{"paper_id":"08dec878-e264-4b05-9adb-9f676ef49a8f","body_text":"1\n1 \n2\n3\n4 Habitat suitability modeling growth and cover trends of staghorn coral (Acropora cervicornis) \n5 outplants in the lower Florida Keys\n6\n7 Glenna Dyson1,5*, Erich Bartels2, Easton R. White3, Ian R. Combs2, Kristen Mello-Rafter1, \n8 Thomas C. Lippmann4,5, Jennifer A. Dijkstra1*\n9\n10\n11\n12 1 Center for Coastal and Ocean Mapping/UNH-NOAA Joint Hydrographic Center, University of \n13 New Hampshire, Durham, New Hampshire, United States of America, 03824\n14 2 Elizabeth Moore International Center for Coral Reef Research & Restoration, Mote Marine \n15 Laboratory, Summerland Key, Florida, United States of America, 33042\n16 3 Department of Biological Sciences, University of New Hampshire, Durham, New Hampshire, \n17 United States of America, 03824\n18 4 Center for Ocean Engineering, University of New Hampshire, Durham, New Hampshire, \n19 United States of America, 03824\n20 5 Department of Earth Science, University of New Hampshire, Durham, New Hampshire USA, \n21 03824\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n2\n22\n23 * Corresponding authors\n24 E-mail: glenna.dyson-roberts@unh.edu (GD); jennifer.dijkstra@unh.edu (JD)\n25\n26\n27 ABSTRACT\n28 The decline of important reef building corals has motivated the development of habitat \n29 suitability models used to identify optimal locations for coral restoration. In the Florida Keys \n30 habitat suitability models incorporate coarse spatial data sampled over large areas, resulting in \n31 recommended outplant sites at distant locations, making it logistically difficult and expensive to \n32 access and regularly monitor. Restoration efforts to date show that outplanting success can vary \n33 widely within a limited space, necessitating improved predictive abilities of coral outplant \n34 success at high spatial resolutions within a restoration site. With the advent of Structure-from-\n35 Motion image reconstruction, fine-scale, site specific, digital terrain models can be created to \n36 support habitat suitability model development. In this study, generalized linear mixed models \n37 used extracted seafloor terrain attributes and environmental variables to identify within site \n38 locations of high Acropora cervicornis growth and healthy coral cover of long-term outplants. \n39 Percent healthy coral cover significantly decreased after two years of outplantation. The \n40 submodel of corals exclusively less than two years old was unable to identify environmental \n41 conditions associated with higher healthy cover. For all corals, outplant recommendations for \n42 higher healthy cover are in deeper waters, away from the coast, in less rough terrain, and closer \n43 to the reef edge. Model results for growth support these recommended outplant sites, in addition \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n3\n44 to concave locations near high slope relief. Finally, our results also indicate that marine heat \n45 waves, but especially marine cold waves negatively correspond with coral growth, and high wind \n46 events positively correspond with coral growth. These model results provide a basis for further \n47 endeavors in modeling endangered organismal success, which are vulnerable to minute \n48 differences in local environmental conditions. \n49 Introduction\n50 In the past half century, corals around the world have been decimated due to mounting \n51 anthropogenic pressures including disease, overfishing, river run off, high intensity thermal and \n52 wind events (1–4). These threats have resulted in increased efforts to promote resilient reefs, \n53 capable of withstanding and recovering from disturbance through the creation of marine \n54 protected areas, genetic engineering of corals towards elevated heat resistance, and outplanting \n55 corals. Stony reef-building corals have been the primary target of outplantation. One such genus, \n56 Acropora, has been targeted due to forming much of the essential reef structure throughout the \n57 Florida Reef Tract (5) and its outplantation facilitates variation in benthic coral composition and \n58 fish species (6). Despite being a good candidate for outplantation, due to asexual propagation and \n59 a fast growth rate (7), the genus has shown to be especially vulnerable to environmental stress \n60 (8,9) and disease (10,11). To support efforts towards enhancing reef resiliency, habitat suitability \n61 models (HSMs) targeting Acropora restoration have been developed, including models of wild \n62 Acropora population occurrence (12), recruitment (13), and global responses to ocean warming \n63 and acidification (14).\n64 These HSMs have the potential to identify optimal sites for Acropora outplantation, \n65 however models for outplant success have been challenging for coral reef managers and \n66 outplanters to implement. Current models often use coarsely resolved environmental data (9), \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n4\n67 many collected through remote sensing efforts, resulting in recommended sites that are \n68 financially and logistically difficult to access for outplantation and recurrent monitoring. \n69 Predicted suitable habitat can be hundreds of square kilometers, although many restoration \n70 activities occur at much finer spatial scales (10s of meters) where coral survival throughout the \n71 same site is highly variable (15). Additionally, current models have primarily used wild \n72 Acropora populations. Coral outplant’s growth and survivorship perform differently than wild \n73 and transplanted coral populations (16,17), making these models imprecise in recommendations \n74 for outplants. Lastly, when coral outplants are surveyed, long-term monitoring is often not \n75 realized nor influential in informing outplanting techniques, and thus sustained recovery of \n76 Acropora outplants is not clearly established (18,19). Consequently, there is a knowledge gap in \n77 HSMs using fine-scale, within site, terrain attributes for long-term coral outplant locations.