Next Article in Journal
An Automatic Detection Method of Slow-Moving Landslides Using an Improved Faster R-CNN Model Based on InSAR Deformation Rates
Previous Article in Journal
Instance Segmentation of LiDAR Point Clouds with Local Perception and Channel Similarity
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 17
Issue 18
10.3390/rs17183241
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals

by
Mia B. Melamed
1,2,
Roberta E. Martin
1,3,*,
McKenna Allen
1,2 and
Gregory P. Asner
1,3
1
Center for Global Discovery and Conservation Science, Arizona State University, Hilo, HI 96720, USA
2
Hawai‘i Department of Land and Natural Resources, Division of Aquatic Resources, Honolulu, HI 96813, USA
3
School of Ocean Futures, Arizona State University, Hilo, HI 96720, USA
*
Author to whom correspondence should be addressed.
Remote Sens. 2025, 17(18), 3241; https://doi.org/10.3390/rs17183241
Submission received: 1 August 2025 / Revised: 31 August 2025 / Accepted: 12 September 2025 / Published: 19 September 2025
Browse Figures
Versions Notes

Abstract

Highlights

What are the main findings?
  • Spectral reflectance analysis of Montiporid corals at ‘Anini Reef revealed that live tissue on colonies with black band disease can be reliably distinguished from healthy colonies even before visual symptoms emerge, with classification accuracy exceeding 85%.
What is the implication of the main finding?
  • Integrating spectroscopy into monitoring and restoration frameworks enables earlier detection and targeted intervention—supporting rapid-response treatment, improved nursery screening, and stronger reef resilience strategies as coral disease risk in-creases under climate change.

Abstract

Coral reefs are essential to the cultural, ecological, and economic well-being of Hawai‘i’s communities, yet they face increasing threats from environmental changes and localized stressors, including coral disease. Detecting coral disease often relies on the visible appearance of lesions; however, in the case of black-band disease (BBD), this visual cue appears too late, as disease progression can cause an average rate of tissue loss of up to 5.7 cm2 per day over two months, followed by partial or full colony mortality. Reflectance spectroscopy offers a promising tool for detecting subtle spectral changes associated with coral health before visible symptoms emerge, yet few studies have applied this method to coral disease. In situ spectroscopy was used to measure the spectral reflectance of health conditions in Montiporid corals at ‘Anini Reef, Kaua‘i, USA. Discriminant analysis revealed that visually identical tissue types—live tissue on colonies with BBD (liveD) and live tissue on colonies without BBD (liveL)—were spectrally distinct. In contrast, BBD lesions (disease) and adjacent tissue that appeared healthy (transition) exhibited similar spectral signatures. Analyses identified three spectrally distinct tissue health conditions with a misclassification rate of 12.8%. These findings highlight the potential of reflectance spectroscopy for early coral disease detection, which could improve response times and support more effective coral reef conservation efforts.

1. Introduction

Scleractinian corals form the foundation of coral reef ecosystems, which are among the most biodiverse ecosystems on the planet. Although coral reefs cover less than 1% of the Earth’s surface, they provide food, shelter, and habitat for nearly a quarter of all marine life. In Hawai‘i alone, coastal protection services have an estimated value of 836 million annually [1,2,3]. However, these ecosystems are under increasing pressure from global climate change and local stressors, including coastal development, worsening water quality, sedimentation, eutrophication, fishing pressure and shifts in ocean temperature and chemistry [4,5]. Over the past two decades, coral disease has emerged as a significant global threat, contributing to widespread declines in key reef-building species [6]. The growing impact of coral disease is closely linked to anthropogenic stressors, which include coastal development, poor water quality, sedimentation, eutrophication, fishing pressure, and changes in ocean temperature and chemistry [4,7]. Under projected warming scenarios, corals are expected to experience disease at alarming rates, threatening the functional stability of coral reef ecosystems and potentially leading to environmental collapse [8,9].
One disease of particular concern for Hawai‘i is black band disease (BBD), a lethal and infectious condition that affects several coral species globally [10]. To date, within the Hawaiian archipelago, BBD has only been documented on the north and east shores of Kauaʻi Island [11], as previously reported [12]. BBD is characterized by a visible black band formed by microbes, primarily cyanobacteria, that rapidly advances across coral colonies, destroying living tissue and leaving behind a bare skeleton [12,13,14]. The disease was first documented on Kaua‘i’s north shore in 2004 (Figure 1) at a prevalence of less than 1%. An outbreak later occurred at ‘Anini Reef in 2012 (Figure 1), where recent surveys have recorded prevalence rates as high as 18% [12,15]. Coral disease detection typically relies on the visible appearance of lesions; however, in the case of BBD, these signs often appear too late for effective intervention. With tissue loss progressing as fast as 5.7 cm2 per day, BBD can cause partial or complete mortality within weeks [12]. For this reason, early detection is crucial in preventing the rapid decline of coral cover due to diseases such as black band disease.
Conventional coral disease monitoring relies on visual assessments, which are time- and labor-intensive and rely on an individual’s ability to recognize disease only after lesions, in this case a black band, appear [16,17,18]. In addition, the characteristic appearance of BBD is seasonally variable, increasing in prevalence during the summer months when ocean temperatures and solar radiation are heightened and lesions subsiding or becoming inactive as temperatures decline in the fall and winter [19,20]. As coral diseases are expected to increase in their frequency and impact, there is a growing need for earlier detection to support effective management. Reflectance spectroscopy offers a promising way to detect subtle spectral changes in coral health before visible symptoms emerge, yet few studies have applied this technique to coral disease [21]. Coral spectral properties are primarily influenced by the composition and density of pigments within the tissue and their interaction with sunlight [22,23]. In situ reflectance spectroscopy captures these properties by measuring the proportion of reflected light from a surface in the visible range of the electromagnetic spectrum—a signal influenced by the coral’s biochemical makeup and skeletal structure. A key contributor to coral optical signals is the endosymbiotic dinoflagellate Symbiodiniaceae (zooxanthellae), which shallow water scleractinian corals rely on to absorb specific wavelengths for photosynthesis. Variations in Symbiodiniaceae density affect light absorption, leading to distinct reflectance spectra, as shown in studies of coral spectral properties [24,25,26]. Reflectance spectroscopy has been successfully applied in coral reef studies, from broad benthic classification to identifying coral functional types, and recent advances include analyzing symbiont chemistry to predict thermally tolerant corals [24,25,26]. While remote sensing has successfully been used to detect stress in terrestrial systems such as agriculture and forestry [23,27,28], as well as coral bleaching, its potential for monitoring coral disease is only beginning to be explored [21].
The current approach to coral disease is largely reactive, intervening only after corals exhibit visible symptoms [29]. While early warning systems based on environmental conditions, such as rising sea surface temperatures and declining water quality, can help predict periods of heightened stress, they do not inform the onset of coral disease [30]. Developing methods that reveal signs of changes in coral health before visual symptoms are present could transform management, enabling more rapid and effective interventions to minimize coral loss, particularly in fast-progressing diseases that cause extensive tissue degradation.
In this study, we tested the applicability of reflectance spectroscopy to detect stress responses to black band disease before symptoms become visible on Montiporid corals at ‘Anini Reef, Kaua‘i. To address the limitations of existing coral disease detection and monitoring, we measured the in situ reflectance from 15 Montiporid colonies and used a linear discriminant analysis to investigate whether tissue health categories were spectrally separable. Our findings build on recent advances in coral spectral ecology and forward spectroscopy as a scalable method for integrating early disease detection into reef monitoring and management. This study contributes to the growing body of research that aims to shift coral disease responses from reactive intervention toward predictive and preventative strategies, an essential shift for sustaining reef resilience in the face of accelerating climate change and intensifying human impacts.

