Tracking a Keystone Species Recovery from a Mass Mortality Event

Marine Environmental Research 156 (2020) 104905

Contents lists available at ScienceDirect

Marine Environmental Research

journal homepage: http://www.elsevier.com/locate/marenvrev

Before and after a disease outbreak: Tracking a keystone species recovery from a mass mortality event

Francesca Gizzi a,*, Jesús Jim�enez a, Susanne Schafer€ a,b, Nuno Castro a, Sonia� Costa a,c, Silvia Lourenço d,e, Ricardo Jos�e f, Joao~ Canning-Clode a,g,h, Joao~ Monteiro a a MARE - Marine and Environmental Sciences Centre, Ag^encia Regional para o Desenvolvimento da Investigaçao,~ Tecnologia e Inovaçao~ (ARDITI), Edifício Madeira Tecnopolo, Caminho da Penteada, 9020-105, Funchal, Madeira, Portugal b GEOMAR, Marine Ecology Department, Helmholtz Centre for Ocean Research Kiel, Düsternbrooker Weg 20, 24105, Kiel, Germany c OOM/ARDITI, Ag^encia Regional para o Desenvolvimento da Investigaçao,~ Tecnologia e Inovaçao,~ Edifício Madeira Tecnopolo, Caminho da Penteada, 9020-105, Funchal, Madeira, Portugal d MARE – Marine and Environmental Sciences Centre, Polit�ecnico de Leiria, Edifício CETEMARES, Av. Porto de Pesca, 2520 – 630, Peniche, Portugal e CIIMAR, Centro Interdisciplinar de Investigaçao~ Marinha e Ambiental, Universidade do Porto, Terminal de Cruzeiros do Porto de Leixoes,~ Av. General Norton de Matos, S/N, 4450-208, Matosinhos, Portugal f Direcçao~ Regional de Pescas, Centro de Maricultura da Calheta, Av. D. Manuel I, 9370-133, Calheta, Portugal g Centre of IMAR of the University of the Azores, Department of Oceanography and Fisheries, Rua Prof. Dr. Frederico Machado, 4, PT-9901-862, Horta, Azores, Portugal h Smithsonian Environmental Research Center, 647 Contees Wharf Road, Edgewater, MD, 21037, USA

ARTICLE INFO ABSTRACT

Keywords: Mass mortality events involving marine taxa are increasing worldwide. The long-spined sea urchin Diadema Long-spined sea urchin africanum is considered a keystone herbivore species in the northeastern Atlantic due to its control over the Disease abundance and distribution of algae. After a firstregistered mass mortality in 2009, another event off the coasts Widespread die-off of Madeira archipelago affected this ecologically important species in summer 2018. This study documented the Population density 2018 D. africanum mass mortality event, and the progress of its populations on the southern coast of Madeira Citizen science Pathogen assessment island. A citizen science survey was designed targeting marine stakeholders to understand the extent and in Madeira tensity of the event around the archipelago. Underwater surveys on population density prior, during and after the mass mortality, permitted an evaluation of the severity and magnitude of the event as well as urchin population recovery. A preliminary assessment of causative agents of the mortality was performed. The event was reported in the principal islands of the archipelago reducing the populations up to 90%. However, a fast recovery was registered during the following months, suggesting that the reproductive success was not compromised. Microbiological analyses in symptomatic and asymptomatic individuals, during and after the event, was not conclusive. Nevertheless, the bacteria Aeromonas salmonicida, or the gram-negative bacteria, or the interaction of different types of bacteria may be responsible for the disease outbreak. Further studies are needed to assess the role of pathogens in sea urchin mass mortalities and the compound effects that sea urchins have in local habitats and ecological functioning of coastal marine ecosystems.

1. Introduction severity of disease-related mass mortality events are increasing in ma rine organisms (Harvell et al., 1999; Epstein et al., 1996; Hayes et al., Over the past decades, several mass mortality events across a range 2001; Lafferty et al., 2004), partly due to climate change, such as ocean of marine taxa (e.g., corals, echinoderms, molluscs, turtles) have been warming (Harvell et al., 2002), since high temperature influences the documented around the globe (Harvell et al., 1999, 2004; Ward and introduction and propagation of pathogens (Scheibling and Lafferty, 2004; Kim et al., 2005; Jackson, 2008). The frequency and Lauzon-Guay, 2010; Scheibling et al., 2013). Disease outbreaks and the

* Corresponding author. E-mail addresses: [email protected] (F. Gizzi), [email protected] (J. Jim�enez), [email protected] (S. Schafer),€ nuno. [email protected] (N. Castro), [email protected] (S. Costa), [email protected] (S. Lourenço), [email protected] (R. Jos�e), jcanning- [email protected] (J. Canning-Clode), [email protected] (J. Monteiro). https://doi.org/10.1016/j.marenvres.2020.104905 Received 1 October 2019; Received in revised form 23 January 2020; Accepted 3 February 2020 Available online 8 February 2020 0141-1136/© 2020 Elsevier Ltd. All rights reserved. F. Gizzi et al. Marine Environmental Research 156 (2020) 104905

following associated mortality often have a compound or cascading ef In the Caribbean, populations of D. antillarum failed to recover (Lessios, fect in local marine habitats and communities (Ward and Lafferty, 2005) and three decades after the event, populations still had not 2004), driving persistent ecosystem changes (Hughes, 1994; Mangel and reached their pre-mass mortality densities (Rodríguez-Barreras et al., Levin, 2005). The length and extent of these ecosystem changes often 2015a). During the mortality event in 2009 in both Madeira and the result from how extensive and intense the mass mortality event is, as Canary Islands, part of D. africanum populations survived and demon well as on the traits of life history and the ecological roles of the affected strated a rapid recovery in their abundance, suggesting that in that taxa (Paine et al., 1998). When mass mortality events affect keystone occasion the event did not compromise the reproductive success of the species, such as key producers, key herbivores or keystone predators, the species (Clemente et al., 2014). Although Clemente et al. (2014) have effects on the community structure and ecosystem functioning can be registered an increase of D. africanum abundance in 2009, the lack of particularly significant( Scheibling et al., 2013; Lessios, 1988; Clemente pre-mortality data did not allow to determine in detail the extent of et al., 2014). Sea urchins have an important role in the community recovery after the mass mortality event. In the current study, we present structure of many benthic coastal habitats worldwide, as they typically data on the mass mortality event of D. africanum around Madeira ar control the abundance and distribution of algae through their grazing chipelago during the summer 2018, including a preliminary assessment activity (Andrew, 1989; Vadas and Hughes, 1990; Vadas et al., 1992; of possible causative agents and analysis of the effect of the die-off on the Valentine et al., 1997). In tropical regions, an increase in sea urchin local urchin population and its subsequent recovery. To better under populations often plays a major role in keeping coral dominated reefs stand the effect of the mass mortality on the abundance of natural from being overgrown by algae. Thus, urchin population die-offs can populations of D. africanum, we analysed population density data often lead to phase-shifts from coral dominated to algae dominated reefs recorded before, during and after the event occurred in 2018. Having (Hughes, 1994; Lessios, 1988; Carpenter, 1990; Knowlton, 1992). In pre-mortality density values as a baseline, this recent outbreak and shallow rocky reefs (common in temperate waters), an increase in sea resulting mass mortality, provided the opportunity to obtain crucial urchin population densities often promotes a phase-shift from a domi information on how an elevated reduction in D. africanum abundance nance of fleshyalgae into urchin barrens or barren grounds - overgrazed affects the continuance of this key herbivore species and to better assess substrates dominated by thin encrusting coralline algae (Scheibling the speed of populations recovery. et al., 2013; Clemente et al., 2014; Filbee-Dexter and Scheibling, 2014; Rodríguez et al., 2017). Clearly, variation in the abundance of sea ur 2. Materials chins, such as higher density or mass mortality, cause alterations of abundance and composition of algae, followed by changes in community 2.1. Study site species composition (Lessios, 1988; Idjadi et al., 2010). Disease out breaks are considered one of the few mechanisms known to cause mass Madeira is a Portuguese archipelago located in the warm-temperate � 0 mortality in sea urchin populations (Feehan et al., 2012) and the fre waters of the Northeast Atlantic Ocean, at approximately 32 38 N and � quency of these events has contributed to the recent characterization of 16 54’ W and consists of two main inhabited islands, Madeira and Porto echinoderms as a “boom-bust” phylum (Jellett et al., 1988; Birkeland Santo, and fiveuninhabited islands (Desertas and Selvagens Islands) that et al., 1989; Uthicke et al., 2009). Bacterial infections have been form one MPA (Marine Protected Area; Fig. 1). Located within Maca described as the cause for different mass mortality events in several sea ronesia, together with the Azores, the Canary Islands and Cape Verde, urchin species, including Paracentrotus lividus in the Mediterranean Madeira is an oceanic archipelago, surrounded by oligotrophic waters, (Boudouresque et al., 2007), Strongylocentrotus franciscanus and where the coastline and shallow waters are dominated by rocky shores S. purpuratus in the PacificOcean (Rogers-Bennett and Lawrence, 2007; and reefs (Canning-Clode et al., 2008). Jurgens et al., 2015) and Diadema antillarum in the Caribbean. The last cited event was the greatest marine animal mass mortality event 2.2. Citizen science survey recorded (Lessios, 1988b, 1995) resulting in 93% decline in population abundance involving the entire Caribbean area (Lessios, 1988b). In In July 2018, signals of a disease outbreak with high mortality of 2009, similar symptoms of disease were observed in D. mexicanum, D. africanum, was first detected at the south coast of Madeira, near causing a localized mass mortality along the Pacific coast of Mexico Funchal, the capital of the island. To have a perception of the geographic (Benítez-Villalobos et al., 2009). In the same year, D. africanum began to extent of the disease and sea urchin mortality, an internet-based survey die-off around the southeast coast of Madeira Island, and the disease was conducted between the end of July and the beginning of September spread across several islands of Macaronesia (comprised by the island 2018. The custom designed survey questionnaire (Supplementary ma systems of Azores, Madeira, Canaries and Cabo Verde), causing massive terial 1), was developed for users to report date, time, location, number mortality in D. africanum populations of Madeira and Canary Islands of dead sea urchins observed (classes 1–5, 6–20, 21–50, 51–100, >100) (Clemente et al., 2014). In the eastern Atlantic, D. africanum is widely and estimated mortality percentage (classes 1–10%, 11–40%, 41–70%, distributed in the Canary Islands (Casanas~ et al., 1998) and in Madeira 71–90%, >90%), and was widely disseminated in early August 2018 on archipelago (Alves et al., 2001, 2003) where it is recognized as a key social media platforms and by email to maximize the number of par ecological species, as it is responsible for facilitating the transition from ticipants. Citizens who detected dead D. africanum during SCUBA diving, macroalgal beds to barren grounds due to its intense grazing activity free diving, snorkelling or spearfishing around Madeira archipelago, (Alves et al., 2001, 2003; Tuya et al., 2005; Hernandez� et al., 2008; were involved to report data on mortality through a short on-line survey. Lessios et al., 2013; Cabanillas-Teran� et al., 2014). Several factors may Survey was active for about one month and data was compiled and influencesea urchin recovery, including competition (Hagen and Mann, plotted on a map using the sf and ggplot R environment packages 1992), resource availability (Menge, 1992), settlement, recruitment (Wickham, 2016) to visualize the severity and spatial extent of the success (Edmunds and Carpenter, 2001; Miller et al., 2003), predation mortality event. (Dayton et al., 1984; Sala and Ballesteros, 1997) and habitat complexity, that facilitates juvenile survival through the provision of shelter from 2.3. Sea urchin population densities predation (Bodmer et al., 2015). Moreover, sea urchin populations are subjected to environmental factors and human-mediated environmental To better assess how the disease and subsequent mortality affected changes (e.g. ocean warming and acidification) that have the potential the populations of D. africanum, two study sites were monitored in the to influence their recovery and survival, increasing metabolic rates to Funchal area: Carlton and Palms (Fig. 1). Site selection was based on supply higher energetic demands and reducing predator avoidance ease of access and on the possibility of conducting density surveys prior behaviour (Carey et al., 2016; Bodmer et al., 2017; Ohgaki et al., 2018). to any symptomatic individuals had been reported in the area.

2 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905

Fig. 1. Madeira archipelago location, monitoring study sites (PA-Palms and CA-Carlton) and extent of the disease outbreak. Circles in the map represent estimated mortality (circles size) and number of dead sea urchins (colour) reported by citizen surveys. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Monitoring was conducted on six occasions with three replicate tran sects at each site. Surveys were performed by scientific divers, before (T0; August 8, 2018), during (T1; 28/30 August 2018) and after (T2-T5; respectively October 17, 2018, November 8, 2018, January 15, 2019 and March 19, 2019) the mass mortality event. During surveys, D. africanum individuals were counted over rocky substrate along 50 � 4 m transects parallel to the coastline at depths between 9 and 11 m. All crevices and small holes were carefully inspected to avoid missing any individuals. The numbers of dead/symptomatic individuals with major loss of spines (Fig. 2) and asymptomatic sea urchins were recorded for each sampling replicate and at each survey time.

2.4. Statistical analysis

In order to compare the population density distribution between the two monitored sites, a two-sample Kolmogorov-Smirnov test was initially performed. Following, data from both sites were merged into a single time series data set. One-sample Kolmogorov-Smirnov’s was used to test data normality and the Levene’s test was used to assess variance homogeneity. Kruskal-Wallis (K–W) test was used as a robust nonpara Fig. 2. Symptoms of a diseased Diadema africanum recorded during the mass mortality of 2018 in Madeira. Photo credit: Susanne Schafer.€ metric alternative to ANOVA for the data that did not assume normal distribution. ANOVA and K–W were used to test: i) overall differences between sites and among times, and; ii) a pairwise comparisons between Statistics v22.0. consecutive times and between T0 and all other times. In particular, K–W was used to compare mean population density between sites and 2.5. Pathogen assessment among different sampling times, and in the pairwise comparisons be tween T0 vs T1 and T0 vs T2. All analyses were performed using PASW In an attempt to identify the pathogen potentially responsible for the

3 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905 mass mortality event, samples for microbiologic analysis were collected concerns regarding total volume of coelomic fluid collected per indi during the peak of the event (September 3, 2018) and approximately 3 vidual, samples were pooled into 2 sets of 5 individuals for both months after (November 27, 2018). During the mortality event, both symptomatic and asymptomatic groups. For the second sampling, with symptomatic and asymptomatic sea urchins (10 adult individuals of knowledge from the previous assessment, pooled coelomic solutions each) were collected by SCUBA diving at depths between 9 and 11 m. were analysed in 5 sets of 2 individuals. Sea urchins showing evident loss of spines were identified as symp tomatic (Fig. 2). After a three-month period, and with no further reports 3. Results of on-going mortality or symptomatic sea urchins, 10 additional in dividuals (asymptomatic) were collected for comparison with prior 3.1. Citizen survey collected samples. The coelomic fluid of each collected individual was sampled in situ by inserting a sterile needle through the peristomal Twenty-three stakeholders participated in the on-line survey be membrane and withdrawn into a sterile syringe. Samples (approxi tween August 1st and September 3rd, 2018, filling a total of 44 ques mately 1 ml/individual) were transferred to a sterile vial and trans tionnaires, from which 42 (39 from Madeira, 2 from Porto Santo and 1 ported in a cooler to the veterinary lab where the samples were from Desertas Islands) were taken in consideration after validation (i.e. immediately processed for bacteriological screening. Coelomic fluid exclusion due to lack of information). The outbreak was spread out samples were dissolved in a saline solution and incubated for 48 h in along all the south coast of Madeira, including the west and east coasts of cultivation chambers at ambient temperature. Grown bacteria colonies the island (Fig. 1). The reported number of dead sea urchins ranged from were isolated, gram coloured and bacteriologic activity was assessed by 1 to 5 to >100 while the reported percentage of mortality ranged from 1 catalase, coagulase and oxidase, followed by species phenotypic iden to 10% to >90% (Fig. 1). The most frequent class reported were >100 tification (Castro-Escarpulli et al., 2015). For the first sampling, due to dead individuals and the most frequent percentage ranges were 11–40%

Fig. 3. Effects of the 2018 disease outbreak in Diadema africanum: a) percentage of symptomatic sea urchin D. africanum in Carlton (CA) and Palms (PA) during the mass mortality event (T1). b) fluctuation in average relative abundance (dashed lines, left y axis) and D. africanum population density (box plot, right y axis) in Funchal (both sites). Box plot with median (solid horizontal line), firstand third quartiles (box outline), minimum and maximum values (whiskers) and outlier (dot). Mass mortality period (T1) highlighted in dark grey.

4 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905

and 71–90%. The results of the on-line survey are shown in the map the recovery and the stability of the system largely unknown. Part of (Fig. 1). these queries was resolved in the present study by reporting data on the sea urchin population densities prior to, during, and following the 3.2. Sea urchin population densities mortality event, permitting to better assess the effects of severe popu lation reduction and measure crucial information on the recovery dy Dead or symptomatic animals were only found during the mortality namics of D. africanum affected by a mass mortality event. This latest event (T1) and both, Carlton and Palms study sites had comparable mass mortality was firstdetected in Madeira archipelago in early August mortality rates (Fig. 3a). The symptoms of the disease presented a pro 2018 and lasted until late September 2018, affecting the main island, gressive deterioration of the epidermis of the body wall and loss of Porto Santo and Desertas (Fig. 1). Apparently, this disease outbreak of spines (Fig. 2). Population densities and their distributions were ho D. africanum was not exclusive to Madeira in 2018, as media reports mogeneous between the two selected sites (Kruskal-Wallis test, p ¼ highlighted a similar phenomenon occurring in the neighbour Canary 0.496; Two-sample Kolmogorov-Smirnov test, p ¼ 0.270), permitting to Islands merge the data from the two sites to perform subsequent analysis. (https://www.canarias7.es/siete-islas/el-erizo-diadema-fulminado- Overall fluctuation in relative abundance was plotted assuming T0 de-los-fondos-BE5251751; last accessed date was August 01, 2019). (prior to reports of mortality) as a starting point (Fig. 3b). Population During this event, anomalous accumulations of moribund sea urchins density was different among sampling times (Kruskal-Wallis test, p ¼ and dead individuals were reported along the coasts of the Madeira 0.034). In particular, different sea urchin densities occurred between T0 islands showing disease symptoms similar to those previously reported and T1 (Kruskal-Wallis test, p ¼ 0.016), between T1 and T2 (ANOVA, p during other mass mortality events in the eastern Atlantic and other ¼ 0.045), and between T0 and T2 (Kruskal-Wallis test, p ¼ 0.004), with areas of the world (Clemente et al., 2014; Benítez-Villalobos et al., 2009; a total mean decrease of 90% in sea urchin relative density within this Scheibling and Stephenson, 1984; Gilles and Pearse, 1986; Rob period (T0 – T2) (Fig. 3b). Two months after the mass mortality event erts-Regan et al., 1988; Girard et al., 2012). In Madeira, the 2018 disease (T3) sea urchin population recovered up to 42% of pre-mortality levels outbreak resulted in a major drop of the population density with a mean and reached 65% of pre-mortality density after 6 months from the event decrease of ~90% from pre-mortality values (Fig. 3b). The most (T5; Fig. 3b). examined sea urchin mass mortality event occurred in 1983 in the Caribbean (western Atlantic), resulting in a 93% decline in abundance of 3.3. Pathogen assessment D. antillarum (Lessios, 1988, 1995), and represented one of the largest documented mass mortalities of marine organisms (Lessios, 1988b). The The bacteriologic analysis revealed in all samples of both sampling recovery process from this large mass mortality event in the Caribbean periods the presence of several bacteria taxa in the coelomic fluid, was slow and spatially patchy (Weil et al., 2005; Miller et al., 2007; including Aeromonas, Sphingomonas, Pasteurella, Brevundimonas, Borde Levitan et al., 2014; Rodríguez-Barreras et al., 2015b; Lessios, 2016). tella and non-identified β-hemolítica colonies (Table 1). In the first Two years after the mass mortality, sea urchin population densities in sampling event, the species Aeromonas salmonicida and Sphingomonas Barbados had only recovered to 37% of their pre-mortality levels (Hunte paucimobilis were found in both symptomatic and apparently healthy sea and Younglao, 1988). After nearly twenty years, populations at St. Croix urchins, but only the symptomatic group presented an unidentified (US Virgin Islands), Curaçao and Panama still struggled to recover, β-hemolitic (gram-negative) bacteria species. For the second sampling reaching no more than 9.6% of their pre-mortality values (Miller et al., period, the analysis revealed the presence of A. salmonicida, in all five 2003; Lessios, 2005; Debrot and Nagelkerken, 2006). Thirty years after groups analysed. Moreover, Pasteurella spp, Brevundimonas vesicularis the die-off, D. antillarum population still had not recovered its pre-mass and Bordetella hinzii were found in one of the five sets (Table 1). mortality densities (Rodríguez-Barreras et al., 2015a) and due to the decrease in grazing activity, the degradation of Caribbean coral reefs 4. Discussion continues, promoting a phase-sifts to algal dominated reefs (Hughes, 1994; Lessios, 1988; Carpenter, 1990; Knowlton, 1992; Beck et al., This study showed a comprehensive analyses on the summer 2018 2014). The lack of recovery of D. antillarum populations may result from mass mortality event affecting the sea urchin D. africanum populations in a combination between low recruitment rates and intense predation, the Madeira Archipelago. The first recorded mortality event was compromising the survival of young sea urchins (Rodríguez-Barreras detected in October 2009 and caused a widespread mass mortality of et al., 2015a, 2015b). The present study showed that the recovery of D. africanum from Madeira (Portugal) to the Canary Islands (Spain), D. africanum was much faster than the one of D. antillarum in the covering an area of more than 400 km and resulting in about 65% Caribbean. Only six months after the event, D. africanum densities had overall reduction of urchin abundance (Clemente et al., 2014; Hernan� already increased their mean value up to 65% of pre-mortality levels dez et al., 2013). However, in that first study the scarcity of (Fig. 3b), suggesting that the reduction in density did not compromise pre-mortality data prevented an in-depth analysis of the incidence and the local reproductive success of the species. D. africanum has a clear severity of the disease outbreak in local populations, leaving the impact, recruitment pattern, with maximum spawning during the period be tween June and August (Hernandez� et al., 2011) and a main peak of � Table 1 recruitment between September and November (Hernandez et al., 2010; Bacteria colonies identifiedin coelomic fluidsamples collected during and after García-Sanz et al., 2014). Diadema has a fast growth rate in the initial the mass mortality event (3 September and 27 November, 2018 respectively). phase post-recruitment, reaching a rate of 5 mm in diameter per month Date Status N Species identified (Ogden and Carpenter, 1987). The mass mortality occurred between August and September, closely after the maximum spawning period. The 3 September Symptomatic 10 (2 groups of 5) Aeromonas salmonicida � Sphingomonas paucimobilis great reproductive potential of the species (Hernandez et al., 2011), β – hemolitic together with the possibility of recruitment from larvae that were in the (gram-negative) water while the adults were dying, may have played a key role in the 3 September Asymptomatic 10 (2 groups of 5) Aeromonas salmonicida observed short-term population recovery. However, since only Sphingomonas paucimobilis populations at depths of 9–11 m were sampled and the overall estimated 27 November Asymptomatic 10 (5 groups of 2) Aeromonas salmonicida Pasteurella spp mortality rates were obtained by the citizen surveys and were quite Brevundimonas vesicularis variable around the island, the existence of refugee populations at other Bordetella hinzii depths and/or from non-surveyed sites may have also had a contribution on the observed recovery. Despite the limitations inherent to citizen

5 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905 science approaches, the web-based survey still provided valuable in 5. Conclusions formation on how widespread the disease outbreak was and gave a general perception of the severity of the mass mortality in the islands This study took an important step towards understanding the effects population of D. africanum. Together with the population density sur of a disease outbreak, mass mortality and recovery of populations of veys, this information allows researchers to identify areas of interest for D. africanum, a major grazer and keystone species in rocky habitats in future investigation, and further examination of how high mortality and sub-tropical Northeast Atlantic islands. Unlike the late 90s mortality subsequent recovery may potentially affect algal cover (i.e. increase in event of D. antillarum in the Caribbean, D. africanum in this study had algal cover due to reduced grazing and low number D. africanum). In shown a recovery to over 60% in about 6 months, suggesting: i) that addition to providing core information on the extent, severity and re reproduction and recruitment have not been compromised by the covery of D. africanum populations, this study also provided a pre massive die-off; and/or ii) the existence of refuge populations. The im liminary assessment on the possible causative agents of the 2018 mass plications of the decrease in population density following the disease mortality event. The occurrence of bacteria in the coelomic fluidsin all outbreaks from 2009 to 2018 are still not fully understood. Further the sets of individuals analysed, suggested that at least one bacterial research and continued monitoring will be required to better understand pathogen was responsible. Healthy sea urchins typically present de the control mechanisms involved and how these will cope with fences against bacteria, including internal fluids with bactericidal increasing pressure from climate change to better predict likely sce properties (Wardlaw and Unkles, 1978; Turton and Wardlaw, 1987) and narios and design conservation and habitat restorations strategies. coelomocytes with phagocytizing activity that maintain the coelomic fluidsterile (Wardlaw and Unkles, 1978; Bang and Lemma, 1962; Smith Authors contribution et al., 1981). Bacterial infections are one of the most frequent reasons for mass mortalities in several sea urchin species. Mass mortality events JM conceived the study and supervised the research; JM, FG, SS and associated with bacterial infections have been previously reported in NC performed all fieldwork; JJ produced the graphs and maps; SC per Strongylocentrotus franciscanus, S. purpuratus (Richards and Kushner, formed the citizen survey on-line; SL and RJ performed bacteriologic 1994), S. droebachiensis (Jellett et al., 1988), (Pearse and Hines, 1979; analysis; FG analysed the data; FG, JJ, SS, SL, RJ, SC, NC, JCC, JM Miller and Colodey, 1983; Scheibling et al., 1999), P. lividus and contributed to the manuscript writing. All authors discussed the results D. mexicanum (Feehan and Scheibling, 2014). The 2009 mortality event and participated to the scientific discussion. of D. africanum in the eastern Atlantic (Canary Islands and Madeira Islands) was associated with a bacterial disease, in particular the Funding opportunistic gram-negative bacteria Vibrio alginolyticos was identified as being responsible (Clemente et al., 2014; Reilly et al., 2011). In the FG was supported by a post-doctoral research fellowship granted by present study, unidentifiedgram-negative bacteria (like V. alginolyticus) Ag^encia Regional para o Desenvolvimento da Investigaçao,~ Tecnologia e were found in the coelomic fluids of symptomatic D. africanum. Addi Inovaçao~ of Madeira (ARDITI) in the framework of project RAGES tionally, all pooled sets of coelomic fluids (during and after the mass [ARDITI-RAGES-2019-001]. JJ was supported by an ARDITI research mortality), showed the presence of Aeromonas salmonicida. Bacteria of fellowship in the framework of project Clean Atlantic [EAPA_46/2016]. the genera Aeromonas are commonly identified in the coelomic fluid of SS was supported through a federal state scholarship by the Kiel Uni infected sea urchins (Unkles, 1977), and there are evidences that versity and is currently funded by an ARDITI research fellowship in the A. salmonicida is a causative agent for the bald-sea-urchin disease in scope of the H2020 project GoJelly, [grant agreement No 774499]. SL S. purpuratus (Gilles and Pearse, 1986). Our findings are supported by was financiallysupported by a post-doctoral grant from ARDITI [Project previous studies (Gilles and Pearse, 1986; Unkles, 1977) and suggest M1420-09-5369-FSE-000001] and is currently funded by Ouriceira that the gram-negative bacteria (like V. alginolyticus) and/or the AQUA project [MAR-02.01.01-FEAMP-0004]. RJ and SC were finan A. salmonicida could be responsible for the 2018 mortality in Madeira cially supported by the Oceanic Observatory of Madeira (OOM) Project archipelago. In fact, sea urchins’ disease outbreaks can be associated to a [M1420-01-0145-FEDER-000001]. NC was supported by an ARDITI large spectrum of opportunistic bacterial families (Becker et al., 2008) Research fellowship in the framework of project MIMAR [MAC/4.6.d/ and are often triggered by shifts in environmental conditions, such as 066] INTERREG MAC 2014–2020 Programme. JCC was supported by a temperature, salinity or pH, favouring the spread of the infectious agent starting grant in the framework of the 2014 FCT (Fundaçao~ para a and influencinghow and to what extent the disease affects a population Ci^encia e a Tecnologia) Investigator Programme [IF/01606/2014/ (Harvell et al., 1999; Girard et al., 2012). Temperature, for example, has CP1230/CT0001]. JM was supported by a post-doctoral research a key role in regulating the transmission and development of bacteria in fellowship [ARDITI-M1420-09-5369-FSE-000001]. Activities within sea urchins (Scheibling and Stephenson, 1984; Jellett and Scheibling, this study were partially supported by MIMAR project cofounded by 1988), and the peak of disease outbreaks often overlap with peaks in sea FEDER Programa [MAC/4.6.d/066, INTERREG MAC 2014–2020 Pro surface temperatures (SST) (Scheibling and Lauzon-Guay, 2010; Feehan gramme], OOM [M1420-01-0145-FEDER-000001] and this is contri et al., 2012; Scheibling and Stephenson, 1984; Miller and Colodey, bution 44 from the Smithsonian’s MarineGEO Network. Finally, this 1983). The mass mortality event in Madeira also supports this pattern, as study also had the support of FCT, through the strategic project [UID/ it started in early August and lasted until the end of September (the MAR/04292/2019] granted to MARE UI&I. period when SST reaches its higher values in Madeira (Martins et al., 2007)). This may indeed explain why both, symptomatic and asymp Acknowledgements tomatic individuals presented A. salmonicida in their coelomic fluids. In-depth microbiological studies are necessary to better understand the The authors thank Azul Diving, Madeira Divepoint and Scuba mechanisms of the bacterial infections in sea urchins and comprehend Madeira for the logistic support, Christine Elisabeth Bauer for field how other variables, such as changes in environmental conditions (e.g. assistance. The authors also thank Margarida Costa and Alexandra Lopes temperature), may influencethe disease expression. Further research is from Laboratorio� Regional de Veterinaria� e Segurança Alimentar essential to validate our hypothesis and to predict real impact on the (LRVSA) for the bacteriologic analysis and Teresa Baptista for the coastal benthic ecosystem in face of climate change. Finally, proper revision of the pathogen assessment results. The authors also wish to management of D. africanum abundance on the temperate rocky sub thank all the participants of the on-line survey. strates of Madeira is recommended to prevent unfavourable ecosystem changes.

6 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905

Appendix A. Supplementary data local scale. J. Exp. Mar. Biol. Ecol. 413, 159–168. https://doi.org/10.1016/j. jembe.2011.12.003. Filbee-Dexter, K., Scheibling, R.E., 2014. Sea urchin barrens as alternative stable states of Supplementary data to this article can be found online at https://doi. collapsed kelp ecosystems. Mar. Ecol. Prog. Ser. 495, 1–25. https://doi.org/ org/10.1016/j.marenvres.2020.104905. 10.3354/meps10573. García-Sanz, S., Navarro, P.G., Tuya, F., 2014. Contrasting recruitment seasonality of sea urchin species in gran canaria, canary islands (eastern atlantic). Mediterr. Mar. Sci. References 15, 475–481. https://doi.org/10.12681/mms.547. Gilles, K.W., Pearse, J.S., 1986. Disease in sea urchin Strongylocentrotus purpuratus: Alves, F.M.A., Chicharo, L.M., Serrao, E., Abreu, A.D., 2001. Algal cover and sea-urchin experimental infection and bacterial virulence. Dis. Aquat. Org. 1, 105–114. spatial distribution at Madeira Island (NE Atlantic). Sci. Mar. 65, 383–392. https:// Girard, D., Clemente, S., Toledo-Guedes, K., Brito, A., Hernandez,� J.C., 2012. A mass doi.org/10.3989/scimar.2001.65n4383. mortality of subtropical intertidal populations of the sea urchin Paracentrotus lividus: Alves, F.M.A., Chicharo, L.M., Serrao, E., Abreu, A.D., 2003. Grazing by Diadema analysis of potential links with environmental conditions. Mar. Ecol. 33, 377–385. antillarum (Philippi) upon algal communities on rocky substrates. Sci. Mar. 67, https://doi.org/10.1111/j.1439-0485.2011.00491.x. 307–311. Hagen, N.T., Mann, K.H., 1992. Functional response of the predators american lobster Andrew, N.L., 1989. Contrasting ecological implications of food limitation in sea urchins Homarus americanus (Milne-Edwards) and Atlantic wolffish Anarhichas lupus (L.) to and herbivorous gastropods. Mar. Ecol. Prog. Ser. 51, 189–193. increasing numbers of the green sea urchin Strongylocentrotus droebachiensis (Müller). Bang, F.B., Lemma, A., 1962. Bacterial infection and reaction to injury in some J. Exp. Mar. Biol. Ecol. 159, 89–112. https://doi.org/10.1016/0022-0981(92) echinoderms. J. Insect Pathol. 4, 401–414. 90260-H. Beck, G., Miller, R., Ebersole, J., 2014. Mass mortality and slow recovery of Diadema Harvell, C.D., et al., 1999. Emerging marine diseases: climate links and anthropogenic antillarum: could compromised immunity be a factor? Mar. boil. 161, 1001–1013. factors. Science 285, 1505–1510. https://doi.org/10.1126/science.285.5433.1505. https://doi.org/10.1007/s00227-013-2382-6. Harvell, C.D., Mitchell, C.E., Ward, J.R., Altizer, S., Dobson, A.P., Ostfeld, R.S., Becker, P.T., Egea, E., Eeckhaut, I., 2008. Characterization of the bacterial communities Samuel, M.D., 2002. Climate warming and disease risks for terrestrial and marine associated with the bald sea urchin disease of the echinoid Paracentrotus lividus. biota. Science 296, 2158–2162. https://doi.org/10.1126/science.1063699. J. Invertebr. Pathol. 98, 136–147. https://doi.org/10.1016/j.jip.2007.12.002. Harvell, C.D., et al., 2004. The rising tide of ocean diseases: unsolved problems and Benítez-Villalobos, F., Díaz Martínez, J.P., Martínez-García, M., 2009. Mass mortality of research priorities. Front. Ecol. Environ. 2, 375–382. https://doi. the sea urchin Diadema mexicanum in La Entrega at Bahias de Huatulco, western org/10.1890/1540-9295(2004)002[0375:TRTOOD]2.0.CO;2. Mexico. Coral Reefs 28, 1017. https://doi.org/10.1007/s00338-009-0541-5. Hayes, M.L., Bonaventura, J., Mitchell, T.P., Prospero, J.M., Shinn, E.A., 2001. How are Birkeland, C., 1989. The influence of echinoderms on coral-reef communities. In: climate and marine biological outbreaks functionally linked? Hydrobiologia 460, Jangoux, M., Lawrence, J. (Eds.), Echinoderm Studies. Balkema, Rotterdam, 213–220. https://doi.org/10.1007/978-94-017-3284-0_19. pp. 1–79. Hernandez,� J.C., Clemente, S., Sangil, C., Brito, A., 2008. The key role of the sea urchin Bodmer, M.D.V., Rogers, A.D., Speight, M.R., Lubbock, N., Exton, D.A., 2015. Using an Diadema aff. antillarum in controlling macroalgae assemblages throughout the isolated population boom to explore barriers to recovery in the keystone Caribbean Canary Islands (eastern subtropical Atlantic): a spatio-temporal approach. Mar. coral reef herbivore Diadema antillarum. Coral Reefs 34, 1011–1021. https://doi.org/ Environ. Res. 66, 259–270. https://doi.org/10.1016/j.marenvres.2008.03.002. 10.1007/s00338-015-1329-4. Hernandez,� J.C., Clemente, S., Girard, D., Perez-Ruzafa, A., Brito, A., 2010. Effect of Bodmer, M.D.V., Wheeler, P.M., Hendrix, A.M., Cesarano, D.N., East, A.S., Exton, D.A., temperature on settlement and postsettlement survival in a barren-forming sea 2017. Interacting effects of temperature, habitat and phenotype on predator urchin. Mar. Ecol. Prog. Ser. 413, 69–80. https://doi.org/10.3354/meps08684. avoidance behaviour in Diadema antillarum: implications for restorative Hernandez,� J.C., Clemente, S., Brito, A., 2011. Effects of seasonality on the reproductive conservation. Mar. Ecol. Prog. Ser. 566, 105–115. http://doi.org/10.3354/ cycle of Diadema aff. antillarum in two contrasting habitats: implications for the meps12034. establishment of a sea urchin fishery. Mar. Biol. 158, 2603–2615. https://doi.org/ Boudouresque, C.F., Verlaque, M., 2007. Ecology of Paracentrotus lividus, p. 243À 285. In: 10.1007/s00227-011-1761-0. Lawrence, J.M. (Ed.), Edible Sea Urchins: Biology and Ecology. Elsevier, Amsterdam. Hernandez,� J.C., Sangil, C., Clemente, S., 2013. Sea urchins, natural events and benthic https://doi.org/10.1016/S0167-9309(07)80077-9. ecosystem functioning in the Canary Islands. In: Fernandez-Palocios,� J.M., Cabanillas-Teran,� N., Martín, J.A., Rodríguez-Barreras, R., Luque, A., 2014. Size-density Nascimiento, L.D., Hernandez,� J.C., Clemente, S., Gonzalez,� A., Diaz-Gonzalez,� J.P. strategy displayed by Diadema africanum linked with the stability of urchin-barrens (Eds.), Climate Change: Perspectives from the Atlantic: Past, Present and Future. in the Canary Islands. J. Mar. Biol. Assoc. U. K. 95, 145–151. https://doi.org/ Servicio de Publicaciones de la Universidad de La Laguna, Tenerife, pp. 487–512. 