To endure, coral reefs must accumulate CaCO3 at a rate greater or equal than the sum of mechanically, biologically, and chemically mediated erosion rates. We investigated the potential role of holothurians on the CaCO3 balance of a coral reef. These deposit feeders process carbonate sand and rubble through their digestive tract and dissolve CaCO3 as part of their digestive process. In aquarium incubations with Stichopus herrmanni and Holothuria leucospilota total alkalinity increased by 97 ± 13 and 47 ± 7 μmol kg−1, respectively. This increase was due to CaCO3 dissolution, 81 ± 13 and 34 ± 6 μmol kg−1 and ammonia secretion, 16 ± 2 and 14 ± 2μmol kg−1, respectively, for these species. Surveys conducted at a long-term monitoring site of community calcification (DK13) on One Tree Reef indicated that the density of sea cucumbers was approximately 1 individual m−2. We used these data and data from surveys at Shark Alley to estimate the dissolution of CaCO3 by the sea cucumbers at both sites. At DK13 the sea cucumber population was estimated to be responsible for nearly 50% of the nighttime CaCO3 dissolution, while in Shark Alley for most of the nighttime dissolution. Thus, in a healthy reef, bioeroders dissolution of CaCO3 sediment appears to be an important component of the natural CaCO3 turnover and a substantial source of alkalinity as well. This additional alkalinity could partially buffer changes in seawater pH associated with increasing atmospheric CO2 locally, thus reducing the impact of ocean acidification on coral growth.
 In order to understand the ability of coral reefs to grow, i.e., deposit CaCO3in an environment that is changing because of large-scale anthropogenic CO2 emissions, it is necessary to identify and quantify the influence of associated biota (e.g., coralline algae, mollusks, sponges, and echinoderms) on the CaCO3 budget of the reef. Estimates of the net deposition of CaCO3 in coral reefs typically involved determination of spatiotemporal changes in total alkalinity of the overlying seawater [Kinsey, 1985; Silverman et al., 2007b; Smith, 1973]. It is generally accepted that hermatypic corals and coralline algae are the primary building blocks of the coral reef framework. Therefore, it is very likely that they are responsible for most of the CaCO3 deposition [Kinsey, 1985]. There is a paucity of data on the role of noncoral invertebrates in the CaCO3 budget of coral reefs [Przeslawski et al., 2008]. Understanding the contribution of associated biota to reef construction and destruction is crucial especially as the ability to maintain a net positive CaCO3 flux into the reef is projected to decrease because of ocean acidification, ocean warming and increased nutrient loading [Hoegh-Guldberg et al., 2007; Silverman et al., 2007a, 2007b, 2009; Tribollet et al., 2009].
 Here we investigated the influence of the deposit feeding activity of two holothurians, Stichopus herrmanni and Holothuria leucospilota on reef CaCO3budget at One Tree Reef (OTR) in the southern Great Barrier Reef (Capricorn-Bunker Group). The influence of these species on reef CaCO3 budget was investigated in laboratory incubations and assessed with respect to their density in the field. This research was part of a detailed community metabolism study conducted at the historic monitoring station, DK13 [Kinsey, 1972]. We also investigated a population of sea cucumbers at an adjacent site at OTR [Eriksson et al., 2010]. Recent global stresses, i.e., ocean warming and acidification and mass bleaching, have been implicated in causing observed decline in coral CaCO3 deposition [De'ath et al., 2009; Hoegh-Guldberg et al., 2007]. As this declining trend progresses according to projections [Silverman et al., 2009] the relative importance of organisms that dissolve CaCO3 to the reefs CaCO3 budget will increase. This is examined here for the sea cucumber populations at OTR.
2. Methods and Materials
2.1. Animal Collection
 This study focused on two species of sea cucumbers that are relatively abundant in the OTR lagoon (Figure 1). Seven specimens of Stichopus herrmanni (mean wet weight 1.78, SE = 0.07; range 1.59–2.06 kg) and Holothuria leucospilota (mean wet weight 0.23, SE = 0.02; range 0.1–0.3 kg) were collected by snorkeling (Figure 1). S. herrmanni was collected from Shark Alley and H. leucospilota was collected from DK13. The animals were transported in buckets filled with seawater and placed in individual aquaria within 10 min of collection.