\n78 This study aims to develop HSMs utilizing fine-scale terrain attributes which may \n79 contribute to within site variations in growth (total and healthy) and healthy cover of a \n80 commonly used stony coral for outplantation, Acropora cervicornis. A. cervicornis has been one \n81 of the primary targeted Acropora species for restoration due to its ecological importance as a reef \n82 builder and its physiological characteristics as a long-lived, fast growing species (20–22). HSMs \n83 developed thus far have shown improved wild A. cervicornis populations in waters with \n84 moderate sea surface temperatures and limited temperature ranges, moderate turbidity to mitigate \n85 temperature fluctuations and UV radiation (9), and higher water flow to deliver food and oxygen \n86 (9,23). From these results, restoration work has called for outplanting deeper, or where there is \n87 higher water flow to decrease the negative impact of high intensity thermal events (24). These \n88 HSMs show A. cervicornis has limited suitable habitats within the Florida Reef Tract, \n89 constrained to the fore and back reef of the lower and upper Florida Keys, the Dry Tortugas, and \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n5\n90 nearshore Broward-Miami reefs (15,23). This study’s use of fine-scale environmental data aims \n91 to delineate optimal coral outplant sites within such regions by modeling outplant growth and \n92 percent healthy cover within the forereef of the lower Florida Reef Tract. Growth (separated into \n93 total and healthy) indicates where at fine-scales there is optimal mass transport (exchange of \n94 nutrients and dissolved gasses) and minimal stress (25,26). Total growth indicates where the \n95 most growth is occurring, whereas healthy growth indicates where the most growth is occurring \n96 despite infection from disease. Percent healthy cover indicates where corals are least exposed to \n97 disease, are more resilient to disease, or where corals are best able to recover from infection (27). \n98 In this study, we use fine-scale three-dimensional terrain attributes derived from \n99 Structure-from-Motion (SfM) photogrammetry, environmental data, and high intensity thermal \n100 and wind events as predictor variables for modeling A. cervicornis growth and percent healthy \n101 cover. The use of SfM is unique in these HSMs for its capacity to develop three-dimensional \n102 terrain models which provide insight in terrain structure and complexity (28–30); the scale of \n103 these models is also several orders of magnitude finer (mm) than spatial data often used in HSMs \n104 (0.25km to 1 km). This study aims to address the utility of SfM and fine-scale data to predict \n105 site-specific optimal coral outplant locations. Model results contribute towards further \n106 understanding of A. cervicornis restoration and the applicability of fine-scale models for other \n107 endangered species. \n108\n109 Methods\n110 Outplanting\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n6\n111 Percent cover and growth of A. cervicornis were determined at eight outplant sites in the \n112 lower Florida Keys. Corals were outplanted by Mote Marine Laboratory at Sand Key, Eastern \n113 Dry Rock, M32, ICC1, and ADAC1 (Fig 1). All sites reside within the Florida Keys National \n114 Marine Sanctuary. Additionally, Eastern Dry Rock and Sand Key are sanctuary preservation \n115 areas. \n116 Coral outplant fragments ranged from 250 to 1000 per outplant site, and the corals’ initial \n117 sizes ranged from 3cm to 50cm. Except for the singular 3cm coral plugs at ADAC1, all other \n118 sites had outplanted clusters of five coral fragments. Corals were affixed to the seafloor using \n119 either nails and cable ties; nails, cable ties, and epoxy; or affixed to plugs with epoxy. To \n120 enhance genetic diversity, a variety of genotypes were outplanted. \n121\n122 Fig 1. Map of coral outplant sites. Eight survey sites within the lower Florida Reef Tract to \n123 west of Key West (S4, S6, T3, T16, M32-2, M32-3, ICC1, and ADAC1). Shaded areas refer to \n124 sanctuary preservation areas.\n125\n126 Image Collection\n127 Eight coral outplant sites were surveyed in June of 2022, and one of the eight sites \n128 (ICC1) was previously surveyed the year before in July of 2021, resulting in nine outplant \n129 surveys (Table 1). Coral outplant ages ranged from 129 to 1,788 days. All sites, except ADAC1, \n130 are on the forereef of the Florida Reef Tract. Images were collected in 10 x 30m plots at all sites. \n131 Coded targets (which also functioned as scale bars to ensure accurately sized orthomosaics) were \n132 systematically placed throughout each site in a grid formation to support co-registration. Images \n133 were taken with two Canon 70D DSLR cameras housed in an Aquatica A70D. Planar images of \n134 the seafloor were collected in a lawn-mower survey pattern to facilitate co-registration and limit \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n7\n135 distortion, resulting in ~ 80% image overlap. Depth measurements taken from a dive computer \n136 accounted for the four corners of each 10 x 30m site.\n137 Table 1. Summary of coral outplant locations and time spent outplanted.\nOutplant ID Location Outplant Date Age of outplants when surveyed\nT3 Eastern Dry Rock 7/8/2017 1788 days\nS4 Sand Key 9/11/2018 1386 days\nS6 Sand Key 3/25/2019 1191 days\nM32-2 Marker 32 8/9/2019 1050 days\nM32-3 Marker 32 5/6/2020 779 days\nICC1(21)\nICC1(22)\nSite C 12/10/2020 237 days \n564 days\nADAC1 Acer 10/5/2021 261 days\nT16 Eastern Dry Rock 2/11/2022 129 days\n138\n139 Three-dimension Terrain Model\n140 Orthomosaics and digital terrain models (DTMs) were constructed using Agisoft \n141 Metashape Professional V1.2.6 software. Orthomosaics and DTMs followed similar protocol to \n142 other researchers and colleagues (31,32), with slight modifications. Permanent control points, \n143 galvanized nails hammered into the bedrock and epoxied at the base to the seafloor, were \n144 georeferenced using a GARMIN system on-board the vessel for image reconstruction. A. \n145 cervicornis cover was quantified from orthomosaics (Fig 2). \n146\n147 Fig 2. A 10 x 10m orthomosaic and a magnified section showing outplanted staghorn coral \n148 in site ICC1.\n149\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n8\n150 Coral Cover\n151 Feature class polygons of healthy coral cover and total coral cover were created. Healthy \n152 tissue cover was identified as orange/beige tissue, without any unhealthy cover (coral tissue \n153 affected by disease or covered with bacterial mats or macroalgae). Total coral cover included all \n154 skeletal tissue cover, both healthy and unhealthy. Percent healthy cover was quantified as the \n155 percentage of healthy coral cover from total coral cover. Coral cover was then classified by \n156 fragment, wild (non-outplanted), or outplanted corals. Total and healthy growth was quantified \n157 as the change in total skeletal cover and healthy tissue cover since outplantation, divided by time \n158 since outplantation. Only one site, ADAC1, had wild coral populations within the survey plot \n159 and were easily identifiable, as they were geographically distant from the outplants and much \n160 larger than the recently outplanted 3mm coral plugs. ADAC1 outplants were all singular plugs; \n161 other sites were composed of five fragment aggregates; older sites (S4, S6, T3) composed of five \n162 fragment aggregates fused together to form a mass of thickets. Average coral fragment cover at \n163 the time of outplantation was utilized to account for variations between initial size in calculating \n164 growth. \n165 Covariates \n166 Bathymetric and Environmental Parameters\n167 DTMs were imported into ArcGIS Pro and projected in UTM 17 zone projection with \n168 WGS84 datum. DTM resolutions were standardized (utilizing the resample tool in ArcGIS Pro) \n169 to 8mm and were georeferenced to align with the orthomosaics to ensure accurate sampling of \n170 terrain attributes around corals. To ascribe seafloor terrain attributes to coral cover, a 0.1m buffer \n171 around each coral polygon was created. Associating seafloor terrain and environmental attributes \n172 - calculated using the Spatial Analyst toolbox and Analysis toolbox in ArcGIS Pro - with total \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n9\n173 coral skeletal cover were extracted using ArcGIS Pro ModelBuilder for each coral outplant and \n174 included as covariates in the HSM (Table 2; Fig S1). \n175 Table 2. Summary of spatial statistics utilized to characterize terrain and environmental \n176 attributes.