This study found that various coral traits disproportionately contribute to three-dimensional structural complexity on coral reefs in Guam. Understanding the coral traits that predict complexity in reefs can help forecast how reef structure and associated ecological functions will be impacted by changes in coral community composition. Coral colony size largely influenced the structural metrics of Surface Complexity, Vector Ruggedness Measure (VRM), and Slope (Table 2, Fig. 4). Increases in the number of larger coral colonies yielded higher structural complexity values, whereas increases in the number of small colonies (e.g., 0–5 cm in size) resulted in lower structural complexity. Similarly, a comprehensive study among Seychelles, Maldives, Chagos, and the Great Barrier Reef showed that maximum colony size was a positive predictor of reef structural complexity12. Notably, our study produced new insight that coral colony size needs to be greater than 10 cm for increases in their abundance to contribute to reef structural complexity positively in Guam. This has an important implication in conservation as a large number of coral recruits are unlikely to contribute to structural complexity in the same way as a few large colonies do, despite having similar total live coral cover. This result also highlights the inability of 2D ‘live coral cover’ to capture nuances in how live coral positively influences structural habitat complexity, and thus caution should be used when solely relying on this metric to assess reef condition. Aside from larger colonies providing greater structural complexity, they are generally more resistant to diseases, dislodgement, and mortality29,44, highlighting the importance of large corals not only for structural complexity but also for reef resilience and function44,46,47. As a result of environmental change and anthropogenic stress, coral reef ecosystems in the Great Barrier Reef, for example, are already showing a shift towards the dominance of smaller coral colonies48. If such an event occurs in Guam or elsewhere, it could result in a significant loss of structural complexity and thus hinder associated ecosystem services.
Coral morphology is a key driver of many biological and ecological processes in reef ecosystems12,46,47,49. In the present study, Laminar-Columnar was the coral morphology with the highest contribution to structural complexity. Laminar columnar is a combination of Laminar and Columnar morphologies, exhibiting multiple layers of plate-like structures with columns protruding from the top, and it is known to be the primary morphology exhibited by the species Porites rus50,51, which is one of the most dominant coral species in Guam51,52,53,54. The unique and complex habitat structure created by this morphology likely explains its contributions to high structural complexity measured by Slope and VRM. On the other hand, increases in the number of colonies with either Mounding or Encrusting-Flat morphology showed negative effects on structural complexity. This finding contradicts a previous study in the Northwestern Hawaiian Islands (NWHI) where positive correlations between the abundances of these morphologies and structural metrics were found55. A possible explanation for this is that in contrast to this study that accounted for twelve morphological types, the study in the NWHI only had four, branching, tabulate, mounding, and encrusting morphologies, as morphologies such as columnar and laminar columnar are rare in the NWHI. This could result in differences in the overall ranges of individual complexity metrics between the two studies, potentially affecting the results of analyses. Another potential explanation is the dominance of small colonies and overall lack of large Fig. 5. Size class distribution for (A) Encrusting Flat, (B) Mounding, (C) Laminar Columnar, (D) Columnar, (E) Laminar, and (F) Branching. colonies in the mounding and encrusting-flat morphologies in the present study (Fig. 5). Further investigations are required among various regions and reef types to unravel the effects of coral morphology and size on reef complexity dynamics.
Profile and Planform Curvature exhibited a positive relationship with the abundance of Branching corals (Table 2, Fig. 4). These findings are also contrasted by another previous study in the NHWI where Branching corals were positively correlated with Surface Complexity and Slope13. In addition to the differences in the morphological types as discussed above, such disparate results can also be explained by methodological differences pertaining to study plot sizes. The plot size in the study in the NWHI was 90m2, whereas the size in the present study was 4m2. Averaging complexity metrics over 90m2 likely results in accounting for more components on the reefs compared to averaging over 4m2 simply due to the larger spatial coverage. As curvature values can sometimes take extreme values56, plot sizes can affect curvature metrics more than other complexity metrics (e.g., slope and VRM) when they are averaged for a plot. In another study, Darling et al (2017) found that proportions of Branching corals on reefs were negatively correlated with reef complexity scores based on visual estimation12. As coral survey methodologies can vary among different studies (e.g., classifications of coral morphology, plot sizes, and calculation/estimation of complexity metrics), generalizing findings from different studies may be difficult, highlighting the importance of being explicit about the survey methodologies when reporting the associations between coral and structural complexity The variability among these findings highlights the importance of conducting a thorough analysis of coral community composition and complexity for any site or region, as there is unlikely to be a singular metric or assumption that can be applied to any location to understand how live corals influence habitat structure.