2.2. Experimental Procedure
 The sea cucumbers were incubated in individual aquaria for approximately 4 h during which time all of them expelled sediment through their cloaca. An eighth aquarium with no sea cucumber served as the control to assess background alkalinity changes due to evaporation. Because of difference in size, the S. herrmanni were incubated in 8 L of seawater while H. leucospilota were incubated in 4 L of seawater. Water samples were collected for total alkalinity (AT), Total ammonia nitrogen (TAN = NH3 + NH4+), total oxidized nitrogen (TON = NO2−1 + NO3−1). The pH of the water was measured at the start and end of each incubation.
2.3. Analytical Methods and Calculations
 AT was measured using the Gran titration procedure [Grasshoff et al., 1983], with HCl (0.05 M), on weighed samples of ∼20 g of seawater using an automated titrator (TitraLab-TIM90, Radiometer, Copenhagen). The precision of this method is better that ±2μmol kg−1. The acid was calibrated daily using Dickson seawater certified reference material [Dickson et al., 2007]. pH (NBS) was measured using a combined electrode PHC-2401 (Radiometer Analytical) connected to a PHM220 pH meter (Radiometer Analytical), with a precision of ±0.001–0.003 pH units. The pH electrode was calibrated using NBS buffers 7 and 10.012 (Radiometer Analytical).
 TAN was measured according to fluorometric method, protocol A described by Holmes et al. . Each sample was spiked with 1 ml of orthophthaldialdehyde (OPA) and was left in the dark for 3 h at room temperature. The fluorometric reading was done at a wavelength of 421 nm (Aquafluor, Turner Designs). The fluorescence readings were converted to ammonium concentration according to the calibration coefficients derived from measurements of 2.5 μmol/l ammonium standard in distilled water. The precision of this method is ±10 nM. Samples of TON were stored in 12 ml polyethylene vials and were acidified with 70 μl of 0.1 N HCl in order to prevent biological degradation. Concentrations were measured with a colorimetric method described by Hansen and Koroleff  using a Flow Injection Autoanalyzer (Lachat Instruments Model QuickChem 8000). The precision of this method is ±0.02 μM.
 The change in AT during the incubation (ΔAT(inc)) was calculated from the difference in the seawater AT values between the beginning and the end of the incubations. In addition we measured the addition of TAN to the seawater from the sea cucumbers digestive process and calculated the input to alkalinity (ΔAT(TAN)) from this source. The effect of TON on the alkalinity was small relative to changes in ΔAT(inc) and ΔAT(TAN) (see below) and therefore not considered in dissolution rate calculations. The difference between the change in alkalinity during the incubation and the change in alkalinity caused by the addition of TAN represents the dissolution of CaCO3 (ΔAT(CaCO3)) in the gut of the sea cucumbers (equation (1)).
The dissolution rate of CaCO3 per day (D) caused by the sea cucumbers was calculated with equation (2), where ρ is seawater density, V is the volume in liters of the water in the incubation aquaria and Ti is the incubation duration in hours. Dissolution is expressed in terms of moles CaCO3, therefore it is half the change in ΔAT(CaCO3) [Broecker and Peng, 1982].
2.4. Sea Cucumber Surveys
 The sea cucumber population was surveyed at DK13 in December 2009 and 2010. The surveys were conducted at night because the animals at this site are diurnally cryptic, emerging from under and around coral bommies and rubble for nocturnally feeding forays. The counts were done using four 50 × 2 m transects and each specimen was identified. Each transect was separated by 3 m and was placed perpendicular to the reef rim. The survey included exposed animals as well as ones under movable rubble. The sea cucumbers at Shark Alley were surveyed in May 2009 using similar transects placed along the base of the reef wall (within 8 m of the wall) at approximately 2–4 m depth. This area is mostly sandy with some rubble and scattered coral bommies. This survey focused on S. herrmanni which is abundant at this site, but all species were recorded. This species is epifaunal and is not cryptic during the day and so this survey was conducted during the day. The survey data were used to determine the contribution of the sea cucumber population to the CaCO3 budget of the reef at this location.
2.5. Data Presentation and Statistics
 All data in this work are presented as mean ± SE. The error in D was calculated by propagating the SD of the changes in AT (ΔAT(CaCO3)) through equation (2). Statistical analyses were done using Matlab with the statistics toolbox (The MathWorks).
