Marine Biology

, Volume 161, Issue 5, pp 1015–1026

Demographic parameters of two populations of red coral (Corallium rubrum L. 1758) in the North Western Mediterranean

  • L. Bramanti
  • I. Vielmini
  • S. Rossi
  • G. Tsounis
  • M. Iannelli
  • R. Cattaneo-Vietti
  • C. Priori
  • G. Santangelo
Original Paper

DOI: 10.1007/s00227-013-2383-5

Cite this article as:
Bramanti, L., Vielmini, I., Rossi, S. et al. Mar Biol (2014) 161: 1015. doi:10.1007/s00227-013-2383-5

Abstract

The demographic and reproductive structure of populations represents the main data set needed for conservation and management plans. Mediterranean red coral Corallium rubrum has been exploited for 2,000 years, but only recently management plans have been solicited by the international community. We examined and compared the demographic features of two red coral shallow populations located in distinct geographic locations: Portofino (Italy) and Cap de Creus (Spain). Adults and juveniles density, growth rates, population size and age structure, fecundity and fertility were examined. Juveniles were the dominant class (33 %) in both populations. The analysis of the gamete content of 653 colonies revealed that the populations have balanced sex ratios and similar fertility and polyp fecundity. The average annual growth rate, determined on 119 colonies by annual growth rings count, was similar in both populations (0.24 mm year−1), decreasing with colony age. Maximum life span of 99 % of the colonies was 60 and 40 years at Portofino and Cap de Creus, respectively. Minimum harvestable size (7 mm basal diameter) was reached in 30–35 years, and the percentage of colonies above it was 6.7 % at Portofino and 2.1 % at Cap de Creus, where juvenile and adult colony densities were significantly lower and the percentage of commercial-sized colonies reduced by 25 % in just a few years. Notwithstanding similar growth and fecundity, the two populations showed different densities and size/age structures suggesting local factors, together with different fishing pressures, have to be taken into account in the management plans for this species.

Introduction

Population dynamics depend largely on demographic histories (Dobson 1998; Ebert 1999, particularly for slow-growing, long-lived species, which have been subjected to dramatic reductions in survival, reproduction and/or recruitment rates (e.g., Fujiwara and Caswell 2001; Santangelo and Bramanti 2006; Arrigoni et al. 2011). Demographic studies can supply clues to improving conservation and management plans, providing information on past dynamics and future trends of populations. In particular, population size frequency distribution, demographic traits (e.g., sex ratio and population density) and life history variables (e.g., growth rates, age and fecundity) can be used to assess population performances, evaluate the current status of a population and to investigate spatial differences between population (Table 1; Ebert 1999; Beissinger and McCullough 2002).
Table 1

Summary table showing how information on population structure (skewness), demography (density and sex ratio) and life history variables (growth rate, age and fecundity) have been used so far in other marine species to define population performance, current status of the populations and spatial differences. In the fourth column some studies in which these applications have been used are reported

Data used

Objective

Population performance

Population status

Spatial differences

Bibliography

Size structure

Skewness can be used as an indicator of young/olds ratio (positive = more youngs; negative = more olds)

Missing age/size classes in the population structure could be indication of a past disturbance

Population dwelling in different locations can have different size structures

Rossi et al. (2012)

Goffredo and Chadwick-Furman (2003)

Linares et al. (2010)

Demography (density, sex ratio)

Unbalanced sex ratio could determine low reproductive output

Low density could determine poor performance

Unbalanced sex ratio could be indicator of disturbance

Low densities are associated to poor population status

Population in different locations can have different sex ratio and/or densities

Bramanti et al. (2009)

Tsounies et al. (2006b)

Priori et al. (2013)

Life history (growth rate, age, fecundity)

Low growth and fecundity rates and short life span (low maximum age) could determine poor performances

A population in good status display long life span (high maximum age) and high fecundity and growth rates

Population in different locations can show different growth rates, fecundity and maximum age

Bramanti et al. (2005)

Torrents and Garrabou (2011)

Santangelo et al. (2007)

Demographic population models, based on life history tables and transition matrixes, combine the informations from population structure, demography and life history traits and have been applied to conservation of several marine species: e.g., the loggerhead sea turtle Caretta caretta (Crouse et al. 1987) and the Right whale Eubalena glacialis (Fujiwara and Caswell 2001). Matrix population models account for recruitment, reproduction and demographic parameters and represent an advanced tool for conservation with respect to classical fishery models based on yield per recruit (Beverton and Hollt 1957). The demographic features and life history variables used to parameterize these models are gathered in life history tables describing the demographic structure of the population by means of size/age structure, fecundity, fertility, survival and recruitment rate.

