Paper Review

Glob Change Biol. 2020;00:1–14. wileyonlinelibrary.com/journal/gcb  |  1© 2020 John Wiley & Sons Ltd

Received: 28 July 2020  |  Accepted: 13 November 2020 DOI: 10.1111/gcb.15455

P R I M A R Y R E S E A R C H A R T I C L E

Ocean acidification locks algal communities in a species-poor early successional stage

Ben P. Harvey1  | Koetsu Kon1  | Sylvain Agostini1  | Shigeki Wada1  | Jason M. Hall-Spencer1,2

1Shimoda Marine Research Center, University of Tsukuba, Shizuoka, Japan 2Marine Biology and Ecology Research Centre, University of Plymouth, Plymouth, UK

Correspondence Ben P. Harvey, Shimoda Marine Research Center, University of Tsukuba, 5-10-1 Shimoda, Shizuoka 415-0025, Japan. Email: [email protected]

Funding information Japan Society for the Promotion of Science, Grant/Award Number: 17K17622; Ministry of Environment, Government of Japan, Grant/Award Number: 4RF-1701; University of Tsukuba

Abstract Long-term exposure to CO2-enriched waters can considerably alter marine biological community development, often resulting in simplified systems dominated by turf algae that possess reduced biodiversity and low ecological complexity. Current un- derstanding of the underlying processes by which ocean acidification alters biologi- cal community development and stability remains limited, making the management of such shifts problematic. Here, we deployed recruitment tiles in reference (pHT 8.137 ± 0.056 SD) and CO2-enriched conditions (pHT 7.788 ± 0.105 SD) at a volcanic CO2 seep in Japan to assess the underlying processes and patterns of algal commu- nity development. We assessed (i) algal community succession in two different sea- sons (Cooler months: January–July, and warmer months: July–January), (ii) the effects of initial community composition on subsequent community succession (by recipro- cally transplanting preestablished communities for a further 6 months), and (iii) the community production of resulting communities, to assess how their functioning was altered (following 12 months recruitment). Settlement tiles became dominated by turf algae under CO2-enrichment and had lower biomass, diversity and complexity, a pattern consistent across seasons. This locked the community in a species-poor early successional stage. In terms of community functioning, the elevated pCO2 commu- nity had greater net community production, but this did not result in increased algal community cover, biomass, biodiversity or structural complexity. Taken together, this shows that both new and established communities become simplified by rising CO2 levels. Our transplant of preestablished communities from enriched CO2 to refer- ence conditions demonstrated their high resilience, since they became indistinguish- able from communities maintained entirely in reference conditions. This shows that meaningful reductions in pCO2 can enable the recovery of algal communities. By understanding the ecological processes responsible for driving shifts in community composition, we can better assess how communities are likely to be altered by ocean acidification.

K E Y W O R D S CO2 seeps, community dynamics, competition, ecosystem function, global change ecology, inhibition, turf algae

2  |    HARVEY Et Al.

1   |   I N T R O D U C T I O N

The oceanic uptake of anthropogenic carbon dioxide emissions is a global environmental issue termed ocean acidification. The ef- fects of ocean acidification are detrimental to a wide range of ma- rine organisms (Harvey et al., 2013; Kroeker, Kordas, et al., 2013), and this affects ecosystem functioning and the goods and services that people derive from marine resources (Gattuso et al., 2015; Hall-Spencer & Harvey, 2019). To better understand the effects of ocean acidification, there has been an effort in recent years to move beyond aquarium-based experiments on single species to- wards in-situ experiments (e.g. Albright et al., 2016, 2018; Brown et al., 2016), long-term mesocosm observations (e.g. Algueró-Muñiz et al., 2017; Moulin et al., 2015), and studies using natural CO2 seeps (e.g. Agostini et al., 2018; Fabricius et al., 2011; Hall-Spencer et al., 2008; Milazzo et al., 2014). These approaches have shown that long- term exposure to ocean acidification conditions projected for the end of the century fundamentally alters the composition of marine biological communities, usually resulting in simplified systems with reduced biodiversity and less ecological complexity (Agostini et al., 2018; Sunday et al., 2017; Vizzini et al., 2017). Many of these studies have been observation-based, and so an understanding of the un- derlying processes responsible for driving these patterns in commu- nity development remains limited. To help assess the future effects of ocean acidification, it would be useful to better understand how community development processes are affected by rising levels of seawater CO2 (Gaylord et al., 2015), and how such changes will influ- ence their associated ecosystem functioning.

Ecological theory suggests that the successional trajectories of ‘disturbed’ marine subtidal communities will be primarily driven by physical stresses, competition for resources through the mecha- nisms of ‘facilitation’ and ‘inhibition’ (Connell & Slatyer, 1977) and the strength of associated bottom-up and top-down interactions (Gruner et al., 2008; Jenkins et al., 1999). One of the difficulties in predicting how community development will be affected by ocean acidification is that the changes in carbonate chemistry can simul- taneously act as both resource and stressor (Connell et al., 2013, 2018; Milazzo et al., 2019). It provides a bottom-up resource to primary producers by enhancing the availability of bicarbonate and CO2 (Connell et al., 2013; Koch et al., 2013), but also acts as a phys- ical stressor to many organisms (including calcified primary produc- ers) via negative effects on their physiology (Harvey et al., 2013; Kroeker, Kordas, et al., 2013). Subsequently, marine communities are expected to be re-organized by the effects of ocean acidification. Ocean acidification alters the initial successional trajectories of algal communities, which lead to dominance by fleshy algae over calcified algae in acidified conditions projected for the end of the century in both temperate and tropical settings (Crook et al., 2016; Kroeker et al., 2012). Enriched CO2 alters competitive interactions, acting as a physical stressor to calcified macroalgae whereas turf algae can use the additional carbon to boost growth, which allows turf algae to attain dominance (also see Connell et al., 2018). Fast-growing oppor- tunistic (r-selected) turf algal species are usually suppressed beneath

macroalgal canopies on temperate reefs (Johnson & Mann, 1988) and by top-down control of grazers in coral reefs (Hughes et al., 2007). In the absence of strong competition or compensatory pro- cesses (e.g. Connell et al., 2018; Ghedini & Connell, 2016; Ghedini et al., 2015), turf species can become dominant thereby changing the ecosystem state.

Under present-day conditions, it has been suggested that despite bottom-up control of primary production being pervasive, top-down control by consumers has a stronger influence on the trajectories of algal community succession (Gruner et al., 2008; Hillebrand et al., 2007). For example, intense grazing by sea urchins and herbivo- rous fish can prevent kelp forest growth resulting in ‘urchin barrens’ dominated by crustose coralline algae (Kelly et al., 2016; Ling et al., 2015). Ocean acidification is expected to reduce bottom-up control on those species which are carbon-limited, as long as sufficient nu- trients are available (Celis-Plá et al., 2015; Gordillo et al., 2003; Li et al., 2012). Top-down control by benthic invertebrates in acidified conditions may also diminish, given that at CO2 seeps the abun- dance and size of many marine fauna are reduced (Garilli et al., 2015; Harvey et al., 2016, 2018), with such examples as the observed num- ber of sea urchin feeding halos being reduced in a CO2 seep (Kroeker et al., 2013). Fish communities include a greater proportion of her- bivorous fish within acidified conditions (during the period of peak macroalgae biomass; Cattano et al., 2020), and so it may be possible in some systems for fish to maintain top-down control (Baggini et al., 2015). Taken together, any strong reductions in bottom-up and/or top-down control are likely to alter community successional tra- jectories and allow r-selected opportunistic species to outcompete other species and dominate under ocean acidification.

Seasonality is an important aspect of shallow-water ecosystems, and yet the consequences of seasonally induced environmental fluc- tuations have rarely been considered in ocean acidification studies (Baggini et al., 2014; Godbold & Solan, 2013). Algal communities in temperate and warm temperate ecosystems experience large seasonal changes in environmental conditions (Figure 1), which result in con- siderable temporal shifts with a period of high recruitment and peak biomass typically occurring in late spring. Thus the responses of algal communities to ocean acidification will likely be strongly influenced by seasonality (Baggini et al., 2014). In the Northern Pacific Ocean, this is further complicated by the occurrence of typhoons, which typically occur between July and October in Japan. Typhoons act as a substan- tial physical disturbance that affects benthic community structure and habitat complexity, such as through the removal of corals (Done, 1992), macroalgae (Cattano et al., 2020) and seagrass cover (Wilson et al., 2020), and can indirectly change the community function of associated species (e.g. fish; Cattano et al., 2020).

Observations at natural CO2 seeps worldwide provide a good understanding of how long-term ocean acidification simplifies the composition of climax communities (Foo et al., 2018; González- Delgado & Hernández, 2018; Hall-Spencer & Harvey, 2019), yet it remains unclear whether these simplified communities develop due to altered successional trajectory, stunted community development (via successional inhibition) or are driven by reduced bottom-up and/

    |  3HARVEY Et Al.

or top-down control. To address these gaps, we deployed recruit- ment tiles in reference (~350 μatm pCO2) and acidified (~900 μatm pCO2) conditions using a natural CO2 seep area as an analogue for end of the century pCO2 conditions (the representative concentra- tion pathway (RCP) 8.5 scenario, 851 to 1370 μatm; IPCC, 2013) to assess the early to mid-successional trajectories of algal communities in two different seasons (cooler months January to July, and warmer months July to January). The study was carried out over these two time periods to investigate whether the effects of ocean acidification on community development are temporally consistent. Following this we carried out a reciprocal transplant of some of those established communities, in order to assess the effects of initial community com- position on subsequent community succession in the reference and acidified conditions. Finally, we assessed the community production of these reciprocally transplanted communities (including the asso- ciated sessile invertebrate communities which contribute in terms of respiration), in order to determine how any changes in community composition will alter their ecosystem functioning.