\nTerrain and \nEnvironmental \nAttribute\nUnit Definition ArcGIS tool Aggregate \nfunction\nDepth Meters Distance between water \nsurface and benthos.\nSurface \nInformation\nMax\nSlope Degree Benthic steepness. Surface \nParameter\nMean\nAspect \nNortherness \nand Easterness\nCosine(Degree)\nSine(Degree)\nDirection the terrain is facing. Surface \nParameter\nMean\nMean \nCurvature\n1/Meters Second derivative of slope. \nMeasures \nconvexity/denudation or \nconcavity/accumulation of \nterrain.\nSurface \nParameter\nMean\nRoughness Unitless Surface relief ratio measures a \nsurface’s unevenness.\nFocal \nStatistics and \nRaster \nCalculator\nN/A\nDistance from \nReef Edge\nMeters Apart from ADAC1, all coral \noutplant sites were along the \nedges of the reef, near drop-\noffs. The Euclidean distance of \nthe A. cervicornis from the reef \ndrop-off\nNear N/A\nDistance from \nCoast\nMeters The Euclidean distance of the \nA. cervicornis from the nearest \ncoast\nNear N/A\nDistance from \nother A. \ncervicornis\nMeters The Euclidean distance of the \nA. cervicornis from the nearest \nA. cervicornis\nNear N/A\n177\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n10\n178 High intensity events\n179 Water temperature and wind data spanning 30 years were cumulatively collected from \n180 three NOAA buoys: Sand Key (January 1991 – September 2005), Sombrero Key (February 1998 \n181 – February 2008), and Key West (February 2005-January 2021). Marine heat waves (MHW) and \n182 marine cold waves (MCW) are defined herein as five consecutive days or more where water \n183 temperatures exceed the 90th percentile (heat wave) or fall below the 10th percentile (cold wave), \n184 based on a 30 year historical base-line (33). This day-specific definition of marine heat and cold \n185 waves accounts for extreme temperatures within a given season, which can disrupt seasonal \n186 biological processes (like reproduction). Consequently, cold waves occurring in summer months \n187 are reflected and vice versa. The duration and intensity of these events were defined by the \n188 number of days within a heat/cold wave event and the number of degrees exceeding or falling \n189 below the 90th/10th percentile, respectively. To avoid collinearity with time, the annual average \n190 intensity of the event per duration of time throughout the corals lifetime was incorporated into \n191 the model. \n192 High wind events were defined as any day(s) when the average daily wind speed \n193 exceeded 12.78 knots, equivalent to the average sustained wind of a tropical storm (34). Like \n194 MHW and MCS, high wind events were similarly transformed to avoid collinearity with time.\n195 Statistical and Model Analysis\n196 Coral fragments were unable to characterize growth and were excluded from the analysis. \n197 All statistical analysis and models were run in R (RStudio Team, 2020). We used a single sample \n198 two-tailed Student’s t-test to determine if there was a significant difference between growth \n199 (both total and healthy) and the null hypothesis of zero growth. To evaluate differences in growth \n200 and percentage of healthy cover across sites and time, a one-way Analysis of Variance \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n11\n201 (ANOVA) was performed, with the null hypothesis of no differences among sites. If differences \n202 were observed a post-hoc Tukey’s HSD test (p < 0.05) was used to determine which sites \n203 differed from one another. \n204 Five generalized linear mixed models (GLMM), operating within a frequentist statistical \n205 framework, were implemented using the glmmTMB package in R (35,36). Three models were \n206 applied to the entire dataset of observed coral outplants: one for percent healthy coral cover, one \n207 for total growth, and one for healthy growth. Dramatically different trends in percent healthy \n208 cover were observed for corals less than and greater than 2 years old, resulting in submodels \n209 being run based on these age groups (young and old coral models). Model selection took the \n210 form of excluding highly correlating covariates. No covariates were excluded in the models for \n211 total growth, healthy growth, and percent healthy cover due to adequate sample size, but \n212 submodels with reduced sample size excluded highly correlating variables. All covariates, except \n213 time, were standardized to z-score values, resulting in model coefficients indicating change in \n214 growth (cm2/yr) for every change in standard deviation of the environmental and terrain \n215 covariate. The percentage composition of a coral outplant with healthy tissue was doubly bound \n216 between 0% and 100% using a beta-distributed GLMM using with a logit link function. Due to \n217 logit link limitations, healthy cover equal to 0% and 100% were transformed to 0.001 or 0.999, \n218 to fall within values of 02yro) corals. Lines depict the covariate estimates \n278 confidence intervals. Asterisks indicate the significance level of estimates.\n279\n280\n281 Healthy and Total Growth Model\n282 Outplanted corals showed total and healthy growth (t-test, tt = 22.72, pt < 0.001; t-test, th \n283 = 14.56, ph < 0.001) (Fig 6). Total and healthy growth rates varied across sites, but did not \n284 change over time (GLMM, zt = 0.97, pt = 0.33; GLMM, zh = -1.46, ph = 0.145) (Fig 6). Since the \n285 time corals were outplanted, total growth across sites averaged 26.03 cm2/yr (CI: 24.19-28.77 \n286 cm2/yr) with an average healthy growth of 16.19cm2/yr (CI: 14.12-18.53 cm2/yr). \n287\n288 Fig 6. Healthy and total coral growth for all sites. The red dotted line at 0cm2/yr depicts where \n289 the growth rate is positive (above) or negative (below). Total and healthy growth rates could be \n290 negative due to fragmentation, reducing the coral size to smaller than its initial size. Healthy \n291 growth could be negative due to disease. \n292\n293 The total growth model and healthy growth model converged according to the hessian \n294 statistic and had AIC values of 4928.68 and 4955.47, respectively. Several variables correlated \n295 with higher total and healthy coral growth. This included distance from shoreline (GLMM, zt = \n296 2.23, pt = 0.026; GLMM, zh = 2.45, ph = 0.014), increased slope (GLMM, zt = 2.27, pt = 0.023; \n297 GLMM, zh = 2.57, ph = 0.010), decreased roughness (GLMM, zt = -2.81, pt = 0.004; GLMM, zh = \n298 -4.22, ph < 0.001), decreased curvature (concavity) (GLMM, zt = -3.84, pt < 0.001; GLMM, zh = \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n15\n299 -3.29, ph = 0.001), increased high wind events (GLMM, zt = 2.03, pt = 0.042; GLMM, zh = 2.59, \n300 ph = 0.009), and decreased marine cold waves (GLMM, zt = -3.73, pt < 0.001; GLMM, zh = -\n301 3.16, ph = 0.0015). Increased marine heat waves correlated with lower total growth only \n302 (GLMM, z = -2.23, p = 0.025) (Fig 7).\n303 Fig 7. Healthy and total growth model covariate estimates for all corals. Lines depict the \n304 covariate estimates confidence intervals. Asterisks indicate the significance level of estimates.\n305 Discussion\n306 Percent healthy cover, unlike growth, decreased after two years from outplantation. A \n307 possible reason for the decline in healthy coral cover after two years of outplanting may be due \n308 to the onset of spawning around 2 years of age, which likely diverts energetic resources from \n309 fighting infection to reproduction and growth (17,38,39). In contrast to results for percent healthy \n310 cover for all corals, the model for corals outplanted for less than two years was unable to identify \n311 correlating terrain attributes that positively related to percent healthy cover, emphasizing the \n312 importance of using long-term coral outplants when developing suitability models. The best \n313 predictors for high healthy coral cover over two years of age were greater depth and less \n314 roughness, indicating these environmental conditions are especially important when informing \n315 outplant locations for A. cervicornis reefs.\n316 Overall, our fine-scale models suggest that depth, distance from the reef edge and the \n317 coast, and seafloor roughness correlated to percent healthy coral cover. For depth, healthy coral \n318 cover increased with depth. Scleractinian corals are sensitive to UV radiation and high \n319 temperatures, with greater temperatures and UV radiation occurring in the upper 10m (24,40,41). \n320 All our study sites were shallower than 10m. Our results revealed depth as a factor for percent \n321 cover of healthy coral, suggesting that even small differences in depth in the upper 10m of the \n322 water column may influence coral outplant success. Corals that were outplanted farther from the \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n16\n323 coastline were predicted to have greater healthy coral cover than those outplanted closer to the \n324 coastline, which supports results found by others (9). This is likely the result of the greater \n325 concentration of anthropogenic factors, such as greater thermal stress, sedimentation, and higher \n326 concentration of pollution, all of which are known to negatively impact coral health (1). Within \n327 the reefscape, corals outplanted closer to the reef edge had a greater percentage of healthy coral \n328 tissue. Reef edges are likely to have greater water flow which reduces bleaching, the build-up of \n329 sediment on the reef, and increases in mass transport (42,43), all factors that mitigate disease \n330 (12,44–46). \n331 Interestingly and in contrast to larger scale studies and our hypothesis, healthy coral \n332 cover positively associated with decreased roughness (i.e., smoother areas). There are a number \n333 of potential reasons for these unexpected results. First, increased surrounding roughness is \n334 indicative of higher coral abundance which may facilitate contact-spread of disease (27,28,34). \n335 Second, complex environmental terrain may support a higher number of corallivores due to the \n336 greater number of available shelters. Corallivores create lesions in the coral tissues, which \n337 provides greater opportunities for disease to colonize scleractinian tissue, including A. \n338 cervicornis (27,47–50). Lastly, high terrain roughness and surrounding corals with rough tissue \n339 results in higher mass transfer due to an increase in micro-turbulence (27,51,52), which may lead \n340 to an increase in nutrient and gas exchange and thus a greater capacity to uptake disease-causing \n341 microbes. \n342 Total and healthy coral growth was hypothesized to decrease over time and both growth \n343 models were assumed to perform differently. These assumptions were made due to evidence that \n344 energetic resources are diverted away from growing when approaching reproductive years or \n345 when fighting off disease (17,38,39), neither of which were evident in this study. Corals in the \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n17\n346 present study did not exhibit any decrease in growth trends, but it is important to acknowledge \n347 that this study did not include monitoring the same corals over time to establish growth trends. \n348 Further studies should monitor annual changes in coral growth or decline since outplantation. \n349 Other than MHW, covariates that significantly correlate to total growth are also significantly \n350 correlated to healthy growth in the same manner and proportion, indicating optimal conditions \n351 for growth did not change after infections. \n352 Total and healthy growth exhibited similar correlations as percent healthy cover with \n353 distance from the coast and roughness. Additionally slope, curvature, MCW, and high wind \n354 events correlated to total and healthy growth. Less terrain roughness and increased distance from \n355 shore positively correlated with higher total and healthy growth and likely occurs for similar \n356 reasons. Steeper slopes positively correlated with healthy and total growth. Steeper slopes may \n357 indicate areas of high-water flow (53,54), suggesting that while healthy coral cover benefits from \n358 water flow, growth is dependent upon the strength of such water flow providing ample nutrients \n359 and dissolved oxygen for mass transport. Higher curvature (convex terrain) correlated with lower \n360 growth and concave terrain correlated with higher growth. Large-scale concave geomorphology \n361 has been shown to negatively correlate to coral growth, likely as a result of less water flow (55). \n362 When zooming in to observe coral biogenic morphology, concavity in coral structure causes \n363 sediment accumulation (56) and an increase ratio in polyp to surface area, causing faster food \n364 depletion (57). Intermediate of these spectrums, at fine spatial scales, convex terrain represents \n365 coral outplant locations next to encroaching soft corals and rocks, limiting water flow and \n366 competition for resources, thereby reducing mass transport (58). Convex terrain was also highly \n367 rough and therefore likely correlated negatively with growth for similar reasons. \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n18\n368 It is not surprising that MHW and MCW negatively correlated to total and healthy coral \n369 growth. MHWs have a long-standing evidential history of adversely influencing coral success \n370 (21,40,59–62), and climate projections show MHWs are the main drivers inhibiting coral growth \n371 (63) and causing coral reef decline (2). While attention has focused on the effect of MHW over \n372 MCW, due to the expected increase in intensity and duration of MHW and decreased prevalence \n373 of MCW due to climate change (2), model results in this study revealed MCW had a greater \n374 negative impact on A. cervicornis total and healthy growth than MHW. One of the most severe \n375 cold-water events in the Florida Keys on record occurred in in the winter of 1976-1977 (64) and \n376 in 2010, causing mass mortalities throughout the Florida Reef Tract (8,65,66). A. cervicornis and \n377 other corals’ high vulnerability to cold water events raises some concerns over selective breeding \n378 and outplanting of heat tolerant corals to combat bleaching (67,68). This form of selection may \n379 leave reefs more vulnerable to the rare occurrence of cold wave events. Future selective breeding \n380 and genetic modifications should take this into account to ensure resilience under varying \n381 environmental conditions. \n382 In contrast to other studies, our model predicts that high wind events may enhance coral \n383 growth. These events have been associated (1–3) with increased lesions (34), reduced growth \n384 (69), and increase coastal run-off (13). However, the benefits of high winds include increased \n385 wave action that can lead to higher incidences of asexual propagation through fragmentation \n386 (7,38) and increased water flow which may mitigate bleaching by bringing in cool deeper waters \n387 (70–72). Our study sites were located at depths of less than 10 m, making it likely that high wind \n388 events had a direct effect on the coral reefs. Although the impact of high wind events on corals \n389 requires further study, our findings indicate they are less of a concern than high intensity thermal \n390 and cold wave events.\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n19\n391 Considerations for future outplanting\n392 Model results of curvature and roughness highlight recurring challenges with coral \n393 restoration; often coral outplanting results in unfavorable environmental conditions for optimal \n394 coral growth and survivorship. Large reef-building coral thickets can increase coral vulnerability \n395 to density induced disease spread (27,73), corallivore presence (48,74), and decreased water flow \n396 penetration into the thickets (42,75). In less rough and convex terrain A. cervicornis had higher \n397 growth and healthy cover trends, yet the coral itself has rough and convex morphology, resulting \n398 in the coral body building unfavorable terrain. Possible courses of actions include outplanting \n399 corals as soon as they are sexually reproductive so outplants may immediately start recruiting to \n400 the coral reef and to increase the adaptive potential of outplants to be more resilient to disease \n401 and optimize mass transport in the long-term.\n402 Coral outplanting and monitoring capabilities can often stand at odds with maximizing \n403 restoration success. Model results within this study and others have demonstrated that the \n404 predicted suitable habitats are in deeper, cooler, waters (at least below 15m deep) that act as \n405 refuge from MHW (24,64,76), and are located farther away from the coast and its associated \n406 anthropogenic impacts (1). However, these locations are especially hard for restoration groups to \n407 regularly access and monitor coral success. Consequently, greater effort could be made to work \n408 collaboratively with restoration groups to develop habitat suitability models which identify \n409 suitable sites useful to coastal managers and practitioners.\n410 The fine-scale digital terrain models developed from SfM can be used to inform several \n411 factors that may be useful in coral restoration areas. These include, but are not limited to, water \n412 flow (which is often used as a surrogate for the relative amount of food and oxygen delivered to \n413 species), presence of corallivores, disease, and areas that may be subject to upwelling and \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n20\n414 therefore mitigate temperature (47,77,78); all factors that help to define ecosystem health and \n415 distribution/abundance of biota (79). Model results of roughness and curvature within this study \n416 support this, where areas of high terrain roughness and convexity are primarily attributed to \n417 encroaching soft and other hard corals and rocks. The insights gained from SfM shows how \n418 higher resolution DTMs are able to bridge the gap from broader environmental conditions to \n419 more localized factors that influence organismal success (42), factors that are otherwise \n420 unavailable from satellite and sonar derived bathymetry (28,29,80,81).\n421 Due to the insulating capacity that deeper and distant waters provide corals from climate \n422 change and anthropogenic influence, auxiliary efforts towards technological developments for \n423 easier monitoring are necessary for long-term restoration work. Current reef restoration has \n424 focused on the shallow reef crest, reef crest, spur and groove terrain, forereef terrace, and deep \n425 reef (82). These sites, apart from the deep reef, range in depth from 0.7m to 10m deep and \n426 constitute over 56,000m2 of restorable area (82). These areas, while accessible, are the most \n427 vulnerable to climate change, necessitating outplanting deeper and farther from the coast. \n428 Opportunities to survey coral reef growth and health more efficiently using a variety of remote \n429 sensing techniques should be explored as it can increase monitoring capabilities. Potential \n430 applications include multibeam sonar, ICESat-2 satellite monitoring, and unmanned aerial \n431 surveillance. Complementing these survey methods with HSMs and deep learning coral detection \n432 methods (83) can provide more accessible monitoring of corals in environments which may be \n433 difficult to monitor recurringly in-situ. \n434 Summary and Conclusion\n435 As climate change progresses and environmental conditions continue to change, the \n436 available habitat for endangered species will continue to shrink. Conservation efforts to restore \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n21\n437 sensitive species populations requires attentive and rigorous research for what supports \n438 organismal success. Within modeling research, this necessitates finer-scale modeling, at the scale \n439 of meters to centimeters. While large-scale modeling is helpful towards understanding suitable \n440 habitat shifts under climate change, site-specific models are necessary to inform active \n441 restoration strategies. Structure-from-motion photogrammetry is unique from other survey \n442 methods in capturing continuous spatial data at the scale of millimeters. While SfM \n443 photogrammetry data collection is more time intensive compared to MBES, satellite, or UAS \n444 imagery data collection, the resulting data resolution is imperative to informing environmental \n445 influences on vulnerable species. \n446 Cumulatively, our results indicate that the variance of success within the fine-scale can be \n447 accounted for and incorporated into outplant decisions. A. cervicornis habitat suitability models \n448 of coral growth and healthy cover inform recommendations to outplant\n449 1) on steeper parts of the reefscape, along the reef edge\n450 2) within marine sanctuaries, preferably in as distant locations as possible from the coast\n451 3) in deeper waters, preferably deeper than 10m\n452 4) where there is minimal surrounding corals and rocks\n453 These outplant recommendations are developed with restoration group capabilities in mind, for \n454 outplant sites they are already frequenting. The creation of these models for A. cervicornis \n455 supports the applicability of fine-scale modeling of other endangered and vulnerable species, to \n456 support conservation efforts. \n457 Acknowledgement\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n22\n458 Thank you to the managers and staff at Mote Marine Laboratory for their support with field work \n459 and commitment to restoring Florida’s reefscape. This study was supported by NOAA grant # \n460 NA20NOS4000196. \n461\n462 References\n463 1. Golbuu Y, van Woesik R, Richmond RH, Harrison P, Fabricius KE. River discharge reduces reef \n464 coral diversity in Palau. Mar Pollut Bull. 2011 Apr 1;62(4):824–31. \n465 2. IPCC. Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the \n466 Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, H. \n467 Lee and J. Romero (eds.)] [Internet]. First. Geneva, Switzerland: IPCC; 2023 Jul [cited 2024 Mar 4]. \n468 Available from: https://www.ipcc.ch/report/ar6/syr/\n469 3. Valentine JF, Heck KL. Perspective review of the impacts of overfishing on coral reef food web \n470 linkages. Coral Reefs. 2005 Jun 1;24(2):209–13. \n471 4. Williams D, Miller M. Coral disease outbreak: pattern, prevalence and transmission in Acropora \n472 cervicornis. Mar Ecol Prog Ser. 2005;301:119–28. \n473 5. Acropora Biological Review Team. Atlantic Acropora Status Review Document [Internet]. National \n474 Marine Fisheries Service, Southeast Regional Office; 2005 Mar p. 152+. Available from: \n475 https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=61b96fc5c6b30c2c3b23e4b7af395\n476 1bd7365a982\n477 6. Calle-Triviño J, Muñiz-Castillo AI, Cortés-Useche C, Morikawa M, Sellares-Blasco R, Arias-\n478 González JE. Approach to the Functional Importance of Acropora cervicornis in Outplanting Sites in \n479 the Dominican Republic. Front Mar Sci [Internet]. 2021 [cited 2024 Jan 23];8. Available from: \n480 https://www.frontiersin.org/articles/10.3389/fmars.2021.668325\n481 7. Lirman D, Schopmeyer S, Galvan V, Drury C, Baker AC, Baums IB. Growth Dynamics of the \n482 Threatened Caribbean Staghorn Coral Acropora cervicornis: Influence of Host Genotype, Symbiont \n483 Identity, Colony Size, and Environmental Setting. PLOS ONE. 2014 Sep 30;9(9):e107253. \n484 8. Kemp DW, Colella MA, Bartlett LA, Ruzicka RR, Porter JW, Fitt WK. Life after cold death: reef \n485 coral and coral reef responses to the 2010 cold water anomaly in the Florida Keys. Ecosphere. \n486 2016;7(6):e01373. \n487 9. van Woesik R, Roth LM, Brown EJ, McCaffrey KR, Roth JR. Niche space of corals along the Florida \n488 reef tract. PLoS ONE. 2020 Apr 7;15(4):e0231104. \n489 10. Precht WF, Gintert BE, Robbart ML, Fura R, van Woesik R. Unprecedented Disease-Related Coral \n490 Mortality in Southeastern Florida. Sci Rep. 2016 Aug 10;6(1):31374. \n491 11. Aronson RB, Precht WF. White-band disease and the changing face of Caribbean coral reefs. \n492 Hydrobiologia. 2001 Sep 1;460(1):25–38. \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n23\n493 12. van Woesik R, Banister RB, Bartels E, Gilliam DS, Goergen EA, Lustic C, et al. Differential survival \n494 of nursery-reared Acropora cervicornis outplants along the Florida reef tract. Restor Ecol. \n495 2021;29(1):e13302. \n496 13. Riegl B, Purkis SJ, Keck J, Rowlands GP. Monitored and modeled coral population dynamics and the \n497 refuge concept. Mar Pollut Bull. 2009 Jan 1;58(1):24–38. \n498 14. Couce E, Ridgwell A, Hendy EJ. Future habitat suitability for coral reef ecosystems under global \n499 warming and ocean acidification. Glob Change Biol. 2013;19(12):3592–606. \n500 15. Ware M, Garfield EN, Nedimyer K, Levy J, Kaufman L, Precht W, et al. Survivorship and growth in \n501 staghorn coral (Acropora cervicornis) outplanting projects in the Florida Keys National Marine \n502 Sanctuary. PLOS ONE. 2020 May 6;15(5):e0231817. \n503 16. Lirman D, Thyberg T, Herlan J, Hill C, Young-Lahiff C, Schopmeyer S, et al. Propagation of the \n504 threatened staghorn coral Acropora cervicornis: methods to minimize the impacts of fragment \n505 collection and maximize production. Coral Reefs. 2010 Sep 1;29(3):729–35. \n506 17. Omori M. Coral restoration research and technical developments: what we have learned so far. Mar \n507 Biol Res. 2019 Oct 14;15:1–33. \n508 18. Boström-Einarsson L, Babcock RC, Bayraktarov E, Ceccarelli D, Cook N, Ferse SCA, et al. Coral \n509 restoration – A systematic review of current methods, successes, failures and future directions. PLOS \n510 ONE. 2020 Jan 30;15(1):e0226631. \n511 19. Fox HE, Harris JL, Darling ES, Ahmadia GN, Estradivari, Razak TB. Rebuilding coral reefs: success \n512 (and failure) 16 years after low-cost, low-tech restoration. Restor Ecol. 2019;27(4):862–9. \n513 20. Devlin-Durante MK, Miller MW, Group CAR, Precht WF, Baums IB. How old are you? Genet age \n514 estimates in a clonal animal. Mol Ecol. 2016;25(22):5628–46. \n515 21. Hughes TP. Life histories and population dynamics of early successional corals. In: Gabrie C, Salvat \n516 B, editors. Moorea, French Polynesia: Antenne Museum-EPHE; 1985 [cited 2023 Aug 4]. p. 101–6. \n517 Available from: http://www.reefbase.org/resource_center/publication/icrs.aspx\n518 22. Irwin A, Greer L, Humston R, Devlin-Durante M, Cabe P, Lescinsky H, et al. Age and intraspecific \n519 diversity of resilient Acropora communities in Belize. Coral Reefs. 2017 Dec 1;36(4):1111–20. \n520 23. Banister RB, van Woesik R. Ranking 67 Florida Reefs for Survival of Acropora cervicornis \n521 Outplants. Front Mar Sci [Internet]. 2021 [cited 2023 Jul 19];8. Available from: \n522 https://www.frontiersin.org/articles/10.3389/fmars.2021.672574\n523 24. Giraldo-Ospina A, Kendrick GA, Hovey RK. Depth moderates loss of marine foundation species \n524 after an extreme marine heatwave: could deep temperate reefs act as a refuge? Proc R Soc B Biol Sci. \n525 2020 Jun 10;287(1928):20200709. \n526 25. Fabricius KE. Factors Determining the Resilience of Coral Reefs to Eutrophication: A Review and \n527 Conceptual Model. In: Dubinsky Z, Stambler N, editors. Coral Reefs: An Ecosystem in Transition \n528 [Internet]. Dordrecht: Springer Netherlands; 2011 [cited 2023 Aug 1]. p. 493–505. Available from: \n529 https://doi.org/10.1007/978-94-007-0114-4_28\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n24\n530 26. Manzello DP. Coral growth with thermal stress and ocean acidification: lessons from the eastern \n531 tropical Pacific. Coral Reefs. 2010 Sep 1;29(3):749–58. \n532 27. Shore A, Caldwell JM. Modes of coral disease transmission: how do diseases spread between \n533 individuals and among populations? Mar Biol. 2019 Feb 26;166(4):45. \n534 28. Burns JHR, Delparte D, Gates RD, Takabayashi M. UTILIZING UNDERWATER THREE-\n535 DIMENSIONAL MODELING TO ENHANCE ECOLOGICAL AND BIOLOGICAL STUDIES OF \n536 CORAL REEFS. Int Arch Photogramm Remote Sens Spat Inf Sci. 2015 Apr 9;XL-5/W5:61–6. \n537 29. Fukunaga A, Burns JHR, Craig BK, Kosaki RK. Integrating Three-Dimensional Benthic Habitat \n538 Characterization Techniques into Ecological Monitoring of Coral Reefs. J Mar Sci Eng. 2019 \n539 Feb;7(2):27. \n540 30. McDowall P, Lynch HJ. Ultra-Fine Scale Spatially-Integrated Mapping of Habitat and Occupancy \n541 Using Structure-From-Motion. PLOS ONE. 2017 Jan 11;12(1):e0166773. \n542 31. Young GC, Dey S, Rogers AD, Exton D. Cost and time-effective method for multi-scale measures of \n543 rugosity, fractal dimension, and vector dispersion from coral reef 3D models. PLOS ONE. 2017 Apr \n544 13;12(4):e0175341. \n545 32. Pierce J, Butler MJ, Rzhanov Y, Lowell K, Dijkstra JA. Classifying 3-D Models of Coral Reefs \n546 Using Structure-From-Motion and Multi-View Semantic Segmentation. Front Mar Sci [Internet]. \n547 2021 Oct 29 [cited 2025 Apr 28];8. Available from: \n548 https://www.frontiersin.orghttps://www.frontiersin.org/journals/marine-\n549 science/articles/10.3389/fmars.2021.706674/full\n550 33. Hobday AJ, Alexander LV, Perkins SE, Smale D, Straub S, Oliver ECJ, et al. A hierarchical \n551 approach to defining marine heatwaves. Prog Oceanogr. 2016 Feb;141:227–38. \n552 34. Goergen EA, Moulding AL, Walker BK, Gilliam DS. Identifying Causes of Temporal Changes in \n553 Acropora cervicornis Populations and the Potential for Recovery. Front Mar Sci [Internet]. 2019 \n554 [cited 2022 Jun 20];6. Available from: https://www.frontiersin.org/article/10.3389/fmars.2019.00036\n555 35. Brooks ME, Kristensen K, Benthem K J ,van, Magnusson A, Berg C W, Nielsen A, et al. glmmTMB \n556 Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed \n557 Modeling. R J. 2017;9(2):378. \n558 36. Brooks ME, Kristensen K, Benthem KJ van, Magnusson A, Berg CW, Nielsen A, et al. Modeling \n559 zero-inflated count data with glmmTMB [Internet]. bioRxiv; 2017 [cited 2023 Apr 4]. p. 132753. \n560 Available from: https://www.biorxiv.org/content/10.1101/132753v1\n561 37. Lüdecke D, Ben-Shachar M, Patil I, Waggoner P, Makowski D. performance: An R Package for \n562 Assessment, Comparison and Testing of Statistical Models. J Open Source Softw. 2021 Apr \n563 21;6(60):3139. \n564 38. Lirman D. Fragmentation in the branching coral Acropora palmata (Lamarck): growth, survivorship, \n565 and reproduction of colonies and fragments. J Exp Mar Biol Ecol. 2000 Aug 23;251(1):41–57. \n566 39. REEFocus. CRF. 2020 [cited 2023 Apr 5]. NEW RESEARCH TRACKS TIMING OF CORAL \n567 SPAWNING ON FLORIDA’S CORAL REEF. Available from: \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n25\n568 https://www.coralrestoration.org/post/new-research-tracks-timing-of-coral-spawning-on-floridas-\n569 coral-reef\n570 40. Asner GP, Vaughn NR, Martin RE, Foo SA, Heckler J, Neilson BJ, et al. Mapped coral mortality and \n571 refugia in an archipelago-scale marine heat wave. Proc Natl Acad Sci. 2022 May \n572 10;119(19):e2123331119. \n573 41. Buma AGJ, De Boer MK, Boelen P. Depth Distributions of Dna Damage in Antarctic Marine Phyto- \n574 and Bacterioplankton Exposed to Summertime Uv Radiation. J Phycol. 2001;37(2):200–8. \n575 42. Stocking JB, Rippe JP, Reidenbach MA. Structure and dynamics of turbulent boundary layer flow \n576 over healthy and algae-covered corals. Coral Reefs. 2016 Sep 1;35(3):1047–59. \n577 43. Stocking JB, Laforsch C, Sigl R, Reidenbach MA. The role of turbulent hydrodynamics and surface \n578 morphology on heat and mass transfer in corals. J R Soc Interface. 2018 Dec 19;15(149):20180448. \n579 44. Brodnicke OB, Bourne DG, Heron SF, Pears RJ, Stella JS, Smith HA, et al. Unravelling the links \n580 between heat stress, bleaching and disease: fate of tabular corals following a combined disease and \n581 bleaching event. Coral Reefs. 2019 Aug 1;38(4):591–603. \n582 45. Nelson HR, Altieri AH. Oxygen: the universal currency on coral reefs. Coral Reefs. 2019 Apr \n583 1;38(2):177–98. \n584 46. Pollock FJ, Lamb JB, Field SN, Heron SF, Schaffelke B, Shedrawi G, et al. Sediment and Turbidity \n585 Associated with Offshore Dredging Increase Coral Disease Prevalence on Nearby Reefs. PLOS ONE. \n586 2014 Jul 16;9(7):e102498. \n587 47. Borland HP, Gilby BL, Henderson CJ, Leon JX, Schlacher TA, Connolly RM, et al. The influence of \n588 seafloor terrain on fish and fisheries: A global synthesis. Fish Fish. 2021;22(4):707–34. \n589 48. Montalbetti E, Fallati L, Casartelli M, Maggioni D, Montano S, Galli P, et al. Reef complexity \n590 influences distribution and habitat choice of the corallivorous seastar Culcita schmideliana in the \n591 Maldives. Coral Reefs. 2022 Apr 1;41(2):253–64. \n592 49. Nicolet KJ, Chong-Seng KM, Pratchett MS, Willis BL, Hoogenboom MO. Predation scars may \n593 influence host susceptibility to pathogens: evaluating the role of corallivores as vectors of coral \n594 disease. Sci Rep. 2018 Mar 27;8(1):5258. \n595 50. Williams GJ, Aeby GS, Cowie ROM, Davy SK. Predictive Modeling of Coral Disease Distribution \n596 within a Reef System. PLOS ONE. 2010 Feb 17;5(2):e9264. \n597 51. Hearn C, Atkinson M, Falter J. A physical derivation of nutrient-uptake rates in coral reefs: effects of \n598 roughness and waves. Coral Reefs. 2001 Dec 1;20(4):347–56. \n599 52. Morrow KM, Pankey MS, Lesser MP. Community structure of coral microbiomes is dependent on \n600 host morphology. Microbiome. 2022 Jul 28;10(1):113. \n601 53. Eugene Shinn. Spur and Groove Formation on the Florida Reef Tract. SEPM J Sediment Res \n602 [Internet]. 1963 [cited 2023 Apr 5];Vol. 33. Available from: \n603 https://pubs.geoscienceworld.org/jsedres/article/33/2/291-303/112994\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n26\n604 54. Hernández-Molina FJ, Serra N, Stow DAV, Llave E, Ercilla G, Van Rooij D. Along-slope \n605 oceanographic processes and sedimentary products around the Iberian margin. Geo-Mar Lett. 2011 \n606 Dec 1;31(5):315–41. \n607 55. Sebens KP, Helmuth B, Carrington E, Agius B. Effects of water flow on growth and energetics of the \n608 scleractinian coral Agaricia tenuifolia in Belize. Coral Reefs. 2003 Apr 1;22(1):35–47. \n609 56. Philipp E, Fabricius K. Photophysiological stress in scleractinian corals in response to short-term \n610 sedimentation. J Exp Mar Biol Ecol. 2003 Feb 26;287(1):57–78. \n611 57. Merks RMH, Hoekstra AG, Kaandorp JA, Sloot PMA. Polyp oriented modelling of coral growth. J \n612 Theor Biol. 2004 Jun 21;228(4):559–76. \n613 58. Chadwick NE, Morrow KM. Competition Among Sessile Organisms on Coral Reefs. In: Dubinsky Z, \n614 Stambler N, editors. Coral Reefs: An Ecosystem in Transition [Internet]. Dordrecht: Springer \n615 Netherlands; 2011 [cited 2024 Mar 6]. p. 347–71. Available from: https://doi.org/10.1007/978-94-\n616 007-0114-4_20\n617 59. Ainsworth TD, Hurd CL, Gates RD, Boyd PW. How do we overcome abrupt degradation of marine \n618 ecosystems and meet the challenge of heat waves and climate extremes? Glob Change Biol. \n619 2020;26(2):343–54. \n620 60. Donovan MK, Burkepile DE, Kratochwill C, Shlesinger T, Sully S, Oliver TA, et al. Local conditions \n621 magnify coral loss after marine heatwaves. Science. 2021 May 28;372(6545):977–80. \n622 61. Foo SA, Asner GP. Sea surface temperature in coral reef restoration outcomes. Environ Res Lett. \n623 2020 Jul 1;15(7):074045. \n624 62. Oliver ECJ, Benthuysen JA, Darmaraki S, Donat MG, Hobday AJ, Holbrook NJ, et al. Marine \n625 Heatwaves. Annu Rev Mar Sci. 2021;13(1):313–42. \n626 63. Cantin NE, Cohen AL, Karnauskas KB, Tarrant AM, McCorkle DC. Ocean warming slows coral \n627 growth in the central Red Sea. Science. 2010 Jul 16;329(5989):322–5. \n628 64. Porter JW, Battey JF, Smith GJ. Perturbation and change in coral reef communities. Popul Biol. 1981 \n629 Nov 2;79:1678–81. \n630 65. Colella MA, Ruzicka RR, Kidney JA, Morrison JM, Brinkhuis VB. Cold-water event of January \n631 2010 results in catastrophic benthic mortality on patch reefs in the Florida Keys. Coral Reefs. 2012 \n632 Jun 1;31(2):621–32. \n633 66. Lirman D, Schopmeyer S, Manzello D, Gramer LJ, Precht WF, Muller-Karger F, et al. Severe 2010 \n634 Cold-Water Event Caused Unprecedented Mortality to Corals of the Florida Reef Tract and Reversed \n635 Previous Survivorship Patterns. PLOS ONE. 2011 Aug 10;6(8):e23047. \n636 67. Buerger P, Alvarez-Roa C, Coppin CW, Pearce SL, Chakravarti LJ, Oakeshott JG, et al. Heat-\n637 evolved microalgal symbionts increase coral bleaching tolerance. Sci Adv. 2020 May \n638 13;6(20):eaba2498. \n639 68. Howells EJ, Abrego D, Liew YJ, Burt JA, Meyer E, Aranda M. Enhancing the heat tolerance of reef-\n640 building corals to future warming. Sci Adv. 2021 Aug 20;7(34):eabg6070. \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n27\n641 69. Muko S, Arakaki S, Nagao M, Sakai K. Growth form-dependent response to physical disturbance and \n642 thermal stress in Acropora corals. Coral Reefs. 2013 Mar 1;32(1):269–80. \n643 70. Carrigan AD, Puotinen M. Tropical cyclone cooling combats region-wide coral bleaching. Glob \n644 Change Biol. 2014;20(5):1604–13. \n645 71. Manzello DP, Brandt M, Smith TB, Lirman D, Hendee JC, Nemeth RS. Hurricanes benefit bleached \n646 corals. Proc Natl Acad Sci. 2007 Jul 17;104(29):12035–9. \n647 72. Page CE, Leggat W, Heron SF, Fordyce AJ, Ainsworth TD. High flow conditions mediate damaging \n648 impacts of sub-lethal thermal stress on corals’ endosymbiotic algae. Conserv Physiol. 2021 Jan \n649 1;9(1):coab046. \n650 73. Anderson RM, May RM. Population biology of infectious diseases: Part I. Nature. 1979 \n651 Aug;280(5721):361–7. \n652 74. Koval G, Rivas N, D’Alessandro M, Hesley D, Santos R, Lirman D. Fish predation hinders the \n653 success of coral restoration efforts using fragmented massive corals. PeerJ. 2020 Oct 2;8:e9978. \n654 75. Reidenbach MA, Koseff JR, Monismith SG. Laboratory experiments of fine-scale mixing and mass \n655 transport within a coral canopy. Phys Fluids. 2007 Jul;19(7):075107. \n656 76. Harrison HB, Álvarez-Noriega M, Baird AH, Heron SF, MacDonald C, Hughes TP. Back-to-back \n657 coral bleaching events on isolated atolls in the Coral Sea. Coral Reefs. 2019 Aug 1;38(4):713–9. \n658 77. Barott KL, Rohwer FL. Unseen players shape benthic competition on coral reefs. Trends Microbiol. \n659 2012 Dec 1;20(12):621–8. \n660 78. Chong-Seng KM, Cole AJ, Pratchett MS, Willis BL. Selective feeding by coral reef fishes on coral \n661 lesions associated with brown band and black band disease. Coral Reefs. 2011 Jun 1;30(2):473–81. \n662 79. Pygas DR, Ferrari R, Figueira WF. Review and meta-analysis of the importance of remotely sensed \n663 habitat structural complexity in marine ecology. Estuar Coast Shelf Sci. 2020 Apr 5;235:106468. \n664 80. Burns JHR, Fukunaga A, Pascoe KH, Runyan A, Craig BK, Talbot J, et al. 3D HABITAT \n665 COMPLEXITY OF CORAL REEFS IN THE NORTHWESTERN HAWAIIAN ISLANDS IS \n666 DRIVEN BY CORAL ASSEMBLAGE STRUCTURE. Int Arch Photogramm Remote Sens Spat Inf \n667 Sci. 2019 Apr 17;XLII-2/W10:61–7. \n668 81. Combs IR. Characterizing the Impacts of Stony Coral Tissue Loss Disease on Coral in Southeast \n669 Florida Using 3D Photogrammetry [Internet] [M.S.]. ProQuest Dissertations and Theses. [United \n670 States -- Florida]: Florida Atlantic University; 2019 [cited 2023 Mar 28]. Available from: \n671 https://www.proquest.com/docview/2333948527/abstract/5040EC761EAC4D06PQ/1\n672 82. Restoring Seven Iconic Reefs: A Mission to Recover the Coral Reefs of the Florida Keys. \n673 83. Trudeau GA, Lowell K, Dijkstra JA. Coral reef detection using ICESat-2 and machine learning. Ecol \n674 Inform. 2025 Jul 1;87:103099. \n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n28\n675 84. Kopp S. New Surface Analysis Capabilities in ArcGIS Pro 2.7 [Internet]. ArcGIS Blog. 2/21 [cited \n676 2023 Apr 10]. Available from: https://www.esri.com/arcgis-blog/products/arcgis-pro/analytics/new-\n677 slope-aspect-curvature/\n678 85. Evans IS. General geomorphometry, derivatives of altitude, and descriptive statistics. In: Spatial \n679 Analysis in Geomorphology. Routledge; 1972. \n680 86. Day M, Chenoweth S. 6.14 Surface Roughness of Karst Landscapes. In: Shroder JF, editor. Treatise \n681 on Geomorphology [Internet]. San Diego: Academic Press; 2013 [cited 2023 Jul 13]. p. 157–63. \n682 Available from: https://www.sciencedirect.com/science/article/pii/B9780123747396001081\n683 87. Maxwell AE, Shobe CM. Land-surface parameters for spatial predictive mapping and modeling. \n684 Earth-Sci Rev. 2022 Mar 1;226:103944. \n685 88. Restoring Seven Iconic Reefs: A Mission to Recover the Coral Reefs of the Florida Keys. Mission \n686 Iconic Reefs; \n687\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \nwas not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprint (whichthis version posted July 15, 2025. ; https://doi.org/10.1101/2025.07.11.664295doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}