3.1. Sea Cucumbers Survey
 The species identified in the surveys of the DK13 site included Holothuria leucospilota, H. atra, H. edulis, H. hilla, Stichopus chloronotus, and S. monotuberculatus (Figure 2). The mean density of all holothurians in 2009 survey was 1.1 m−2 (SE = 0.1, n = 4; range 1.0 –1.4 m−2), while in 2010 was 1.0 m−2, (SE = 0.1, n = 4; range 0.8–1.1 m−2) (Figure 2). The most abundant species at DK13 was H. leucospilota (mean density 0.8 m−2 in both 2009 and 2010). The total number of sea cucumbers in the community at DK13 station was not significantly different between 2009 and 2010 (ANOVA, p >0.84).
 The species identified in Shark Alley included H. atra, H. edulis. H. isuga, H. leucospilota, H. whitmaei, S. chloronotus and S. herrmanni (Figure 3). The mean density of all holothurians in 2009 was 0.3 m−2 (SE = 0.04, n = 12; range 0.1–0.6 m−2) (Figure 3). The most abundant species in Shark Alley was S. herrmanni, mean density 0.2 m−2. (SE = 0.04, n = 12; range 0–0.4 m−2).
 The average increase in alkalinity was 97 ± 13 and 47 ± 7μmol kg−1 (mean ± SE) for S. herrmanni and H. leucospilota, respectively (Figure 4). The TAN concentration increase during the experiments was 16 ± 2 and 14 ± 2 μmol kg−1 (mean ± SE) for S. herrmanni and H. leucospilota, respectively (Figure 4). The average change in TON during H. leucospilota incubations was 0.01 ± 0.03 μmol kg−1. Therefore, the net change in alkalinity associated with CaCO3 dissolution neglecting the small change in TON is 81 ± 13 and 34 ± 6 μmol kg−1 (mean ± SE) for S. herrmanni and H. leucospilota, respectively (Figure 4). The alkalinity increased at a rate of 1.17 ± 0.45 and 4.78 ± 1.75 mmol d−1, for H. leucospilota and S. herrmanni, respectively. For these species TAN secretion was 16.6 ± 3.3% and 29.3 ± 7.1% of the total alkalinity addition, respectively. The pH decreased for both sea cucumbers species. For S. herrmanni the pH decreased by 0.42 ± 0.01 from an initial value of 8.06 over the duration of the experiments and for H. leucospilota the pH decreased by 0.20 ± 0.01 from an initial value of 8.18 over the duration of the experiments.
3.3. CaCO3 Dissolution Rates
 The average dissolution rate of CaCO3 in the gut of H. leucospilota and S. herrmanni individuals calculated according to equation (2) were 0.41 ± 0.20 and 1.99 ± 0.85 mmol d−1, respectively.
 One Tree Reef supports a high density of sea cucumbers (approximately 1 individual m–2) with an unusually high density of Stichopus herrmanni [Eriksson et al., 2010; Lee et al., 2008]. The density of Holothuria leucospilota recorded at DK13 is similar to that recorded for other reefs [Kerr et al., 1993; Mangion et al., 2004; Roberts and Bryce, 1982]. Sea cucumber density had not changed significantly between measurements conducted at DK13 during December 2009 and December 2010 of 1.1 and 1.0 individual m–2, respectively, suggesting that the community is stable.
 The average dissolution rate of CaCO3 of the most abundant species, H. leucospilota at DK13 was D = 0.41 ± 0.20 mmol d−1 individual–1. For the larger species S. herrmanni collected from Shark Alley the average dissolution rate of CaCO3 was higher by a factor of ∼5, D = 1.99 ± 0.85 mmol d−1 individual–1. The net diel CaCO3 deposition rate measured at DK13 using the low tide slack water method [Kinsey, 1972] was 80 mmol m−2 d−1 (J. Silverman et al., manuscript in preparation, 2011). Measurement of water movement using fluorescent dye during low tide indicated that water was moving from the rim of the reef along grooves in the reef flat toward the lagoon at an average speed of less than 1 cm s−1 or 36 m h−1. Therefore, in order to estimate the total effect of an individual species in a consistent manner on the alkalinity change at DK13, which was measured at 3 h intervals (1.5 h before low tide and 1.5 h after low tide), it is necessary to sum the number of individual animals along the water path traveled in those 3 h and multiply it by the CaCO3 dissolution rate measured in the incubations. Thus, considering the density of H. leucospilota at DK13 (84.5 individuals per 100 m2 – 100 transect length × 1 m transect width, 2009 survey) the rate of CaCO3 dissolution is therefore 34.9 ± 17.8 mmol m−2 d−1, which is 48.4 ± 11.5% of the 72 ± 12 mmol m−2 d−1 of nighttime dissolution. It should be noted that the total number of sea cucumber individuals from all species in a 100 m2 transect averaged 112.8 indicating their effect may be larger. We do not have current data on the community CaCO3 precipitation of Shark Alley. However, it is probably lower than that of DK13 based on past estimates that showed a net diel CaCO3 precipitation in DK13 was 133 ± 54 mmol m−2 d−1 while in DK17 near Shark Alley it was 10 ± 10 mmol m−2 d−1 [Kinsey, 1979]. Considering the density of S. herrmanni at Shark Alley (14.2 individuals per 100 m2) the CaCO3 dissolution is 31 ± 14.7 mmol m−2 d−1. Therefore, considering the observed decrease in net calcification at DK13 between 1970s and 2009, it is likely that the CaCO3 budget of the OTR lagoon in the region of Shark Alley and DK17 is already negative.