Gorgonians are long-lived, engineering and habitat-forming species (Gili and Coma 1998) frequently affected by anthropogenic pressure of different origins (e.g., Cupido el al. 2009; Tsounis et al. 2012, and references herein). Moreover, due to their high economic value, some species have suffered intensive exploitation (Tsounis et al. 2010, 2013). Conservation of harvested, long-lived, slow-growing species is one of the most difficult challenges faced by ecologists. Determining the best approach for the sustainable exploitation of precious octocorals is still a controversial matter (Bruckner 2009; Santangelo and Bramanti 2010; Tsounis et al. 2010, 2013; Iwasaki et al. 2012). Ad hoc studies must therefore be addressed at understanding the complex dynamics of the populations in order to evaluate their conservation and management status (Edmunds and Elahi 2007; Santangelo et al. 2007; Bramanti et al. 2009).

The Mediterranean red coral Corallium rubrum is a long-lived, slow-growing precious gorgonian endemic to the Mediterranean and neighboring Atlantic areas (Zibrowius et al. 1984), frequently forming dense monospecific patches within the Mediterranean coralligenous assemblages. Harvested for over two thousand years due to its highly valued axial skeleton used as raw material for jewelry, this species has played and still plays an important economic and cultural role in European and Eastern Asian countries (Cicogna and Cattaneo-Vietti 1993; Tsounis et al. 2010). The long history of red coral harvesting is a paramount example of longtime, highly specialized exploitation, selectively addressed toward a single species.

Following management needs, an operational distinction has been made between shallow and deep C. rubrum populations (Santangelo and Abbiati 2001; Costantini et al. 2010). The shallow populations live between 10 and 50 m depth and have been long overharvested due to the ease of accessibility. They are mainly characterized by dense patches of small-sized colonies (Garrabou and Harmelin 2002; Tsounis et al. 2006a) with low economic value but high naturalistic and touristic value (Bramanti et al. 2011). The deep populations (50−200 m) are composed of sparse large colonies (Rossi et al. 2008; Priori et al. 2013) with differing genetic structure (Costantini et al. 2011) and are nowadays the main target of fishery (Santangelo and Bramanti 2010; Priori et al. 2013; Tsounis et al. 2013). There is some evidence that the long-term harvesting pressure on shallow red coral populations may have caused a shift in their structure toward higher colony densities and smaller size (Garrabou and Harmelin 2002; Tsounis et al. 2006a; Bramanti et al. 2009; Linares et al. 2010, 2012). However, the lack of demographic data on pristine populations limits our knowledge of the past population structure (Knowlton and Jackson 2008).

The last 20 years of scientific research on C. rubrum showed that the exploitation of shallow populations must be avoided due to the risks represented by climate change (Santangelo et al. 2007; Bramanti et al. 2013). The exploitation of deep dwelling populations needs a careful management based on sound population and fishing data and reliable analysis of demographic trends (Santangelo and Abbiati 2001). Moreover, different populations can show different demographic characteristics linked to the different environment in which they live and/or to different histories (Table 1). Basing on this assumption, the management of C. rubrum has to be thought taking into account the differences between populations.

Due to the international concern for conservation, recent attempts have been made to regulate the trade of the species of the genus Corallium, including all species of this genus in the annex II of CITES (Convention for International Trade of Endangered Species; Bruckner 2009; Tsounis et al. 2010). Due to the lack of data necessary to determine whether or not the population decline meets CITES criteria, all such attempts have failed (Tsounis et al. 2013). According to the queries raised in the course of international meetings (Bruckner 2009; IFPC 2009, 2012; GFCM 2010, 2011), comparative demographic data are urgently needed in order to set management and conservation plans suitable for sustainable harvesting of the Mediterranean red coral. On the basis of the works of Santangelo et al. (2004, 2007) and Bramanti et al. (2009), the demographic approach to the study of the population dynamics has been pinpointed as a key instrument in the management of C. rubrum (Bramanti et al. 2010).

Since the seminal work of Huges (1984), the demographic approach has been widely applied to corals (e.g., Hughes and Jackson 1985; Babcock 1991; Fong and Glynn 1998; Goffredo and Chadwick-Furman 2003). These authors used a combination of demographic data such as fecundity, growth and survival rate in order to estimate population growth and describe the dynamics of different species of scleractinian corals. However, only in few cases the projection capacity of demographic models has been used to propose conservation and management measures (Santangelo et al. 2007; Bramanti et al. 2009, 2010).

