr. Casper J. van der Kooi, professor of plant physiology at the Groningen Institute for Evolutionary Life, has done a tremendous amount of work to advance quantitative analysis of light interaction to explain flower coloration.

The petals of a flower are made up of several layers consisting of outer epidermal layers and inner pigmented and light scattering layers of the mesophyll. Pigment layers selectively absorb incident light within a specific wavelength, while light scattering structures and vacuoles backscatter the incident light in all directions, this diffuse reflectance creates a consistent visual effect when viewed from multiple angles.

In his recent work partnered with Dr. Doekle G. Stavenga, professor of computational physics at the Zernike Institute for Advanced Materials, University of Groningen, Dr. van der Kooi studied the Chilean Bellflower, Nolana paradoxa, which has a distinct color differentiation between the vivid, saturated purple of its adaxial (upper) surface and the unpigmented abaxial (lower) surface.

Previous efforts to explain the interaction of light within a plant to produce flower color relied on geometrical optics, however plant structures are not homogenous, making direct optical analysis inconvenient. Furthermore, these methods require knowledge of essential optical parameters such as a refractive index and absorption coefficients of the component structures; however, optical constants of botanical samples such as these are simply unavailable in the current body of knowledge. Alternatively, the Kubelka-Munk theory for absorbing and diffusely scattering media allows the absorption and scattering coefficients to be derived from measured transmission and reflectance spectra.

Stavenga and van der Kooi have used this method with apparent success in previous work on a non-invasive method for estimating chlorophyll content using the Kubelka-Munk theory and treating a plant leaf as a stack of absorbing and reflective plates. In this related spectra analysis work sets the basis for future quantitative analysis and comparison of the coloration strategies of flowers.

Methods and Results

main spectraTo describe flower coloration, Drs. van der Kooi and Stavenga developed an optical model treating a flower petal as a stack of layers, and then applied the Kubelka-Munk theory for diffuse scattering and absorbing media to the layers. This method uses the combined stack reflectance and transmission spectra and, with the number and relative thickness of the layers known, can estimate the reflectance and transmission spectra for each layer.

To measure the physical structure and distribution of pigments, these researchers examined cross sections of bellflower petal under magnification and found the Nolana paradoxa petals have a pigmented, strongly scattering adaxial mesophyll layer and an unpigmented moderately reflective abaxial layer.

To capture spectrophotometric measurements, researchers employed a deuterium-halogen lamp (AvaLight-D(H)-S) to deliver light via optical fiber to an integrating sphere (AvaSphere-50-Refl). With the corolla (the sum group of petals of a flower) positioned in the sphere and illuminated directionally, the reflected light was then captured by a second optical fiber and collected with the AvaSpec-ULS2048XL-USB2 spectrometer with a 2048 pixel back-thinned CCD image sensor.

Stavenga and van der Kooi then interpreted this measured transmission and reflectance spectra for the corolla using the Kubelka-Munk absorbing stack layer model to estimate the transmission and reflectance spectra of the observed layers of the petal interior.

Subsequent measurements sought to confirm the results of applying the layer stack theory by measuring the absorption spectra for the pigmented adaxial layer and the reflective properties unpigmented abaxial layer, individually. These subsequent experiments performed on the isolated layers appeared to confirm the spectral analysis derived from using the stack layer modeling.

pigment-testConclusion and Continuing Research

Casper van der Kooi and Doekele Stavenga has greatly advanced our understanding of how plants use light. This team’s growing body of work, together with other frequent contributors, has investigated the physical interaction of light that gives flowers their colorful appearance and how competition for pollinators affects the spectral display of flowers.

This research into the efficacy of the Kubelka-Munk method can have potentially wide-reaching impact on both the study of plants with potential agricultural applications, as well as potential relevance to other applications which concern stack metrology.




van der Kooi, CJ; ElzengaJTM; Staal M; Stavenga DG. (2016) How to Colour a Flower: on the Optical Principles of Flower Coloration. Proceedings of the Royal Society B283:20160429. LINK

van der Kooi, CJ; Pen, I.; Stall, M. (2016) Competition for Pollinators and Intra-communal Spectral Dissimilarities of Flowers. Plant Biology Journal 18:1 10.111/plb12328 LINK

Ozawa, A.; Uehara, T.; Sekiguchi, F.; et al. (2009) Spectral Analysis of Scattered Light from Flowers' Petals Optical Review 16:458. doi:10.1007/s10043-009-0088-2 LINK

van der Kooi, C.; Wilts, B.; Leertouwer, H.; Staal, M.; Elzenga, T.; Stavenga, DG. (2014) Iridescent Flowers? Contribution of Surface Structures to Optical Signaling. New Phytologist 203:2 10.1111/nph.12808 LINK

The Anatomy of Flower Color (2016) Phy.org May 10, 2016 Accessed February 24, 2017. LINK



Related items

  • Miniature Spectrometers, a Powerful Tool for Laser Characterization

    Regardless of whether one is producing low-cost vertical cavity surface emitting laser diodes (VCSELS) for consumer electronics or using a tunable mode-locked Ti:Sapphire laser for multi-photon fluorescence microscopy, having accurate spectral data during the manufacturing process is critical for ensuring quality, repeatable product.  Unfortunately, the overwhelming number of laser engineers are still relying solely on large and costly optical spectrum analyzers (OSAs) for spectral testing.  While these devices provide excellent spectral resolution and dynamic range, the size and cost often limit their ability to be deployed at multiple points throughout the manufacturing environment [1].

  • November Spectra of the Month: NIR Spectra In Vivo

    Biomedical Applications was November's theme at Avantes. Near-infrared is the preferred wavelength range for biomedical applications because of the therapeutic optical window in the near-infrared. Using the high-powered AvaSpec-NIR256-1.7-EVO, Avantes engineers collected NIR spectra of two sample locations: fingernails, and ear cartilage, from three volunteer subjects.

  • The Role of NIR Spectroscopy in the Treatment of Osteoarthritis

    Medical researchers studying osteoarthritis are exploring the use of near-infrared reflection spectroscopy to characterize bone tissue and improve standards of care for arthroscopic surgeries. There are a large number of uses for near-infrared spectroscopy in the field of biomedical research including non-invasive monitoring, blood testing, protein identification, and more. The near-infrared range is so important because it offers the best quality of reflection spectra measurements, with the least scattering and absorption between 650 nm and 1900 nm.  Spectral measurements in the near-infrared are non-invasive and non-destructive, making this particularly attractive for the potential development of clinical applications.  

    The Avantes family of NIR instruments offers the best mix of sensitivity and resolution for medical applications with NIR spectroscopy. Our next-generation EVO electronics provide lowers noise and offer the latest high-speed communications options for fast data when working in vivo.