 As seen at the DK13 site H. leucospilota and other sea cucumber species have nocturnal feeding and activity patterns [Hammond, 1982]. This suggests that their influence on the CaCO3 budget of the reef may be greater at night than during the daytime. The experiments presented here were conducted during the daytime suggesting that the rates reported here are likely underestimating the total diel rate of CaCO3 dissolution by sea cucumbers at OTR. This is supported by the fact that the dissolution of CaCO3 in the gut is a consequence of the feeding by the animal which is mainly nocturnal [Hammond, 1981, 1982; Jansen and Ahrens, 2004]. It should also be noted here that CaCO3 dissolution in coral reefs is caused also by the activities of endolithic organisms within the reef framework and by microbial activity in the sediments [Sanders, 2003; Tribollet et al., 2009].
 To endure, coral reefs must accumulate CaCO3 at a rate that is greater or equal than the sum of mechanically, biologically and chemically mediated erosion rates. The ability of coral reefs to fulfill this basic requirement is projected to and may already be severely compromised as a result of ocean acidification [Silverman et al., 2009; Tribollet et al., 2009]. The possibility that inorganic dissolution of metastable carbonate minerals in coral reefs can offset the effect of increasing atmospheric CO2 by increasing total alkalinity of seawater at least locally has been tested in a modeling study and shown to be ineffective [Andersson et al., 2003, 2005]. In a later experimental study it was shown that boring activity by endoliths increased under more acidic conditions because of increasing atmospheric CO2, however this study also concluded that the additional alkalinity would have a negligible effect on the carbonate system of seawater [Tribollet et al., 2009]. In DK13 the contribution of sea cucumbers to alkalinity (both carbonate and ammonium) is greater by a factor of 2–3 than the alkalinity production rates assessed for endolithic organisms in the hard substrate and carbonate sediments of coral reefs.
 Assuming that the population of sea cucumbers is stable as indicated by the 2 year survey (Figure 2) and long-term observations of the sea cucumber populations at OTR, an unfished reef, the relative importance of these deposit feeders to the reef's CaCO3budget will increase. This is especially relevant to OTR because it is a high-nutrient reef [Bell et al., 2007] that supports high rates of phytoplankton productivity and therefore it is likely that bioeroder activity is increased substantially [Glynn, 1997]. It should be noted that even in reefs with limited water exchange with the ocean bioeroders will not be sufficient to buffer the system from global effect such as ocean acidification as evident from the decrease in community calcification measured in OTR (J. Silverman et al., manuscript in preparation, 2011). However, it is very likely that this contribution may and is already reducing the potential impact of ocean acidification on calcification rates in OTR.
 Calculations from our measurements suggest that nearly 50% of the observed reef nighttime CaCO3 dissolution can be explained by sea cucumber digestive processes. While quantitatively uncertain, our results nonetheless indicate that the digestive physiology of sea cucumbers' may play an important role in the life of a coral reef. For the feeding biology of aspidochirotids this involves two key and contrasting processes, reef dissolution as documented here and enhanced productivity due to nutrient cycling [Uthicke, 2001a, 2001b; Uthicke and Klumpp, 1998]. There is an urgent need to understand the impact of removal of sea cucumbers through commercial harvest and other benthic invertebrates on reef health and resilience at a time when reefs face an uncertain future because of climate change.
 This research was supported by the Moore foundation. We thank the managers at One Tree Island Research Station a facility of the University of Sydney, Jennifer Reiffel and Russell Graham. We also wish to thank Christian Andreassi, Sebastian Schmidt-Roach, Victor Perez-Landa, Jon Fabricus-Dyg, and Mads Lichtenberg for assisting with the sea cucumber surveys and to Richard Stump for the sea cucumber collection.