Cluster: A New Application for Spatial Analysis of Pixelated Data for Epiphytotics.

Area of Science:

• Plant Pathology
• Ecology
• Digital Image Analysis

Background:

• Spatial analysis of plant disease epiphytotics is crucial for understanding pathogen ecology and disease dynamics.
• Current data collection methods for spatial analysis are time-intensive and require significant investment.
• Optimizing plant disease management strategies relies on accurate spatial pattern analysis.

Purpose of the Study:

• To develop a novel, user-friendly approach for spatial analysis of pixelated data in digital imagery.
• To create a stand-alone desktop application, Cluster, to facilitate this analysis.
• To enable statistical testing of spatial distribution patterns of disease clusters.

Main Methods:

• The Cluster application allows users to isolate target clusters by defining nontarget colors and adjusting a threshold.
• It calculates cluster attributes such as percent area, centroids, and orientation angles.
• Stochastic simulations and t-tests are employed to statistically assess the randomness of cluster distribution.

Main Results:

• Cluster quantifies cluster attributes including pixel number, dimensions, orientation, and length/width ratio, outputting data as a spreadsheet.
• The application facilitates manual or automatic deselection of anomalous clusters based on size.
• Statistical testing compares observed inter-cluster distances against simulated random distributions.

Conclusions:

• The Cluster application provides an efficient and statistically robust method for spatial analysis of plant disease patterns.
• It aids in testing hypotheses related to pathogen ecology and disease dynamics across various spatial scales.
• The tool is available as a free download for Apple computers, supporting research in plant pathology and ecology.