On the occurrence, structure and distribution of deep-water Cystoseira (Phaeophyceae) populations in the Port-Cros National Park (north-western Mediterranean)

Introduction

The large literature dealing with shallow benthic algal assemblages provides considerable knowledge about their species composition, structure, dynamics and distribution. However, deep-water algal assemblages are less well known, especially in temperate waters (e.g. Spalding et al., Citation2003). Furthermore, the few existing studies of deep-water algal communities are focused on the species composition, zonation patterns (e.g. Littler et al., Citation1986; Vadas & Steneck, Citation1988; Spalding et al., Citation2003), or describe benthic algae associated with macrofauna (Sears & Cooper, Citation1978; Lewbel et al., Citation1981; Lissner & Dorsey, Citation1986; Vadas & Steneck, Citation1988; Foster & Schiel, Citation1992). Few studies are focused on deep algal assemblage structure (Ballesteros, Citation1990; Ballesteros et al., Citation1998, Citation2003; Spalding et al., Citation2003), dynamics (Ballesteros, Citation1991a; Garrabou et al., Citation2002) or biogeographic patterns (Hanisak & Blair, Citation1988).

Deep-water algal communities deserve more attention due to their floristic uniqueness (i.e. most of the species are restricted to deep water environments) and ecological and biogeographical interest (Sears & Cooper, Citation1978; Lewbel et al., Citation1981; Lissner & Dorsey, Citation1986; Vadas & Steneck, Citation1988; Foster & Schiel, Citation1992). Furthermore, as they are usually abundant close to the limits of their physiological light compensation point for growth (Markager & Sand-Jensen, Citation1992) deep-water algal species have low community dynamics, with low growth rates, long life-spans and low recruitment rates (Ballesteros, Citation1991b; Ballesteros et al., Citation1998; Garrabou & Ballesteros, Citation2000; Garrabou et al., Citation2002).

Because of their slow demographic changes, these communities may be easily affected by perturbations such as grazing or mechanical losses, changes in water turbidity, or competition with alien invasive species, and thus can be excellent indicators of anthropogenic disturbances (Markager & Sand-Jensen, Citation1992, Thibaut et al., Citation2005).

Species of the genus Cystoseira C. Agardh (Fucales, Phaeophyceae) dominate several Mediterranean rocky-bottom communities (Giaccone, Citation1973). These species are ecosystem engineers (sensu Lawton, Citation1994) as their canopy provides habitat for numerous other algae and invertebrates (Sauvageau, Citation1912; Funk, Citation1927; Rull & Gómez-Garreta, Citation1989; Ballesteros, Citation1990, Citation1992; Serio, Citation1994; Ballesteros et al., Citation1998; Bulleri et al., Citation2002). Although most species of algae inhabit shallow waters, some species only thrive in relatively deep waters where they constitute highly structured and diverse communities (Feldmann, Citation1937; Giaccone & Bruni, 1972; Ballesteros et al., Citation1998). Nevertheless, the composition, structure, dynamics and geographic distribution of these deep-water assemblages are poorly known because of the difficulties of sampling such inaccessible habitats.

Most endemic species of the genus Cystoseira in the Mediterranean and the communities they constitute are highly threatened (Boudouresque et al., Citation1990) and recent studies show that their abundance and distribution has been greatly reduced in the last 60 years (Cormaci & Furnari, Citation1999; Thibaut et al., Citation2005; Serio et al., Citation2006). Anthropogenic disturbances are generally considered to be factors in their decline, although the ultimate causes are seldom understood (Thibaut et al., Citation2005; Serio et al., Citation2006). The recent finding of a deep-water stand of Cystoseira spp. in the Port-Cros National Park (France) (Hereu et al., Citation2003) encouraged us to study the presence, composition and population structure of deep-water assemblages dominated by Cystoseira species in this Marine Protected Area. Our aims were: (i) to report on the abundance and distribution of deep-water Cystoseira species and communities throughout the Port-Cros National Park, (ii) to describe their population structure, and (iii) to incorporate our findings in order to update the known geographical distribution of these species.

Materials and methods

Port-Cros National Park encompasses two small islands of the Hyères archipelago (north-western Mediterranean, France), situated 5 miles offshore (Fig. 1). Twenty-one sites considered likely to be populated by deep water Cystoseira assemblages were a priori selected using recent topographic and bathymetric cartography (Belsher et al., Citation2005). The selected sites had rocky or pebble bottoms, and were preferably isolated rocks or seamounts highly exposed to predominating currents and situated at depths between 25 and 50m (Fig. 1).