2. Materials and Methods

2.1. Coral Tissue Spectral Reflectance Sampling

Montiporid coral colonies (n = 15) were sampled in the shallow fringing backreef of ‘Anini Reef on Kaua‘i Island, Hawai‘i, USA (Figure 1). Sampling sites across the reef were identified based on BBD prevalence surveys conducted biannually since 2019 by the State of Hawai‘i’s Kaua‘i Division of Aquatic Resources [11]. Tissue health conditions were categorized based on visual assessment of live tissue and the absence or presence of the “black band” that is characteristic of this disease. An example is shown in Figure 2, and additional examples are provided in Figure A1. Tissue health categories on Montiporids affected by BBD are defined here: “disease” is a black band disease lesion, “transition” is the tissue immediately adjacent to a BBD lesion, and “liveD” is live tissue sampled from a colony with a visible BBD lesion. Live tissue on Montiporid colonies without a visible BBD lesion was listed as “liveL”.
Reflectance spectra were measured using a handheld spectroradiometer (Analytical Spectral Devices HH2-Pro, Panalytic Inc., Boulder, CO, USA) with a tungsten halogen light calibrated to the solar spectrum (Keldan Inc., Zurich, Switzerland) [24]. Reflectance was calculated from each spectral radiance measurement using a calibration panel (Spectralon®, Labsphere Inc., Durham, NH, USA) by calculating the ratio between the radiance of the sample and the incident radiance. Samples were collected in situ by divers at depths ranging between 1.4 m and 2.6 m on 12 August 2024, between 9:00 and 14:00. Lighting conditions were similar throughout the time period, with the additional calibrated light source contributing most of the incoming light to support uniform conditions.
Spectral signatures were collected from two species of Montiporid with encrusting morphology. A total of eight colonies were sampled in the dominant species, contracting BBD, Montipora capitata. Samples from two colonies of the wide-ranging and less abundant species Montipora patula were also included, but separation between species was not evaluated. Figure 2 shows an example of the spectral measurements collected across the colony from the diseased area extending through the live tissue. The tables included in the figure summarize the number of spectra for each tissue health category and the number of colonies sampled. Colonies were located as part of a larger DAR survey and were selected to maximize sampling time for this initial pilot study (Figure 1).
The reflectance probe was held at a constant 3 cm from the target, with an effective surface sampling area of 1.39 cm2. Corals affected by BBD were sought to contain an area of each tissue health category that allowed a minimum of three tissue health samples across the colony (Figure 2). The reflectance sampling field of view was 1.3 nm full-width at half-maximum (FWHM) from 400–700 nm.

2.2. Data Processing and Analysis

2.2.1. Data Cleaning and Standardization

Noise was trimmed at both ends of the spectra, leaving a wavelength range of 420–680 nm. Of the 220 samples, two live tissue samples were identified as areas of necrotic tissue and were excluded. Cross-spectrum brightness normalization was applied to all spectra prior to analysis. Brightness normalization (BN) is calculated as
B N = ∑ i R i 2
where R is the reflectance value at wavelength (nm) i for a given pixel. For this study, R included wavelength bands of 420–680 nm. BN was used to account for differences in the overall brightness of the spectra, thereby minimizing unwanted variations in intensity while retaining the features pertinent to the spectral signatures [31].

2.2.2. Discriminant Analysis

Linear discriminant analysis (LDA) was used to evaluate how well tissue health categories could be predicted from spectral data. LDA was chosen for this exploratory analysis because it utilizes a principal component analysis of the spectral data to establish the differences between known categories and then uses derived patterns to predict or classify new observations within those known groups, facilitating an end-to-end evaluation of the feasibility of discrimination among categories. LDA is particularly helpful in exploratory cases with smaller, unbalanced sample sizes because, unlike clustering or purely variance-based approaches such as principal component analysis, partial least squares analysis, or direct identification machine learning models (i.e., support vector machines or random forest models), LDA leverages predefined class information to optimize category separation [32].
Because the spectral data (wavelength bands = 261) exceeded the number of samples (n = 218), wide LDA was implemented. This approach applies a pseudoinverse to address the singular within-class covariance matrix that arises in these situations. All statistical analyses were performed in JMP version 18 [32]. Two separate LDA models were developed using spectral data between 420–680 nm, with the LDA variables assigned as tissue health categories. In the first model, four tissue health categories were evaluated independently. In the second, the BBD and BBD-transition categories were grouped into a single “infected” category. For both models, the predictor variables were brightness-normalized reflectance values for wavelength bands from 420–680 nm. Model performance was assessed using misclassification rates, with lower values indicating stronger discriminatory power. Additionally, entropy R-squared from the LDA was used to quantify the proportion of uncertainty in group membership explained by the model, where a value of 1 indicates that classifications are perfectly predicted [32].

3. Results

3.1. Spectral Signatures of Tissue Health Categories

Spectral signatures of Montipora capitata and Montipora patula displayed similar patterns in reflectance and absorption features; therefore, the spectral reflectance data were combined and analyzed at the genus level (Figure A2 and Figure A3). The mean BN reflectance varied in relative magnitude and shape across tissue health categories, revealing spectral signatures for liveD, liveL, disease, and transition (Figure 3). For comparison, Figure A4 shows the mean spectral reflectance (%). On average, spectra for liveD and liveL intersected near 555 nm and again around 650 nm (Figure 3), revealing shifts in reflectance patterns for live tissue on colonies with and without BBD. Both liveD and liveL had reflectance peaks near 580 nm, 605 nm, and 640 nm, though liveL maintained a higher intensity throughout this range. LiveD exhibited a more gradual incline before reaching peak brightness reflectance and broader reflectance peaks compared to liveL.
Overall, liveL exhibited the highest reflectance and the deepest absorption features, with the least variation from the mean reflectance across the spectral range. Meanwhile, liveD absorption features below 550 nm and above 650 nm were weaker. Moreover, greater spectral variability was observed for liveD, with the most deviation from the mean reflectance across the spectral range. In contrast, disease and transition spectra demonstrated the least change in slope, with muted reflectance peaks and absorption features. Similar features for all four tissue health categories included a maximum reflectance located near 600 nm and the strongest absorption well occurring near 670 nm. A weak anomalous peak at ~480 nm, due to ambient light reflecting the blue water signal [24], did not impact spectral separability among tissue health categories.