2   |   M AT E R I A L S A N D M E T H O D S

2.1  |  Experimental design

To investigate our core question of how ocean acidification in- fluences early community succession of algal communities, ex- periments using recruitment tiles were carried out using an acidified area of the Shikine Island CO2 seep, Japan (34°19′9ʺN, 139°12′18ʺE), and a nearby reference pCO2 area in an adjacent bay

(~600 m away by the shortest route). Both the reference and acidi- fied locations (hereafter ‘350 μatm’ and ‘900 μatm’, respectively) have had their carbonate chemistry and biology well character- ized previously (Agostini et al., 2015, 2018; Cattano et al., 2020; Harvey et al., 2018, 2019; Kerfahi et al., 2020; Witkowski et al., 2019), and we present 2 months of additional original pHT (Figure S1) and temperature data collected at the ‘900 μatm’ location with a Durafet sensor (SeaFET, Sea-Bird Scientific) using the same ap- proach as Agostini et al. (2018). Salinity was measured concurrently using Hobo conductivity loggers (U24-002-C), and discrete sam- ples for total alkalinity were collected throughout the study period, with total alkalinity measured using an auto-titrator (916 Ti-Touch, Metrohm). In summary, the ‘350 μatm’ location had a mean pHT of 8.137 ± 0.056 (SD) and the ‘900 μatm’ location had a mean pHT of 7.781 ± 0.105 (SD), and the mean carbonate chemistry of the two locations is presented in Table 1. Long-term temperature data were recorded over a 1-year period by deploying a temperature logger (HOBO Pendant Temperature/Light 64K Data Logger) at ~6 m depth in each site. The ‘900 μatm’ elevated pCO2 location repre- sents an end of the century projection for reductions in pH (the RCP 8.5 scenario; IPCC, 2013), and was not confounded by differences in temperature, salinity, dissolved oxygen, total alkalinity, nutrients or depth relative to reference sites used for comparison (Agostini et al., 2015, 2018; Harvey et al., 2019). Our basalt recruitment tiles were 130 × 130 × 15 mm and were secured using individual anchor bolts (8.5 mm width, 70 mm length) drilled into rock by SCUBA divers at ~6 m depth (Nemo Underwater Drill). The tiles at each location were deployed haphazardly across a c. 400 m2 area (with at least 5 m be- tween individual tiles), fixed to upward-facing substrata.

F I G U R E 1  Conceptual representation of the recruitment tile treatments. (a) Tiles were deployed for 6 months during the ‘Cold Period’ or ‘Warm Period’ in either reference pCO2 (350 μatm; blue line) or acidified conditions (900 μatm; red line). (b) Tiles from the ‘Warm Period’ were then used as part of a reciprocal transplant either being transplanted into 350 or 900 μatm conditions for a further 6 months

4  |    HARVEY Et Al.

2.1.1  |  Seasonal experiment

For the first experiment, five recruitment tiles were individually de- ployed in each location (350 and 900 μatm) during the cooler months of January 2017–July 2017 (hereafter termed ‘Cold Period’), with eight recruitment tiles deployed in each location (350 and 900 μatm) during the warmer months of July 2017–January 2018 (hereafter termed ‘Warm Period’). Mean seawater temperature (±SD) during the ‘Cold Period’ was 18.14 ± 1.81°C at 350 μatm and 18.07 ± 1.63°C at 900 μatm, and during the ‘Warm Period’ was 22.86 ± 2.97°C at 350 μatm and 22.67 ± 2.83°C at 900 μatm. See Figure 1a for a con- ceptual overview of the experimental design.

2.1.2  |  Reciprocal experiment

For the second experiment, tiles from the ‘Warm Period’ of the seasonal experiment were used (each tile had an algal community following 6 months recruitment). Sixteen tiles from the seasonal ex- periment were reciprocally transplanted into the 350 and 900 μatm locations for a further 6 months to assess the effects of initial com- munity composition on subsequent community succession in refer- ence and acidified conditions (four tiles in each combination, see Figure 1b).

2.2  |  Community analysis

For both the seasonal experiment and the reciprocal transplanta- tion experiment, following the 6 month experimental period each tile was brought into the laboratory and photographed (Nikon D7200, Nikon). For each 12 month tile, two photos were taken to image the upperstorey community, and (after removal of the up- perstorey community by hand) the understorey community. These data were then combined for analysis. Community composition was assessed using ImageJ (Abràmoff et al., 2004) by overlaying 64 points on a grid, and recording the abundance of each algal functional group underlying that point. Functional groups used followed Steneck and Dethier (1994), with algal grouping being

based on their morphology, thallus size and complexity (micro- algae, filamentous algae, foliose algae, corticated foliose algae, corticated macrophytes, leathery macrophytes, articulated cal- careous algae and crustose algae). The habitat complexity of each tile was determined by combining the abundance of each algal functional group with a rank (0–5) based on the biogenic habitat complexity provided by that functional group. The ranking was based on categories assigned by Steneck and Dethier (1994): mi- croalgae = 1, filamentous algae = 2, foliose algae = 3, corticated foliose algae = 3.5, corticated macrophytes = 4, leathery macro- phytes = 5, articulated calcareous algae = 4 and crustose algae = 3. The habitat complexity score was then normalized to between 0 and 1.

2.3  |  Community production

Community production and respiration of individual tiles were as- sessed by measuring, with an Orion 4-Star pH and dissolved oxy- gen meter (Thermo Scientific), the changes in dissolved oxygen concentrations during an incubation within a 2.5 L seawater con- tainer (15 cm wide × 20 cm length × 10 cm height) in a tempera- ture-controlled water bath. Magnetic stirrers (M-1 Controller and MS101A Stirrer, AS-One) were used to continuously mix the sea- water within each container throughout measurements. Seawater for each treatment (pHNBS 8.05 ± 0.01 SD vs. pHNBS 7.83 ± 0.01 SD) was collected from the same location off-shore (oceanic pHNBS 8.05) and treatments of pHNBS 7.83 ± 0.01 SD were acquired by bubbling with pure CO2 (Fukurow pH Controller; Aqua Geek). Community production and respiration were measured over a 150 min period; first determining oxygen production (60 min light period, c. 200 μmol m−2 s−1), and after a 30 min dark period, oxygen consumption (60 min dark period). Net community production and community respiration were measured during the light and dark periods, respectively, with gross primary production calculated as the net community production minus community respiration. All three measurements are presented as mmol O2 h

−1 m−2. To assess whether any changes in respiration rates were driven by an altered sessile invertebrate community for the 12 month communities, the

T A B L E 1  Summary of the carbonate chemistry for the 350 and 900 μatm locations. The pHT (350 μatm, n = 1964; 900 μatm, n = 10,818), salinity (350 μatm, n = 1964; 900 μatm, n = 10,818) and total alkalinity (AT; 350 μatm, n = 56; 900 μatm, n = 47) are measured values. All other values were calculated using the carbonate chemistry system analysis program CO2SYS: Seawater pCO2, dissolved inorganic carbon (DIC), bicarbonate (HCO3

−), carbonate (CO3 2−), carbon dioxide (CO2), saturation states for calcite (Ωcalcite) and aragonite (Ωaragonite).

Values are presented as mean, with standard deviation below

Location pHT Salinity (psu)

AT (μmol kg−1)

pCO2 (μatm)

DIC (μmol kg−1)

HCO3 −

(μmol kg−1) CO3

2− (μmol kg−1) Ωcalcite Ωaragonite

‘350 μatm’ 8.137 34.504 2264.29 316.057 1962.694 1740.629 211.979 5.087 3.301

0.056 0.427 15.34 47.466 34.376 55.084 22.221 0.534 0.348

‘900 μatm’ 7.788 34.351 2268.33 841.148 2125.785 1984.889 115.150 2.771 1.805

0.106 0.484 19.45 291.762 39.381 52.510 21.308 0.512 0.336

Note: ‘350 μatm’ carbonate chemistry data are sourced from Agostini et al. (2018), ‘900 μatm’ carbonate chemistry data are averaged across Agostini et al. (2018) and original data collected between 2017/05/31 and 2017/08/08.

    |  5HARVEY Et Al.

community composition and total cover of the sessile fauna (lo- cated on the underside of the tile) was assessed, following the same procedure as the community analysis (based on the percent cover of the sessile faunal species).

2.4  |  Statistical analysis

Statistical analyses were conducted using R (version 3.6.0; R Core Team, 2019), with the ‘vegan’ (Oksanen et al., 2019), ‘mvabund’ (Wang et al., 2012) and base ‘stats’ package used for statistical anal- ysis, and the ‘ggplot2’ (Wickham, 2016) and ‘ggpubr’ (Kassambara, 2019) packages used for figure production. For each of the analyses performed, the package and specific function used in R are listed below as ‘package::function’.

