Manipulating structural complexity to bolster restoration efforts on Hawaiian coral reefs

Methods

Summary

RESEARCH METHODS/APPROACH:

Module Creation 

Alongside Co-I Ehrenberg, two undergraduate students will be trained and tasked with creating silicone molds for 17 designs to capture the full range of rugosity and fractal dimension levels found on coral reefs in Hawaii and casting concrete modules (Fig. 2). Modules will have base dimensions of approximately 22 x 22 cm, with a ridge footprint of 15 x 15 cm, and heights ranging from 0 to 22 cm depending on the design. Designs may differ from those in Fig. 2B-D to better reflect the pyramid modules currently used by restoration initiatives in Hawaii, and our project partners: Kuleana Coral Restoration and the State of Hawaiʻi’s Division of Aquatic Resources Hawaiian Coral Restoration Nursery. All new designs will be approved by the project partners before proceeding with production. 

The first 6 months of the project will be focused on design, mold creation, and concrete casting. The Geometric Ecology Lab (GEL) at HIMB has experience with 3D design using the software Blender and 3D printing to create plastic models and negatives to create silicone molds. In March 2023, the GEL received a large format 3D printer for mold creation, which will be used to print four sets of mold negatives for each design, speeding up casting processes. Commercial grade, crack resistant, Quikrete will be hand mixed and poured into silicone molds. After air bubbles have been released from the wet concrete mix, the modules will be placed in a cool and dry environment to cure for 48 hours. Once hardened, modules will be removed and stacked on drying racks for further curing and storage. Following casting and curing processes, modules will be laser scanned to calculate the rugosity and fractal dimension using the R package habtools (developed by GEL). We anticipate investigating branching (BR), massive (MA), and encrusting (EN) growth forms that cover high, medium, and low structural complexities with four replicate modules to be made for each design for both in-situ and ex-situ testing, resulting in a total of 408 modules.

Colony Selection and Fragmentation 

In collaboration with Kuleana Coral Restoration, coral colonies in Kāneʻohe Bay and West Oʻahu will be located and recorded via handheld Garmin GPS unit, photographed, and tagged (Fig. 3). Maps will be created in QGIS to plan field collection of coral fragments (Fig. 3). Selected branching, massive, and encrusting growth forms of Montipora capitata, Montipora patula, Porites compressa, Porites lobata, Pocillopora meandrina, Leptastrea purpurea, and Pavona varians, taken from Kāneʻohe Bay and West Oʻahu, will be micro-fragmented. Fragments of the same genotype will be kept together to ensure that all fragments attached to a module are genetically identical, following methods used by the Hawaiian Coral Restoration Nursery. Select tips of branching colonies, with ~1.5 cm diameter and ~1.75 cm length, will be separated from the donor colony using surgical bone cutting shears. Using a hammer and chisel, massive and encrusting colonies will be carefully removed from the benthos and taken ashore to be cut into ~2.25 cm^2 fragments with a coral saw. Throughout the collection, fragmentation, and adhesion to module processes, colonies and fragments will be out of the water for no more than 10 minutes at a time to reduce stress on the live coral tissue.

Ex-situ:

One set of 204 modules will be randomly placed in 34 tanks at HIMB, plumbed to open circulation systems, with each tank housing six modules. Ehrenberg and two undergraduate students will attach genetically identical fragments to modules at different positions along module ridge slopes using water activated cyanoacrylate glue (Fig. 4). On all slope surfaces, one fragment will be glued at the center of the ridge with two fragments placed ~3.25 cm from the center (Fig. 4).  Distance down the slope will be determined by the ridge height, for example, on 4 cm tall modules the fragments will be adhered 1.5 cm down the slope from ridge tips, while fragments on the 6 cm tall module will be placed at 2.5 cm. Flat control design modules will have 12 fragments, single ridge modules will require 6 fragments, two ridge modules will host 12 fragments, three ridge modules will have 18 fragments, and there will be 24 fragments on the four ridge modules. Post-attachment, modules will be visually monitored for a one week acclimation period, at which time if fragments become detached or experience mortality due to fragmentation, they will be replaced.

Once acclimation of the modules has been completed, all modules will be laser scanned, then photographed using photogrammetry techniques. On a rotating pedestal with reference markers, the module will be placed and imaged with a Canon EOS Rebel SL3 camera on a tripod. Photogrammetry for each module will take 5 minutes, with modules being returned a designated saltwater tank post imaging. There will be a one week period between photogrammetry and scanning to reduce stress on coral fragments.