3.2. Separability Analysis

3.2.1. Spectral Separability of Four Tissue Health Categories

The LDA included four tissue health categories (liveL, liveD, disease, and transition), resulting in a misclassification rate of 26.6% and an entropy R2 = 0.38 (Table 1). LiveL had the highest prediction accuracy (93%), though it was sometimes confused with liveD (7%) but never confused with disease or transition (0%), meaning that with confidence, liveL is spectrally separable from disease and transition and is mostly separable from liveD. The variation in the reflectance and absorption for liveD was evident in its confusion with disease (7%), liveL (8%), and transition (8%). Higher rates of confusion between disease and transition were likely major contributors to the misclassification rate (26.6%). These results show that while there is uncertainty in classification, nearly 73% of the tissue health predictions were assigned with confidence based on their assigned category.
Visualization of the first two canonical axes showed multivariate spectral clustering of tissue health categories while exhibiting substantial overlap between the disease and transition groups, illustrating what is more likely to be three tissue health categories as opposed to four (Figure 4). Wilks’ Lambda test results indicate that at least one of the group means is significantly separable from the other tissue health groups (Λ = 0.12; F (57, 585.24) = 10.45, p < 0.001). See canonical details in Table 2.

3.2.2. Spectral Separability of Three Tissue Health Categories

Based on the overlap in discriminant space and confusion rates in the disease and transition categories, a second LDA was performed, combining these categories into a general infection category (infected). An LDA using the three tissue health categories, infected, liveD, and liveL, assessed whether this grouping improved class separability by reducing redundancy between highly similar categories. Classification performance showed a marked improvement, reducing the overall misclassification rate from 26.6% to 12.8% (Table 1). This suggests the previous misclassification rate was mostly due to the spectral similarity between disease and transition. The model correctly identified 94% of the infected samples, which were sometimes confused with liveD (6%) but never confused with liveL (0%). Prediction accuracy for liveD improved from 77% to 81%. Confusion rates remained the same for liveL.
Visualization of the two canonicals provided the maximum separation for the three tissue health categories (Figure 5). This refined classification showed a slightly stronger separation, as indicated by Wilks’ Lambda test results (Λ = 0.13; F (38, 394) = 18.07, p < 0.0001) (Table 3).

4. Discussion

4.1. Spectral Separability of Tissue Health Categories

Spectral reflectance measurements of tissue health categories affected by black band disease (BBD) on Montiporid corals at ‘Anini Reef, Kaua‘i, USA, revealed patterns that diverged from categories defined by visual assessment and prior knowledge of the disease. Four tissue health categories were initially, visually, identified: live tissue on healthy colonies (liveL), live tissue on diseased colonies (liveD), tissue adjacent to the lesion (transition), and the BBD lesion itself (disease); (Figure 2). Discriminant analysis of the spectral data revealed spectrally distinct tissue health groups, with clear separation between live tissue on colonies without BBD (liveL) and live tissue on colonies with BBD (liveD) (Figure 4 and Figure 5, Table 1). In contrast, the BBD lesions (disease) and adjacent live tissue (transition) exhibited overlapping spectral signatures (Figure 4).
The spectral separability of the two live tissue types, liveL and liveD, was evident in differences in absorption wells and reflectance peaks. Here, the spectral signature of liveL closely aligned with previous coral reflectance studies [24,26]. Before sampling liveD, the live tissue was assessed to ensure measurements did not include areas that may have been stressed, such as paling. This supported the comparison of liveD and liveL, two live tissue types that appeared similar, despite originating from colonies in different states of health. Previous studies have demonstrated that reduced absorption and reflectance features are linked to declines in endosymbiont density, where decreased pigment concentrations in Symbiodiniaceae lead to diminished absorption in the blue and red regions of the spectrum [22,24,25,33]. To capture representative measurements of live tissue on diseased colonies, liveD spectral reflectance samples were taken at locations progressively farther from the lesion margin. These results provided insight into the extent of stress experienced by the coral colony in response to BBD in tissue that appeared healthy. Based on our results, we asked the question, what might be driving the observed variation in reflectance for liveD?
The spectral differences in liveD may be linked to microbial processes associated with BBD. Microbial studies have shown that bacterial communities of BBD lesions are stratified and diverse in function, with cyanobacteria capable of migrating beneath the coral tissue ahead of the black band and sulfur-reducing bacteria near the skeleton producing compounds that are toxic to coral tissue [34,35,36]. Additionally, there may be successional stages in BBD development associated with shifts in different bacterial assemblages for Montipora corals. In a study conducted in the central Great Barrier Reef, Montipora corals exhibiting a cyanobacterial patch without a distinct black band progressed to active BBD lesions within one month in 19% of cases [37].
Although the transition tissue appeared visually similar to the liveD tissue, it was classified as a distinct category due to microbial processes known to cause sublethal damage in areas adjacent to the lesion [34,36]. Spectral signatures of both transition and disease were characterized by general flattening of the reflectance curve and diminished absorption features (Figure 3), patterns similar to those seen in bleached or dead coral [38]. What is interesting is that these two tissue types are functionally different: the BBD lesion is defined by a dark microbial mat with necrotic tissue, while the adjacent tissue (transition) appears visually healthy at the surface, indicating coral tissue pigment, and Symbiodiniaceae are abundant with no signs of paling. Interestingly, liveD spectra closest to lesions often resembled transition or disease, reinforcing the idea of physiological stress extending outward from the band. The integration of microbial analyses with spectral measurements could provide validation of this theory.
It is well established that various pigments, proteins, and compounds within the coral tissue, along with skeletal morphology, contribute to the shape and magnitude of spectral signatures. However, without complementary chemical or microbial data, the drivers of the spectral similarities and differences observed here remain unclear [33,39,40]. Spectroscopy has been widely applied in coral reef studies to distinguish broad benthic types, classify coral functional groups, and assess symbiont chemistry to predict thermally tolerant corals [25,41,42]. Reflectance studies on corals have largely focused on detecting stress or identifying healthy corals, where “stress” and “health” refer to bleaching—the loss of endosymbionts in response to elevated ocean temperatures [38,43,44]. In situ reflectance spectroscopy studies focused on coral disease are uncommon. One study [21] used hyperspectral reflectance in a controlled laboratory setting to study Caribbean Yellow Band Disease (CYBD) and bleaching of Orbicella faveolata. Spectral discrimination among coral disease states was demonstrated, where the health groups, CYBD, bleaching, and asymptomatic, were distinct. This research provided early evidence supporting the use of hyperspectral remote sensing to detect disease-related changes in coral optical properties, with implications for understanding the role of pigments in coral disease etiology [21]. Although these advances have been made, a significant knowledge gap remains in applying reflectance spectroscopy to coral disease, despite its potential to detect subtle, early-stage changes associated with coral health.