For the seasonal experiment, differences in community compo- sition (based on percentage cover) between pCO2 (two levels: 350 and 900 μatm) and Season (two levels: ‘Cold Period’ and ‘Warm Period’) were assessed using non-metric multidimensional scaling (nMDS; vegan::metaMDS) and a permutational analysis of variance (PERMANOVA) based on Bray–Curtis dissimilarity (vegan::vegdist and vegan::adonis). To test for differences in the percentage cover of the specific algal functional groups, we employed a two-way generalized linear model (GLM; Family: Binomial (Link = Logit)), with pCO2 (two levels: 350 and 900 μatm) and Season (two levels: ‘Cold Period’ and ‘Warm Period’) as fixed factors (stats::glm).

For the reciprocal transplant experiment, differences in commu- nity composition (based on percentage cover) between Community Origin (two levels: 350 and 900 μatm), Community Destination (two levels: 350 and 900 μatm) and Time (two levels: 6 and 12 months) as fixed factors were assessed using an nMDS (vegan::metaMDS) and a PERMANOVA based on Bray–Curtis dissimilarity (vegan::vegdist and vegan::adonis). As the starting communities at 6 months could influence the resulting communities at 12 months on a particular tile, we accounted for this repeated measure in the PERMANOVA by using the ‘strata’ argument within vegan::adonis. To test for differ- ences in the percentage cover of the specific algal functional groups at 12 months, we employed a two-way GLM (Family: Binomial (Link = Logit)), with Community Origin (two levels: 350 and 900 μatm) and Community Destination (two levels: 350 and 900 μatm) as fixed factors (stats::glm).

Differences in community production and respiration of the 12 month communities were tested using a two-way GLM (Family: Gaussian (Link = Identity)), with Community Origin (two levels: 350 and 900 μatm) and Community Destination (two levels: 350 and 900 μatm) as fixed factors (stats::glm). To assess whether changes in respiration were driven by an altered sessile invertebrate commu- nity for the 12 month communities, we assessed for differences in the community composition (based on percentage cover) and total percentage cover of the sessile fauna. Differences in the community composition of the sessile fauna, with Community Origin (two lev- els: 350 and 900 μatm), and Community Destination (two levels: 350 and 900 μatm) as fixed factors, were assessed using a PERMANOVA

based on Bray–Curtis dissimilarity (vegan::vegdist and vegan::adonis). Differences in the total percentage cover of the sessile fauna were tested using a two-way GLM (Family: Gaussian (Link = Identity)), with Community Origin (two levels: 350 and 900 μatm) and Community Destination (two levels: 350 and 900 μatm) as fixed fac- tors (stats::glm).

Post hoc comparisons for all PERMANOVAs were achieved using a Bonferroni-corrected pairwise PERMANOVA. For the re- ciprocal transplant, only comparisons chosen a priori were tested; these were between Time (two levels: 6 and 12 months) for each of the four combinations of Community Origin (two levels: 350 and 900 μatm) and Community Destination (two levels: 350 and 900 μatm). The assumptions of the generalized linear models (GLMs) were met, with the response variable independent, the mean–variance relationship suitable (assessed by plotting the residuals vs. fits; mvabund::manyglm) and dispersion parameters not being under- or over-dispersed (assessed using quasibinomial error distributions in R; stats::glm). Post hoc testing of the GLM for the gross primary production was achieved using TukeyHSD (multcomp::glht).

3   |   R E S U LT S

3.1  |  Effects of pCO2 and season on community composition

Overall community composition following 6 months settlement was highly separated by nMDS, being significantly affected by both ‘pCO2’ and ‘Season’ (PERMANOVA: pCO2*Season, F1,25 = 4.351, p = 0.012; Figure 2a; Table 2). Community composition was simi- lar between seasons at 350 μatm (post hoc Bonferroni-adjusted, p = 0.102), but greatly differed between seasons at 900 μatm (post hoc Bonferroni-adjusted, p = 0.018). The typical newly settled com- munity at 350 μatm during the ‘Warm Period’ largely comprised of a low coverage (<10%) of microalgae and turf algae (Figure 2b) as well as corticated foliose algae, and a high coverage (~75%) of both corticated macrophytes (Figure 2c, including Chondracanthus tenellus (Harvey) Hommersand, 1993) and crustose coralline algae (Figure 2d, including Lithophyllum sp. Philippi, 1837). The community at 350 μatm during the ‘Cold Period’ was similar, the only difference being a significantly lower cover (approximately half) of corticated macrophytes relative to the ‘Warm Period’ (Figure 2c).

Relative to 350 CO2, at 900 μatm the newly settled community during the ‘Warm Period’ showed increased coverage by microalgae and turf (from almost zero to ~30% coverage; Figure 2b), as well as increased corticated foliose algae (tenfold increase; including the species Zonaria diesingiana J. Agardh, 1841), but a threefold de- crease in corticated macrophytes (Figure 2c) and a fourfold decrease in crustose coralline algae (Figure 2d). This resulted in a community with roughly similar coverage of the microalgae and turf, corticated macrophytes and crustose coralline algae. During the ‘Cold Period’, the 900 μatm communities showed a high coverage of microalgae

6  |    HARVEY Et Al.

and turf algae (~72%; Figure 2b), but an absence of corticated foliose algae. Both the corticated macrophytes (Figure 2c) and crustose cor- alline algae (Figure 2d) were similar in coverage between the ‘Warm Period’ and ‘Cold Period’ at 900 μatm CO2.

3.2  |  Effects of early-stage composition and pCO2 on subsequent community succession

After performing a reciprocal transplant of the established 6 month communities for a further 6 months, it was found that

regardless of the conditions under which the communities were established (i.e. ‘Community Origin’), the communities converged to form similar communities based on the pCO2 conditions they were currently residing in (i.e. ‘Community Destination’; Figure 3; Table 3).

At 12 months the 350 μatm CO2 recruitment tiles typically had a corticated macrophyte upperstorey (Figure 4b, including Codium coactum Okamura, 1930 and Chondracanthus tenellus), with an un- derstorey of microalgae and turf algae (Figure 4a), as well as crustose coralline algae (Figure 4c, including Lithophyllum sp.). Both the ‘350 to 350 μatm’ and the ‘900 to 350 μatm’ communities had a higher total cover than at 6 months (significant increase by ~40% cover for both). This was due to an increased cover of understorey microalgae and turf algae in the ‘350 to 350 μatm’ community (Figure 4a), and increased corticated macrophytes (Figure 4b; both as understorey and canopy) for both communities. Overall the ‘350 to 350 μatm’ community was most complex, followed by the ‘900 to 350 μatm’ community (Figure 4d).

At 900 μatm CO2, a typical community after 12 months lacked canopy algae and was predominantly comprised of turf algae (Figure 4a, including Biddulphia biddulphiana (J.E. Smith) Boyer 1900) with some corticated foliose algae (including Zonaria diesing- iana), and minimal cover of corticated macrophytes (Figure 4b) and

F I G U R E 2  Community composition and percentage cover of algal functional groups. (a) nMDS of community composition based on algal functional groups. Treatments are displayed by pCO2 (‘350 μatm’ – blue; ‘900 μatm’ - red) and season (‘Warm Period’ – triangles; ‘Cold Period’ – crosses). (b–d) Percentage cover (%) of (b) microalgae and turf algae, (c) corticated macrophytes, and (d) crustose coralline algae following 6 months settlement at either 350 μatm pCO2 (‘Warm Period’ – darker blue, ‘Cold Period’ – lighter blue) or 900 μatm pCO2 (‘Warm Period’ – darker red, ‘Cold Period’ – lighter red). Two-way GLM (pCO2*Season) results are presented in the top-right of (b)–(d). See Table S1 for more detailed statistics

(a)

(c)

T A B L E 2  PERMANOVA summary of pCO2 (350 μatm vs. 900 μatm) and Season (‘Cold Period’ vs. ‘Warm Period’) for the algal communities. For p-values, * (p < 0.05), ** (p < 0.01), *** (p < 0.001)

Term df Sum Sq.

Mean Sq. F p

pCO2 1 1.995 1.995 32.58 0.001***

Season 1 0.378 0.378 6.18 0.003**

pCO2*Season 1 0.266 0.266 4.35 0.012*

Residuals 22 1.347 0.061

Total 25 3.986

    |  7HARVEY Et Al.

crustose coralline algae (Figure 4c). The community that was trans- planted into 900 μatm (‘350 to 900 μatm’) showed a decline in total cover (significant decrease by ~45% cover), and a large reduction in complexity (Figure 4d) due to the loss of corticated macrophytes and crustose coralline algae (Figure 4b,c). The community that was

consistently maintained at 900 μatm (‘900 to 900 μatm’) showed an increased coverage by turf and microalgae (Figure 4a), but overall displayed similar levels of total cover (~100% cover at 6 months and ~120% cover at 12 months) and complexity (Figure 4d) relative to the initial 6 month community.

F I G U R E 3  nMDS of community composition based on algal functional groups. Communities are grouped as those exposed to 350 μatm throughout (darker blue), transplanted from 900 to 350 μatm (lighter blue), transplanted from 350 to 900 μatm (lighter red) and exposed to 900 μatm throughout (darker red). The initial starting 6 month communities are displayed with open symbols and dashed lines, and the 12 month communities following the transplant are displayed with solid symbols and lines

Term df Sum Sq. Mean Sq. F p

Origin 1 0.910 0.910 13.05 0.001***

Destination 1 0.669 0.669 9.59 0.001***

Time 1 1.267 1.267 18.17 0.001***

Origin*Destination 1 0.141 0.141 2.03 0.002**

Origin*Time 1 0.490 0.490 7.03 0.023*

Destination*Time 1 1.16 1.16 16.66 0.003**

Origin*Destination*Time 1 0.074 0.074 1.06 0.541

Residuals 23 1.604 0.070

Total 30 6.136

T A B L E 3  PERMANOVA summary of Community Origin (‘Origin’: 350 vs. 900 μatm), Community Destination (‘Destination’: 350 μatm vs. 900 μatm), and Time (‘6 months’ vs.’12 months’) for algal communities grown on settlement tiles off Shikine Island, Japan. For p-values, * (p < 0.05), ** (p < 0.01), *** (p < 0.001)

F I G U R E 4  Percentage cover (%) of functional groups following a further 6 month settlement: (a) microalgae and turf algae, (b) corticated macrophytes, (c) and crustose coralline algae. (d) structural complexity of the communities. Communities are grouped as those exposed to 350 μatm throughout (darker blue), transplanted from 900 to 350 μatm (lighter blue) or 350 to 900 μatm (lighter red) and exposed to 900 μatm throughout (darker red). Two-way GLM results are presented in the top-right of each. See Table S2 for more detailed statistics

(a) (b)

(c) (d)

p = 0.59 p < 0.01

8  |    HARVEY Et Al.

3.3  |  Effects of early-stage composition and pCO2 on community production

Net community production was reduced for the 350 μatm com- munities compared to the 900 μatm communities (Figure 5a). This was due to greatly increased community respiration rates for those communities at 350 μatm (Figure 5b), which resulted in the ‘350–350 μatm’ communities showing no net community production, and the ‘900–350 μatm’ communities having a nega- tive net community production (Figure 5a). Overall, the resulting communities had similar levels of gross community production (Figure 5c). A significant interaction was observed for gross com- munity production (Figure 5c); however, all post hoc comparisons were non-significant.

3.4  |  Effects of early-stage composition and pCO2 on sessile invertebrate community

To assess whether increased respiration was driven by an altered sessile invertebrate community for the 12 month communities, the community composition and total cover of the sessile fauna was assessed. The sessile invertebrate community composition of the 12 month communities was altered by both ‘Community Origin’ and ‘Community Destination’ (PERMANOVA: F1,14 = 15.07, p < 0.001 and F1,14 = 3.68, p = 0.042, respectively; Figure S2), with a greater cov- erage of ascidians (Didemnidae) and hydrozoans (Leptomedusae) in the elevated pCO2 conditions (‘350–900 μatm’ and ‘900–900 μatm’) and a greater coverage of polychaetes (Serpulidae) in the reference pCO2 conditions (‘350–350 μatm’ and ‘900–350 μatm’; see Figure S3). However, the overall coverage of sessile fauna did not significantly differ between the treatments (GLM: ‘Community Origin’ t = 1.17, p = 0.27 and ‘Community Destination’ t = −0.18, p = 0.86), with no interaction (GLM: ‘Community Origin’*‘Community Destination’ t = −0.07, p = 0.95).

4   |   D I S C U S S I O N

Ocean acidification alters the competitive abilities of algae and so is expected to change the structure, composition and functioning of both coastal and open ocean marine habitats (Cornwall et al., 2017; Hall-Spencer & Harvey, 2019). Observations at volcanic CO2 seeps in the photic zone have shown profound ecosystem shifts towards sim- plified non-calcareous communities that are often algal dominated with lower biodiversity and reduced ecological complexity (Agostini et al., 2018; Connell et al., 2018; Foo et al., 2018; González-Delgado & Hernández, 2018; Kroeker, Gambi, et al., 2013). The underlying processes by which ocean acidification affects the structure of shallow-water marine communities are not clearly established and require additional investigation, although significant progress has been made in recent years (see Kroeker, Gambi, et al., 2013; Kroeker et al., 2011, 2012; Porzio et al., 2013; Teixidó et al., 2018; Vizzini et al., 2017). We found that an enriched CO2 environment had a positive effect on r-selected, fast-growing microalgae and turf algal species in the early stages of community development. This species- poor, low complexity, early-successional stage was then locked-in as it inhibited the settlement and growth of corticated macrophytes. We highlight the potential ecological processes responsible for this change in a temperate rocky reef community exposed to enriched CO2 conditions.

Ocean acidification altered competitive dominance after 6 months with substrata in the acidified conditions being dominated by microalgae and turf algae, rather than the corticated macrophytes and crustose coralline algae that dominated in reference pCO2 condi- tions. This pattern was temporally consistent, showing that regardless of the growing season ocean acidification truncates the normal suc- cessional trajectories of communities (Baggini et al., 2014; Kroeker et al., 2012). The benefits of seawater acidification to opportunistic

F I G U R E 5  (a) Net community production (mmol O2 h −1 m−2), (b)

Community respiration (mmol O2 h −1 m−2), and (c) Gross community

production (mmol O2 h −1 m−2). Communities are grouped as those

exposed to 350 μatm throughout (darker blue), transplanted between locations (900 to 350 μatm, lighter blue; and 350 to 900 μatm, lighter red), and exposed to 900 μatm throughout (darker red). See Table S3 for more detailed statistics

(a)

(b)

(c)

p p p

p p p

p p p

    |  9HARVEY Et Al.

species, such as turf algae, over others (including calcareous species) is well established (Connell et al., 2013, 2018; Cornwall et al., 2017). Lowered carbonate saturation is a stressor to calcified macroalgae (Brinkman & Smith, 2015; Enochs et al., 2015; Fabricius et al., 2011; Kamenos et al., 2016) whilst turf algae and other fast-growing oppor- tunistic species can use the additional carbon from ocean acidifica- tion to grow and compete for resources (Harvey et al., 2019; Kroeker et al., 2012; Porzio et al., 2013), thereby attaining dominance (Connell et al., 2018). The shift from typical coastal habitat-forming species (such as corals and kelp forests) to turf algal dominance causes a loss of structural complexity and associated ecosystem services (O’Brien & Scheibling, 2018; Rogers et al., 2014).

Before turf algae overgrew the recruitment tiles in acidified con- ditions (between 6 and 12 months, see Figure 3b), both crustose coralline algae and corticated macrophytes recruited onto the sub- strata. This suggests that the divergence in community composition was not due to limited recruitment or a physiological intolerance to the acidified conditions, but was driven by altered competitive in- teractions (Crook et al., 2016; Kroeker et al., 2012) and/or loss of compensatory processes (Connell et al., 2018; Ghedini et al., 2015). Although bottom-up control helps stimulate algal growth on coral and rocky reefs, grazing pressure determines whether turf algae dominate (Mumby et al., 2006). The grazing pressure of large benthic invertebrates in acidified sites is thought to be lowered due to phys- iological impacts (Calosi et al., 2013; O’Donnell et al., 2010) which cause their size and abundance to be reduced compared to the ref- erence pCO2 areas (Connell et al., 2018, White Island, New Zealand; Harvey et al., unpublished, this site and Vulcano, Italy). Similarly, the observed number of sea urchin feeding halos has also previously been found to be reduced in the Ischia CO2 seeps (Kroeker, Gambi, et al., 2013). Fish communities also play a key role in top-down con- trol and, in our site, communities included more herbivorous fishes than the surrounding non-acidified areas (Cattano et al., 2020). Clearly, the increased turf algae supported a greater herbivorous fish population than in the reference conditions, but those same herbivores alone are not able to control the increased growth of the boosted turf algae (also see Baggini et al., 2015; Connell et al., 2018). Although outside the scope of this study, this may support the no- tion that the fish are preferentially consuming different algae other than the turfs and further reinforcing the ecological shift.

After 12 months, assemblages in reference pCO2 conditions continued to gain species through time and had developed more structurally complex communities with clearly defined understorey and canopy species. The assemblages in the elevated pCO2 became arrested in terms of their successional development due to competi- tion for space by the turf algae. A similar overgrowth and dominance by turf algae was observed on recruitment tiles in the Ischia CO2 seep (Kroeker et al., 2012; Porzio et al., 2013). This community de- velopment in our study resulted in a similar community composition at 6 and 12 months with only the abundance of the turf algae being increased. At 12 months, communities on the tiles were visually in- distinguishable from the surrounding rocky substrata (Figure S4). Similar declines in macroalgal diversity with increasing pCO2 have

also been demonstrated in Methana, Greece (Baggini et al., 2014). The simplification of marine ecosystems has been observed across CO2 seeps (Agostini et al., 2018; Brown et al., 2018; Cigliano et al., 2010; Fabricius et al., 2011; Kroeker, Gambi, et al., 2013; Vizzini et al., 2017), with such changes leading to a functional biodiversity loss in the system (Teixidó et al., 2018). It is likely that such sim- ple systems are maintained by reinforcing feedback loops (sediment trapping, changes in physicochemical environment, and recruit- ment inhibition) that facilitate turf algal dominance. Turf algae can inhibit successional development by reducing primary substratum availability (Airoldi, 1998; Connell & Russell, 2010) and by trapping sediment which alters settlement surface chemistry and reduces the survival of other recruits (Airoldi, 2003; Gorman & Connell, 2009). Such dominance by short-lived species, which then locks the system in place, can lead to decreased stability in the system (Stachowicz et al., 2007), with implications for the functioning of the system under future ocean acidification (Teixidó et al., 2018).