Modules placed on a rotating pedestal with reference stickers will be scanned using a Creaform 3D HandyScan laser scanner. As the pedestal rotates, all surfaces of the module will be captured in high resolution, with scans saved for later processing. Scanning will take approximately 15 minutes, including a 5 minute break between scans to allow fragments to recover in a saltwater tank. The undergraduate students will assist with initial scanning and photogrammetry, monitoring, and regularly scheduled tank maintenance throughout the span of the study. Modules will be monitored monthly for the first three months, then every two months for the remainder of the one-year study period to record mortality and changes in complexity through visual observations and photogrammetry techniques. At the end of the study, final laser scans and photogrammetry will be captured and modules will be deployed to an appropriate reef area for further monitoring (if permitting allows).

In-situ: 

Following the ex-situ design, the same number of modules with attached coral fragments will be deployed at a study site in West Oʻahu, near the Ko Olina Resort area. To accommodate all 204 modules, the investigative team will consult with Kuleana Coral Restoration to select a 30 m^2 area; we will allow 6 months post site selection for permitting. Due to limited tank space at HIMB and logistical issues with transporting corals long distances, the one week acclimation period will not occur for the in-situ portion of the study; however, laser scans and photogrammetry for each module will be captured at the start and end of the study. 

 Before module deployment, a baseline survey will be conducted, consisting of a photogrammetry and Stationary Point Count (SPC) survey to record initial reef complexity and fish/invertebrate species. A modified version of a commonly used spiral photogrammetry survey methodology will be used in this study. Before imaging the study area, two surveyors will lay out four weighted floats at each corner, one reference marker at the center, and three scale bars in

a triangular formation within the plot (Fig. 5). Depth measurements will be taken at the reference marker and all scale bars. One surveyor, positioned at the center of the study area will hold a spool of line attached to a Canon EOS Rebel SL3 camera in a Ikelite underwater housing. The second surveyor will continuously image the benthos, while swimming in a spiral pattern out from the center of the plot (Fig. 5). Once the outer boundary of the plot is reached, the second surveyor will turn around and swim the spiral back in the opposite direction. When the center of the plot is reached, the center surveyor will detach the spool from the camera and move to allow the second surveyor to image the center. The camera will be set to the following settings: continuous shooting mode, 18 mm focal length, 1/180 shutter speed, F8 aperture, and 800 ISO. 

The SPC survey will follow a modified version of the standard NOAA methods, with the use of only one 7.5 m diameter cylinder to encompass the study area (Fig. 6). Two divers or snorkels, back to back, will be situated in the center of the cylinder, with each diver recording only their half of the cylinder (Fig. 6). Unlike the NOAA SPC, the first 10 minutes of the SPC will consist of recording the abundance and species present in the cylinder. In the last 10 minutes, surveyors will follow NOAA SPC protocols for measuring and recording fish in the survey area. Next, modules will be placed approximately 10 cm apart, at random, evenly throughout the study site (Fig. 6). Post deployment, surveyors will wait for an appropriate surface interval to allow reef organisms to acclimate to the modules, before returning to conduct another photogrammetry and SPC survey. Kuleana Coral Restoration will assist with module deployment at the study site and provide support in the form of divers and vessel access. Ehrenberg will train all surveyors on the appropriate methods before site surveys occur. Sites will be resurveyed every month for the first three months, then every two months for the remaining duration of the experiment and consist of the following surveys: SPC, photogrammetry, and visual assessments of fragment percent mortality. Continued monitoring of the modules will occur past the end of the study period to investigate the long-term influences of module complexity on coral fragment success and associated reef organisms.

Challenges

Primarily, obtaining permits to deploy structures in Hawaiian waters, from the State of Hawaiʻi's Division of Aquatic Resources, may delay the timeline of this project. To overcome this challenge, I plan to use previous accepted permit applications submitted by our lab (Geometric Ecology Lab) or attempt to alter existing permits. 

Pre Analysis Plan

Image and Data Processing 

Module and study site images will be stitched together in Agisoft Metashape to generate orthomosaics and digital elevation models (DEMs). For the in-situ study site, baseline and post out-plant DEMs will be compared to determine the overall change in structural complexity following NOAA standard operating procedures. Using QGIS and modified methods outlined by NOAA, the perimeter of all fragments on each module will be traced and tracked over time to measure growth, fusion, and mortality rates. The fragment boundaries traced from starting and ending laser scans will be compared to orthomosaics to determine accuracy of data obtained from photogrammetry surveys. In the remaining six months of the project timeline, all data will be analyzed to compare the starting and ending overall reef structural complexity, biodiversity (in-situ study), module complexity, fragment survivorship, growth rates, and fusion rates. Statistical analyses will be conducted and figures will be created in R Studio. Publications will be drafted and findings will be presented to local coral restorations groups, such as Kuleana Coral Restoration, to develop collaborative relationships for future restoration efforts, research opportunities, and outreach events.