Long-term study of the combined effects of ocean acidification and warming on the mottled brittle star, Ophionereis fasciata Open Access

Climate change is driving changes to the oceans, increasing sea surface temperature (ocean warming, OW) and decreasing pH (ocean acidification, OA), alongside the impacts of ocean pollution, environmental degradation and overfishing (Bindoff et al., 2019; Caldeira and Wickett, 2003; Orr et al., 2005; Sabine et al., 2004). Simultaneous changes to these physical and chemical properties make it a challenge to assess their combined impact on the biological processes of marine biota (Bednaršek et al., 2021; Byrne and Przeslawski, 2013; Gattuso et al., 2015; Kroeker et al., 2013; Whiteley, 2011). The study of multiple environmental drivers (e.g. OW and OA) and their effect on marine ecosystems remains limited, partly as a result of the logistical challenges of multi-factorial experiments (Boyd et al., 2018; Widdicombe et al., 2010). However, new approaches that link biological processes and abiotic environmental conditions (Boyd et al., 2018) have led to a rapid increase in studies assessing responses of marine organisms to a changing ocean. For example, studies evaluating the response of echinoderms to OW and/or OA grew from 19 papers published by 2010 (Dupont et al., 2010), to 53 by 2013 (Kroeker et al., 2013; Wittmann and Pörtner, 2013), 249 in 2023 (Márquez-Borrás, 2023) and 299 in 2024 (Table S1).

Two-thirds of the studies assessing the response of echinoderms to OA and/or OW have used echinoids (around 213 published works), with almost half of these performed on pre-metamorphic stages (Table S1). In contrast, no published study has considered effects of a changing ocean on crinoids, and only a few studies have been performed on asteroids (50 studies) and holothuroids (33 studies). Ophiuroids, which are the largest class of echinoderms (O'Hara et al., 2018; Stöhr et al., 2012), are globally diverse and ecologically important in most marine habitats but are poorly studied in the context of a changing ocean, with only 12 species studied over 19 publications (OW, OA or both; Fig. S1A).

The mottled brittle star (Ophionereis fasciata) is an endemic New Zealand ophiuroid found in the intertidal and in subtidal mudstone channels, lying under boulders on patches of sand, small stones and shells (Mills and O'Hara, 2013; Pentreath, 1971). The geographic range of O. fasciata includes most of New Zealand, predominantly in the southern region with a distribution extending as far north as the Bay of Islands (Mills and O'Hara, 2013). Some physiological and developmental aspects of the mottled brittle star have been investigated (Falkner et al., 2006, 2015; Selvakumaraswamy and Byrne, 2000), including a study assessing the impact of temperature on its metabolic rate (Pentreath, 1971). However, to date, no study has examined the combined effects of low pH and high temperature on this widespread New Zealand species.

Here, we examined the effect of high temperature (21–24°C) and low pH (7.75–7.6) on O. fasciata under conditions experienced only in its tidally influenced habitat (short term) as well as changes expected to occur in the near future in its subtidal habitat due to OW–OA (Atkins, 2014; Evans and Atkins, 2013; Foley and Carbines, 2019; Law et al., 2018a; Pentreath, 1971; Vance et al., 2020). We measured survivorship, respiration rate, growth rate, righting response and calcification over 15 weeks using a mesocosm experiment with individual (temperature, pH) and combined drivers. We hypothesized that as an intertidal species that experiences natural daily fluctuations in temperature and pH, this brittle star would be partially resilient to mid-temperatures and mid-pH. However, it was predicted that O. fasciata would display strong stress responses as conditions got closer to its maximum threshold. Moreover, simultaneous exposure to high temperature and low pH was expected to exacerbate deleterious effects on the mottled brittle star. We address the knowledge gap of the impact of a changing ocean on ophiuroids and propose that this class is an ideal alternative echinoderm for future global change research.

Ophionereis fasciata Hutton 1872 individuals (disc diameter 11–18 mm) were collected by SCUBA from 2–8 m depth at Matheson's Bay, New Zealand (36°18′13.56″S, 174°47′58.09″E) during the late austral summer of 2021. Animals were transported to the University of Auckland in 50 l sealed plastic containers with ambient seawater and placed in aerated tanks (130 l) with flow-through filtered seawater (FSW; 1 μm filter) previously adjusted to average ambient conditions during winter–spring (pH≈8.01, 18°C). A 3 week acclimation was performed prior to experiments in a temperature-controlled room with a 12 h:12 h light:dark cycle to minimise the initial stress response. During the acclimation, and later experimental period, the brittle stars were fed with commercial coral food and algal pellets (Vitalis, Worldfeeds, Thorne, UK) once a week (7 mg of coral pellets and 8 mg of algal pellets per individual per feeding). After the acclimation period, each specimen was photographed with a scale to provide a photo ID for individual identification over time, as well as length measurements (disc diameter, longest arm length).