In terms of community dynamics, both the reference and ele- vated pCO2 conditions appeared to overwhelm any ecological re- sistance that would have otherwise resisted ecosystem change. This was demonstrated by the established algal communities that were transplanted from reference to elevated pCO2 conditions con- verging (in terms of community composition) to almost match the community formed under elevated pCO2 conditions (and vice versa). This suggests that acidification-driven ecological shifts to simplified turf algae communities will occur regardless of the state that the community is in, and means that the community successional trajec- tory is not fixed from the initial bare substratum during primary or secondary succession. The prevention of such shifts by ecosystem management will require a focused effort on resilience building in order to mitigate the future degradation of ecosystems (Billé et al., 2013; Falkenberg et al., 2013). In contrast, the convergence of the communities transplanted from elevated pCO2 conditions to the ref- erence pCO2 conditions could mean that recovery from a degraded state is possible. This would likely be due to sufficient compensatory processes at our reference pCO2 location, and/or the turf algae los- ing its competitive edge in the absence of elevated pCO2. Therefore, a combination of conservation strategy and meaningful reductions in atmospheric CO2 emissions could achieve substantial recovery of the abundance, structure and function of shallow coastal marine ecosystems (Duarte et al., 2020).

Despite possessing highly divergent communities, gross oxygen production was similar between all of the transplanted tiles. Net ox- ygen production, however, was positive in the acidified conditions, but balanced between productivity and respiration for the reference pCO2 communities due to elevated respiration. Ecosystems that are more developed and stable will tend towards rates of oxygen production and respiration being equal, tending to not accumulate further biomass. Early-stage ecosystems will tend to have a higher productivity per biomass, but will be lacking in terms of biomass and species diversity (Cooke, 1967). This further supports the concept that the algal community developing under elevated pCO2 is ar- rested into a typical early-stage community dominated by r-selected

10  |    HARVEY Et Al.

species. Previous studies in CO2 seeps have generally focussed on the primary production or photophysiology of individual species of algae (e.g. Celis-Plá et al., 2015; Porzio et al., 2018, 2020) with the aim of assessing their physiological response to ocean acidifica- tion, rather than the effects on overall community net production (making comparisons difficult). The sessile invertebrate communi- ties differed in community composition between the reference and elevated pCO2 sites, but not in percentage cover, suggesting that they were not a sizeable contributor toward such large changes in net oxygen production. Instead, the decreased net oxygen was likely driven by the greater algal biomass (as well as low surface to vol- ume ratio) of the more highly structurally complex reference com- munity compared to the high surface to biomass ratio of the turf algae. Taken together, this suggests that the greater net production stimulated by ocean acidification does not translate into enhanced ecosystem benefits, such as increased community cover, biomass, biodiversity or structural complexity, as well as an altered sessile in- vertebrate community.

Natural analogues provide a number of benefits for advancing our understanding about the responses of shallow-water marine communities to ocean acidification conditions, but they are not perfect analogues. Carbonate chemistry at some CO2 seeps can be highly variable (Rastrick et al., 2018), and areas in close proximity to CO2 vents can be enriched in some metals and toxins (Vizzini et al., 2013; Zitoun et al., 2020). It is possible to reduce such con- founding factors by avoiding toxic areas and only selecting sites a suitable distance away, since contamination from hydrothermal flu- ids can be quickly diluted by mixing with seawater (Agostini et al., 2015; Pichler et al., 2019). The gas being released at our study site is 98 ± 3 (SD) % CO2, and although concentrations of hydrogen sulfide are detected at the main vent, they are below detection limits ~50 m away from the main vents (Agostini et al., 2015) and the study site used in this study is more than 300 m away from the main vent. The possibility remains that other trace elements or heavy metals (as yet unmeasured) may be present either in the water or bioaccumulated in biota, as has been shown in other CO2 seeps (Mirasole et al., 2020; Mishra et al., 2020; Vizzini et al., 2013; Zitoun et al., 2020), which could influence the response of marine organisms to ocean acidification. An additional consideration for CO2 seeps is that they demonstrate the consequences of future ocean acidification but in the absence of concurrent ocean warming (Hughes et al., 2017), and temperatures will mediate the response of organisms and communities to future ocean acidification. Such an issue can be addressed by comparing CO2 seep systems under different thermal regimes and assess the consistency of responses (Johnson et al., 2012), or by manipulating temperature along CO2 gradients (Alessi et al., 2019). Despite these caveats, the use of CO2 seeps is still invaluable for providing a window into the future state of organisms, communities and ecosystems to future ocean acidifi- cation (Rastrick et al., 2018).

In conclusion, ocean acidification can set the course of suc- cessional development in algal communities, benefitting turf algae whilst causing reduced algal biomass, diversity and complexity.

Altered carbonate chemistry can enable opportunistic r-selected species to competitively exclude other species and lock the com- munity in a species-poor early successional stage. The ecological process responsible for this shift in community composition was not simply altering community trajectory during primary succes- sion, as the same shift occurred in preestablished communities. This highlights that without reducing atmospheric CO2 emissions we may increasingly observe the loss of large algal habitats and the spread of fast-growing, small opportunistic species that can utilize additional inorganic carbon. By understanding the ecological pro- cesses responsible for driving shifts in community composition, we can begin to better assess how communities are likely to be altered by ocean acidification. Finally, our results show that the recovery of shallow-water marine communities is possible if meaningful reduc- tions in CO2 emissions are implemented, as encouraged by the Paris Agreement.

A C K N O W L E D G E M E N T S We thank the technical staff at ‘Shimoda Marine Research Center, University of Tsukuba’ for their assistance aboard RV Tsukuba II and at the study site, and the Japan Fisheries agencies of Nijima/Shikine Island (Tokyo prefecture) for their support. This project was heavily supported by the ‘International Education and Research Laboratory Program’, University of Tsukuba. This work was also supported by JSPS KAKENHI Grant Number 17K17622, and we acknowledge funding support from the Ministry of Environment, Government of Japan (Suishinhi: 4RF-1701).

Some of the images used within the graphical abstract are courtesy of the Integration and Application Network, University of Maryland Center for Environmental Science (ian.umces.edu/symbo ls/).

C O N F L I C T O F I N T E R E S T The authors declare no conflicts of interest.

A U T H O R C O N T R I B U T I O N S B.P.H. conceived the idea, designed the methodology, analysed the data and led the writing of the manuscript. B.P.H. and K.K. carried out the image analysis. B.P.H. and S.A. performed the oxygen production measurements. All authors assisted with field work, contributed critically to the drafts and gave final approval for publication.

D ATA AVA I L A B I L I T Y S TAT E M E N T Biological data (Figures 2–5 and Tables 2 and 3) and associated car- bonate chemistry data (Figure 1; Figure S1; Table 1) are stored on Zenodo (https://doi.org/10.5281/zenodo.4280018).

O R C I D Ben P. Harvey https://orcid.org/0000-0002-4971-1634 Koetsu Kon https://orcid.org/0000-0003-1379-0702 Sylvain Agostini https://orcid.org/0000-0001-9040-9296 Shigeki Wada https://orcid.org/0000-0001-6893-7498 Jason M. Hall-Spencer https://orcid.org/0000-0002-6915-2518

    |  11HARVEY Et Al.

R E F E R E N C E S Abràmoff, M. D., Magalhães, P. J., & Ram, S. J. (2004). Image processing

with ImageJ. Biophotonics International, 11(7), 36–42. Agostini, S., Harvey, B. P., Wada, S., Kon, K., Milazzo, M., Inaba, K., &

Hall-Spencer, J. M. (2018). Ocean acidification drives community shifts towards simplified non-calcified habitats in a subtropi- cal-temperate transition zone. Scientific Reports, 8, 11354. https:// doi.org/10.1038/s4159 8-018-29251 -7

Agostini, S., Wada, S., Kon, K., Omori, A., Kohtsuka, H., Fujimura, H., Tsuchiya, Y., Sato, T., Shinagawa, H., Yamada, Y., & Inaba, K. (2015). Geochemistry of two shallow CO2 seeps in Shikine Island (Japan) and their potential for ocean acidification research. Regional Studies in Marine Science, 2(Suppl.), 45–53.

Airoldi, L. (1998). Roles of disturbance, sediment stress, and substratum re- tention on spatial dominance in algal turf. Ecology, 79(8), 2759–2770.

Airoldi, L. (2003). The effects of sedimentation on rocky coast assem- blages. Oceanography and Marine Biology: An Annual Review, 41, 169–171.

Albright, R., Caldeira, L., Hosfelt, J., Kwiatkowski, L., Maclaren, J. K., Mason, B. M., Nebuchina, Y., Ninokawa, A., Pongratz, J., Ricke, K. L., Rivlin, T., Schneider, K., Sesboüé, M., Shamberger, K., Silverman, J., Wolfe, K., Zhu, K., & Caldeira, K. (2016). Reversal of ocean acid- ification enhances net coral reef calcification. Nature, 531(7594), 362–365.