The estimation of colony age and growth rates is required to assess the population age structure (Tsounis et al. 2006a) and to make inferences of future population trends. While size is a basic descriptor of coral population structure and demographic models can be based on size rather than age (e.g., Huges 1984), age determination allows for a temporal resolution of the model, by estimating the most probable age for each size class. The calculated growth rate gives then a time step for the reiteration of size-structured population models (Caswell 2001). To this end, a method for age determination of C. rubrum colonies was applied to the populations object of this study. Species displaying skeletal calcareous structures such as mollusks, cnidarian and bryozoans can be analyzed by sclero chronological techniques in order to have information on the age (e.g., Knutson et al. 1972; Millstein and O’Clair 2001). Coral reef species, in particular, are the most widely studied due to the presence of annual alternance of low- and high-density bands (Knutson et al. 1972; Barnes and Lough 1993). We applied a method developed and validated by Marschal et al. (2004) for the aging of C. rubrum colonies. The method have been previously applied to other C. rubrum populations (Torrents et al. 2005; Gallmetzer et al. 2010; Priori et al. 2013).

The study of population size frequency distribution can give insights to estimate the response of coral populations to the environment (Bak and Meesters 1998) and to climate change (Bak and Meesters 1999). Rossi et al. (2012) used the analysis of the size structure of a population of Eudendrium racemosum in order to estimate the trophic impact of this species, while Linares et al. (2010) compared the skewness values of the size frequency distribution of C. rubrum in different marine protected areas. The size frequency distribution can also give information on the disturbances that affected a population in the past, as showed by Linares et al. (2008a) for Paramuricea clavata and Eunicella singularis.

Recruitment, a crucial basic demographic parameter determining population growth rate (Benayahu and Loya 1985; Grigg 1988), is an operational term defining the initial sighting of a recently settled juvenile (Caley et al. 1996). In the present paper, the term juveniles refers to the first age class in the population and includes settlement and survival of newly settled individuals. Due to the difficulty in identifying very small individuals (Bramanti et al. 2003), this class has been examined only in a few C. rubrum populations (Garrabou and Harmelin 2002; Bramanti et al. 2005; Bramanti et al. 2007; Santangelo et al. 2012).

The present research aims at analyzing the demographic and reproductive features of two shallow red coral populations living in two different locations in the NW Mediterranean, separated by 550 km: Portofino (Italy) and Cap de Creus (Spain). Demographic and reproductive features will be analyzed in terms of: (1) juveniles and adult colonies density; (2) colony growth rate; (3) size and age structure; (4) sex ratio, fertility and fecundity. A comparison between the reproductive and demographic features of the two population will be performed. The data obtained are the first step for the parameterization of robust demographic models based on sound information on the demographic parameters (Santangelo et al. 2007). The ultimate aim is to shed light on the current status of the populations of this precious species by means of comparative demographic evidence.

Materials and methods

The study areas

This study was carried out in two locations in the NW Mediterranean Sea, located 550 km from each other (Fig. 1a): Portofino Promontory marine reserve (Eastern Ligurian Sea 44°18.18′N; 09°12.83′E, Italy; Fig. 1b) and Cap de Creus marine reserve (Costa Brava; 42°29.21′N; 3°30.18′E, Spain; Fig. 1c).
Fig. 1

a Western Mediterranean Sea. b Portofino marine reserve. c Cap de Creus marine reserve. Black dots indicate sampling Sites

Extending along a 13-km coastline, the 3.46 km2 Portofino marine reserve was established in 1999. In 2002, it was included in the SPAMI list (Special Protected Area of Mediterranean Interest) due to the species richness of its coralligenous species assemblage. The southern slope of the promontory is characterized by a heterogeneous habitat with patches of different species of gorgonians, among which C. rubrum is the most common. Red coral harvesting, as well as many other fishing activities, has been banned inside the marine reserve since 1999.

The marine reserve of Cap de Creus, established in 1998, extends for 30.5 km2 along 40 km of coastline and was included in the SPAMI list in 2001. National and regional laws regulate harvesting and fishing activities inside the marine reserve; in particular, C. rubrum can be collected, only manually and in some locations, by licensed fishermen. The yearly quota of raw C. rubrum that each of 10–12 licensed fishermen can harvest is 400 kg. The harvesting season lasts from May to October and the size limit is 7 mm in basal diameter (recently proposed by FAO; GFCM 2011), although illegal red coral harvesting is frequently conducted throughout the reserve (Tsounis et al. 2013).

Values of chlorophyll a, water transparency and total organic matter (TOM) in the two locations are summarized in Table 2. According to Durrieu de Madron et al. (2011), these values place both populations in the same eco-region.
Table 2

Average annual values of environmental characteristics recorded in Portofino (Bavestrello et al. 1995) and in Cap de Creus (Rossi and Gili 2005)

 

Portofino

Cap de Creus

Water transparency (m)

14.8

15.25

Chlorophyll a (μg l−1)

0.6

0.7

TOM (mg l−1)

3.3

3.9

Sampling

Sampling was carried out by SCUBA diving between 30 and 35 meters depth in both marine reserves (Locations) in June–July 2009. Adult colony and juveniles densities were examined according to a nested ANOVA sampling design; in each Location (fixed factor, 2 levels: Portofino and Cap de Creus), four Sites (random factor, 4 levels, nested in area) were randomly selected. In each Site, three 20 × 20 cm quadrats (replicates) were selected randomly. Overall, 24 replicates were sampled, and the percentage of variance explained by each factor of the nested ANOVA was calculated (R software, package ‘stats,’ command ‘aov’). Before performing the test, all the assumptions of ANOVA were verified.

All adult colonies in each quadrat were individually detached by a chisel, collected by hand and placed in plastic jars. As the sampling was carried out in marine reserves, both the extent and number of samplings were limited. Once adult colonies had been collected (thus revealing the underlying smaller colonies they concealed), the quadrats were photographed with a digital camera (Canon G10 with IKELITE housing and TTL strobe) and juveniles were identified through image analysis of pictures. As red coral juveniles, measuring on average 0.6 mm (basal diameter; Bramanti et al. 2005, 2007), are difficult to be identified, the pictures were taken in macro mode (resolution 40 megapixels) and examined at higher magnification in the laboratory. This method is extremely accurate; however, it could be biased due to the difficulty of handling smaller colonies (2–3 mm), some of which may have been lost during sampling (Santangelo and Bramanti 2010). Part of the smaller colonies (1–2 mm) and juveniles do not detach from the substrate due to their small dimensions, and they are counted on photographs.

Removed adult colonies (n = 620; 477 from Portofino and 143 from Cap de Creus) were fixed immediately after collection in 4 % formaldehyde–seawater-buffered solution to be subsequently processed for laboratory analyses to determine size, age, sex, fertility and fecundity of female colonies. Due to the lower density of colonies at Cap de Creus, an additional sample of 33 colonies was taken from that Location at the end of June 2009 and used only for the sex ratio analysis.

Population size structure

Colony basal maximum and minimum diameters were measured with a caliper (precision ±0.05 mm) at the inflection point (1 ± 0.2 mm from colony bases) and averaged. The size structure of the population was obtained by dividing all the colonies into 1 mm diameter classes. Skewness, a measure of the asymmetry of a distribution (Linares et al. 2008a; Gori et al. 2011; Rossi et al. 2012), was then calculated, and the hypothesis of symmetry (skewness = 0) was tested by Agostino’s test (R software, package ‘moments,’ command ‘agostino.test’),

Colony age and growth rate

Colony age was determined according to the thin section—organic matrix staining method (Marschal et al. 2004). After decalcification and staining, thin sections were photographed under a stereomicroscope. Each image was calibrated according to the proper magnification and processed by the graphic software ImageJ (Rasband 1997). Age was estimated as the average number of growth rings recorded on three linear transects, each drawn in a different sector of a thin section (Fig. 2). The first visible growth ring usually appears after 4 years at a distance from the medullar region (Marschal et al. 2004; Vielzeuf et al. 2008); hence, the age of colonies was calculated by adding 4 to the number of growth rings counted. Only a subsample of the collected colonies (n = 119) was suitable for this method due to holes made by boring sponges as well as to the irregular morphology of some colonies that hindered annual growth rings count (Priori et al. 2013). The surface of the circular crown area (CCA) delimited by two subsequent growth rings was also measured on a subsample of the colonies (n = 33). Each CCA is the annual increase in surface of the colony base section; thus, each colony yielded a number of CCA measurements proportional to its age. Finally, the annual increment of the overall CCA of the 33 colonies (i.e., the sum of the CCAs) was plotted against age. As age determination by growth rings count can have some uncertainties due to the first years of growth, CCAs were grouped into 3-year intervals. The annual growth rate was calculated by dividing the averaged basal diameter by colony age. The relationship between colony diameter and age was fitted to the subsample of colonies whose age was determined by the annual growth ring count (n = 119 colonies). This relationship was then applied to divide all the sampled colonies into size/age classes according to their average annual growth rate (Santangelo et al. 2007). The relationship between the age and the mean annual growth rate of colonies was examined via Pearson’s linear correlation coefficient r (CurveExpert Pro software).
Fig. 2

Thin section of a C. rubrum colony after decalcification and staining (Marschal et al. 2004) where growth rings are visible. a Transect lines along which rings count have been performed (see ‘Materials and methods’ section). b Medullar region

Reproductive features

Overall, 653 colonies (620 from the replicated quadrats and 33 additional from Cap de Creus) were examined, and the sexual status of all the colonies sampled in the quadrats was determined. The number of polyps examined in each colony varied proportionally according to colony size, ranging from five polyps in the smaller colonies (<3 mm of basal diameter) up to 18 in larger ones (>5 mm basal diameter; Santangelo et al. 2003). Male and female colonies were distinguished via microscope analysis of gonads (Santangelo et al. 2003; Tsounis et al. 2006b; Torrents and Garrabou 2011; Priori et al. 2013). Polyps were dissected and observed under a stereomicroscope, at 10−80× and then examined by an optical microscope at 250–1,000× to identify oocytes and sperm sacs. Colonies in which no gametes were found were considered sterile. Sex ratios were tested using a chi-square test (R software, package ‘stats’; command ‘chisq.test’). After sex identification, the number of oocytes in each polyp was counted along with the average polyp fecundity calculated as the ratio between the overall number of oocytes found and the number of polyps analyzed.

Results

Population density

Overall, out of the 24 replicated quadrats, 620 adult colonies were found (n = 477 for Portofino and n = 143 for Cap de Creus), corresponding to a density of 39.7 ± 17.8 and 11.9 ± 6.9 (mean ± SD) colonies 400 cm−2 for Portofino and Cap de Creus, respectively (Fig. 3a). The observed differences in adult colonies density between the two populations, were confirmed by the nested ANOVA (Table 3a) which revealed that the main variance component is due to the factor Location (60.47 %), while the factor Site (nested in Location) accounts for 1.54 % of the variance.
Fig. 3

Adults (a) and recruits (b) average density (colonies 400 cm−2) in Portofino (black columns) and Cap de Creus (gray columns). Error bars represent SD

Table 3

Nested ANOVA for C. rubrum adult colonies (a) and juveniles (b) densities

Source

SS

DF

MS

F

p

F versus

Variance %

a: Adult colonies

Location

3,626.04

1

3,626.04

16.58

0.01

Site (area)

60.47

Site (location)

1,313.58

6

218.76

0.84

0.56

RES

1.54

RES

4,175.33

16

260.96

   

37.99

TOT

9,113.96

23

     

b: Juveniles

Location

847.87

1

847.87

41.1

0.0007

Site (location)

97.39

Site (location)

123.77

6

20.62

1.59

0.215

RES

1.61

RES

208.16

16

13.01

   

1.00

TOT

1,179.80

23

     

Factors are Location (Portofino and Cap de Creus) and Site (four randomly chosen Sites inside each Location. In both cases, the main variance percentage was explained by the factor Location

Overall, 278 juveniles were found: 210 at Portofino and 68 at Cap de Creus, corresponding to 17.5 ± 4.7 and 5.6 ± 2.8 (mean ± SD) recruits 400 cm−2 (Fig. 3b), respectively. Similarly to the adult density analysis, a significant difference between the two Locations was revealed by the nested ANOVA, showing Location accounted for 97.39 % and Site for only 1.61 % of the variance (Table 3b). Moreover, the variance component associated with residual was remarkably higher for adult colonies than for juveniles (37.99 vs. 1.0 %; Table 3b), indicating a larger variability occurred in the density of adult colonies. Both adult and juveniles densities were about threefold higher in Portofino than in Cap de Creus, and thus, the ratio adult/juveniles was similar in both populations (~2.15/1).

Colony size, growth rate and age

The average basal diameter of colonies was 4.42 ± 1.69 mm at Portofino and 4.08 ± 1.24 (mean ± SD) mm at Cap de Creus, with a maximum of 14.48 mm in the former and 11.5 mm in the latter population. The size distribution at Portofino (Fig. 4, black bars) is characterized by a significantly positive skewness (skewness = 0.61, z = 3.01, p = 0.003), due to the larger colonies there, while Cap de Creus distribution (Fig. 4, gray bars) was not significantly skewed (skewness = 0.129, z = 0.426, p = 0.67). The frequency of adult colonies larger than 7 mm in diameter (the minimum legally harvestable size) was 6.7 and 2.1 % at Portofino and at Cap de Creus, respectively. Juveniles are the dominant class in both populations in 79 % of the sampled quadrats, representing 30.6 and 32.2 % of the overall colonies at Portofino and Cap de Creus, respectively.
Fig. 4

Size (basal diameter) distribution in the Portofino (black bars) and Cap de Creus (gray bars) populations of C. rubrum

Overall, 119 colonies (n = 75 for Portofino and n = 44 for Cap de Creus), encompassing all diameter classes (1 mm each), were sectioned and stained (Marschal et al. 2004; Priori et al. 2013), and the annual growth rings were counted (Fig. 2). Colonies bored by endobiont sponges (unsuitable for age determination) were 50.3 % at Portofino and 41.5 % at Cap de Creus. The average colony growth rate was 0.241 ± 0.061 and 0.237 ± 0.062 mm year−1 (mean ± SD) at Portofino and at Cap de Creus, respectively. As these values did not differ significantly (t test, t = 0.9; df = 43, p = 0.27), they were pooled and the average colony growth rate of both populations was 0.24 ± 0.06 mm year−1 (mean ± SD). A significant negative, linear correlation was found between the mean annual growth rate of a colony and its age (r Pearson = −0.43; n = 119; p < 0.001). This finding indicates a decrease in the growth rate of colony basal diameter with age. The yearly increase in the circular crown area was constant during the life of the colonies (Fig. 5a). The yearly average colony growth in diameter is well fitted by the power curve
$$y = 0. 5 5 2x^{0. 6 8 9} \left( {R^{ 2} = 0. 5 7;N = 1 1 9;p = 0.0 1;\;{\text{Fig}}.{\text{ 4b}}} \right)$$
(1)
where x is the age (years) of the colony and y the diameter (mm). When data on the initial 4 years of colony growth (Bramanti et al. 2005; Santangelo et al. 2012) are included in the fit, the significance of the power curve representing colony growth remarkably increases (y = 0.646x0.652; R2 = 0.88; N = 180; p = 0.003; Fig. 4c).
Fig. 5

a The growth of the circular crown area (CCA) of colony base is fitted by a linear equation (R2 = 0.99), indicating an increase in CCA constant along colony life span. b Relationship between colony age (determined by annual growth rings count) and basal diameter. c Relationship between colony age (determined by annual growth rings count) and basal diameter including data of the first 4 years of colony growth (Bramanti et al. 2005; Santangelo et al. 2012)

Based on the relationship between age and size reported above [Eq. (1); Fig. 5b], all the colonies sampled in the two Locations (n = 620) were divided into annual size classes determined on the basis of their average annual growth in diameter (i.e., thereby assigning the colonies to the most probable age group, based on their size; Santangelo et al. 2007). Juveniles, the dominant class in both populations, were also included in the size/age structures. Both size/age structures were well represented by monotonic decreasing curves with negative exponents (y = 210 e−0.109x and y = 68 e−0.115x for Portofino and Cap de Creus, respectively, Fig. 6a, b). The size/age distributions revealed a 33 % longer life span at Portofino, where 99 % of the colonies have a maximum estimated age of 60 years, while at Cap de Creus 99 % of the colonies have a maximum estimated age of 40 years. The size distribution of the Portofino population also showed significant skewness due to the presence of larger colonies (Fig. 4). The maximum estimated age was 110 and 79 years at Portofino and Cap de Creus, respectively.
Fig. 6

Size/age structure in the two locations. a Portofino. b Cap de Creus. Black dots represent colonies >3 mm (average basal diameter). Gray dots represent colonies <3 mm that may be underestimated due to the sampling method

Reproductive features

Overall, 7,294 female polyps (n = 6,318 for Portofino and n = 976 for Cap de Creus) were dissected and examined. The mean fecundity of female polyps was 0.64 ± 0.33 at Portofino and 0.42 ± 0.30 at Cap de Creus (oocyte polyp−1; mean ± SD) with no significant difference between the two Locations (t test; t = 1.9; df = 39; p = 0.067). No significant relationship between polyp fecundity and colony size/age was found (r Pearson = 0.18; n = 146; p = 0.39).

About 74 % (353/477) of the adult colonies examined at Portofino and 78.6 % (143/181) of those examined at Cap de Creus were found fertile with no significant difference between the two Locations (chi-square test: chi-square = 0.15; p = 0.69). Overall, 24 % of colonies are sterile (162/658) representing 26.3 % of the 1−3 mm diameter classes and 9 % of the 4−6 mm classes. No sterile colonies were found in larger classes (>6 mm). The sex ratio of both populations did not diverged from randomness, being the male/female ratio 0.95/1 (172/181) and 0.74/1 (61/82) at Portofino and at Cap de Creus, respectively (chi-square = 0.39; p = 0.53 for Portofino, and 0.1, p = 0.75 for Cap de Creus).

Discussion

In this paper, we analyzed the demographic and reproductive features of two populations of the precious gorgonian C. rubrum in two different geographic locations: Portofino (Italy) and Cap de Creus (Spain). Results showed a significant higher density of adult and juveniles in Portofino.

Juveniles density found at Portofino was similar to that found in other shallow populations of the Eastern Ligurian Sea, similarly characterized by high colony density (Bramanti et al. 2005; Santangelo et al. 2012). Low juvenile colony densities, similar to the values we found in Cap de Creus, have been also found in the red coral population of Medes marine reserve, located at about 20 km from Cap de Creus and similarly composed by low density patches (Tsounis et al. 2006a; Santangelo et al. 2012).

