Morphological Diversity and Its Relationships with Coral Growth and Fate

Abstract

Corals are modular organisms that can grow, shrink, divide, and fuse, making age an unreliable indicator of demographic condition. Coral population models have therefore relied mainly on colony size. Yet colonies of similar size can differ greatly in shape, which may retain information about environmental interactions and past disturbance. We used underwater structure-from-motion photogrammetry to generate high-resolution orthomosaics and followed 796 records of Pocillopora acuta colonies in Kenting National Park, Taiwan, over two years. Colony area, circularity, and perimeter-to-area ratio were linked to survival, growth, shrinkage, fission, and fusion. Models combining area and circularity best explained most demographic processes. Large, circular colonies survived better; large, irregular colonies were more likely to undergo fission; and small, circular colonies showed faster growth rate. These findings demonstrate that shape is not redundant with size but represents an important additional dimension for predicting coral population dynamics.

Keywords: coral demography; morphology; modular organisms; population dynamics; structure-from-motion

Methodology and Implementation Steps

Three surveys were conducted from October 2020 to December 2022 across three plots in Kenting National Park, Taiwan (84.39 m² in total; Fig. 2). Divers collected overlapping images with a dual-camera setup. Structure-from-motion (SfM) photogrammetry was used to reconstruct high-resolution three-dimensional reef models and geometrically corrected orthomosaics. Acrylic plates of known area were used to validate area measurements, with relative errors below ±2%. Pocillopora acuta colonies were delineated in ArcGIS, assigned unique IDs, and matched among survey dates. Each transition was classified as mortality, growth, shrinkage, fission, or fusion (Fig. 3). We quantified planar area as size, circularity as a size-independent measure of shape, and perimeter-to-area ratio as an integrated size-and-shape metric. Proportional area change was calculated from final area divided by initial area, with the summed area used for fission and fusion events. Generalized linear models tested single-trait models, additive size-plus-circularity models, and their interaction. Model performance was evaluated with Akaike information criterion and McFadden pseudo-R². This workflow integrates underwater imaging, 3D reconstruction, GIS-based colony tracking, and statistical modeling into a repeatable demographic monitoring approach.

Innovation and Cross-Disciplinary Collaboration

The study establishes image-derived morphological indicators that improve predictions of coral survival, fission, and proportional area change (Fig. 4), filling the missing dimension of shape in modular-organism demography. The framework can support long-term reef monitoring, assessment of nonlethal damage after disturbances, and selection of colonies for restoration. Combined with AI, remote sensing, and population models, it can be extended into automated risk maps and decision tools for precision coral restoration.

Expected Results and Contributions

The study establishes image-derived morphological indicators that improve predictions of coral survival, fission, and proportional area change (Fig. 4), filling the missing dimension of shape in modular-organism demography. The framework can support long-term reef monitoring, assessment of nonlethal damage after disturbances, and selection of colonies for restoration. Combined with AI, remote sensing, and population models, it can be extended into automated risk maps and decision tools for precision coral restoration.