Chemical and mechanical bioerosion of boring sponges from Mexican Pacific coral reefs

The growth and maintenance of coral reefs is the result of an equilibrium between the deposition and the erosion of carbonate(Hutchings, 1986). A significant part of the erosion process is of a biological nature and is termed bioerosion (Neumann,1966). In coral reefs, bioerosion is driven by a high diversity of organisms but, usually, the dominant groups around the world are boring sponges (MacGeachy and Stearn,1976; Hudson,1977). The activity of these sponges and other bioeroders can weaken the reef by causing dislodgment of the framework, especially during wear and tear caused by waves (Stearn and Scoffin, 1977; Macdonald and Perry, 2003) and in some Caribbean coral reefs they are thought to be responsible for reef destruction (Rose and Risk, 1985; Ward-Paige et al., 2005).

As the sponge penetrates the coral, the substrate is gradually destroyed as a result of the sponge hollowing out an extensive system of cavities and tunnels (López Victoria et al.,2006; Calcinai et al.,2007). These excavations are produced by the expulsion of small lenticular calcareous chips (15–100 μm in diameter) through the aquiferous system of the sponge (mechanical boring)(Rützler and Rieger,1973). Boring sponges can remove large amounts of calcareous material from the reef framework by chip production (up to 22 kg CaCO₃ m^–2 year^–1), generating up to 40% of the sediment deposited on some reef ecosystems(Neumann, 1966; Rützler and Rieger, 1973; Fütterer, 1974; Rützler, 1975).

In addition, boring sponges are able to dissolve part of the carbonate during the bioerosion process (chemical boring)(Rützler and Rieger,1973; Pomponi,1977). Traditionally, it was considered that the amount of chemically dissolved material was minimal compared with the material removed mechanically (Rützler and Rieger,1973; Rützler,1975; Acker and Risk,1985). However, it was recently shown that Pione cf. vastifica (Hancock 1849) dissolves three masses of reef CaCO₃framework per each part of carbonate removed mechanically, suggesting that chip production represents only a small fraction of the sponges bioerosion capacity (Zundelevich et al.,2007).

The coral communities of the Eastern Pacific coast are distributed between 30 deg. N and 5 deg. S (Glynn and Ault,2000), and the Mexican Pacific coast comprises 46% of their total distributional range. The present study was carried out in the National Park Isla Isabel (México), which harbors a typical coral community from the Eastern Pacific, formed mainly by corals of the genus Pocillopora(Reyes-Bonilla, 1993).

We addressed the hypothesis that boring sponges may have a major role in the incorporation of dissolved CaCO₃ to the water column.

This work had four goals: (1) to study the species richness (S) and the frequency of invasion (IF) by boring sponges on living colonies of Pocillopora verrucosa (Ellis and Solander 1786) at Isla Isabel; (2)to quantify the carbonate dissolution rate of Cliona vermiferaHancock 1867 and Cliona flavifodina Rützler 1974, two of the most abundant and amply distributed boring sponges from the east Pacific Ocean(Carballo et al., 2004; Carballo et al., 2008). This is especially important to ensure the validity and generality of previous results for Pione cf. vastifica(Zundelevich et al., 2007),which would allow us to establish a confident prediction about the real importance of boring sponges in coral reef environments; (3) to quantify the sediment production of C. vermifera and C. flavifodina; and(4) to quantify the bioerosion rate of these species.

The Mexican Pacific coast harbors only a few important reef structures due to coastal upwelling and high river discharge(Reyes-Bonilla, 1993). Isla Isabel (21 deg. 50′35″N–105 deg. 53′04″W) is an oceanic island located 32 km off the coast(Fig. 1) that harbors a diverse coral community (14 species), dominated by corals of the genus Pocillopora, a common Eastern Pacific species(Reyes-Bonilla, 1993). Since the island is protected by natural reserve there is no significant human impact.

The species richness and abundance as frequency of invasion (IF) of boring sponges was quantified by examining living colonies of P. verrucosa(Carballo et al., 2008).

Sampling was carried out by scuba diving, collecting 25 fragments of the basal portion of live corals approximately 90 cm³ along 10 transects 25 m long perpendicular to the shoreline(Carballo et al., 2008). The samples were examined in the laboratory for the presence of boring sponges. Each sample was broken into small pieces. When a sponge was found, a sample of its tissue was digested with sodium hypochlorite, and its spicular characteristics were examined under an optical microscope. The identification was done to the highest taxonomic level using the available literature(Carballo et al., 2004; Carballo and Cruz-Barraza,2005; Bautista-Guerrero et al.,2006). The IF for each species (mean ± s.e.m.) was recorded as the proportion (%) of fragments where boring sponges were found.

The variation in the alkalinity of the seawater has been accepted as a reliable and rapid methodology to quantify small changes in the dissolved carbonate concentration in the water(Smith and Kinsey, 1978). It was initially utilized to calculate bioerosion rates by litophagid molluscs(Lazar and Loya, 1991) and calcification in corals (Chisholm and Gattuso, 1991). Recently, this technique has been successfully tested to calculate the sponge bioerosion rate of Pione cf. vastifica(Zundelevich et al.,2007).

Five fragments of dead corals invaded by the boring sponge C. vermifera and five fragments invaded by C. flavifodina were selected and transported to the laboratory. Boring sponges were allowed to recover for 24 h. Before the beginning of the experiment, each coral fragment was externally cleaned with a soft brush to remove particles that could confound the measurements of sediment production. It was then placed in a plastic container with 2.5 l of filtered, aerated seawater. Another five coral fragments without sponges were cleaned and used as controls(Zundelevich et al.,2007).

Since the size of the chips produced by boring sponges range from 15 to 100μm (Rützler and Rieger,1973), the water in each plastic container was passed through a sieve (150 μm mesh) to eliminate bigger particles before sieving it with a pre-weighed and dried 0.45 μm glass fiber filter. All particles within the size range of the chips produced by the sponge were retained in the filters. The particles were then gently washed with distilled water before being dried,heated to 500°C for 2 h and weighed again. The mass of the particles was determined from the difference between the final and initial masses of the filters.

Later, all particles of each filter were washed and homogenized in 1 ml of distilled water. The mean proportion of sponge chips and non-sponge chips (%)was then counted from multiple samples in a hematocytometer cell-counting slide using an optical microscope.

Sediment production (mean ± s.e.m.) was estimated as the mass of the sponge chips (kg) subtracted from the total mass of the particles retained in the filter.

When the experiment was finished, the fragments of dead coral were fixed with a 4% solution of formalin. To determine the amount of sponge in the fragments, these were cut into 0.5 cm sections, and both sides of the sections were photographed on a 1 cm² grid. The photographs were digitalized and analyzed on a computer to calculate the area of sponge in each section,using the grids as a reference, thus obtaining the total amount of sponge per m².

Since the carbonate dissolution and the production of sponge chips were quantified during a 24 h period, the bioerosion rates (mean ± s.e.m.)were extrapolated to kg CaCO₃ m^–2year^–1.

Significant differences in dissolution rate, sediment production and bioerosion rate between species were detected by analysis of variance (ANOVA)(Cochran's C-test).

12 species belonging to four genera (Aka, Cliona, Pione and Spheciospongia) were identified, and 56% of the samples examined were invaded by boring sponges. C. vermifera was the most abundant species(30±2.2%, mean ± s.e.m.), followed by Thoosa mismalolli(Carballo et al., 2004)(18.8±1.6%). The rest of the species, including C. flavifodina(0.7±0.3%), had low abundance (Fig. 2).

The dissolution rate oscillated from 0.6 to 2.4 kg CaCO₃m^–2year^–1 (1.2±0.4,mean ± s.e.m.) for C. vermifera, and from 0.1 to 1.1 kg CaCO₃ m^–2 year^–1 (0.5±0.2,mean ± s.e.m.) for C. flavifodina, respectively, while the production of chips varied from 2.1 to 4.7 kg CaCO₃m^–2 year^–1 (3.3±0.6, mean ±s.e.m.) and from 3.8 to 5.6 kg CaCO₃ m^–2year^–1 (4.6±0.5, mean ± s.e.m.), respectively(Fig. 3). However, there were no significant differences between the species for both factors (ANOVA). Thus,the total bioerosion rate (sum of chemical and mechanical phases) was similar for both sponges. It ranked from 2.9 to 7.1kg CaCO₃m^–2 year^–1 (4.5±0.9, mean ±s.e.m.) for C. vermifera and from 3.9 to 6.3 kg CaCO₃m^–2 year^–1 (5.1±0.5, mean±s.e.m.) for C. flavifodina(Fig. 3).

Studies that quantify the species richness and abundance of boring sponges in coral reef ecosystems are relatively scarce(Highsmith et al., 1983; Holmes, 2000; MacDonald and Perry, 2003; López-Victoria and Zea,2005; Venkataraman and Wafar,2005), especially in the Eastern Pacific Ocean(Carballo et al., 2008). Thus,the present study increases knowledge about the diversity and abundance of coral boring sponges in this important region.

The results showed that coral reefs from Isla Isabel harbor a more diverse boring sponge community (12 species) compared with those from southern México (seven species) (Carballo et al., 2008) but are similar to what was found in coral reefs affected by anthropogenic (eutrophication) and natural disturbances(bleaching) (MacGeachy, 1977; Hutchings and Peyrot-Clausade,1988; Holmes,2000; Venkataraman and Wafar,2005).

In addition, a large proportion of the living colonies from Isla Isabel were invaded by boring sponges (56%). This percentage is higher than has been recorded from southern reefs (35%)(Carballo et al., 2008) and is similar to what has been found in eutrophic reefs or in reefs impacted by natural processes, such as El Niño(Sheppard et al., 2002; Venkataraman and Wafar,2005).