Notwithstanding the large differences in juveniles and adults densities, a similar ratio between these two life stages was found in both populations, suggesting density-dependent correlation between these two life stages could occur (Garrabou and Harmelin 2002; Santangelo et al. 2007). The colony size distributions of the two populations revealed that the percentage of colonies above the minimum legal harvestable size (7 mm in diameter) at Cap de Creus is threefold lower than in Portofino (2.1 vs. 6.7 %) and fourfold lower than in year 2003 (Tsounis et al. 2007). Notwithstanding the sampling method was different from the one used in our study (in Tsounis et al. 2007, a photographic sampling of adult colonies was carried out), the comparison with our results suggested a 20 % reduction in the average diameter of colonies has occurred in the Spanish reserve over the course of only a few years.

Due to the presence of large colonies at Portofino, the size distribution was significantly skewed. This pattern was not found at Cap de Creus, where large colonies were much more scarce and the maximum size was lower by 1/3. An even more evident shift toward large colonies has been observed by Tsounis et al. (2006a; Linares et al. 2012) at Medes islands (Costa Brava, Spain) and by Linares et al. (2010) at Scandola (Corsica, France), Carry-le Ruet (Marseille, France) and Banyuls-sur-Mer (Languedoc-Roussillon, France), where harvesting is strictly forbidden and controlled. These findings underline that in some marine reserves, the enforcement has an effect on the size of red coral colonies.

The average colony growth rate found in this study (0.24 ± 0.06 mm year−1) was consistent with the values reported for other shallow populations (Garrabou and Harmelin 2002; Marschal et al. 2004; Torrents et al. 2005; Gallmetzer et al. 2010) and slightly lower than the one found by Priori et al. (2013) for deep dwelling populations (0.26 mm year−1). All these findings suggest some constancy in the average colony growth rate of shallow populations.

The significant negative correlation found between the mean annual growth rate of the colonies and their age indicates a decrease in the growth of colony basal diameter with age. Moreover, it has been found that during the first 4 years, colony diameter grows 2−3 times faster than in subsequent years (Cattaneo-Vietti et al. 1993; Bramanti et al. 2005, 2007). Thus, a decreasing curve should better describe the trends of colony growth with age. A reduction in basal diameter growth with time is furthermore confirmed by the constant growth of the circular crown area (CCA) we found.

The not significantly skewed size/age distributions in Cap de Creus, where 99 % of colonies have a maximum age of 40 years, confirmed a shorter (66 %) life span than those in Portofino, where the maximum age of 99 % of the colonies was 60 years. In both populations, the minimal legal harvestable size (7 mm) corresponds to an estimated age of 30−35 years.

The dominant age group in both populations is represented by juveniles. Although occasional pulse peaks of recruitment have been described in red coral populations (Garrabou and Harmelin 2002; Santangelo et al. 2012), our results indicate that the dominance of juveniles over the older classes is a pattern common to both the populations studied. The monotonic decreasing pattern of the size/age structure is another common feature of the two populations examined and is also similar to that found in other red coral populations living in shallow waters (Santangelo et al. 2007). The low recruitment rate found in the Cap de Creus population could likely be determined by a low larval production due to low density and small sizes of local colonies. This could be amplified by low larval retention rates and low larval supply from neighbor populations (Ledoux et al. 2010; Costantini et al. 2011; Guizien et al. 2012). Despite main life history parameters (colony growth rate, sex ratio and colony fertility and fecundity) were similar in both populations, adults and juveniles densities as well as population life span were remarkably lower for the population at Cap de Creus. The environmental characteristics reported in Table 2, in particular chlorophyll a, are good proxies for phytoplankton and zooplankton (Estrada 1996). As phytoplankton and zooplankton represent between 22 % and 57 % of the diet of C. rubrum colonies (Tsounis et al. 2006c), the similar values of these characteristics in the two locations could explain the similar values of growth rates and fecundity found in the two populations. The observed differences in the demographic structures could then be due to the different mortality rates at which the larger/older colonies have been and are still subjected, probably due to harvesting that selectively affects large colonies. Red coral harvesting at Cap de Creus is a historic local tradition, resulting in intense longstanding harvesting continuing until today (Rossi et al. 2008; Tsounis et al. 2010, 2013). In contrast to Cap de Creus, red coral harvesting at Portofino has always been less intense, decreasing some years before the establishment of the marine reserve in 1999. Moreover, due to the smaller extension of Portofino reserve, enforcement may be easier and more effective. Lastly, Cap de Creus population is subject to lower infestation by boring sponges (Corriero et al. 1997), producing colonies with axial skeletons more suitable for jewelry, thus increasing the commercial value of colonies, which could have encouraged a more intense harvesting at Cap de Creus. According to Tsounis et al. (2007), overharvesting may have caused a threefold reduction in the colony life span in Cap de Creus population, a difference similar to that found between the two populations object of the present study. The low genetic diversity found at Cap de Creus represents a further evidence of a recent bottleneck likely due to overharvesting (Costantini et al. 2011).