4.2. Recommendations

Our findings lay the groundwork for future studies to expand the application of reflectance spectroscopy as a tool for detecting black band disease in Montiporid corals, with a focus on Hawaiian corals. Early detection of disease is critical for coral resource management, supporting timely and spatially targeted responses. Whether through localized treatment efforts or quarantine strategies, such information enables practitioners to mitigate the risk of disease transmission to nearby colonies and prioritize conservation interventions more effectively. With emerging tools aimed at predicting coral disease outbreaks and the expansion of in situ and ex situ coral reef restoration efforts in Hawai‘i, early detection methods can play a critical role in supporting effective response and management strategies [30,45,46].
An example of small-scale management applications includes in-field coral health assessments at permanent monitoring sites in areas prone to disease, as well as coral conservation efforts that collect corals of opportunity for ex situ nursery care. Corals collected showing signs of disease can be screened to assess the extent of affected tissue, helping to determine how much live tissue should be removed. For corals that appear healthy, screening can identify subtle signs of stress and/or disease and these colonies can be further assessed through microbial tests to determine if they are fit for outplanting, i.e., disease-free. If disease treatments, such as topical ointments, are available, colonies suspected of being stressed (i.e., based on spectral reflectance) can be closely monitored and treated promptly at the first sign of lesion development.
The demonstrated spectral separability among coral tissue conditions associated with disease may serve as a prerequisite for assessing the feasibility of predictive models. Additional sampling is needed to produce statistically robust models capable of guiding management practices. In terrestrial systems, spectroscopy has successfully detected physiological stressors such as drought and disease in plants [23,24,27,47]. Applying a similar framework to marine ecosystems, coral disease prediction models with unknown classification may be achieved through the integration of in situ and nursery-based spectral and microbial data to link variations in physiological function due to disease with spectral traits [24,25].

5. Conclusions

This study highlights the value of using reflectance spectroscopy as a management tool for coral disease detection. While four tissue health categories were originally identified through visual inspection, spectral analysis revealed three spectrally distinct categories, with a misclassification rate of 12.8%. Discriminant analysis indicated that visually similar tissue types, live tissue on BBD-affected colonies (liveD) and live tissue on colonies without BBD (liveL), were spectrally distinct. Conversely, BBD lesions (disease) and adjacent tissue that appeared healthy (transition) exhibited similar spectral signatures. By enabling early identification of sublethal stress before visible lesions appear, spectroscopy offers managers the ability to prioritize colonies for monitoring, treatment, or restoration. Such applications could enhance rapid-response strategies in the field, improve screening protocols for nursery and outplant corals, and reduce the risk of disease spread across reefs. Given the projected rise in coral disease under climate change, incorporating spectroscopy into monitoring frameworks could enhance early detection, improve intervention response times, and support more effective coral reef conservation and resilience efforts.

Author Contributions

Conceptualization, M.B.M., R.E.M. and G.P.A.; methodology, M.B.M., R.E.M. and G.P.A.; formal analysis, M.B.M. and R.E.M.; investigation, M.B.M., R.E.M., M.A. and G.P.A.; resources, M.B.M., R.E.M. and G.P.A.; data curation, M.B.M. and R.E.M.; writing—original draft preparation, M.B.M. and R.E.M.; writing—review and editing, M.B.M., R.E.M., M.A. and G.P.A.; funding acquisition, R.E.M. and G.P.A. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Dorrance Family Foundation, grant number G-10759-300, and Hawai‘i Department of Land and Natural Resources, contract C43115.

Data Availability Statement

The data presented in this study are openly available in Zenodo at https://doi.org/10.5281/zenodo.16513875 (created 27 July 2025).

Acknowledgments

We acknowledge with gratitude Heather Ylitalo-Ward for her invaluable knowledge and continuous mentorship, and Nicholas Vaughn for his analytical expertise during this study, which have contributed to the success of this work. This paper is based on the following master’s thesis: Tierney, M.B.M., 2025. Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals. Master’s thesis, Arizona State University, Tempe, Arizona, USA. (https://hdl.handle.net/2286/R.2.N.201555, accessed on 1 July 2025).

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
BIBrightness index
BNBrightness normalization
BBDBlack band disease
CYBDCaribbean Yellow Band Disease
DARDivision of Aquatic Resources
LDALinear discriminant analysis

Appendix A

Remotesensing 17 03241 g0a1
Figure A1. Additional examples of measurements of Montiporid corals with (left column; a–c) and without black band disease (BBD) (right column; d–f). Colored circles show an example of where reflectance of disease (red), transition (orange), and liveD (blue), defined as live tissue on corals with BBD, as well as liveL (green) without disease, were sampled. Note: These example photos were taken before spectral collection without full illumination. A constant light source was used for spectral measurements to minimize any differences in illumination of the coral surface.
Figure A1. Additional examples of measurements of Montiporid corals with (left column; a–c) and without black band disease (BBD) (right column; d–f). Colored circles show an example of where reflectance of disease (red), transition (orange), and liveD (blue), defined as live tissue on corals with BBD, as well as liveL (green) without disease, were sampled. Note: These example photos were taken before spectral collection without full illumination. A constant light source was used for spectral measurements to minimize any differences in illumination of the coral surface.
Remotesensing 17 03241 g0a1
Remotesensing 17 03241 g0a2
Figure A2. Spectral reflectance (%) of live tissue on colonies without visible black band disease lesions (liveL), by species. Montipora capitata (spectra = 11; colonies = 2) in blue, and Montipora patula (spectra = 18, colonies = 3) in yellow. The mean spectral reflectance of each category is shown as a solid line. The shaded regions indicate one standard deviation.
Figure A2. Spectral reflectance (%) of live tissue on colonies without visible black band disease lesions (liveL), by species. Montipora capitata (spectra = 11; colonies = 2) in blue, and Montipora patula (spectra = 18, colonies = 3) in yellow. The mean spectral reflectance of each category is shown as a solid line. The shaded regions indicate one standard deviation.
Remotesensing 17 03241 g0a2
Remotesensing 17 03241 g0a3
Figure A3. Brightness-normalized (BN) reflectance of live tissue on colonies without black band disease (BBD) (liveL) by species. Montipora capitata (spectra = 11; colonies = 2) in blue, and Montipora patula (spectra = 18; colonies = 3) in yellow. The mean BN reflectance of each category is shown as the solid line. The shaded regions indicate one standard deviation.
Figure A3. Brightness-normalized (BN) reflectance of live tissue on colonies without black band disease (BBD) (liveL) by species. Montipora capitata (spectra = 11; colonies = 2) in blue, and Montipora patula (spectra = 18; colonies = 3) in yellow. The mean BN reflectance of each category is shown as the solid line. The shaded regions indicate one standard deviation.
Remotesensing 17 03241 g0a3
Remotesensing 17 03241 g0a4
Figure A4. Mean spectral reflectance (%) signatures, colored by the four tissue health categories, of 15 Montiporid colonies for disease in red (n = 49), transition in orange (n = 28), liveD in blue (n = 112), defined as live tissue on corals with a visible black band disease (BBD) lesion, and liveL in green (n = 29), defined as live tissue on corals without a visible BBD lesion. The shaded region is one standard deviation.
Figure A4. Mean spectral reflectance (%) signatures, colored by the four tissue health categories, of 15 Montiporid colonies for disease in red (n = 49), transition in orange (n = 28), liveD in blue (n = 112), defined as live tissue on corals with a visible black band disease (BBD) lesion, and liveL in green (n = 29), defined as live tissue on corals without a visible BBD lesion. The shaded region is one standard deviation.
Remotesensing 17 03241 g0a4