Fig. 1. Map of the Port-Cros National Park with submerged rocky bottoms (shading areas). Labels indicate numbered sampling sites.

Display full size

SCUBA dives were undertaken at all sites during April and May 2005, identifying all Cystoseira species in situ and collecting some plants as voucher specimens. Samples were kept in 4% formaldehyde:seawater and stored in personal herbaria. The density and size structure of Cystoseira populations were measured only when plants formed a canopy (i.e. densities higher than 1.25 ind 0.25m⁻²). Densities were measured by sampling haphazardly located 0.25m² quadrats at each site (n>20, except Port-Man n=11). Size structure of the populations was evaluated by measuring the length of the main axis of all plants within the quadrats, a parameter that (i) is easy to measure underwater, (ii) is highly correlated with average yearly biomass and other morphological features of the plant, and (iii) has weak seasonal variation as it is independent of the phenological state of the plant (Ballesteros et al., Citation1998).

Results

We found three different species of Cystoseira: C. zosteroides, C. spinosa var. compressa and C. funkii and determined their distribution and abundance across sampling sites (Table 1).

Cystoseira zosteroides (Turner) C. Agardh

Cystoseira zosteroides was recorded at 15 of the 21 sites visited, distributed all around the islands (Table 1). In the north and west parts of the archipelago it was present only as isolated plants, except on Port-Man rocks (site 16), where it reached a density of 1.02 ind/0.25m² (Table 1). At the Pointe du Vaisseau (site 13) and Ilot de la Gabinière (site 11), this species formed dense populations with 4.5 and 3.82 ind 0.25m⁻², respectively, and in Ilot de la Gabinière it was mixed with isolated plants of C. funkii. In these sites, the C. zosteroides assemblage covered an extensive area (>100m²) on the flat basal part of the rocks surrounded by detritic sand. In the Pointe de la Croix (site 12) we also found a dense assemblage dominated by C. zosteroides, but within a narrow area of less than 10m². Because of this limited distribution we did not measure population parameters (Table 1).

The population size structure in Pointe du Vaisseau (site 13) displayed a log-normal distribution (Table 2), with a peak within the size classes between 5 and 6cm, and a maximum length up to 14cm (Fig. 2). At this site we observed the introduced invasive alga Caulerpa racemosa var. cylindracea (Sonder) Verlaque, Huisman et Boudouresque (10–20% coverage) growing within the population of C. zosteroides. The size structure in Gabinière (site 11) was not log-normal (Table 2), but displayed a peak length around 7–8cm (Fig. 2). In the Roche des Catalans (site 7), an isolated seamount positioned between the islands of Port-Cros and Bagaud at 29–40m depth, C. zosteroides co-occurred with C. funkii and isolated individuals of C. spinosa var. compressa, as described previously by Hereu et al. (Citation2003). This mixed assemblage, with a total density of 3.89 ind 0.25m⁻², covered a rocky platform of around 90m² between 32 and 35m depth (Table 1). Cystoseira funkii was the most abundant species, with densities of 2.66 ind 0.25m⁻², while C. zosteroides density was 1.19 ind 0.25m⁻². Size structure of C. zosteroides showed a log-normal distribution (Table 2), but the peak was at 11–12cm, with a maximum size of 21cm (Fig. 2). In North Bagaud (site 4) and in Roche du Chien (site 10) Cystoseira zosteroides was present but at very low densities in assemblages dominated by C. spinosa var. compressa (Table 1). In the Port-Man rock (site 16), a flat isolated rock at 38–47m depth, an assemblage dominated by C. zosteroides with a density of 1.02 ind 0.25m⁻² was found. This site was unique in the North part of the island, with a dense assemblage dominated by a Cystoseira species.

Fig. 2. Size distributions of the main axis with overlaid log-normal distributions of measured colonies of Cystoseira zosteroides, C. funkii and C. spinosa in sampled sites. Distributions with less than 10 measures are not shown.

Display full size

General pattern of distribution

Plant densities and species composition were not homogeneous; densities were low in the north side of the island, where all species were present, while in the south and west sides of the islands high-density populations were common. The distribution of the different species also changed: C. zosteroides was dominant on the SE side of the island, while C. spinosa var. compressa and especially C. funkii dominated in the SW and W areas (Fig. 3).

Fig. 3. Map of the Port-Cros National Park with numbered study sites, where densities of the three Cystoseira species are represented. Diameters of the circles are proportional to the maximum densities measured: small circles: 0–1.25ind 0.25m⁻²; medium circles: 1.25–2.5ind 0.25m⁻²; big circles: >2.5ind 0.25m⁻²; triangles: presence. Specific values are shown in Table 1.