Three different temperatures were chosen to test the thermal tolerance of the mottled brittle star (18, 21 and 24°C), together with three different levels of pH (8.01, 7.7 and 7.6). The temperature of 21°C and pH 7.7 correspond to the mean maximum/minimum sea surface conditions at Leigh (2 km away from Matheson's Bay) during summer. The temperatures of 24°C and pH 7.6 correspond to extreme concurrent conditions currently observed only in tide pools, where intertidal animals are exposed and survive over the short term. The effects of simultaneous exposure to high temperature and low pH were assessed with the control treatment set as 18°C and pH 8.01, representing current average conditions experienced by O. fasciata during autumn–winter (Atkins, 2014; Evans and Atkins, 2013; Vance et al., 2020).

Experimental pH treatments (8.01, 7.7 and 7.6) were generated by continually bubbling mixed pure gaseous instrument-grade CO₂ (99.98%, BOC) with dried CO₂-free atmospheric air into three 50 l mixing tanks (MixTs) containing FSW using calibrated mass flow controllers (Smart Trak2, Sierra Instruments, Monterey, CA, USA) (based on Fangue et al., 2010; Hudson and Sewell, 2022; Sewell et al., 2021). Each MixT was used to provide the pH treatment waters for the static 10 l experimental tanks (ETs); i.e. 10 l of water was transferred from the MixT into the three independent replicate ETs of each pH treatment, each containing seven randomly assigned O. fasciata individuals. Both the MixTs and the ETs had plastic lids to minimise gas exchange with the atmosphere; pH was maintained in each ET by continuous bubbling of CO₂ through tubing connected to the appropriate mass-flow controller.

The three replicate pH treatment ETs were assigned to the three temperature treatments using a randomized block design (Fig. S1C). Insulated water-baths were used to keep the ETs at a constant temperature, with submersible water pumps placed in each bath to keep homogeneous temperatures. During water changes, detailed below, FSW from the appropriate MixT was pre-heated to the treatment temperature by placing containers of water into the water-bath until the target temperature was reached.

After laboratory acclimation, brittle stars inside the ETs were gradually introduced first to temperature (1°C per day) and then to pH treatments (0.07 units per day) to minimise stress responses. After reaching the desired conditions (week 0), the animals were held in the ETs for 15 weeks with a weekly adjusted light:dark regime reflecting diurnal sunrise/sunset. This duration under stable treatment levels (experienced by this species in its subtidal habitat) was selected to assess a new physiological state instead of a shock response. ETs were cleaned every third day to remove faeces and uneaten food, resulting in the gradual change of one-half of the experimental water on each occasion. ETs were refilled using water from the appropriate pH MixT, pretreated to the experimental temperature, as described above. Nitrogenous waste (ammonia, nitrites and nitrates) was monitored and controlled throughout the study with a test kit (API Marine, Saltwater test kit). PVC plates (dark grey, 14 cm×14 cm×0.4 mm) were placed in each ET, where the brittle stars could hide. All the ETs were monitored daily for signs of stress or mortality of the brittle stars. Individuals that were dead or showed visible signs of stress (e.g. degenerating/ragged disc, excessive mucus secretion and/or multiple arm autotomy) were removed, and the date, treatment and photo ID were recorded.

Ophionereis fasciata experiences daily highly variable temperature and pH conditions in its intertidal habitat; such variation is attenuated in magnitude and time (seasonally) for subtidal populations. In this sense, our experimental design took into account the expected stable conditions for the collected specimens (subtidal), assessing O. fasciata's potential plasticity (given the depth range of this species) and the effects of temperature/pH changes after a long-term exposure.