References

  1. Knowlton, N.; Brainard, R.E.; Fisher, R.; Moews, M.; Plaisance, L.; Caley, M.J. Coral Reef Biodiversity. In Life in the World’s Oceans; Wiley: Hoboken, NJ, USA, 2010; pp. 65–78. ISBN 978-1-4051-9297-2. [Google Scholar]
  2. Storlazzi, C.D.; Reguero, B.G.; Cole, A.D.; Lowe, E.; Shope, J.B.; Gibbs, A.E.; Nickel, B.A.; McCall, R.T.; van Dongeren, A.R.; Beck, M.W. Rigorously Valuing the Role of U.S. Coral Reefs in Coastal Hazard Risk Reduction; U.S. Geological Survey: Reston, VA, USA, 2019. [Google Scholar]
  3. Wild, C.; Hoegh-Guldberg, O.; Naumann, M.S.; Colombo-Pallotta, M.F.; Ateweberhan, M.; Fitt, W.K.; Iglesias-Prieto, R.; Palmer, C.; Bythell, J.C.; Ortiz, J.-C.; et al. Climate Change Impedes Scleractinian Corals as Primary Reef Ecosystem Engineers. Mar. Freshw. Res. 2011, 62, 205. [Google Scholar] [CrossRef]
  4. Good, A.M.; Bahr, K.D. The Coral Conservation Crisis: Interacting Local and Global Stressors Reduce Reef Resiliency and Create Challenges for Conservation Solutions. SN Appl. Sci. 2021, 3, 312. [Google Scholar] [CrossRef]
  5. Hoegh-Guldberg, O.; Mumby, P.J.; Hooten, A.J.; Steneck, R.S.; Greenfield, P.; Gomez, E.; Harvell, C.D.; Sale, P.F.; Edwards, A.J.; Caldeira, K.; et al. Coral Reefs under Rapid Climate Change and Ocean Acidification. Science 2007, 318, 1737–1742. [Google Scholar] [CrossRef]
  6. Bruckner, A. The Global Perspective of Incidence and Prevalence of Coral Diseases; NOAA Fisheries Coral Reef Conservation Program: Silver Spring, MD, USA, 2009. [Google Scholar]
  7. Hughes, T.P.; Barnes, M.L.; Bellwood, D.R.; Cinner, J.E.; Cumming, G.S.; Jackson, J.B.C.; Kleypas, J.; Van De Leemput, I.A.; Lough, J.M.; Morrison, T.H.; et al. Coral Reefs in the Anthropocene. Nature 2017, 546, 82–90. [Google Scholar] [CrossRef]
  8. Burke, S.; Pottier, P.; Lagisz, M.; Macartney, E.L.; Ainsworth, T.; Drobniak, S.M.; Nakagawa, S. The Impact of Rising Temperatures on the Prevalence of Coral Diseases and Its Predictability: A Global Meta-analysis. Ecol. Lett. 2023, 26, 1466–1481. [Google Scholar] [CrossRef]
  9. Maynard, J.; van Hooidonk, R.; Eakin, C.M.; Puotinen, M.; Garren, M.; Williams, G.; Heron, S.F.; Lamb, J.; Weil, E.; Willis, B.; et al. Projections of Climate Conditions That Increase Coral Disease Susceptibility and Pathogen Abundance and Virulence. Nat. Clim. Change 2015, 5, 688–694. [Google Scholar] [CrossRef]
  10. Morais, J.; Cardoso, A.P.L.R.; Santos, B.A. A Global Synthesis of the Current Knowledge on the Taxonomic and Geographic Distribution of Major Coral Diseases. Environ. Adv. 2022, 8, 100231. [Google Scholar] [CrossRef]
  11. Allen, M. (Kauaʻi Division of Aquatic Resources, Līhuʻe, HI, USA). Personal Communication, 2024.
  12. Aeby, G.S.; Work, T.M.; Runyon, C.M.; Shore-Maggio, A.; Ushijima, B.; Videau, P.; Beurmann, S.; Callahan, S.M. First Record of Black Band Disease in the Hawaiian Archipelago: Response, Outbreak Status, Virulence, and a Method of Treatment. PLoS ONE 2015, 10, e0120853. [Google Scholar] [CrossRef]
  13. Oberle, F.K.J.; Storlazzi, C.D.; Cheriton, O.M.; Takesue, R.K.; Hoover, D.J.; Logan, J.B.; Runyon, C.; Kellogg, C.A.; Johnson, C.D.; Swarzenski, P.W. Physicochemical Controls on Zones of Higher Coral Stress Where Black Band Disease Occurs at Mākua Reef, Kaua‘i, Hawai‘i. Front. Mar. Sci. 2019, 6, 552. [Google Scholar] [CrossRef]
  14. Runyon, C.; Aeby, G.S.; Callahan, S.M. Kauai Montipora Coral Disease Prevalence and Environmental Drivers; University of Hawai’i: Honolulu, HI, USA, 2015; pp. 1–10. [Google Scholar]
  15. Allen, M. Spatial Patterns and Environmental Drivers of Black-Band Disease (BBD) in Montiporid Corals Across Anini Reef, Kauai, HI. Master’s Thesis, Arizona State University, Tempe, AZ, USA, 2025. [Google Scholar]
  16. Apprill, A.; Girdhar, Y.; Mooney, T.A.; Hansel, C.M.; Long, M.H.; Liu, Y.; Zhang, W.G.; Kapit, J.; Hughen, K.; Coogan, J.; et al. Toward a New Era of Coral Reef Monitoring. Environ. Sci. Technol. 2023, 57, 5117–5124. [Google Scholar] [CrossRef]
  17. Combs, I.R.; Studivan, M.S.; Eckert, R.J.; Voss, J.D. Quantifying Impacts of Stony Coral Tissue Loss Disease on Corals in Southeast Florida through Surveys and 3D Photogrammetry. PLoS ONE 2021, 16, e0252593. [Google Scholar] [CrossRef]
  18. Jordán-Dahlgren, E.; Jordán-Garza, A.G.; Rodríguez-Martínez, R.E. Coral Disease Prevalence Estimation and Sampling Design. PeerJ 2018, 6, e6006. [Google Scholar] [CrossRef]
  19. Kuta, K.; Richardson, L. Ecological Aspects of Black Band Disease of Corals: Relationships between Disease Incidence and Environmental Factors. Coral Reefs 2002, 21, 393–398. [Google Scholar] [CrossRef]
  20. Sato, Y.; Bourne, D.G.; Willis, B.L. Dynamics of Seasonal Outbreaks of Black Band Disease in an Assemblage of Montipora Species at Pelorus Island (Great Barrier Reef, Australia). Proc. R. Soc. B Biol. Sci. 2009, 276, 2795–2803. [Google Scholar] [CrossRef]
  21. Anderson, D.A.; Armstrong, R.A.; Weil, E. Hyperspectral Sensing of Disease Stress in the Caribbean Reef-Building Coral, Orbicella Faveolata-Perspectives for the Field of Coral Disease Monitoring. PLoS ONE 2013, 8, e81478. [Google Scholar] [CrossRef] [PubMed]
  22. Hedley, J.D.; Mumby, P.J. Biological and Remote Sensing Perspectives of Pigmentation in Coral Reef Organisms. Adv. Mar. Biol. 2002, 43, 277–317. [Google Scholar] [PubMed]
  23. Weingarten, E.; Martin, R.E.; Hughes, R.F.; Vaughn, N.R.; Shafron, E.; Asner, G.P. Early Detection of a Tree Pathogen Using Airborne Remote Sensing. Ecol. Appl. 2022, 32, e2519. [Google Scholar] [CrossRef]