Display full size

Discussion

The deep-water macroalgal assemblages in Port-Cros National Park are abundant and diverse. Although a large number of studies have been performed describing benthic assemblages of Port-Cros National Park (e.g. Augier & Boudouresque, Citation1967, Citation1974, Citation1975, Citation1976, Citation1978), deep-water Cystoseira communities have not previously been recorded.

Cystoseira zosteroides is relatively common in Port-Cros, especially on the east side of the island. It was previously recorded from Port-Cros on detritic bottoms around the island (Pérès & Picard, Citation1963) and observed rarely (Coppejans, Citation1977; Hereu et al., Citation2003). This species is distributed throughout the Mediterranean (Serio, Citation1995a) and can be the main species of communities occurring on rocky bottoms exposed to strong unidirectional currents and light ranging between 1% and 0.3% of surface irradiance (Giaccone & Bruni, Citation1973).

Cystoseira spinosa var. compressa was considered as a rare species in Port-Cros, being recorded as C. spinosa by Pérès & Picard (Citation1963) from detritic bottoms, and by Augier et al. (Citation1971), Belsher et al. (Citation1976) and Augier & Boudouresque (Citation1978) in several deep rocky bottoms between 36 and 45m depth. We found it mainly on the south and west sides of Port-Cros. This species is known to thrive on rocky reefs at depths ranging from 20 to 50m (e.g. Giaccone, Citation1973, Ballesteros et al., Citation1998). Cystoseira spinosa is distributed throughout the Mediterranean (Serio, Citation1995b) and can be the main taxa in communities that occur on Mediterranean rocky bottoms with light ranging between 2 and 10% of surface irradiance (Giaccone & Bruni, Citation1973).

Cystoseira funkii has been commonly misidentified as Cystoseira jabukae Ercegovic, but its taxonomic position has been validated by Verlaque et al. (Citation1999) who reported its distinctive features. The distribution of this species covers the Western Mediterranean and the westernmost part of the Eastern Mediterranean (Verlaque et al., Citation1999; Hereu et al., Citation2003; Thibaut et al., Citation2005). In this study we located several sites with dense stands of C. funkii, mainly distributed at the southwest side of the islands.

Our data indicated high spatial variability in the composition and structure of deep-water Cystoseira communities at relatively small geographic scales (hundreds to thousands of metres) in contrast to other studies that describe relatively homogeneous patterns in deep-water algal composition and distribution (Hanisak & Blair, Citation1988; Spalding et al., Citation2003). Unfortunately, at this time we do not know the causes of this heterogeneity. In this respect, depth (between 32 and 48m) and substratum (rocky bottoms) are similar at our sampling sites and anthropogenic disturbances (i.e. sewage, anchoring and small-scale, ‘artisanal’ fishing) and herbivory are low, or almost absent at all sites. The number of inhabitants of Port-Cros is around 40 people in winter, reaching a maximum of around 140 in summer and sewage treatment was implemented from 1999. Anchoring zones are outside the locations hosting Fucales, and although small-scale local ‘artisanal’ (net) fishing is permitted, its pressure is very low (a maximum of six fishermen in summer, one in winter) and is strictly regulated (PNPC, Citation2005). Diving is limited to few sites with organized mooring and is a significant distance from the majority of sampling sites. Densities of the herbivorous sea urchin Paracentrotus lividus Lamarck, 1916 are low in Port-Cros (Hereu et al., Citation2005) as its Mediterranean distribution is usually limited to shallower water environments (Chelazzi et al., Citation1997). In addition, many of the rocks were deep and isolated by surrounding sediment, thus having a low probability of being colonized by sea urchins. Sea urchins were absent at the study sites (authors' pers. obs.).

We hypothesize that differences in composition and structure of Cystoseira populations in Port-Cros National Park could be related to currents. The Port-Cros area lies within a branch of the liguro-provençal current, a general east to west geostrophic current that is confined between the Hyères archipelago (including Porquerolles, Bagaud, Port-Cros and Levant islands) and the continental coast. This current flows through the channels between the islands generating a small-scale pattern of high currents and counter-currents around the islands (Jeudy de Grissac, Citation1982). In fact, C. zosteroides, a species that thrives in strong currents (Giaccone et al., Citation1994), mainly occurs at the east side of Port-Cros where the current is funnelled by the presence of Levant Island to the East of Port-Cros. Cystoseira spinosa and C. funkii were mainly found in areas not directly exposed to the predominant east to west geostrophic current, in accordance with observations by Giaccone et al. (Citation1994). Other causes, such as patchy recruitment, or other local processes affecting these isolated populations, could also be important in determining composition and structure. However, these hypotheses need to be tested in the future by recording and describing accurately the current patterns at the different sites.