10.1017/S0025315414001246. Hughes, T.P., 1994. Catastrophes, phase-shifts, and large-scale degradation of a Canning-Clode, J., Kaufmann, M., Molis, M., Wahl, M., Lenz, M., 2008. Influence of Caribbean coral reef. Science 265, 1547–1551. https://doi.org/10.1126/ disturbance and nutrient enrichment on early successional fouling communities in science.265.5178.1547. an oligotrophic marine system. Mar. Ecol. 29, 115–124. https://doi.org/10.1111/ Hunte, W., Younglao, D., 1988. Recruitment and population recovery of Diadema j.1439-0485.2007.00210.x. antillarum (echinodermata; echinoidea) in Barbados. Mar. Ecol. Prog. Ser. 45, Carey, N., Harianto, J., Byrne, M., 2016. Sea urchins in a high-CO2 world: partitioned 109–119. effects of body size, ocean warming and acidificationon metabolic rate. J. Exp. Biol. Idjadi, J.A., Haring, R.N., Precht, W.F., 2010. Recovery of the sea urchin Diadema 219, 1178–1186. https://doi.org/10.1242/jeb.136101. antillarum promotes scleractinian coral growth and survivorship on shallow Carpenter, R.C., 1990. Mass mortality of Diadema antillarum. 1.Long term effects on sea Jamaican reefs. Mar. Ecol. Prog. Ser. 403, 91–100. urchin population-dynamics and coral reef algal communities. Mar. Biol. 104, Jackson, J., 2008. Ecological extinction and evolution in the brave new ocean. Proc. Natl. 67–77. Acad. Sci. U.S.A. 105, 11458–11465. https://doi.org/10.1073/pnas.0802812105. Casanas,~ A., Hanek-Larsen, H., Haroun, R.J., 1998. Developmental stages of blanquizal Jellett, J.F., Scheibling, R.E., 1988. Effect of temperature and prey availability on growth due to herbivory by the sea urchin Diadema antillarum Philippi in the Canary Islands. of Paramoeba invadens in monoxenic culture. Appl. Environ. Microbiol. 54, Boletim do Museo Municipal do Funchal 5, 139–146. 1848–1854. Castro-Escarpulli, G., et al., 2015. Identification and typing methods for the study of Jellett, J., Wardlaw, A., Scheibling, R., 1988. Experimental infection of the echinoid bacterial infections: a brief review and mycobacterial as case of study. Arch. Clin. Strongylocentrotus droebachiensis with Paramoeba invadens: quantitative changes in Microbiol. 7, 1–10. the coelomic fluid. Dis. Aquat. Org. 4, 149–157. Clemente, S., Lorenzo-Morales, J., Mendoza, J.C., Lopez,� C., Sangil, C., Alves, F., Jurgens, L.J., Rogers-Bennett, L., Raimondi, P.T., Schiebelhut, L.M., Dawson, M.N., Kaufmann, M., Hernandez,� J.C., 2014. Sea urchin Diadema africanum mass mortality Grosberg, R.K., Gaylord, B., 2015. Patterns of mass mortality among rocky shore in the subtropical eastern Atlantic: role of waterborne bacteria in a warming ocean. invertebrates across 100 km of northeastern Pacific coastline. PloS One 10, Mar. Ecol. Prog. Ser. 506, 1–14. https://doi.org/10.3354/meps10829. e0126280. https://doi.org/10.6084/m9.figshare.1372470. Dayton, P.K., Currie, V., Gerrodette, T., Keller, B.D., Rosenthal, R., Tresca, D.V., 1984. Kim, K., Dobson, A., Gulland, F., Harvell, C., 2005. Diseases and the conservation of Patch dynamics and stability of some California kelp. Ecol. Monogr. 54, 254–289. marine biodiversity. In: Norse, E., Crowder, L. (Eds.), Marine Conservation Biology. https://doi.org/10.2307/1942498. Island Press, Washington, pp. 149–166. Debrot, A.O., Nagelkerken, I., 2006. Recovery of the long-spined sea urchin Diadema Knowlton, N., 1992. Thresholds and multiple stable states in coral reef community antillarum in Curacao (Netherlands Antilles) linked to lagoonal and wave sheltered dynamics. Am. Zool. 32, 674–682. https://doi.org/10.1093/icb/32.6.674. shallow rocky habitats. Bull. Mar. Sci. 79, 415–424. Lafferty, K.D., Porter, J.W., Ford, S.E., 2004. Are diseases increasing in the ocean? Annu. Edmunds, P., Carpenter, R., 2001. Recovery of Diadema antillarum reduces macroalgal Rev. Ecol. Evol. Syst. 35, 31–54. https://doi.org/10.1146/annurev. cover and increases abundance of juvenile corals on a Caribbean reef. Proc. Natl. ecolsys.35.021103.105704. Acad. Sci. U.S.A. 98, 5067–5072. https://doi.org/10.1073/pnas.071524598. Lessios, H.A., 1988. Population dynamics of Diadema antillarum (Echinodermata: Epstein, P., 1996. Emergent stressors and public health implications in large marine echinoidea) following mass mortality in Panama. Mar. Biol. 99, 515–526. https:// ecosystems: an overview. In: Sherman, K., Jaworski, N.A., Smayda, T.J. (Eds.), The doi.org/10.1007/BF00392559. Northeast Shelf Ecosystem: Assessment, Sustainability, and Management. Blackwell, Lessios, H.A., 1988b. Mass mortality of Diadema antillarum in the Caribbean: what have Cambridge, MA, pp. 417–438. we learned? Annu. Rev. Ecol. Systemat. 19, 371–393. https://doi.org/10.1146/ Feehan, C.J., Scheibling, R.E., 2014. Effects of sea urchin disease on coastal marine annurev.es.19.110188.002103. ecosystems. Mar. Biol. 161, 1467–1485. https://doi.org/10.1007/s00227-014-2452- Lessios, H.A., 1995. Diadema antillarum 10 years after mass mortality: still rare, despite 4. help from a competitor. Proc. R. Soc. Lond. B Biol. Sci. 259, 331–337. https://doi. Feehan, C., Scheibling, R.E., Lauzon-Guay, J.S., 2012. An outbreak of sea urchin disease org/10.1098/rspb.1995.0049. associated with a recent hurricane: support for the ‘killer storm hypothesis’ on a

7 F. Gizzi et al. Marine Environmental Research 156 (2020) 104905

Lessios, H.A., 2005. Diadema antillarum populations in Panama twenty years following Rodríguez-Barreras, R., Duran,� A., Lop�ez-Morella, J., Sabat, A.M., 2015a. Effect of fish mass mortality. Coral Reefs 24, 125–127. https://doi.org/10.1007/s00338-004- removal on the abundance and size structure of the sea urchin Diadema antillarum: a 0443-5. field experiment. Mar. Biol. Res. 11, 1100–1107. https://doi.org/10.1080/ Lessios, H.A., 2016. The great Diadema antillarum die-off: 30 years later. Annu. Rev. Mar. 17451000.2015.1064140. Sci. 8, 267–283. https://doi.org/10.1146/annurev-marine-122414-033857. Rodríguez-Barreras, R., Mercado-Molina, A., P�erez, M.E., Sabat, A.M., 2015b. Arrested Lessios, H.A., 2013. ‘Natural’ population density fluctuations of echinoids. Do they help recovery of Diadema antillarum population: survival or recruitment limitation? predict the future? In: Fernandez-Palacios,� J.M., de Nascimento, L., Hernandez,� J.C., Estuar. Coast Shelf Sci. 163, 167–174. https://doi.org/10.1016/j.ecss.2015.06.021. Clemente, S., Gonzalez,� A., Díaz-Gonzalez,� J.P. (Eds.), Climate Change Perspectives Rogers-Bennett, L., 2007. The ecology of Strongylocentrotus franciscanus and from the Atlantic: Past, Present and Future. Servicios de Publicaciones, Universidad Strongylocentrotus purpuratus. In: Lawrence, J.M. (Ed.), Edible Sea Urchins: Biology de La Laguna, pp. 341–359. and Ecology. Elsevier, Amsterdam, pp. 393–425. https://doi.org/10.1016/S0167- Levitan, D., Edmunds, P.J., Levitan, K.E., 2014. What makes a species common? No 9309(07)80083-4. evidence of density-dependent recruitment or mortality of the sea urchin Diadema Sala, E., Ballesteros, E., 1997. Partitioning of space and food resources by three fishof the antillarum after the 1983–1984 mass mortality. Oecologia 175, 117–128. https://doi. genus Diplodus (Sparidae) in a Mediterranean rocky infralittoral ecosystem. Mar. org/10.1007/s00442-013-2871-9. Ecol. Prog. Ser. 152, 273–283. https://doi.org/10.3354/meps152273. Mangel, M., Levin, P.S., 2005. Regime, phase and paradigm shifts: making community Scheibling, R.E., Lauzon-Guay, J.S., 2010. Killer storms: north Atlantic hurricanes and ecology the basic science for fisheries. Philos. Trans. R. Soc. Lond. B Biol. Sci. 360, disease outbreaks in sea urchins. Limnol. Oceanogr. 55, 2331–2338. https://doi.org/ 95–105. https://doi.org/10.1098/rstb.2004.1571. 10.4319/lo.2010.55.6.2331. Martins, A.M., Amorim, A.S.B., Figueiredo, M.P., Souza, R.J., Mendonça, A.P., Scheibling, R.E., Stephenson, R.L., 1984. Mass mortality of Strongylocentrotus Bashmachnikov, I.L., Carvalho, D.S., 2007. sea surface temperature (AVHRR, droebachiensis (echinodermata: echinoidea) off nova scotia, Canada. Mar. Biol. 78, MODIS) and ocean colour (MODIS) seasonal and interannual variability in the 153–164. https://doi.org/10.1007/BF00394695. macaronesian islands of Azores, Madeira, and Canaries. P. 67430A. In: Bostater Jr, C. Scheibling, R., Hennigar, A., Balch, T., 1999. Destructive grazing, epiphytism, and R., Mertikas, S.P., Velez-Reyes,� X.N.M. (Eds.), Remote Sensing of the Ocean, Sea Ice, disease: the dynamics of sea urchin-kelp interactions in Nova Scotia. Can. J. Fish. and Large Water Regions 2007. Proc. of SPIE. https://doi.org/10.1117/12.738373. Aquat. Sci. 56, 2300–2314. https://doi.org/10.1007/BF00394695. Menge, B.A., 1992. Community regulation: under what conditions are bottom-up factors Scheibling, R.E., Feehan, C.J., Lauzon-Guay, J.S., 2013. Climate change, disease and the important on rocky shores? Ecology 73, 755–765. https://doi.org/10.2307/ dynamics of a kelp-bed ecosystem in Nova Scotia. p. 361-387. In: Fernandez-� 1940155. Palacios, J.M., de Nascimento, L., Hernandez,� J.C., Clemente, S., Gonzalez,� A., Miller, R.J., Colodey, A.G., 1983. Widespread mass mortalities of the green sea urchin in Díaz, J.P. (Eds.), Climate Change Perspectives from the Atlantic: Past, Present and Nova Scotia, Canada. Mar. Biol. 73, 263–267. https://doi.org/10.1007/ Future. Servicio de Publicaciones, Universidad de La Laguna. BF00392252. Smith, V.J., 1981. The echinoderms. In: Ratcliff, N.A., Rowley, A.F. (Eds.), Invertebrate Miller, R.J., Adams, A.J., Ogden, N.B., Ogdenand, J.C., Ebersole, J.P., 2003. Diadema Blood Cells, vol. 2. Academic Press, London, pp. 513–562. antillarum 17 years after mass mortality: is recovery beginning on St. Croix? Coral Turton, G.C., Wardlaw, A.C., 1987. Pathogenicity of the marine yeasts Metschikowia Reefs 22, 181–187. https://doi.org/10.1007/s00338-003-0301-x. zobelli and Rhodotorula rubra from sea urchin Echinus esculentus. Aquaculture 69, Miller, R.J., Adams, A.J., Ebersole, J.P., Ruiz, E., 2007. Evidence for positive density- 199–202. https://doi.org/10.1016/0044-8486(87)90027-5. dependent effects in recovering Diadema antillarum populations. J. Exp. Mar. Biol. Tuya, F., Sanchez-Jerez,� P., Haroun, R.J., 2005. Influenceof fishingand functional group Ecol. 349, 215–222. https://doi.org/10.1016/j.jembe.2007.05.014. of algae on sea urchin control of algal communities in the eastern Atlantic. Mar. Ecol. Ogden, J.C., Carpenter, R.C., 1987. Species profiles: life histories and environmental Prog. Ser. 287, 255–260. https://doi.org/10.3354/meps287255. requirements of coastal fishes and invertebrates (south Florida): long-spined black Unkles, S.E., 1977. Bacterial flora of the sea urchin Echinus esculentus. Appl. Environ. sea urchin. In: U S Fish Wildl. Serv. Biol. Rep., 82. US Army Corps Engineers, Microbiol. 34, 347–350. pp. 82–84. TR EL. Uthicke, S., Schaffelke, B., Byrne, M., 2009. A boom-bust phylum? Ecological and Ohgaki, S.-I., Tetsuya, K., Kobayashi, N., Tanase, H., Kumagai, N.H., Ishida, S., evolutionary consequences of density variations in echinoderms. Ecol. Monogr. 79, Nakano, T., Wada, Y., Yusa, Y., 2018. Effects of temperature and red tides on sea 3–24. https://doi.org/10.1890/07-2136.1. urchin abundance and species richness over 45 years in southern Japan. Ecol. Vadas, R.L., 1990. Comparative foraging behaviour of tropical and boreal sea urchins. In: Indicat. 96, 684–693. https://doi.org/10.1016/j.ecolind.2018.03.040. Hughes, R.N. (Ed.), Behavioural Mechanisms of Food Selection. Springer-Verlag, Paine, R.T., Tegner, M.J., Johnson, E.A., 1998. Compounded perturbations yield Berlin, pp. 477–514. https://doi.org/10.1007/978-3-642-75118-9_24. ecological surprises. Ecosystems 1, 535–545. https://doi.org/10.1007/ Vadas, R., Elner, R.W., 1992. Plant-animal interactions in the north-west Atlantic. p. 33- s100219900049. 60. In: John, D.M., Hawkins, S.J., Price, J.H. (Eds.), Plant-animal Interactions in the Pearse, J., Hines, A., 1979. Expansion of a central California kelp forest following the Marine Benthos. Clarendon Press, Oxford. mass mortality of sea urchins. Mar. Biol. 51, 83–91. https://doi.org/10.1007/ Valentine, J.F., Jr Heck, K.L., Busby, J., Webb, D., 1997. Experimental evidence that BF00389034. herbivory increases shoot density and productivity in a subtropical turtlegrass Reilly, G.D., Reilly, C.A., Smith, E.G., Baker-Austin, C., 2011. Vibrio alginolyticus- (Thalassia testudinum) meadow. Oecologia 112, 193–200. https://doi.org/10.1007/ associated wound infection acquired in British waters. Guernsey 16, 3. July 2011. s004420050300. Eurosurveillance. Ward, J., Lafferty, K., 2004. The elusive baseline of marine disease: are diseases in ocean Richards, D., Kushner, D., 1994. Kelp Forest Monitoring 1992 Annual Report. National ecosystems increasing? PLoS Biol. 2 https://doi.org/10.1371/journal.pbio.0020120, park service, channel islands national park. Tech Rep CHIS-94-01. 0542–0547. Roberts-Regan, D.L., Scheibling, R.E., Jellett, J.F., 1988. Natural and experimentally Wardlaw, A.C., Unkles, S.E., 1978. Bacteriocidal activity of coelomic fluid from the sea induced lesions of the body wall of the sea urchin Strongylocentrotus droebachiensis. urchin Echinus esculentus. J. Invertebr. Pathol. 32, 25–34. https://doi.org/10.1016/ Dis. Aquat. Org. 5, 51–62. 0022-2011(78)90170-2. Rodríguez, A., Clemente, S., Hernandez,� J.C., Brito, A., García, I., Becerro, M.A., 2017. Weil, E., Torres, J.L., Ashton, M., 2005. Population characteristics of the sea urchin Nutritional, structural and chemical defenses of common algae species against Diadema antillarum in La Parguera, Puerto Rico, 17 years after the mass mortality juvenile sea urchins. Mar. Biol. 164 (6), 127. https://doi.org/10.1007/s00227-017- event. Biol. Trop. 53, 219–231. 3156-3. Wickham, H., 2016. ggplot2: Elegant Graphics for Data Analysis. Springer.