Using image analysis to study the succession of coral reef micro-communities

Coral reefs are extremely important ecosystems that support a wide range of species and provide many ecological services. Many young corals react to cues from reef micro-communities that may alter where they choose to grow based on the presence and abundance of recruitment inhibitors and facilitators. Much research has been conducted on coral reef succession at the macroscale, but little research has been conducted on succession at the microscale. This research seeks to answer the question: do benthic micro-communities change over time, in light vs dark? To study this, settlement tiles were placed at three replicate sites on the White Hole coral reef site in Roatán, Honduras and left for one year to allow micro-communities to develop. Monthly images were taken of the top and bottom surfaces of tiles between June 2021 and September 2021. Morphospecies on each tile surface were identified and outlined by hand in ImageJ. Differences in community composition were calculated for each tile using Bray-Curtis distances and compared using permutational analysis of variance tests. There was no significant difference in community composition over time, but there were significantly different communities on the tops and bottoms of the tiles, likely driven by changes in the relative abundance of light. However, the data used in this project is a portion of the total data collected. Once more tile images are included in the analysis, it is expected that change in community composition over time will be significant. The implications of these results could help improve future restoration methods.

Title Slide:
Hello, my name is Teagan Corpening, and I will be presenting my research, “Using image analysis to study the succession of coral reef micro-communities”. This work was performed in the Salerno Lab.

Introduction:
Coral reefs are important ecosystems because of the many ecological services they provide, including coastal protection and providing habitat for a wide range of species. Protecting coral reef ecosystems is integral to protecting the biodiversity of marine ecosystems and preventing shoreline degradation. Understanding coral reefs at the microscale is important because of the connection between micro-communities of algae and other invertebrates that live on the sea floor and large-scale coral reef health. Coral recruitment and cover are heavily influenced by micro-communities. By better understanding the influences on coral recruitment, restoration methods can be improved.
I am working with Jordan Sims, who is a PhD student working on a larger settlement project in Roatan, Honduras.
My project seeks to better understand coral reef succession at the micro-scale by studying how micro-communities change through time and how environmental conditions, primarily light, drive community succession. My project seeks to answer the question: Do benthic micro-communities change over time, in light vs dark? This was done by observing community succession focusing on six different functional groups (CCA, fleshy macroalgae, filamentous algae, turf algae, other invertebrates, and biofilm). The presence of different functional groups is important to differentiate between because some groups, like turf algae, are coral recruitment inhibitors while other groups, like CCA, are coral recruitment facilitators.

Study Site:
This study was carried out in Roatan, an island off the coast of Honduras in Central Latin America. My study site is a reef on the northwestern coast of Roatan called White Hole.

Methods:
[1] Ceramic tiles were placed on three platforms across the White Hole reef site in Roatan Honduras. For this project, tile images at each platform were taken once per month between June and September of 2021. Images were taken of both the top and the bottom surfaces of each tile. The primary difference between the top and bottom surfaces of the tiles was the amount of light available to the communities. The tops of the tiles received more light than the bottoms of the tiles.
Here’s an image of one of our tile platforms once it has been secured to the reef. Settlement tiles were secured to the platform using zip ties.
[2] Here are representative unedited tile images after the tiles were removed from one of the platforms. On the left, we have the bottom of the tile and on the right, we have the top of the same tile.
[3] I cropped the unedited images to remove the unnecessary pieces including the collection bag and the sides of the tiles.
[4] These are the labelled tile images. I performed labeling by hand in ImageJ. I identified 85 different morphospecies and traced the boundaries of each morphospecies. The morphospecies were then combined into 6 different functional groups. After labeling, I used ImageJ to calculate the relative abundance of each functional group in each image.
functional groups:
These are examples of the six different functional groups that I combined the morphospecies into. These groups include Crustose Coraline Algae, filamentous algae, fleshy macroalgae, turf algae, biofilm, and other invertebrates.

Results:
[1] Differences in community composition over time and between tile surfaces were calculated using Bray-Curtis distances. To compare the micro-community composition on the tops and bottoms of the tiles and the micro-community composition through time, PERMANOVA tests with 999 permutations were performed. There was no significant change in community composition over time, but there were significantly different communities on the tops and bottoms of the tiles.
[2] On this graph, we can better visualize the drivers of the differences in functional group composition between tops and bottoms of the tiles. Bottom surfaces had a significantly higher percent cover of biofilm than top surfaces. We can also see a higher percent cover of other invertebrates on bottom surfaces. However, top surfaces had higher percent cover of turf algae and fleshy macroalgae.

Discussion:
The tops and the bottoms of the tiles were determined to have different community composition. This is most likely because of the differences in light. These results are supported by other studies that have also seen changes in benthic cover and functional group productivity influenced by the abundance of light (Vooren, 1981). The greater amount of light available to the communities on top surfaces is likely what is driving the observed increase in algal growth.
This study only used a portion of the tile images and data that were collected in the larger settlement project. There are four additional timepoints from this reef site and another full set of images from a replicate reef site. For the next steps of this project, I’ll continue to analyze more tile images so that more timepoints and replicates can be included in statistical analysis in the future. We expect to see a significant pattern of change emerge as more tile images are included in the project’s scope.
Acknowledgements:
Thank you to Dr. Salerno for being my mentor, to Jordan for helping me with the analyses, and to Jennifer Keck for taking the tile photos.