  24. Asner, G.P.; Drury, C.; Vaughn, N.R.; Hancock, J.R.; Martin, R.E. Variability in Symbiont Chlorophyll of Hawaiian Corals from Field and Airborne Spectroscopy. Remote Sens. 2024, 16, 732. [Google Scholar] [CrossRef]
  25. Drury, C.; Martin, R.E.; Knapp, D.E.; Heckler, J.; Levy, J.; Gates, R.D.; Asner, G.P. Ecosystem-scale Mapping of Coral Species and Thermal Tolerance. Front. Ecol. Environ. 2022, 20, 285–291. [Google Scholar] [CrossRef]
  26. Hochberg, E.J.; Atkinson, M.J.; Apprill, A.; Andrefouet, S. Spectral Reflectance of Coral. Coral Reefs 2004, 23, 84–95. [Google Scholar] [CrossRef]
  27. Sapes, G.; Schroeder, L.; Scott, A.; Clark, I.; Juzwik, J.; Montgomery, R.A.; Guzmán Q, J.A.; Cavender-Bares, J. Mechanistic Links between Physiology and Spectral Reflectance Enable Previsual Detection of Oak Wilt and Drought Stress. Proc. Natl. Acad. Sci. USA 2024, 121, e2316164121. [Google Scholar] [CrossRef] [PubMed]
  28. Zahir, S.A.D.M.; Omar, A.F.; Jamlos, M.F.; Azmi, M.A.M.; Muncan, J. A Review of Visible and Near-Infrared (Vis-NIR) Spectroscopy Application in Plant Stress Detection. Sens. Actuators Phys. 2022, 338, 113468. [Google Scholar] [CrossRef]
  29. Aeby, D.G.S.; Hutchinson, M.; MacGowan, P. Hawaii’s Rapid Response Contingency Plan; The Division of Aquatic Resources, Department of Land and Natural Resources: Honolulu, HI, USA, 2008.
  30. Caldwell, J.M.; Liu, G.; Geiger, E.; Heron, S.F.; Eakin, C.M.; De La Cour, J.; Greene, A.; Raymundo, L.; Dryden, J.; Schlaff, A.; et al. Multi-Factor Coral Disease Risk Forecasting for Early Warning and Management. Ecol. Appl. 2024, 34, e2961. [Google Scholar] [CrossRef] [PubMed]
  31. Feilhauer, H.; Asner, G.P.; Martin, R.E.; Schmidtlein, S. Brightness-Normalized Partial Least Squares Regression for Hyperspectral Data. J. Quant. Spectrosc. Radiat. Transf. 2010, 111, 1947–1957. [Google Scholar] [CrossRef]
  32. SAS Institute Inc. JMP® Pro, Version 18; SAS Institute: Cary, NC, USA, 2024.
  33. Wangpraseurt, D.; Larkum, A.W.D.; Ralph, P.J.; Kühl, M. Light Gradients and Optical Microniches in Coral Tissues. Front. Microbiol. 2012, 3, 316. [Google Scholar] [CrossRef]
  34. Carlton, R.G.; Richardson, L.L. Oxygen and Sulfide Dynamics in a Horizontally Migrating Cyanobacterial Mat: Black Band Disease of Corals. FEMS Microbiol. Ecol. 1995, 18, 155–162. [Google Scholar] [CrossRef]
  35. Miller, A.W.; Blackwelder, P.; Al-Sayegh, H.; Richardson, L.L. Insights into Migration and Development of Coral Black Band Disease Based on Fine Structure Analysis. Rev. Biol. Trop. 2015, 60, 21. [Google Scholar] [CrossRef]
  36. Miller, A.W.; Richardson, L.L. Fine Structure Analysis of Black Band Disease (BBD) Infected Coral and Coral Exposed to the BBD Toxins Microcystin and Sulfide. J. Invertebr. Pathol. 2012, 109, 27–33. [Google Scholar] [CrossRef]
  37. Sato, Y.; Willis, B.L.; Bourne, D.G. Successional Changes in Bacterial Communities during the Development of Black Band Disease on the Reef Coral, Montipora hispida. ISME J. 2010, 4, 203–214. [Google Scholar] [CrossRef]
  38. Holden, H.; Ledrew, E. Spectral Discrimination of Healthy and Non-Healthy Corals Based on Cluster Analysis, Principal Components, and Derivative Spectroscopy. Remote Sens. Environ. 1998, 65, 217–224. [Google Scholar] [CrossRef]
  39. Ferreira, G.; Bollati, E.; Kühl, M. The Role of Host Pigments in Coral Photobiology. Front. Mar. Sci. 2023, 10, 1204843. [Google Scholar] [CrossRef]
  40. Roth, M.S. The Engine of the Reef: Photobiology of the Coral Algal Symbiosis. Front. Microbiol. 2014, 5, 422. [Google Scholar] [CrossRef]
  41. Harrison, D.E.; Asner, G.P. Sensitivity of Spectral Communities to Shifts in Benthic Composition in Hawai’i. Remote Sens. Environ. 2024, 304, 114050. [Google Scholar] [CrossRef]
  42. Hochberg, E.J.; Atkinson, M.J. Spectral Discrimination of Coral Reef Benthic Communities. Coral Reefs 2000, 19, 164–171. [Google Scholar] [CrossRef]
  43. Teague, J.; Willans, J.; Megson-Smith, D.A.; Day, J.C.C.; Allen, M.J.; Scott, T.B. Using Colour as a Marker for Coral ‘Health’: A Study on Hyperspectral Reflectance and Fluorescence Imaging of Thermally Induced Coral Bleaching. Oceans 2022, 3, 547–556. [Google Scholar] [CrossRef]
  44. Yamano, H.; Tamura, M.; Kunii, Y.; Hidaka, M. Spectral Reflectance as a Potential Tool for Detecting Stressed Corals. J. Jpn. Coral Reef Soc. 2003, 5, 1–10. [Google Scholar] [CrossRef]
  45. Eaton, K.R.; Clark, A.S.; Curtis, K.; Favero, M.; Hanna Holloway, N.; Ewen, K.; Muller, E.M. A Highly Effective Therapeutic Ointment for Treating Corals with Black Band Disease. PLoS ONE 2022, 17, e0276902. [Google Scholar] [CrossRef]
  46. ‘Āko‘ako‘a: Championing Coral Reef Restoration in Hawai‘i. Available online: https://www.akoakoa.org (accessed on 19 July 2025).
  47. Sankaran, S.; Mishra, A.; Ehsani, R.; Davis, C. A Review of Advanced Techniques for Detecting Plant Diseases. Comput. Electron. Agric. 2010, 72, 1–13. [Google Scholar] [CrossRef]
Remotesensing 17 03241 g001
Figure 1. Regional context of the study site on ‘Anini reef, located along the north shore of Kaua‘i Island, Hawai‘i (A). Zoom of ‘Anini reef (B): red boxes denote survey areas on ‘Anini reef where spectral reflectance measurements of Montipora capitata and Montipora patula corals, with and without black band disease, were collected.
Figure 1. Regional context of the study site on ‘Anini reef, located along the north shore of Kaua‘i Island, Hawai‘i (A). Zoom of ‘Anini reef (B): red boxes denote survey areas on ‘Anini reef where spectral reflectance measurements of Montipora capitata and Montipora patula corals, with and without black band disease, were collected.
