Localized Impacts of Hurricane Irma on Diadema antillarum and Coral Reef Community Structure

Abstract

Strong physical disturbance from hurricanes can disrupt coral reef ecosystems and precipitate a regime shift toward algal dominance, particularly in the absence of grazing pressure to regulate algal growth post-storm. Here, we examine the influence of Hurricane Irma on a keystone grazer, Diadema antillarum, and the surrounding coral reef benthic community in the Florida Keys. D. antillarum densities and test diameters, as well as percent cover of coral reef benthic groups, were measured at 10 sites in the middle and upper Keys before and after Irma. Significant decreases in mean D. antillarum density and median test diameter were observed following the storm. There was a correlation between the magnitude of decline in D. antillarum density and the magnitude of sediment deposition on reefs, suggesting that abrasion or burial from sediment transport may have contributed to D. antillarum mortality. We detected significant decreases in the percent cover of sponges and hydrocorals following the storm, but no change in scleractinian coral cover, which was very low (3% mean cover) at the onset of the study. Macroalgal cover increased at sites in the upper Keys and decreased at sites in the middle Keys. There was no relationship between post-storm D. antillarum density and the change in percent cover of macroalgae or turf-algal-sediment matrix (TAS), likely due to low overall abundance of the grazer. We predict that coral reefs will remain in an algal-dominated ecosystem state due to, among other factors, increasing frequency of strong hurricanes that impact the D. antillarum population.

Introduction

Abrupt and persistent changes in ecosystem structure and function, or regime shifts, are becoming increasingly common in ocean ecosystems due to intensifying anthropogenic stressors that diminish ecosystem resiliency, such as overharvesting of marine resources, climate change, and pollution (Nyström et al. 2000; Folke et al. 2004). The resiliency of many coastal ecosystems, including kelp forests (Estes and Palmisano 1974), seagrass beds (Gunderson 2001), and coral reefs (Hughes 1994) has been linked to the intensity of herbivory, with regime shifts occurring following changes in the abundance of herbivore populations (Bellwood et al. 2004). These coastal ecosystems can exhibit multiple stable equilibria, or alternative stable states, maintained by a suite of feedback mechanisms that make regime shifts difficult to reverse (Scheffer et al. 2001; Beisner et al. 2003).

On Caribbean coral reefs, intense grazing pressure by diverse assemblages of herbivorous fishes and invertebrates regulates algal abundance and stabilizes a coral-dominated ecosystem state by indirectly facilitating coral settlement (Edmunds and Carpenter 2001; Hughes et al. 2007; Mumby et al. 2007b), growth (Lirman 2001; Idjadi et al. 2010), and survival (Hughes et al. 2007). Following historical overharvesting of herbivorous fishes, the long-spined sea urchin Diadema antillarum became one of the last abundant grazers on Caribbean reefs (Hughes 1994). This loss of functional redundancy decreased ecosystem resiliency and precipitated a regime shift to algal dominance following a disease-induced mass mortality of D. antillarum in 1983–1984 (Hughes et al. 1994; Nyström et al. 2000). Regime shifts on Caribbean reefs from a coral-dominated to an algal-dominated ecosystem state also have been attributed to declines in coral health and abundance caused by bleaching, disease, and hurricanes (Hughes et al. 1994; Knowlton 1992; Jackson et al. 2014).

The 1983–1984 mass mortality event reduced D. antillarum population abundance by two orders of magnitude (93–100% decline) throughout the Caribbean (Lessios 1988). Most D. antillarum populations have shown very limited recovery and have not returned to pre-mass mortality levels (Lessios 2016). The feeding preferences and grazing behavior of D. antillarum make the species particularly effective at maintaining the coral-dominated state (Adam et al. 2015). Accordingly, on reefs where D. antillarum populations are recovering, algal cover has decreased while coral recruitment, growth, and survival have increased, and a reversal to a coral-dominated community has been initiated (Edmunds and Carpenter 2001; Myhre and Acevedo-Gutiérrez 2007; Furman and Heck 2009; Idjadi et al. 2010). D. antillarum has yet to recover in the Florida Keys, potentially due to recurrent disease outbreaks (Forcucci 1994), high mortality at early life-history stages (Miller et al. 2009; Feehan et al. 2019), low reproductive success (Feehan et al. 2016), lack of suitable refuge habitat from predators, and physical disturbances such as strong storms (Chiappone et al. 2013).

Hurricanes are acute perturbations that can temporarily disrupt coral reef community structure through dislodgement, abrasion, and burial of benthic organisms from intense wave energy, resulting in mortality (Woodley et al. 1981). Although hurricanes are part of a natural disturbance regime on coral reefs, human activities have changed the capacity of the reef ecosystem to recover from hurricane disturbance (Connell 1978; Nyström et al. 2000). The disturbance regime itself is also being altered, since the frequency of high-intensity hurricanes is predicted to increase due to anthropogenic climate change (Knutson et al. 2010; Cheal et al. 2017). Post-hurricane recovery to a coral-dominated state can be facilitated by grazers regulating algal growth in newly opened spaces, but herbivore abundance may also be impacted by hurricane disturbance through direct (e.g., mortality) or indirect (e.g., reduced refuge habitat) storm effects (Bellwood et al. 2004; Mumby et al. 2006).

In the Caribbean, the impact of hurricanes on D. antillarum can be unpredictable. Local D. antillarum population density has been documented to decrease (Woodley et al. 1981) or remain stable (Aronson 1993; Jorgensen et al. 2008) following hurricane disturbance, and the factors driving these varied responses are not clear. Prior to the Caribbean-wide D. antillarum mass mortality event in 1983–1984, Hurricane Allen in 1980 reduced D. antillarum density from 7.98 to 3.85 urchin m−2 on shallow (5–8 m) but not deep (10–20 m) Jamaican fore reefs (a more wave-exposed habitat relative to the back reef) within 2-months post-storm (Woodley et al. 1981). In contrast, persistence of D. antillarum on shallow (1–3 m) back reefs was documented 2.5 months after Hurricane Dean in Southern Mexico (7.29 urchin m−2 pre-storm and 12.6 urchin m−2 post-storm) (Jorgensen et al. 2008). D. antillarum densities on shallow (1–2 m) back reefs were also unchanged following Hurricane Gilbert (1988, Jamaica) and Hurricane Hugo (1989, U.S. Virgin Islands) (Aronson 1993). Since Hurricanes Allen, Dean, and Hugo were all extremely powerful (Category 4–5) hurricanes with similar potential to impact coral reef communities, discrepancies in storm impacts on D. antillarum may be attributed to local reef characteristics (e.g., depth and wave exposure). Additionally, physical structures on reefs dissipate wave energy and can minimize hurricane disturbance (Steiner 2003); thus, reef three-dimensional structure may be important in determining localized impacts on marine organisms. Studies at local scales will therefore be important for understanding hurricane impacts on D. antillarum in order to predict and manage Caribbean coral reefs given the importance of this grazer in mediating coral-algal regime shifts.