Corals from Isla Isabel are not currently affected by anthropogenic impacts but it seems that they were affected by an El Niño event in the past,which considerably increased the availability of substrate by sponge colonization (information provided by the Isla Isabel National Park authority). The availability of dead coral substrata is a key factor in the development of the assemblages of boring sponges(Lopez-Victoria and Zea,2005). The most abundant boring sponge was C. vermifera(30% of the fragments analyzed), the main coral-boring sponge from the Mexican Pacific ocean (Carballo et al.,2008).

This study provides the bioerosion rate of two coral boring sponges from the Eastern Pacific, and confirms the high capacity of these sponges to dissolve carbonate during the erosion process.

C. flavifodina and C. vermifera dissolved 0.5 and 1.2 kg CaCO₃ m^–2 year^–1, respectively,which represents almost one-tenth and one-third of carbonate dissolved per part removed mechanically. P. cf. vastifica dissolved 0.3 kg CaCO₃ m^–2 year^–1, which represents three parts of CaCO₃ per part of carbonate removed in the form of sediment (Zundelevich et al.,2007).

Since previous studies did not calculate the dissolution rate during the bioerosion process, comparisons with different sponges cannot be made. The dissolved fraction, 27.0% in C. vermifera, 10.2% in C. flavifodina and 75% in P. cf. vastifica, is greater than the 2–3% reported in early studies (see Table 1), which confirms the importance of the chemical phase in the bioerosion process.

Some studies suggested that the chemical phase is relevant to the size of the sediments produced by the sponges(Rützler and Rieger,1973; Zundelevich et al.,2007). The results of this study suggest that boring sponges could exert an important role as CaCO₃ recyclers, even benefiting corals in healthy communities by accelerating the reincorporation of carbonate in the water column.

However, there was no difference in sediment production between the two sponges, 3.3kg CaCO₃ m^–2 year^–1for C. vermifera and 4.6kg CaCO₃ m^–2year^–1 for C. flavifodina. By contrast, P. cf. vastifica produced a lower amount of sediment; 0.08 kg CaCO₃m^–2 year^–1(Zundelevich et al., 2007). The dissolved carbonate together with the sediment production accounted for a bioerosion rate of 5.1 kg CaCO₃ m^–2year^–1 for C. flavifodina and 4.5 kg CaCO₃ m^–2 year^–1 for C. vermifera, which are in the range reported for sponges of the genus Cliona (Acker and Risk,1985) (Table 1). For example, while the bioerosion rate of Cliona peponaca did not exceed 3.3 kg CaCO₃ m^–2 year^–1(Bak, 1976), Cliona caribbaea reached 8 kg CaCO₃ m^–2year^–1 (Acker and Risk,1985) and Cliona orientalis ranked from 3.4 to 10.0 kg CaCO₃ m^–2 year^–1(Schönberg, 2002). Higher bioerosion rates, such as 23 kg CaCO₃ m^–2year^–1, have been recorded in fragments of Cliona lampa transplanted on experimental substrata(Neumann, 1966). However, as discussed by Rützler (Rützler,1975), the initial phase of substrata colonization in boring sponges may be conducive to higher bioerosion rates, as could have happened in Neumann's experiment (Neumann,1966).

Other source of variation may be the density of the substrata(Neumann, 1966; Calcinai et al., 2007). In an experimental study of the sponge C. orientalis, Schönberg recorded bioerosion rates from 3.4 to 10.3 kg CaCO₃m^–2 year^–1 in different species of corals,and 17.6 kg CaCO₃ m^–2 year^–1 in the denser shell of individuals of Tridacna squamosa(Schönberg, 2002).

The calcification rates in coral communities have been estimated to be between 1.1 and 4.0 kg CaCO₃ m^–2year^–1 (Kinsey,1985; Silverman et al.,2007). The bioerosion rates recorded in the present study confirm that coral destruction is similar to the corals' rate of construction capability per unit area. These results also suggest that in localities heavily invaded by boring sponges, bioerosion alone could reach a critical level, resulting in the disequilibrium of the ratio of reef accretion/destruction.

Given the frequency of invasion of Cliona vermifera in live colonies of Pocilloporid corals (30%), an important participation of this species in the bioerosion of coral reefs from the Mexican Pacific coast is expected.

The authors wish to acknowledge the following sources of funding: CONACYT SEP-2003-C02-42550, CONABIO FB666/S019/99, CONABIO FB789/AA004/02 and CONABIO DJ007/26. This work was carried out under permission of SAGARPA (Permit numbers: DGOPA.02476.220306.0985 and DGOPA.06648.140807.3121). We thank C. Ramírez-Jáuregui for help with the literature, C. Robles-Carrillo, G. Pérez-Lozano and C. Vega-Juárez for their assistance in the field sampling, and the Director of the Isla Isabel National Park, J. Castrejón, for the permission granted for the collection of samples in the National Park Isla Isabel.