Responses of the temperate calcareous sponge Grantia sp. to ocean acidification

Abstract Sponges are important components of marine systems globally, and while sponges have generally been shown to tolerate ocean acidification (OA), most earlier studies have focused on demosponges with siliceous skeletons. In contrast, little is known of how calcareous sponges, with calcite or aragonite skeletons, may react to OA conditions. Here we measured tissue necrosis and respiration rate of the temperate New Zealand calcareous sponge Grantia sp. to simulated OA. Our treatment conditions were based on the IPCC RCP8.5 (pCO2 1131.9 ± 113 μatm) scenario over a 28 day experiment, and responses were compared to current day control conditions (pCO2 512.59 ± 23 μatm). Sponge respiration rate was not significantly different between the control and treatment sponges and there was no evidence of tissue necrosis over the course of the experiment. Overall, our study is consistent with earlier studies on demosponges, showing calcareous sponges to be resilient to OA.


Introduction
Atmospheric carbon dioxide concentrations (CO 2 atm ) are currently around 417 ppm, which is over 100 ppm more than prior to the industrial revolution (NOAA, 2021).CO 2 atm levels are predicted to rise beyond 1000 ppm by the year 2100 if no action is taken to reduce CO 2 emissions (IPCC, 2021).Due to its buffering abilities, the ocean has absorbed approximately a quarter of the CO 2 that has been released into the atmosphere from anthropogenic activities, resulting in decreased ocean pH (IPCC, 2021).Under the worst-case scenario, by 2100 ocean pH is predicted to decrease by a further 0.3-0.5 units (Orr et al., 2005; The Representative Concentration Pathway [RCP] 8.5; IPCC, 2021).This changing ocean chemistry is leading to an increase in hydrogen ions (H + ) in a phenomenon known as ocean acidification (OA), which has the potential to impact a wide range of marine organisms (Caldeira and Wickett, 2003;Kroeker et al., 2010;, 2013).
There are a number of ways that marine organisms can be impacted by OA, although most focus has been on calcifying organisms, particularly molluscs, echinoderms, corals and crustaceans (Medeiros and Souza, 2023).These impacts include changes to growth, metabolism, acid-base balance, calcification, survival rates, settlement, reproduction and cell signalling processes (Espinel-Velasco et al., 2018;Melzner et al., 2020;Medeiros and Souza, 2023).Of particular concern is the dissolution of calcium carbonate structures, particularly as a result of biominerals being prone to dissolution when seawater is undersaturated with respect to calcium carbonate (Melzner et al., 2020).While much of the focus of such dissolution has focused on coral reefs (Cornwall et al., 2021), many other organisms with calcium carbonate structures are at risk.Furthermore, disruption of acid-base regulation ability and cell signalling pathways may actually interfere with the process of calcification either directly or indirectly (see Melzner et al., 2020 for review).
Sponges are abundant benthic organisms in temperate, polar, and tropical marine ecosystems (Bell et al., 2020).As sedentary suspension feeders, sponges are integral in linking benthic and pelagic ecosystems, filtering large volumes of water per unit body mass, retaining small particles and nutrients, and providing a link to higher trophic levels (Turton et al., 1997;Bell, 2002;Perea-Blázquez et al., 2010;De Goeij et al., 2013).Due to their sessile nature and the presence of an aquaferous system, sponges are continuously exposed to ambient seawater, potentially increasing their vulnerability to alterations in seawater chemistry (Bergquist, 1978;Aguilar-Camacho and McCormack, 2017).However, despite this, some sponges can tolerate changes in salinity (e.g.Leamon and Fell, 1990) and have been shown to both osmoregulate (Brauer, 1975) and control intracellular pH (e.g.Webb et al., 2019).These abilities of sponges may enable them to cope with changes in the pH of the surrounding seawater, although this might come with increased metabolic demand.
A number of studies have shown tropical sponges to be resilient to climate change impacts (Fabricius et al., 2011;Wissak et al., 2012;, 2014;Fang et al., 2013;Bennett et al., 2016;Bell et al., 2018;Agostini et al., 2018), but far less known about temperate sponges (but see Goodwin et al., 2013;Bates and Bell, 2018).Importantly, the majority of all previous sponge climate studies have focused on demosponges with siliceous skeletons, with much less know about climate impacts on calcareous sponges species that have spicules made of calcium carbonate (see Bell et al., 2018).Class Calcarea make up approximately 5-8% of the phylum Porifera, with many species still undescribed (Uriz, 2006;Smith et al., 2013).Calcareous sponges are often abundant in caves and other dark environments where they can be more abundant than demosponges (e.g.Bell, 2002;Rapp, 2006;Fromont et al., 2012).Calcareous sponges are the only sponge class with spicules formed of calcium carbonate (Jones, 1970).Therefore, these sponges may be more vulnerable to OA impacts compared to demosponges.However, earlier studies have found the spicules of calcareous sponges to be surrounded by an organic sheath (Jones, 1955), which has the potential to prevent dissolution of spicules once formed.
To date, only two experimental laboratory studies have considered the impact of OA on non-tropical calcareous sponges (Peck et al., 2015;Ribeiro et al., 2020).Peck et al. (2015) found that Leucosolenia sp.increased in abundance 2.5-fold in low pH (pH 7.7, pCO 2 not given) conditions compared to controls and suggested that sponge resilience may be a result of the organic sheath described above (see Jones, 1955).More recently, a 9-day laboratory study by Ribeiro et al. (2020) examined the effect of temperature, pH, and their combined effects on the skeleton and microbial community of the calcareous sponge Sycettusa hastifera (Ribeiro et al., 2020; treatment conditions: control: 22 °C, pH 8.1, pCO 2 301 ± 30 μ atm; low pH: 22 °C, pH 7.6, pCO 2 1182 ± 119 μ atm; high temperature: 26 °C, pH 8.1, pCO 2 278 ± 319 μ atm; combined effects: 26 °C, pH 7.6, pCO 2 1156 ± 167 μ atm).These authors found that none of the treatments caused any significant sponge degeneration, but they did find that higher temperature affected spicule shape, and sponges in the reduced pH treatment had smaller spicules in their exterior layer (Ribeiro et al., 2020).
These limited earlier studies show the potential resilience of calcareous sponges to OA, but longer experiments with more species are needed before any generalisations can be made.We hypothesise that calcareous sponge respiration rates will increase as a result of OA impacting cellular processes, particularly the maintenance of intracellular pH.To test this we conducted a 28-day experiment, to measure respiration rate (as a measure of metabolic demand) and tissue necrosis in the calcareous sponge Grantia sp.Sponges were exposed to pH 7.6 (treatment, pCO 2 1131.9 ± 113 μ atm) and pH 8 (control, pCO 2 512.59 ± 23 μ atm), consistent with the IPCC RCP8.5 scenario for 2100 (IPCC, 2013).