The stability of physicochemical parameters (pH, temperature, dissolved oxygen, salinity) was continuously monitored throughout the entire 15-week experimental period. Temperature was continuously monitored (recorded every 10 min) in each water-bath using submersible HOBO temperature loggers (Onset HOBO MX2201). pH on the total scale (pH_T) was determined spectrophotometrically using m-Cresol Purple (Sigma-Aldrich) every 2 days from one random ET from each of the nine treatments based on SOP6b, calibrating with Tris standard (SOP3a; provided by Dr Kim Currie and Judith Murdoch, NIWA, New Zealand) (Dickson et al., 2007). Dissolved oxygen, pH (calibrated with pH_T) and salinity were measured daily with a YSI meter (Xylem, Yellow Springs, OH, USA). Finally, every week, 500 ml seawater samples were removed from one random ET per treatment and immediately poisoned with saturated HgCl₂ for total alkalinity analysis (TA). TA was determined using potentiometric open-cell titration (T50 Titrator, Mettler-Toledo, Greifensee, Switzerland) as per Fangue et al. (2010). Seawater carbonate speciation within ET was calculated from measured parameters (salinity, temperature, TA, pH_T) using the seacarb (v.3.3.1) package in RStudio (https://CRAN.R-project.org/package=seacarb). All experiments were carried out during the dormant period of reproductive activity of O. fasciata to reduce the impact of the gametogenic cycle on metabolic demand and energy allocation.

The weight-specific oxygen uptake rate (Ṁ_O2; mgO₂ h⁻¹ g⁻¹) of individual brittle stars was determined using intermittent-flow respirometry. Double-bottomed Plexiglas chambers (volume 475 ml) were filled with fresh FSW from the appropriate exposure treatment for each run, hermetically sealed, and submerged in a water-bath to maintain the experimental temperature. A magnetic stirrer was placed in the bottom of the chamber to recirculate the water and homogenise the oxygen concentration. Ṁ_O2 experiments were conducted at least 3 days after the weekly feeding to minimise any confounding effect of the gut contents.

A subsample of five intact specimens from each ET was individually transferred into respiration chambers (separated from the magnetic stirrer by a mesh cage). The experimental set-up (sealed respiration chambers inside the water-bath) was placed on magnetic drive plates and covered with black plastic to provide a darkened habitat similar to that in natural conditions. The animals were allowed 30 min to acclimate and position themselves in the chamber prior to measurement of respiration. After this acclimation period, dissolved oxygen (DO) concentration (%) was recorded every 15 s for 90 min with a calibrated fibre optic oxygen sensor (PreSens oxygen sensor spot, type SP-PSt3-NAU, Precision Sensing GmbH, Regensburg, Germany) placed in the lid of the respiration chambers that was connected to an Oxy-4 Mini multichannel fibre optic oxygen transmitter (PreSens). Oxygen saturation did not fall below 85% in the chambers at any time. If values approached 85%, the chambers were flushed with fresh treatment FSW to re-oxygenate the respirometry chamber before recordings were continued.

Two 90 min runs were performed (and Ṁ_O2 averaged) for each specimen with a 5 min flush cycle between runs. At the end of the second run, the wet mass of each brittle star was measured (nearest 0.01 g) after gently blotting them dry with a paper towel. Every 2 days, recordings from chambers with FSW but no brittle star were used to estimate background respiration rates by microbial action; a correction was applied to experimental calculations. Ṁ_O2 experiments were run on weeks 0, 5, 10 and 15 after reaching the treatment conditions, using the same specimens from each ET when possible. Respiration rate was calculated as oxygen consumption rate (Ṁ_O2; corrected by volume) using the R package respR (Harianto et al., 2019). For comparison across different studies, Q₁₀ values were used to describe the temperature sensitivity of respiration. We used the standard equation: Q₁₀=(R₁/R₂)10/T₁−T₂, where R₂ is the Ṁ_O2 at T₂ (21 or 24°C), and R₁ is the Ṁ_O2 at T₁ (18°C) for each of the measured intervals (e.g. 18–24°C).

At week 0, autotomy was induced on one arm of two selected specimens from each ET by applying gentle pressure with a scalpel blade across an inter-vertebral plane of the arm, about one-third of the distance from the arm tip. For organisms with recent injuries (due to stress or sublethal predation before collection), the injured arm was autotomised at a level corresponding to one-third of the lost tissue. Progress of regeneration and disc size were monitored weekly for 15 weeks post-amputation. Length of arm lost (LL), regeneration length (RL) and differentiated length (DL) were measured for each specimen. Disc size, longest arm, LL, DL and RL were measured using image analysis software (Photoshop v.23.5.1) calibrated with scaled photographs. Regeneration rate (RR) and differentiation rate (DR) were calculated as the slope of the significant simple linear regression between the RL or DL (mm), respectively, and time (days). Individuals generally showed a reduction in disc size; this was recorded as a percentage of the total disc diameter at the beginning of the experimental period.