Here, we examine the initial (within 2.5 months) impacts on D. antillarum and the surrounding reef community of Hurricane Irma, a powerful Category 4 storm that made landfall in the lower Florida Keys on 10 September 2017 (Cangialosi et al. 2018). We test the hypothesis that disturbance from Hurricane Irma has the potential to reduce the already low-density D. antillarum population on degraded coral reefs in the Florida Keys. These reefs have exceptionally low three-dimensional structure due to low coral cover (Schutte et al. 2010). We expected that large D. antillarum would suffer from a lack of coral habitat more so than small D. antillarum, which are able to hide in smaller reef crevices (Randall et al. 1964), and we therefore predicted that the size structure of the population would shift toward smaller individuals following the storm. Hurricane impacts on benthic communities can vary with proximity to the storm (Woodley et al. 1981), and we expected that reefs closer to the eye of Irma (middle Keys) would be more damaged than reefs farther from Irma’s path (upper Keys). Since sediment abrasion is a potential source of hurricane-induced mortality for benthic organisms such as corals and sponges (Woodley et al. 1981), we predicted that D. antillarum decline would be greatest at locations with higher magnitudes of sediment transport. We also predicted that greater algal colonization would occur in locations of lower post-storm D. antillarum abundance, given the role of D. antillarum as a keystone grazer. To test these predictions, we examined D. antillarum density and size structure, and coral reef community structure before and after the storm at 10 sites spanning the middle and upper Florida Keys.

Methods

Study Sites and Experimental Design

The impacts of Hurricane Irma on D. antillarum and the surrounding benthic community were assessed at a total of 10 inshore patch reef, rubble reef flat, and offshore bank-barrier reef sites, spanning 3—6-m maximum depth and 80 km (linear distance) of the middle (n = 7 sites: DS, EFM, EW, TP, TR, WS, WT) and the upper (n = 3 sites: PP, PR, PS) Florida Keys (Table 1; Fig. 1). Sites were sampled 1 to 3 months prior to Hurricane Irma (10–15 June and 2–3 August 2017) as part of a separate study examining recovery of D. antillarum following historical disease-induced mass mortality events (C.J. Feehan unpublished data). To assess storm impacts, all sites were resampled 2.5 months (21–29 November 2017) after the storm, aside from a single location in the middle Keys (WT) that was sampled 4 months after the storm (9–11 January 2018) due to poor visibility in November (Fig. 1). We employed a stratified random sampling design with 2 belt transects (60 × 2 m) placed in the east-west direction ~ 10 m apart at each site. Permanent transects were not fixed to the substratum for repeat sampling, but site landmarks were used to place the transects in the same general location on the reef during each sampling period. Fewer sites were sampled in the upper Keys than in the middle Keys due to logistical constraints of transporting divers and gear to the upper Keys from the Florida Wildlife Research Institute in the middle Keys.

Table 1 Information about 10 sites sampled in the middle and upper Florida Keys, including site abbreviation, maximum depth, location, and GPS coordinates.
Fig. 1
figure1

Map of the Florida Keys region (a) with inset box (b) showing 10 sites in the middle (red triangles) and upper (red diamonds) Keys sampled to assess storm impacts on Diadema antillarum and the surrounding coral reef community. Color contours indicate significant wave heights (SWH, m) from WAVEWATCH III for the 3-h period during which Hurricane Irma made landfall (WW3DG et al. 2016). The solid black lines in (a) and (b) indicate the hurricane track

Physical Oceanographic Effects of the Storm

We examined time series of bottom ocean temperatures at 3 sites spanning the study region (EW, TR, and PP; Fig. 1) as an indicator of ocean mixing caused by the passage of Hurricane Irma. Based on the storm track, we expected greater physical impacts in the middle versus upper Keys. Temperature data were collected at hourly intervals with HOBO Pendant® Temperature/Light 8K data loggers (Onset Computer) attached to concrete blocks on sand bottom adjacent to the reef at each site. Mean daily temperatures were plotted over a 6-week period from 1 September to 13 October 2017 encompassing landfall of the hurricane.

To examine Hurricane Irma’s impact on ocean wave heights, significant wave heights (mean height of highest one third of waves) in the study region were examined. Data were extracted from the National Oceanographic and Atmospheric Administration (NOAA) wave model hindcasts for the 3-h period during which Hurricane Irma made landfall in the Florida Keys (http://polar.ncep.noaa.gov/waves/index2.shtml). The wave hindcasts were performed with the third generation WAVEWATCH III spectral wave model (WW3DG et al. 2016). Data from the US East Coast grid used in this study had a 4-arc minute spatial resolution. Wind forcing was from the National Weather Service (NWS) Global Forecasting System (GFS) analysis winds. Conclusions drawn from the wave height model are limited by low spatial resolution (4-arc minute). Thus, the map generated from the model can be used to assess only general east-west differences in wave heights.

Sea Urchin Density and Size Structure

D. antillarum density (urchin m−2) was estimated before and after Hurricane Irma by divers with SCUBA thoroughly searching belt transects (120 m2 per transect) for individuals under coral heads and reef overhangs, amongst macroalgae, and inside rubble and substrate holes (see the “Study Sites and Experimental Design” section). All individuals observed within the belt transects were counted and test diameters were measured with long-jaw calipers to the nearest 1.0 mm. D. antillarum population size structure was examined as size frequency distributions of test diameters before and after Hurricane Irma from individuals observed within the belt transects, in addition to individuals encountered opportunistically adjacent to the transects (to increase sample size for size frequency distributions at low sea urchin abundance). Only counts of individuals within the belt transects were used to determine D. antillarum density. Urchins less than 10-mm test diameter size are cryptic and difficult to find (Hunte and Younglao 1988) and therefore may be underreported.

