A Protocol for Extracting Structural Metrics From 3D Reconstructions of Corals

Abstract

The 3D structure of individual coral colonies provides insights into their ecological functioning. While structure from motion techniques make it possible to reconstruct 3D models of coral colonies based on overlapping images, the extraction of relevant metrics of complexity in a reproducible way remains challenging. We present a method and associated scripts for the 3D reconstruction of coral colonies from in-situ images and the automatic extraction of eleven structural complexity metrics, designed to be run in widely-used software packages. The metrics are designed to capture aspects of complexity relating to the colony’s size and shape that are related to their ecological function. We explored the potential ecological applications of some of these metrics using linear models, comparing aspects of complexity among colonies of different size and morphotaxa (combined information on morphology and taxa). Our results showed that a metric as simple as colony diameter explained 95% of the variation in shelter provisioning capability when paired with information on colony morphotaxa. Further, the habitat provisioning of colonies of comparable size was similar among the six of the seven morphotaxa examined. During the current period of rapid uptake of photogrammetry among ecologists, the results of our study provide a basis to use data derived from 3D models to further explore the nuances of the relationship between structure and function of corals at the colony scale in a replicable and standardised way.