Animals from all ETs were tagged with calcein, a fluorochrome that leaves a green fluorescent mark on the growing edge of calcified elements. Brittle stars were submerged for 24 h in a solution of 15 mg l⁻¹ calcein after the acclimation phase, following the protocol used by Medeiros-Bergen and Ebert (1995) for Ophionereis annulata. Calcein stock solution (3 g l⁻¹) was made by dissolving 0.3 g of the fluorochrome in 100 ml of tap water and buffered to a pH of 7 using sodium bicarbonate (NaHCO₃). Twenty specimens at a time were incubated in aerated 20 l FSW buckets with 100 ml of stock solution. Following the incubation, the animals were washed in FSW several times.

After 15 weeks of experimental exposure, mid-positioned teeth, mid-ventral proximal arm spines, and vertebrae from the base of the arm (before the eighth segment) from four selected specimens from each ET were isolated. Soft tissue was removed by soaking two segments or a portion of the disc in 10% sodium hypochlorite solution. Isolated ossicles were then rinsed 3 times with both distilled water and 95% drum ethanol and air dried for 24–48 h at room temperature. Ossicles were examined under a Leica microscope fitted with a GFP filter cube (450–490 nm).

Growth measurements were made on one tooth, one spine and two vertebrae of each brittle star based on the fluorescent marks at the time of tagging. Vertebral growth (%) was determined from the average distance between the fluorescent marks (experimental growth) and the distal border of the vertebrae, divided by the distance between the centre and the fluorescent marks (pre-experimental growth) (based on Ravelo et al., 2017). Tooth growth (%) was calculated as the average distance between the fluorescent mark and the edge of the tooth (articulation point with the dental plate) divided by the total length of the tooth. Arm spine growth (%) was calculated as the average distance between the tagged growth layer and the external edge of the spine, divided by the total length of the spine. All measurements were made using image analysis software (Photoshop v.23.5.1) calibrated with scaled photographs.

The fine-skeletal morphology of the outer surface of ossicles (mid-ventral spine, lateral arm plate, tentacle scale and mid-tooth) in four O. fasciata from each ET was used to determine differences in calcification/dissolution among treatments after 15 weeks of experimental exposure. Soft tissue was removed as described for the calcein-stained ossicles. Ossicles were mounted on aluminium stubs (Ted Pella, Inc., Redding, CA, USA) using carbon adhesive tabs and coated with gold for imaging with an FEI Quanta scanning electron microscope (SEM) in the Research Centre for Surface and Materials Science at the University of Auckland. From each brittle star, two teeth, a tentacle scale and a lateral plate were analysed, with images captured at two different magnifications, (1) the whole ossicle and (2) a close magnification to observe the outer stereom in the central region. Similarly, two arm spines from each individual were used to obtain images capturing the whole spine and a higher magnification to capture the tip, mid-region and base stereom. All images were examined for any visible differences between treatments, including uniformity/geometry of the stereom, signs of degradation, thickness of the mesh and pore sizes.

After the 15 week experimental exposure, all specimens were categorised as intact or stressed depending on the state of the disc and arms. Stressed individuals included those brittle stars with a damaged disc, evidence of multiple autotomy events and/or excessive mucus secretion. The mean number of specimens in each category (intact, stressed, dead) and the mortality values per treatment were analysed.

On weeks 0, 5, 10 and 15, the righting response of five selected specimens from each ET was measured. Righting is a behavioural response of brittle stars to return to a more natural, ventral-down position when flipped over, and is a proxy for physiological state (Brothers and McClintock, 2015; Schram et al., 2011; Wood et al., 2010). Each individual was gently placed in the middle of a tray filled with enough experimental FSW to completely cover the specimen. Righting time, from the release moment to the time when the organism placed all arms flat on the surface, was measured 3 times, recording the average time to 0.01 s using a digital stopwatch. The righting time was tested before measuring body mass and Ṁ_O2 runs to avoid overstressing the individuals.