Albright, R., Takeshita, Y., Koweek, D. A., Ninokawa, A., Wolfe, K., Rivlin, T., Nebuchina, Y., Young, J., & Caldeira, K. (2018). Carbon dioxide addition to coral reef waters suppresses net community calcification. Nature, 555, 516–519. https://doi.org/10.1038/ natur e25968

Alessi, C., Giomi, F., Furnari, F., Sarà, G., Chemello, R., & Milazzo, M. (2019). Ocean acidification and elevated temperature negatively affect re- cruitment, oxygen consumption and calcification of the reef-building Dendropoma cristatum early life stages: Evidence from a manipula- tive field study. Science of The Total Environment, 693, 133476.

Algueró-Muñiz, M., Alvarez-Fernandez, S., Thor, P., Bach, L. T., Esposito, M., Horn, H. G., Ecker, U., Langer, J. A. F., Taucher, J., Malzahn, A. M., Riebesell, U., & Boersma, M. (2017). Ocean acidification ef- fects on mesozooplankton community development: Results from a long-term mesocosm experiment. PLoS One, 12(4), e0175851. https://doi.org/10.1371/journ al.pone.0175851

Baggini, C., Issaris, Y., Salomidi, M., & Hall-Spencer, J. (2015). Herbivore diversity improves benthic community resilience to ocean acidi- fication. Journal of Experimental Marine Biology and Ecology, 469, 98–104. https://doi.org/10.1016/j.jembe.2015.04.019

Baggini, C., Salomidi, M., Voutsinas, E., Bray, L., Krasakopoulou, E., & Hall-Spencer, J. M. (2014). Seasonality affects macroalgal commu- nity response to increases in pCO2. PLoS One, 9(9), e106520.

Billé, R., Kelly, R., Biastoch, A., Harrould-Kolieb, E., Herr, D., Joos, F., Kroeker, K., Laffoley, D., Oschlies, A., & Gattuso, J.-P. (2013). Taking action against ocean acidification: A review of management and policy options. Environmental Management, 52(4), 761–779. https:// doi.org/10.1007/s0026 7-013-0132-7

Brinkman, T. J., & Smith, A. M. (2015). Effect of climate change on crus- tose coralline algae at a temperate vent site, White Island, New Zealand. Marine and Freshwater Research, 66(4), 360–370.

Brown, N. E. M., Milazzo, M., Rastrick, S. P. S., Hall-Spencer, J. M., Therriault, T. W., & Harley, C. D. G. (2018). Natural acidification changes the timing and rate of succession, alters community struc- ture, and increases homogeneity in marine biofouling communities. Global Change Biology, 24(1), e112–e127. https://doi.org/10.1111/ gcb.13856

Brown, N. E. M., Therriault, T. W., & Harley, C. D. G. (2016). Field-based experimental acidification alters fouling community structure and reduces diversity. Journal of Animal Ecology, 85(5), 1328–1339. https://doi.org/10.1111/1365-2656.12557

Calosi, P., Rastrick, S. P. S., Graziano, M., Thomas, S. C., Baggini, C., Carter, H. A., Hall-Spencer, J. M., Milazzo, M., & Spicer, J. I. (2013). Distribution of sea urchins living near shallow water CO2 vents is dependent upon species acid-base and ion-regulatory abilities. Marine Pollution Bulletin, 73(2), 470–484.

Cattano, C., Agostini, S., Harvey, B. P., Wada, S., Quattrocchi, F., Turco, G., Inaba, K., Hall-Spencer, J. M., & Milazzo, M. (2020). Changes in fish communities due to benthic habitat shifts under ocean acidi- fication conditions. Science of The Total Environment, 725, 138501. https://doi.org/10.1016/j.scito tenv.2020.138501

Celis-Plá, P. S. M., Hall-Spencer, J. M., Horta, P. A., Milazzo, M., Korbee, N., Cornwall, C. E., & Figueroa, F. L. (2015). Macroalgal responses to ocean acidification depend on nutrient and light levels. Frontiers in Marine Science, 2, 26. https://doi.org/10.3389/ fmars.2015.00026

Cigliano, M., Gambi, M. C., Rodolfo-Metalpa, R., Patti, F. P., & Hall- Spencer, J. M. (2010). Effects of ocean acidification on invertebrate settlement at volcanic CO2 vents. Marine Biology, 157(11), 2489– 2502. https://doi.org/10.1007/s0022 7-010-1513-6

Connell, J. H., & Slatyer, R. O. (1977). Mechanisms of succession in nat- ural communities and their role in community stability and organi- zation. The American Naturalist, 111(982), 1119–1144. https://doi. org/10.1086/283241

Connell, S. D., Doubleday, Z. A., Foster, N. R., Hamlyn, S. B., Harley, C. D., Helmuth, B., Kelaher, B. P., Nagelkerken, I., Rodgers, K. L., Sarà, G., & Russell, B. D. (2018). The duality of ocean acidification as a resource and a stressor. Ecology, 99(5), 1005–1010. https://doi. org/10.1002/ecy.2209

Connell, S. D., Kroeker, K. J., Fabricius, K. E., Kline, D. I., & Russell, B. D. (2013). The other ocean acidification problem: CO2 as a re- source among competitors for ecosystem dominance. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1627), 20120442. https://doi.org/10.1098/rstb.2012.0442

Connell, S. D., & Russell, B. D. (2010). The direct effects of increasing CO2 and temperature on non-calcifying organisms: Increasing the potential for phase shifts in kelp forests. Proceedings of the Royal Society B: Biological Sciences, 277(1686), 1409–1415.

Cooke, G. D. (1967). The pattern of autotrophic succession in lab- oratory microcosms. BioScience, 17(10), 717–721. https://doi. org/10.2307/1294089

Cornwall, C. E., Revill, A. T., Hall-Spencer, J. M., Milazzo, M., Raven, J. A., & Hurd, C. L. (2017). Inorganic carbon physiology underpins macroalgal responses to elevated CO2. Scientific Reports, 7, 46297. https://doi.org/10.1038/srep4 6297

Crook, E. D., Kroeker, K. J., Potts, D. C., Rebolledo-Vieyra, M., Hernandez- Terrones, L. M., & Paytan, A. (2016). Recruitment and succession in a tropical benthic community in response to in-situ ocean acidifi- cation. PLoS One, 11(1), e0146707. https://doi.org/10.1371/journ al.pone.0146707

Done, T. (1992). Effects of tropical cyclone waves on ecological and geo- morphological structures on the Great Barrier Reef. Continental Shelf Research, 12(7–8), 859–872.

Duarte, C. M., Agusti, S., Barbier, E., Britten, G. L., Castilla, J. C., Gattuso, J.-P., Fulweiler, R. W., Hughes, T. P., Knowlton, N., Lovelock, C. E., Lotze, H. K., Predragovic, M., Poloczanska, E., Roberts, C., & Worm, B. (2020). Rebuilding marine life. Nature, 580(7801), 39–51. https:// doi.org/10.1038/s4158 6-020-2146-7

Enochs, I. C., Manzello, D. P., Donham, E. M., Kolodziej, G., Okano, R., Johnston, L., Young, C., Iguel, J., Edwards, C. B., Fox, M. D., Valentino, L., Johnson, S., Benavente, D., Clark, S. J., Carlton, R., Burton, T., Eynaud, Y., & Price, N. N. (2015). Shift from coral to macroalgae dominance on a volcanically acidified reef. Nature Climate Change, 5(12), 1083–1088. https://doi.org/10.1038/nclim ate2758

Fabricius, K. E., Langdon, C., Uthicke, S., Humphrey, C., Noonan, S., De’ath, G., Okazaki, R., Muehllehner, N., Glas, M. S., & Lough, J.

12  |    HARVEY Et Al.

M. (2011). Losers and winners in coral reefs acclimatized to ele- vated carbon dioxide concentrations. Nature Climate Change, 1(3), 165–169. https://doi.org/10.1038/nclim ate1122

Falkenberg, L. J., Connell, S. D., & Russell, B. D. (2013). Disrupting the effects of synergies between stressors: Improved water quality dampens the effects of future CO2 on a marine habitat. Journal of Applied Ecology, 50(1), 51–58. https://doi.org/10.1111/1365-2664. 12019

Foo, S. A., Byrne, M., Ricevuto, E., & Gambi, M. C. (2018). The carbon dioxide vents of Ischia, Italy, a natural system to assess impacts of ocean acidification on marine ecosystems: An overview of re- search and comparisons with other vent systems. Oceanography and Marine Biology: An Annual Review, 56, 237–310. https://doi. org/10.1201/97804 29454 455-9

Garilli, V., Rodolfo-Metalpa, R., Scuderi, D., Brusca, L., Parrinello, D., Rastrick, S. P. S., Foggo, A., Twitchett, R. J., Hall-Spencer, J. M., & Milazzo, M. (2015). Physiological advantages of dwarfing in surviv- ing extinctions in high-CO2 oceans. Nature Climate Change, 5(7), 678–682. https://doi.org/10.1038/nclim ate2616

Gattuso, J.-P., Magnan, A., Billé, R., Cheung, W. W. L., Howes, E. L., Joos, F., Allemand, D., Bopp, L., Cooley, S. R., Eakin, C. M., Hoegh- Guldberg, O., Kelly, R. P., Pörtner, H.-O., Rogers, A. D., Baxter, J. M., Laffoley, D., Osborn, D., Rankovic, A., Rochette, J., … Turley, C. (2015). Contrasting futures for ocean and society from differ- ent anthropogenic CO2 emissions scenarios. Science, 349(6243), aac4722. https://doi.org/10.1126/scien ce.aac4722

Gaylord, B., Kroeker, K. J., Sunday, J. M., Anderson, K. M., Barry, J. P., Brown, N. E., Connell, S. D., Dupont, S., Fabricius, K. E., Hall- Spencer, J. M., Klinger, T., Milazzo, M., Munday, P. L., Russell, B. D., Sanford, E., Schreiber, S. J., Thiyagarajan, V., Vaughan, M. L. H., Widdicombe, S., & Harley, C. D. G. (2015). Ocean acidification through the lens of ecological theory. Ecology, 96(1), 3–15.