Benthic Community Structure

To examine the impacts of Hurricane Irma on benthic community structure, video was captured along the same belt transects used for D. antillarum surveys (see the “Sea Urchin Density and Size Structure” section) with a GoPro HERO5 camera held by divers swimming approximately 1 m above the bottom. Still images were extracted from each video at a rate of 0.2 images per second to avoid sampling the same area of the reef twice. Using randomly generated numbers, 20 images from each image sequence (out of 30–40 images total) were selected for point count analysis in ImageJ (NIH) to estimate average percent cover of benthic organisms within a transect (Aronson et al. 1994). A standardized grid of 50 points was overlaid on each image, and the type of benthic group under each point was categorized. Each image was standardized to an area of approximately 0.7 m × 2.0 m using the transect tape for scale. Benthic groups included hydrocorals, scleractinian corals, dead coral skeleton, sponges, coralline algae, non-coralline calcareous and fleshy brown (e.g., Dictyota sp.), green (e.g., Halimeda sp.), and red (e.g., Wrangelia sp.) macroalgae (hereafter “macroalgae”), turf-algal-sediment matrix (hereafter TAS), sediment, and unknown (e.g., due to an obscured portion of the image). Average percent cover of each benthic group ((number of points overlying a group/50 total points) × 100) for the analyzed images was calculated for each transect (n = 20 images per transect). Video was captured on only one belt transect at WS in the middle Keys after Hurricane Irma due to poor visibility (Fig. 1).

Statistical Analysis

To examine the effect of Hurricane Irma on D. antillarum density (urchin m−2), 1-way analysis of variance (ANOVA) was used with “Storm” (fixed factor, 2 levels: before and after the storm) as the independent grouping variable. Sites where D. antillarum were absent both before and after Hurricane Irma were excluded from the analysis to minimize zero values in the dataset (DS, TR, and PP; Fig. 1). The effects of “Location” (upper and middle Keys) and potential “Storm” by “Location” interactions were not tested due to the limited sample size in the upper Keys (n = 2 sites) after the removal of zero-urchin sites from the analysis. The two transects sampled within each site were pooled for the analysis.

To examine whether size frequency distributions of D. antillarum test diameters (pooled among sites) differed before (n = 98 urchins) and after (n = 25 urchins) Hurricane Irma, we used the Kolmogorov-Smirnov (K-S) 2-sample test. Test diameter data was not normally distributed (Shapiro-Wilks, p < 0.001), so a non-parametric Mann-Whitney U test was used to evaluate differences in median test diameter before and after the storm.

Changes in coral reef community structure following Hurricane Irma were evaluated with 2-way multivariate analysis of variance (MANOVA) on the percent cover of community groups with “Storm” (fixed factor, 2 levels: before and after storm) and “Location” (fixed factor, 2 levels: upper and middle Keys) as independent grouping variables. The unknown category and rare community groups (< 1% cover), including coralline algae (mean percent cover = 0.38%) and dead coral skeleton (mean percent cover = 0.37%), were excluded from the analysis to minimize zero values to meet the assumptions of MANOVA (homogeneity of variances), and because these groups did not contribute substantially to benthic community structure. Following detection of a marginally non-significant interaction in MANOVA (see the “Results” section), univariate tests with 2-way ANOVA were conducted for each community group separately with “Storm” (fixed factor, 2 levels: before and after storm) and “Location” (fixed factor, 2 levels: upper and middle Keys) as independent grouping variables. The two transects sampled within each site were pooled for the analyses.

To test for a correlation between sediment transport and D. antillarum mortality, we evaluated the relationship between the change in the percent cover of sediment (difference between pre-Irma and post-Irma percent cover) and change in D. antillarum density (difference between pre-Irma and post-Irma density) with a non-parametric Spearman’s rank correlation, with transects pooled within sites. Sites where D. antillarum were absent before and after Hurricane Irma were excluded from the analysis, given that there was no change in density at these sites (DS, TR, and PP; Fig. 1). To test for a correlation between grazer abundance and post-storm algal colonization, Spearman’s rank correlations were also used to investigate the relationship between post-storm D. antillarum density (urchin m−2) and change in percent cover of macroalgae and TAS, with transects pooled within sites.

Percent cover data were normalized with arcsine transformation prior to the analyses, and normality of transformed data was confirmed using a Shapiro-Wilk test for normality on each community group. The assumption of homogeneity of variances was tested on the transformed percent cover data with a Bartlett test and no violations were detected. All statistical analyses were conducted in R, using the packages “dplyr” (Wickham et al. 2018) and “car” (Fox and Weisberg 2019).

Results

Physical Impacts of Hurricane Irma

Ocean mixing caused by the passage of Hurricane Irma was indicated by a rapid (within 24 h) 3 °C drop in bottom sea temperature, which was consistent at 3 sites spanning the middle and upper Florida Keys (Fig. 2). Significant wave heights during landfall of Hurricane Irma from the WAVEWATCH III model indicate that the eye of Hurricane Irma passed directly over the study region, with no clear east-west pattern in wave heights among study sites and similar magnitudes of waves in the upper and middle Keys (Fig. 1).

Fig. 2
figure2

Mean daily bottom sea temperature (°C) at 3 sites in the Florida Keys over 6 weeks from 1 September to 13 October 2017. The vertical dotted line indicates when Hurricane Irma made landfall in the Florida Keys. See Fig. 1 for site locations

Sea Urchin Density and Size Structure

There was a significant difference in mean D. antillarum density before and after Hurricane Irma (1-way ANOVA: F1,12 = 8.08, p = 0.015). Overall mean density of D. antillarum declined from 0.052 to 0.014 urchin m−2 (Fig. 3).

Fig. 3
figure3

Mean density (urchin m−2; + SE) of Diadema antillarum before and after Hurricane Irma at 7 sites where D. antillarum were present before the storm

Size frequency distributions of test diameters were significantly different before and after Hurricane Irma (K-S test: p = 0.003). D. antillarum population size structure was dominated by large individuals with 60–80 mm test diameter before Hurricane Irma, but shifted to dominance by smaller individuals with 30–40 mm test diameter after the storm (Fig. 4). Median test diameter was significantly reduced from 65 mm before the storm to 35 mm after the storm (Mann-Whitney U test: p = 0.0012).

Fig. 4
figure4

Size frequency distributions of Diadema antillarum test diameters (mm) at 7 sites in the Florida Keys before (top) and after (bottom) Hurricane Irma (n = 98 urchins pre-Irma, n = 25 urchins post-Irma). A bin size of 5 mm was used to group test diameters

Benthic Community Structure

Prior to Hurricane Irma, the benthic community at sites in both the middle and upper Keys was dominated by TAS, hydrocorals, sediment, and macroalgae (Fig. 5). We observed a marginally non-significant interaction between “Storm” (before and after storm) and “Location” (upper and middle Keys) on benthic community structure (2-way MANOVA: F1,16 = 2.43, p = 0.095; Table 2), a marginally non-significant effect of “Location” (2-way MANOVA: F1,16 = 2.43, p = 0.096), and no effect of “Storm” (2-way MANOVA: F1,16 = 1.56, p = 0.25). 2-way ANOVAs indicated that the marginally non-significant interaction was driven by differences in storm impacts on macroalgae in the middle versus upper Florida Keys (2-way ANOVA, Storm × Location: F1,10 = 7.99, p = 0.012; Table 3). Examining these locations separately with an a posteriori ANOVA indicates that there was a significant decrease in macroalgae in the middle Keys (12.6 ± 5.6% pre-Irma, 6.5 ± 4.8% post-Irma; ANOVA: F1,12 = 4.8, p = 0.049) and a marginally non-significant increase in macroalgae in the upper Keys (8.5 ± 2.2% pre-Irma, 17.9 ± 7.0% post-Irma: ANOVA: F1,4 = 5.87, p = 0.07) (Table 3, Fig. 5).

Fig. 5
figure5

Percent cover of benthic community groups (%) before and after Hurricane Irma at sites pooled in the middle and upper Florida Keys (n = 7 and 3 sites, respectively). “TAS” is turf-algal-sediment matrix. “Unknown” indicates that the community group could not be identified

Table 2 Two-way multivariate analysis of variance (MANOVA) of the effect of “Storm” (fixed factor, 2 levels: before and after Hurricane Irma), “Location” (fixed factor, 2 levels: upper and middle Keys), and the interaction between “Storm” and “Location” on percent cover (%) of benthic community groups (n = 7 sites middle Keys, 3 sites upper Keys)
Table 3 Pre-Irma and post-Irma mean percent cover (%, ± SD) of benthic community groups in the middle and upper Florida Keys. Also shown are the results of univariate 2-way analysis of variance (ANOVA) of the effect of “Storm” (fixed factor, 2 levels: before and after Hurricane Irma), “Location” (fixed factor, 2 levels: upper and middle Keys), and the interaction between “Storm” and “Location” on percent cover (%) for each community group (n = 7 sites middle Keys, 3 sites upper Keys). Italicized values are significant at α = 0.05

There was no significant interaction between “Storm” and “Location” in the univariate 2-way ANOVA tests for any other community group (Table 3). For hydrocorals, there was a significant decrease in percent cover in both the middle (28.4 ± 12.7 % pre-Irma, 17.4 ± 13.4% post-Irma) and upper (24.2 ± 15.9% pre-Irma, 10.1 ± 7.4% post-Irma) Keys following the storm (ANOVA: F1,16 = 4.87, p = 0.042; Table 3, Fig. 5). There was no effect of “Location” on hydrocoral percent cover (Table 3). Sponge cover also decreased significantly in the middle (5.2 ± 2.8% pre-Irma, 2.5 ± 1.7% post-Irma) and upper (1.3 ± 0.9% pre-Irma, 0.5 ± 0.3% post-Irma) Keys after the storm (ANOVA: F1,16 = 4.90, p = 0.042; Table 3, Fig. 5). “Location” had a significant effect on percent cover of sponges (ANOVA: F1,16 = 8.47, p = 0.01; Table 3), with a greater percent cover of sponges in the middle Keys as compared with the upper Keys both before and after the storm (Fig. 5). For scleractinian corals, TAS, and sediment, there was no significant effect of either “Storm” or “Location” on percent cover (Table 3, Fig. 5).

Relationship Between Urchin Abundance and Benthic Community

A Spearman’s rank correlation detected a significant relationship between the change in percent cover of sediment and the change in D. antillarum density, with greater decreases in D. antillarum density correlating with increased sediment (R2 = 0.57, p = 0.049; Fig. 6). There was no significant relationship between D. antillarum density following Hurricane Irma and the change in percent cover of macroalgae (R2 = 0.07, p = 0.467) or TAS (R2 = 0.17, p = 0.239).

Fig. 6
figure6

Relationship between the change in Diadema antillarum density (urchin m−2) and change in percent cover of sediment (%) at 7 sites in the Florida Keys. The Spearman’s rank correlation R2 value and p value, and regression line (y = − 0.003x–0.0161) are shown

Discussion

In the Florida Keys, a reduction in D. antillarum density following Hurricane Irma has contributed to a multi-decadal decline of this population due to recurrent disease outbreaks (Forcucci 1994; Kissling et al. 2014). Given the important role of D. antillarum as a grazer on Caribbean coral reefs, the lack of recovery of D. antillarum following historical disease outbreaks and its continued decline are of great concern for coral reef ecosystem managers (Ladd et al. 2018). Our results indicate that recovery of the Florida Keys population is impeded at least in part by strong storms, conditions that are expected to intensify with climate change (Knutson et al. 2010; Cheal et al. 2017). Chronically low D. antillarum abundances that lead to poor reproductive output (Feehan et al. 2016), coupled with increasing frequency of intense storms, may diminish the likelihood of local population recovery.

The significant relationship between an increase in sediment percent cover and decline in D. antillarum density suggests that sediment transport may have influenced D. antillarum mortality through abrasion and burial. Impacts of sediment transport on D. antillarum during storms have not previously been reported, and sedimentation as a mechanism for D. antillarum mortality was not explicitly tested in the present study. However, strong storms can displace large volumes of sediment and deposit it in new locations (Gagan et al. 1990; Hubbard 1992), which can negatively impact benthic reef organisms through tissue damage from sediment suspended in the water column during transport and by physically burying the benthos in a layer of newly deposited sediment (Blair et al. 1994). Alternatively, changes to sediment cover and D. antillarum density may be correlated simply due to patterns of water movement during the storm that both resuspended sediments and dislodged urchins due to intense wave action (Verling et al. 2005).

The decrease in median D. antillarum test diameter following the storm suggests size-specific mortality that may be related to a lack of structured habitat for large sea urchins on degraded Florida Keys reefs. Smaller D. antillarum that are able to seek refuge in reef crevices (Randall et al. 1964) may have been protected from direct sediment abrasion or dislodgement during the passage of Hurricane Irma. Physical structure offered by scleractinian corals is important refuge habitat for juvenile and adult D. antillarum (Weil et al. 1984; Lee 2006). However, coral cover was very low at our sites (3% mean cover) and continues to decrease throughout the region due to recent coral disease outbreaks (Walton et al. 2018). Reefs in Southern Mexico, where high survivorship of D. antillarum was documented following Hurricane Dean, had greater coral cover than the degraded reefs in our study (12.6% mean cover; Jorgensen et al. 2008), suggesting that habitat complexity associated with coral cover may be important for sheltering of D. antillarum from storm damage. Coral-depauperate reefs throughout the Caribbean region may offer inadequate refuge habitat to D. antillarum due to their increasingly flattened, low-complexity state (Alvarez-Filip et al. 2009), leaving large D. antillarum that require proportionately larger refugia more vulnerable to storm damage.

Given that the eye of Hurricane Irma passed through the lower Keys, we expected to observe differences in physical oceanographic conditions between the middle and upper Keys based on their proximity to the storm path. Contrary to this prediction, our hindcast estimates of significant wave heights did not detect a clear difference between locations during the passage of the storm. Similarly, bottom sea temperature data showed a comparable 3 °C drop in temperature at 3 sites spanning the study region. These findings are consistent with the lack of a storm by location interaction in ANOVA for several benthic community groups, including hydrocorals and sponges, suggesting that storm impacts on these groups were relatively homogenous across a spatial scale of 80 km. This observation is likely due to the overall close proximity of both the middle and upper Keys to the core of the storm (within 120 linear km) and the high intensity of the storm (Category 4) (Cangialosi et al. 2018).

Spatial variation in storm impacts was evident when comparing changes with the abundance of macroalgae in the middle Keys, where abundance decreased, and upper Keys, where abundance increased. Hurricanes can result in both macroalgal blooms (Woodley et al. 1981; Rogers et al. 1991) and losses (Blair et al. 1994; Mumby et al. 2005) in the Caribbean, and successive changes to the composition of macroalgal species on reefs following hurricane impacts are well documented (Harmelin-Vivien 1994). Our physical oceanographic data and examination of other benthic groups, including D. antillarum, indicate that Irma’s impacts were similar throughout the middle and upper Keys; therefore, it is unlikely that differences in physical storm impacts were a major driver of the observed patterns of macroalgal abundance. Instead, differences in the ability of macroalgae to withstand storm impacts could have influenced post-storm macroalgal abundances. Variation in attachment strength and morphology of different algal species (e.g., weakly attached and erect Wrangelia sp.) may influence the composition of algae following a hurricane (Blair et al. 1994). However, initial macroalgal community composition at our sites is not known due to the limitations of image analysis for species-level identification. Bottom-up control of macroalgal abundance (i.e., nutrient availability; Lapointe 1997) also could have influenced the colonization of macroalgae following Hurricane Irma, but changes to nutrient concentrations associated with Irma are not known.

Grazing pressure from fish and invertebrate herbivores can also regulate the distribution of macroalgae on reefs following disturbance (Woodley et al. 1981; Carpenter 1986; Hughes 1994). However, we detected no relationship between post-Irma D. antillarum density and the change in cover of macroalgae or TAS, indicating that grazing by low densities of D. antillarum at our sites was not sufficient to control macroalgal or turf proliferation. Unlike reefs where D. antillarum population recovery has resulted in localized decreases in macroalgal abundance (e.g., Edmunds and Carpenter 2001; Idjadi et al. 2010), D. antillarum densities at our sites in the Florida Keys remain low, with little sign of population growth (Chiappone et al. 2013). A diverse community of herbivores in addition to D. antillarum can be important for regulating algal abundance (Francis et al. 2019), although abundances of other grazers, including other sea urchin species and herbivorous fishes, are not known for our sites.

Although hurricanes have historically contributed to significant losses in the abundance of scleractinian corals throughout the Caribbean, the role of hurricane disturbance in coral decline has waned in recent decades relative to other stressors such as sedimentation and disease (Gardner et al. 2005). Correspondingly, we did not detect an effect of Hurricane Irma on scleractinian coral abundance in our study. The degree of mechanical damage to scleractinian corals can depend on coral morphology, and fragile branching species tend to be more affected by storms than sturdy massive species (Woodley et al. 1981). Since scleractinian coral abundance was initially very low at our sites and the majority of corals were massive species (data not shown), it is unsurprising that overall scleractinian coral abundance was not significantly impacted by Hurricane Irma. However, reports of localized coral mortality following Hurricane Irma, especially in the middle and lower Keys, indicate that the present study may not fully capture Irma’s impacts on scleractinian corals in the region (NOAA 2018).

In the absence of D. antillarum grazing to facilitate coral recruitment, the Florida Keys reef ecosystem will likely remain coral-depauperate. Regime shifts between coral- and macroalgal-dominated states are influenced by processes regulating grazing pressure and coral abundance (Mumby 2009), and dramatic increases in grazing intensity or coral recruitment may shift the community towards coral-dominance (Mumby et al. 2007a). Hydrocorals, macroalgae, and TAS dominated the benthic community at our sites prior to Hurricane Irma. Following Irma, the benthic community may remain dominated by relatively fast-growing, soft-bodied community groups including hydrocorals and fleshy macroalgae, which have become increasingly dominant taxa on Florida Keys reefs (Ruzicka et al. 2013). While these soft-bodied benthic groups are commonly susceptible to hurricane damage (Woodley et al. 1981; Hubbard et al. 1991; Blair et al. 1994; Jorgensen et al. 2008), recovery can occur within weeks of storm disturbance following reattachment to substrate and rapid growth (Woodley et al. 1981). These benthic taxa do not provide the complex refuge habitat needed to support D. antillarum recruits, which contributes to the habitat limitation that hinders D. antillarum population recovery (Miller et al. 2009; Bodmer et al. 2015). Disturbance from strong storms like Hurricane Irma may reinforce the coral-depauperate state by depressing grazing pressure from D. antillarum populations during the critical post-storm recovery period, which prevents reestablishment of scleractinian corals and by extension further inhibits D. antillarum population recovery.

The scope of the present study is limited to shallow reefs in the middle and upper Keys and the overall status of D. antillarum in the Florida Keys remains poorly understood. Large-scale assessments of D. antillarum in the Florida Keys are needed to determine the current status of the population. D. antillarum abundances may increase locally following Hurricane Irma over timeframes of years as pelagic larvae arrive and recruit to reefs. However, this assumes a source of larvae and post-settlement survival, which can be intermittent and low, respectively (Miller et al. 2009; Feehan et al. 2019). Thus, recovery of D. antillarum in the Florida Keys may depend on active management, including mitigating anthropogenic climate change that is driving an increase in storm intensities, and recovering physical reef structure to improve D. antillarum survival during storms and provide habitat for recruits.

References

  1. Adam, T.C., D.E. Burkepile, B.I. Ruttenberg, and M.J. Paddack. 2015. Herbivory and the resilience of Caribbean coral reefs: Knowledge gaps and implications for management. Marine Ecology Progress Series 520: 1–20.

    Article  Google Scholar 

  2. Alvarez-Filip, L., N.K. Dulvy, J.A. Gill, I.M. Côté, and A.R. Watkinson. 2009. Flattening of Caribbean coral reefs: Region-wide declines in architectural complexity. Proceedings of the Royal Society B: Biological Sciences 276 (1669): 3019–3025.

    Article  Google Scholar 

  3. Aronson, R.B. 1993. Hurricane effects on backreef echinoderms of the Caribbean. Coral Reefs 12 (3-4): 139–142.

    Article  Google Scholar 

  4. Aronson, R.B., P.J. Edmunds, W.F. Precht, D.W. Swanson, and D.R. Levitan. 1994. Large-scale, long-term monitoring of Caribbean coral reefs: Simple, quick, inexpensive techniques. Atoll Research Bulletin 421: 1–19.

    Article  Google Scholar 

  5. Beisner, B.E., D.T. Haydon, and K. Cuddington. 2003. Alternative stable states in ecology. Frontiers in Ecology and the Environment 1 (7): 376–382.

    Article  Google Scholar 

  6. Bellwood, D.R., T.P. Hughes, C. Folke, and M. Nyström. 2004. Confronting the coral reef crisis. Nature 429 (6994): 827.

    CAS  Article  Google Scholar 

  7. Blair, S.M., T.L. McIntosh, and B.J. Mostkoff. 1994. Impacts of Hurricane Andrew on the offshore reef systems of central and northern Dade County, Florida. Bulletin of Marine Science 54 (3): 961–973.

    Google Scholar 

  8. Bodmer, M.D., Rogers, A.D., Speight, M.R., Lubbock, N., Exton, D.A. 2015. Using an isolated population boom to explore barriers to recovery in the keystone Caribbean coral reef herbivore Diadema antillarum. Coral Reefs 34 (4):1011-1021

  9. Cangialosi, J. P., Latto, A. S. and Berg, R. 2018. National Hurricane Center tropical cyclone report: Hurricane Irma (AL 112017). National Hurricane Center.

  10. Carpenter, R.C. 1986. Partitioning herbivory and its effects on coral reef algal communities. Ecological monographs 56 (4): 345–364.

    Article  Google Scholar 

  11. Cheal, A.J., M.A. MacNeil, M.J. Emslie, and H. Sweatman. 2017. The threat to coral reefs from more intense cyclones under climate change. Global Change Biology 23 (4): 1511–1524.

    Article  Google Scholar 

  12. Chiappone, M., L.M. Rutten, S.L. Miller, and D.W. Swanson. 2013. Recent trends (1999–2011) in population density and size of the echinoid Diadema antillarum in the Florida Keys. Florida Scientist 76: 23–35.

    Google Scholar 

  13. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199 (4335): 1302–1310.

    CAS  Article  Google Scholar 

  14. Edmunds, P.J., and R.C. Carpenter. 2001. Recovery of Diadema antillarum reduces macroalgal cover and increases abundance of juvenile corals on a Caribbean reef. Proceedings of the National Academy of Sciences 98 (9): 5067–5071.

    CAS  Article  Google Scholar 

  15. Estes, J.A., and J.F. Palmisano. 1974. Sea otters: Their role in structuring nearshore communities. Science 185 (4156): 1058–1060.

    CAS  Article  Google Scholar 

  16. Feehan, C.J., M.S. Brown, W.C. Sharp, J.S. Lauzon-Guay, and D.K. Adams. 2016. Fertilization limitation of Diadema antillarum on coral reefs in the Florida Keys. Ecology 97 (8): 1897–1904.

    Article  Google Scholar 

  17. Feehan, C.J., W.C. Sharp, T.N. Miles, M.S. Brown, and D.K. Adams. 2019. Larval influx of Diadema antillarum to the Florida Keys linked to passage of a Tortugas Eddy. Coral Reefs 38 (2): 387–393.

    Article  Google Scholar 

  18. Folke, C., S. Carpenter, B. Walker, M. Scheffer, T. Elmqvist, L. Gunderson, and C.S. Holling. 2004. Regime shifts, resilience, and biodiversity in ecosystem management. Annual Review of Ecology, Evolution, and Systematics 35: 557–581.

    Article  Google Scholar 

  19. Forcucci, D. 1994. Population density, recruitment and 1991 mortality event of Diadema antillarum in the Florida Keys. Bulletin of Marine Science 54 (3): 917–928.