Effects of treatments on Ṁ_O2, arm RR, arm DR, percentage growth rate, survival and righting time were assessed using a two-way analysis of variance (ANOVA) with temperature and pH as fixed factors. The assumptions of ANOVA, homogeneity of variance and normality, were tested with Bartlett's test and Shapiro–Wilk test, respectively. Post hoc analyses were performed using Tukey's HSD test (P=0.05) to detect significant differences between treatments. All statistical analyses were performed in R (v.4.2.1) using RStudio (2022) with the default stats package. Data for Ṁ_O2, RR, DR were square root transformed, while righting time was log transformed. The differences between treatments were analysed with a two-way ANOVA using temperature and pH as fixed factors for each dataset (week 0 and 15). Additionally, differences over time for Ṁ_O2 and righting time were analysed with a one-way ANOVA using treatment time (week) as a fixed factor. Raw data are available in Dataset 1.

Seawater temperature, salinity and pH/carbonate parameters within the ETs remained relatively stable throughout the 15 week experimental period (Table 1). The original target pH for the low pH conditions was not achieved (7.6 instead of 7.4), resulting in a similar pH_T between the medium and low pH treatments. The pH/P_CO2 treatments ranged from pH 7.59 to 7.93 and from 529 to 1344 μatm. The aragonite saturation only approached 1 in the high P_CO2 treatment at control temperature.

No mortality was recorded during the 3 week acclimation period or during the subsequent thermal or pH ramping periods. After the 15 week exposure period, elevated temperature had a significant negative effect on the survival of O. fasciata (Fig. 1A, Table 2). Only individuals from the high-temperature treatment displayed a clear pattern of higher mortality, with no statistically significant difference between pH treatments and no interaction between drivers (Fig. 1A, Table 2).

The differences in survival among temperature treatments were significant between 18 and 24°C and between 21 and 24°C, but not between 18 and 21°C (Tukey's HSD, P<0.05; Fig. 1A). Mortality in the mid-temperature treatment was seen around week 14, while mortalities were steadily observed throughout 15 weeks for the highest temperature, starting from week 0 (Fig. 1A). Signs of stress showed the same pattern as survival, with a decreasing proportion of intact animals as the conditions of both temperature and pH were more acute (Fig. 1B–E, Table 2).

At the start of the experiment (week 0), Ṁ_O2 in O. fasciata was significantly affected by temperature and pH, with no significant interaction between drivers (Table 2). Ṁ_O2 at week 0 decreased by up to 26% with warming and 12% with low pH (Tukey's HSD, P<0.05; Fig. 2A). After 15 weeks of experimental exposure, the mottled brittle star exhibited a significant difference in Ṁ_O2 as a result of elevated temperature and decreased pH, with a significant interaction between drivers (Table 2). Ṁ_O2 increased by up to 23% with warming and decreased up to 56% with low pH, and there was a combined effect of the two drivers (Tukey's HSD, P<0.05; Fig. 2B).

Time of exposure to the different treatments had a significant effect on the Ṁ_O2 of the mottled brittle star (Table 2). The mean Ṁ_O2 of animals from most treatments was at its highest point after reaching temperature and pH conditions (week 0). In the following days, Ṁ_O2 progressively declined to reach its lowest values by the end of the experimental period (Fig. S1D). Q₁₀ for O. fasciata was 2.18 for the temperature range 18–24°C, 1.62 for 18–21°C, and 2.31 for 21–24°C, with a decreasing trend as exposure time increased, after reaching a peak in week 10 of the experimental period (Fig. 2C).

After 15 weeks of experimental exposure, O. fasciata exhibited a significant difference in RR due to elevated temperature, but not pH, with no interaction between drivers (Table 2). There was an inverse relationship between RR and experimental temperature, with animals in the control and medium temperature treatment showing significantly higher RR than those in the highest temperature treatment (Tukey's HSD, P<0.05; Fig. 3A). Functional recovery (DR) was negatively affected by increased temperature but not pH, with no interaction between drivers (Table 2). Animals in the control and medium temperature treatment showed increased DR compared with those at high temperature (Tukey's HSD, P<0.05; Fig. 3B).

Mean percentage growth of vertebrae was significantly higher in brittle stars exposed to medium temperature (Table 2) compared with those in the control and high temperature treatments (Tukey's HSD, P<0.05; Fig. 4A). Ophionereis fasciata exhibited no differences in growth rates between pH treatments or the interaction between drivers for any of the ossicles assessed (vertebrae, arm spines or teeth; Table 2). Temperature, but not pH or the interaction between them, had a significant effect on the reduction of disc size (Table 2). Reduction in size was magnified as temperature increased (Tukey's HSD, P<0.05; Fig. 4B).