Apart from these differences, these assemblages had common characteristics in population structure that reflect low community dynamics. Plant sizes within stands of the three species were log-normal distributed, with a relatively high presence of old individuals. Ballesteros et al. (Citation1998, Citation2003) described a similar pattern for deep-water Cystoseira assemblages off Scandola (Corsica), hypothesizing that the log-normal distribution may result from intraspecific competition, where a canopy of older individuals inhibits juvenile recruitment. Recruitment would occur in pulses after a disturbance strong enough to remove part of the population (catastrophic storms, enhanced grazing). It has been reported from terrestrial studies that plant populations at equilibrium generally follow negative exponential distributions (Lorimer, Citation1985; Edmond et al., Citation1992; Berg & Harmig, Citation1994); however, in communities with strong herbivore pressure on small plants (Crisp & Lange, Citation1976; Frelich & Lorimer, Citation1985), or with strong inter-annual variations in recruitment rates (Tyrrell & Crow, Citation1994) the type of distribution can be normal or log-normal. An alternative explanation could be that recruitment is inhibited by filamentous algae such as the alien taxon Womersleyella setacea (Hollenberg) R.E. Norris (Ballesteros et al., Citation1998), an invasive species that has been found at our sampling sites (authors’ unpublished data). Nevertheless, density-dependent competition is the scenario that could better explain the log-normal size distribution. This distribution has also been found in other sites (Arrighi, Citation1995; Ballesteros et al., Citation1998, Citation2003) and seems to be a general pattern for these deep-water Cystoseira communities. Log-normal size structure distribution is also related to low population dynamics, since recruitment is low and there is an accumulation of medium and high size class individuals, with long-life spans. This pattern is supported by Ballesteros et al. (Citation2003) who describe very low annual growth rate of the main axis both for C. spinosa var. compressa and C. zosteroides (0.70cm year⁻¹ and 0.46cm year⁻¹ respectively), indicating that the longest specimens can reach ages of more than 50 years old. This finding agrees with some studies that describe a decrease of turnover and growth rates with depth (Bak & Luckhurst, Citation1980; Hughes & Jackson, Citation1985; Huston, Citation1985; Ballesteros, Citation1991b; Garrabou et al., Citation2002). These low dynamics reinforce the idea of the fragility of these communities since algae living in very deep waters are usually at their physiological light compensation point for growth, with no surplus energy to balance grazing, or mechanical losses (Markager & Sand-Jensen, Citation1992).

In conclusion, we have shown that deep-water Cystoseira species in Port-Cros have characteristics that are species and site specific. All species tend to have log-normal distributions in size structure and most species distributions vary at different sites. Because of the logistic problems of obtaining data due to the depth of the sampling sites, size-frequency distributions provide the only easily available indicators to assist in understanding the underlying dynamics of growth, survival and recruitment, since long-term data are lacking for most algal assemblages. Similar approaches and analyses have been used for other organisms, such as corals (Meesters et al., Citation2001), where they can detect differences in populations structure over gradients of environmental and anthropogenic disturbances.

Recent studies comparing present and historical data show a decline of Fucales in different areas of the Mediterranean (Thibaut et al., Citation2005; Serio et al., Citation2006). Furthermore, deep-water Cystoseira communities have disappeared over the last 30–40 years from several localities where they were previously known to exist, both on the continental coast of France and Spain (Thibaut et al., Citation2005; authors’ pers. obs.) and on isolated islands (Cormaci & Furnari, Citation1999; Serio et al., Citation2006). These assemblages still exist in the National Park of Port-Cros, suggesting that this area has not been affected by the causes of decline in other localities. Nevertheless, because the distribution of these deep-water communities is still largely unknown and Port-Cros has benefited from protected status for more than 40 years, further research is needed to determine if Port-Cros National Park is an exceptionally isolated site, or if these communities are also common in other places. Moreover, the previously unnoticed existence of these deep-water, mature and vulnerable algal populations in one of the best studied Mediterranean areas emphasizes the necessity of increasing basic research activities in coastal deep waters in order to increase knowledge of Mediterranean biodiversity and underwater seascapes, as a necessary step to guarantee conservation. Monitoring endemic species must be implemented in the localities where they are known to exist, even in protected, highly restricted areas, since non-controlled environmental changes such as the introduction of invasive species, or climate change will almost certainly have large effects on these assemblages in the future.