Materials and methods
Grantia sp. is a common New Zealand sponge that has an offwhite colouration and a network of delicate erect singular tubes, with oscula that are slightly hispid (Dendy, 1924).This species is typically found in rocky caves and on vertical reef walls that experience high levels of turbulence and water flow.A 28-day laboratory experiment was undertaken to investigate how our study species is affected by the IPCC worst case scenario RCP8.5 conditions.We compared sponges kept in three replicate treatment tanks (pCO 2 1131.9 ± 113 μ atm) to three control tanks (pCO 2 512.59 ± 23 μ atm) (Bates and Bell, 2018; Figure S1).Eighty-one Grantia sp. were collected via SCUBA from Breaker Bay, Wellington, New Zealand (41°19 ′ 58 ′′ S, 174°49 ′ 52 ′′ E) (see Bates and Bell, 2018 for location).
Sponges were attached to ceramic tiles and haphazardly placed in tanks, with a total of 13 sponges in each tank (6 tanks in total) and acclimated for two weeks, after which a secondary acclimation period began where the pH was reduced in the treatment tanks by 0.1 units per day until the desired pH/pCO 2 was reached to simulate a gradual OA decline (e.g.Johnson et al., 2019).A Neptune APEX controller (Neptune System LLC, USA) was used to keep the tanks within 0.1 unit of the target value via the slow-release bubbling of CO 2 from a solenoid box connected to a 6.8 kg CO 2 cylinder.The sponges were kept at the average mean temperature (13 °C) for the Wellington South Coast (Greater Wellington Regional Council, 2021), which was also similar to the temperature at the time of collection.This temperature was maintained inside a 100L header tank (see Figure S1) using a heater and chiller that were controlled automatically using the APEX system.The water from a main tank was transferred into individual secondary header tanks (see supplemental information for more information), which then flowed into a corresponding tank where the 13 sponges per tank were housed (n = 39 per treatment).Temperature and pH were monitored weekly (Supplemental Table S1) throughout the experiment using the APEX probes and real-time measurements of pH(T) (mV) using a HQ40d portable pH multi-parameter (HACH, USA).To determine the carbonate chemistry of the system pHT [H + ], total alkalinity (TA), temperature and salinity were measured throughout the experiment to complete the calculations using dissociation constant of water and Henry's law (Zeebe, 2012) (Tables 1 and S1).
At T0, baseline respiration measurements were made for three randomly selected sponges per tank (meaning a total n = 9 per treatment), thereafter respiration rates were measured on day T3, T6, T14, T17, T24, and T28 (Tend).Randomisation was achieved using the numbers on the tiles and random number tables.We used a similar method as Bates and Bell (2018), Cummings et al. (2020), andMicaroni et al. (2021).See supplemental information for further details.
All statistical analyses were performed in R version 3.1.3(R Core Team, 2013).The effects of pCO 2 (ambientcontrol), RCP8.5 (treatment) and time (day of measurement throughout the 28 day trial, T0-T28) were tested using linear mixed-effects models (LMM) with normally distributed errors and random intercepts (lmer, lme4 package; Bates et al., 2015).Treatment and time were considered fixed effects and experimental tank was considered a random effect.An experimental tank effect was included to address pseudoreplication and possible tank effects (Hurlbert, 1984).For all the models, fixed-and random-effect terms were tested using the function anova and ranova (R package lmerTest, Kuznetsova et al., 2017).The goodness of fit, normality and homoscedasticity of the errors were checked for all models by inspecting plots of the normalised residuals and the quantile-quantile plots (see Figure S3).For mixed-effect models there were slight concerns about normality of residuals, so the analyses were supplemented with a more robust non-parametric model (univariate PERMANOVA, Anderson, 2001Anderson, , 2014)).