Aside from being positively correlated with Branching morphology, Profile and Planform curvature also showed a negative relationship with Encrusting-Columnar morphology (Table 1). Curvature is a non-monotonic measure, with “zero” meaning a flat surface, whereas values less and greater than zero indicating increases in complexity; negative values indicate a convex surface and positive values indicate concave surface57. Our results thus suggest that an Encrusting-Columnar morphology in the present study had more convex surfaces, and Branching morphology had more concave surfaces. This is supported by visualization of Profile Curvature for the two morphology types where convex surfaces around Branching morphology yielded extreme positive values (Fig 6). The ability of curvature metrics to separate these two seemingly similar morphology types is a key finding that has not been previously documented due to the lack of Encrusting-Columnar morphology in the previous studies at other geographic locations utilizing photogrammetry techniques13, 55.
Despite the high coral biodiversity of coral reefs surrounding the Island of Guam, coral abundance, genus diversity and richness had significant inverse relationships with Surface Complexity, VRM, and Slope (Table 1). Profile and Planform Curvature, on the other hand, did show a weak yet positive relationship with coral richness and diversity (Table 1). While diversity is often considered a fundamental feature of ecosystem function49, these results suggest that the dominance of one or few genera may be a stronger predictor of structural complexity, which is known to influence ecosystem functionality in a positive manner2. A study done in the Caribbean, for example, found that the dominance of one or two coral genera showed the highest contribution to reef structural complexity compared to more diverse sites5. In our study, however, dominance of a particular genera did not seem to explain changes in complexity. For instance, while the genus Porites accounted for 49.6% of the coral colonies surveyed for this study, Porites alone did not exhibit a positive relationship with any complexity metric. As Porites corals exhibited nine out of the twelve possible morphologies, it is possible that the variability in Porites morphologies muted the effect of this genus on structural complexity when analyzed collectively. It is also important to note that coral abundance in the present study was measured by colony density per 4-m2 plot. Having more colonies, and possibly more species, in each plot can lead to individual colonies in the plot being smaller in size, which has negative effects on structural complexity (Table 2).
Only two of the 32 surveyed genera showed a positive relationship with a complexity metric: Pocillopora with Profile and Planform Curvature and Millepora with Surface Complexity and Slope (Table 1, Fig. 4). In the case of Pocillopora, this relationship could be linked to morphological traits, as Pocillopora's dominant morphology is Branching, which also exhibited a positive relationship with Profile and Planform Curvature (Table 1, Fig. 4). This is consistent with a recent study in Oahu, Hawaii where the associations between structural complexity and specific coral species were found to be closely tied to species morphology58. Millepora, on the other hand, exhibited some variability in both sizes and morphologies, but they lacked small colonies (i.e., 0–5 cm in size, Fig. 7). This might have contributed to the observed positive associations with Surface Complexity and Slope, indicating the importance of considering colony size when modeling habitat complexity on coral reefs. Collectively, our results overall indicate that coral genera alone are not a strong predictor of habitat complexity in Guam’s reefs.
These findings also highlight the increased importance of accounting for coral morphological traits as it suggests that taxonomy-based metrics alone does not provide sufficient insights. Future studies should focus on examining interactions between specific morphology types and their sizes in the effects on reef structural complexity and consider species data in relation to morphological traits.
Environmental parameters also influenced structural complexity in Guam reefs, as we observed sites with higher wave exposure to have significantly lower structural complexity values (Fig. 3B). This could be explained by the wave-forcing, resulting in shifts in the abundance and size of various morphologies. Hard corals exhibit strong morphological plasticity in response to hydrodynamic force15,45,36. This occurs because certain morphologies, such as those that grow vertically and have a smaller attachment, have a higher risk of dislodgement and mechanical damage in higher wave energy environments44. For our study sites in Guam, the sites with the highest wave energy tended to be dominated by smaller coral colonies with a mounding morphology, whereas the sites with less wave energy were dominated by larger colonies with columnar and laminar morphologies (Fig S4 & S5). Lastly, as expected, structural complexity was much lower at pavement-type reefs than aggregate and rock & boulder reefs (Fig. 3A), highlighting the role that underlying substrate can play in the reef’s architectural complexity56.
Coastal ecosystems are presently among the most heavily impacted ecosystems on earth59. Natural and anthropogenic disturbances are rapidly reconfiguring coral assemblage composition, urging the need to identify key physical drivers of habitat complexity given their direct link to ecosystem functionality2. Using a trait-based approach, we examined how physical coral traits influence 3D photogrammetry-derived structural complexity metrics and compared the results with those based on either taxonomic classifications or abundance/diversity metrics. Our main conclusion is that the traits of coral colony size and morphology were the best predictors of habitat complexity in Guam’s reefs and should thus be included in coral reef monitoring programs. This study offers important insights and foundation for future studies assessing the impact of changing reef habitats on reef-associated organisms under climate change.