Ghedini, G., & Connell, S. D. (2016). Organismal homeostasis buffers the effects of abiotic change on community dynamics. Ecology, 97(10), 2671–2679. https://doi.org/10.1002/ecy.1488

Ghedini, G., Russell, B. D., & Connell, S. D. (2015). Trophic compensation reinforces resistance: Herbivory absorbs the increasing effects of multiple disturbances. Ecology Letters, 18(2), 182–187. https://doi. org/10.1111/ele.12405

Godbold, J. A., & Solan, M. (2013). Long-term effects of warming and ocean acidification are modified by seasonal variation in species re- sponses and environmental conditions. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1627), 20130186.

González-Delgado, S., & Hernández, J. C. (2018). The importance of natural acidified systems in the study of ocean acidification: What have we learned? Advances in Marine Biology, 80, 57–99. https://doi. org/10.1016/bs.amb.2018.08.001

Gordillo, F. J., Figueroa, F. L., & Niell, F. X. (2003). Photon-and carbon-use efficiency in Ulva rigida at different CO2 and N levels. Planta, 218(2), 315–322.

Gorman, D., & Connell, S. D. (2009). Recovering subtidal forests in hu- man-dominated landscapes. Journal of Applied Ecology, 46(6), 1258– 1265. https://doi.org/10.1111/j.1365-2664.2009.01711.x

Gruner, D. S., Smith, J. E., Seabloom, E. W., Sandin, S. A., Ngai, J. T., Hillebrand, H., Harpole, W. S., Elser, J. J., Cleland, E. E., Bracken, M. E. S., Borer, E. T., & Bolker, B. M. (2008). A cross-system synthesis of consumer and nutrient resource control on producer biomass. Ecology Letters, 11(7), 740–755.

Hall-Spencer, J. M., & Harvey, B. P. (2019). Ocean acidification impacts on coastal ecosystem services due to habitat degradation. Emerging Topics in Life Sciences, 3(2), 197–206. https://doi.org/10.1042/ ETLS2 0180117

Hall-Spencer, J. M., Rodolfo-Metalpa, R., Martin, S., Ransome, E., Fine, M., Turner, S. M., Rowley, S. J., Tedesco, D., & Buia, M.-C. (2008). Volcanic carbon dioxide vents show ecosystem effects of ocean acidification. Nature, 454(7200), 96–99. https://doi.org/10.1038/natur e07051

Harvey, B. P., Agostini, S., Kon, K., Wada, S., & Hall-Spencer, J. M. (2019). Diatoms dominate and alter marine food-webs when CO2 rises. Diversity, 11(12). https://doi.org/10.3390/d1112 0242

Harvey, B. P., Agostini, S., Wada, S., Inaba, K., & Hall-Spencer, J. M. (2018). Dissolution: The Achilles’ heel of the triton shell in an acidifying ocean. Frontiers in Marine Science, 5, 371. https://doi.org/10.3389/ fmars.2018.00371

Harvey, B. P., Gwynn-Jones, D., & Moore, P. J. (2013). Meta-analysis reveals complex marine biological responses to the interactive effects of ocean acidification and warming. Ecology and Evolution, 3(4), 1016–1030. https://doi.org/10.1002/ece3.516

Harvey, B. P., McKeown, N. J., Rastrick, S. P. S., Bertolini, C., Foggo, A., Graham, H., Hall-Spencer, J. M., Milazzo, M., Shaw, P. W., Small, D. P., & Moore, P. J. (2016). Individual and population-level responses to ocean acidification. Scientific Reports, 6, 20194. https://doi. org/10.1038/srep2 0194

Hillebrand, H., Gruner, D. S., Borer, E. T., Bracken, M. E. S., Cleland, E. E., Elser, J. J., Harpole, W. S., Ngai, J. T., Seabloom, E. W., Shurin, J. B., & Smith, J. E. (2007). Consumer versus resource control of producer diversity depends on ecosystem type and producer com- munity structure. Proceedings of the National Academy of Sciences of the United States of America, 104(26), 10904–10909. https://doi. org/10.1073/pnas.07019 18104

Hughes, T. P., Barnes, M. L., Bellwood, D. R., Cinner, J. E., Cumming, G. S., Jackson, J. B. C., Kleypas, J., van de Leemput, I. A., Lough, J. M., Morrison, T. H., Palumbi, S. R., van Nes, E. H., & Scheffer, M. (2017). Coral reefs in the Anthropocene. Nature, 546(7656), 82–90. https:// doi.org/10.1038/natur e22901

Hughes, T. P., Rodrigues, M. J., Bellwood, D. R., Ceccarelli, D., Hoegh- Guldberg, O., McCook, L., Moltschaniwskyj, N., Pratchett, M. S., Steneck, R. S., & Willis, B. (2007). Phase shifts, herbivory, and the resilience of coral reefs to climate change. Current Biology, 17(4), 360–365. https://doi.org/10.1016/j.cub.2006.12.049

IPCC. (2013). Climate change 2013 – The physical science basis: Working group I contribution to the fifth assessment report of the IPCC (No. 0521880092; p. 1535). Cambridge University Press.

Jenkins, S. R., Hawkins, S. J., & Norton, T. A. (1999). Interaction between a fucoid canopy and limpet grazing in structuring a low shore inter- tidal community. Journal of Experimental Marine Biology and Ecology, 233(1), 41–63.

Johnson, C. R., & Mann, K. H. (1988). Diversity, patterns of adaptation, and stability of Nova Scotian kelp beds. Ecological Monographs, 58(2), 129–154. https://doi.org/10.2307/1942464

Johnson, V. R., Russell, B. D., Fabricius, K. E., Brownlee, C., & Hall- Spencer, J. M. (2012). Temperate and tropical brown macroalgae thrive, despite decalcification, along natural CO2 gradients. Global Change Biology, 18(9), 2792–2803.

Kamenos, N. A., Perna, G., Gambi, M. C., Micheli, F., & Kroeker, K. J. (2016). Coralline algae in a naturally acidified ecosystem persist by maintaining control of skeletal mineralogy and size. Proceedings of the Royal Society B: Biological Sciences, 283(1840), 20161159. https://doi.org/10.1098/rspb.2016.1159

Kassambara, A. (2019). ggpubr: “ggplot2” based publication ready plots. R Package Version 0.2.4. https://CRAN.R-proje ct.org/packa ge=ggpubr

Kelly, E. L. A., Eynaud, Y., Clements, S. M., Gleason, M., Sparks, R. T., Williams, I. D., & Smith, J. E. (2016). Investigating functional redundancy versus complementarity in Hawaiian herbivorous coral reef fishes. Oecologia, 182(4), 1151–1163. https://doi.org/10.1007/s0044 2-016-3724-0

Kerfahi, D., Harvey, B. P., Agostini, S., Kon, K., Huang, R., Adams, J. M., & Hall-Spencer, J. M. (2020). Responses of intertidal bacterial biofilm communities to increasing pCO2. Marine Biotechnology. https://doi. org/10.1007/s1012 6-020-09958 -3

Koch, M., Bowes, G., Ross, C., & Zhang, X.-H. (2013). Climate change and ocean acidification effects on seagrasses and marine macroalgae. Global Change Biology, 19(1), 103–132. https://doi.org/10.1111/ j.1365-2486.2012.02791.x

    |  13HARVEY Et Al.

Kroeker, K. J., Gambi, M. C., & Micheli, F. (2013). Community dynam- ics and ecosystem simplification in a high-CO2 ocean. Proceedings of the National Academy of Sciences of the United States of America, 110(31), 12721–12726. https://doi.org/10.1073/pnas.12164 64110

Kroeker, K. J., Kordas, R. L., Crim, R., Hendriks, I. E., Ramajo, L., Singh, G. S., Duarte, C. M., & Gattuso, J.-P. (2013). Impacts of ocean acidifica- tion on marine organisms: Quantifying sensitivities and interaction with warming. Global Change Biology, 19(6), 1884–1896. https://doi. org/10.1111/gcb.12179

Kroeker, K. J., Micheli, F., & Gambi, M. C. (2012). Ocean acidification causes ecosystem shifts via altered competitive interactions. Nature Climate Change, 3(2), 156–159.

Kroeker, K. J., Micheli, F., Gambi, M. C., & Martz, T. R. (2011). Divergent ecosystem responses within a benthic marine community to ocean acidification. Proceedings of the National Academy of Sciences of the United States of America, 108(35), 14515–14520. https://doi. org/10.1073/pnas.11077 89108

Li, W., Gao, K., & Beardall, J. (2012). Interactive effects of ocean acid- ification and nitrogen-limitation on the diatom Phaeodactylum tri- cornutum. PLoS One, 7(12), e51590. https://doi.org/10.1371/journ al.pone.0051590

Ling, S. D., Scheibling, R. E., Rassweiler, A., Johnson, C. R., Shears, N., Connell, S. D., Salomon, A. K., Norderhaug, K. M., Pérez-Matus, A., Hernández, J. C., & Clemente, S. (2015). Global regime shift dynam- ics of catastrophic sea urchin overgrazing. Philosophical Transactions of the Royal Society B: Biological Sciences, 370(1659), 20130269.