    Google Scholar 

  20. Fox, J. and Weisberg, S. 2019. An {R} Companion to Applied Regression, Third Edition. Sage.

  21. Francis, F.T., K. Filbee-Dexter, H.F. Yan, and I.M. Côté. 2019. Invertebrate herbivores: Overlooked allies in the recovery of degraded coral reefs? Global Ecology and Conservation. https://doi.org/10.1016/j.gecco.2019.e00593.

  22. Furman, B., and K.L. Heck. 2009. Differential impacts of echinoid grazers on coral recruitment. Bulletin of Marine Science 85 (2): 121–132.

    Google Scholar 

  23. Gagan, M.K., A.R. Chivas, and A.L. Herczeg. 1990. Shelf-wide erosion, deposition, and suspended sediment tranport during Cyclone Winifred, central Great Barrier Reef, Australia. Journal of Sedimentary Research 60 (3): 456–470.

    Article  Google Scholar 

  24. Gardner, T.A., I.M. Cote, J.A. Gill, A. Grant, and A.R. Watkinson. 2005. Hurricanes and Caribbean coral reefs: Impacts, recovery patterns, and role in long-term decline. Ecology 86 (1): 174–184.

    Article  Google Scholar 

  25. Gunderson, L.H. 2001. Managing surprising ecosystems in southern Florida. Ecological Economics 37: 371–378.

    Article  Google Scholar 

  26. Harmelin-Vivien, M.L. 1994. The effects of storms and cyclones on coral reefs: A review. Journal of Coastal Research 12: 211–231.

    Google Scholar 

  27. Hubbard, D.K. 1992. Hurricane-induced sediment transport in open-shelf tropical systems; An example from St. Croix, US Virgin Islands. Journal of Sedimentary Research 62 (6): 946–960.

    Google Scholar 

  28. Hubbard, D.K., K.M. Parsons, J.C. Bythell, and N.D. Walker. 1991. The effects of Hurricane Hugo on the reefs and associated environments of St. Croix, US Virgin Islands—A preliminary assessment. Journal of Coastal Research 8: 33–48.

    Google Scholar 

  29. Hughes, T.P. 1994. Catastrophes, regime shifts, and large-scale degradation of a Caribbean coral reef. Science 265 (5178): 1547–1551.

    CAS  Article  Google Scholar 

  30. Hughes, T.P., M.J. Rodrigues, D.R. Bellwood, D. Ceccarelli, O. Hoegh-Guldberg, L. McCook, N. Moltschaniwskyj, M.S. Pratchett, R.S. Steneck, and B. Willis. 2007. Phase shifts, herbivory, and the resilience of coral reefs to climate change. Current Biology 17 (4): 360–365.

    CAS  Article  Google Scholar 

  31. Hunte, W., and D. Younglao. 1988. Recruitment and population recovery of Diadema antillarum (Echinodermata; Echinoidea) in Barbados. Marine Ecology Progress Series 45 (1): 109–119.

    Article  Google Scholar 

  32. Idjadi, J.A., R.N. Haring, and W.F. Precht. 2010. Recovery of the sea urchin Diadema antillarum promotes scleractinian coral growth and survivorship on shallow Jamaican reefs. Marine Ecology Progress Series 403: 91–100.

    Article  Google Scholar 

  33. Jackson, J.B.C., M.K. Donovan, K.L. Cramer, and V.V. Lam. 2014. Status and trends of Caribbean coral reefs. Washington, DC: Global Coral Reef Monitoring Network, International Union for the Conservation of Nature Global Marine and Polar Program.

    Google Scholar 

  34. Jorgensen, P., J. Espinoza-Ávalos, and H. Bahena-Basave. 2008. High population density survival of the sea urchin Diadema antillarum (Philippi 1845) to a category 5 hurricane in southern Mexican Caribbean. Hidrobiológica 18 (3): 257–260.

  35. Kissling, D.L., W.F. Precht, S.L. Miller, and M. Chiappone. 2014. Historical reconstruction of population density of the echinoid Diadema antillarum on Florida Keys shallow bank-barrier reefs. Bulletin of Marine Science 90 (2): 665–679.

    Article  Google Scholar 

  36. Knowlton, N. 1992. Thresholds and multiple stable states in coral reef community dynamics. American Zoologist 32 (6): 674–682.

    Article  Google Scholar 

  37. Knutson, T.R., J.L. McBride, J. Chan, K. Emanuel, G. Holland, C. Landsea, I. Held, J.P. Kossin, A.K. Srivastava, and M. Sugi. 2010. Tropical cyclones and climate change. Nature Geoscience 3 (3): 157.

    CAS  Article  Google Scholar 

  38. Ladd, M.C., M.W. Miller, J.H. Hunt, W.C. Sharp, and D.E. Burkepile. 2018. Harnessing ecological processes to facilitate coral restoration. Frontiers in Ecology and the Environment 16 (4): 239–247.

    Article  Google Scholar 

  39. Lapointe, B.E. 1997. Nutrient thresholds for bottom-up control of macroalgal blooms on coral reefs in Jamaica and southeast Florida. Limnology and Oceanography 42 (5.2): 1119–1131.

    CAS  Article  Google Scholar 

  40. Lee, S.C. 2006. Habitat complexity and consumer-mediated positive feedbacks on a Caribbean coral reef. Oikos 112 (2): 442–447.

    Article  Google Scholar 

  41. Lessios, H.A. 1988. Mass mortality of Diadema antillarum in the Caribbean: What have we learned? Annual Review of Ecology, Evolution, and Systematics 19: 371–393.

    Article  Google Scholar 

  42. Lessios, H.A. 2016. The great Diadema antillarum die-off: 30 years later. Annual Review of Marine Science 8: 267–283.

    CAS  Article  Google Scholar 

  43. Lirman, D. 2001. Competition between macroalgae and corals: Effects of herbivore exclusion and increased algal biomass on coral survivorship and growth. Coral Reefs 19 (4): 392–399.

    Article  Google Scholar 

  44. Miller, M.W., K.L. Kramer, S.M. Williams, L. Johnston, and A.M. Szmant. 2009. Assessment of current rates of Diadema antillarum larval settlement. Coral Reefs 28 (2): 511–515.

    Article  Google Scholar 

  45. Mumby, P.J. 2009. Phase shifts and the stability of macroalgal communities on Caribbean coral reefs. Coral Reefs 28 (3): 761–773.

    Article  Google Scholar 

  46. Mumby, P.J., N.L. Foster, and E.A.G. Fahy. 2005. Patch dynamics of coral reef macroalgae under chronic and acute disturbance. Coral Reefs 24 (4): 681–692.

    Article  Google Scholar 

  47. Mumby, P.J., J.D. Hedley, K. Zychaluk, A.R. Harborne, and P.G. Blackwell. 2006. Revisiting the catastrophic die-off of the urchin Diadema antillarum on Caribbean coral reefs: Fresh insights on resilience from a simulation model. Ecological Modelling 196 (1-2): 131–148.

    Article  Google Scholar 

  48. Mumby, P.J., A.R. Harborne, J. Williams, C.V. Kappel, D.R. Brumbaugh, F. Micheli, K.E. Holmes, C.P. Dahlgren, C.B. Paris, and P.G. Blackwell. 2007a. Trophic cascade facilitates coral recruitment in a marine reserve. Proceedings of the National Academy of Sciences 104 (20): 8362–8367.

    CAS  Article  Google Scholar 

  49. Mumby, P.J., A. Hastings, and H.J. Edwards. 2007b. Thresholds and the resilience of Caribbean coral reefs. Nature 450 (7166): 98–101.

    CAS  Article  Google Scholar 

  50. Myhre, S., and A. Acevedo-Gutiérrez. 2007. Recovery of sea urchin Diadema antillarum populations is correlated to increased coral and reduced macroalgal cover. Marine Ecology Progress Series 329: 205–210.

    Article  Google Scholar 

  51. NOAA. 2018. NOAA and partners assess reef, aid recovery following Hurricane Irma. https://floridakeys.noaa.gov/whatsnew/releases/2018/20180110-noaa-and-partners-assess-reef-aid-recovery-following-irma.html. Accessed 22 Aug 2018.

  52. Nyström, M., C. Folke, and F. Moberg. 2000. Coral reef disturbance and resilience in a human-dominated environment. Trends in Ecology & Evolution 15 (10): 413–417.

    Article  Google Scholar 

  53. Randall, J.E., R.E. Schroeder, and W.A. Starck. 1964. Notes on the biology of the echinoid Diadema antillarum. Caribbean Journal of Science 4 (2-3): 421–433.

    Google Scholar 

  54. Rogers, C.S., L.N. McLain, and C.R. Tobias. 1991. Effects of Hurricane Hugo (1989) on a coral reef in St. John, USVI. Marine Ecology Progress Series 78 (2): 189–199.

    Article  Google Scholar 

  55. Ruzicka, R.R., M.A. Colella, J.W. Porter, J.M. Morrison, J.A. Kidney, V. Brinkhuis, K.S. Lunz, K.A. Macaulay, L.A. Bartlett, M.K. Meyers, and J. Colee. 2013. Temporal changes in benthic assemblages on Florida Keys reefs 11 years after the 1997/1998 El Niño. Marine Ecology Progress Series 489: 125–141.

    Article  Google Scholar 

  56. Scheffer, M., S. Carpenter, J.A. Foley, C. Folke, and B. Walker. 2001. Catastrophic shifts in ecosystems. Nature 413 (6856): 591.

    CAS  Article  Google Scholar 

  57. Schutte, V.G., E.R. Selig, and J.F. Bruno. 2010. Regional spatio-temporal trends in Caribbean coral reef benthic communities. Marine Ecology Progress Series 402: 115–122.

    Article  Google Scholar 

  58. Steiner, S.C.C. 2003. Stony corals and reefs of Dominica. Atoll Research Bulletin 498: 1–15.

    Article  Google Scholar 

  59. The WAVEWATCH III © Development Group (WW3DG). 2016. User manual and system documentation of WAVEWATCH III© version 5.16. Tech. Note 329, NOAA/NWS/NCEP/MMAB, College Park, MD, USA, 326 pp. + Appendices

  60. Verling, E., D.K. Barnes, and A.C. Crook. 2005. Smashing tests? Patterns and mechanisms of adult mortality in a declining echinoid population. Marine Biology 147 (2): 509–515.

    Article  Google Scholar 

  61. Walton, C., N.K. Hayes, and D.S. Gilliam. 2018. Impacts of a Regional, Multi-Year, Multi-Species Coral Disease Outbreak in Southeast Florida. Frontiers in Marine Science 5: 323.

    Article  Google Scholar 

  62. Weil, E., F. Losada, and D. Bone. 1984. Spatial variations in density and size of the echinoid Diadema antillarum Philippi on some Venezuelan coral reefs. Bijdragen tot de Dierkunde 54 (1): 73–82.

    Article  Google Scholar 

  63. Wickham, H., François, R., Henry L., and Müller, K.. 2018. dplyr: A Grammar of Data Manipulation. R package version 0.7.6. https://CRAN.R-project.org/package=dplyr. Accessed 8 Jan 2019

  64. Woodley, J.D., E.A. Chornesky, P.A. Clifford, J.B.C. Jackson, L.S. Kaufman, N. Knowlton, J.C. Lang, M.P. Pearson, J.W. Porter, M.C. Rooney, and K.W. Rylaarsdam. 1981. Hurricane Allen’s impact on Jamaican coral reefs. Science 214 (4522): 749–755.

    CAS  Article  Google Scholar 

Download references

Acknowledgments

The authors thank Brian Reckenbeil, Elliot Hart, and Mike Bollinger for field support, and Dara Yiu and three anonymous reviewers for helpful comments on the manuscript.

Funding

The research was financially supported by Montclair State University and a Florida State Wildlife Grant (SWG) awarded to WCS and CJF. JNK was financially supported by a research assistantship from Montclair State University.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Julia N. Kobelt.

Additional information

Communicated by Ken Dunton

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kobelt, J.N., Sharp, W.C., Miles, T.N. et al. Localized Impacts of Hurricane Irma on Diadema antillarum and Coral Reef Community Structure. Estuaries and Coasts 43, 1133–1143 (2020). https://doi.org/10.1007/s12237-019-00665-4

Download citation

Keywords

  • Sea urchin
  • Coral
  • Hydrocoral
  • Macroalgae
  • Algal turf
  • Turf-algal-sediment matrix
  • Sponge
  • Storm
  • Tropical cyclone