Scanning electron micrographs of arm spines, lateral arm plates, tentacle scales and teeth revealed differences between treatments, with malformation and traces of degradation more evident in ossicles from animals reared at elevated temperature and low pH (Fig. 5). This was true for the whole spines/teeth and under high magnification along the length of these ossicles. Qualitative analysis of arm spines, tentacle scales and teeth showed higher degradation in ossicles from lower pH at control temperature. A more complex pattern was displayed in ossicles exposed to medium and high temperature when comparing different pH levels, particularly under a quantitative analysis (Table 3). For the lateral arm plates, differences between treatments were only evident at high magnification of the spine articulation (Fig. 5M,R). Moreover, combined effects of temperature and pH were not clear quantitatively on lateral arm plates (Table 3). All assessed ossicles displayed an increasing gradient in degradation/malformation as treatments were more extreme in terms of temperature, pH or a combination of the two (Figs S2–S4).

At week 0, the righting time of O. fasciata was significantly influenced by increased temperature but not by decreased pH, with a significant interaction between drivers (Table 2). Righting time decreased by up to 14% with warming, while the complex response to the combined effects decreased righting time by up to 43% (Tukey's HSD, P<0.05; Fig. 6A). Time of exposure to the different treatments had a significant effect on the righting time of the mottled brittle star, with a general increasing trend as time passed (Table 2, Fig. 6B). Moreover, after 15 weeks of experimental exposure, neither warming nor acidification had a significant effect on the righting time response of O. fasciata (Table 2).

Long-term exposure of O. fasciata to high temperature and low pH conditions affected survival, respiration and growth rates, as well as calcification/dissolution and the righting response, as described below for each of these proxies. Temperature changes clearly impacted these aspects of the ecophysiology of mottled brittle star, while changes in pH had a more subtle or no effect. Our results indicate that exposure to a combination of high temperature and low pH produces complex responses for several traits.

Ophionereis fasciata experiences temperatures up to 28°C (and pH as low as 7.4) inside tide pools during the summer for short periods (Pentreath, 1968), as confirmed during a pilot study for this research (Márquez-Borrás, 2023). In its natural environment, this species is not exposed to temperatures beyond 22°C (or pH 7.7) for longer than a few hours (Atkins, 2014; Evans and Atkins, 2013; Law et al., 2018b; Vance et al., 2020). Our results indicate that the sublethal temperature threshold of this species lies below 22°C, with animals showing high mortality when exposed for several weeks to any temperature above that. This thermotolerance limit is lower than that of the southern Australia sister species Ophionereis schayeri (25°C), consistent with the higher sea surface temperature (SST) (28°C) experienced by the latter (Christensen et al., 2011, 2023). However, both species already live close to their upper thermal thresholds, as reported for other intertidal species (Balogh and Byrne, 2021; Catarino et al., 2012; Somero, 2010). Additionally, the target low pH of 7.4 was not achieved, which could partially explain the subtle effects of low pH on O. fasciata. However, given the long-term exposure to less extreme treatments (7.7 and 7.6), similar effects to those observed for temperatures experienced by this species (for short-term periods) would be expected when the time of exposure exceeds that observed in its natural habitat.

Warming was the main driver of mortality risk in O. fasciata, while no statistical differences were found in mortality as pH decreased. These results support the idea that low pH leads to sub-lethal effects, while high temperatures are more likely to result in lethal effects (Przeslawski et al., 2015). Additionally, our results confirmed that time of exposure has an important impact on the response and survival of the mottled brittle star, in line with the results for other echinoderms (Balogh and Byrne, 2021; Brothers et al., 2016; Christensen et al., 2017, 2023; Delorme and Sewell, 2016; Khalil et al., 2023; Suckling et al., 2015; Uthicke et al., 2021).

As a single driver, higher temperatures increased Ṁ_O2 of O. fasciata after 15 weeks of exposure, as expected for marine ectotherms and ophiuroids in particular (Christensen et al., 2023; Fang et al., 2015a,b; Pribadi et al., 2016; Rohr et al., 2018; Seebacher et al., 2015). Meanwhile, the Ṁ_O2 of this species was reduced as pH decreased, supporting the proposed robustness of echinoderms to OA in terms of metabolic rates (Appelhans et al., 2014; Asnicar et al., 2021; Burnham et al., 2022; Carey et al., 2014; Catarino et al., 2012; Christensen et al., 2017; Collard et al., 2014; Dorey et al., 2013; Hu et al., 2014; McElroy et al., 2012; Pan et al., 2015). The combined effect of temperature and pH at week 15 on O. fasciata is similar to that observed by Christensen et al. (2017) in two brittle star species (Hemipholis cordifera and Microphiopholis gracillima). Echinoderm metabolic rates in response to simultaneous high temperature and low pH exposure have a high level of variability between species with antagonistic (Harianto et al., 2021; Kelly et al., 2013; Rich et al., 2018; Wood et al., 2010), additive (Carey et al., 2016; Christensen et al., 2011; Uthicke et al., 2014) and synergistic responses (Catarino et al., 2012; Khalil et al., 2023; McElroy et al., 2012). Time of exposure to the experimental conditions is an additional factor in the variability mentioned above (Harianto et al., 2021), as also observed in this study.

Combined high temperature and low pH have a positive or neutral effect on the growth rates of brittle stars (Christensen et al., 2017; Wood et al., 2010, 2011). Ophionereis fasciata displayed results similar to those described for the Arctic species Ophionereis sericeum, where high temperature and low pH interacted antagonistically, resulting in a neutral combined effect (Wood et al., 2011). In the present study, the mottled brittle star displayed negative effects for four of the six growth-rate proxies when exposed to high temperature; meanwhile, none was significantly affected by pH or the interaction between the two drivers. These results suggest that the negative effects on growth rates due to high temperatures are partially ameliorated when O. fasciata is exposed to both drivers, as observed in other echinoderms (Brown et al., 2014; Byrne et al., 2013; García et al., 2015; Keppel et al., 2015). As a single driver, high temperature has been observed to increase growth rates in ophiuroid species worldwide (Christensen et al., 2017, 2023; Clark et al., 2007; Weber et al., 2013; Wood et al., 2010), in contrast to the results described for the mottled brittle star. In this sense, high temperatures increase foraging time, allowing animals to gather more food (Christensen et al., 2023), and hence the display of faster growth rates. Foraging efficiency might have been affected in O. fasciata exposed to high temperatures by the excess mucus production, as supported by the lethargic behaviour observed in these animals. In this sense, animals exposed to higher temperatures showed a more lethargic behaviour with a notably slower reaction to handling and light stimuli as well as a more flexible and soft body. Alternatively, faster growth rates might be displayed below the temperature threshold, slowing down afterwards, as observed for the vertebrae growth rates of O. fasciata.

As high-Mg calcifiers, echinoderms are considered to be particularly vulnerable to OA (Byrne and Fitzer, 2019; Dubois, 2014; Figuerola et al., 2021), and although less studied, high temperatures are also known to impact the calcification/dissolution process of some species (Duquette et al., 2018; Wolfe et al., 2013; Wood et al., 2011). However, the interaction of these drivers has a neutral effect on most species of echinoderms (Byrne et al., 2014; Christensen et al., 2017; Manríquez et al., 2017; Uthicke et al., 2014; Wood et al., 2010). Our results indicate a negative effect of increased temperature and decreased pH, which seems to be additive for the degradation of most ossicles, as described for a sea urchin species (Wolfe et al., 2013). In contrast, skeletal malformation is clearly increased by high temperature and low pH as single drivers, but no evident interaction between these was identified in O. fasciata. In all cases, decreased pH was the main driver of negative effects on calcification/dissolution, as expected, considering the correlation between pH and calcification in marine species (Byrne and Fitzer, 2019; Di Giglio et al., 2020; Dubois, 2014; Figuerola et al., 2021).

The control pH treatment in this study (7.93) was lower than the average conditions in O. fasciata's natural habitat (8.01), which might indicate that minor changes to pH impact calcification/dissolution processes. In this sense, the limited soft tissue present in most ophiuroids has been proposed as a factor making the skeleton of this group more susceptible to OA (Azcárate-García et al., 2024; Wood et al., 2010). As the pH of the world's oceans has already decreased by ∼0.1 pH units since the pre-industrial era (IPCC, 2019), this may partially explain why ossicles from animals in the control group displayed some degradation.

Righting time is adjusted over time in response to combined high temperature and low pH as a physiological strategy associated with reduced movement and activity (Sparks, 2018; Wood et al., 2010). Our results support this hypothesis, with a clear increasing trend in righting times over time as treatments were more acute. Moreover, righting time results displayed the same trends at the beginning and end of the experiment, although not significant at the end. Only four studies have assessed the righting time response of echinoderms in response to the combined effects of increased temperature and decreased pH to date (Manríquez et al., 2017; Sparks, 2018; Uthicke et al., 2021; Wood et al., 2010). The faster righting time at higher temperatures in O. fasciata is similar to that observed in a New Zealand sea star, Antarctic brittle and sea stars, as well as in Australian sea cucumbers (Buccheri et al., 2019; Sparks, 2018; Wood et al., 2011), but contrasts with that reported for New Zealand sea urchins and temperate or polar sea stars (Arribas et al., 2022; Delorme et al., 2020; Kidawa et al., 2010; Peck et al., 2008). Righting time appears to be a species-specific indicator of physiological state and should be analysed considering species lifestyle and habitat. In this sense, it has been proposed that positive effects on the righting time could be displayed by species naturally experiencing a larger range of temperatures, such as O. fasciata (Brothers and McClintock, 2015; Buccheri et al., 2019).

Regional climate models project in the near future (2100) mean summer SSTs around 22°C, with the most extreme temperatures rising to a maximum of 26°C and pH dropping by ∼0.36 (Foley and Carbines, 2019; Law et al., 2018b). These trends are close to the thresholds currently experienced by O. fasciata for short periods due to daily and seasonal fluctuations in its intertidal and shallow subtidal habitat (Pentreath, 1968; Vance et al., 2020). The broad depth range of this species might allow it to avoid higher temperatures through vertical migration, but this might be limited by the seasonal and daily variations of temperature/pH, and by its thermal threshold and migration capabilities. Further investigation should compare the performance of different New Zealand populations and investigate the impact of a changing ocean on other aspects of the life of O. fasciata, including its reproduction and development.

Echinoderms have been recognised as a model for studying OW–OA effects on marine biota, particularly sea urchins (Table S1). Two-thirds of these studies have focused on echinoids as a result of a combination of logistical factors. These include their presence in most marine environments, knowledge of their basic biology and physiology (Bednaršek et al., 2021; Kroeker et al., 2013), the ease of obtaining adults and gametes of this class (partially explaining the number of studies assessing the pre-metamorphic response to OW–OA), and their vulnerability to a changing ocean as marine calcifiers (Hughes et al., 2011; Melzner et al., 2009; Pörtner et al., 2004). Furthermore, particular species have been used as models for OW–OA studies in North America (Strongylocentrotus purpuratus, 25 studies), Europe (Paracentrotus lividus, 29), Australia (Heliocidaris erythrogramma, 18) and Antarctica (Sterechinus neumayeri, 18) (Márquez-Borrás, 2023) (Fig. S1B, Table S1). Echinoderms from other classes have not been studied at all (crinoids), or as much as echinoids, particularly Ophiuroids (Table S1).

Here, we propose that brittle stars represent an excellent alternative model system for testing the effects of a changing ocean on marine taxa for several reasons. Firstly, brittle stars are globally distributed in different habitats and represent the largest class of echinoderms in terms of species richness (O'Hara et al., 2018; Stöhr et al., 2012), making them ideal for comparative purposes. Many species, such as O. fasciata, also have wide geographical ranges, which allows within-species comparisons of population tolerance over different levels of variability in oceanic conditions (Christensen et al., 2011; Silva Romero et al., 2021). Secondly, like other echinoderms, brittle stars can regenerate lost appendages/arms, allowing researchers to study the formation of calcified material under controlled conditions (Byrne and Fitzer, 2019; Wilkie, 2001). Measuring growth rates and calcification in regenerating tissues in ophiuroids is simplified given their calcified and differentiated arms, and their faster and more efficient regeneration (Ben Khadra et al., 2018). Thirdly, species such O. fasciata are relatively easy to manipulate in the laboratory and field and are usually smaller than sea urchins, decreasing logistical difficulties in mesocosm experiments (Carey et al., 2014; Christensen et al., 2011, 2023; Hu et al., 2014; Wood et al., 2011). Fourthly, the biology of many genera has been studied across the globe. For example, studies of the genus Ophionereis include their diet (Pentreath, 1970; Yokoyama and Amaral, 2008), reproduction (Byrne, 1991; Selvakumaraswamy and Byrne, 2004; Yokoyama et al., 2008) and arm regeneration (Pentreath, 1971; Sides, 1987; Yokoyama and Amaral, 2010), among others. Finally, our experimental design confirmed that this species, like many other species of brittle stars, can be maintained in groups of several individuals. Such behaviour allowed us to increase the sample size without a significant impact on the replicate system of the mesocosm.

We thank Peter Schlegel and Esther Stuck for their help with animal collection; and Catherine Hobbis for technical assistance with the SEM.