Remotesensing 17 03241 g001
Remotesensing 17 03241 g002
Figure 2. Example measurements of Montiporid corals with (A) and without black band disease (BBD) (B). (A) The table shows the number of spectral reflectance measurements per tissue health category, the total number of spectral reflectance measurements (n = 189), and the total number of colonies = 10, sampled from Montiporid colonies with a visible BBD lesion. Number of colonies sampled by species: Montipora capitata = 8; Montipora patula = 2—image of Montipora capitata with BBD at ‘Anini reef, Kaua‘i. Colored circles show an example of where the reflectance of disease (red), transition (orange), and liveD (blue), defined as live tissue on corals with BBD, was sampled. (B) The table shows the number of spectral reflectance measurements of liveL (green) (n = 29), defined as live tissue on corals without a visible BBD lesion, and the total number of colonies = 5 sampled from. Number of colonies sampled by species: Montipora capitata = 2; Montipora patula = 3—image of Montipora patula without BBD at ‘Anini reef, Kaua‘i. Green-colored circles exemplify where liveL samples were collected. The scale bar is used to reference the size of the lesion or colony. The distance between colored circles and the size of circles are not to scale.
Figure 2. Example measurements of Montiporid corals with (A) and without black band disease (BBD) (B). (A) The table shows the number of spectral reflectance measurements per tissue health category, the total number of spectral reflectance measurements (n = 189), and the total number of colonies = 10, sampled from Montiporid colonies with a visible BBD lesion. Number of colonies sampled by species: Montipora capitata = 8; Montipora patula = 2—image of Montipora capitata with BBD at ‘Anini reef, Kaua‘i. Colored circles show an example of where the reflectance of disease (red), transition (orange), and liveD (blue), defined as live tissue on corals with BBD, was sampled. (B) The table shows the number of spectral reflectance measurements of liveL (green) (n = 29), defined as live tissue on corals without a visible BBD lesion, and the total number of colonies = 5 sampled from. Number of colonies sampled by species: Montipora capitata = 2; Montipora patula = 3—image of Montipora patula without BBD at ‘Anini reef, Kaua‘i. Green-colored circles exemplify where liveL samples were collected. The scale bar is used to reference the size of the lesion or colony. The distance between colored circles and the size of circles are not to scale.
Remotesensing 17 03241 g002
Remotesensing 17 03241 g003
Figure 3. Brightness-normalized (BN) spectral reflectance signatures, colored by the four tissue health categories of 15 Montiporid colonies. Disease (n = 49), transition (n = 28), liveD (n = 112), defined as live tissue on corals with a visible black band disease (BBD) lesion, liveL (n = 29), defined as live tissue on corals without a visible BBD lesion. The mean BN reflectance is shown as the solid line. The shaded regions represent one standard deviation. Dotted lines indicate wavelengths of interest: 555, 580, 605, 650, and 670 nm.
Figure 3. Brightness-normalized (BN) spectral reflectance signatures, colored by the four tissue health categories of 15 Montiporid colonies. Disease (n = 49), transition (n = 28), liveD (n = 112), defined as live tissue on corals with a visible black band disease (BBD) lesion, liveL (n = 29), defined as live tissue on corals without a visible BBD lesion. The mean BN reflectance is shown as the solid line. The shaded regions represent one standard deviation. Dotted lines indicate wavelengths of interest: 555, 580, 605, 650, and 670 nm.
Remotesensing 17 03241 g003
Remotesensing 17 03241 g004
Figure 4. Canonical plot from LDA of all spectral reflectance samples (n = 218) from 15 Montiporid colonies. The first and second canonical axes explain 76.97% and 21.91% of the variation, respectively. Points represent spectral reflectance samples and are colored by tissue health. The solid ellipse is the 50% prediction region, and the inner dashed ellipse is the 95% confidence region for the mean of each tissue health category.
Figure 4. Canonical plot from LDA of all spectral reflectance samples (n = 218) from 15 Montiporid colonies. The first and second canonical axes explain 76.97% and 21.91% of the variation, respectively. Points represent spectral reflectance samples and are colored by tissue health. The solid ellipse is the 50% prediction region, and the inner dashed ellipse is the 95% confidence region for the mean of each tissue health category.
Remotesensing 17 03241 g004
Remotesensing 17 03241 g005
Figure 5. Canonical plot from LDA of all spectral reflectance samples (n = 218) from 15 Montiporid colonies. Canonical 1 explains 78.2% of the variance, Canonical 2 explains an additional 21.7% of the variance. Points represent spectral reflectance samples and are colored by tissue health. The solid ellipse is estimated to contain 50% of the population, and the inner dashed ellipse is the 95% confidence region for the mean of each tissue health category.
Figure 5. Canonical plot from LDA of all spectral reflectance samples (n = 218) from 15 Montiporid colonies. Canonical 1 explains 78.2% of the variance, Canonical 2 explains an additional 21.7% of the variance. Points represent spectral reflectance samples and are colored by tissue health. The solid ellipse is estimated to contain 50% of the population, and the inner dashed ellipse is the 95% confidence region for the mean of each tissue health category.