Milazzo, M., Alessi, C., Quattrocchi, F., Chemello, R., D’Agostaro, R., Gil, J., Vaccaro, A. M., Mirto, S., Gristina, M., & Badalamenti, F. (2019). Biogenic habitat shifts under long-term ocean acidification show nonlinear community responses and unbalanced functions of asso- ciated invertebrates. Science of The Total Environment, 667, 41–48. https://doi.org/10.1016/j.scito tenv.2019.02.391

Milazzo, M., Rodolfo-Metalpa, R., Chan, V. B. S., Fine, M., Alessi, C., Thiyagarajan, V., Hall-Spencer, J. M., & Chemello, R. (2014). Ocean acidification impairs vermetid reef recruitment. Scientific Reports, 4, 4189. https://doi.org/10.1038/srep0 4189

Mirasole, A., Scopelliti, G., Tramati, C., Signa, G., Mazzola, A., & Vizzini, S. (2020). Evidences on alterations in skeleton composition and mineralization in a site-attached fish under naturally acidified conditions in a shallow CO2 vent. Science of The Total Environment, 143309. https://doi.org/10.1016/j.scito tenv.2020.143309

Mishra, A. K., Santos, R., & Hall -Spencer, J. M. (2020). Elevated trace elements in sediments and seagrasses at CO2 seeps. Marine Environmental Research, 153, 104810. https://doi.org/10.1016/j. maren vres.2019.104810

Moulin, L., Grosjean, P., Leblud, J., Batigny, A., Collard, M., & Dubois, P. (2015). Long-term mesocosms study of the effects of ocean acid- ification on growth and physiology of the sea urchin Echinometra mathaei. Marine Environmental Research, 103, 103–114. https://doi. org/10.1016/j.maren vres.2014.11.009

Mumby, P. J., Dahlgren, C. P., Harborne, A. R., Kappel, C. V., Micheli, F., Brumbaugh, D. R., Holmes, K. E., Mendes, J. M., Broad, K., & Sanchirico, J. N. (2006). Fishing, trophic cascades, and the process of grazing on coral reefs. Science, 311(5757), 98–101.

O’Brien, J., & Scheibling, R. (2018). Turf wars: Competition between foun- dation and turf-forming species on temperate and tropical reefs and its role in regime shifts. Marine Ecology Progress Series, 590, 1–17.

O’Donnell, M. J., Todgham, A. E., Sewell, M. A., Hammond, L. M., Ruggiero, K., Fangue, N. A., Zippay, M. L., & Hofmann, G. E. (2010). Ocean acidification alters skeletogenesis and gene expression in larval sea urchins. Marine Ecology Progress Series, 398, 157–171.

Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P. R., O’Hara, R. B., Simpson, G. L., Solymos, P., Stevens, M. H. H., Szoecs, E., & Wagner, H. (2019). The vegan package. Vegan: Community ecology package. R Package Version 2.5-5. https:// Cran.r-Proje ct.Org/Packa ge=vegan

Pichler, T., Biscéré, T., Kinch, J., Zampighi, M., Houlbrèque, F., & Rodolfo- Metalpa, R. (2019). Suitability of the shallow water hydrothermal system at Ambitle Island (Papua New Guinea) to study the effect of high pCO2 on coral reefs. Marine Pollution Bulletin, 138, 148–158. https://doi.org/10.1016/j.marpo lbul.2018.11.003

Porzio, L., Arena, C., Lorenti, M., De Maio, A., & Buia, M. C. (2020). Long-term response of Dictyota dichotoma var. intricata (C. Agardh) Greville (Phaeophyceae) to ocean acidification: Insights from high pCO2 vents. Science of The Total Environment, 731, 138896. https:// doi.org/10.1016/j.scito tenv.2020.138896

Porzio, L., Buia, M. C., Ferretti, V., Lorenti, M., Rossi, M., Trifuoggi, M., Vergara, A., & Arena, C. (2018). Photosynthesis and mineral- ogy of Jania rubens at low pH/high pCO2: A future perspective. Science of The Total Environment, 628–629, 375–383. https://doi. org/10.1016/j.scito tenv.2018.02.065

Porzio, L., Garrard, S. L., & Buia, M. C. (2013). The effect of ocean acid- ification on early algal colonization stages at natural CO2 vents. Marine Biology, 160(8), 2247–2259. https://doi.org/10.1007/s0022 7-013-2251-3

R Core Team. (2019). R: A language and environment for statistical comput- ing. R Foundation for Statistical Computing. Retrieved from https:// www.R-proje ct.org/

Rastrick, S. S. P., Graham, H., Azetsu-Scott, K., Calosi, P., Chierici, M., Fransson, A., Hop, H., Hall-Spencer, J., Milazzo, M., Thor, P., & Kutti, T. (2018). Using natural analogues to investigate the effects of climate change and ocean acidification on Northern ecosys- tems. ICES Journal of Marine Science, 75(7), 2299–2311. https://doi. org/10.1093/icesj ms/fsy128

Rogers, A., Blanchard, J. L., & Mumby, P. J. (2014). Vulnerability of coral reef fisheries to a loss of structural complexity. Current Biology, 24(9), 1000–1005. https://doi.org/10.1016/j.cub.2014.03.026

Stachowicz, J. J., Bruno, J. F., & Duffy, J. E. (2007). Understanding the effects of marine biodiversity on communities and ecosys- tems. Annual Review of Ecology Evolution and Systematics, 38, 739–766.

Steneck, R. S., & Dethier, M. N. (1994). A functional group approach to the structure of algal-dominated communities. Oikos, 69(3), 476–498.

Sunday, J. M., Fabricius, K. E., Kroeker, K. J., Anderson, K. M., Brown, N. E., Barry, J. P., Connell, S. D., Dupont, S., Gaylord, B., Hall-Spencer, J. M., Klinger, T., Milazzo, M., Munday, P. L., Russell, B. D., Sanford, E., Thiyagarajan, V., Vaughan, M. L. H., Widdicombe, S., & Harley, C. D. G. (2017). Ocean acidification can mediate biodiversity shifts by changing biogenic habitat. Nature Climate Change, 7, 81–85. https:// doi.org/10.1038/nclim ate3161

Teixidó, N., Gambi, M. C., Parravacini, V., Kroeker, K., Micheli, F., Villéger, S., & Ballesteros, E. (2018). Functional biodiversity loss along nat- ural CO2 gradients. Nature Communications, 9(1), 5149. https://doi. org/10.1038/s4146 7-018-07592 -1

Vizzini, S., Di Leonardo, R., Costa, V., Tramati, C. D., Luzzu, F., & Mazzola, A. (2013). Trace element bias in the use of CO2 vents as analogues for low pH environments: Implications for contamina- tion levels in acidified oceans. Estuarine, Coastal and Shelf Science, 134, 19–30.

Vizzini, S., Martínez-Crego, B., Andolina, C., Massa-Gallucci, A., Connell, S. D., & Gambi, M. C. (2017). Ocean acidification as a driver of com- munity simplification via the collapse of higher-order and rise of lower-order consumers. Scientific Reports, 7(1), 4018. https://doi. org/10.1038/s4159 8-017-03802 -w

Wang, Y., Naumann, U., Wright, S. T., & Warton, D. I. (2012). Mvabund– An R package for model-based analysis of multivariate abundance data. Methods in Ecology and Evolution, 3(3), 471–474.

Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer-Verlag.

Wilson, S. S., Furman, B. T., Hall, M. O., & Fourqurean, J. W. (2020). Assessment of hurricane Irma impacts on south Florida seagrass communities using

14  |    HARVEY Et Al.

long-term monitoring programs. Estuaries and Coasts, 43(5), 1119– 1132. https://doi.org/10.1007/s1223 7-019-00623 -0

Witkowski, C. R., Agostini, S., Harvey, B. P., van der Meer, M. T. J., Sinninghe Damsté, J. S., & Schouten, S. (2019). Validation of car- bon isotope fractionation in algal lipids as a $p$CO2 proxy using a natural CO2 seep (Shikine Island, Japan). Biogeosciences, 16(22), 4451–4461. https://doi.org/10.5194/bg-16-4451-2019

Zitoun, R., Connell, S. D., Cornwall, C. E., Currie, K. I., Fabricius, K., Hoffmann, L. J., Lamare, M. D., Murdoch, J., Noonan, S., Sander, S. G., Sewell, M. A., Shears, N. T., van den Berg, C. M. G., & Smith, A. M. (2020). A unique temperate rocky coastal hydrothermal vent system (Whakaari–White Island, Bay of Plenty, New Zealand): Constraints for ocean acidification studies. Marine and Freshwater Research, 71(3), 321–344.

S U P P O R T I N G I N F O R M AT I O N Additional supporting information may be found online in the Supporting Information section.

How to cite this article: Harvey BP, Kon K, Agostini S, Wada S, Hall-Spencer JM. Ocean acidification locks algal communities in a species-poor early successional stage. Glob Change Biol. 2020;00:1–14. https://doi.org/10.1111/ gcb.15455