Quantitative description and discrimination of butterfly wing patterns using moment invariant analysis

Area of Science:

• Entomology
• Image analysis
• Biometrics

Background:

• Objective quantification of insect wing patterns for individual and population characterization has been limited.
• Existing methods struggle with objective measurement of subtle pattern variations.
• Wing pattern analysis is crucial for insect identification, population genetics, and evolutionary studies.

Purpose of the Study:

• To demonstrate the statistical differences in wing patterns using moment invariant data sets.
• To assess the efficacy of moment invariants in objectively quantifying insect wing patterns.
• To explore the application of this method for differentiating insect groups.

Main Methods:

• Digitized images of the speckled wood butterfly (Pararge aegeria) wings were analyzed.
• Moment invariant data sets were automatically generated from the wing images.
• Multivariate analyses of variance (MANOVA) were performed on the moment data.

Main Results:

• Moment invariants successfully detected pattern differences between wing surfaces, even with simulated fading and damage.
• MANOVA reliably differentiated groups of butterflies based on sex, geographical origin, and culture history.
• The method proved effective in objective wing pattern quantification.

Conclusions:

• Moment invariants offer a robust and objective method for quantifying insect wing patterns.
• This technique has significant potential for insect identification, population studies, and biodiversity assessment.
• Further improvements to the moment methodology could enhance its applicability in entomological research.