Remotesensing 17 03241 g005
Table 1. Linear discriminant analysis (LDA) score summary and confusion matrix classification results from brightness-normalized spectra for the four tissue health categories from 15 colonies. Columns represent the predicted classification rate, rows are the actual classification, and diagonal values represent correct classification. LDA score summary and confusion matrix results for analysis of three tissue health categories (infected, liveD, liveL) are indicated by “*”.
Table 1. Linear discriminant analysis (LDA) score summary and confusion matrix classification results from brightness-normalized spectra for the four tissue health categories from 15 colonies. Columns represent the predicted classification rate, rows are the actual classification, and diagonal values represent correct classification. LDA score summary and confusion matrix results for analysis of three tissue health categories (infected, liveD, liveL) are indicated by “*”.
SourceCount (n)Number MisclassifiedPercent MisclassifiedEntropy R2
Exploratory2185826.610.38
28 *12.84 *0.50 *
ActualPredicted Rate
Count (n)Four CategoriesDiseaseTransitionLiveDLiveL
49Disease0.670.330.000.00
28Transition0.360.500.140.00
112LiveD0.070.080.770.08
29LiveL0.000.000.070.93
Three CategoriesInfectedLiveDLiveL
77Infected0.94 *,10.06 *,10.00 *
112LiveD0.11 *0.81 *0.08 *
29LiveL0.00 *0.07 *0.93 *
1 Values are rounded to two decimal places, except where the sum of the row exceeds 1; in this case, the largest value in the row was rounded up, and the smallest value was left unchanged.
Table 2. Canonical details from linear discriminant analysis (LDA) on brightness-normalized spectral reflectance of four tissue health categories: disease, transition, liveD, and liveL. Calculated from the overall pooled within-group covariance matrix. The asterisk next to the p-value indicates statistical significance at p < 0.05. Abbreviations are as follows: Cum Percent = cumulative sum of the proportions; Canonical Corr = canonical correlations between the covariates and the groups defined by the categorical X; Approx. F = approximate F value; NumDF = numerator degrees of freedom; DenDF = denominator degrees of freedom; Prob > F = p-value.
Table 2. Canonical details from linear discriminant analysis (LDA) on brightness-normalized spectral reflectance of four tissue health categories: disease, transition, liveD, and liveL. Calculated from the overall pooled within-group covariance matrix. The asterisk next to the p-value indicates statistical significance at p < 0.05. Abbreviations are as follows: Cum Percent = cumulative sum of the proportions; Canonical Corr = canonical correlations between the covariates and the groups defined by the categorical X; Approx. F = approximate F value; NumDF = numerator degrees of freedom; DenDF = denominator degrees of freedom; Prob > F = p-value.
EigenvaluePercentCum PercentCanonical CorrLikelihood RatioApprox. FNumDFDenDFProb > F
3.1176.9776.970.870.1210.4557585.24<0.001 *
0.8921.9198.880.690.514.4236394<0.001 *
TestValueApprox. FNumDFDenDFProb > F
Wilks’ Lambda0.1210.4557585.24<0.001 *
Table 3. Canonical details from linear discriminant analysis (LDA) on brightness-normalized spectral reflectance of three tissue health categories: infected, liveD, and liveL. Calculated from the overall pooled within-group covariance matrix. The asterisk next to the p-value indicates statistical significance at p < 0.05. Abbreviations are as follows: Cum Percent = cumulative sum of the proportions; Canonical Corr = canonical correlations between the covariates and the groups defined by the categorical X; Approx. F = approximate F value; NumDF = numerator degrees of freedom; DenDF = denominator degrees of freedom; Prob > F = p-value.
Table 3. Canonical details from linear discriminant analysis (LDA) on brightness-normalized spectral reflectance of three tissue health categories: infected, liveD, and liveL. Calculated from the overall pooled within-group covariance matrix. The asterisk next to the p-value indicates statistical significance at p < 0.05. Abbreviations are as follows: Cum Percent = cumulative sum of the proportions; Canonical Corr = canonical correlations between the covariates and the groups defined by the categorical X; Approx. F = approximate F value; NumDF = numerator degrees of freedom; DenDF = denominator degrees of freedom; Prob > F = p-value.
EigenvaluePercentCum PercentCanonical CorrLikelihood RatioApprox. FNumDFDenDFProb > F
3.0678.2278.220.870.1318.0738394<0.001 *
0.8521.77100.000.680.549.3818198<0.001 *
TestValueApprox. FNumDFDenDFProb > F
Wilks’ Lambda0.1318.0738394<0.001 *
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Melamed, M.B.; Martin, R.E.; Allen, M.; Asner, G.P. Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals. Remote Sens. 2025, 17, 3241. https://doi.org/10.3390/rs17183241

AMA Style

Melamed MB, Martin RE, Allen M, Asner GP. Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals. Remote Sensing. 2025; 17(18):3241. https://doi.org/10.3390/rs17183241

Chicago/Turabian Style

Melamed, Mia B., Roberta E. Martin, McKenna Allen, and Gregory P. Asner. 2025. "Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals" Remote Sensing 17, no. 18: 3241. https://doi.org/10.3390/rs17183241

APA Style

Melamed, M. B., Martin, R. E., Allen, M., & Asner, G. P. (2025). Reflections: Spectral Investigation of Black Band Disease in Hawaiian Corals. Remote Sensing, 17(18), 3241. https://doi.org/10.3390/rs17183241

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop