Discussion

  The contributions above constitute an unprecedented joint effort of the Centre of Advanced Studies of Blanes to answer a question of scientific and social concern: Which was the impact of the extreme storm of 26 December 2008 on the marine biota of the Catalan Coasts? Some of the studies were initially designed well enough to properly answer this question (quantitative monitoring programmes), some others lacked the ideal spatial-temporal sampling effort, and some others did not have information on the immediately ‘before the storm condition’ to allow for a rigorous assessment of the impact of the storm.

On spite of these limitations, this project represents the most accurate and broadest approach to the study of the impact of an extreme storm on Mediterranean coastal natural communities to date. It has provided a wealth of information about the mechanisms and factors that ultimately determine the damage caused by an extreme hydrodynamic effect on the coastal biota. Despite of the highly diverse response of the various populations and communities studied, in the overall the analysis has revealed a remarkable resistance, in some cases, and resilience, in some others, of the shallow water coastal biota to the extreme perturbation of Sant Esteve’s Day 2008.

Wind, waves and shear stress

Using the integrated wave power during the storm duration (see Chapter 1), three different areas were identified: (i) the northernmost part of the Catalan coast (Costa Brava) – characterised by the Roses and Palamós buoys – where the greatest wave power, the largest wave heights, and the longest storm duration were recorded, and where the storm first began. Is in this area where the storm can be classified as extreme following the storm classification of Mendoza and Jiménez (2008). This storm was also the largest ever recorded by those two buoys. Maximum shear stress was above 70 Nm-2 at 5 m depth; (ii) the central coast – characterised by the Tordera and Llobregat buoys – where the wave power decreased down to about half of that in the northern area, with smaller wave heights (although still large), a slightly shorter duration and slightly later impact. Maximum shear stress was between ca. 50 and 70 Nm-2 at 5 m depth; and (iii) the southernmost area (the rest of the Catalan coast), where the wave power was only one third of that of the northern area, with relatively small wave heights. Maximum shear stress below ca. 50 Nm-2 at 5 m depth.

It seems that, in recent times, the previous comparable storm in the Catalan coast took place in 1947, i.e., 61 years before the extreme event of Sant Esteve 2008 (Camuffo et al. 2000). Since 1958, when storms begun to be recorded instrumentally, only an event from 10 to 16 November 2001, with maximum wave heights of 10m, could be compared in intensity with the one object of this project. No temporal patterns or correlations with global atmospheric events have been found to be associated with these type of events, making them unpredictable.

Against intuition, the most important damage was not a direct consequence of the  hydrodynamic shear stress, but of the impact and abrasion caused by the relative movement of the substrate particles surrounding the organisms. Those growing on stable rocky substrates resisted quite well the impact while sessile organisms growing among boulders or on the sand suffered the highest damage due to abrasion, burial, dislodgement or uprooting. What follows is a brief synthesis and discussion of the various types of impacts, mechanisms and factors identified by the various contributors to this report.

Impact: a complex concept

The impact on the biota can be understood as any change in abundance, biomass, or functioning of a certain assemblage of organisms following a disturbance. The definition of this last concept, disturbance, although a constant topic of debate, is generally assumed to be any unpredictable, episodic event, caused by external agents, that changes temporarily or permanently the reference state of an ecosystem, community, or population (Montefalcone et al. 2011). These changes are usually referred to as ‘perturbations’. The terms impact and perturbation are often used indistinctly.

The changes in those parameters (abundance, biomass, or functioning) following a disturbance, can be very evident or very difficult to detect and quantify. Such difficulty often lies in the complexity of the interactions in the biota and in the possible delay of the actual effects of the perturbation. For instance, it has been shown how the algal cover has been denuded in up to 94% in some locations (Hereu et al., Chapter 9). The immediate impact is undeniably extreme in such locations, but depending on the recovery capacity of the various algal species, the ‘effective’ impact could be very small or null if, after a few weeks or months, the denuded surfaces have been totally recolonized by the same species (see further). An example at the other end are the populations of Paramuricea clavata, a long-lived soft coral with low mortality rates but also very low rates of recruitment. The 13.4% (up to 20.6% in some stations) of individuals of P. clavata lost as a consequence of the storm can be, therefore, considered as a potentially very severe impact with possible unsustainable long-term effects on the population (Coma et al., Chapter 3). Examples closer to this same end are those represented by other long-lived, low-recruiting benthic sessile species like the seagrass Posidonia oceanica (Alcoverro et al., Chapter 11), the deep growing algae Cystoseira zosteroides (Navarro et al., Chapter 10), and the fan mussel Pinna nobilis (Garcia-Rubies et al., Chapter 12).

Facilitation, i.e., a positive effect on the abundance of the initially affected population can be another outcome of the perturbation. In the same case mentioned above about algal populations, it was also shown how sea urchin populations were also devastated by the storm. If recolonization rates of the algal cover is faster than that of sea urchins, the area recolonized by the algae, in the absence of their main grazers, might be even higher than that occupied before the storm (Hereu et al., Chapter 9 and Hereu and Garcia-Rubies, Chapter 3). This example highlights the importance of the continuation of the monitoring studies after passage of the disturbance. Depending on the temporal scale at which we observe the effects of the storm, these can be from severely negative (short-term) to positive (longer term) for a certain population.

One more example of facilitation serves also to illustrate how complex can be at times to neatly establish the impact of a disturbance. In the study by Uriz et al. (Chapter 6) it is shown how grazing on photophilic communities facilitates growth of bioeroding sponges because they contain photosynthesising zooxanthellae that compete for light. Grazers and sponges abundances were positively correlated before the storm as grazers and canopy-forming seaweeds were negatively correlated. The storm destroyed both the algal and grazer components and any previous significant correlation between them and the sponge population. So the evolution of this last one will depend on the dynamics of recovery of the former, illustrating how impacts can take place through cryptic/hidden/unobvious cascading effects.

One of the contributions of Garcia-Rubies et al. (Chapter 15), has provided a good example of the Intermediate Disturbance Hypothesis (Connell 1978). The diversity of local species in the North Costa Brava (Medes-Montgrí) has been maximized in some locations as a consequence of the storm. At very low or high levels of disturbance diversity tends to decrease following inter-specific competition or extinctions (respectively). At intermediate levels of disturbance, diversity can result maximized because both competitive K-selected and opportunistic r-selected fish species can coexist.

An enhanced recruitment in some fish species has also been observed as an indirect impact of the storm. This is the case of Serranus cabrilla (Garcia-Rubies et al., Chapter 15), where a significantly higher recruitment success could be attributed to a severe rarefaction of large individuals in shallow waters following the storm (Garcia-Rubies, 1997 found that adults prevent recruitment in already saturated populations inside protected areas). A similar effect could be attributed to the rarefaction of natural predators of other species (Symphodus spp.). 

An apparent increase in adults of Phycis phycis (forkbeard, brótola de roca) and Palinurus elephas (spiny lobster, langosta) has also been reported in this study as suggested by an increase in catches by the artisanal fleet of Palamós, Costa Brava (Gordoa and Dimitriadis, Chapter 17). However, the coupling of the life cycle of these species and the timing of the response in the catches after the storm cannot easily explain the observed increase in the abundances. Gordoa and Dimitriadis hypothesise that habitat destruction and the concomitant increase in accessibility to fisherman may be behind the increase observed (see next section). Once again, if the evolution of these catches is not followed in a sufficiently long time frame, we may conclude a positive (false) or a negative (real) impact of the storm on these populations when combined with human activities.

Mechanisms of damage

As mentioned before in this discussion, the main mechanism of damage observed in this work was not the direct hydrodynamic force (shear stress) exerted by the waves on the organisms. Instead, it was the movement transmitted by this hydrodynamic force to the particles of various sizes in the substrate underneath or near the organisms (sand, cobbles, and rocks). Direct impact of big-sized rocks, abrasion by finer-sized particles, burial by sand, dislodgment or uprooting, and removal from habitat have been all observed in the different studies in this report.

The only case where a direct effect of the shear stress was observed was that of the soft coral Paramuricea clavata. Both detachment off the rocks or direct damage of their delicate polyps was observed (Coma et al., Chapter 4). Although not rigorously assessed, the also fragile gill and filtering elements of the fan mussel Pinna nobilis may have been directly damaged by the energy of the waves (Garcia-Rubies et al., Chapter 12).  

Among the indirect mechanisms, the one with the most striking effect was the removal of many fish species from their habitat as being washed ashore. Thousands of individuals were found on the beach of L’Estartit (North Costa Brava) and others; among them Anthias anthias and Chromis chromis, which mostly move freely in the water column with little contact with the bottom, were the most abundant. More sedentary fishes were also found on the beach, including labridae (mainly Coris julis) and Serranidae (mainly Serranus cabrilla, but also some specimens of Epinephelus marginatus). Whether the individuals were or not killed by the storm before being stranded cannot be established. High fish mortality directly provoked by wave action has been rarely reported (Pfeffer & Tribble, 1985, but see Robins, 1957) and direct mortality does not seem extremely important in many cases (Walsh, 1983; Fenner, 1991, Adams & Ebersole, 2004) but habitat shifts after the storm have been observed (Kaufman, 1983) and the main changes in fish assemblage have been documented to be mostly related to changes in habitat composition (Kaufman, 1983; Fenner, 1991; Aronson et al., 1991).

In some areas of Medes-Montgrí, horizontal rocky substrata covered almost completely by a leafy photophilic algal community, appeared virtually denuded by abrasion after the storm as a consequence of the relative movement of boulders (Hereu et al., Chapter 9). The same was observed for the deep-growing algae Cystoseira zosteroides (Navarro et al., Chapter 10) or for sea urchin populations (Hereu and Garcia-Rubies, Chapter 3). A vivid demonstration of the crucial role of the stability of the substrate is given by the contrasting effects observed in the upper sub-littoral algal community studied by Torras and Garcia-Rubies (Chapter 8). They found no overall significant effect of the storm. Even though the degree of exposure of this community was maximal, no destruction was observed. The differential factor explaining this phenomenon is the virtual lack of mobile particles in the shallow-most community.

Burial and dislodgment or uprooting were the main mechanisms of damage and death for two common dwellers of Mediterranean shallow sandy bottoms such as the seagrass Posidonia oceanica (Alcoverro et al., Chapter 11), the giant mussel Pinna nobilis (Garcia-Rubies et al., Chapter 12), and, probably, Paracentrotus lividus. Specially in shallow environments, sand was eroded by the action of the waves dislodging seagrass entire plants and fan mussels that were subsequently detached for the bottom and transported elsewhere. Once these organisms loss there original insertion on the substratum, they are no longer viable and die. The eroded sand is moved to deposition areas where it can suffocate both P. oceanica and P. nobilis. The study by Alcoverro et al. (Chapter 11) reported that P. oceanica showed a relatively low tolerance to burial. Beyond just 4 cm of burial, the plant quickly began to lose 50% of their shoots, and beyond 6 cm, mortality was nearly complete. Similarly, the study performed in Giverola Cove (Costa Brava) showed that out of the 53% of the individuals of P. nobilis that disappeared, a similar proportion died due to partial or to total burial (Garcia-Rubies et al., Chapter 11). Within the study on P. oceanica (Alcoverro et al., Chapter 11) a reduction by 46% of the P. lividus population living in the meadow at Fenals is also reported. The absence of damaged individuals in the area suggests that the reduction observed might be a consequence of burial.

In this study only the “immediate” direct and indirect effects on nearshore populations was addressed. However, also secondary effects may become evident in the future. Among them, ecosystem state shifts may occur in the rocky bottoms denuded by the storm; a delayed reduction in sea urchins abundance may occur following the observed reduction in the reproductive output as larger sizes of the population have been more affected; a reduction in abundance may also have take place due to the deterioration of the habitat that has made a species more vulnerable to predation; in turn, some species may result favoured by a higher accessibility to preys (this might the case of the artisanal fishermen of Palamós and the increased catches of spiny lobster). See also Comments on ecosystem recovery three years later.

Factors

This project has provided a wealth of excellent examples of factors that modulate the extent of the impact caused by an extreme event. The most obvious and determinant was the distribution of the intensity of the storm, that resulted in much higher shear stress in the Costa Brava than anywhere else along the Catalan Coast (Fig. 1, General Introduction). Figure 2 in Hereu et al. (Chapter 9) provides a perfect example of the direct link between the shear stress and the impact on the benthic communities (algal communities in this case). Very roughly, they found an increase in the algal cover removal of 50% every increase of 50 Nm-2 of maximum shear stress.

Apart from the latitudinal distribution observed, the degree of exposure, i.e., orientation respect to the dominant wave direction and the depth played also a crucial role. A clear example of the role of orientation was given by the monitoring of the Pinna nobilis populations in the two locations studied (Garcia-Rubies et al., Chapter 12). In Giverola Cove, the population living in the area around the northern-most transect experienced 4 times less mortality than the one living the southern-most transect. Both transects were separated only around 20 m, but a projection of the coastline at the north of the cove provided a remarkable higher protection to the individuals growing in the northern-most transect (see Fig. 1 in Garcia-Rubies et al., Chapter 11). The example in Medes Islands was even more extreme. All 21 individuals labelled in Salpatxot (almost totally exposed to the east) disappeared after the storm, while none of the individuals in Embarcador (totally at the leeward of the waves, in the west side of the Meda Gran) died or was lost.

As a norm, a higher depth offered higher protection to the organisms, although sometimes the vicinity of an unstable bottom to deep populations resulted in higher impacts than to shallower ones (see further). For example, mobile species such as Anthias anthias and Chromis chromis were the most abundant of marine fish species found stranded, both species inhabiting shallow areas of the water column. More sedentary fishes with a much closer relationship with the bottom (category 5, after Harmelin, 1987) were also found, but in much lower amount, on the beach, including labridae (mainly Coris julis) and Serranidae (mainly Serranus cabrilla, but also some specimens of Epinephelus marginatus). Only two species of the most bottom-related species (category 6), were found on the beach. Although not consistent for all the locations sampled, Posidonia oceanica meadows growing below 10-15 m showed a 3-4 times lower impact than those growing at around 5m of depth (Alcoverro et al., Chapter 11).

The characteristics of the bottom surrounding the organisms, particle size, composition, and slope, were determinant. The large differences in the impact on photophilic algal communities in Medes-Montgrí were explained by the type of substrate (Hereu et al., Chapter 9). While most of the stations on the Montgrí coast were characterised by bottoms constituted by big rocky boulders (Falaguer, Dui, Molinet), some sampling stations at the Medes Islands (Ferranelles, Pedra de Deu) had a homogeneous rocky substrate. The storm impact was much less severe on populations growing on these last type of substrates than in the former. As discussed before, abrasion by the moving particles was the main mechanism explaining this observation.

A very illustrative example of the influence of the nature of the substrate is given by the pioneering study on Cystoseira zosteroides by Navarro et al. (Chapter 10). The depth and the physical and biological characteristics of the substrate around the population all played an important role in determining the impact of the storm in this species. The population in the Tascons site displayed only a 14% mortality, probably due to its deeper situation and the geomorphology of the bottom, devoid of boulders. In Ferranelles, however, the abundant boulders over a flat area devastated the algal cover. Some six months after the storm, however, the new surface made available was already largely colonized by new propagules of C. zosteroides. In contrast, in the location Tomas Ras, which bottom was still covered by encrusting corallines filamentous understory algae and some adult individuals of C. zosteroides. After the storm, there was a much lower density of recruits. In this example, depth was controlling the intensity of the disturbance, the physical features of the substrate (slope, size of particles, and mobility) modulated the impact, and the community composition that persisted after the storm controlled the post-event recruitment.

The size and morphology of the organism, and its means of attachment to the bottom played certainly a role. For instance, the small, flat and excavating individuals of the sponge Cliona viridis resisted perfectly the hit of the storm even in very shallow environments (Uriz et al., Chapter 6). In contrast, the arborescent and relatively fragile body of the gorgonian Paramuricea clavata offered much less resistance to the shear stress of the storm even though the population studied was growing at 16-20 m of depth (with up to 20.6% of mortality; Coma et al., Chapter 4).

Hereu and Garcia-Rubies (Chapter 3) put in evidence very neatly how the bigger sizes of the sea urchin Paracentrotus lividus were more affected by the storm than the smaller ones. In the overall, the number of individuals between 4.5 and 5 cm added up about 240 before the storm and dropped down to less than 20 after the storm. The small sea urchins were probably able to hide better in small crevices of the rocky substrate being less vulnerable to detachment and abrasion.  This must have been also the case for the upper sublittoral blennioid assemblage, which withstood remarkably well the effects of the storm (Garcia-Rubies and Macpherson, Chapter 16). Conversely but understandably, a bigger size resulted advantageous for Pinna nobilis. The individuals that perished by the storm, were significantly larger than the survivors (Garcia-Rubies et al., Chapter 12). At the population level, a significant reduction of the reproductive output of the population (bigger sizes of sea urchins), may affect the succession of sublittoral macroalgal communities, or bring to a critical threshold the populations of species threatened with extinction.

The shallow photophilic communities in the absence of a mobile substrate around them, resisted fairly well the intensity of the storm probably owing to evolutively selected body shapes, textures, flexibility, and means of attachment (e.g., Torras and Garcia-Rubies, Chapter 8, and Villamor et al., Chapter 7). An example of this robustness that stands out is that represented by sponge populations in general. Maldonado (Chapter 5) provided a very illustrative example by monitoring labelled sponge populations of 3 different sites of the Costa Brava. He showed how despite of large differences in exposure, depth distribution, and body shapes and textures, the mortality among sponge species was low (6.6% on average). He suggests that collectively, the results of his study suggest that most of the monitored sponges have evolved effective anchoring mechanisms to withstand the forces generated during heavy, periodic storms.

This project has also highlighted the importance of population dynamics in modulating the overall effect of the storm. For instance, the low natural mortality rate of large colonies, such the gorgonians, plays a crucial role in attenuating the effects on the population of long episodes of low recruitment and against high mortality rates of small colonies. This is why even a small increase in mortality rate of large colonies may produce unsustainable long-term effects on the population (Coma et al., Chapter 4). It is also obvious that the lower the turnover of the organism, the lower the recovery capacity and therefore more severe the effects of an extreme perturbation. In these terms, among the species studied, the soft coral P. clavata, the giant mussel P. nobilis and the seagrass P. oceanica are the species to be considered as the most severely affected by the storm of Sant Esteve’s Day 2008.

Comments on ecosystem recovery three years later

Observations and surveys carried on in 2010 and 2011 confirm the slow (years: algal cover, sea urchins), very slow or null recovery (decades: seagrass meadows, gorgonians, fan mussels) predicted for those populations and communities that were severely impacted by the storm. B. Hereu reports interesting observations regarding the algal community and sea urchin populations in Medes-Montgrí. The rocky bottoms denuded by the storm was recolonized (100% of it) but by pioneering species such as Dictyota spp. or Padina pavonica.  The perennial species that characterise the mature algal community of the Costa Brava rocky bottoms (with a thick basis of Corallinaceae and Cystoseira spp.) will still take years to regain their space. Hereu highlights the importance to distinguish between specific and functional diversity. While the later can be considered as fully recovered, the former, i.e. the community with the original species, will still need to undergo a slow succession process even in such a dynamic assemblage. This will be specially true for the deep-growing Cystoseira zosteroides as a consequence of its low growth rate (adult individuals may exceed 50 years of age).

With respect to the population of Paracentrotus lividus, B. Hereu reports profound changes in its structure two years after the event. Despite the recruitment peak observed in 2010 at some sites, they believe that the affected populations will not recover for many years because of the relatively low growth rate and the limited migration capacity of this species. Because inside the marine reserve (Medes) the predation pressure on juveniles is higher, and the recruitment peaks are more attenuated respect the non protected area (Montgrí), they predict that the recovery of sea urchins populations inside the reserves will be slower than outside the reserves.

I. Uriz reports total normality of the shallow water sponge community. They comment that the ‘a priori’ more vulnerable Hemimycale spp. seem to be totally unaffected or even in a slightly better condition than in previous years.

R. Coma reports that, although the data obtained during the two last years on the state of the populations of P. clavata have not been yet analysed in detail, no obvious changes (recovery or recruitment) can be reported.

In the case of the seagrass P. oceanica, A. Gera reports that the forecast mortality in areas buried or uprooted after the storm unfortunately has been realized in most of the cases. All those shoots buried more than 10 cm showed a 100% of mortality a few months after the storm. Those shoots buried 5-10 cm, showed an apparent dependence on sediment dynamics and on the size of the patch (in this last case probably associated to the possible resources subsidy from the unburied shoots next to the buried ones). For this range of burial, some meadows are still very affected (Canyelles Cove, at 18 m), in some others mortality seems to have diminished (Fenals Cove at 8 m), and in some others new recruitment seems to have taken place (Giverola Cove at 7 m). Those shoots buried 0-5 cm, present a very wide range of impact after the storm as a function of the site and the granulometry of the sediments. Data are still too preliminary and need a detailed analysis before any sound conclusion can be drawn.

5. E.Jordana confirms that no changes in the soft bottom macroinvertebrate communities attributable to the storm have been detected in the data collected in 2010. Some changes observed both in the granulometry and the organic matter content of the sediment or in the recruitment of some bivalves, are within the normal interannual range of natural variation.
   

Literature cited

Adams A.J., Ebersole J.P. (2004) Resistance of coral reef fishes in back reef and lagoon habitats to a hurricane. Bulletin of Marine Science 75: 101-113.

Aronson R.B., Sebens K.P., Ebersole J. P. (1991) Hurricane Hugo’s impact on Salt River submarine canyon, St. Croix, U.S. Virgin Islands. In Ginsburg,  R.N. and Smith,  F. G. W. (eds). Proc. Colloq. on global aspects of coral reefs, Miami, 1993.

Camuffo D., Secco C., Brimblecombe P., Martin-Vide J (2000) Sea storms in the Adriatic sea and the western Mediterranean during the last millennium. Climate Change 46: 209–223.

Connell J. H. (1978) Diversity in tropical rain forests and coral reefs. High diversity of trees and corals is maintained only in a nonequilibrium state. Science 199: 1302-1310.

Fenner D.P. (1991) Effects of Hurricane Gilbert on coral reefs, fishes and sponges at Cozumel, Mexico. Bulletin of Marine Science 48: 719-730.

Garcia-Rubies A. (1997) Estudi ecològic de les poblacions de peixos litorals sobre substrat rocós a la Mediterrània occidental: efecte de la fondària, el substrat, l’estacionalitat i la protecció. PhD. Thesis, Universitat de Barcelona, Spain.

Harmelin J.G. (1987) Structure et variabilité de l’ichthyofaune d’une zone rocheuse protegée en Mediterranée (Parc National de Port-Cros, France). P. Z. N. I. Marine Ecology 8: 263-284.

Kauffman S.A. (1993) The origin of order: self-organization and selection in evolution. Oxford University Press.

Kaufman L.S. (1983) Effects of Hurricane Allen on reef fish assemblages near Discovery Bay, Jamaica. Coral Reefs, 2: 43-47.

Montefalcone M., Parravicini V., Bianchi C.N. (2011) Quantification of coastal ecosystem resilience. In: Wolanski E., McLusky D.S. (Eds), Treatise on Estuarine and Coastal Science, Elsevier, USA, 10(3): in press.

Pfeffer R.A., Tribble G.W. (1985) Hurricane effects on an aquarium fish fishery in the Hawaiian Islands. Proceedings of the Fifth International Coral Reefs Congress, Thaiti, 27 May-1 June, 1985. Vol. 3, 331-336.

Robins R.C. (1957) Effects od storms on the shallow-water fish fauna of southern Florida with new records of fishes from florida. Bulletin of Marina Science 7: 266-275.

Walsh W.J. (1983) Stability of a coral reef fish community following a catastrophic storm. Coral Reefs 2: 49-63.