Results
There were no signs of tissue necrosis in any of the control or treatment sponges over the course of the experiment.Respiration rate of Grantia sp.(Figure 1) fluctuated over the course of the experiment for both the control and treatment sponges.The LMM analysis determined that treatment had no significant effect on the sponge respiration (F 1,102 = 2.18, p = 0.14; Table S2).However, there was a significant effect of time on respiration rate (F 1,102 = 9.27, p = 0.003; Table S2).Both treatment and controls experienced an approximate 50% reduction in mean respiration rate over the experiment based on a comparison between T0 and T28 (Figure 1).The PERMANOVA model confirmed the results of the mixed-effect model (Table S3).

Discussion
While there is a considerable amount of data available on the impacts of OA and warming on demosponges, we know comparatively little about calcareous sponges.Grantia sp.showed no signs of tissue necrosis over the course of the experiment.We also found no impact of our OA simulation on the respiration rate of sponges, although there was an effect of time.Our results are consistent with the emerging picture that calcareous sponges show tolerance to changes in the pH of the external environment.
Reduced ocean pH can cause serious physiological stress to marine organisms, resulting in alterations to respiration rates as they use their energy to activate detoxification and survival mechanisms (Sokolova et al., 2012).Several recent reviews have examined the impacts of OA on the physiology and ecology of marine invertebrates.Both Shi and Li (2023) and Medeiros and Souza (2023) identified a number of broader direct and indirect impacts on marine invertebrates including impacts on calcification, behaviour, immunity, energy budget, metabolism, growth, development, genetics, oxidative stress, and disruption of acidbase balance.These reviews highlight the wide range of invertebrates that are negatively impacted by OA.However, results for our study species, and the previous studies on OA impacts on calcareous sponges (see Peck et al., 2015;Ribeiro et al., 2020) contrast with reported impacts on other organisms with calcareous skeletons.For example, calcium carbonate accretion by reef building corals in tropical ecosystems is expected to be reduced 156% by 2100 under the worst case RCP 8.5 scenario (Cornwall et al., 2021).However, no negative physiological impacts were found for our study species.A number of previous studies have suggested that sponges may be winners in future warmer and more acidic oceans (e.g.Bell et al., 2013), however, this hypothesis is based largely on tropical demosponges, with far less known about the climate change impacts on temperate species, and particularly calcareous sponges.Our results support this hypothesis for temperate calcareous sponges, albeit based on a single species, at least in responses to OA.
We found no significant differences in the respiration rate of Grantia sp. between the control and treatment sponges, however, we did find a significant effect of time in both treatments.Both control and treatment sponge respiration rates declined over the course of the 28 day experiment however, the sponges did not show any other signs of deterioration over this period, suggesting that they stayed healthy in experimental set up.We believe the decline in respiration rates over time was likely due to lower food in our experimental system compared to natural environments.Our respiration results contrast with those for a noncalcareous temperate sponge, Tethya bergquistae.For this sponge, respiration increased in response to low pH conditions compared to the controls (treatment pH 7.6; pCO 2 1514 ± 13.6 μ atm).(Bates and Bell, 2018).Intracellular pH (pHi) directly impacts basic cell functions (Casey et al., 2010), therefore, the increase  in respiration rate of the T. bergquistae could be a result of cellular acidosis as seen with increased temperature in corals and their symbionts (Gibbin et al., 2015), and with decreased pH in corals and other marine invertebrates (Orr et al., 2005).However, consistent with our study, Bates and Bell (2018) did not find a significant change in the respiration rate of a second demosponge species, Crella incrustans, which was exposed to the same conditions.Species specific differences might occur due to differences in acid-base regulation abilities (Pörtner et al., 2004), which may explain why T. bergquistae has a different response than either C. incrustans or our study species, although little is known about acid-base regulation specifically in sponges and needs further investigation.

Conclusion
We found no evidence for negative impacts of reducing ocean pH on the calcareous sponge Grantia sp., with our study providing evidence that despite calcareous sponges having calcareous spicules they are resilient to the short term impacts of changing pH.However, the exact mechanisms of tolerance still require further investigation, particularly around how sponges either regulate internal pH conditions or how they tolerate changes to pH without increasing energy expenditure.

Table 1 .
(Robbins et al., 2010)) and calculated (**) seawater parameters represented as the mean (±SD) of measurements taken during the acclimation period and weekly during the experiment (n = 4 sampling periods) Measured values of temperature, salinity, pH (pHT ), and total alkalinity (TA) were measured in all aquaria weekly.pCO 2 (μ atm) and seawater saturation states (Ωca and Ωar) were calculated by entering the recorded values of temperature, pHT, salinity and TA into the CO 2 calc software(Robbins et al., 2010).