Based on the present data, we are unable to assess whether harvesting is the only cause of the density and size decline recorded in Cap de Creus population. However, our findings suggest that the legislation and the enforcement in this location are likely insufficient to sustainably manage the local overharvested red coral population. If actual harvesting rates (to which an unquantified effect of poaching must be added) will be allowed to continue, it may lead the population to local extinction, with related economic loss (Bramanti et al. 2011).

As several shallow red coral populations dwell in non-protected areas (Bramanti et al. 2011), a unique Mediterranean legislation banning red coral harvesting in the Mediterranean Sea above 50 m bathymetry (as proposed by GFCM in 2011) together with an increased local enforcement could supply the best tool to ensure the proper protection to this species. Our findings provide additional field evidence in support of the recent GFCM proposal.

Other additive factors have to be considered to understand why a potential local extinction of red coral shallow patches in Cap de Creus could be forecasted (Tsounis et al. 2007). First, mass mortality events strongly affecting octocorals in the Mediterranean Sea (Cerrano et al. 2000; Garrabou et al. 2001, 2009; Bramanti et al. 2005) could be an important factor of mortality, especially for populations living at the edge of the summer thermocline (Linares et al. 2008b; Cupido et al. 2009). Moreover, it has been shown that the synergistic effects of overharvesting and heat wave-induced mass mortalities could reduce the resilience of shallow red coral populations if these two stressors will increase in frequency (Santangelo et al. 2007; Garrabou et al. 2009). Second, once red coral patches have been depleted, competition for space could arise: The low recruitment of red coral, together with a reduction in density, may lead to successional stages in which Leptopsammia pruvoti could occupy the space previously occupied by red coral colonies (Cattaneo-Vietti et al. 1993; Giannini et al. 2003).

Population dynamic models based on demographic traits such as fecundity, fertility, growth and recruitment rates can be used to project the population structure overtime. These models are widely applied in marine, freshwater and terrestrial ecosystems (Tuljapurkar and Caswell 1997). Finally, our data underline the role of local factors in the population dynamics of C. rubrum and represent a sound basis in the development of demographic models for populations dynamics of this species.

Acknowledgments

The authors are grateful for the invaluable support of Portofino MPA and Cap de Creus MPA. L Bramanti was supported by a Marie Curie Intra European Fellowship within the European Community’s Seventh Framework Programme (CORGARD Project No. 221072), and S Rossi was funded by a Ramón y Cajal Contract (RyC-2007-01327). The research leading to these results has received funding from the European Community’s Seventh Framework Programme under grant agreement 265103 (Project MedSeA). Part of the project was funded by the Italy–Spain exchange program UNIPI-CSIC, PRIN 2011 and BENTOLARV (CTM2009-10007). Thanks to Fischer Italia and SSI Italia for support and to TDI and Aldo Ferrucci for the CCR Megalodon units used for the underwater sampling. A thanks also to the European diving center and to Toni Garcia for the logistic help in the field work and to A. Cafazzo, D. Brown and K. Guizien for the revision of the English and comments on the manuscript. Data and results form part of PhD thesis of I. Vielmini at Pisa University.

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • L. Bramanti
    • 1
    • 2
    • 8
  • I. Vielmini
    • 3
  • S. Rossi
    • 4
  • G. Tsounis
    • 5
  • M. Iannelli
    • 6
  • R. Cattaneo-Vietti
    • 7
  • C. Priori
    • 8
  • G. Santangelo
    • 8
  1. 1.CNRS, UMR8222, LECOB, Observatoire OcéanologiqueUniversité Pierre et Marie CurieBanyuls-sur-MerFrance
  2. 2.Instituto de Ciencias del Mar, Barcelona (ICM-CSIC)BarcelonaSpain
  3. 3.AquaTTDublin 2Ireland
  4. 4.Environmental Science and Technology InstituteAutonomous University of BarcelonaBarcelonaSpain
  5. 5.Leibniz Center for Tropical Marine Ecology GmbH (ZMT)BremenGermany
  6. 6.Department of MathematicsUniversità di TrentoPovo, TrentoItaly
  7. 7.Dipartimento per lo Studio del Territorio e delle sue Risorse (Dip.Te.Ris)Università di GenovaGenoaItaly
  8. 8.Department of BiologyUniversità di PisaPisaItaly